/9’c'9~ we} 4-} I P0 "STAYING TOGETHER": KINSHIP AND COMMUNITY IN FIJI By James Nest Turner A DISSERTATION Submitted to Michigan State University in partiaI fulfilIment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of AnthropoIogy 1983 (@1983 JAMES NEST TURNER All Rights Reserved ABSTRACT “STAYING TOGETHER": KINSHIP AND COMMUNITY IN FIJI By James Nest Turner This study examines the principles of organization which regu-. late the life of a Fijian Village. The title is taken from the Fijian phrase tjkghggtg, translated here as "staying together.“ Dis- cussion focuses on how social units whose members are regarded as people of different "types,f who claim different origins, and who acknowledge different ancestral deities and totemic associations, nonetheless come to regard themselves as a community. The bonds which unite them are shown to include mutual dependence on village lands; propinquity and the resultant coactivity; common allegiance to a chief; ritual interdependence; and ties of uterine kinship and marriage. In developing this general theme, individual chapters explore a variety of topics. Among them are the use and distrfibution of land; the production, sharing, and exchange of food; rituals which renew the charter of the political order; the role of kava in ritual and recreational contexts; the web of kinship; and principles of hiararchy. All of these can be seen to be an integrating factor in community life, James Nest Turner but they also define differences between individuals as members of a variety of groups and categories.' Understanding the nature of these groups and categories is a principal concern of this study. This requires that we unravel the intertwined strands of propinquity and descent. Those who are united in one context may be differentiated in another, and it will be seen that the contexts in which boundaries are defined frequently involve exchange. The data on which this study is based were collected during seventeen months of fieldwork in a village in Naitasiri province in the mountainous interior of Viti Levu, the largest of the Fiji Islands. Much of the information was.acquired through participant observation of community life. Topics of interest first encountered in informal ’settings were then pursued through more structued interviews includ- ing household surveys and genealogical research. Oral accounts were combined with archival research to study such processes as chiefly succession. ACKNOWLEDGMENTS The field work on which this study is based was supported by a predoctoral research grant from the National Science Foundation, but it benefited, as well, from the assistance of many individuals. I wish to thank Luke Waqa of the Fijian Affairs Board for helping to arrange our stay in the village. I am also grateful to S. Tuinaceva, Archivist, Margaret Patel, Librarian, and the staff of the National Archives of Fiji for their assistance. I owe a debt of gratitude as well to Hon. Livai Nasilivata and Adi Kelera Royalotama for their hospitality and support. To the people of Nairukuruku many thanks. Without their tolerance and hospitality this study would not have been possible. I am especially endebted to Varinava Vatulawa, Taito Nakaloubula, Penisoni Haisiga, and Ratu Saimone Nawaqadau for their patient assistance. I also wish to thank my chairman, Charles Morrison, and the other members of my committee, Harry Raulet, John Hinnant, and Robert McKinley, for their comments and encouragement. Finally, special thanks to my wife, Diane, and daughter, Megan, who shared it all, good times and bad. ii TABLE OF CONTENTS Page LIST OF TABLES . . . . . . . . . . . . . . . . v LIST OF FIGURES . . . . . . . . . . . . . . . vi Chapter I. INTRODUCTION . . . . . . . . . . . . . . 1 II. LAND . . . . . . . . . . . . . . . . . 23 Footnotes--Chapter II . . . . . . . . . . 56 III. "EATING TOGETHER" . . . . . . . . . . . . 57 Footnotes--Chapter III . . . . . . . . . . 84 IV. “TRUE FOOD“ . . . . . . . . . . . . . . 86 Footnotes--Chapter IV . . . . . . . . . . 125 V. THE WATER OF LIFE . . . . . . . . . . . . 126 Footnotes--Chapter V . . . . . . . . . . . 169 VI. GROUPS AND CATEGORIES . . . . . . . . . . . 171 Footnotes--Chapter VI . . . . . . . . . . 225 VII. THE WEB OF KINSHIP . . . . . . . . . . . . 227 Footnotes--Chapter VII . . . . . . . . . . 283 VIII. PRINCIPLES OF HIERARCHY . . . . . . . . . . 284 Footnotes--Chapter VIII . . . . . . . . . 328 IX. NEIGHBORS . . .' . . . . . . . . . . . . 330 Footnotes--Chapter IV . . . . . . . . . . 369 Chapter Page x. CONCLUSION . . . '. . . . . . . . . . . . 37o APPENDIX--A BRIEF NOTE ON FIJIAN ORTHOGRAPHY . . . . . . 329 BIBLIOGRAPHY . . . . . . . . . . . . . . '. . 331 iv LIST OF TABLES Table Page 1. Kinship Terminology . . . . . . . . . . . . 249 Figure 1. ‘0 m \l 01 01 h 00 N C O O U C O O N N H H H H H H H H H H H o o oo \1 0‘ 01 b u N H O 0 0 O O O O O 0 O O 0 0 LIST OF FIGURES Map of Village Lands . Map of Village Lands by Soil Type Genealogical Genealogical Genealogical Genealogical Genealogical Genealogical Structure of Mataqali Nabubuco Structure of Mataqali Nakorowaqa Structure of Mataqali Siko (Lineage VA") . Structure of Mataqali Siko (Lineage VB") . Structure of Mataqali Siko (continued) Diagram Illustrating Matagali Endogamy Classification of the M855 . Classification of Kin in the Maternal Patriline . Direct Exchange Kin Terms Organized as Reciprocal Sets . . . Terminological Equivalence of Alternate Generations Generational Distance Equivalence of Cross Kin of Alternate Generations Classification of Kin by a Male Ego . Classification of Kin by a Female Ego A Relationship of Vasu Vata A Relationship of Veivakawai (Case 1) A Relationship of Veivakawai (Case 2) A Relationship of Veivakawai (Case 3) vi Page 35 40 179 207 213 214 215 219 221 222 237 254 261 262 263 266 269 278 279 280 281 Figure Page 22. The Principle of Seniority of Descent . . . . . . 288 23. The Division of Labor within the layusa . . . . . 293 24. Order of Succession to Chiefship . . . . . . . 315 25. Map of the Village . . . . . . . . . . . . 333 vii CHAPTER I INTRODUCTION This ethnographic account is based on information collected during seventeen months of fieldwork in a village in Naitasiri Province in the mountainous interior of Viti Levu, the largest of the Fiji Islands. The goals of the study are two-fold. One is quite general; the other more focused. 0n the one hand, it is an attempt to describe the life of the community as fully as possible. Like most ethnographies, it treats in detail some aspects of the way of life which it describes while others are barely mentioned. The topics selected for special treatment partly reflect my own interests and partly those of my informants. No doubt to some extent they also reflect the workings of chance, through which I was exposed to certain events and processes, rather than others. Underlying the discussion of these topics is an interest in those factors which promote cohesion in community life and thus produce a counterbalance to divisions based on descent, rank, and traditions of separate origin. This is a common goal of anthropological research, that of fleshing out the concept of social integration. The more focused issue deals specifically with social organiza- tion. It concerns the problematic nature of those units of local organization known as mataoali. Are they descent groups, groupings 1 of cognatic kin, or local groups in which membership is based on coresidence and cooperation, rather than principles of kinship? Or are they in some way all of these? Despite the existence of consider- able local variation in Fiji with respect to matters of kinship and descent, researchers working in different areas have encountered the same questions. On the island of Moala, Sahlins found the standard anthropological terminology for descent groups inapplicable to matagali and other units of organization, partly because they were not "precisely comparable? to units described elsewhere, and partly because they seemed to combine several principles of organization. Whether common descent, co-residence, aspects of rank, the existence of core lives, or other features of Moalan local groups [including matagali] should be used to label them is debatable; indeed, what is decisive at one level is not decisive at another level of local organization (Sahlins 1962:240). As Sahlins and others have noted, the analysis of Fijian social structure has been complicated by the coexistence of two types of models (Cf. Belshaw, 1965; France 1969; Walter 1978a). On the one hand, there are locally defined models of organization. These are built on people's own understandings of the composition, inter- relations, and purposes shared by the groups and categories to which they belong. The principles of recruitment to these units and the terms by which they are known are subject to some local variation. On the other hand, there is the standardized segmentary lineage model enshrined in the gghjurg_version of Fijian social structure by the Native Lands Commission. The official model imposes a single system of hierarchically arranged social units, a single set of terms for those units, and with the exceptions of Lau, Cakaudrove, Macuata Provinces, a single principle of recruitment, patrilineal descent. (In these three provinces affiliation is optative within a defined range of cognatic descent.) Males and females are registered in "The Book of Descendants," 511913.91 M, at birth or in early child- hood. Wherever else they may go in their lifetime, Fijians retain their membership in these jurally constituted units and retain the rights in land which go with membership. We will discuss this dual system more fully later. For now it is enough to note an important difference between the folk and jural models. To the extent that the activities of daily life are organized in terms of group membership, it is in terms of the local and not the gg jug; model that they will be structured. Because it does provide a structure for the organization of social action, the local model must reflect demographic and political shifts. It does so through the processes of segmentation, coalescence, and political realignment. The gg jugg model has less relevance for daily life since its principle object is the registration of legally defined rights in land. In fact, as we shall see, local tenurial relations reflect principles quite different from those encoded in the Native Lands regulations. Unlike local models of social structure which can be readjusted in the face of demographic change, the 9g jurg model is essentially frozen. It reflects the situation that the Native Lands Commission found in an area (or sometimes an order which it imposed) at a particular point in time. This dual system has complicated the ethnographer's task in two ways. In the past uncritical acceptance of the basic assumptions underlying the gg;jgrg_version of the Fijian social organization may have obscured understanding of the SL9. flagg principles obtaining in particular areas (e.g., Capell and Lester 1941:29). Moreover, upon learning of the ethnographer's interest in local group organization, informants may refer him or her to elders knowledgeable with respect to the largely irrelevant gg_jugg_model. Armed with the names of groups and membership lists obtained in the capital, the ethnographer may be dismayed to learn that many persons are unsure of the names of the groups in which they are registered. In time, however, the ethno- grapher will learn how people really are grouped, for what purposes, and what names they assign to these various units. Perhaps a more serious impediment to the understanding of Fijian social organization has arisen from a methodological short- coming on the part of anthropologists themselves. I refer to a failure to maintain the analytical distinction between local groupings based on coresidence and cooperation and those which are based on descent (Cf. Groves, 1963:287-288; Walter 1978a). The composition of the former type of grouping may reflect, to varying degrees, the outcome of individual choices, and those choices may be described in statistical terms. But the composition of the later type of group or category is a reflection of the ideology, rather than choice. Consider, for example, the distinction made by Geddes (1945: 36) between "true" and "composite? yavusa. A yavusa is a higher level descent category consisting of several mataqali. Geddes con- siders such a grouping to be a true ygyu§a_if its constituent mataggli claim a common founder, acknowledge a single ancestral deity, have the same totemic affiliations, and acknowledge the same chief. A composite yayu§a_is an association of several (or segments of several) true ygvusa. In the case of composite yavu§a_traditions of common origin, a common ancestral deity, and shared totems are absent. The grouping is unified only by allegiance to a single chief. Geddes' study was conducted in Serua province on the southern coast of Viti Levu. It is not possible to say whether the difference he is describing is an aspect of the local folk model of group organi— zation or whether his comments are based on a comparison of yavu§g_ as recorded in the official version of local social organization. What is clear is that he is describing two very different types of groupings, one of them based on descent principles, the other on territorial and political association. Fijians may refer to both as {yayu§a, but if the researcher is to understand the interplay of descent, propinquity, and chiefly ties each must first be understood separately. In considering the structure of Fijian descent groups, care must be taken to avoid confusing associations based on these other principles with those based on common descent. In the area in which I worked, the rule of descent is patri- lineal. While individuals, families, and even groups of related families can change their local group affiliation, membership in descent groups and categories is immutable, except through adoption. Adoption is rare, and the only examples which I encountered involved illegitimate children or the transfer of children between close agnates. Difficulties in sorting out the relationship between local groups and groupings based on descent arise because local groups typically consist of an agnatic core from which the name of the local group is derived and to which affinal and cognatic kin may be attached. Moreover, an agnatic descent group and the "composite" local group of the same name are referred to by the same terms. "Mataqali,” the term used to refer to middle-range groupings of either type, simply means "category," "type," or "species" and can be used for other types of classes, as well as these social groupings based on descent or coresi- dence. By most accounts the situation in Fiji's eastern islands is more fluid. Sahlins reports that on Moala, yaygsa (dispersed descent categories which Sahlins also refers to as Astocks") are "simply all the people who descend jg_the paternal line from a particular semi- devine ancestor" (1962:226; emphases mine). But he also claims that Adescent is bilaterally reckoned and honored and by this means cor- porate kin groups that are in the main patrilineally established are welded together" (1962:168). Similarly, Walter found that on the Exploring Islands of northern Lau, Athe ideology of recruitment is nonunilineal" (1978a:356). He refers to mataoali qua descent groups asramages, rather than clans since the latter term connotes unilineal recruitment. Hocart reports that on Lakeba, the chief island of Lau‘ Usually a man "follows“ his father's clan, but many men live with their mother's people, even though both clans may be in the same village, next to one another. If a man lives with his wife's people, the children follow the mother's clan. A man will also live with the clan into which he has been ado ted, but may leave it to go back to his father's (1929:17). There are ambiguities here to be sure, but the situation seems to be similar to that described by Sahlins on Moala. Of the major studies of the social organization of the eastern islands, only Thompson's report on southern Lau describes a system in which descent group membership is determined by an unambiguously patrilineal rule (1940: 35). The principle of bilateral descent and, assuming their exis- tance, ambilineal descent groups make it even more difficult to sort out the principles of descent and coresidence, but untangle them we must. In the Matailobau area it is possible to say with certainty that groupings which combine cognatically related patrilines are based on affinity, uterine kinship, coresidence, or chiefly ties but not common descent. It is apparently not as easy to draw the same con- clusion on Lakeba, the Exploring Islands, or Moala. In his analysis of Moalan social organization Sahlins clearly distinguishes between dispersed, patrilineally defined descent gate: .ggrie§_and local kin ggggps. He attributes the frequent inclusion of cognatic kin in Moalan local groups to the fact that 'descent is bilaterially reckoned and honored? (1962:168) and states that ”Common cognatic descent thus justifies the adhesion of lines" (1962:244). He does not refer to such groupings as cognatic descent groups, though given such statements one might expect him to classify, them as such. Nor is it clear what relationship exists between this recognition of cognatic descent as a criterion for local group membership and the agnatic principle which is said to be operative at the level of descent categories. Perhaps the meaning of the statements quoted above would be better served if the word "kinship" were substituted for "descent" (which seems to be used as a gloss for the Fijian word "kgwaf). Sahlins distinguishes between two types of matagali on Moala. When the core lines of these middle-range groupings consist of members of the same ”stock" or ygyusg, Sahlins refers to them as "agnatic matagali.“ When the core lines consist of members of more than one 133353, he refers to them as "cognatic matagali" (1962:244). Con- sider these cognatic matagal'. Their core lines are derived from different descent categories with different founding ancestors and different totems. In a word, their identities are different, and this difference of identity is a matter of agnation, for we are told that Moalan stocks are patrilineal descent categories. Yet while these core lines are different, they are also "the same? (tautauvata). But their sameness appears to be more a matter of association than of bilateral descent. They stay together, work together, and eat together. How do we describe this situation? It would seem useful, indeed crucial, to maintain a clear distinction between groupings defined in terms of descent and those defined on the basis of coresi- dence and coactivity. The latter type are usually also based on affinal and uterine ties among their members, but this need not entail the recognition of cognatic descent. Indeed the existence of ideology of bilaterally transmitted common substance (gig, "blood"; Sahlins, 1962:168) is not sufficient evidence that the concept of cognatic descent (as opposed to the recognition of bilateral kinship) even exists on Moala. What of the differences between the cognatically related patrilines which make up these composite matagali? What importance attaches to agnation? In cases where one line is much smaller than the other(s) and its inclusion a fairly recent event, what ambiguities of status attach to its members? Do the sister's son and his descendants play an equal role in the ceremonial duties of the matagali? Do they speak as equals in matters under discussion? These are some of the questions relevant to an analysis of the relationships between descent categories and local groups and between agnation and uterine kinship. At the heart of this study is the need to understand the bases for, and the consequences of, coresidence or, to use the Fijian idiom, "staying together" (tiko vata). Among other matters this involves a consideration of the relationship between people and the land from which they derive their livelihood. People who share a place come to be bound to one another through their mutual dependence on the land. There is an identification of "breed" with fborder' to borrow Burrows' (1939) terms or, in Silverman's (1971) more graphic, if. less elegant, vocabulary, "mud" with "blood." The study is organized around a number of tOpics, each of them the subject of a chapter: the distribution of land; the production, sharing, and exchange of 10 food; a first fruits ceremony which renews the political charter, while at the same time it promotes the fertility of the land and the people who depend on it; the ritual significance of kava; and the principles of agnation, cognation, rank, and coresidence. Each is vital to an understanding of the texture of village life, and each is relevant, as well, to our more particular focus. Ethnography is, of course, the product of human interaction, and it becomes more meaningful for us, or takes on a more human dimension, if we know something about the setting in which that inter- action took place. Having said a little about the foci of this mono- graph, it remains to provide the reader with some information about the conditions under which the study was conducted and some informa- tion, as well, about Fiji itself. The latter will provide a context for the historical discussion which follows. Having escaped a Midwestern winter, my wife and I and our nine-month old daughter had waited for several weeks in Fiji's capital, Suva, for the final preparations for our stay in the village to be completed. The time was spent making the inquiries which any new arrival in a country must make with a few added complications asso- ciated with the nature of our visit. And there were the purchases which had to be made, mostly household equipment, a task hampered somewhat by our unfamiliarity with the conditions of village life. When the time finally came for us to leave for the interior, the three of us and our baggage barely fit into the car which we had hired. 11 The road to the interior winds through the suburbs of the city, past the large homes of the wealthy with well-kept lawns, and a spectacular view of the sea. Further out, the pavement gives way to a narrow and winding gravel road, its surface pitted and rock-strewn in places by heavy tropical rains. The forest cover is lush, in places coming right down to the road. Occasionally there are breath- taking views: a thin, plume-like waterfall plummeting down a sheer rock face; a sudden view of a river winding through its flood plain. The road twists and turns, climbs and drops. On that first trip our driver nervously pressed the windshield with his hand. A month earlier its predecessor had been shattered by a flying rock. The road passes by several villages, and here and there we pass school children smiling and waving, or men and women, alone or in groups, traveling to and from their gardens. Distant mountains come into sight. The road becomes rougher as the view becomes more open and the countryside more beautiful. Finally, the road begins to descend in a series of sharp turns to the Wainimala valley, the river a glimmering ribbon in the distance below. From the point where the road reaches the valley floor, the journey to Nairukuruku village must continue by boat. During our stay in the village, my family and I traveled to and from Suva numerous times by boat and bus, but the three- to four- hour trip never became a routine journey for us, not because it was especially arduous, but because the two worlds we were traveling between were so different. At each end of the trip there would be a little hesitation, a mental changing of gears, before we adjusted to 12 the flow of life around us. But we were not yet fully aware of those differences on the afternoon of our arrival. Nairukuruku stands on a low bluff directly above a gravel bar which separates the course of the river. Sand quarried from that bar by the men of the village was used to mix the cement for the steps that lead up from the river at either end of the village. When we arrived on that first day, a crowd of smiling children rushed up and helped carry our things to the house that had been assigned to us. The village is organized along three parallel paths, the middle one of which is paved with cement. The grass within the village is trimmed regularly, either laboriously by machete or by one of the several power mowers owned by villagers. Ornamental plants and flowering trees and bushes abound, and on all sides of the village, there is a sprinkling of banana trees and other food plants. The general lushness of the vegetation and the visually pleasing houses give the village a picturesque quality. (For a person reared in a temperate climate, however, the seasonably heavy rains sometimes seemed a heavy price to pay for that luxuriant flora.) The surround- ing countryside is equally beautiful. In early morning, mist rolls down the hillside (in the local dialect the word for mist, "kabg,f also means morning), and lowing cattle come to the river to drink. Across the river and several miles away there is a single spectacular peak. After rainstorms it seemd to me that waterfalls could be seen coursing down its sheer rock face. Massive and pyramidal, its aspect 13 is continually changing as clouds give way to sun and sun to shad- ows. The fifty or so houses in the village are made from a variety of building materials. The most common house type has bamboo walls. Lengths of bamboo are stacked one upon the other between hardwood foundation posts and inner posts made from the very dense and beauti— fully grained wood of the tree fern. Once the bamboo is in place, it is lashed to the inner posts with vines. The walls are then covered with reeds and, finally, with the leaves of the makjtg_tree (Parinari laurina). The floors are earthern, but over them lies a layer Of dried grass and, finally, a covering of padamus mats so that the surface is soft and comfortable to sit on. When we first arrived in the village, only two.houses still had the traditional steeply pitched thatch roof. Before we left, both houses had been replaced so that all the homes in the village had iron roofs. Corrugated iron is a poorer insulation, but easier to maintain. Several houses in the village have walls of plaited bamboo attached to frames of sawn timber. Raised above the ground on founda- tions of cement blocks or hardwood, those houses have wooden floors and numerous windows. They are probably the most comfortable type of house in hot weather, but when the nights are cool and the iron roofs are damp with condensation, families sometimes take refuge in their cook houses. Two houses in the village are a variation on this type. Their walls are covered with corrugated roofing irons, perhaps less comfortable, but very durable. Finally, there are several 14 houses made completely from sawn timber. Because they have several windows in addition to the two or three doors that all houses have, they are cooler than some other house types, but with their floors raised above the ground and their iron roofs, they are also less comfortable in cool, damp weather, a fact we were soon to become aware of, for the night of our arrival turned out to be the coolest night of our stay in the village. . The house assigned to us had been built by a woman employed outside of the village. Its inhabitants included the woman's aged mother, her unmarried middle-aged brother, and numerous young peOple from all over the village, for the house functioned as a REE£.(3 sleeping house for unmarried boys and youths). The principal reason for its selection as our home during our stay was that it did not house a family, and those who normally stayed there could find accommo- dations elsewhere. Like all other village residents, the owner of the house (i.e., the male head of the household) was a member of a kin group or matagali, and because we were his guests, our household was attached to that matagali. It is common for husbands and wives to belong to different matagali before marriage, and so my wife was given the status of a natal member of another group, though she was counted as a member of our host's group through marriage. I stood in the relationship of younger brother (352133) to our host and his siblings, and he and his male siblings referred to my wife as "brother's spouse" (9152). Through these initial links, relation- ships of fictive kinship could be traced to all other village resi- dents. 15 Our position as titular members of this kin grouping was not unique. All other outsiders (e.g., ministers, school teachers) who reside in the village for any length of time are also attached to one of the five groupings which are considered to be the "owners" or "natives" (j £22521) of the village. Our inclusion was primarily a matter of ritual responsibility. We were expected to allign ourselves with our host's matagali on all occasions of ritual significance and to contribute as members toward the exchange obligations of the group. It is true that much of our socializing was with other households of the matagali and their kinsmen, but since the group was linked by affinity and uterine kinship to all other groups, we were expected, as they were, to mix with everyone. Our affiliation in no way limited interaction with informants. During the first weeks of our stay, those who could spoke to us in English. But as the weeks passed, and as our skills in the language increased, peOple began to speak to us only in Fijian, usually in the local dialect. Because our daughter was still an infant with consequent laundry problems, we employed a young woman during the first half of our stay. Prior to leaving for the field, I had shared child care responsibilities with my wife, but for what- ever reason (whether participant observation was a factor is a matter of debate) during our stay in the village most of the child care responsibilities fell to my wife. The impact our daughter had on our fieldwork is not an easy factor to assess. The presence of at least one child had the effect of "normalizing? us in our informants' eyes, 16 and for our daughter's part, she enjoyed her stay in the village. In fact, she became culturally Fijian; she spoke only Fijian, and all of her mannerisms were Fijian. To be truthful, there were times when my wife and I felt ambivalent toward” this. We began to understand how immigrants feel when their children adopt the host culture. To the extent that it was possible, we conformed in our domes- tic arrangements to the community around us. Unlike other households we did not cook on a wood fire in a cookhouse; procuring fuel would have been too time consuming. Instead, we cooked on a small kerosene stove. All of the women of the village had similar stoves, but because of the cost Of kerosene, they used them much less. To an extent our diet also differed from other peoples, partly from choice and partly from necessity. For breakfast we usually ate oatmeal because it was fast, as well as nutritious. We ate more rice and tinned fish than other families did, and until we planted edible hibiscus (yegygg_or .bele) in the yard, we ate fewer greens. But we ate the same root crops that other households did, tapioca (tavioka), taro (yoke), and yams. (ygg), both wild and domestic. These were given to us, gifts for which we reciprocated in kind with articles purchased in Suva. I assisted people in their gardens and even planted my own patch of tavioka. In time it supplied not only our needs, but I was also able to give some of it away. As our little garden was harvested, I replanted and expanded it on behalf of the family that had lent us the land. Minutes after our first arrival in the village we were taken by our host to the house of the chief. The elders of the community 17 were assembled to meet us, though in retrospect I am unable to say whether every mataqali in the village was represented. The purpose of this meeting was the acceptance of our presentation of some kava root (yggggg) and a whale's tooth (32222) by the chief. This formal presentation, or.i sevusevu, was made in order to secure the chief's permission for our stay in the village. This made our position as guests a “matter of the land" (k§_vakavanua). Such guests as we were placed a considerable burden, in a moral sense, on a Fijian village, and in turn, incur obligations toward the community at large. Given the fact that we were strangers with different customs who spoke their language imperfectly and who witnessed and sometimes probed into their personal affairs, our hosts treated us with tolerance and generosity. Whenever meat was distributed, and such distributions only occur on ritual occasions, a portion was set aside for us. In turn, whenever money was collected for some community project, we contributed, and we contributed as well at all village funerals. When the time finally came for us to leave the village, we again made a formal presentation of kava and whale's teeth and made a cash donation toward the financ- ing of the village electrical system. Whether the books were ulti- mately balanced is not for me to say, but certainly what we received in intangibles was beyond value. During our stay, I tried whenever possible to participate in the work of the village. My inexpert contribution to garden planting, housebUdeing,the quarrying of sand, or whatever the task at hand might be, was, I think, appreciated for what it was--an attempt to 18 participate in the life of the community. Much of what I learned was acquired through these informal sessions or from the conversa- tion at the kava sessions which inevitably follow any cooperative activity. In this way I became aware of topics which could then be investigated further through more formal interviews. The most demand- ing of these, both for me and my informants, were devoted to geneologi- cal research, but the genealogies obtained were also extremely reward- ing in terms of insights into the social life I was witnessing. Certain events, such as weddings and funerals, also provoked numerous questions and recording such events, and following up on them became endurance feats. For a community of its size, the village seemed an extremely busy place, and I often despaired of ever keeping up with it. Staying indoors to write was sometimes a necessity, but I always felt out of touch with events when I did so and a bit defensive as well. Staying up at night was out of the question, since we were always tired from the day's events. Each day began at around half past six. Our schedule had to conform to that of the community at large, and our daughter was always eager to begin exploring the day. As a result, we were often tired with the fatigue that accompanies a task which never ends. From time to time we sought refuge in the anonymity of Suva, but when we left the village for the last time, my wife and I both felt an emptiness and a kind of insecurity. After leaving the village, we devoted our remaining weeks in Fiji to library research in the National Archives. From time to time information thus acquired is incorporated into the following 19 account. Accordingly, I want to include in this section a very brief overview of the early period of European contact in Fiji. Hopefully, it will provide the necessary historical background. The first Europeans to sight the Fiji Islands were aboard the two ships commanded by Abel Janszoon Tasman. In 1643, in the face of a driving rainstrom, they sighted the eastern tip of Vanua Levu, Fiji's second largest island, and Cikobia, the northernmost island in the group. In 1774 Captain Cook recorded the position of Vatoa in southern Lau, and fifteen years later, on his epic boat Voyage from Tahiti to Timor, Captain Bligh of the Boga y recorded the positions of several more islands. Having learned of the Fijian's fierce reputation in Tonga while serving under Cook, Bligh did not dare land, and at one point his boat was pursued by canoes from the Yasawas. The first Europeans to live among Fijians were the survivors of Ships wrecked upon the islands' numerous reefs. They brought with them epidemic diseases and a knowledge of muskets. Because of the latter, they were sometimes the deciding factor in Fijian con- flicts and were consequently sought after by chiefs. The first deliberate contacts with Fijians came after Europeans discovered the presence of sandalwood on Vanua Levu. During the first decade of the nineteenth century, fortunes were made in the China trade, but by the end of that period, the large stands of sandalwood were gone. In order to secure cargoes of the aromatic wood traders began to assist helpful chiefs hiattacks upon their vivals. In the second and third 20 decades of the century sandalwood was replaced by béche-de-mer as the desired item of trade, and muskets, powder, and lead were the price that European captains had to pay in order to secure a cargo. By the time that the first resident missionaries arrived in 1835, Fiji had been transformed by the introduction ofthe musket. The first decade of the missionary presence saw an intensification of warfare which reached its peak in the conflict between the chiefdoms of Bau and Rewa on Viti Levu. These two polities, the most powerful in the islands, were linked by affinal and uterine ties among their chiefly elite, and the conflict was internecine. In the end it was Bau, centered on a tiny islet off the southeastern coast of Viti Levu which won the upper hand, but the Bauan victory owed much to their Christian Tongan allies. The Bauan leader, Cakobau, had already begun to refer to himself as the Tui Viti, the "king of Fiji." The subjugation of his Rewan adversaries did not end Cakobau's troubles. His ultimate victory had only been won at the expense of indebtedness to King George of Tonga. The Tongan ruler's kinsman, Ma'afu, was the leader of a confederacy based in Lau which rivaled the power of Bau, but a growing European community centered in Levuka town on the island of Ovalau proved to be the more serious threat. The 18605 brought a flood of Europeans into the islands eager to secure land on which to grow cotton made relatively scarce by the American Civil War. In 1871, in an attempt to control events, the aging Cakobau. proclaimed himself King of Fiji and formed a European- style government. At first there was little support among either 21 Fijians or Europeans, but when Ma'afu, the Tongan leader of the Lau Confederacy supported the new government, Fijian chiefs began to follow suit. Europeans were more reluctant, despite the fact that the new government was to be dominated by Europeans. Fijian representation was limited to the Privy Council which had no veto power over the decisions reached by the European Legislative Assembly. The Cakobau Government was a short-lived experiment. Plagued by debt, faced with dissension on the part of Europeans, and challenged by armed resistance by the Hill Tribes of Viti Levu, the Government was dissolved and an interim government formed. Cakobau made a second offer of Cession to Great Britain (the first offer had been made in 1858), and in October of 1874 it was accepted. Fiji became a Crown Colony of Great Britain. The complex events which saw Bau rise to political supremacy in the islands insured the prominence of the Bauan dialect and the Bauan' version of Fijian custom as well. Today Bauan is the lingua franca of the islands, and Bauan social organization, is interpreted by Colonial administrators steeped in nineteenth century evolutionary theory, served as a model for the codified version of Fijian society adapted by the Native Lands Commissions. The effects of this have already been touched on and will be discussed again later. The events of this brief, but turbulent, period during which the Cakobau Government struggled to extend its sovereignty into the interior of Viti Levu are still meaningful to the people of the area 22 in which I worked. During that time, Nairukuruku village was firm in its support of the Christian government, and this has affected its political position within the area ever since. To the extent that historical data is introduced in the discussion which follows, it is this period with which it deals. CHAPTER II LAND As one walks along the track leading southward from Nairukuruku village, one climbs several hundred feet to the top of a ridge which parallels the river. If it has rained recently, and in the months from November to March, it probably will have, the clay path will be as slick as wet soap. This, atong with water filled potholes of unde- termined depth, a few traverses across slippery, wet rocks, and a couple of steep climbs, requires one to concentrate on the immediate problem of footing. If the sun is hot, one will enjoy the opportunity to rest for a minute in the shade of the trees that grow around the first house in Taulevu village, Nairukuruku's nearest neighbor. I remember doing so on one day firparticular. The sky was a dramatic purple, and heavy clouds were piling up around the distant hills to the northwest. From my vantage point on the ridge, I looked back along the river, swollen and muddy from the runoff of the most recent shower. Far below and across the water lay a very flat piece of land bounded by the horseshoe curve of the river. The air was heavy. The upstream struggle of a motor boat seemed unrelated to the high-pitched scream of the engine. Wisps of mist rose from among the stands of banboo, grasses, and scrub that covered the flood plain. The natural 23 24 vegetation encompassed the irregularly shaped, but neatly defined, gardens and the tall, stately palms scattered among them. Its fertility periodically renewed by the flooding of the river that bounds it and the creeks that flow through it, this piece of land (see map in Figure 1) plays an extremely important part in the life of the village. The gardens they plant there provide the villagers with the bulk of their subsistence and yield the surplus of root crops which is their principal source of cash income. Accord- ingly, I want to begin my account here, with a discussion of the land and the villagers' relation to it. I will attempt to trace some of the factors that have affected this relation over time, but the account will be brief and selective. A. B. Brewster, an administrator who spent forty years in the hill country of Viti Levu, collected genealogies and oral histories of some of the principal tribes of the interior. He concluded that Although they relate to a period of almost unbroken warfare, but little blood was actually shed. They harried and chased each other, frequently burning villages, which were speedily replaced by others. The land was but sparsely populated, and there was plenty of uninhabited country for the Fijians to take refuge in. The impression on my mind after some study of their legends and folklore stories is that life in the hills in the olden times was like a huge game of hide and seek (1922:59). One who travels along the major rivers of the interior can not help but be impressed by the apparent vastness of Viti Levu. The fast-moving streams glide past sheer cliffs, and the forests have a sort of brooding silence. When the twistings and turnings of the river allow one a view of the horizon, it is one dominated by moun- tains and forests. It would seem that there have always been enough 25 empty spaces to allow for the population movements Brewster describes. However, moden surveys have shown that only about 30 percent of Viti Levu's total area consists of soils suitable for cultivation without major modification (Twyford and_Wright 1965:219). Considerations of precontact population densities and their relation to the carrying capacity of the land lie outside the scope of this discussion, but we can at least note some of the existing limitations on human habitation. Today, in those places where the Wainimala valley widens into broad alluvial flats, dairy cows graze on the grassy slopes above the rivercn'the rich soil is devoted to village gardens. These pockets of alluvial soil are among the best on the island. But despite their superior fertility, it is unlikely that large tracts of alluvial soils were much utilized prior to pacification. Their low relief, a key factor in the renewal of their fertility, also makes them difficult to defend. Villages tended to be located on hilltops, and there was probably a greater dependence on colluvial soils and patches of alluvial soils along smaller streams. Traveling through the upper Sigatoka Valley in the 18705, St. Johnston saw taro being cultivated on irrigated terraces (1889:263) and Sahlins (1962) describes a more rudimentary form of irrigation practiced on Moala. These more intensive methods are not practiced in the Wainimala area. Ample rainfall and the absence Of a marked dry season makes it unnecessary, and their ancestors must have cultivated their hillside gardens much as the inhabitants of Nairukuruku do today. By the 18705 the European presence had affected some changes in man-land relations in the interior of Viti Levu. The mechanisms 26 were multiple and complex and began considerably earlier, but by 1873 their impact became apparent. In March of that year warfare broke out in Naitasiri district downriver from Nairukuruku, and the conflict threatened the growing European community along the upper Rewa. Whales' teeth (33223) had been given by the Viria people (apparently the village of Naqali was heavily involved) to the Waikalou people of Serea village along with the request that they assist in an attack on the village of Naigunugunu (Outward Correspondence, Provincial Secretaries Office [Naitasiri]. W. S. Carew: FLetter to Minister of Native Affairs, March 30, 18739). The Viria people and their Waikalou allies were assisted by the Tai Vugalei people of Tailevu province. Fifty-seven of the inhabitants of Naigunugunu were killed in the attack. The Viria and Waikalou people warned white settlers that if there were any reaction by the Cakobau government or the European community, all of the settlers would be driven from the area (Outward. Cor., Prov. Sec. 0f. [Naitasiri]. W. S. Carew: FLetter to P. Storck, Warden, Naitasiri, April ,, 1873"). The Government responded by laying seige to the rebel strong- hold. Cakobau himself accompanied the government troops along with his daugher, Adi Kuila, and the chief magistrates of the provinces of Tailevu, Naitasiri, and Rewa. The government forces prevailed; the rebel population was marched off, and their chiefs were sentenced to five years hard labor. One of them, the Tui Waikalou, died while serving his sentence. Later in the same year W. S. Carew, then Provincial Secretary for Naitasiri, wrote to the Minister of Finance in the Cakobau 27 Government stating that "No land available for [revenue plantations] can be obtained nearer than Viria it being with the exception of about 400 acres wholly in the hands of whites. [O]f a river frontage of about 50 miles only about 5 are now in the hands of the natives" (Outwd. Cor., Prov. Sec. 0f. [Naitasiri]. W. S. Carew: “Letter to Minister of Finance, August 26, 1853"). Carew suggested that the lands of the Viria people be used for plantation purposes, for as he noted, "nothing will have to be paid for this land it being formerly the property of the rebels lately conquered by the king." Carew's letter reveals the extent to which European settlement had affected land tenure relations below Viria. But in addition to the transfer of lands to Europeans, Fijians along the upper Rewa were also putting their remaining land to new uses. There appears to have been a diffusion of the preeminent cash crop, cotton, to districts not yet controlled by Bau. In a letter to the Fiji Times, December 11, 1869, a trader calling himself "Englishman" writes that he "was informed by one of the planters [on the upper Rewa] . . . that the reason he could get no labor at one time was that the head chief of this district (after returning from a visit to the Vunivalu of Bau) put all his people on to clear ground and plant cotton." Cotton could be used to obtain muskets. The district in question had since turned against Bau, and VEnglishmanB reasoned that its inhabitants must have legitimately obtained cotton to sell. He was writing in response to letters to the editor (Fiji Times, November 20, 1869) con- cerning unscrupulous traders on the upper Rewa who were said to be buying cotton that had been stolen from white-owned plantations. 28 In the 18705 the frontier of European settlement was still some distance downriver from Matailobau, but it was closing rapidly. In October of 1872 Carew met with representatives of all the prin- cipal tribes of Naitasiri, including delegates from Matailobau and the neighboring district of Soloira. He reported that they readily agreed to obey the laws of the Cakobau government, to give up prison- ers, and to pay taxes to the best of their abilities. Carew explained to them that by doing so they would benefit from the protection of the entire kingdom "in the event of any unprovoked attack being made upon them by their heathen neighborsi (Outwd. Cor., Prov. Sec. Of. [Naitasiri]. W. S.lCarew: "Letter to Minister of Native Affairs, October 9, 1872). As we have seen the Waikalou of Soloira supported Viria in their rebellion against the Cakobau government, but the villages of Matailobau remained true to their pledge. The ruling kin group in this small confederacy of villages used their influence with the representatives of the Cakobau, ad interim, and colonial governments to secure their position vis-a-vis external rivals. Indeed, Carew referred to Nairukuruku, the ranking village of Matailobau, as the Government's bai ki colo or war fence in this area of Viti Levu. In its first legislative session the Cakoban government imposed a poll tax of-E 2 for Europeans and-E 1 for Fijians. Difficulty in col- lecting the tax led to the passage in 1872 of an act authorizing the government to sentence Fijian defaulters to hard labor On Europeanéowned plantations. In 1873 the government sought to extend its powers with an 29 act which would allow it to place rebelious districts under martial law and to hire out to Europeans those Fijians convicted under these provisions. The bill failed to pass, but regulations published on August 24, 1873, provided that Fijians sentenced to more than one year hard labor could be hired out to private persons. The Govern- ment's attempt to apply these regulations to the conquered peoples of Viria and Waikalou led to bitter disagreement between Cakobau's Chief Secretary, J. B. Thuston and Commodore Goodenough of the Royal Navy (Scarr 1973:274-5). Had this legislation been systematically enforced, its impact on Fijian society would have been great. At that time the payment of a tax in currency would have, almost necessarily, required working for European-owned enterprises. But in the interior of Viti Levu, the effects of the revenue policy were mitigated by the unsettled politi- cal state of the hill tribes and the government's own lack of power. I do not have the data necessary to assess the degree to which the people of the area were involved in the labor market, but I do know that as early as 1873 fifty men from Matailobau district were employed on a plantation on the island of Taveuni (Outwd. Cor., Prov. Sec. Of. [Naitasiri]. W. S. Carew: "Letter to the Secretary for Native Affairs, Tailevu, June 14, 1873"). The interim government which administered Fiji during the period immediately following the signing of the Deed of Cession abolished the poll tax, remitted all taxes still in arrears, and released all Fijians serving sentences on European plantations for 30 default. The poll tax was replaced by a labor tax which required every male Fijian between the ages of 16 and 60 to contribute twenty days labor on public work projects in their own provinces, an obli- gation which could be commuted by cash payment. While the revenue policy of the interim government was pri- marily a reaction to the supposed disruption of Fijian society caused by the poll tax, the tax in kind instituted by Sir Arthur Gordon, the first governor of the Colony, was intended to be an instrument of progressive change. It was Gordon's intention to stimulate the production of marketable commodities as the Fijian contribution to the development of the Colony. Since the tax would be assessed on a communal basis, it was hoped that it would promote the cohesiveness of village communities. And since the proceeds from the sale of com- modities produced in excess of the amount assessed would be returned to the producers, it was hoped that there would be an incentive to increase production. The tax was assessed ona provincial basis, the amount to be paid by each province being determined by the Legislative Council of the Colony. The government then solicited bids for the purchase of specified commodities. Each province was required to produce suf- ficient quantities of a particular commodity or commodities, such that when sold at the agreed price, they would yield an income equal to the tax assessment for that province. (Maize was the crap grown for tax purposes in Matailobau.) Within each province a share of the tax burden was apportioned to each district, and within each district, 31 the District Council set the amount to be produced by each village. Thus the responsibilities for the supervision of production fell on local Fijian officials, the bell, or district chief, and Efliéflénfli 5959, or government appointed village headman. While in some respects Gordon's tax program might be con- sidered a success (i.e., with respect to the Colony's financial situa- tion), the Fijian reaction to it was mixed at best. Gordon was annoyed by the use of the term "government gardens“ by European critics of the tax in kind (Legge 1958:243), but it would seem that Fijians shared the view that the system amounted to forced labor. In the correspondnece of European administrators one finds occasional complaints about lax supervision on the part of local Fijian offi- cials, a classic response to the demands of their intercalary posi- tion in the revenue system. The tax in kind has long since been replaced by other forms of taxation, but other of Gordon's policies have affected Fijians' relations to their land in a more lasting way. Article IV of the Deed of Cession granted the British sovereign and her successors absolute ownership of all lands which could not be shown to have become the bona fide property of Europeans or other foreignors or not in actual use by Fijians or required for their future maintenance. At the time the Deed of Cession was signed, 854,000 of Fiji's four and one-half million acres were claimed by Euopeans, though of this 854,000 only 16,524 acres were actually under cultivation (Legge, 1958:170). 32 In order for Gordon's program for their development to be successful, it was necessary that Fijians retain control of their lands. This meant that European claims had to be limited, and Gordon sought to do so. The legality of his position was challenged by lawyers in Fiji who pointed out that since the Deed of Cession excluded lands acquired by Europeans through bona fide transactions, the Colonial government did not have the right to. dispose of those lands, challenge the validity of European claims to them, or even the right to grant them to the cliamants. The investigation of any claim involved the prior issue of whether the land in question was or was not excluded in the Deed of Cession and, thus, was a matter only the courts could decide. The legal advisors of the Colonial Office eventually conceeded the validity of this position. Gordon sought a way out of this dilemma through special leg- islation, the Lands Claims Ordinance. This provided for the investi- gation of all European claims by a Lands Claims Commission which was to submit recommendations to the Governor in Council. The Governor's decisions on these recommendations were subject to a special board made up of the Governor, the Chief Justice of the Colony, and the Commissioner of Native Affairs. Of the 1,683 applicants considered by the Commission, only 517 were granted as claimed and another 390 . were granted ex gratia. In total 414,615 acres were confirmed to have been transferred to Europeans prior to Cession, about half of the acreage originally claimed (Legge 1958:194). Having dealt with the immediate problem of European claims, Gordon sought to limit any future alienation of Fijian lands. Under 33 Ordinance XXI of 1880 native lands were made inalienable except in cases of purchase by the CrOwn for public purposes. The boundaries of Fijian lands were to be determined and recorded. If a land-owning unit were to die out, its land would pass to the Crown, but, con- versely, where population increase made existing holdings inadequate, additional lands were to be provided by the Crown. The Native Lands Ordinance of 1880 and amending legislation empowered a series of Commissions to investigate traditional patterns of land tenure, on the one hand, and to record the boundaries of Fijian lands and register their owners on the other. The commission- ers' first efforts met withFijian indifference, but as the implications of registry became clearer and Fijian sensitivity on the land issue increased, indifference gave way to suspicion and, occasionally, open opposition. The effectiveness of the early commissions was also hampered by the dual nature of their goals. It was soon apparent that it would be extremely difficult to determine what the rules of land tenure had been in the past. Moreover, these rules differed widely from place to place, and Fijians were more interested in establishing a tenurial system that provided for present needs than they were in the investigation of traditional patterns. Despite the difficulties encountered by the Native Lands Commissions, several of Gordon's successors remained committed to the reestablishment of what was thought to be a common pattern of traditional tenure relationss There was a brief hiatus during the administration of Im Thurn, who put an end to the Commission's work and supported 34 legislation permitting the sale of Fijian land to Europeans, but ultimately one of the Commission's goals was realized. The boundaries of Fijian lands were recorded and the owners registered but only after the Commission abandoned its other goal, that of discovering and recording traditional patterns of land tenure. Indeed, the Commis- sion found it expedient to send clerks into an area well in advance of its hearings in order to assist Fijians in classifying themselves in terms of the segmentary lineage model which the Commission came to regard as the structure of Fijian society. The second step in the process was to identify the boundaries of the lands claimed by those neatly classified groups. Despite the evidence of local variation in tenurial relations throughout the group, Fijian lands were registered as the property of matageli, kin groups at the middle level in the segmentary model. The map shown in Figure 1 illustrates the results of this process with respect to Nairukuruku village. A few points are worth considering. Mataggli Siko is the largest kin group in the village account- ing for 22 of the village's 49 households with a resident population of 98. As the map in Figure 1 indicates, the members of matagali Siko are the registered owners of a total of 131 acres (plots #2 and #5). Of these, 73 acres (i.e., plot #5) includes the village itself and the comparatively infertile land around it. An additional 25 acres or so (i.e., plot #24) are reserved for matagali Siko and Nawaita jointly as subdivisions of the more inclusive kin group, yavusa Siko. 35 .mucmg mmmppw> mo aux--.“ «snow; ., .2 n2. ‘ , x we amzii.m mg:m_u \ 352.. :.s&::3 as M. «an? 98:388- ui... 3 I3— 5}: a... 3.3 Boo—32.3 7:32 a .23.... v... 3.3 3.38 fi— 40 41 hundred yards inland. The sand was deposited by the heavy flooding brought by cyclone Bebe. This land registered in the name of matagali Nakorowaiwai of 199999 Nabubuco, was formerly devoted to banana plan- tations. Today it is covered by brush and is left uncultivated, though cattle graze on it. Thus y_al9s9 Nabubuco- is, in a less favorable position than Nakorowaqa with respect to the most desirable alluvial soils. As has been indicated, disparities exist among land-owning units with respect to both the man/land ratio and the availability of a full range of soil types. The question then arises: to what extent has the 99”j999 tenurial system hampered the adjustment of these disparities? The answer up until now has been Anot at all," for mechanisms exist whereby a cultivator can obtain rights to the land which he requires. The category of land in greatest demand is fertile plots on the alluvial flats. As we have seen, Nakorowaqa is favorably posi—. tioned in this regard, and they have made surplus land available to others. In a sample of 25 cultivators, 14 men of other matagali have gardens on Nakorowaqa land. In response to my question concerning the ownership of the land they planted on, all 14 answered in terms of the 99Hj9§9_system. That is, they gave the name matagali Nakorowaqa or one of its senior men. Of these 14, five have either married Nakorowaqa women or have sisters married to Nakorowaqa men. Another six men are classificatory cross cOusins to men of matagali Nakorowaqa. Seven of the 14 are members of matagali Siko, the largest mataguali in the village. 42 In the same sample, 13 men cultivate on land registered in the name of matagali Nolovia of Taulevu village, though in response to my question concerning ownership most responded with the name of a senior man of that matagali. Some said that they simply asked for the right to plant on the land in question, though one man indicated that members of matagali Siko had rights in the land that were of long standing. One man claims that he presented a whale's tooth (99999) and kava root (199999) when making his petition. He lives outside of the village on Nalovia land and plants extensively there, including bananas. Nairukuruku is his mother's natal village; his father was from another village in the area. Villagers are aware of the general principles of.the legally enacted tenurial system and tended to answer my questions in terms of those principles. When the opportunity arises, they exploit the system to their advantage. But there exists another set of principles regulating the distribution of land, one which minimizes the dispari- ties that inevitably arise under the 99mi999 system. I will refer to this alternative system as ftraditional,§ a word which I mean to refer to informally regulated 9ractice (99”f9999) as opposed to a formally legislated code (d_e_ _j_u_r_e). (Whether the system so labeled is also ftraditional“ in the sense of antedating European contact is another question, one which we are unable to answer.) A similar system has been described elsewhere by Rutz (1977 1978) for the Waimaro people _ of the Waidina valley. Both can be seen as the outcome of two prin- ciples. Any land which is known to have been cultivated has an 43 owner, and the use of a garden site confers ownership on the cultiva- tor. A plot of land belongs to the last person to cultivate it until he transfers it to someone else or his use of the land is for- 2 In the Waimaro dialect land which has been cultivated but gotten. has since been allowed to go fallow is termed veimada, a term which also occurs in Matailobau. As Rutz notes (1977:167) knowledge of rights in veimada varies in direct proportion to the intensity of their use and their economic value, two factors which are themselves related. Thus people are more apt to remember rights to former garden sites on the alluvial flats, where gardens are in constant use, than they are hillside sites which require much longer fallow periods. The fallow lands of individual cultivators, as well as their gardens currently in production, are usually dispersed over several types of soil. Cultivators acquire their first garden sites from senior kinsmen. Informants told me that their children or children's children could plant on their veimada, assuming that they themselves did not require it. There was disagreement as to whether one's brothers and their children would have the same degree of access. They did agree that other kin, including uterine kin, would have to formally petition (kerekere) and perhaps present kava (y99999) in order to acquire one's fallow lands. One can be said to own gardens and fallow sites on land which, in terms of the 99 1999 system, is owned by a matagali other than one's own. Moreover, one can transfer those rights to another. Such 44 transfers are transactions between individuals, not the corporate land owning units of the 99Hj999_5y5tem. It was noted earlier that the traditional system minimizes the types of inequalities which inevitably arise under the 99 1959 system. I do not mean by this that the 99 £9999 system is merely an adaptation to an overly rigid tenurial system, a way of making it work. Rather, it is an alterna— tive way of defining tenurial relations, one which is admirably suited for meeting the needs of individual cultivators in a system of shifting cultivation. It would be incorrect to say that these are conflicting sets of rules, for they are applied under different circumstances. But the potential for conflict is present. The traditional system regu- lates the distribution of land among individual cultivators within villages and, to a limited extent, between villages. Garden sites held on this basis can be cultivated for subsistence or commercial purposes; indeed, the same garden may yield taro for sale as well as for home consumption. The transfer of rights to fallow lands are transactions between kinsmen, an expression of the ethic of reciprocity and solidarity that is the essence of those relationships. Land transactions under the 99“j999_code are a matter of bisinesi (from 3 In the English Abusiness“) and reflect an opposing set of values. some situations villagers do assert their rights under the legal code. One case which came to my attention concerned rights to royalties to be paid by the Fiji Electric Authority. The FEA was in the process of building the transmission line linking the hydroelectric project at 45 Monasavu with Suva, the capital. A dispute arose between two men of two different villages upriver. One man's claim was apparently based on prior cultivation of the land in question. Recourse to the records of the Native Lands Trust Board settled the matter in favor of the other man whose matagali were the registered owners. It is difficult to say whether instances of conflict between traditional claims based on use and 99_j959_ownership will be more commonplace in the future. One factor which might influence this is the intensification of commercial production. As the infrastruc- ture develops, the villages along the Wainimala are becoming more heavily involved in production for sale. Recently the Sawani-Serea road was extended to the Monasavu hydroelectric project in the heart of the island. The road now passes through the nearby village of Naivucini, about twenty-five minutes by boat from Nairukuruku. Middle- men come here fairly regularly to buy truck loads of taro and other produce for resale in Suva and other urban centers. In November of 1980 the Ministry of Agriculture gave the village a grant of F$3,900. (It is my understanding that the European Economic Community were the original source of the funds.) The money was to be spent on four pairs of oxen, wire, and staples for fencing, four plows and harrows, fertilizer, pesticide, and 10,000 taro suckers for planting.4 Each matagali was to fence off its land on the alluvial flats, and within each of these blocks land near the river and some higher ground away from the river was to be fenced off for cattle. The central portion of each mataqali's land would thus be 46 protected for planting. Each matagali was to plow one communal field and plant taro. In time secondary crops such as rice, maize, melons, and 9919 (a 2 variety of edible hibiscus) would be added. The produce grown in these communal gardens would be marketed and the proceeds would to to the matagali as a whole. Additionally, indi- vidual cultivators would have access to the oxen and would continue to plant their own gardens to manage as they saw fit. The initial stages of the project were to be directed by the village committee (599191H9195999) under the leadership of the village chairman. The oxen were purchased, the fences built, and the matagali fields plowed. The first crops were planted in January; I left the village two months later and do not know what the outcome has been. The plan, as explained to me by the agricultural officer administer- ing it, was a deliberate attempt to foster the cohesiveness of matagali. The income derived from the communal gardens would be available for reinvestment if the membership so desired. It is diffi- cult to say what effect the success of this project would have on the traditional system. Will matagali continue to recognize the usufructory rights of nonmembers? The Waidina valley is ecologically very similar to the Waini- mala basin, and there are strong historical and cultural ties between the peoples of the two areas. Rutz reports that when an all-weather road was extended into the Waidina valley, there was an intensifica- tion of production for sale and a consequent shortening of fallow cycles. In villages along the road, matagali whose lands lie adjacent 47 to the road or are cross-cut by it began to assert their 99_jure claims at the expense of the veimada rights of nonmembers. He also found that In those villages away from the road, where there is a high degree of variability between the amount of land and the size of lineage populations, there has been a reluctance on the part of disadvantaged villagers to abandon their veimada rights. Tensions have arisen between those who are members of lineages with much land and those whose lands are small. The latter are afraid of exclusion from desirable lands, while the former have argued that the legal code of meta ali boundaries will give them the security they need . in order to make capital investments in land (Rutz 1978:31). Such tensions have not yet appeared in Nairukuruku, but it is conceiv- able that an intensification of market production and internal demo- graphic pressures could combine to produce them. Patterns of land use associated with the traditional system result in detailed knowledge of the fertility, drainage, and other qualities of small plots. When questioned about the whereabouts of V these gardens (veiwere) and fallow lands (veimada), people may speak of their relationship to named places (vanua) that are dis- tributed here and there over the village lands. When asked where he is going, a man may respond with a place name such as Balabala (#tree fern"), and others will know that he is going to his garden. A few of these named vanua are old village sites, but the majority are simply garden lands which for one reason or another have assumed impor- tance as reference points. Vanua names may be included in the formal names of the matagali with which they are associated. Ceremonial presentation of 48 valuables is a central feature of Fijian social life, and every matagali has a formal name which is used when making speeches of offering and acceptance.5 For example, in such a context matagali Nakorowaqa is properly referred to as "Nakorowaqa-DelanaN; Nabubuco as "Nabubuco-Naikakau"; and Navitilevu as "Navitilevu-Mataiqereqere." In each case the second name in these hyphenated forms is the name of a vanua. Delana, which means "hill," is one of the matagali of ‘999999_Nakorowaqa listed in the records of the Native Lands Trust Board and was associated in the past with a tract of garden land across the river. Delana has since died out as a social unit. Simi- larly in formal speeches matagali Nawaita is referred to as FNawaita- Naitabuivalu." In the records of the Native Lands Trust Board Naitabuivalu appears as a matagali in M Siko, the 199959 to which Nawaita also belongs. Like Delana, Naitabuivalu was associated with a tract of land, but it too has died out in Nairukuruku. As one informant put it, matagali may die but the land lives on and, thus, continues to be important to the organization of the village. The formal names of matagali Nakorowaqa and Nawaita thus commemorate their link to social units which once stood in a sibling relationship to them as well as their link to the land with which these units were associated. We can say, then, that relationships to land are an important aspect of a kin group's identity, but land can also be said to be a mediating factor in the relationship between kin groups. The very fact of coresidence and mutual dependence on a tract of land is an 49 important form of bond between groups. Along with coresidence there is the bond of comnon allegienace to a chief, the ruler of the lands on which they reside. Groups sharing such a bond are said to be 99999, 9999, literally “land (or place together);" Such a relationship may be based on current coresidence, but it can also be based on past association. For example, matagali Nakorowaqa, Nabubuco, and Navitilevu occupied the area around Nairukuruku before the coming of 999999_Siko. Today they jointly receive first fruits in the annual ceremony marking the maturation of the yam crop. Navitilevu is a chiefly matagal'. Despite matagali Siko's Current ascendancy and Nakorowaqa and Nabubuco's liege ties to Siko, in some contexts their prior association with Navitilevu takes precedence. Similarly matagali Vusovuso and Nasautoka are said to have been vanua vata in the past near Naicuvacuva in tikina Wainibuka. Nasautoka, a chiefly matagali, is found in Nairukuruku's closest* up-river neighbor, the village of Navuniyasi. As a consequence of this association in the past, when they require the services of a chief, Vusovuso calls upon Nasautoka. Relationships basedcwipropinquity involve the responsibility of participating in one another's ceremonial obligations. Thus, all deaths and marriages concern the village at large, not just the kin groups of the principals. Similarly, neighbors outside of the village proper support one another in meeting their ritual obligations. Several months after our arrival in the village, a man of matagali Siko petitioned an elder of matagali Nakorowaqa for the privilege of build- ing a house near his gardens which already lay on Nakorowaqa land. 50 The Nakorowaqa elder also lived outside of the village, not far from the proposed housesite. The two men call one another "cross cousin" (199919). The house was built by men of the two matagali, Siko and Nakorowaqa. From that time on,the Siko man accompanied his Nakorowaqa host to all funerals, weddings, and other village functions, and on such occasions the two men presented their gifts together. In some instances, this placed the Siko man and his younger brother, who lived in the village, on opposite sides in the transaction. When I asked why on such occasions the Siko man joined with his "host? rather than his own matagali, I was told that he did so because he "stayed with” the other man. As I have mentioned already, matagali Nakorowaqa and Nabubuco are referred to collectively as the 1_199591. 1199991? may be trans- lated as "owner,“ but when used in reference to land, it can also connote ownership by virtue of being indigenous to the land in ques- tion. It does so in this case. According to tradition the ancestors of Nakorowaqa and Nabubuco lived in separate hamlets on their lands before the arrival of the ancestors of the chiefly matagali Siko. Because of their prior claim to the garden lands on which the village depends, these two matagali receive the first fruits of the yam crop (1 9999), a ceremony which marks the traditional coming of the new year. (Indeed, the word for Vyear,f‘999991, also refers to the year's yam crop.) In all the neighboring villages up and down the river the same pattern obtains; the matagali considered indigenous to the village lands receive the 1 sevu. This emphasis on the sacred aspect of 51 their relationship to the land does not overlook the fact that, according to their own traditions, the 1 199991_may have originated elsewhere. Whatis important is that their ancestors are believed to have been in the area before other groups arrived. The first fruit ceremoney will be examined more closely in a later chapter. For now I want to examine the significance of the 1 199991'5 relationship to the land, but the path taken will be some- what circuitous. The role of the Fijian chief in public ceremonial may be described as passive, even otiose. He sits at the head of the kava circle, and if there areruiother high ranking men present, his exalted position may leave him physically isolated. He sits, looks grave, and speaks very little. Oh those solemn occasions which Fijians describe as "heavy" (9191), a functionary known as the 9919 91_99999 (literally, "face of the land") speaks for the chief. The two positions are linked; the role of the chief requires that of the 9999 91 99999_and vice versa. It should be noted that in Nairukuruku the role of 9919 91 99999 is said to be properly filled by matagali Nakorowaqa and Naberbuco. They are of the land, and it is fitting that they also fill the role of "face of the land.9 In his “northern States of Fiji? Hocart argues that the mean- ing of the term 9919_91_99999_is broader than that of Chief's master of ceremonies or herald. He sees its more basic meaning to be the titular head of any clan (matagali), including chiefly clans (1952:16). That argument need not concern us here, but what is of interest is the 52 question which Hocart asks next. Why are these clan heads or, if we adopt the more restricted meaning of the term, these speakers for the chief referred to as "faces of the land“? Hocart says that the answer is clear if we look at the way in which the term Avanua" (land) is used on the island of Vanua Levu. The term "land" is there used in two senses, of a country, and of the sacred plot of earth after which that country is named. That plot is the place where the founder-god "came-up.“ When a man of Wainunu said "My land is in x," he does not mean "x is my country," as anyone might understand who had a superfie. cial knowledge of the customs and beliefs; he means "x is my holy ground." It comes to mean more or less the same, since the country is attached to the holy ground, but the point of views is different (Hocart 1952:17). Hocart views the relationship between "face of the land" and "land 999_sacred site" as one of the equivalence via metonymy. #The difference is the same as between a shrine (sava) and the face of a shrine. ..... Since it is the front that is always approached by the worshippers the shrine is to all intents and purposes its front" (1952:17). But if the term Fface of the land" refers to a sacred site, the place of origin of the founder-god of a matagali, how then does it come to be applied also to the leader of that matagali? Hocart con- cludes that the clan head is called "face of the land? (mata 91_vanua) because he is in charge of the sacred plot. Indeed "The chieftain is not only in charge of the sacred land: he is identified with it? (Hocart 1952:17). The details of Hocart's etymological arguments need not con- cern us. The correctness or incorrectness of his conclusions are less important than the tone of his argument. By that I mean, Hocart's 53 emphasis on the sacred as a basic feature of a kin group's relation- ship to its land. In the Wainimala valley that is very much the case. The relationship of the 1 999991 to the land is mediated by their relationship to their 99_(ancestral deity) whose domain it is.6 It is the99 who insures the fertility of the soil and the success of the crops. Exactly what the relationship is between the 99_and Na Kalou, the Christian God, is left unresolved. Both are necessary for the well-being of the community. Na Kalou can be approached through prayer and collective worship, but the 99_of the place can only be approached through his living representatives, the elders of the matagali 1 Latel- If one intends to present a complete account of the signifi- cance of land in contemporary Fijian society, it is not enough to treat it as a physical phenomenon and source of livelihood. Nor is it enough to add a discussion of the legal and social dimensions which effect its distribution and use. A recognition of the fact that man's relationship to the land is mediated by his relationship to the supernatureal does not complete the picture. ,For in Fiji today land has also become a potent symbol in the political arena. A minor- ity in their own country, Fijians view the control of their lands as the most important factor in their survival as a people. For the Indian community security of tenure on leased property is an equally vital issue. One evening during the funeral of a village elder discussion turned to the land question. The hour was late. Senior men from 54 several villages were gathered around the kava bowl in a temporary shed (vakatuniloa) beside the dead man's house. Most of the conver- sation which occurs while drinking kava (9999991) is anecdotal in nature. Men exchange stories (1 talanoa) about one another and other persons known to those present. Often the humor of the story depends on the listeners'knowledge of the character and foibles of the prin- cipals. A good story is worth repeating to newcomers, perhaps becoming more elaborate in the process. At funerals, however, espe- cially in the presence of senior men, conversation tends to be more subdued. On this occasion someone posed the question, fwhich is more powerful (kaukauwa), land (99999) or money (1_1999)?" One by one those present offered their opinion on the subject. When it came my turn to respond, I was at a loss for an answer. The question was characteristically phrased in terms of absolutes, an Aeither/orf choice, but to answer the question I felt that I needed to relate it to some context. Was I being asked to comment on the position of Fijians vis—a-vis other segments of the population? Was I being asked whether that position would be enhanced if there were more of the one (money) even if that meant less of the other (land)? And what did they mean by Amore powerful?9 Were they talking about political clout? Needless to say no one else present had any difficulty in answering. Most said that land was more powerful. They argued that Fijians with money but no land would be in a vulnerable position, 55 whereas that would not be the case if the situation were the reverse. Money can be lost or spent, but the land remains. One man, a retired school teacher, I believe, opted for money, arguing that today's economy is based on it, and power lies in the hand of those who have it. His comments provoked an effort on the part of the majority to convince him of his error. One or two men said that land and money are equally powerful, an answer which satisfied no one but themselves. It became clear to me in the course of the discussion that the majority of thse present were not talking merely about the pragmatics of political and economic power. They were also talking about the relative value of symbols; about what they considered to be contrast- ing ways of life; and about their identity as Fijians. In affirming the value of LAND, they were not only assessing its political and economic significance, they were also affirming the value of a way of life and identifying themselves, as Fijians, with it. It is no accident of language that the Fijian term for custom is vakavanua, "the way of the land." FOOTNOTES--CHAPTER II 1These are mtagali Siko and Nawaita. The land commission includes matagali1Navoka in avusa Siko for the purposes of land registration, but it is not included in 9avusa Siko as that unit is defined locally. As will be explained later each 9avusa is asso- ciated with a particular species of plant and fish or animal. The plant of matagali Siko 9nd Nawaita is the wil.d sugar cane or vico and their fish is the vo. The totems of mata ali Navoka are differ- ent. Today there is OETy one member of matagai avoka in the village, a woman married to man of matagali 1$iko. 2Belshaw (1964: 185) found the same principle obtaining in Nadroga, and Sahlins' comments (1962: 219) suggest that a similar situation may have obtained on Moala. 3The legal code permits the leasing of matagali lands by private persons, but not their sale. 4One of the five matagali in the village already had a team of oxen, a plow, and harrow. 5The most commonly exchanged valuables include whales"teeth. (tabua), padanus mats (ibe), and barkcloth (masi). Kava root (9agona) an nd feasts (magiti ) are—ETso formally presente ted 6"Vu" is a word with the very general meaning of "the basis or root (source) of a thing. " Thus a Lu ni kau (kau= a stick or piece of wood) is a tree; a vu ni wai (wai= water, and by extension, medicine or wai ni matET'1s 3THOctor, the source of medicines, and a Lu ni volSTTVOla = to write) is a scribe or secretary. Some authors report a distinction between kalou Lu and vu. Hocart identifies the kalou as the souls of the dead and— Lu as ancestor spirits who are equivalent to the pre-Christain gods (1952: 9). Sahlins, on the other hand, reports that on Moala the kalou Lu are ancestral spirits and supernatural guardians while the La are ances- tors with more human attributes (1962: 228). This distinction was not made by my informants. The word “LO" was applied by them to both the founding ancestral deities, who were never men, and the spirits of human ancestors. The meaning was made clear by context (Cf. Geddes 1945:43 and Spencer 1941:9). 56 CHAPTER III "EATING TOGETHER" Nairukuruku, like many Fijian villages, is a community based on the association of kin groups which claim to have come into the area from separate points of origin. They differ in terms of rank, ritual specifization, and totemic associations, but these differences are countered by the bonds of kinship, affinity, and common political allegiance which have evolved over time. Perhaps the most important source and expression of their unity is their mutual dependence on village lands for their livelihood. The food derived from their gardens is literally the source of their substance, and thus the shar- ing of food becomes a potent symbol of solidarity and the exchange of food an eloquent expression of differentiation within an encompassing unity. Soon after our arrival in the village we became aware, mostly through direct statements, that people had definite ideas about the relationship between food and well-being and about ethnic differences in food habits. The nutritional virtues of locally produced foods, principally root crops and boiled greens, were repeatedly extolled and contrasted with the weak, bland (malumaluma) food considered to be basic to the European diet. In keeping with Fijian notions about the key role of carbohydrates, this is thought to depend very heavily on 57 58 bread and potatoes. There was a concern that we appreciate the local food, and when I met Fijians from other places, they would invariably ask me or my hosts whether we ate Fijian food. Men were proud of the fact that they could work well into the afternoon before eating their midday meal. It was common knowledge that Europeans had to eat promptly at noon and that they tend to be measured in their food intake, one more expression of what was believed to be Europeans' highly regulated approach to life. We became aware also of the importance of commensality and the giving and receiving of food. There was a great deal of interest in, and sometimes jealousy over, who gave what to whom, not just because of the things that were given, but also because such exchanges are expressions of solidarity and kinship. Finally, we became aware of the value placed on producing food as an activity in and of itself. All men are not equally active and diligent with respect to food production, and their gardening abilities are regarded as important personal traits. Old men continue to plant and harvest food as long as they are able, at least in part because that is a man's proper function. In focusing on the significance of food in Fijian life, this chapter and the next will touch on many other topics: the division of labor and the relationship between the sexes; the symbolic associa- tions of the principal crops, as well as the techniques used in cultivating them; and the role of food in exchange. We will see that not only is food exchanged between social units, but also that, 59 ultimately, the production of food is dependent on exchanges between humanity and the spirit world. Let us begin by exploring the physical setting in which food is most commonly shared, the interior of a Fijian house, for an under- standing of the symbolic value of space within the house is necessary in order to discuss the etiquette of commensality. A Fijian house typically consists of a single rectangular room. Despite the absence of any physical barrier, this space is symbolically divided into two areas with a gradation between them. In many homes this division is marked by a long strip of decorated barkcloth hung from a roof beam across the width of the house. Photo- graphs of family members are often hung along this decorative strip, sometimes fastened with cowry necklaces or strings of plastic flow- ers. All households have separate cookhouses (9919919999) where most of the food is cooked and the related equipment stored, but families usually eat in the house. And so at one end of the main building, near the door, in fact, there is often a food cabinet and some equip- ment for the preparation and serving of food. There may be a small kerosene stove, wash basins, pots and pans, and so on. At the oppo- site end of the house is the sleeping area of the household head and his wife. There may be a frame bed covered with decorative mats or perhaps simply sleeping mats in one corner. There are likely to be numerous pillows, perhaps stacked in this end of the house, but these are just as likely to be distributed here and there throughout the room, handy for anyone who wants one. If the family has a chest of 6O drawers, it will probably be at this end of the house. Extra clothes may be stored there or in suitcases stacked along the wall. Valuables may be secreted in these or under the bed, a place off limits to all but family members. Fringed mats overhang the length of the bed, and separate cloths are then tucked under the mats to overhang the end of the bed. Thus the bed is also a storage area protected from view. The whole sleeping area can be shut off by curtains when privacy is desired. These are an indispensable feature of household furnishing, as important as the mats which cover the floor. I These, then, are the two poles within a Fijian house, the sleeping area at one end and the cooking area at the other. They are spoken of as the upper (9919) and lower (99) ends respectively. The lower end of the house is preeminently the domain of women. It is there that the female head of the household sits when serving meals, washing dishes, and so on. Despite the fact that it includes the sleeping area of the married couple, the upper end of the house is a male domain. If you enter a house at some time when both the head of the household and his wife are at home, you are likely to find him and his guests siting or lying near the upper end and his wife at the lower end, perhaps listening to their conversation while plaiting a mat or busying herself with some other household task. When only family members are present, the use of space tends to be more relaxed. There is something of the opposition of the sacred and profane in this structuring of domestic space. In part this separation of the upper end of the house is an expression of respect for the privacy 61 of possession, but there is also the fact that it is not just a sleeping area. It is also a place in which conjugal relations take place, and a measure of separateness attaches to it for that reason. I once attended a 999999 (kava) session at which the host's wife's brother was also present. The wife's brother is middle-aged and married. When someone enters a 999999 party at which other people are already seated, it is customary to offer them a seat. One honors the newcomer by asking him to sit 919919_("above"), toward the upper end of the house. On this occasion the wife's brother sat slightly below the midpoint of the house, lower, in fact, than other men younger than he. I asked him to move up, but he declined saying that he could I In not because his sister slept in the upper end of that house. fact, on another occasion he did sit at the upper end of his sister's house, but his comments on the first occasion are nonetheless signifi- cant. It is not the association with sexuality and, hence, procrea- tion which explains why the upper end of the house is male, for neither the act of generation nor generative power are conceived of as exclusively male. It is, rather, this fact which explains why the upper end of the house is kept separated from the profane area of cooking and ordinary coming and going. .Because it is tinged with the sacred, it is conceived of as the fhigher“ end of the house. Using the same spatial idiom men are higher than women just as chiefs are higher than commoners, and thus the upper end of the house is man's proper place. 62 All houses have more than one door. Most in the village have three and a few have four. Doors are functionally specialized. The one at the lower end of the house is used by most people most of the time. Cooked food is always brought through this door, and women who are not members of the household should always use it. Unless beckoned by someone within, most men who are not household members also use the lower door. Two additional doors are usually placed at about the midpoint of the longer sides of the house. These “upper" doors tend to be freely used by household members, but, properly speaking, these doors are for the use of the household head and elders, especially those of chiefly rank. If there is a fourth door, it will be at the upper end of the house directly opposite the lower door. I have never seen anyone other than the most senior men of chiefly rank use these doors. Before one enters a house other than one's own, good manners dictate that they announce their presence. The formula (1999) that is used in doing so differs for men and women.2 Upon entering, men will be expected to go towards the upper end of the house, but where they sit with respect to their host will depend on their respective age and rank, as well as kinship. An older man, especially if he be chiefly, will be asked to sit above the host. The older brother, father, or mother's brother of the household head would do so as well while younger kinsmen would normally sit below their host. Among men of approximately equivalent age, especially those related as cross counsins, relative positioning is much more relaxed, though a good host will always ask such a guest to imove up." 63 Not to belabor the point, then, space within a Fijian house is conceived of hierarchically and becomes an idiom for expressing social hierarchy. This is especially marked in the case of seating arrange- ments at meal time. When the meal is ready to be served, a long cloth is laid on the floor down the length of the house. The seating arrangement reflects the dual principles of gender and seniority. The eldest male of the household sits at the upper end of the cloth, younger males below him on either. side,and children of both sexes below them. If the household is a large one, two settings might be necessary in order to serve everyone. The female head of the house- hold, perhaps assisted by an adolescent daughter, sits at the lower end of the cloth. The pots of food sit beside her on a bit of cloth or board to protect the mats from soot and heat. A If this were an afternoon (vaka siga levu) or evening meal (vak99 yakavi), it would probably be centered around boiled taro or some other root crop. Within the category of "food" (999999), Fijians distinguish between "true food" (kakana dina) and the relish which accompanies it (99.9999 1 991). The subcategory of "true food" includes the traditional root crOpse-taro (£199), yams (999) and sweet potatoes (999919)--but also the more recently introduced tavioka (manioc), some tree crops such as bread fruit (919 9999) and plan- tains (9991), and wild yams (9E919e_i1<99). Some of these are seasonal (domesticated and wild yams, bread fruit), while others are not (taro and tavioka). At certain times of the year, such as just after the harvest of the new yam crop or in the peak of the wild yam 64 season, very little taro is eaten, but over the course of the year, more taro is eaten than any other form of "true food." Still the various forms of kakana dina provide a sufficient variety to alleviate boredom, and then, too, several varieties of taro are cultivated, each characterized by differences in taste and texture. For the morning meal (1 katalau) some form of cooked flour is often substituted for kakana dina. Roti soaked in sweetened coconut cream, fried cakes (9anikeke), or cabin crackers along with heavily sweetened tea is a common breakfast menu. Rice may also be substituted for "true food." Kakana dina is boiled without salt; salt is placed on the cloth, and people sptnkle it over their food to taste. Small citrus fruit (9911) and hot peppers (999999) are normally the only other condiments provided. Occasionally breadfruit is baked by placing it directly in the fire, and when cooked like that, one can understand why it was given its English name. On festive occasions taro and other foods are baked in an earth oven (1999).3 Baked taro is indeed delicious, but when boiled it tastes exceedingly bland to the uniniti- ated. In time one learns to appreciate the taste. Not only does kakana dina play a central part in most meals, it is also spatially centered. It is placed on serving platters dis- triubted at intervals along the cloth, and people help themselves to it. But the relish (1 991) is served by the female head of the household. If there has been time to fish, the relish may consist of boiled fish or prawns in broth, and, if not, perhaps tinned fish or, more rarely, tinned beef or fresh meat. But on a daily basis, 65 the most common form of relish is boiled greens, either taro leaves (999999) or the leaves of a variety of hibiscus (9919 or, in the local dialect, 999999). These may be simply boiled in water, but usually at least once a day they will be cooked in coconut cream and served as a soup. Before mealtime, one can hear the rhythmic grating of coconut all over the village. Both before and after the meal fingerbowls are provided. If the relish is cooked in broth, spoons may be distributed. Knives used for cutting off portions of yam or taro are the only other uten- sils. A bottle of water and a single glass may be placed on the cloth for those who wish to quench their thirst after eating. The meal begins with a prayer (9999) of thanksgiving offered by the senior male present. There is usually little conversation during the meal, mainly requests for this or that. In the household ' of the paramount chief and on public occasions when he or some other senior man of chiefly rank is present, all of those present clap when the man of rank has finished his meal. As people finish eating, they pass their plates down to the hostess, thank her for the meal, and retire, making their places available for others. After eating, it is customary to recline (9999) for a few minutes. This is viewed as not merely a matter of comfort, but also as an aid to digestion and, therefore, conducive to good heatlh. Except in cool or inclement weather, all the doors of the house are left open while meals are eaten, and therefore, the members of the household are visible to all who pass by when they sit down to 66 their meals. Because the sharing of food is a statement of solidarity among kinsmen, and since all who live in the village should behave as kinsmen towards one another, good manners dictate that any passerby be invited to join the commensal unit. In the morning the invitation is likely to take the form of an invitation to drink tea. (1991. 19991j.11i") At other times during the day the invitation may take the form of a command to eat (1991. .99991"), or one might simply call out the name of the meal being eaten. Our daughter, who was two years old at the time, became very conscientious about this point of etiquette. (Her performance was doubly appreciated because she extended, whenever possible, the proper kin term for the person to whom she was calling.) Such invitations are a matter of convention. The polite response is to thank the person for the invitation and walk on. Only those whose kinship ties are especially close (e.g., a sibling, parent, sibling's spouse, or spouse's sibling) are likely to drop in at meal time, and such close kin really need no invitation. But if others do visit at mealtime, no fuss is made. If the person wants to eat, room will be made at the cloth. If not, he or she will simply wait until the meal is finished. There is usually quite a bit of visiting during and after the morning meal, a time when people talk over their plans for the day. When guests do share a meal, the good host urges them to "eat big? (kana vakalevu).4 The female head of the household is very much in charge of "putting the meal on the table.f She fills the plates one by one and hands them up the cloth in order of seniority. People may request more 67 after they have eaten what she has given them, but no one ever com- ments on the portions given. A woman herself does not eat until everyone else has been served, often after all but the youngest house- hold members have finished eating. While others eat, she and any helpers she may have sit beside the cloth waving a hand or cloth to scatter the flies so plentiful during the hot season. At both large gatherings and more intimate family meals, it is not uncommon for senior males to pass the food left on their plates to a junior kinsman, a child, or even their wives. Sahlins reports that on Moala it is believed that eating food remaining on the plate of an elder kinsman results in a magical swelling of the throat or stomach, an ailment known as 9919_(1962:110). In the Wainimala area a woman's brother and a person's mother's brother (9999) are the only categories of relatives from whom one must not accept leftovers. I was told, in fact, that in the past this prohibition extended even to taro that had been carried by one's 9999 in the customary fashion, on a pole slung over his shoulders. Eating from the same plate or drinking from the same cup as the mother's brother indicates disrespect and may result in mystical sanctions of an unspecified sort. It is doubtful that similar prohibitions ever extended to parallel kin of adjacent generations (i.e., fathers and children). On those occasions when leftovers do change hands, they pass from a senior to a junior kinsman, and, thus, a clear statement of the related principles of seniority and rank is inherent in the act. On the surface it would appear that a similar sort of statement is 68 implicit in the deference of woman as they wait upon their families. Eating last they may be said, in a way, to be eating leftovers. But it would be wrong to interpret the position of women in the commensal unit as simply one of subservience. Though that aspect seems to be underscored at mealtime, there is another side to women's role in the household economy. While the responsibility for food production is shared by her husband and any other household members able to con- tribute, food preparation is primarily the woman's responsibility. It is an important aspect of the domestic economy over which women exercise considerable autonomy, and they derive satisfaction from seeing that their families eat well. There is, of course, a good deal of drudgery involved in the preparation of meals. Children of both sexes, adolescents, and even the male head of household may assist in the collection and preparation of fire wood, but ultimately the cook fire is a woman's responsibility. Someone must be up at dawn every morning to clean the pots, start the fire, and prepare breakfast. If there are no adolescent girls in the household, a woman must perform these tasks herself. Usually deci- sions about the harvesting of food are made by the male head of house- hold in conjunction with his wife who also assists him in transporting the food from the garden to the village.5 When a married couple go to the garden to harvest food, there is often a division of labor. After completing any other garden work that might be required, the husband harvests the taro while his wife gathers greens (9999). Providing the relish (1_991) for the daily meals is primarily the woman's responsibility. On one occasion my family and I 69 accompanied a party attending a funeral in a village upriver. When it came time for the boat to return to Nairukuruku, a woman who had accompanied the party decided to spend the night in the host village. (She had been born there and was closely related to the family of the deceased.) She gave my wife fifty cents, about the price of a large tin of fish in the village store, and asked my wife to give it to her husband "for his 9999." On a daily basis women bring in most of the animal protein that is consumed. Cattle are only killed for ceremonial purposes; chickens and ducks are also reserved for special octasions; and the killing of a wild pig is a comparatively infrequent occurrence. The most common sources of animal protein are fresh water fish and shell- fish. The amount of fresh fish consumed by a household depends on a number of factors including the number of small children a woman has to care for and her own attitudes toward work. Women use a number of fishing techniques depending on the size of the stream, the water level, and the number of women involved. When fishing in the Wainimala, a broad, fast-moving stream, women usually use cane poles or thow lines. When the water level is low, dip nets may be used though they are more appropriate for use in smaller streams and ponds. Young girls combine work with play, wading in the river and dislodging cobbles to capture fresh-water prawns as they attempt to escape. After the floods of the rainy season, the retreating waters leave large numbers of fish trapped in shallow ponds. One day I accompanied an informant to his gardens, and after the work 70 was over, we returned to the river by a path that was unfamiliar to me. Off in the distance I could hear what seemed to be many voices chattering and laughing. We passed through a grove of trees and came upon an ox-bow pond full of women and girls wading and splashing about with dip nets. I am sure the fish were justifiably terrified, but their would-be captors werezcertai nl y enjoying themselves. The fish that such efforts yield are often very small in size, but they are sweet-tasting and are boiled, the broth being drunk as soup, and nothing is wasted. Occasionally large numbers of women cooperate in fish drives, and on those occasions, the catch can be impressive. When a person dies, it is customary to 9999_fishing in a designated stream or on a defined section of a given stream. In December of 1980 a highly respected man of chiefly rank died, and a stretch of the Wainimala was placed under a 9999, This section of the river included a shallow pond-like cove. It was from this cove that the majority of the fish were taken when, one hundred days after the death, this section of the river was fished to provide the relish for the feast marking the end of the official mourning period. About thirty women, two men, and several youths participated. As with all significant undertakings, the fishing was pre- ceeded by the drinking of 999999 (kava). It was mixed at the site by a young man of the dead man's lineage and presented by him to two men from Taulevu village, for this stretch of the river is actually within the boundaries of Taulevu, and the.cove in question was only a 71 few hundred yards from a house of that village. The 999999_lifted the 1999 prohibiting fishing. When the 999999 was completed, the women began entering the chest-deep water and sealing off the cove.~by placing their dip nets side-by-side. Women and girls kept arriving some on foot, others more or less floating down the river from Nairukuruku, until there were enough to complete the enclosure. A boat stood by behind the line of nets to receive the catch. Spirits were high. Girls and youths swam to the head of the cove and then, yelling and splashing, drove the fish into the nets. In the first several drives, fish were being caught so rapidly that they were flying into the boat from all directions. The species of fish being caught (991999) has long, bony spines in the dorsal fin which prick like needles. Those in the boat dodged them as best they could. Occasionally a women would hold a fish in her mouth so as not to leave her place in the line. The catch from each drive gradually diminished, but by the end of the morning, it totaled about 100 fish weighing between I and 1; pounds.6 Perhaps in part as a consequence of the sexual division of, labor, men and women appear to have different attitudes towards food, or at least express different attitudes, and these attitudinal differ- ences would appear to have physical consequences. Men talk very little about food. When they drink 999999, which they do frequently, men do not eat. If a man plans to drink in the evening, he may take an afternoon tea, perhaps with boiled greens and taro. Often this is unnecessary, for if he is late returning home from his gardens, a man 72 may not eat his midday meal until later in the afternoon. Late in the evening when the 999999_bowl is empty, if their wives cannot be per- suaded to serve a hot meal, men may eat a snack of cold food before retiring or go without. As guests at weddings and funerals, occas- ions on which meals are always served, men seem to positively vie with one another for the honor of being the last to leave the 9a_q_09_a_ bowl to eat. It is often necessary for a representative of the host group to ask them to do so, pointing out that the women are tired of waiting to serve them. Men do not rush through their meals, but neither do they linger. As a consequence of all this and their phy- sically active lives, most men remain trim into old age. Perhaps a lifetime of preparing food, waiting on others, and expressing themselves through the giving of food results in a differ- ent attitude toward it on the part of women. Women spend more time talking about food, and at meal times if they eat later, they also eat longer. They are also more apt to eat between meals than are men, if for no other reason than the fact that their leisure time is spent in the house or cookhouse where food is available. As a consequence, while they certainly lead physically active lives, with few exceptions, they tend to get heavier as they get older. In part this may reflect 'the belief that health and appearance are enhanced by maintaining a stout (levulevu) physique. In part, too, it may be a consequence of childbearing. For the first six weeks after giving birth, a woman rarely ventures out of the house and only does the ligher tasks. The bulk of the work she would normally perform falls on other female 73 members of the household or extra-household kin, an older daughter or younger sister perhpas. If no such assistance is available, her husband may take over some of the work a woman would normally perform. Her main concern during her confinement is the care and feeding of the new baby. Some foods are avoided by nursing mothers because they are believed to be harmful to the infant. Other foods are believed to help produce rich milk. Among the latter is 99_9999, shredded taro stems boiled in coconut cream. §9_are also believed to help in tightening up the uterus after childbirth. The English aphorism "A man's house is his castle? is appli- cable to Fiji as well, but if a man's house is his castle, a Fijian woman's cookhouse (9919 91_9999) is hers. Or perhaps it would be better to think of it as an important node in a woman's network of communication. While women certainly visit one another in their homes, at church, and so on, a lot of socializing also takes place in the cook house. In old age women's networks begin to shrink; they tend to spend more and more time in their cook houses beside the fire, receiv- ing visitors but venturing out less and less. It is there that a woman has produced the cooked fbod which she has given, not only to the members of her own household but to other households as well, an impor- tant form of communication about the value she placed on relation- ships. At meal time women and children criss-cross the village care- fully bearing dishes of cooked food covered with a plate or cloth. When they arrive at their destination, the women of that household 74 will transfer the food to one of their own plates and refill the dish with food from their own household's meal. Or perhaps the gift will be duly noted and reciprocated at another time. One informant stressed that these gifts of cooked food should only be given before the meal. To give food after the donor household has eaten is tantamount to giving leftovers, and the recipents would be insulted. These interhousehold presentations of cooked food (1_takitaki) may occur at any meal on any day but are most common at the morning and midday meals on Sunday when more time is available. No one goes to the garden, children do not go to school, and meals tend to be more elaborate than on other days. Any category of cooked food might be given. Sometimes the gift will consist of boiled yams or taro, and the recipients may make a counterpresentation of the same type of food. Probably more common, however, are gifts of pastries at the morning meal and relish dishes (1_991) at other meals, food in which the cook can take pride. It is women who initiate these gifts of cooked food and women who reciprocate them. I recall one Sunday afternoon when a woman from a nearby household brought over a portion of boiled eel in broth. She had caught the eel herself the day before. She told us that the eel had fed a lot of households--her own, ours, her brother's, the households of her husband's two sisters, and one other. Of all the households in the village, these were the households with whom she and her family interacted most frequently. Another woman in a nearby household sent food several times a week to her unmarried brother 75 and his children. A third woman regularly exchanged food with her parents' household with whom her only son often stayed for days at a time. Some women were more active than others in these exchanges. But most households followed this pattern; exchanges were most fre- quent with one or more nearby households, just one aspect of the ongoing flow of reciprocity, but, it is important to stress, one which is controlled by women. From time to time events occur which require food contribu- tions from a number of households. When a man asks others to assist him in building or repairinga house or planting a garden, it is under- stood that he will provide a meal and 999999_afterwards. The women of related households and immediate neighbors may assist by contribut- ing cooked kakana dina. Other events such as a small feast in honor of a child's baptism are a matagali concern, and women who have married into the kin group as well as those who had been born into it will contribute cooked food. Some events, such as funerals, will involve contributions of cooked food from all the households of the village. Any funeral will attract mourners from other villages, and the funeral of a senior man of chiefly rank may attract several hundred persons. The visitors (991991) will have to be fed three times a day during the four-day funeral. The burden of providing for them falls most heavily on the household of the deceased and the deceased's matagali, but all other households should contribute cooked food (i.e., kakana dina) as well. The degree to which they actually do so may depend on the public 76 perception of the willingness on the part of the "owners of the death" (1M9199t9) to make outlays. In the case of a funeral such contributions are a matter of kinship and coresidence, but occasionally other events occur which are more stricly a civic matter. The village is the site of the dis- trict school serving its OWn children and those of two neighboring villages. Similarly, the minister (1 talatala) presiding over the village congregation is also responsible for these neighboring villages, and, moreover, the village is the administrative center of the local diocese (tabacakacaka vakalotu) of the Wesleyan Church of Fiji. And finally, as the home of the paramount chief, Nairukuruku is the highest ranking village in lower Matailobau district. For these reasons from time to time the village is called upon to host events related to the running of the church and school. The responsi- bility for organizing such events falls most heavily on the relevant committee (e.g., the school committee or 999191_91.9999 9911) but will ultimately involve the entire village. If the visitors are to be fed, the paramount chief, in consultation with other elders, will decide on the contribution to be made by each household. It will then fall on the village chairman (119999_91_9999) and town crier (999999 91 9999) to implement their decisions.7 These assessments may be in the form of raw or cooked food, depending on the nature of the event. We have mentioned the fact that personal, informal presenta- tions of cooked food are largely controlled by women. In contrast, formal, public presentations of raw food (999191) are made by men. (The contrast between raw and cooked food is less significant than the 77 fact that the presentation of 999191 is formal and public, whereas the interhousehold exchanges discussed earlier are not.) The Fijian word 1999191" is usually translated as “feast.9 However, rather than being eaten by donor and recipient together, a 999191 is usually given uncooked and is divided among the recipients to be eaten later. A ,999191 normally has three components. It includes kakana dina, some form of relish (1 c_oi_)_, and a container of La_q9n_a_ (an unprocessed 999999 root or, on especially solemn occasions, a whole shrub may be substituted for mixed 999999,) Only taro and domesticated yams are suitable for ceremonial presentation; breadfruit, tavioka, or plantains cannot be substituted. The kakana dina may be raw or cooked, but the 1_999_is always uncooked flesh. Prior to Christianity it would have been human flesh; today it is usually beef or, more rarely, pork. A variety of occasions can provide the motive for the giving and receiving of 999191. A 999191_is normally presented on the fourth day of a funeral (99'99, literally "fourth night") and in the context of subsequent exchanges associated with mourning. A 999199_is also presented on the fourth day of wedding celebrations. They may also be presented when special visitors arrive or depart or may be given as compensation to those who have assisted one in major undertakings such as house building. The presentation of a 999191 is always a transaction between two groups, often part of an ongoing series of exchanges. All of the presentations which I observed took place while men were already gathered around the 9agona bowl. As the older men drink, young men 78 carry in the carcases of the cattle. When all is made ready, the senior men who choose to witness the transaction separate into two groups, donors and recipients, facing each other on either side of the 999191. The carcases of the animals to be presented lay on palm fronds, often with their legs towards one group and their backs towards the other. The other items to be presented, 999999_and kakana dina, lay beside them. I have referred to the donors and recipients as groups, but on most occasions on which 999191_change hands, they would be more accur- ately described as "groupings.“ For example, funerals are organized around a series of exchanges which include the giving and receiving of ,999191. These exchanges take place between the matagali of the deceased, the 999991_91_9999_or "owners of death,9 and that of his mother and mother's brothers. The latter matagali stands in a special relationship to the deceased. They are his 9999, a term which in other contexts refers to the general category of "relatives" but which, in the context of death, assumes this more restricted meaning. All of this will be discussed in depth later, but here I want to make the point that these two matagali, that of the deceased and that of the .9999, are two nuclei to which all who wish to honor the deceased by their presence and contributions attach themselves. (This assumes that these are two different matagali. Since there is no rule of matagali exogamy this need not be the case.) Many will have more direct or salient ties to one or the other of these principal matagali, but there are always some who have roughly equivalent ties 79 to both. They may choose between the two or even contribute to the ritual obligations of both sets of principals. As a form of respect those gathered to witness the exchange of a 999191 sit or squat; only those who watch from a distance remain standing. When everyone is settled in their places, a senior man representing the donors comes forward and, kneeling beside the 999191. announces his intention to present it by calling out f119_9119 The recipients respond by repeating the phrase. In his offering speech the donors' representative recalls the reasons for giving the feast and refers to the relationship which binds donors and recipients. The message is clear; the 999191_is being offered to repay a debt or strengthen existing ties. While making offering speeches, men often fuss with the arrangement of things being presented. For example, in offering a 999191 the speaker may pick at one of the carcases removing bits of grass and brushing away flies. Iwipart this may be a cultur- ally patterned nervous gesture, but I think there is also some sig- nificance in the fact that the speaker is touching the thing to be given and thus associating himself with it. The structure of ritual prestation is similar for all forms of ritual exchange, and we will examine this structure more closely later on. For now, let me note that when the donors' representative has finished speaking, a representative of the recipients must make an acceptance speech. The word normally used to describe the making of such speeches is "touching? (9999). As we shall see later, when men are drinking 999999, it is considered polite for newcomers to present 9agona or cigarettes to those already present. Usually there 80 is no offering speech. The donor simply passes them to the man host- ing the gathering or to some senior man. The latter may, in fact, be at some distance from the gift when he makes his acceptance speech, but he extends his hands as if to touch the gift and says, fiflu .tagéutikg.ga_. . ." ("I am touching the . . . .") I was once present when a man related a bit of news to a senior kinsman in the formal manner. He clapped lightly and then announced, "Sir, I have a bit of news.“ It is normal on such occasions for the person receiving the news to make a formal acceptance speech (ulia) which he begins by saying flg,_gu ulia tiko na_;i tukutuku . . . ." ("Yes,I am comment- ing on the news. . . "). On this occasion, however, the senior kins- man extended his hands as if to touch the news and, in fact, said that this was what he was doing. "19, au tara tiko n§_j_tukutuku v3] turgga, ng_j_tukutuku gj_vanua," ("Yes, I am laying my hands on the chiefly news, on the news of the land."). This layingtwiof hands or touching symbolically transfers that which has been associated with the donor to the recipient. In the case of a 933131 though,there is a bit of a logistical problem. Since the Eggujj_usually consists of a number of large items, it is impossible for the recipients' representative to touch them all, and the mere extending of hands would not always be in keeping with the solemnity of the occasion. Then, too, the man representing the recipients is usually a senior man of chiefly rank. His dignity requires that he remain in place. The problem is often solved by younger men of the recipient group coming forward and placing their right hands on each item being received. 81 When the transaction is finished, a representative of the donors confers with the mean who will be responsible for the actual division. He gives the latter a list of the individuals and groups that are to receive a portion and some idea of the size of the portion to be assigned to each. For one of the last magiti I witnessed, one presented by the matagali of the decreased on the fourth day of a funeral, these decisions were made by one of the decreased's sons. On this occasion the magiti consisted of the carcases of four cattle. After the formal speeches of presentation and acceptance, one cow was taken to the house of the principal wgkg_to be divided by him among those who had supported him in the initial exchange. A second carcase, along with a whale's tooth (tabug) was presented by the sons of the deceased to men who had assisted them in various ways throughout the funeral. The remaining two carcases were carried across the lane to the yard. of the deceased's house. There they were cut into portions to be presented to those who had supported them through their contribu- tions. A younger son of the deceased cut the carcuses up with swift, deft strokes of a machete while his (the son's) mother's brother looked on, checking and rechecking the list of recipients. Portions were readjusted and recounted until both men were satisfied with the divi- sion. The wife's brother then read off the names of the recipients in turn. As he read off each name, his nephew, bending down, touched the portion with his right hand and called out the recipient's name in] oud voice. This publicly witnessed laying on of hands occurs whenever a magiti is divided. 82 While all of this was going on, as is usually the case, a large number of men sat nearby drinking yaggga, Shares designated for older men or men who were not present were carried off to the appropriate house by boys and youths. When their names were called, younger men went to claim their shares, clapping their hands in thanks before accepting it. They then left for home or, if from another village, for the house where they were staying, to divide up the meat they had been given among those houses they wished to share with. Through this process of division and redivision meat is dis- tributed throughout the village whenever a major magiti is presented. Occasionally collective meals are formally presented to those foregathered to share them, and in the offering speech the food may be referred to as a magiti. Stricly speaking, it is not but is, rather, an instance oquaga_vata_or "eating together." As in all human societies commensality is here an expression of mutual partici- pation and solidarity. This is quite consciously so during the holi- day season. As part of their observance of Christmas (§jga.nj_§ugu) and New Years (vakatawase) mataqali members eat their meals together for a period of one week to ten days. Commensality is an important feature of household membership, not just on ritual occasions, of course, but on a daily basis, and even adolescent sons, who often sleep elsewhere, eat at least one meal a day with their families. The public act of eating together (ga_kanavata) by bride and groom is one of the most important stages in the marriage ceremoney, itself the founding link in the web of kinship. Commensality also plays an 83 important part in mourning the dead, a process which, for the deceased, in one sense marks the dissolution of that chain. The fourth (bgfva), tenth (£11111), and hundreth night (.b_o_' draudrau) after the death are especially important, but the immediate kin of the deceased will also gather to drink yaggga and eat together on each tenth night until they are released from formal mourning on the hundreth night. Death does not immediately cut the decreased off from sharing food with the living. 0n the fourth night after the death, food is taken to the grave site and left for the spirit of the deceased. It must be food suitable for a feast. In the one case that I know of, it included boiled chicken, but I was told that beef tongue or heart would also be suitable. Throughout the four days of the funeral proper, specially selected mourners maintain a vigil in the house of the deceased (flaw). They are known as the muggy ("chiefs of the four nights") or turaga lekaleka ("short chiefs,§ i.e., in the duration of their reign) because they sit in the upper end of the house and take precedence in drinking yagona over all others present. 0n the fourth night, the same night on which food is carried to the grave, a meal for one person is laid out while the EEIEEQ. lekaleka sleep. The first man to awaken eats the food without waking the others. In the instance that I know of, the funeral of a senior man of chiefly rank, this food came from the same meal which provided the serving offered to the spirit of the deceased at the grave site. FO0TNOTES--CHAPTER III 1Brothers and sisters should avoid even the hint of sexual intimacy. After puberty brothers and sisters should not sleep under the same roof. 2Prior to entering a house or any other place where people are formally assembled, a man calls out "Eu - - 9," The first syllable is said sharply, and the second is drawn out. To this those already present respond with "Q_- - duo." Women, on the other hand, call out lQu_- -." The single syllaETe is drawn out and falling in pitch. After nightfall, both sexes call out "§a_bgfsaka" (sa bogi saka, "it is night sir") to which the proper response is "§a_§gf5"— 3Construction of the earth oven, preparation of the food for baking, and supervision of the cooking process are all men's work. Men also make fancy puddings on special occasions. While men might occasionally help out, the preparation of the boiled food that forms the day-to-day diet is women's work. 4These conventional patterns of generosity apply only to meal times. The offering of food in other settings is less a matter of form and more a matter of personal expression. Meals do not normally include fruit, but some fruits are highly valued as snacks. Most of the fruit that is eaten grows either in the gardens or the bush and is consumed outside of the village. I have seen people who were eating fruit approaCh a group and share the fruit with some of those present but not others. Sometimes people are given a share only after they ask. Sometimes, too, fruit is concealed in order to avoid demands which could not, with good grace, be refused. 5Men carry taro tied in bundles and slung over either end of a pole which is then balanced on the shoulder. Women carry taro and other produce in sacks which are tied with long strips of bark and carried on the back. Women never carry anything on their shoulders, and men rarely carry anything any other way. 6Men also fish, but the methods they employ are different from those of women. Men take fish by diving under water and impaling them on spears made of heavy wire sharpened at one end. These are propelled sling-shot fashion by a piece of heavy rubber. When we first arrived in the village, the Wainimala was muddy due to the on- going construction of a hydroelectric project on a tributary upriver, and men did very little fishing. But by January 1981 the water was 84 85 clear enough to permit diving. I once accompanied a man who spent an hour or so fishing on his way home from the garden. His catch included two eels and nineteen other fish. At one point he had located a nesting area and brought up a fish on every other dive or so. 7The literal meaning of turaga.ni koro is "village chief." Functions performed by the turaga gl_ oro during the colonial era included the planning and coordinating of communal work, matters which are handled today by the village chairman in consultation with the traditional chief. The duties of the contemporary turaga nj_koro amount to little more than relaying their decisions. CHAPTER IV "TRUE FOOD" So far we have considered food as a focal point of cultural values and as something which is shared and exchanged. In the process we have touched on the etiquette of commensality and, very briefly, on beliefs about the relationship between food and health. It remains to consider food from the standpoint of production. Discussion will be essentially limited to those traditional cultigens classified as "true food" (kakana dina), for they are not only the staple items of diet; they are also those foods whose symbolic value is richest. As we shall see their production is not simply a technical matter. It also involves ritual observances. The most important of these, the first fruits ceremony or_i.§e!u, not only promotes health, growth, and prosperity; it is also a restatement of the political charter upon which the village and chiefdom are founded. Of those traditional cultigens classified as kakana dina, taro (Colocasia esculenta) is the most important, both in terms of subsistence and as a source of cash income. Some taro (9951) is grown on hillside plots or in kitchen gardens near the village, but every household in the village also plants taro in the alluvial soil across the river. There the natural plant cover consists of a mixture of tall grasses and scrub with scattered stands of bamboo. 86 87 Until very recently, only a few gardens were planted in plowed soil. In most cases preparation for planting simply involves cutting down the grass. Burning is used only as a means of dispoisng of bamboo and other debris and not as a means of fertilizing the soil. The taro is planted among the grass stubble which, at first, dries out in the hot sun, but soon sprouts afresh and has to be cleared away several times during the first stages of the taro's growth. The soil is prepared for planting by removing the stubble and matted grass from a circular area 1; to 2 feet in diameter. A digging stick (kau) is used to loosen and turn over the soil. Usually no more than three strokes are required. On the final stroke the soil is turned over and all the clods are broken up. Then, with the stick projecting from the soil at about a 45° angle, it is raised rather forcefully to an upright position and the soil tamped down around it. The stick is then worked back and forth before being extracted from the SOil. This process leaves a hole about one foot deep into which the taro cutting or sucker is placed. It is important that the hole be deep enough because the taro corm will grow upwards, and if the planting is too shallow, there will not be enough room for it to develop properly. Taro is not seasonal and it is reproduced vegetatively either from cuttings (l!!!) or suckers (sulisuli),immature plants derived from the parent plant through vegetative reproduction. Accordingly the planting and harvesting of taro go on simultaneously throughout the year. It is harvested by loosening the soil with a digging stick, 88 care being taken not to cut the conn in the process. The corm is then extracted by pulling on the stem. The immature plants or suckers which cluster around the harvested plant are tamped back into the soil. The leaves are then trimmed off the harvested corm leaving two or three stems attached to it. These are used to tie the corms into bundles of three to six, depending on size. Men will carry two to six of these bundles balanced on a carrying stick born on the shoulder. In the process of harvesting a load of taro, a certain amount of debris accumulates in the form of leaves, stems, and rejected corms. This is never left scattered about the garden, but is always gathered up in one place and covered with grass specially cut for the purpose. When I first witnessed this, I inquired about its purpose and was told that it was done "because our grandfathers did it." The result of this practice is compost heaps sprinkled here and there over the garden. This compost is never used (i.e., spread over the garden area) so that any effects on the fertility of the soil are limited to the immediate.area around the heap. In any case, the practice is not thought of as a technical procedure, but as a ritual observance done out of respect for the taro and the ancestors who insure its fertility. Failure to observe the taboo will offend the ancestors (3g) and result in supernatural sanctions directed at the gardens. The failure of a newly constructed fence to keep cattle out of the gardens of one matagali was blamed on the failure ofits members to observe this taboo. The depredations of wild pigs directed at the gardens of another matagali were similarly explained. 89 Before taro is carried into the village, it is washed in the river and the rootlets stripped off. It may be stored in the cook- house before use, and therefore, this cleaning would seem to be moti- vated by practical considerations. But it is also done because it is the proper way to treat taro. Taro has sacred connotations which cassaVa (tavioka), for example, does not have. Taro was cultivated by the ancestors and can be given in ceremonial prestations. Tavioka is a new world cultigen introduced by Europeans and is not subject to any special treatment. When tavioka is harvested, the stems (tape) from which cuttings will be taken for replanting may be left scattered here and there about the garden. Unlike tavioka, taro can be stored for several days, though a household will seldom harvest much in advance of its needs. .As the taro corms are required, they will be cut away from the stems (i.e., the lower portion of the petiole or vgra), but in the cutting a portion of the corm is left on the vura_to provide food for growth when planted. Over the course of the week the cuttings or vu[a_begin to accumulate in the cookhouse. A man will usually plant these little by little, either expanding an existing garden or beginning a new one which is added to a bit at a time. Occasionally, a gardner may harvest as much as several hun- dred weight of taro at one time. When this happens, he has enough suckers to plant an entire garden. He will usually call upon men from other households to assist him in this. In one such instance, a man harvested about 100 weight of taro at the request (kerkere) of a 90 classificatory cross cousin. The cross cousin, a native of the village, lived and worked in an urban center on the other side of the island. He and his wife, also a native of the village, intended to take the taro with them when they returned to the western side of the island on the following day. As the taro was harvested, the suckers were gathered up, trimmed, placed in the shade, and covered with grass. Several days later the gardner, assisted by several classificatory cross cousins, members of his wife's natal matagali, cleared a new garden and planted these suckers. On this particular day the sun was very hot, the work group was comparatively small, and the pace of the work was accordingly slow, much of it being done later in the afternoon. When the work group is a large one and there are expec- tations of a lively yaggna session afterwards, men work with enthusi- asm. If the sun is hot, the work will proceed in spurts, and if there are enough men, they can work in shifts. Everyone knows what to do, and either no directions are given or so many are being given by so many different people that the effect is the same. Each man does as he sees fit, and the work proceeds smoothly. In the case discussed above, between 275 and 300 new taro were planted. In return for their labor, those who assisted were given a lunch cooked at the garden site by the owner's wife and one large bundle of taro each (valued at about F$4). Those who assist in planting a new garden may be compensated in other ways. On another occasion a man from a neighboring village requested the assistance of his cross cousins in matagali Nakorowaqa in planting a garden on 91 Nakorowaqa land. In return for their assistance and the privilege of planting on their land, he presented them with a magiti (one cow) and'yaggna. There is an expanding urban market for root crops in Fiji, especially for taro, and over the course of the last decade village households have become increasingly involved in production for sale. As noted earlier, this has been made possible by the extention of the road system into the interior of the island. This growing market orientation has probably effected taro production in several ways. In the recent past bananas were the main source of cash income, but heavy flooding destroyed the banana plantations and laid down deep deposits of sand._ No effort was made to restore banana production. Instead, villagers invested in cattle which have not added substan- tially to their cash income, but do make an important contribution to subsistence in the form of meat. The introduction of cattle has been correlated with a corresponding reduction in pig production. The cattle forage freely on scrub and unimproved grass and are given a minimal amount of care. Pigs, on the other hand, require constant attention and represent, in effect, another mouth to feed from the family gardens, a growing proportion of whose produce is now destined for sale. During our stay, the village livestock included only one domestic pig. I I have no data to support the assumption, but it is reasonable to suppose that the switch from bananas to taro as the principal source of cash income has required an increase in taro production. since gardens would now have to supply not only household subsistence 92 requirements plus the surplus necessary to meet traditional exchange obligations, but would also have to supply a salable surplus. This switch has also brought about a change in the relative amounts of the several varieties of taro grown. Today a variety called samga_is numerically preponderant in village gardens. It takes longer to mature and produces fewer suckers than some other varieties, but it has the highest market value. Given the need to strike a balance between subsistence needs and marketing activities, households may utilize the less salable taro for their daily needs and substitute alternative carbohydrate foods such as plantains, breadfruit, cassava, and wild yams. No one plants enough plantains to make them more than an occasional food, and while breadf'ruit' and wild yams are heavily utilized, both are seasonal. From the middle of August through the end of November, 1980, there _was an especially heavy dependence on wild yams in the area. Every Friday parties left for the forest (vgjkag) to dig wild yams, sometimes with horses to carry them back on. Wild yams almost entirely replaced taro as an item of diet. There seems to have been two reasons for this. Wild yams were especially plentiful that season, and large amounts of taro had been sold. Consequently, taro may have been in shorter supply than would have normally been the case. In early December bread fruit replaced wild yams as an alternative food source. On a year-round basis cassava is the most common alternative to taro in the diet, and most households plant some cassava. But there is considerable variation among households with respect to the frequency 93 with which it is eaten. In some households cassava is eaten almost daily, more frequently than taro, in fact, which is reserved for week- end meals.. Other households eat very little cassava. Among such fac- tors as taste preference, one of the things which seems to be related to the substitution of cassava for taro is the size of the household's work force relative to the total number of household members. Cassava does not require as much care as taro does, and it will grow on soil that would not normally be utilized for taro. Households with a number of young children and an adult work force limited to the married couple may find it desirable to substitute cassava more fre- quently than households with larger work forces and fewer dependents. But the substitution of cassava in the diet involves a certain degree of nutritional trade-off since its food value is less than that of taro. Also it should be kept in mind that while cassava, plantains, breadfruit, and wild yams can replace taro in meeting subsistence needs, they cannot be substituted for taro in ritual prestations. Domesticated yams are substitutable in the context of ceremonial exchange, but as we shall see in a moment, yams are comparatively scarce . The initial clearing of the land, as well as the planting and harvesting of taro, are men's work, but much of the intervening work of cultivation is performed by women. About three weeks after planting, grass and other competing vegetation will have to be cleared away from the developing plants, a task usually performed by women. A knife is used, and the debris is left to lie where it falls.1 The dirt around the stem is then loosened with the knife and scooped out by 94 hand to form a circular hole around the plant perhaps two inches deep and eight to ten inches in diameter. The dirt on the bottom of this depression is then tamped down. I was told that this process, known a$.££lih!252“flé.92529 aids in the formation of a well-rounded corm. It also forms a well which gathers rain water and nutrients. Following this procedure, there will be several successive weedings at variable intervals until the leafy structure of the plant is well established. During the later stages of growth, weeding ceases. At a glance a stand of mature taro may appear to be free of weeds. It ‘ towers over other plants and absorbs the sunlight it needs, but, in fact, before a plant is harvested, a tangle of grass may have to be cut away. At some point during the leaf-growth stage of the maturing taro, a second procedure known as bulubulutaki (“hurrying“) is per- formed. The circular well around the base of the plant is filled in and the earth is mounded around the stem of the plant. This covers the developing corm and also provides support for the stem and leaves (Cf. Firth 1965:67). The weeding of taro and the processes of keli.vaka_and bulubulutaki require patience and care. Though men may assist their wives in these tasks, they are thought of as being appropriate to women. Indeed, there is a belief that taro requires a woman's atten- tion to thrive. Men without wives may call upon the assistance of female kin; for example, a widower may ask a married daughter to weed and cultivate his taro. A man may formally request (kerekere) this 95 assistance or offer the woman some form of compensation. This is one way in which women can earn cash for their own use. In the last century Brewster noted the eXiStence-of an asso- ciation between women and taro in this area of Viti Levu. Yams, on the other hand, were associated with maleness (1922:93,163). While there are, as noted, certain processes in the cultivation of taro which are said to be women's work, the cultivation of yams is done by men. Aside from notions about the appropriateness of certain tasks for men or women, I encountered no overt statements about yams or taro as male and female symbols respectively. Still Brewster's comments are significant, for the dichotomy occurs elsewhere in the Pacific. Julia Hecht reports that in Pukapuka land is divided into two categories: 'te lei, or lowland swamps, and tgnwenga, higher land on periphery of the islets. Taro is cultivated by women in the damp soil of the interior; coconuts are tended by men on the drier land of the periphery. The dryness of the higher coconut "land" is also associated with males and the periphery in varying contexts; while the wetness of the lower "swamp" is associated with females and the center or interior in context. These linked associations of male, dry, up, outside and female, wet, down, inside are played out in Pukapukan social patterns and informal cul- tural concepts (Hecht 1977:186). This association of "wetcrops,9 such as taro, with the female principle and "dry crops,§ such as yams, with maleness also occurs in New Caledonia, but there shape-~phalliC'hithe case of yams, roughly triangular in the case of taro--is also used as a rationale for this association (Leenhardt 1937). The same set of oppositions--yams 96 cultivated by men on dry hillside plots, taro cultivated by women in irrigated fields--is reported for the Yami of Botel Tabago island off the southern tip of Fomosa (Barrau 1965:340-341). What is there about taro which suggests femininity to peoples who cultivate it? Perhaps more basic than the shape of the taro corm or the dampness in which it thrives is taro's remakable regenerative quality. Unlike yams which are seasonal and whose leaves and vines undergo a transformation from green lushness to withered dryness as they mature, taro can be planted anytime. As previously noted, plant- ing and harvesting are coterminus and continual. Taro can be grown from cuttings or from the suckers which cluster around the parent plant, and even discarded material may take root and sprout. In the Trobriand Islands the terminology associated with the growth and regen- eration of taro draws a clear parallel to human reproduction. The principle corm is calledana; a ("mother“), and the suckers are its latu-la or "children." The suckers which develop first are tuwa-la or ”older sibling" to those which come later. When the taro is har- vested, the rotting remains of the seed corm (i.e., that portion attached to the cutting from which the new plant is generated) may be found. It is likened to the afterbirth (bag) of the plant which developed from it (Malinowski 1935:105-6). The regenerative quality of taro may explain why the petiole of the plant (i.e, the bé, or upper petiole; it is the 1333, or lower petiole plus corm, which provides the cutting for planting) is thought to be an especially suitable food for women who have just given birth. 97 It should also be noted that childbirth is one of the class of events which are observed by a type of gathering known as bj§39§_(3auan . veisiko). Guests arrive bringing garden produce, gifts for the new- born child such as talcom powder and baby lotion, or household supplies such as kerosene, flour, or sugar. In return they are provided with 199991, tobaco, and hospitality. The word Tbisabé" seems to be a compound formed of the word bj§g_meaning Pto fall gently (of rain)," and Ea, the upper petiole of the taro plant. The literal meaning seems to be that the objects of the visit, in this case the mother and child, are "showered" with bi. Indeed one of the things which might be given at a Qléééé are bundles of raw taro. On these, as on all other occasions at which raw taro is given, it includes the lure. as well as the corm. It is, thus, not only something that can be cooked and eaten; it can also provide cuttings from which new taro plants can be grown. Like taro, domesticated yams are a traditional cultigen deemed suitable for ceremonial prestation. Indeed yams have even greater ritual importance than taro, for it is yams which are the focus of the first fruits ceremony (j sevu) on which the health and prosperity of the community depends. But despite their ritual significance, yams make up a comparatively small proportion of the total diet. Firth found a similar situation in Tikopia. He was puzzled at first by the importance which the Tikopia attributed to the ritual sequence known as the "Work of the Yams.f Though a valued food source, the yam is not the most important in the Tikopian diet. In terms of 98 the quantities consumed, taro and breadfruit are far more important. Firth considered the possibility of whether the greater difficulty and uncertainty involved in the cultivation of yams might not explain the ritual attention devoted to them. He notes, however, that “In Tikopia . . . which lacks the large yams of mainland New Guinea, no great technical attention was given to the yam," and it is Firth's impression that the yield was no more variable than that of bread- fruit (Firth 1967:164). If one assumes that there is some relation- ship between the degree of ritual attention paid to a food source and the contribution which it makes toward total subsistence, the ritual importance of the yam would appear to be annomalous in both Fiji and Tikopia. The Tikopia explain the ritual importance of the yam in terms of its association with the preeminent ancestral deity, the Atua i Kafika. Firth agrees that "there is an unexplained historical factor [i.e., this association with the Atua i Kafika] which has determined the situation" (1967:194). But he also notes that, for the Takiopia, the yam symbolizes all vegetable foodstuffs so that "It is the repre- sentative quality, not its bulk, that is being celebrated? (1967:16). He adds that a practical consideration, the yam's storability, might also be significant. Firth reports that the Tikopia sometimes worry that the yam crop will be insufficient for the ritual associated with the "Work of the Yams." There are fears that if the crop were too small, the Atua i Kafika would punish them for their meanness. 99 If the true yam crop should suffer disaster, as occasionally happens, then taumako, a species of yam with a prickly vine, regarded by the—Iikopia as a different food type, is called into service. In the last resort ulaka [Alocasia sp.] or even taro would be used, but this wou d only be done in extreme circumstances and might not be acceptable to the deity (1967:153). Their contribution to the Tikopian diet aside, it would almost appear that, rather than the rite being celebrated to promote the growth of yams, yams are grown in order to hold the rite. I think we have here a clue as to why the yam is of preeminent ritual importance in Fiji as well as Tik0pia. The Tikopian Work of the Yams" is a ritual cycle with individual rites associated with the preparation of seed yams, planting, and harvesting, but its overall purpose seems to be to secure favorable weather and a bountiful food supply (Firth 1967: 159-160). The Fijian-1'sevu_has a similar purpose. What is it about the yam which makes it an appropriate offering to the gods who can extend or withold prosperity? This is one issue to be addressed in this section, but in order to do so we will first have to consider the pragmatics of yam cultivation. Unlike taro, yams are seasonal. They are planted during the "dry" season,'hithe months from July through September, and take about seven months to mature so that their growth period spans the change of seasons. They are first harvested for the j sevu or first fruits ceremony in February, but the remainder of the crop may be left in the ground through March and April. There is considerable variation among households with respect to the number of yams planted. In a sample of 24 households, the 100 number of yams planted in 1980 ranged from 10 mounds to more than 500 with seven households not planting any yams at all that year. Of these latter, three were households headed by men under 30 years of age. Along with such factors as food preferences, the number of yams planted reflects the availability of seed yams and.the ability to muster and reciprocate a work force. More frequently than not, yams are planted on the alluvial flats across the river in close proximity to taro and other crops. In July 1980 I assisted in the preparation and planting of such a garden. It measured approximately 40 by 50 feet, and when it was ready for planting it contained 430 mounds. The first step in the preparation of the plot was the clearing away of all vegetation. Knives were used, and the dead vegetation was swept off by hand. Piles of rubbish were burned off to one side, and in the process the clearing was extended a little, but this was a by-product of the burning, not its primary purpose. The result of this meticulous clearing was a plot so clean that it resembled an earthen floor.. I was told that this care was necessary because if there were any competing vegetation, the yams would not produce luyega_(i.e., tubers, lit. "children") and would wither and die. ‘While several men were engaged in the process of clearing, two others were busy digging up the soil and turning it over, leaving circular patches perhaps 21 feet in diameter and a little more than one foot apart. Some of the mean who had been clearing the plot now began to till these patches of soil by hand. They dug their hands into 101 the soil, churned it up, and broke up any clods left by the men with digging forks. The result was fine-grained, neatly formed mounds regularly spaced throughout the garden. In all, eight men worked on the preparation of that garden. An additional man remained in the village preparing seed yams while another occupied himself with other garden tasks including the harvest- ing of a yaggna (kava) plant whose root was grated and served to the work party later in the day. The work party arrived at the garden site at about 9:45 a.m. and returned to the village at 1:30 p.m. so that approximately 30 man hours were required in the initial prepara- tion of the garden. Of the seven men who assisted the owner, one was an actual cross cousin (gaggle) while two others were classificatory cross cousins, and three others were classificatory mother's brothers (mgmg). One was a classificatory younger brother (335133). The work force was drawn from five different matagali, and no one from the owner's matagali participated. In return for their assistance, the work force were served an afternoon meal which included stewed chicken and tinned beef. This was followed by the drinking of 139932, also supplied by the garden owner. Despite the fact that the seed yams were prepared that morning the garden was not planted that day or the next. Most of the male work force of the village was engaged in the quarrying of sand from a gravel bar in the river. This was being used in mixing concrete to form into blocks for the foundation of a new district school. It was not until two days later that the garden was planted. The work force consisted 102 of the owner, his wife, his cross cousin, his sister's daughter, and me. The two women assisted in carrying the seed yams, but when we arrived at the garden site, they busied themselves with weeding taro and left the yam planting to the men. The first step in this process was to place the pieces of seed yams, or sets, on top of the mounds, all of them oriented in the same direction. At this point a few comments about the preparation of the seed yams are necessary. Two major varieties of yams (Dioscorea alata) were planted in this garden, a multilobed, vaguely handshaped variety called beka ("bat") and taniela, an elongated, sweet potato-shaped variety with two subvarieties, one white-fleshed (taniela vulavula), the other with a purplish flesh (taniela damudamu). In addition to these, a few kawai (Dioscorea esculenta) were also planted. The man preparing the sets selected the large yams and cut them into several pieces, usually three or four, which he then cut in half longitudinally, in the process removing some of the inner flesh from each piece. This went into a pot to be cooked for the meal to be served to the work party. On each piece destined for planting, he cut a notch marking the end nearest the uluga_or "head" of the yam, the proximal end of the tuber where it attaches to the stem. On the day that the garden was planted, the sets were placed on top of the mounds with these notches all pointing in the same direction. The planting procedure was simple. Both hands were used to dig a hole about six to eight inches deep into which the seed yam was placed with the notched hark; consistently on one side of the mound 103 and slightly lower than the other end of the set. The soil was then crumbled through the fingers to cover the set and rebuild the mound. The whole process took perhaps forty-five minutes to complete. An additional hour was spent expanding the cleared area, a process which was continued on the following day. (When the garden was completed, it contained 580 mounds and was the largest in the village that year.) Another half-hour was spent transplanting cabbage spouts among the yam mounds. There was some concern as to whether the sprouts would survive in the hot sun, and in their speches dedicating the yaggga which followed the planting, the owner and his cross cousin asked for rain so that the cabbage and yams would flourish. It rained off and on throughout the next day, the first good rain for some time.. Sometime during the following week, the owner constructed a sturdy structure of bamboo to support the yam vines when they appeared. It resembled a house frame, as long and wide as the garden with regularly spaced bamboo cross bars running from side to side. When the vines appeared, they were entwined around long-stemmed reeds (gasau), eventually climbing to the bamboo frame itself. The support- ing structure built for this garden, in keeping with the garden's size, was more elaborate than is usually required. Sometimes yams are planted in a circular pattern at the foot of a tree. Bamboo poles are simply leaned against the tree, and the yam vines entwine around them. A household's gardens may include several of these wigwam-like structures. In the case of the yam garden described above, the only addi- tional work involved was an occasional weeding with a hoe among the 104 mounds. This was done two or three times during the first three months of growth. (By this time all of the cabbage had been harvested, the yam vines were well established, and it became difficult to move among the mounds. The domestic yam, with its seasonal planting and harvesting, is the focal point of the traditional annual cycle. Indeed in con- temporary Fijian usage the word for the yam crop, yabakj, also means “year." A full consideration of the yam's ritual significance requires that we also consider the relationship between the living and their ancestral deities, between chief and "the people of the land,' between culture and nature, and between men and women. For these relation- ships are all given ritual expression in the first fruit ceremony, or j_§eyu, that marks the ripening of the yam crop. The warmth and rains of December and January stimulate the growth of vegetation, yams included, and by early February they have matured and are ready to be dug. And yet it is imperative that no one harvest yams for their own use until the j_taukej_nj_gele, "the owners of the soil," have been presented with the first fuits. Thus the distinction between the j tagkei_and the members of other matagali is given expression through the giving and receiving of yams. In most contexts, matagali Nakorowaqa-1. and Nabubuco and treated as the joint "owners" of the village lands. During my first weeks in the village, I was told that up until the recent past (i.e., the generation of the fathers of senior men) these two matagali had been a single social unit. They had separated in the 19205 as a consequence 105 of an assault on a man of one of them committed by a man of the other. The victim suffered serious injuries, but recovered, lived several more years, and fathered children. However, his eventual death was blamed on an infection of the blood believed uihave been caused by the attack. Despite the fact that his wounds headed on the surface, the effects of the injury remained within. It is not clear whether the two groups separated at the time of the attack or later. There was apparently one unsuccessful attempt to revenge the injury, but follow- ing the separation, amicable relations were restored. While it does not prevent cooperation, friendly interaction, and intermarriage (including that of the assailant's son's son and the sister of the victim's namesake), there is a notion that there is an imbalance in their relations, an unanswered injury. As a result of their prior association, the two matagali recognize joint obligations to the descendants of women who married out of the group before the separation occurred. Thus upon hearing of the death of an old man in a village upriver, the son of a Nakorowaqa woman, the two matagali met together as the weka (mother's patrikin) of the deceased and pooled their gifts to be presented at the funeral. Incidently, this case is important for understanding the nature of matagali, a matter which we will consider in more detail later. When people told me that Nakorowaqa and Nabubuco had been a single matagali, I assumed at first that their unity had been based on putative descent from a common ancestor. But unlike matagali that are linked by patrilineal descent, the vegetable and fish totems of 106 matagali Nakorowaqa and Nabubuco are different. They also have dif- ferent 1g or kalgg_!§ (ancestral deities). As I was to learn later, there was never any question of a putative common ancestry linking the two groups. People refer to Nakorowaqa and Nabuluco as having been one matagali in the past because the two groups 'stayed together" and functioned as a single unit in exchange relations. Mataqali Nabubuco's origins e 2519 ("upriver.") in the district of Nabubuco are remembered. It is matagali Nakorowaqa who are the true j_taukei, and yet the two matagali sit together as recipients of the first fruits. They are linked by a kind of siblingship and a consequent structural equivalence. In all the villages of the district, there are one or more matagali who receive first fruits by virtue of their prior claim to the land. Upriver in the village of Saumakiya it is matagali " Navitilevu and matagali Vutia who receive first fruits. There are two men of matagali Navitilevu resident in Nairukuruku, one of whom is the sister's son of a senior man of matagali Nabubuco. In certain contexts, their membership in their natal matagali is salient, but for most purposes, including the 1 sevg, they function as members of matagali Nabubuco. Like many Fijian rituals, the first fruit ceremony lasts four days; for the elders of the 1“592522.1t amounts to a four-day vigil. Within the j_§evg_there are three identifiable stages. The first stage, na i sevu ni kalou, or "i seveu of the ancestral spirit,f involves the eating of yams by the mean of those matagali designated 107 as j_taukei. The second stage, na i sevu ni vanua, or the "j_sevu of the land," involves the presentation of yams to the chief. The koile, a type of wild yam, to the men gathered in the place of vigil. In the first j_sevu that I witnessed (1980) these stages fell on the first, second, and third days respectively. The j_sevu.ni koile would normally have occurred on the fourth day, but in that year the fourth day fell on a Sunday, and the Sabbath is strictly observed. In our second i sevu the order of events was the same, but the spacing was different. In 1981 the i sevu _n_i_ kalou and the i sevu ni vanua both occurred on the first day, in the afternoon and evening respec- tively. In our second year in the village the j_§evg was not held until the first week in March, almost two weeks later than it had been the year before. As February drew to a close, I began to wonder about the timing of the event and asked a Nabubuco man when it would be held. He replied that the date was a matter to be decided by the seniormost man of matagali Nakorowaqa "because they (Nakorowaqa) are the b_e_t_e_ (priests)" with respect to the 1 sexy, I asked how the elder would determine the date and was told that the vi (ancestral spirit) would visit him in his sleep and indicate the proper time. Later I asked the elder's son about the timing of the event and was told that indeed it had been his father's decision. The son made no mention of communi- cation with ancestral spirits, but said that his father knew that the 1 sevu should beheld in February or March before the j_ sevu i lotu 108 or “i_§evu_of the church." Each year the Wesleyan Church of Fiji sets aside one Sunday for the presentation of first fruits to the Christian God. People bring in the largest taro, yams, sugar cane, and tavioka that their gardens have produced. These are placed in the church on the Saturday afternoon before the service and are later distributed among the sick and elderly. The traditional j_§evg_should already have been held before this. Thus, just as the first fruits of the yam crop are eaten by the j_tagkej_before they are offered to the chief, so are they offered to the ancestral spirits before they are offered to the Christian God. On the day the j sevu_is to begin, the men of the matagali j_tagke1 gather in the house of an elder of either matagali. In 1980 the j_§egg_was held in a Nabubuco house and in 1981 in a Nakorowaqa house. For the elders this is the beginning of their vigil, and younger men are present to assist them in this. As with all Fijian rituals, yaggna is a central feature of the proceedings. If not only provides refreshment, but also serves as a medium for communication with the spirits, a point we will discuss in more detail later. Young men prepare and serve the yaggna, and middle- aged men join their elder kinsmen around the kava circle. The atmos- phere is informal. Men smoke and talk or listen to the radio. There is nothing to indicate the importance of the event. By this time yams will have already been dug, and around noon on the first day women arrive with plates of cooked yams and other food. The meal is prepared and the yams presented by women who have 109 either married into or were born into the matagali designated “owners of the soil." A cloth is laid and places set, but the yams remain on platters at the lower end of the cloth. At the first.i sevu I witnessed.when word was given that the meal was ready, a middle-aged man came forward and turning off the radio, he hushed the child of his elder brother's son. Kneeling beside the yagona bowl, he addressed the elders as follows. Chiefly indeed are Nabubuco. Chiefly also Nakorowaqa. This is a bowl of a ona for you, the chiefly brothers. There is word that our preparations await outside [i.e., the relish had not yet been brought inside]. For the sake of our [inclusive] j_sevu today pardon us [exclusive] for laying hands upon the yam gardens. Let there be no problems should we pluck them. I ask that our kin be healthy. Be of good soul you two and make our [inclusive] yam crop fruitful. And let the harvests be big for all the times that follow. The zagona is being offered in a chiefly manner to Delana; it is being offered in a chiefly manner to Nabu- buco. That is the length of my speech. Chiefly is Nabubuco, chiefly also Nakorowaqa and all the chiefly zavu $6 . The opening and closing lines of this speech follow a formula applicable to any formal presentation. While delivering the salutation the speaker pauses slightly after each phrase honoring the matagali present. As he delivers the closing lines, his speech accelerates until it reaches a crescendo at the very end. He then claps respect- fully. The benedictions which make up the substance of the speech are common to ritual transactions, but in this case they are linked to the harvesting of the yam crop. The speaker is addressing the two elders and through them, and in them, the ancestral deities of the 110 two matagali. For though I was given no direct statement on the matter, it is reasonable to conclude that, in the context of the j sexg, the elders stand in place of their ancestral deities. This is the meaning of the i _s_efl £1 k_a_l_9_u_, the "1M of the ancestral deity;“ in offering the ygggga_and the accompanying feast to the elders, the spokesman is also offering it to the kalgg_!§_of the i taukei. One old man said of the j_§evg_as a whole "Keimami vakoyacora tiko na_neimami vakavinavinaka vu'a na_kalou nj_dela nj.yavu kei Na_ Kalou mai Lomalagi." (We are fulfilling our thanksgiving to the god of our ancestors and to God in heaven.) This embodiment in ritual context of the kalou v§_in the persons of living elders was an important element of the traditional belief system. In 1884 the Governor's Commissioner to Colo East, the province to which the district of Matailobau then belonged, wrote the following in a monthly report to the Colonial Secretary: There has been some little sickness amongst the people [of the village of Rokorokoyawa in the diStrict.of Wainimala],.. nothing at all serious, and some half dozen or so of the old men presented as offerings to their ancestral spirits their houses and contents. They made these presentations to their elder living relatives avering that the ancestral god dwelt in them, or that they were the present living incarnation of the family deity (991, Sec, 9:, 87-734). In the traditional religious system gods were communicated with (e.g., through sacrifice) yja_their living representative, be he chief, priest, or elder. It is significant that the role of the j_taukei in the.i sevu is likened to that of a bete or priest, for traditionally priests were possessed by the gods that they served. I have spoken of the.i taukei as receiving first fruits, for this was how their role 111 in the j segg was described to me. But rather than looking on them as the recipients of the j_§evu, it may be more correct to think of them as transmitters of the offering--firstly to their ancestral deity, and secondly to the chief. In this their role appears to be similar to that of the Moalan village priest described by Sahlins (1962:343). I have mentioned the fact that no one must dig yams for their own use before the j_§eyg, During the planting season of our second year in the village, I planted a few tubers in a small garden I had made just outside the village. Seven or eight months later the vines were dry, and the yams were ready to harvest. Apparently I had planted two species of yam, kawai: (Dioscorea esculenta) and 292.9122. or "true yam" (Dioscorea alata). Somehow over the intervening months I had forgotten that I planted the latter. Consequently, when the tUbers were ready to be dug, I asked several people if it would be alright to harvest kawai before the j_§evu, They all agreed that it would be. As I was washing off a spade that I had borrowed, a woman came by and asked what I had in the sack that I brought back from the garden. I told her that I had dug up some kawai. She opened the sack to look at them and found xgahgjna, Several other people looked at them also. They were not angry; what I had done was regarded as a mistake, albeit a foolish one. But they were concerned. As the day passed, the consensus seemed to be that we should not cut up and cook the yams but, rather, that they should be returned to the earth. I did so. Not only is it tabu for persons of other matagali to dig yams for their own use before the i sevu, but it is also tabu for the 112 "owners of the soil" to do so. On two occasions I was told of a man of the i_taukei (on one occasion it was his son who related the incident to me) who had given in to his daughter's whim to eat yams before the j_§evg, That night he saw 3g, the spirits of his ancestors, in a dream. They tried him for his offense, and within three or four days he was dead. Returning now to the events of the first day of the j_§§!g, following the speeches of presentation and acceptance, yaggna_is served in the "chiefly manner" (vakaturaga). The meal is then served, the men eating in order of seniority. They rest for a bit after the heal and then resume drinking throughtout the afternoon and evening. When the elders become sleepy, they retire into the upper part of the house. 1 The second day is much like the first. Men of the matagali j_tagkej gather after breakfast and begin drinking yaggna, Men of other matagali may drop by, honoring their hosts with a presentation of yaggga, At noon women of the matagali j_tagkei bring a simple meal for the two elders who are keeping the vigil. In the evening of that second day the center of events shifts to another house, one large enough to accommodate all of the senior men of the village. This is the j_§egg_nj.yanga, the presentation of yams to the chief. In 1981 it was held in the house of a Nabubuco man, the son-in-law of the paramount and brother-in-law of the middle- aged son of the Nakorowaqa elder. When I entered the house, I found most of the senior men of the village seated in a semi-circle at the 113 upper end of the house. The chief was seated at the head of the circle and the Wesleyan minister (i talatala) at his right hand. An eating cloth was stretched down the length of the house with places already set. Women of all matagali were gathered at the lower end of the house preparing to serve the meal. Plates of boiled yams were already arranged at the lower end of the cloth. Not everyone was in attendance. Most of the younger men remained in the Xélé.fli.§£!!. drinking‘yaggna, When everyone was settled in the second house, younger men of matagali Nabubuco removed a section of the dining cloth and began to mix (lose) yaqona in a large wooden bowl (tanoa). When all was pre- pared, an elder of matagali Nabubuco came forward and with downcast eyes addressed those assembled before him, but especially the chief, as follows: In the chiefly manner Nabena Delanavulele Ethe chief's linea e], the chiefly child, the Onwer of the Flood the chief's title]. In the chiefly manner the Top of Taulevu [a reference to another lineage within the chief's matagalil. In the chiefly manner to him, the head of the diocese of Matailobau-Wainimala. [He goes on to acknowledge the representatives of other matagali present.] A small bowl of a ona, a small feast, lie before you [plural] today. Before offering the feast to you [plural], you are seeing this--it is your [plural] feast only. I am the mouth of the young that I may offer to you, the chief, the.1 sevu on behalf of our [the inclusive form of the first person plural was used] staying on the land or soil. I am praising in a respectful and believing manner you, the chief, that you will accept it with a kind heart. Let there be no stain on them, the young, likewise on myself who offers the feast. I am entrusting [to you] our [inclusive] staying on the land. That is the length of my speech, etc. 114 When he had finished, he clapped lightly and took a place at the lower end of the kava circle. At that point the chief's father's brother's son, a man whom the chief calls “younger sibling," came for- ward, and in a quavering voice, a form of speech indicating solemnity and respect, he accepted the feast on behalf of those assembled. I am touching your feast and yagona. [He stretched out his hands as if to touch them.] Let it be a feast of life, a yagona of our staying [here]. Health and life to you sib- lings and children. Love us [the inclusive form of the pronoun was used] along with God. Let our staying together continue. Grant us strength. Let your days be prosperous. May your kin be blessed. Let your days continue. Let the earth bear fruit. His speech ended with a closing formula usually incorporated in acceptance speeches irrespective of the nature of the transaction: "Nana! E . . . dina!" In this context, "mana" is an evocation of efficiacy, or perhaps an avowal of the potency of the utterance. "g dina" may be translated as "indeed" or "it is true." The speeches of dedication and acceptance were followed by a formal or "chiefly yagona." Young men of the matagali j_taukei manned the tance and served as cup bearers. When it is served in a formal manner,men drink yagona in pairs. The first cup of a pair usually goes to a man of chiefly rank, the highest ranking member of his group present at the time. I use the generalized term "group" because the relevant unit may be a lineage, a matagali, a village, a chiefdom, the government of Fiji, or the Wesleyan Church. In the context of the yagona the unit is treated as a distinct sacradotal entity and its representative honored as a chiefly personage. For each man who drinks 115 as a chief there must be someone to drink a second cup (rage). The criteria for selecting this second are various and contextual, and it is a matter which need not concern us here. In this instance the first cup went to the chief, the para- mount of the vanua_(land, chiefdom) of Nabena, the highest ranking person present. Normally a senior man of matagali Nawaita, the bati or warriors to the chief, would rage for him, but no senior men of that group were present as yet. On other occasions a senior man of matagali Nakorowaqa or Nabubuco, heralds or matanivanua to the chief, might have drunk the second cup. But on this occasion it was they who were presenting the 129929, Instead, a senior man of the chief's own matagali but of a different lineage drank the gage, The next cup was served to the Wesleyan minister as the highest ranking representa- tive of the Church, and his helper, the catechist (vakatawa), drank the rage for him. The third chiefly cup went to a senior man of the chief's own matagali, but of another lineage, the chiefly line of Taulevu village. An untitled man of the same matagali, whose origins were also in Taulevu, drank as his second. The fourth and final chiefly cup went to a man of matagali Navitilevu, the chiefly group in the village of Saumakiya. As already noted Navitilevu have a traditional association with the j_t§gkej_of Nairukuruku, and in their own village it is they who receive the lugevu, A senior man of matagali Nakoro- waqa clapped his hands in anticipation of receiving the second cup, but he was preempted by the paramount's father's brother's son, the same man who had delivered the acceptance speech. Before drinking, he 116 lifted his cup, and turning to two classifcatory brothers who had just had a loud and heated disagreement, he said "dou bula veitacini“ (to your health brothers). At this point the original cup bearer retired and the formal ygggna_was brought to a close. Several more young men than took the cupbearer's place and began to serve the rest of the men present. Following a single round of m, the men ate in order of seniority. The meal consisted of boiled yams, boiled eel, and other fish and was served by women and girls of all of the village's matagali. When the men had finished, the women ate. As with most rituals the j_§evu_nj_ganga_has several levels of meaning. 0n the one had, it is clearly a statement about the nature of the political bonds between the chief and the j tagkei_with the latter functioning as representatives of the community as a whole. I say “political" though I might just as well say "ritual," for the two are inseparable. It is not as the chief's heralds (mata_nj,yanua) but as the owners of the soil and priests of the j_§eyu_that Nakorowaqa and Nabubuco. present the yams to the chief. They do this only after having first offered them to their ancestral deity and eating them in communion with that deity. It is made clear that they are offering the first fruits to the chief in recognition of his rule and in exchange for, and as a token of gratitude for being allowed to stay on the land. There would appear to be irony in this, for the ancestors of the i tagkei were on the land before the chiefly matagali arrived. The irony is only apparent, however. Through the act of investiture 117 (veibuli), the jutagkei surrendered sovereignty (lewé) to the chief's ancestor, but that did not sever their ritual connection to the land. On the contrary, it confirmed it and gave it new meaning in their ritual obligations to the chief. Still it should be pointed out that in the speech made in offering the first fruits, the speaker refers to "our? staying on the land using the inclusive form of the first person plural pronoun. 'That is, the i sevu is offered not only on behalf of the speaker and the community he speaks for, but also on behalf of the chief he addresses. It is offered on behalf of their (inclusive) "staying on the land." On another level the 1 sevu nj_vanua, like the 1 sevu hi kalou and the first fruits ceremony in its entirely, is a ritual of increase, an appeal for health, well-being, and prosperity. It exemplifies Hocart's dictum that “Ritual, as long as it retains its meaning, is a cooperation for life" (Hocart 1936:37). In the 1 taukei's offering, and in the chief's acceptance of that offering, a positive action is being taken to secure the prosperity (sautfl) of the community. As the recipient of the first fruits, the chief is god-like. Like other elders, in the context of the ritual he is identified with an acestral deity. Though a direct statement of that identification may not be forthcoming, the belief that the well-being of the community depends on this chiefly ritual is very much alive. Never in living memory has the j sevu failed to be held. 0n the surface the i sevu nj_vanua is a communal celebration of the passing of the old year and the arrival of the new, for such 118 it was in the traditional annual cycle and such it remains. As a celebration of transition, it performs the functions that such cele- brations do generally. It both requires and reinforces cooperation and communal identity. Quarreling is definitely out of place. (Recall the admonition to the quarreling classicatory siblings.) In this it differs from the celebration of the calendrical New Year, January 1, since the latter often involves the use of alcohol which may precipi- tate fights among young men. 0n the third day of the.i.§evu men gather in the appointed house to drink yaggna, There is nothing to distinguish the affair from a hundred other occasions, except for the fact that the elders of the matagali jhtagke1_continue their Vigil. Though they generally speak very little, they are the focal point of the gathering, sitting in the upper end of the house. Meanwhile women of all matagali who wish to participate set off in small groups for the bush (veikag) to gather a particular type of wild yam. This yam (Dioscorea bulbifera, called koile in the local dialect, ripens in February-March and can be considered a biological marker of the season. The variety of koile the women are looking for is very bitter and can only be eaten after the causative agent is leached out of it. ‘kgile grows at the base of trees, its Vines climb- ing up the trunk into the sunlight. After the vine has been pr0perly identified, the soil is loosened with a knife, though a sharpened stick is used for the actual digging. The flesh of the koile may be tinged with purple, blue, or yellow, or it may be whitish in color. When it is brought into the 119 village it is washed and prepared for cooking, the first step in the processing necessary to render it edible. The actual cooking (i.e., boiling) may take place on the afternoon of the third day or the morn- ing of the fourth. The next step in the process, the leaching, is more time consuming. Before the village had piped water, this was done in small streams in the veikangith sections of bamboo serving as pipes to direct the flow of water over the 59113, A special apparatus is constructed for the leaching process. The main component is a large basket. This is laid on a square of clean. cloth (e.g, a slit flour sack) which is, in turn, laid upon a thick layer of broad leaves from a plant known as ga§§u_nj_g§ ("duck reed,' gang; ingjga). The whole structure is supported by a frame- work of “logs" from a banana or plantain plant. The cooked koile is grated into the basket which is continually washed by running water. The sediment is filtered out of the cloudy water and collected on the cloth. When the leaching process is finished, the cloth is folded into a packet, tied with vines, and suspended to allow the excess water to drip off. The main event of the fourth and final day of the j sevg_is the presentation of the |_<9_i_le. Men gather in the yilenigfl around mid-morning and begin drinking yaggna, In the early afternoon women arrive with platters 0f.52112 and other food. In its prepared form kgjle_resembles gruel in texture and consistency and may be oatmeal- colored, purplish, or light yellow. Its taste can only be described as nonexistent; the leaching has not only removed its bitterness, but 120 any other flavor it might have had. A few platters of kgile_may be prepared with the legume-like‘j_§jbj, the large disc-like fruit of the _w_5_ _la_i_ (Entada phaseoloides). This adds a pleasant tang to the other- wise bland k_oil_e_. Despite its lack of flavor, people eat the kgile in large quan- tities and with a good deal of enjoyment. This is normally the only timeof the year in which it is eaten, and it carries with it the connotations of the season. Moreover, since it represents the first fruits of the forest and has been offered to the vfi in the persons of the elders, it has a sacramental quality though there is nothing solemn about the occasion. In contrast to the presentation of the 1591119151123, the iseflnijgojie is a much more relaxed and light- hearted affair due, I suspect, to the central role played by women as much as anything else. In 1981 the magiti presented by the women included 18 large platters of 59113, a couple plates of yams, a plate of greens, and one large tin of fish. (The fish, greens, and yams were provided for the elders of the matagali j_t§gke1_who had kept the four-day vigil.) The men present included most of the j_tagkei, the chief, the chief's younger brother and his parallel cousin (classificatory younger brother), the head of another segment of matagali Siko, senior men of the other matagali in the village, and the Wesleyan minister. The magiti_was presented to the chief on the women's behalf by an elder of matagali Nabubuco and accepted for the chief by his younger brother. The presentation was followed by a yaggna vakaturaga or 'chiefly yaqona." When this was completed, a bucket of yagona was brought in 121 and presented to the women as a vakanakanaka (thanksgiving) by an elder of matagali Nabubuco. The yaggna was accepted on the women's behalf by a chiefly lady. At this point the platters of koile_were distributed among those present and several were sent out to the cook house (vale ginkggg) where the younger men had gone in order to make room for the women. This was followed by a huge tub of breadfruitf pudding presented by the j_talatala and his family.. People ate and joked, and there was an atmosphere of relaxation as if after the accomplishment of an important piece of work as was indeed the case. Though there is a basic similarity between them the 1.§£!!”fll kgiLe and the _i_s_elu__n_i_ka_loy_ differ in some important ways. In the 1 sex! n1 kalgg_yams are presented not only to the elders of the matagali j tauke1_but also to the kalgg_y§_of those matagali. The xgaudina_or "true yamf is planted in gardens which are transitional between the domestic sphere of the village and the veikag or bush which encompasses them. Though women weed them from time to time, it is men who plant yams. But they are also cared for and nourished by the kalggflgfi associated with the vanga_on which they are planted. Just as men and their ancestral deity share in the care and nurture of yams, so do they share the first fruits of their efforts. ‘Egile contrasts with 133 dina in several ways. It is not domesticated, it is inedible in its "natural" state (cooking alone does not render it palatable), and it is women who collect it for the 13.5913. To say thatitis "of the M" (i.e., uncultivated) is to say that it is planted, not by men, but by Na Kalou, the Christian 122 God or the kalgg_y§_associated with the land on which it grows. Attribution depends upon the religious orthodoxy of one's informant which may have a situational component.. But in any case, it is not to the Christian God that the j_§eyg_nj_kgile_is offered, but to the elders and their chief and through them and in them, their ancestral deities. Once the j_sevu nkaalou and j_sevu nj.vanua have been offered, it is permissible to dig yams for one's own use. Similarly, once the j_§egg_njukgile has been offered, it is permissible to dig wild yams (v_ua_ n_i M), though, in fact, it is not until August-September when the leaves and vines of the wild yams have dried up, that they are dug in any quantity. By this time many people will have planted their yam gardens for the following year, and domestic yams will be scarce. To summarize, then, the j §e1g_is offered to secure the health and well being of the community and the fertility of the forest as well as their gardens. The question remains, however: What is it about yams which make them the focus of these concerns? Why, for instance, are they and not taro considered appropriate offerings to the spirit beings who can extend or withhold prosperity? The most obvious answer is that yams are seasonal and are, therefore, suitable symbols for the ebb and flow of life and an understandable focus for concern about the struggle between life forces and death. Hocart saw this concern as theiprimary issue of Fijian religion and correctly 50, I think. But yams are a suitable offering in another way, and to understand why,it is useful to draw upon the Tik0pian material. 123 In Tikopia the yam is not only associated with the principal deity, the Atua i Kafika, it is also considered to be a representa- tion of his body. In the most dramatic segment of the Work of the Yams, men compete to be the first to swallow a mouthful of scalding hot yam. In so doing, they are eating the body of the Atua in Kafika and are doing so, moreover, in the deity's presence, for in the con- text of the ritual the Atua i Kafika is embodied in the person of the chief of the clan, the Ariki i Kafika. It is not surprising that the Tikopians have extended the name for this rite, teflkai tapu, or Pthe sacred food" to the Christian sacrament of Communion (Firth 1967:156). In the Fijian j.sevu the community present first fruits of the yam crop to anancestral deity incarnate in the person of a chief or elder, and after making the presentation, they eat the yams. Though here there is no identification between the flesh of the yam and the flesh of the deity, this too is a form of communion. In both Tikopia and Fiji, then, the cult of the yam focuses on ancestral deities, and those deities are approached through the act of communion. The yam is an appropriate vehicle for that commun- ion because it represents both continuity and the sequence of genera- tions. Maurice Leenhardt writes that in New Caledonia Of couse the yam grows a new each season, but it is considered to be one inits continuity. . . . In times of mourning, the bereaved are offered a yam and told, "you see, it has a tip; it has a head." And the bereaved know that whereas the tip dies in the soil the head sends out new shoots from which new yams are born. The yam is offered as a symbol of life perpetuating itself, and this image of life springing forth again is consolation for their sorrow. In short, the tuber, which is buried as a plant when the fields are sown, dis- appears to be replaced by new tubers. It is the image of the 124 dead man. Thus a husband returned to his widow in a dream; “I am an old yam,“ he said (1947:62-63). Combined with the yam's male connotations discussed earlier, [these qualities of continuity and generational succession makes it an appropriate symbol of the "long line of men" which stretches back to the founding ancestors. It may be suggested that in offering yams to their ancestral spirit, the men of the matagali j_tagkei_are offering something which is a suitable representation of themselves. But they do so on behalf of the village at large, for the fortunes of the entire community-~male and female, chief, and commoner--depend on the insevg, It should be kept in mind, however, that the male role with respect to the first fruits of the domestic yam is paralleled by the female role in the 1 sevu i koile. FO0TNOTES--CHAPTER IV 1The machete (1 sele) is the gardner's most important tool. Indeed men and women never go the garden or forest (veikau) without one, and even small children can use them with consideraBIe skill. The blades can be purchased in the village store, but men prefer to make their own knife handles. The end result is a heavy bladed knife about 3 feet long which, when swung with both hands, can cut through small trees with ease. The blade is sharpened before each day's use, and through repeated filing gradually becomes shorter, narrower, and lighter. Eventually, it will be hafted on a shorter handle and become a woman's knife suitable for the more precise task of weeding among the developing taro. 125 CHAPTER V THE WATER OF LIFE Enough has already been said in the references scattered throughout the previous chapters to indicate the importance of kava (139933) in Fijian culture. This importance is manifested in various contexts: in the rituals of kinship and chiefship; in the treatment of illness; in the atonement for misdeeds and the repair of social relations; before any major undertaking and after the completion of any joint work; in recognition of arrivals and departures; in all public assemblies; in the extension of hospitality; and in social gatherings of all kinds. It is not only a refreshing beverage used in both ritual and recreational contexts; it also has its darker uses. It is a source or conduit for power which can be used for evil, as well as good. If taro and other forms of kakana dina are the staff of Fijian life, then yaggna is “the water of life." The justification for including a chapter devoted to food in a study whose primary foci are kinship and social organization is that, for human beings food is never just a collection of nutrients. Food is also a class of symbols which, through sharing and exchange, express in especially potent ways the very relationships which wewish to study. The same can be said of yagona, for it is never drunk alone. It is always shared, but in most 126 127 cases the sharing is preceded by an act of exchange. However abbre- viated the ceremony, yaggng_is usually presented and received, and the structure of this transaction can be taken as paradigmatic for all ceremonial exchange. Moreover, the etiquette of yaggna_drinking reflects principles of age, rank, and kinship. Indeed, yaggna_is so much a part of Fijian life that some dis- cussion of its use is necessary, if for no other reason than to avoid distorting the description of that way of life. In all the time that I spent in the village, I can recall only one day in which yaggna_was not being drunk in some quarter of the village. It was during the dry season when cool night time temperatures lessen enthusiasm. Many men were out of the village for one reason or another, and a few of those who remained gathered at the village cooperative store when it opened at five in the afternoon. Normally dried yaggna is on sale there, but on this particular afternoon none was available, and none of the men present had thought to bring a fresh root back from the gardens. They stood around for a few minutes assessing the situation before resigning themselves to a "dryf evening and going home in disap- pointment. Yaggna_is not restricted to the village setting. It is also very much a part of urban life. It is on sale in the market places of all urban centers. It is sold in bulk, in powdered form, or as a beverage. Small packets of powdered yaggna_are even on sale in the‘ modern supermarkets of the major trading companies. The yaggna_bowl is a regular feature in offices around the capital, and casual drinking is an.inobtrusive feature of the working routine. 128 The word yaggna refers to both the plant, Piper methysticum, and the beverage infused from it. The plant, a member of the pepper family, is a shrub with broad, heart-shaped leaves and smooth fleshy stems, pale green in color in the variety most common in village gardens. The stems are significantly thicker at the nodes giving them a jointed appearance. The plant consists of a number of unbranched stems joined at the base in a knotty mass. When the plant is harvested, the stems are cut off about 1 foot from this basal crown, for it is the fleshy subterranean stem and the fibruous roots which are used in making the beverage. The plant is said to grow to a height of 6 to 9 feet when fully mature (three to four years), but the tallest shrub that I was was about 5 feet tall. The longer the plant is left in the ground, the more potent the beverage will be, but local demand is high, and plants are often harvested after two or three years. Yaggna_grows well in pockets of black colluvial soil on well- drained hillsides, and some men plant it in such soil, but yaggna_is also grown in the sandy soils of the alluvial flats where it is often intercropped with taro. It is even grown in the village itself, behind houses, beside paths, and on the slope above the river. 199933. is propagated from cuttings (5353) bearing nodes. The soil is pre- pared for planting by clearing a circular plot 1 - 11 feet in diameter. The soil is turned over to the depth of 1 foot with a digging fork (mataiva) or other digging implement and is then tilled by hand so that all the clods are broken up into fine-grained soil. About six to eight cuttings about 8 - 10 inches long are poked into the soil 129 close together and at an acute angle to the soil surface. These are then completely covered with grass to protect them from the sun. As noted above yaggna is frequently intercropped with recently planted taro. Once the cuttings are covered with grass, they are difficult to see, and to prevent them from being damaged when the taro is weeded, a piece of reed may be stuck in the ground to mark their loca- tion. The taro will be harvested long before the yaggna_fully matures. The number of yaggna_a man plants depends on numerous factors including his own drinking habits, whether it is being grown for con- sumption and presentation only or whether it is also being cultivated for sale, and estimates of future requirements. For example, a man once showed me a large stand of yaggna_in his gardens and explained to me that he was maintaining it (margrgya) because he had 'a problem" (139a). His father was in his 805 and would die soon, and the funeral would require a large quantity oflyaggna. The parts of the plant actually used in the preparation of the beverage are the extreme basal portion of the stems (legena) and the rootstock or rhizome and the roots (wakana). The former fetches a higher market price and the beverage brewed from it is more valued. When the plant is brought back from the garden, the root mass is washed in the river to free the soil adhering to it and is later washed again under the tap as the legena_is prepared for processing. This involves pairing away the thicker bark and cutting out any dark or diseased portions much as one would cut away the eyes and imper- fections of a potato. The flesh of the lewena_is starchy and ranges in color from white to yellowish or orange. 130 The next steps in the processing of yaggga_depend on whether the beverage is going to be drunk immediately or stored for future use. If it is to be used immediately, it is grated into a pulpy mass by rubbing it back and forth on slabs of coarse grained rock (ygga, “to grate"; 1.1351, "grater").. This is a laborious process, and at weddings, funerals, and other public events which require large quan- tities of yagona, special sheds (vakatunuloa) are constructed to shelter the young men and youths to whom this task falls. When a sufficient amount of grated root is accumulated, it is shaped into ‘balls (bgli) about 5 or 6 inches in diameter to be infused in water. If the yaggna is being stored for future use, rather than being grated, the legena is sliced into thin cross-sections. These, along with the EEKEEE.3T9 then placed on a sheet of roofing iron and placed in the sun to dry (vakasigane; siga = "sun“). When fully dry (madgkj) the yaggna_will be a fraction of its former weight, and greyish in color. Before being mixed with water, the dried yaggna has to be powdered. Cast iron mortars and pestles are used for this, and almost anytime of the day or night one can hear the dull fwhump" of the pestle and then the bell-like ring of the mortar as it is struck on the up-stroke. laggna_is only bought and sold in the dry form. Men may dry a small portion of wakana from time to time to present to their hosts when drinking in the houses of others. But men only dry yaggna_in quantity for commercial purposes. Most of this is sold locally. Once in a while a man may sell a bale of dried yagona to the cooperative 131 store, but interestingly enough the store buys most of its yaggna_ from men of other villages. It is more common for yaggna to be sold directly to village consumers. The dried yaggna is ground into powder and then wrapped in little paper packets which sell for 20 to 50 cents each. When going to another house to drink yaggna, particularly if the house is of another matagali, it is customary to present some yaggna_ to the host. (Such a presntation of yaggna is known as an j_sevusevu. Guests who habitually arrive without bringing something are said to have given an.i sevusevu dredre. Dredge means "to laugh or smile." In other words, such guests bring nothing but their smiles.) Even if the company are drinking freshly grated 199939, these small packets of powdered yaggna would be an acceptable presentation, and so there is a dependable demand for them.1 Fresh and dried yaggna_are never mixed together in the same bowl, and rather than mix a second bowl of ngggflg, the host may keep any Jgggflgi presented for his own future use. It is difficult to describe yaggna_to someone who has not tasted it. It is astringent and produces an almost involuntary pucker- ing of the mouth. It also has a numbing effect on the tongue and lips. Even habitual drinkers will admit that yaggna_is bitter (gage; the word also means "poisonous.")Lester (1941:105) reports that on the northwest coast of Viti Levu the word "9992? means both Ybeverage" and "bitter" and suggests that the word 'yaggna,“ used throughout the islands for both the plant and the beverage, may have come from this area . 132 If yaggna is bitter, it is not unpleasantly so, and it is refreshing on a hot day. Actually, there is a significant differ- ence in tasts between the beverage derived from the freshly grated root (bulabula, "live," or drokadroka, "raw") and that infused from the powedered dry root (madgkj).. Since 999351 has minute particles of the powdered root suspended in it, bulabula is thinner and has a cleanerst'a‘ste.. 0n the other hand, 91929.. has a mellowness about it that I found more pleasant than the fraw" taste of yaggga_brewed from the fresh root. While the two types of beverage affect the drinker in similar ways, they differ in the strength of what, for want of a better term, I will call their inebriating effect. (The term is misleading since, unlike alcohol, yaggna_does not depress the central nervous system. But the term does reflect Fijian usage.) The Fijian term for this effect is m, a shortened form of @9111 m. "_M_a_t_e_" is a word which refers to diminished vitality, whether total (death) or partial (sickness). The term fmateni" is also applied to alcohol intoxication. Fijians are explicit about the parallels between yaggna and alcohol, but if they recognize similarities, they also recognize differences. They point out that alcohol can lead to rau- cous or even violent behavior, but when one drinks yaggna, they become pacific. Despite this difference, alcoholic beverages, as a general category, are referred to as yaggga_ni_vavalagi or "white man's yaggna," 123933 bulabula, or that prepared from the freshly grated root, is considered to have the stronger effect and is preferred for 133 that reason. One informant compared dried yaqona (maduki) to beer A and said he preferred to drink rum (i.e., yaqona bulabula). Some men will only drink yaqona bulabula. The physiological effects produced by yaqona depend on the degree of concentration of its active ingredients and, to an extent, on individual susceptibility. The level of concentration of the active substances depends on the variety of yaqona_plant, the age of the plant, the type of soil it was grown on, and, no doubt, other factors affecting plant growth as well. These differences are appre- ciated. I recall sitting one afternoon in a vakatunuloa (an open- sided shelter made of roofing irons and bamboo) where young men were grating yaggga_for a funeral. A young man picked up a root and began to prepare it for grating when he suddenly stopped. The lewena_was a rich orange in color, and this was taken as an indication of its superior strength. He sniffed at it and perhaps nibbled at it as well. He asked where it had been grown, and upon being told, he decided to set it aside. Rather than mixing it with the pulp of inferior roots, it would be grated and infused separately. Meyer (1979) found that when administered to laboratory animals, various active substances found in the resin of the yaggna root produce muscular relaxation. Larger doses of some of these substances produce ataxia and ascending paralysis without affecting the consciousness of the subject. The substances were found to be most effective when administered intravenously. The oral dosage necessary to produce paralysis was found to be ten times higher. 134 These same substances were found to be more effective than the muscle relaxant mephenesin in preventing convulsions and death in mice subjected to toxic doses of strychnine. Some of the kava pyrones were also found to inhibit contractions in isolated ileum or uterus. In addition to being powerful muscle relaxants, they were also shown to reduce edema, act as a local anesthesia, and have mild anti-pyretic properties (Meyer 1979:137). .Yagona's soporific properties have been known for a long time. The active substances studied by Meyer are relatively insoluble in water. Accordingly they were dissolved in peanut oil for intra- peritoneal and oral administration and in polyethylene glycol for intravenous injections and experiments on isolated organs (Meyer 1979: 133). Since kava is mixed with water by the peoples who have tradi- tionally used it, experiments with its water soluble ingredients have perhaps greater relevance. Buckley found that a lyophilized steam distillate of kava root (LE—1) and two water soluble subfractions of that distillate depressed spontaneous motor activity in mice (1979: 149). Pfeiffer reports that crude kava root administered to nine uncontrollable epileptic subjects produced 'a better degree of seizure control" but also produced tinting of the skin and sclera as side effects (Pfeiffer et al. 1979:156). Despite the testimony of habitual drinkers of yaqona, some authors have expressed doubts about the reputed physiological effects of the beverage. For example, Ford concludes that "with the evidence available, it seems that early reports on the physiological effects 135 of kava drinking were greatly exagerated“ (1979:172). He notes that kava does produce a slight numbing of the tongue and throat, but he attributes the ataxia which some Europeans have experienced (e.g., Hocart 1929:59) to the smoke of strong native tobaco in poorly venti- lated interiors coupled with the effects of sitting cross-legged for long periods of time, an explanation echoed by Holmes (1979:107). Ford feels compelled to ask why yaqona plays such an all-pervasive role in Fijian culture. He finds no explanation in the physiological effects of the beverage. He writes that “it seems more likely that in considerable measure the importance of kava to the people of western Polynesia and Fiji is derived from the part it plays in their life“ (1979:172). There would appear to be a certain degree of cir- cularity in this formulation, but what is of interest here are Ford's doubts about the physiological effects of kava drinking. Writing in the same volume Gajdusak describes in some detail the reactions he and other members of his team experienced after drinking kava on the island of Tongariki in the New Hebrides. He reports that A few peculiar paresthesiae of the face, legs, and arms-- especially of the legs--a slight feeling of numbness, tingling, coldness and then weakness, accompanied early by shorter flashes of warmth or flush, occurred during the first half hour after Lingestion. . . . There is, with higher doses, a pleasant, relaxing, paresthesia-enjoying, refreshing state of somnulence without mental dulling which eventually leads to sleep at times, members of our team have taken large doses--a large coconut shell full--and real weakness, even a paresis making walking impossible, has been present for several hours after ingestion (1979: 122). 136 Gajdusek reports that he also drank kava in Fijian villages without these effects and attributes the difference to the fact that on Tongariki the fresh root is prepared by chewing which, combined with a certan amount of salivary digestion, breaks up the woody fibers more effectively than grating and emulsifies the active ingredients (Cf. Steinmetz 1960:43). Moreover, the beverage drunk on Tongariki is apparently much more concentrated than Fijian yaqona, My experiences with kava differ from those of Ford or Gajdusek in Fiji or Holmes in Samoa. While I do not think these differences are explained by the fact, it is likely that individuals do differ in their degree of tolerance to the active ingredients in kava. Over the 17 months of my stay in the village, I drank mega about three times a week for several hours at a time. On certain occasions such as funerals or the j_§evg_discussed in the last chapter, I spent much of the day drinking yaggna as did most of the men of the village. The effect of the beverage seemed to depend on a number of factors. I found the fresh root to be more powerful than dried yaqona, and if I drank too much yaggna bulabula, the effect was the same as if one had eaten too much chocolate or other rich food (i.e., a sensa- tion ranging from dyspepsia through an intense burning sensation). As already noted, yaggna plants are recognized as differing in potency, but my palate never became sufficiently cultivated to detect the differences an experienced yaqona drinker would. The most important factor affecting the potency of the beverage is the amount of water used in making it. If the dried root is being used, the powder is placed on a piece of cheesecloth (sulu‘ni_xagona) 137 about 12-18 inches square which is then carefully folded and gathered in the hand to prevent the powder from escaping into the beverage. This is then kneaded under the surface of the water until the person preparing the drink thinks it has reached the proper consistency. If the grated fresh root is being used, the ball of pulp may also be put in a cloth, or a strainer (ithEE) might be used. The j_lgba_con- sists of many strands of fiber derived from the bark of the veg tree (Hibiscus tiliaceous). The j_lgba is preferred because it is believed that more of the juice containing the active ingredients is extracted when it is used. In fact, however,only one i lgba_was used during my stay in the village. It was used throughout the period of mourning following the death of a highgranking man. When the j_lgba_is used, the pellet of yaggna pulp is kneaded just as it when a cloth is used, but in the process, a lot of particulate matter gets into the beverage. This is removed by pressing the strainer tightly against the front surface of the yaggna bowl and passing it through the liquid. The j_lgba_is than shaken out (vakamirake) by giving it a vigorous snap. The process is repeated a number of times until the liquid is suffi- ceintly clear. When the‘person mixing the yagona (k_olos.e_yaqona) deems it to be of the proper consistency, he signals to the highest ranking of the senior men present that it is ready to be evaluated. The lose yaqona then takes a coconut shell cup full of the liquid, lifts it above the yaqona_bowl and pours the liquid out. The color and relative clarity of the liquid are indications of its strength. The person preparing the 138 drink may be told to dilute it with water of to repeat the kneading process (Log; _i_;_a_l_e n_a m). Younger men often have a tendency to make the yaggna too strong for the elders' liking, and some old men occasionally keep a bottle of water beside them to further dilute their drink. The degree to which one experiences the effects of the yaggna_ depend, of course, on how much of it one drinks at a sitting. In theory, at least, one is able to tell the cup bearer whether one wants the coconut shell cup full (sjnaj), half-full (veimama), or just a little of the drink (vaka lailai). 'In practice, however, the quantity served is determined by the cup bearer, and one usually drinks what is offered. If the gathering is informal, and if the cupbearer is one's cross-cousin (333319), or if a cross-cousin is nearby, it is useless to demure when a full cup is offered, for in that case an even larger cup is likely to be substituted. Further protest is likely to be cut short when the brimming cup is placed in the lap of the reluctant drinker. Such antics are a common expression of the classic joking behavior characteristic of the interaction of cross cousins. But it should be noted that there is also an element of respect in the giving of a full cup. Drinking it can become an ordeal, but the offering is also a statement about the person'a ability to drink vakalevu (in a big way). Some elders habitually bring their own small cups which the cupbearer fills from the common cup. If one only drinks vakalailai (a little), one can drink for a long time without experiencing the yagona's effect. 139 Another factor which has to be taken into consideration is the length of time between rounds. In the early stages of a yaggna_ session, these may come every fifteen minutes or so and are generally worked into the ebb and flow of conversation. "A little yaggna, a lot of talk; a lot of yaggna, little talk," summarizes the situation neatly. As the hours roll on and the company begin to feel the numbness which yagg 3 brings, faces take on a sullen expression, the talk slows down, the interval between rounds becomes longer, and each round is a little more of an ordeal (at least for the ethnographer). Perhaps a brief comment should be made about the importance of talk in the context of yaggna_drinking and the importance of the yaggga_circle as a forum for talk. Aside from relatively infrequent public events such as funerals, weddings, and meetings (all of which will involve yaggga_drinking), yaggna_is the only thing which brings numbers of men together on a daily basis. To the outsider,Fijian villagers seem to have an infinite capacity for talk. But, of course, it is precisely being an outsider which makes it seem so, for most of their conversation has to do with everyday events and the characters and foibles of local actors.2 From time to time national and inter- national events are also topics of converstaion, but generally speak- ing, the further removed they are from the local arena, the less imme- diacy they are perceived to have, and the less there is to say about them. Informal conversation around the yaggna bowl fulfills the impor- tant function of keeping the community informed about current events, future events, and the actions of others. If a consensus is not 140 reached, at least all of those present have had an opportunity to form an opinion. It is interesting to contrast the importance of conversation in Fijian kava drinking with Gajdusek's description of the way in which kava is used in the island of Tongariki in the New Hebrides. Usually, half of a coconUt: shell or a bowl of the same capac- ity is used to prepare the kava and the full contents--about 100 ml.--drunk slowly in one draught. . . . A kava drinker usually eats immediately after taking the kava; the kava is prepared while the evening meal is being cooked. The effects come on in a half-hour more or less, and the drinking is thus usually postponed until food is ready. Those who have drunk kava find a comfortable place to sit, often beside a dying fire in the dark house, where they remain hunched over and avoiding light and sound disturbances of all sorts. Con- versation ceases, and slowly they fall into a kava-induced stupor, which is not true sleep. This stage occurs about an hour after drinking. . . . A few hours after they have drunk kava they arise and walk to their own houses to fall asleep promptly again; others remain where they have first "fallen." In early morning they appear fresh and without any “hangover"-like sequelae (1979:121-1). Fijian yagona drinkers often reach the same point, but it takes them longer to do so. And when drinking in a recreational context, they will also have been entertained and will have enjoyed the camari- derie of the experience. Possibly some important communication will have taken place as well. I have already described the numbing effect which yaqona has on the tongue and lips. This is experienced fairly soon after drinking a single cupful. The first several cupfuls also seem to have a relax- ing effect. While the physiological basis for this is quite different, the effect is the same as inbibing a moderate amount of alcohol. Con- sersation flows freely. After about six to eight cupfuls, I usually 141 began to feel pleasantly drowsy. At this stage I would not experience any impairment of motor activity, but my face would feel somewhat numb, and my appetite seemed to be sharpened. After about ten to fifteen cupfuls, I began to feel extremely drowsy. At this point falling asleep while standing or walking not only seemed possible, but actually seemed desirable. Along with the drowsiness, there was often a certain degree of ataxia. This was most noticeable, or perhaps only more memorable, on those occasions when I had to walk home from other villages. There is a small creek named Waisucuwalu which runs between Nairukuruku and the village of Taulevu. Over this creek there is what once must have been a solid bridge constructed of creosote-treated planking, but it had fallen into disrepair. Most of the cross-planks were missing, and one of the two main beams had become so rotten as to be unsafe. People routinely crossed on the one remaining beam, studded as it was by the spikes that once held the cross beams. I did not relish crossing that bridge under normal conditions, but after drinking yaggna_for most of . the day it became an impossibility. Luckily my companions, being more sure footed to begin with as well as more experienced yaggna_drinkers, did not share my difficulty. When the characteristic torpor sets in, the yaggna_drinker begins to crave peace and quiet. Bright light is irritating, and sound of any kind is disturbing. Conversation becomes burdensome. When sleep does come, it is extremely restful. But by the time I had reached this point, I would have already experienced some less pleasant side effects. I found yaqona to be not only a strong diuretic, 142 but also, when drunk in large quantities, an infallible purgative, and both of these qualities probably contributed to the extreme thirst I felt after drinking for several hours. Also the photophohia experienced seemed to be associated with a feeling of pressure behind the eyes and tension at the base of the skull. I sometimes woke up the next day with the same headache and still thirsty. I have discussed the physiological effects of yaqona_drink- ing at some length because, unlike Ford (1979), I think they have to be considered in any attempt to explain the importance of the bever- age in both recreational and ritual contexts.3 The use of 229922.10 religiousritual seems to be historically prior to its widespread use in secular contexts (Cf. Lester 1941), and so it is perhaps to these - functions that an argument ought to be addressed. In his "Northern States of Fiji," Hocart makes a distinction between the founder gods of Fijian tradition, the "gods of the land," and shamanistic gods appealed to in cases of sickness and warfare. Though yaggna_was as central to the cult of the former as it was to the latter, it is significant that shamanistic practitioners used it as a means of inducing the trance state in which direct contact with the supernatural was believed to be possible. "In the shamanistic cycle everything is precise; the mechanism is clear: when the shaman drinks the offered kava, the spirit comes down the curtain of the temple and 'jumps' or 'embarks' on him, so that he goes into a fit and prophesies" (Hocart 1952:12). Kava was used for similar purposes elsewhere in Oceania. Handy (1940) notes its use to "induce the desired passivity or trance" by seers and makes the point that, in function, it is 143 similar to a hypnotic though its physiological effect is quite dif- ferent. Unlike hypnotics kava does not have a depressing effect on the cerebral cortex (Meyer et al. 1979:177). When drunk in sufficient quantities, yaggna has identifiable physiological effects, and this marks it as a thing of power (mega). It is this power which is tapped when yaqona_is used to make contact with the supernatural. Ultimately, of course, the mana_of 129202 is not reducible to scientifically measurable physiological effects; it is a cultural construct. But the construct does have a basis in the physiological alterations which the drinker experiences. We will have occasion to return to this matter of mana_later in our discussion of yaggna_and its uses, but now I want to turn to a consideration of the etiquette of yaggna_drinking. I should note at the outset that it is difficult to make general statements about actual behavior because this depends on many factors. the foremost of which are the relative ages and rank of the drinkers. For example, when youths (cauravou) or young men gather to drink among themselves, the emphasis is on equivalence and camaraderie. There is a general relax- ing of compartment and, along with it, a suspension of some points of etiquette that would be observed if senior men were present. With this is mind, let us begin with a description of the formulae which accompany the mixing of yaggna. 133933 may be prepared in a variety of containers: a galvan- ized bucket, a porcelain basin, or a traditional pottery bowl. More frequently, however, it is mixed in a special wooden bowl (tanoa) 144 carved from a single piece of hardwood from the greenheart of India tree (Afzelia b1jgga). Variations in the size of 1angg_are discussed in terms of the number of legs which they have. Most of those in the village had four legs, a few had six, and one or two had eight. On one side of the bowl there is a wooden lug through which a sennit rope is passed. This forms a loop which can be used to suspend the bowl when not in use. One or more white courie shells (6911), symbols of chiefly rank, may be fastened at the end of the loop. When the bowl is in use, the loop is stretched out toward the highest ranking man present. In actual fact, most of the tanga in the village were without the sennit rope (w§.n1“1angg), but they were, nevertheless, . treated as if the rope were present. That is, the lug to which the rope would have been attached was always oriented toward the person of highest rank. If for some reason one needed to cross the room, one did it in a crouch, for it is offensive to be higher than a seated person, especially elders and men of rank. Nor did one cross the room directly but, instead, went first to the 1anga_and lightly touched its rim with an outstretched right hand before crossing the rope that was not there. The highest ranking man present sits at the apex of an arc formed by lower ranking men to his right and left. The bowl sits [before him, and behind the bowl sits the person responsible for mixing and pouring the yaggna.(kg_lose yaqona). Slightly behind the 195g. Iygggng_and to either side sit other young men who will serve as cup- bearers and clap at appropriate times during the ceremony. They show their respect for the chiefs by wearing sulu,cn~kneelength cloths 145 wrapped around their waists, rather than the short pants they wear while working in the gardens. Behind these functionaries sit the youths and younger men who, strictly speaking, are present to minis- ter to (veiqaravi) the elders. The orientation of the ygggng circle is consistent with what was said earlier about the floor plan of the Fijian house. That is, the highest ranking man sits in the upper- most part of the house and younger, lower ranking men sit below him. On most occasions any women who are present sit behind the young men --in fact, near the door in the lower end of the house.4 On certain occasions senior women of chiefly rank may be honored by being seated among men in the upper end of the house. I only witnessed this twice, at the funerals of persons towards whom the women's families, as chiefs, stood in a special relationship. (If the deceased was married, man his widow also sits in the upper end of the house, but usually slightly apart from the ygggna circle.) When all the principals are seated and the yagggg_is ready to be mixed, the lose yaqona, sitting cross-legged behind the tanoa, leans slightly forward and, placing the tips of the fingers of both hand on the forward rim of the 1anga,he announces in a quiet voice and with downcast eyes, "Vakarau lose saka na_yaqona vakaturoga.§ ('Sir, the chiefly yaggng_is ready to be extracted [wrung]".) To this one of the elders may or may not respond with the quiet command figgge," At this point another young man (32.1111Pwa1, or water pourer) kneel- ing or squatting beside the 1anga_pours water over the hands of the lose yaqona and over the yagona itself until the tanoa is filled to 146 the desired level. The lose yaqona kneads the cloth containing the yaqona (sulu _n1 1330—03.) under the surface of the water. When the kneading process is over, he holds the cloth high above the bowl and at arms length and wrings out the last drops of water. He then takes the cloth and wipes the rim of the 1angg over its entire circumference. It is at this point that he asks the presiding chief to judge whether 'the‘yaggna_has been properly mixed. If neither water nor additional kneading are deemed necessary, the lose yaqona announces "Lose 911. saka ng_yaqona vakaturaga." (Sir, the chiefly yaqona is wrung out [extracted].“) At this point there may be a brief pause until the person who is offering the 139933 takes up his position beside the Igngg, Kneeling beside it, he delivers a speech of offering and dedication which is followed, in turn, by an acceptance speech on behalf of the person sitting as chief of the yaggng, The acceptance speech may be made by this person himself, or, in the case of a high ranking elder, the yaggna_will be accepted on his behalf by another. Finally, there is a speech given by a third person in which he asks for blessings on behalf of both the giver and receiver of the yaggna_as well as those assembled to witness the transaction. This benediction is known locally as the caucaunake. When the speeches have been completed, the lose yaqona says in a clear voice "Talo saka ga_yaqona vakaturaga." ("Pour out, sir, the chiefly yaqona.") Immediately afterwards he, the young man who is to serve as cupbearer, and those directly behind them clap (cobo) three 147 times with cupped hands. The lose yaqona then takes a coconut shell cup and, after first stirring up the ygggng_(a sediment quickly settles to the bottom of the bowl), pours it into a similar cup held by the cup bearer (DE gflnygggna). The cupbearer then rises and carries the cup to the presiding chief. If the occasion is especially important (9161 or "heavy"), or if the person receiving the yaggng is especially exalted, this is done in a very formal manner. The cup- bearer sits down before the chief, and with his right forearm extended and supported ina stylized manner by his left hand, he offers the chief the cup. The cupbearer calls back to the lose yagona and those who sit behind him to 9929 39y! (clap three times). The chief may offer a brief toast before drinking, and then, taking the cup in both hands, he drains it in a single draught and throws it spinning down on the mat. The cupbearer then exclaims that the cup is dry ("A . . . 52521"), and he, along with the other young men ministering to the elders clap loudly. The same procedure is followed with the second drinker, except that the person who drinks as a second for the chief only receives two claps after the cup is offered. If the occasion is a formal one, others may be honored by the recognition of their chiefly status, and they will also receive £999 tolu. Finally, when all those who are drinking as the chiefly representatives of various social units have been served along with their seconds, the cupbearer sits down beside the 1agga, and the lose yaqona announces 9§a_maca saka vakadua na_yaq9na vakaturaga.f ("Sir, the chiefly yaqona is entirely dry.")5 The lose yaqona and his assistants then 148 give three claps. At this point, the original cupbearer retires and another takes his place to serve the general assembly in order of seniority. Thus the difference between the formal yaqona vakaturaga, or chiefly ygggng, and the drinking of ordinary men is underscored by a change of personnel. Except during the foraml yaqona vakaturaga which sometimes preceeds the serving of the general assembly, it is proper to clap once before accepting a cup of m and then several more times after draining the cup. When the cup is taken, the cupbearer also claps, three times for the man sitting as the chief of the yaggn§_and twice for others. The aseembly at large should join in, at least, to give fee 3 191g" for the chief. The reverberating sound of men clapping in unison carries a long way in the still night air, and it is one of those distinctive sensory experiences which, when recalled, so evoke the field experience. This clapping in unison is followed by a staccato of clapping which marks the draining of the cup. All of the details of the formal procedure described above will not be observed on all occasions, but it does provide a model of the proper etiquette of my; drinking. When younger men, age mates, gather to drink some of the finer points may be ignored. Often no one will be willing to take up the position of the chief of the 139933 circle. In the case of young men, this is a matter of diffidence and a statement of the egalitarian tone of the gathering. In the case of middle-aged men, it is more likely to be motivated by the wish to avoid the embarrassment of having to move should a higher ranking man arrive. 149 Prior to the coming of the 191g, the Christian faith, only elders drank ygggna, Young men's involvement was restricted to preparing and serving it.6 Today youths begin smoking tobacco and drinking ygggna at about 16 or 17. Most of their drinking is done with their age mates or slightly older men. They also occasionally drink when serving ygggna for the elders. Some youths routinely accompany their fathers and fathers' brothers to mix and serve yaggng,' Young men in their twenties prefer the company of other young men because of the relaxed atmosphere. When drinking with older brothers or with senior men related to them as fathers or mothers' brothers, their relative youth places them in a subservient role, and they have to remain respectful observers and general factota, rather than full participants. The preferred drinking companions are one's 1aggle, or cross cousins, because these are people with whom one is not only permitted certain liberties, but people with whom joking is a matter of etiquette. The antics of gaggle provide a certain amount of comic leavening for what might otherwise be a fairly somber and reserved affair. The same function is performed by a few individuals who are invariable objects of good humored fun, plied with bowl after bowl filled vaka sinai, their sulu tugged, and their persons prodded in unprotected places at unsuspecting moments. Enough has already been said to indicate that the seating arrangements reflects the principles of rank and seniority. The old- est member of the chiefly matagali present will be accorded the status of chief of the yaggga, The 1 talatala, or Wesleyan minister, and the 150 headmaster of the district school will also be accorded a very high position in the yaggn§_circle if they are present. But once these positions have been taken, seniority often takes precedence over chiefly status. That is, elders of commoner matagali usually sit higher than younger men of matagali Siko or one of the other chiefly matagali of the area. And on such occasions, it is not their possible status as members of matagali which function as ma1a_n1_gagu§_or.bgt1 which is.;salient. It is, rather, their age which is being respected. Another factor which affects the seating arrangement on both formal and informal occasions are the kinship ties among the partici- pants. A man (ego) will never sit next to his father (11111111931) nor to any classificatory father who is senior to him (that is, senior to ego). Nor will he normally sit next to an elder brother Q31 1gkgna), for he, too, is a person to whom one owes respect. Brothers frequently sit on opposite sides of the room; this is especially true in the case of elders, for either of whom a lower position in the circle would be inappropriate. The behavior of classificatory brothers is more variable. If anything this is a more difficult rela- tionship than that of full siblings, particularly if the fbrothers" are close in age and members of the same matagali. The interaction between classificatory brothers related through women (e.g., mothers' sisters' sons) appears to be on an easier footing. The relationship of greatest respect is that of mother's brother (mgmg) and sister's son (ga1uyu). Not only will the sister's son avoid sitting next to his mother's brother, but he must also avoid 151 drinking from the same cup. (Aside from a few old men who carry their own small bowls with them, everyone drinks from the same coco- nut shell cup. In addition to this communal cup the lose yaqona also has a cup which he uses to stir and pour the ygggna, and there are usually one or two additional cups so that it is not difficult to avoid the cup from which one's memo drinks.) Avoidance of the meme in the context of ygggn§_drinking is not limited to the actual mother's brother but includes classificatory mother's brothers as well, even where there is little difference in age between the 9mother's brother" and his "sister's son.f Avoidance of the classificatory mother's brother may also extend to avoidance of the mgmgfs cup. The relationship between veivugoni, wife's father and daugher's husband, is another relationship of respect (veidokai).7 A man owes his wife's father the same degree of respect due the mother's brother, for indeed, the wife's father is structurally equivalent to the mother's brother.8 He may be referred to as meme and, in fact, may be related to ego as meme prior to ego's marriage. Because this is a relationship of veidokai (mutual respect) a man does not sit next to his wife's father while drinking yaggng. (In most cases the differences in age between wife's father and daughter's husband would be sufficient reason for them to be seated in different areas in the room.) Even toward the lower end of the ygggna circle, the seating arrangement is never random. However, recognizing the pattern requires a thorough knowledge of the actors and their interrelation- ships. On occasion readjustments are required. It often happens 152 that, when one man gets up and leaves, the two men on either side of him are related to one another as classificatory mgmg and 131312, They may remain as they are with a gap between them; or one of them may move to another spot; or a third party might sit down between them relieving them of their embarrassing proximity. As noted earlier, the importance of ygggng_stems, at least in part, from its status as a thing of man; (power). As an attribute of persons man; is clearly differentiated from physical strength (kaukauwa) or political power (lewa), though both of these can be manifestations of gang. In translation mgna_is frequently disting- uised from other forms of power as being a supernatural aspect of persons and things. The dichotomy evoked is truer to the categories of modern Western thought than it is to the Fijian concept, but this is not the place for a full disquisition of the topic. What is of interest here is that gang, as an aspect of persons, can be nur- tured or enhanced by the drinking of yaggng, for yaggng_is itself a repository _n1an_a. Informants stated this to me explicitly. M is also a source of or path to knowledge (ygkg) of the deeper nature of things. For this reason persons who would cure illness or see beyond the normal confines of space and time must drink 139993, These are not goals to be achieved by everyone. To do so requires personal .EEEE.35 well. But any person's well being is enhanced by drinking ygggng, providing it is not drunk to excess. The reason why curers and seers utilize ygggna in their practice is that it enhances communication with the realm of the 153 spirits. One manifestation of this attribute is that the transmission of prayers or benedictions is more direct or efficacious when they are spoken over mg. It is difficult to say how this notion and other aspects of local custom and belief relate to the doctrines of orthodox Chris- tianity. The Wesleyan faith is sincerely believed in, and the church plays an important part in the life of the village. But that faith does not preclude the holding ofother beliefs not sanctioned by Christian orthodoxy anymore than the utilization of Western medicine precludes the use of traditional methods of curing. Just as there are believed to be ailments which Western medicine is powerless to cure, so are there spiritual beings and forces effecting human con- cerns which are best approached in traditional ways. The power of mnat only enhances the efficacy of the requests for health and prosperity spoken over it; that power can also be used to harm others. Another name for sorcery or drag £1_kgu_is ggyanyaggng, that is, "to pour yaggng," Earlier I mentioned the fact that yggggg is never drunk in solitude. Even when sorcery is being practiced, ygggng_is drunk in the company of others, but in that case, it is drunk by two people only. One of these prepares the beverage, and the other utters the curse while pouring out the ygggng, presumably to the effecting spirit though this was never stated. What people stressed when describing sorcery to me was not its aggressive aspect but, rather, its anti-social character. Drinking yaggng_"alone" (that is, with only one other person) and the sorcery itself were seen as 154 part of an on-going pattern of anti-social behavior. The other thing which was stressed was the efficacy of the ygggng_itself. Though one person mentioned the burying of leaves (the term for sorcery, drag Q1 53!, means literally the "leaf of the tree“), all stressed that it was the pouring out of xggQfl§_Wthh made the curse effective. Note that "not only is drinking alone an inversion of normal behavior, but so also is the pouring out of ygggng, rather than ingesting it. People did not hesitate to attribute the practice to others. Because sorcery requires one other person to be effective and that person must be someone trusted, married couples are typically the subjects of accu- sations. While it was asserted that the sorcerer's curse is inevitably successful, it was also stated that "pouring ygggng" is likely to have a harmful effect on its practitioners. I was told of one case in which a man in a neighboring village died suddenly. Sorcery was sus- pected. This was confirmed when, before the end of the four-day fun- eral, a heretofore healthy man died and his wife was stricken with a permanent affliction. This was interpreted as proof of their guilt. A more positive aspect of the power of ygggga_is revealed in the role which it plays in rituals of reconciliation (1.§g§g). In this instance, "power? is not intended as a gloss for the Fijian CO"? cept of mggg_but is, rather, understood in its commonplace sense as the "ability to act or produce an effect whether physical, mental, or moral.“9 The power to effect reconciliation is not unique to yaggna; it is also latent in tabua or whales' teeth, the most sacred form of 155 chiefly wealth (1 yau vakaturagg). As symbols yagona and tabua evoke, in every powerful ways, the moral obligations of kinship (veiwekani), the bond between chief and commoner (veiqalaki), and Fijian tradition in general (i valavala vakavanua). In offering them, the donor expresses repentence, reminds the offended party of the nature of their relationship, and expresses his desire for a return to normal relations. If his anger is still strong within him, the offended party may refuse to accept the offering, but such a refusal is not made lightly. Perhaps the following case would be a useful example of the role of yaggng in rituals of reconciliation. I include it here not only because it illustrates the symbolic value of yaggng, but also for what it says about beliefs in mystical danger, the role of adepts (xi g1 331) in dealing with those dangers, and the tension beween the potential for status rivalry (Goldman 1970) among chiefly lineages, on the one hand, and the ideal of matagali solidarity on the other. In the third month of our stay in the village, a highly respected man of chiefly rank died. He had suffered a severe stroke some months previously and had been partially paralyzed. The stroke left him bedridden, but in the weeks just before his death, his condi- tion had begun to improve. On the day of his death, he had ventured out with the assistance of his daughter and had sat for a while in the sun. Some minutes after he had returned heme, the midday quiet of the village was broken by the heart-wrenching sound of keening. The reaction was immediate. Men, women, and children ran out of 156 their houses to see what had happened. There was shock and an imme- diate out-pouring of grief. Within minutes the adult males of households with traditional ties to the dead man's lineage had taken up their vigil in the 131g._1.__gg. In the days that followed, I learned that the deceased was the third member of his sibling set to die suddenly in the space of a few years, leaving his lineage almost bereft of its senior males. His children and brothers' children were determined to know the cause for this. It was acknowledged that there might have been natural causes involved, a familial predisposition, perhaps aggravated by diet, but there were other factors that had to be considered also. Some days after the death members of the deceased's lineage visited a village some distance up river to consult a curer and adept. I first became aware of their return when I noticed a crowd of women around the empty house next to ours. It was a large structure and served as a village hall when one was needed. I went over and near the doorway met the dead man's youngest son. His face was stern and troubled. He told me that the adept had determined that their troubles stemmed from the fact that their ancestor had usurped the chiefly powers ofhis elder kinsman, an ancestor of the current chief's lin- eage. It was expressed to me later in biblical terms: the sins of the father fall on the children. The natue of the "sin" in this case, the usurpation of the position of the elder by a younger kinsman is a recurrent theme in Fijian culture. The dead man's son said that in order to prevent future troubles, his lineage (which I will call lineage B) had to "give back 157 their chiefs“--that is, give public acknowledgment of the fact that an usurpation had occurred in the past. He added in English that "Fiji custom is too hard," but he said that he and his brothers weretu_ra_g_a_, chiefs, and so there could be no veering from their responsibilities. On the appointed day the 1113111411 (curer) and his party arrived to direct the members of the dead man's lineage in the per- formance of the ritual. He was formally received in the same empty house, and if it were possible or proper to stand the phrase Fstanding room only" would have applied, for the house was literally filled with youths and men of all ages. The only group who were not present were the members of the chief's lineage, they who were to receive the 1_§ggg (hereafter, lineage A). An 1_sevusevu (i.e., formal presen- tation of ygggna) was given to the y§_n1hwa1 and his party by an elder of matagali Nabubuco as mat§_nj_yanga, or spokesman, for the lineage of the deceased. A bucket of yaggna_was then made available to the y§_gfl_wa11§_party in order that they could return the 1_sevusevu. Following a few rounds of m, the 11191 w_a_i_'_s spokesman reported to the gathering the reason for the adept's presence, the nature of his findings, and the steps that had to be taken to pre- vent future mishaps. He then turned to the events of the day. That is, how the 1 5939 should be conducted, who would speak, and who would drink in what order. When this was concluded, the ygun1nwa11§ spokesman asked for the tabua that were to be presented along with yagona to the chief's 158 lineage. Three were to be given, the same as the number of men who had died. These were placed perhaps three feet in front of the adept and his spokesman. At this point, something strange happened: the spokesman began to talk to something beside the three whale's teeth. His remarks were addressed to a 1§ygrg, the adept's spirit familiar. Necks craned to see what was going on, none more than my own. The spokesman explained to the spirit what was to be given at the 1_§grg_ --three tabua, yaqona, and a magiti--so that the tévoro would not stay, but would return to his own village satisfied that things had been done properly. I1aggng_was then drunk, but only two cups were served, one to the ygnn1 E91 and one to his spokesman. When this was concluded, people began to file out, and the adult males of the dead man's lineage began removing their shirts. Some appeared to do so reluctantly. This was not a matter of modesty, but rather, because to do so was to humble oneself, and this was a difficult thing. Those who had been inside the building formed a column along the sidewalk. Some men hurriedly wrapped borrowed §ulu_ on over their short pants, for this was a "heavy" (bTbi) matter which required proper dress. The members of other matagali led the way, and the males of lineage B followed carrying three empty tanga_and a bucket of yaggn_, Women were seated at various places along the route watching the procession to the paramount's house where the 1_§grg_was to take place. When we arrived at our destination, I sat with other men close to me in age in the shade of the chief's house. It was difficult to see and hear from there, for directly in front with their backs to 159 us sat the men of lineage 8. They sat facing the three tanoa, and in front of the 129923 bowls lay the carcase of a single cow pro- vided by the senior surviving male of lineage B. Directly across from them sat the chief and other senior men of lineage A. To the chief's left the more junior men of lineage A sat on a separate mat. Directly across from the latter sat the senior men of other matagali. The three fl , the m_ag_i_t1, and the m were presented to the chief by the spokesman for the 1§.n1nwa11§ party. They were accepted quietly and with saddened deameanor by the chief himself. The concluding benediction was offered by the chief's younger "brother," actually his father's brother's son. The most senior man of lineage 8 then began to serve as cup bearer to the senior men of (lineage A who drank in order of seniority. The younger men of lineage A were then served by junior men of lineage B, and finally, when all the members of lineage A had drunk, the members of other matagali were served. When everybody had drunk, the ritual was completed, and people began to leave. As they did so, the chief went up to the eldest son of the deceased, shook his hand, and kissed it. The seniority of the chief's line had been publicly acknowledged for a reason, and in a manner, which saddened him. His action, an unusual one for a senior man to direct toward a junior kinsman, expressed the feeling that the death of the young man's father had been a loss to the entire matagali and the community at large. 160 The majority of the crowd returned to the large, empty house to resume drinking. On my way there, I noticed several women gathered around the house of the senior male of lineage 8. Inside the adept was drinking with the men of the dead man's lineage. He drank there for the rest of the day. I was told that he had asked for no payment for his services, just food and yaggna, I was also told that when he threw down a ygggna cup after draining it, it would spin for "three minutes." I was intrigued by the fact that the y§_n1hwa1 had located the source of lineage B's afflictions in events that had taken place long before even their eldest members were born and, moreover, offered a genealogical explanation. Later, when I began collecting genealogies, I found that people had a wide knowledge of lateral ties in their own and in their parents' generation, but knowledge of lineal relation- ships was comparatively shallow, and beyond the parental generation, knowledge of collateral ties was even more restricted. As I tried to piece together the relationships between the two lineages involved, I was more able to appreciate the 11111w_a1'_s_ interpretation of the situation. I commented to an informant that the adept was a compara- tively young man to know so much about the past (he appeared to be in his early forties). The informant agreed that he was young, but stressed that the knowledge did not come to the y§_n1_w21_in the usual ways. It came from his ability to see beyond the limits of ordinary men, and ygggng played an important part in the acquisition of that power. 161 In later chapters we will examine more thoroughly the rela- tionships which lay behind this event, but for now one final note about this “ritual of reconciliation." It was mentioned earlier that three 399g; were exchanged, and ygggna_was served from three tanoa. At first I thought that this might be related to the fact that three men of lineage B had died. But several days later, there was another .1.§grg directed by the same gflwn1uwg1g-this one involving two other matagal'. Again three Egbga_were given, and again ygggna_was served from three 12393, On this second occasion, the largest of the three 1anga_was only partially filled. Yaqona was poured from it for the paramount and for the members of the sibling group to whom the 1 ggrg was being offered. I!gggna_from the two smaller 1;ngg_was poured and served to other members of that sibling group's matagali. One inform- ant referred to the first tang; as the yaggna_n1_soro (yaqona of atonement) and the other two as madrali. Capell (1941) lists the meaning of Fmadrali" as "offerings or thanksgivings to the gglgu, not sacrifices for atonement.' But Spencer reports that in Navosa, "the 1 mandrali is certainly not usually, if ever, made as a thanksgiving ceremony; its main function is clearly to propitiate . . .f (1941:13). There an.1,madrali in the form of yaggna is offered in cases of illness believed to be caused by offending the £19! (spirits). Another informant referred to the first 1139; as an 1_vuluvulu. The verbal form "vuluvulu" means to wash one's hands as from a finger bowl after eating. The noun "1_vuluvulu" is the water 162 in which one washes. Thus the ygggng offered to the offended party washed away the wrong which had been commited and made things clean (savasava) again. We could go on indefinitely, describing the various contexts in which ygggng is used, the functions it performs, and the various facets of meaning which it conveys, but we might gain more insight into the power of yaggn§_by considering at greater length a matter touched on earlier. It was mentioned that ygggna_is something which is always shared, and, at least with respect to its licit uses, the act of sharing is normally preceeded by an acts of exéhange. Someone presents the 139923, and someone accepts it. In the majority of cases, the transaction is acknowl edged. by formal speeches, however abbreviated. When one party formally presents yaggna_to another-- as one does, for example, when visiting a kinsman not seen for a long time-~the transaction is known as an 1’sevusevu. In a sense, then, every time ygggna is drunk, an 1_sevusevu has taken place, though this is not made explicit, nor would the term "1_sevusevu? be extended in this manner. Earlier in this discussion the claim was made that the struc- ture of the 1 sevusevu could be taken as a paradigm for all ceremonial exchanges, whether of magiti, tabua, or of other ceremonial goods.f Rather than claiming paradigmatic status for the 1 sevusevu, perhaps it would be better to simply say that there is a structure common to all forms of ceremonial exchange. With only slight differences in introductory formulae, what I will call the composition of ceremonial transactions is the same in all cases. 163 The oratorial formulae associated with all such transactions fall into three types: speeches of presentation (£39931), speeches of acceptance (tari), and a concluding speech delivered by a third party (caucaunake). The object or objects which are handled and exchanged form a fourth conponent of the structure, so that any transaction explicitly involves donors, recipients, witnesses, and the objects exchanged. In any speech of ceremonial presentation, the donor honors the matagali present by including their formal names in the intro- ductory formulae. He also stresses the nature of the tie between the (group he represents and the recipients of the transaction. Speeches of all three types (presentation, acceptance, and benediction) may include references to a veritable litany of relationships--e.g., veitacini (siblingship), veiluveni (the parent/child relationship), veiwekani (people related through women), veivakawai (the children of veiwekani).10 When referring to the object(s) being given, the . speaker describes them, if not in disparaging terms, at least empha- sizing their inadequancy. If, for example, the object being given is a whale's tooth or 1abgg, he may refer to it as dua batina lailai, "a small tooth." The recipients respond by saying EM! ("it is large.") The spokesman for the donors may also ask to be forgiven for anything that might have been forgotten or done incorrectly. To this the recipents may respond with polite reassurances. To emphasize the solidarity that ought to exist between donors and recipients, the speaker might say, "let the string of this 1gbga, bind us together."11 164 There are many factors which the donors might take into con- sideration in choosing a spokesman, including seniority, speaking ability, and individuals' relationships to the recipients. If the occasion is one which could be described as "heavyf (b1b1) or "diffi- cult" (Qfiégié): such as the funeral of a man of high rank, the donors may elicit the assistance of a man of chiefly rank to make the offer- ing on their behalf. According to the cultural model, the recipient ofga ceremonial presentation should be a man of chiefly rank. In fact, many trans- actions are quite informal and, moreover, there may not be any men of chiefly rank present. But on those occasions when a gift is presented by a senior man of chiefly rank, a chiefly man should accept it. If they are themselves commoners, both donors and recipients may invoke traditional ties to chiefly houses in order to secure the services of a man of rank. For example, a man whose younger brother had been visiting wanted to honor him and his wife with a festive meal. The younger brother worked for the government and lived in another province. He had no children of his own, but had adopted his elder brother's younger son and cared for two of his brother's older children while they attended school. That afternoon and evening both brothers were drinking ygggng_at the house of a third man. (This man, an elder of chiefly rank, had just presented a mag111 to his visiting daughter, and a large number of men were gathered there to witness the transaction and drink.) When the elder brother received word that the meal was ready, he invited two young men of chiefly rank to accompany them. 165 Both of these men belonged to the kin group that I referred to earlier as lineage B, a group to which the brothers' house had traditional ties. The affair was a small one limited to the members of the household and a handfull of other men including the elder brother's wife's brothers. When everything was prepared, the elder brother formally presented yggggg to his younger brother. One of the chiefly men accepted it on the younger brother's behalf, and the younger brother then spoke a benediction over it. Since every formal presen- tation requires a counterpresentation, a bucket of ygggga_was made available to the younger brother which he formally presented to his elder brother. It was accepted on the elder brother's behalf by the second chiefly man, and the elder brother than offered a benediction. Ideally, the third speech accompanying a formal transaction should be offered by a ggge, an ”elder." This designation refers, not so much to the speaker's chronological age, though other things being equal, an older man would be more suitable than a younger one. Rather, the term fggge? refers to a man's status as a member of a matagali whose traditional duties toward the chief are those of E132 g1_ygggg_or herald. There is almost something proprietary about the term, and indeed chiefly lines can speak of Your gage," There is an interesting contrast there with the chiefly honorific gone turaga or Hchiefly child." Again, there is no chronological referrent, for the "chiefly child" is likely to be quite an old man. But here, as elsewhere in Fiji, according to the mythic account of the origins of the 166 polity, the chiefly line are newcomers (child-like?) in relation to the "owners of the soil," and it is the latter who provide 9the face of the land" (ga1g_g1_!§ggg). While men of the matagali designated m§1ghg1_!gggg often do offer these concluding speeches, men of other matagali may do so as well. One of the factors which determine who will offer the bene- diction is the nature of the relationship between the possible speak- ers and the principals in the transaction. The concluding speech will often be delivered by the man with the closest ties to the donor or recipient. For example, when a man living and working elsewhere arrived to present his gift at the funeral of an old man of his own matagali, the gift was accepted by his eldest brother and the bene- diction spoken by a third brother. These concluding speeches do several things. Through them ' third parties bear witness to the transactions, and in the process fpraise? those present by referring to them by the formal names of their respective matagali. But most importantly they express the wish that the 1151ge (to use that word in its archaic sense) of the valuables being exchanged and the act of exchange itself, confer health and prosperity on the principals and all others present. For this reason, I have referred to these speeches as benedictions. The following speech is fairTy typical, and I think the text will make clear my reasons for so characterizing them. I am [hereby] proclaiming the chiefly i sevusevu. Chiefly indeed are matagali Nasautoka (etc.). “Yagona of life, yagona of strength, yagona of love. . . . In the chiefly 167 manner Nabena Delanavulele [the formal name of the local segment of the chiefly mata ali Siko], the chiefly Child, the Owner of the Flood [thegparamount chief of lower Matailobau]. I sevusevu of life, i sevusevu of strength. Permanent be oUr_[mUtEaT] love. .ETevatEd—EE the chiefly abode. Let the Faith increase. Mana! §_. . . g1gg1 The meaning of gene has already been discussed. "g g1ngf means fit is true," but “true" in this context is practically synonymous with . ‘gagg, By uttering the closing formula the speaker confers (claims?) [gang upon (for?) the utterance. Some men deliver these speeches with force, their voices rising to a crescendo as they speak the closing formula. Other deliver them softly with downcast eyes, but not always from diffidence or shy- ness, for some men project a quiet intensity as they speak. The ' speech is, after all, a sort of prayer. It is often delivered over yggggg as was the speech presented above, and, of course, all that has been said about the efficacy of words spoken over ygggng_applies to them. But the speech, which calls down blessings on all who wit- ness the transaction it acknowledges, has an efficacy of its own in any case. Ceremonial exchanges may be initiated for many reasons. They occur, for example, at all themajor transitions in the life cycle. And they are socially productive. Prestations may initiate social relationships, effect transitions in existing relationships, or restore them after a dispute. They are also compelling. A request accompanied by the presentation of yggggg or 1gbgg_cannot be lightly <1ismissed. "Women's goodsF such as pandanus mats (1be) and barkcloth (nasj) are also given in ceremonial exchange at weddings, funerals, 168 and other occasions. They are not things of m333_to the same degree as y3g333 and 13333, but they are valued. There is, therefore, pres- tige to be derived from their manufacture and exchange. Moreover, they are emblematic of the productive powers of women.12‘ Unlike 13333, the supply of which is fixed, mats and barkcloth are reproducible. Barkcloth is not manufactured in this area, but the cultivation and processing of pandanus and the weaving of mats are important womanly tasks. So far we have discussed power as something inherent in y3ggg3_(and.t3333), but the very act of exchange is itself productive, efficacious, or a source of power. For in the act of giving and in . the fulfillment of traditional obligations, one pleases the ancestors and secures their blessing. Conversely, failure to fulfill traditional exchange obligations can provoke their wrath. Moreover, according to the cultural model, it is a chief who accepts the ceremonial presta- tion, and in giving to a chief, one is also honoring those whom the chief represents, both human persons (133333) and the ancestral spirits of the group (13). But it is the Herald (3313.31_y3gg3), fthe face of the land," a representative of the autochthones, who calls down the blessings conferred by the act of exchange. Thus, the act of ceremonial prestation is a statement of the ritual division of, labor, a matter which we will begin to explore in the next chapter. FOOTNOTES 1Small packets of locally produced cigarettes and plug tobacco are also acceptable presentations. Both can be bought at the cooperative store,but there are also a number of individuals, including women and girls, who sell cigarettes for a few cents cheaper than the store. 2Fijians recognize several categories of talk which are likely to accompany a ona drinking. The type of anecdotal narrative (veitalanoa) describe a ove is often interspersed with the repartee and banter (veiwali) characteristic: of the interaction of cross cousins. Occasionally, there is a discussion (veivosaki) of a serious tapic (Cf. Arno 1976:60). 3By making this contrast, I do not mean to imply that any instance of yagona drinking is completely devoid of ritual nor that y3gona drinking in a ritual context may not also have a recreational component. 41 was not present on any occasion when a ona was mixed by and for women, and so I do not have any first-hand knowledge of such gatherings. Even though they occur fairly regularly, as a male, my presence would have been inappropriate. 5The same phrase will be repeated later when all the yagona has been drunk and the assembly is about to break up. 6Preparation involved mastication of the root. The chewed pulp was then expectorated and mixed with water. Apparently chewing produced a more potent beverage since some active ingredients insolu- ble in water are released through the emulsification produced by mastication (Steinmatz 1960:24; Gajdusek 1979:122). The chewing of kava seems to have been introduced into Fiji by Tongams (Brewster 1922:18; Lester 1941:226). The more traditional Fijian method involved the use of stone pounders and was reestablished without resistance when colonial officials prohibited the chewing of kava root to minimize the spread of contagious diseases. 7This statement requires some clarification. The reciprocal formulation, veivu oni, refers to the relationship between spouse's parent and chila's spouse. Therefore, it also includes the relation- ship between a woman and her husband's father which is likewise one of 169 170 veidokai. But the relationship between a person of either sex and Uieir spouse's mother, also. one of veivugoni, does not require the same patterns of respect. 8However, the relationship which exists between the two men prior to the establishment of the affinal tie can mitigate the patterns of respect required of veivu oni. For example, a young man of mata ali Nabubuco was related to His wife's father as a cross cousin (tava e prior to his marriage. The behavior expected of veivugoni and veitavaleni are diametrically opposed, and in this case each modifies the other. The respect aspect of the in-law relationship is toned down. The young man can talk directly to his wife's father without the use of a go-between as is typical of veivugoni, but their con- versation must be restricted to veivosaki,or discussion. They must not joke (veiwali) in the manner of cross cousins. They may drink fiagona from the same cup, but they must not engage in banter while aing so. The younger man must be respectful in his compartment. 9As defined here power is not merely an "aspect of social relations [marking] the relative equality of the actors or operating units" (Adams 1975; Cf. Weiner 1976:243). In this discussion, it is defined more broadly as an aspect of thin s as well as person . As an attribute of any entity, power is revealed through interaction with an environment. In that sense, it is relational, but its manifesta- tion need not be either conscious or volitional. The justification for adopting a more general definition of power is that it reflects Fijian thinking on the matter. A swift canoe, a whale's tooth, the war club of a manslayer, and a powerful chief all have something in common; power (mana) is inherent in all of them. The relative control exerted by human actors over elements of the environment (including whale's teeth and other objects charged with mana) is only one form or manifestation of power (Cf. Weiner 1976:243). 10The relationships of veiwekani and veivakawai will be dis- cussed at some length later. 11A tabua is always suspended on a cord of sinnet, hemp rope, pandanus fiber, or some other cordage attached by wire or thread which passes through small holes drilled at either end of the tooth. The tabua is held in the left hand while the cord in held in the right. When the spokesman for the recipients picks it up, he may raise it briefly to his lips before speaking. 12Women may own tabua and present them or yagona in their own name, but men must speak for them. When mats or other women's goods are given, they are always carried by women and, indeed, given by women, but it is men who make the speeches of presentation. Similarly, it is men who make all decisions about the redistribution of all wealth, including mats and other women's goods, received in the con- text of funerals. CHAPTER VI GROUPS AND CATEGORIES In Chapter II we briefly considered the relationship between matagali as jurally constituted groups and specific tracts of land. In subsequent chapters there have been numerous references to the role of matagali in ceremonial exchange, and implicit in much of that discussion was the notion that these are social divisions, with spe- cific ritual functions in a system which has, simultaneously, political, religious, and economic aspects. It will be the task of this chapter to consider the nature of these and other social divisions in greater detail. But as others have pointed out, this is a matter of some difficulty. In concluding his discussion of the terminology he applied to units of social organization on the island of Moala, Sahlins writes that This terminology must suffice in lieu of a received technical classification because, frankly, I do not know of any group- ings or organization precisely comparable to the Moalan. The Oceanic-historical basis of Moalan local society is quite clear. The kin units have points of resemblance to Polynesian ramages, Oceanic nonunilineal-residential kin groups, and Mangaian "stem lineages.“ Whether common descent, co-residence, aspects of rank, the existence of core lines, or other features of Moalan local groups should be used to label them is debatable; indeed, what is decisive at one level is not decisive at another level of local organization (1962:240). At the heart of theproblem is the Fijian concept of "matagali." While the term has .specific referent in the legal code of the Fijian 171 172 Administration, in everyday usage it can be highly ambiguous. Its most general meaning is "category" or "type.“ Thus, one can speak of a particular brand of cigarettes, a style of house, a variety of taro, or any other groupings of similar things as a matagali. It is when the term is applied to groupings or categories of human beings that our problems begin. It becomes difficult sometimes to pin down exactly what is meant when such phrases as fmatagali vata" ("mataqali together," i.e., same matagali) or "dua tale matagal " (Yanother? or "different matagali") are applied to social categories. In literal terms all that is being said is that two social groups or categories areeither equivalent or different. The difficulty comes in pinning down the defining features of the categories being compared and then determining what is meant when they are said to be similar. Are they being defined in terms of customary behavior in general, ritual func- tion in particular, coresidence and cooperation, or descent from a common ancestor? One might think that contextual considerations would dispel any ambiguity, but such is not always the case. Hocart (1952) despaired of ever defining what a matagali 13. Instead he advocated fucusing on what a matagali does, and more spe- cifically, he sought the essence of the concept (as applied to social categories) through etymological considerations. He pointed out that the root fg3l_? means "'having ties with another,‘ generally of vassalage" and can, therefore, be glossed as "subject" or 1"vassal." Thus one village or confederation of villages (13333) subject to another will be known as its gali. "Mata? with the possessive 173 suffix ":33? means the eye of a person or thing. Thus, "matafg17siga" or "eye of the day" is a poetic term for the sun. Through synecdoche "matana" also means the entire face or front of a person or thing. According to Hocart "face" also refers to a sacred spot, a rave, a temple. The obvious conclusion is that a face [mata] is a group of kinsmen with a common sacred place, and a vassal face [mata + qali] is such a cult group which has ties with another cult group. The translation [he concludes] is "vassal shrine" (1952:23). If Hocart were correct, we would expect matagali to be depend- ent on or subject to some other individual, group, or category who do not themselves constitute a matagali. He reports that on Vanua Levu it is said that "The chief is the stem, the matanggali [sic] the branch,“ and on the island of Lakeba the nobility insisted that they did not constitute a matagali (1952:24). On this point it is neces- sary to note that, as applied to the type of social division under discussion here, the term matagali was not universal throughout the group until it was made so by the Native Lands Commission (Cf. Quain 1948:187). And even where it existed, it may have referred to social divisions quite different from the segmentary partrilineal model of the Lands Commission.1 Moreover, the terms for the structural units in the official model became equated with local terms for units organized on a different basis than envisioned in the official model (Cf. Walter 1978a:359). Perhaps the protestations of the Lakeba chiefs can be explained by the resulting confusion.2 In any case, it seems to me that Hocart's analysis fails on another point. His comments deal only with the term Hmatagalif as it is applied to divisions of society. He says nothing about the more 174 general sense of the term (i.e., "category," "type," or fspecies“ of anything.) But clearly the various social groups and categories Fijians call “matagali" are simply a subtype of “matagal " in this more general sense. The root 193311) has another, more fundamental meaning in addition to the one considered by Hocart. As a verb, it means "to twist together" as in rolling fibers of the coconut husk to make sinnet, and when combined with the prefix fvei-," it means intertwined. As a noun, 931133, it refers to a bundle of ten coco- nuts, first tied in pairs using narrow strips of husk. One of the senses of the root m313_listed by Capell (1941) is fa company of . people pursuing a given trade or occupation, a group on guild; 33137 .3333, a choir; 3313:3313, hierarchy of priests . . . a team; 3313: kirikiti, a cricket team." An alternative to Hocart's anlaysis is suggested: a matagali is a group or category of "intertwinedf--that is, interrelated individuals. Perhaps its use in reference to social categories is logically as well as etymologically prior to its appli- cation to other types of categories, but we need not commit ourselves on that point. Nor, indeed, does our analysis hinge on these etymologi- cal speculations. It remains to describe the nature of the "inter- twining" characteristic of the social divisions we have been calling matagali. Despite all that has been said about the ambiguity of the term, if one were to go up to any adult and ask them how many matagali there are in the village, he would receive an unambiguous answer. But if one were to follow that query by asking the same individual, 175 the identity of their own matagali, he or she may require more infor- mation before knowing how to respond. There is a simple explanation for how certainty is possible on the structural level, while at the same time ambiguity may attach to questions about individual affilia- tion. In the context of these questions, the term “matagali? really refers to two types of groups. In the first instance, the meaning of the term would be clear, but in the second, informants may be uncer- tain about which type they are being asked about. While analytically distinct, these two types of groups tend to be equated in practice. On the one hand, the term “matagali? refers to an agnatically constituted descent category. More specifi- cally, in the context of the questions asked by our hypothetical investigator, “matagali? refers to a localized descent group, several generations of siblings linked by their descent in the male line from a putative common ancestor and identified with a particular village. When I say that they are "identified? with a village, I mean that they are considered to be 1 133531, not only native to the village, but also joint owners of some portion of the village lands. I stress that a localized segment of a matagali is identified with a territory because a person need not reside in the village to quality as a member. (Given the rule of virilocal residence and frequent village exogamy many female members of the descent group will reside in other villages as will a handfull of males living uxorilocally. Other mem- bers may live and work in urban centers. All retain their membership in the localized segment of the matagali as do the children of male 176 members resident elsewhere. Membership carries with it certain obli- gations and privileges which members living elsewhere may fulfill and activate to varying degrees. As defined here a localized segment of a descent group is not some sort of hybrid creature based partially on descent and partially on coresidence. Some, but not all, of its members will actually co-reside in the community with which it is identified. It is the sharing of a common identity and cooperation for common purposes which makes these grgggg; it is recruitment through a rule of descent which makes them descent groups; and it is identification with a specific territory which makes them localized descent groups. From time to time it may be necessary to distinguish the coresidential core from other members. From now on these will be referred to as the "resident members? of the descent group. The term "matagal " can also be understood to refer to a coresidential, cooperating group which participates as a unit in ceremonial exchange and contributes to the general maintenance of the village. The core of these groups consist of agnatically related male heads of households, but they may include a number of nonagnatess as well. Given the high incidence of descent group (i-e., matagali) exogamy, the wives of these men are likely to belong to other descent groups, and in-marrying women will account for the majority of non- agnates. But these local groups may also include other nonagnatic members, for example, men related to members of the agnatic core as sisters' husbands or sisters' sons along with their children. Indeed, 177 all residents of the village are assigned to a group even if no actual kin relationship exists between them and the agnatic core. Thus, the families of such functionaries as the headmaster of the district school, school teachers, and the Wesleyan minister are included in one or another of these groups (Cf. Nayacakalou 1978:83). Since the term "matagali" refers indifferently to groups based on both descent and coresidence plus cooperation asking a person the identity of his/her matagali may lead to misunderstanding. The most common response would be to name the group to which he or she belongs by virture of patrilineal descent. But an informant may qualify this with the statement that they Pstay withf some other group, or, in the case of a married woman, she may add as a qualifi- ction the name of her husband's descent group. A number of authors have reported cases of uncertainty on the part of Fijians when questioned about their own matagali affiliation. They have sometimes assumed that their informants' vagueness on this issue stems from what Sahlins has called the "dual system" of Fijian social organization (1962:235). As noted earlier, this dual system is the product of the administrative imposition of a uniform segmen- tary lineage system throughout the group, a model of Fijian society often quite different from previously existing forms of organization. But as Walter (1978) rightly points out, some of this confusion stems from the failure to make the distinction between mataqali qua descent groups and the local groups identified with them. Consider my informantS' statements that up until the compara- tively recent past matagali Nakorowaqa and matagali Nabubuco had been 178 mataqali vata ("matagali together") or 333.33_matagali (one matagali). Yet the two groups maintained traditions of separate origins, claimed different ancestral spirits, and had different totems. Their state- ments were also at variance with the official model of local descent group organization encapsulated in A1_!313“31H§333, the official registry of matagali membership. But the statement that Nakorowaqa and Nabubuco had constituted a single matagali in the past was not really in conflict with the official model on this point, for what my informants' were really saying was that the two group had "stayed together? and functioned as a single unit for the purposes of cere- monital exchange. In a word they had been unified as a local group, but their agnatic cores had remained distinct descent groups. A quarrel in which a man of one group seriously injured a man from the other precipitated a split, and today they function as two separate local groups, but honor their former association and are considered "brothers." The composite nature of mataqali qua local groups is illus- trated by matagali Nabubuco. According to the official registry, Nabubuco is one of two matagali within a higher level descent group, or y31333, of the same name. The other matagali within the y31333, (matagali,Nakorowaiwai, is much the larger of the two, but the name ”Nakorowaiwai" is never used in the village (see Chapter II). The diagram on the next page illustrates the genealogical relationships among some of the living members of these two descent groups (see Figure 3). 179 a, a t m 3:33 fiddle Ma , 3% 3331 180 Matagali Nakorowaiwai and Nabubuco, as constituents of a single y31333, are presumed to share a common agnatic ancestor, but the exact nature of the connection is no longer remembered. What is salient is the affinal tie represented by the union of ego's mother and father. Today the agnatic line known officially as matagali Nabubuco is represented in the village by a single household, that of 1. This household has become attached to the local group headed by Ego who stands in a cross cousin relationship to 1 and his siblings. Despite the fact that, in his eighties, he is one of the oldest men in the village, Ego is still vigorous and takes a keen interest in community affairs. The local group of which he is the titular head Ais perhaps the most cohesive in the village. In addition to the single affinally and (putatively) agnatically linked household of matagali Nabubuco. this local group not only includes Ego's sister's son (2) and his household, by descent a member of matagali Navitilevu, but also the household of a junior patrilineal kinsman of this man (i.e., 3). This cooperating coresidential group is known locally as Fmatagali Nabubuco," presumably taking its name from the y313§3_rather than from the smaller attached matagali of the same name. If one were to ask 1 or any of his siblings, or for that matter any one in the village, they would tell you that 1 is a member of matagali Nabubuco, as he is indeed according to the 31_1313_g1nk3w3, But if you press the matter, you will learn that he, in fact, belongs to a different matagali (i.e., descent group) than the other cluster of kinsmen who call themselves "matagali Nabubuco.Y While he traces 181 descent from the same founding ancestral spirit and claims the same totems, he and his siblings are considered the owners of a separate plot of garden land. Because at sometime in the past his descent group gave a small piece of land to the chiefly matagali to serve as a burial ground, it is known locally as Solia (“to give"). Another indication of their separateness is the fact that 1 and his siblings are t3m313ug1.1313 (lit. "persons of the house") or retainers to one lineage within the chiefly matagali, while the other sibling groups which make up the agnatic core of the local group have traditional ties to another lineage. When people tell you that 1 and his siblings are members of matagali Nabubuco, and then tell you in almost the same breath that this group constitutes a separate matagali within a y3ygg3_which includes other members of fmatagali Nabubuco," they are not contra- dicting themselves nor are they trying to rationalize ggnf3313 organi- zation in terms of the ideal model represented by the official regis- tration. They are using the term "matagal " in two different senses. They are saying, in a word, that 1's household counts itself as a member of a particular local group, but that it is agnatically dis- tinct from the core members of that group. For most purposes, it is local group membership which is salient, but the fact that 1's agnatic line is distinct (through presumably linked) is not forgotten. In some contexts, this difference in descent is significant. Incidentally, the genealogy given in Figure 3 also illus- trates one reason why matagali Nakorowaqa and the agnatic core of 182 Nabubuco (i.e., Ego's line) are considered "brother" matagali. Ego's father and 4's father's father (5) married sisters. As we will see later, the children of sisters are classified as siblings. Thus Ego called 4's father 133333 (“my elder brother"), and the two men were _va_su 1132 ("v_a_§g_ together") to 1'5 patriline.3 Given the tendency for marriages between patrilines to be repeated, it may well be that .there were earlier, now forgotten alliances between both matagali Nakorowaqa and Ego's patriline, on the one hand, and 1's patriline on theother. The marriage between 1's sister and 6, a Nakorowaqa man, could be seen as an expression of this tendency for the descend- ants of cross-sex siblings to repeat alliances. Despite all that was said earlier about the flexibility of the term, if you were to ask anyone from the village how many matagali it contained, you would be told that it was composed of five such groups bound together by political association and kinship. By this it is meant that there are five agnatic descent groups whose village this is considered to be. Resident members of these descent groups provide foci or core structures for an equal number of local groups with economic, political, and ritual functions. The personnel of these local groups include members of about twenty other descent groups, and though some of these matagali have close ties of long standing with the natives of the village, all of them are 1313g1 (strangers) there. As noted earlier,the"true natives" (331 taukei dina) of the area are matagali Nabubuco and, through association, with the former, matagali Nabubuco. A few hundred yards above the village 183 the river narrows considerably, and on the side opposite the village a sheer rock face looms above the river, its summit covered by a tangle of vegetation. Here, near a 13333 known as Naikakau, there is said to be a stone house beneath the waters, the abode of a female spirit. According to one account she is the mother of two sons, the 13 (ancestor spiritS) of matagali Nakorowaqa and Nabubuco. According to another account this female spirit seems to be associated with matagali Nabubuco, for Naikakau is a 13333 (place, land) of that matagali, but her relation to the two founding spirits is unclear. In any case the ancestral spirits of these two matagali are said to be brothers. The y§_of matagali Nakorowaqa is sometimes seen as a snake-like chain of lights winding its way through the gardens at night. Such apparitions are said to be a warning of impending mis- fortune to his human descendants. The 13 of matagali Nabubuco is said to assume human shape, but like his Nakorowaqa counterpart, he has never been a human being. Prior to the coming of the chiefly group, the human descend- ants of these two ancestral spirits are said to have lived in dis- persed homesteads and small hamlets near their gardens and with easy access to fortified positions to which they could retreat when attacked. In composition these ancestral matagali were probably simi- lar to what I have been calling flocal groups,? the constitutent units of the contemporary village. That is, they probably consisted of an agnatic core with some attached members of other descent groups, principally in-marring women and uterine kin. 184 The chiefly group trace their arrival in the immediate area to the exploits of an ancestor named Rokowai. Rokowai seems clearly to have been an historical figure, but his story has taken on mythical proportions. He is said to have been born in the village of Navunida- kua, one of six siblings. At one point Rokowai angered his older brothers by taking their 1331 (plantains) and leaving his trash about. In other words, not only did he take their food without asking, but he did so in a brazen manner without trying to conceal the fact. Such insolence on the part of a younger sibling could not go unpunished, and so his older brothers plotted to kill him. He was warned in time by his mother (were the older siblings perhaps only half-siblings, the children of another woman?) and fled. In anger his brothers set fire to Rokowai's house. (Today that event is said to be commemor- ated in the house name Nakamerusi.) fgggjf is a kind of vine (33), Smilax vitiensis Smilacacene, perhaps used in house construction. #5333? means Fburned.“ The meaning of the name is said to be that the house was only burned to the 3331, not down to the y3!3_or founda- tion.) Rokowai and his retainers fled for a time to a place called Wainileka. Among the followers of Rokowai were the ancestors of what is today matagali Nawaita. They serve the chiefly descendants of Rokowai as their warrior vanguard (3311) and are of the same y313§3_ (see below). Rokowai married a woman from a koro makawa (ancient village) called Navunimoli near the site of the later village of Waidracia. Three children (read fsons") are said to have been born 185 to him in the area of the present village of Taulevu. Two of them stayed there, and one, Ratu Meli Davetanivalu, came to the lands of the present village of Nairukuruku. At this point we enter the his- torical period, for Ratu Meli is probably the chief whom Commodore Goodenough met on his brief visit to Nairukunuku in 1874 (Goodenough 1876:218-219). Ratu Meli is credited with converting to Christianity the people of the imnediate area, a fact comemorated in the frontis- piece of the hymnals used in the parish church and on a monument raised on the y3yg, or foundation, of an earlier church. At this point the composition of the communities in the imme- diate area are said to have changed. That is, the settlement pattern shifted from dispersed homesteads on matagali lands to villages com- posed of several matagali acknowledging the rule of a chief. Two of Ratu Meli's sons are said to have presided over communities composed of the ancestors of matagali Nakorowaqa and Nabubuco respectively. The former are sometimes referred to as §g1_Ulira, that is Fpeople of Ulira,“ an old village site on their lands, and the latter as 331 Naikakau. The village of Nairukuruku is seen as the product of the coming together of these four groups-~the ancestors of the present day matagali Nakorowaqa, Nabubuco, Siko, and Nawaita. The ancestors of the fifth group, matagali Vusovuso, are said to have come later. They had been vanua vata ("place, land together") with matagali Nasautoka, the chiefly line in the village of Navuniyasi, Nairukuruku's nearest neighbor upriver. Their lands lie between the two villages 186 along Wainikovo Creek. A powerful descendant of Rt. Meli's, one Rt. Viliame Robatiratu, known locally as "g3_turaga levu" (Pthe great chief“) is said to have sent a contingent of matagali Vusovuso to the village of Naqara. Naqara had asked for a chief and Rt. Viliame sent them a younger kinsman along with members of matagali Vusovuso to serve as his retainers. Today their descendants residing in Naqara plant their gardens there, for they live too far from Wainikovo to make use of their own lands. In addition to the tracts of garden land registered in their name, each of the five descent group considered native to the village is said to own the creeks which drain their lands.4 Some of these are fished more often than others, and while no one can be denied access to them, members of other matagali must ask permission to do so or demonstrate in some other way a respect for the owner's rights. When a resident 'of' the village dies, it is customary to place a _t3b_u on a stand of taro and either an entire stream or a portion of it. On the hundreth day after the death, the taro will be harvested and the stream fished to provide the food for a large, communal meal, marking the end of formal mourning. A matagali may place a t333_on its own stream, or, in honor of their relationship to the deceased, some other matagalinewroffer theirs for this purpose. Thus, when an elder of matagali Siko died, his classificatory cross counsins of. matagali Nakorowaqa offered to place a 1333_on their creek Wainatibi. In addition to their economic importance, actually quite minor in comparison to the much larger Wainimala, these creeks have 187 another significance, for their environs are said to be the abode of certain female spirits. Those of yavusa Siko have given their name to the creek they live along, Wainiyalewa ("the stream of the women") or simply Nagoneyalewa ("the girsl"). There is a strong strain of dualism in this, for the female spirits always travel in pairs. One is said to be good, and one evil; one is said to be red-skinned, the other dark. They are not named. One elder claimed that the female spirits of his matagali resembled two of his daughters, and the spirits of another matagali were said to resemble two sisters, women of that group. Apparitions by these female spirits, like all direct contacts with the spirit world, are frightening experiences. In all of the instances that were related to me, the female spirits had a normal human appearance. Indeed, sometimes they took on the appear- ance of specific women, but during or after the encounter, something would happen that indicated their true nature, often resulting in an emotional and physical shock to the victim. Brewster recounts such an encounter which occurred in the last century. Then from the body of the assembly a young man . . . modestly observed, "1, Sir, have seen tevoro ["devils," spirits]. I was one of the carriers on the Overland Mail which I took at night between our Government station at Vunidawa and Nambila. It arrived one evening at seven as usual from Suva. The native magistrate gave the bag to me and told me to go on with it, and not to light my torch until clear of the station, as it was forbidden to carry naked lights there. I went on in the dark as far as Ndeleitonga, the next village, and then lit up. When I got to the part of the road where the big precipice overhangs the river, I suddenly felt my body begin to glow and burn all over, and when I reached the great ndakua trees I saw two gigantic goddesses coming along the road one after the other and talking to each other. . . . I was in the most mortal terror and down I fell. 188 I tried to yell but my voice failed me. Then I managed to get on my knees to say my prayers, but words would not come. Then I prayed inwardly three times and the goddesses fled, and I saw them enter the door of the big Mbure at Tau Levu. I staggered past it to Nairukuruku the next village and managed to fling myself into the house of Manasa. Directly he saw me he said this boy has seen a tevoro, and asked me if it were not so, but I would not answer Him. So he made a decoction of lemon leaves and bathed me with it and gave me some medicine, which revived me, and I told him what I had encountered. "Oh, yes," said he, "they were the Alewa Kalou (goddesses) whose haunt is at the great ndakua tree" (1922:212-213). While the places known to be frequented by these spirits are not really avoided, they have to be respected. One man told me that at one time he had a garden quite near the creek frequented by the female spirits of his matagali, and for a long time he was bothered by persistent bad luck. Repeatedly, children born to him and his wife died in infancy, and finally he himself had an illness which medical doctors could not diagnose but which left him virtually bed- ridden. Finally, he consulted a ygug13331, a Fijian curer, and the adept suggested a connection between the patient's misfortunes and‘ the location of his garden. The man stopped gardening there, pro- vided the spirits with a m3g111, and his luck changed. At one point I asked an informant why it was that feamle spirits are associated with streams, and he responded that the 13, the ancestral deities of the various descent groups, who are always male, are localized on high ground above the streams. This contrast between male and female spirits and their haunts suggests New Guinean parallels (Cf. Rappaport 1968), but the informant gave it a very Fijian interpretation. He said that the male 13_is like a 1gr3g3. 189 (chief), and therefore he stays 3.3353 (above). The female spirits are there to serve the 1111393, catching fish and drawing water, and their place is 3Hr3_(below). The informant said that it was not clear whether the male yg_and the female spirits are related and, if so, whether the relationship is one of siblingship (veitacini) or affinity (veiwatini). The health and well being of matagali members and the produc- tivity; of their gardens are affected by their relationship to their ancestors and to their ancestral deity. Their lives and fortunes will be adversely affected by failure to discharge kinship obligations, especially exchange obligations. (What form the retribution will take and the precise nature of the agency involved are not specified. In cases of persistent illness or misfortune supernatural sanctions may be suspected, but the services of an adept are often required to clarify the issue.) Conversely, fulfilling one's obligations is pleasing to both the ancestors and the ancestral deity. In that sense exchange is a creative, life-promoting activity. We have already dis- cussed the fact that apparitions of the 13 are interpreted as a warn- ing to matagali members of impending misfortune. Such apparitions can also signal the 13's displeasure. The ancestral deity of y3yg§3_ Siko sometimes embarks upon or is manifested in the form of wild pigs. (He is also associated with a stone said to resemble a pig located, I believe, near the village of Navunidakua.) At one time plans were made to build a new church of cement blocks. It was a project of considerable expense, and several men from the village worked in 190 New Zealand to earn money especially for that purpose. But some time passed during which construction was at a standstill, and during that time wild pigs increased in numbers and began spoiling gardens quite close to the village. These depredations were interpreted as manifestations of the yfi_of‘y3yg§3_5iko. When the project was resumed, and the original intention was realized, the pigs' forays ceased. In this area each matagali is associated with a species of plant (533) and a species of fresh water fish (1_1313). Matagali of the same 131333 (e.g., Nawaita and Siko) are §3u_1313 ("tree together"). That is, they are associated with the same plant spe- cies just as they share a single y§_or ancestral deity. This rela- tionship between natural species and social "species" is, then, a matter of y3ygg3_membership, but it is talked about as it it were an attribute of matagali. But keep in mind that "matagalif also has the general meaning of "a kind of people." Not all groups that are k3g_ vata are also yavusa vata, but sharing an association with a particu- lar plant species provides a kind of bond between groups (e.g., matagali Nakorowaqa shares its plant totem, the 1331, with matagali Nauluvatu of Taulevu village.) The chart on the following page shows the plants and fish associated with the five descent group considered to be the constitu- ent units of the village. Sahlins reports that in Moala the 53133H!3_(ancestral spirit) of a yavusa might appear to the members of the yavusa in the form of 191 Mataqali Plant Fish Nakorowaqa vesi (a hardwood tree, kanace (Mu ilidae ___7__. __£L______ Intsia bijuga or green- sp. Heart 0 In 1a Yavusa Siko vico (a wild cane, .16 Siko Nawaita Erianthus maximus) - Nabubuco 333 (a tree with fra- ikadroka grant flowers, Fagraea (Dules sp.) berteriana) Vusovuso drala dadarikai (a freusater eel) a particular animal or plant which is spoken of as his vehicle (wagawaga). He goes on to say that "A number of secondary carnate forms are associated with each ancestor spirit and stock, forms of animals or plants: typically a tree, a bird, and a fish for each stock" (1962:228). In Matailobau there is no notion that these plant and animal totems (i.e., of the second type discussed by Sahlins) are carnate forms of the 13, It should be noted, however, that Brewster writes that "the original ancestors of the chiefly families could transform themselves into their animal totems, and resume their human shape when they desired to do so" (1922:105-6). Presumably his remarks refer to the tribes of Colo East, including the Matailobau area. He also mentions that it is forbidden to eat one's own totem, though he notes that there are exceptions, and that, moreover, where it still existed the prohibition was beginning to be ignored (1922: 106). 192 On this matter of totemism Hocart's remarks are very helpful. He draws a distinction between what he calls manumanu kalou or spirit animals and 1 cavu 31 yaca or vuti yaca. The former correspond to what Sahlins refers to as wagawaga or vehicles of the kalou 13, Hocart refers to them as "the vessels [wagawaga] or spirits or ghosts. They may not be eaten by the clan [matagalil or tribe lyavusa] to which they belong" (1914:737). The vuti yaca typically form a set of species which, according to Hocart, in its most com- plete manifestations include a fish, a plant, a variety of yam, a variety of taro, and a variety of banana. Unlike the manumanu kalou the vutiyadha [sic] have no connection with spirits: the natives were most positive on this point [as were my informants], answering in negation even such leading questions as whether the spirit entered them in the negative. They are simply "our“ fish, or ”our" plant, or “our" yam. . . . They belong to the tribe or clan because they were especially assigned to it by their ancestor (vu) or snatched up by him at the general distribution at Nakauvandra [the mythic place of origin of many yavusa]. The vutiyadha 13 in some laces considered the special food gfuthe tribe (1914:738; emp a- sis mine). Hocart‘traces conflicting statements about the nature of Fijian totemism to a failure to distinguish between the manumanu kalou and the vuti yaca or 1_cavu g1_yaca. Part of the confusion stems from the word "totemism" itself. As Hocart (and others) noted, it is both convenient (as a glossifim*many different concepts) and dangerous because it leads to an assumption of identity among the concepts to which it is applied. I suspect that such a confusion may be at work in Brewster's comments on "totemism." Some of his remarks seem to 193 refer to the vuti yaca while others (e.g., the statement that one is forbidden to eat one's totem) seem to refer to the manumanu kalou. Quain reports that in inland Macuata Province on Fiji's second largest island, Vanua Levu, it is believed that violation of the tabu against eating the manumanu kalou will result in illness. But one may experience illness after eating the manumanu kalou of any group to which one can trace a relationship. One learns through a process of trial and error which tabus he or she must observe. 9[T]he individual must frequently suffer from a breach of taboo before he discovers which he must observe. After the illness has appeared, possible breaches will be reviewed in hopes of finding a clue; if a gift presentation to an ancestor removes the disease, the line of inheritance of the taboo is established pragmaticallyf (1948:184). To my knowledge no such ambiguity in the inheritance of food tabus has been reported for inland Viti Levu. It seems to have been the animal form assumed by one's own 13 (i.e., the ancestral deity of one's own y31333) which was forbidden (Brewster 1922:106-7). Therefore,-it was interesting to me that the only case that I recorded in which illness was attributed to the eating of the animal associated with an acestral spirit involved the illness of a young man who had unknowingly eaten the manumanu kalou of his mother's people (but Cf. Brewster 1922:109). The wild pig, a form which the ancestral deity of y313§3 Siko sometimes assumes, is an example of what Hocart calls a manumanu kalou. (So is the snake-like form which the 13 of matagali Nakorowaqa assumes, 194 though in this case the creature corresponds to no natural species.) Today the members of y3gg§3_Siko do not refrain from eating wild pork, though such a tabu may have existed in the past. There is one context in which pork, whether domestic or wild, is forbidden, though this tabu seems unrelated to the fact that wild pigs are asso- ciated with their 13, When they are in villages or 13333_with which they are veibatiki or 3311 to one another (e.g., Serea or Naivucini, both of which are Waimaro villages and therefore ftriballyf distinct from themselves), they are forbidden to eat certain foods, among them 5 Such relationships, sometimes compared to another form of pork. relationship known as veitabani, involve the mutual recognition of food tabus as a prominent feature (Cf. Hocart 1913:106). The two sides or fdivisions" (1333) in the relationship are said to be veitabuki (vei, indicating a reciprocal relationship + 1333, for- bidden): Eating one of the forbidden foods in the presence of the other side may lead to choking. Moreover, the other side may cry out because "their? food has been eaten. (Note that in this case pork would be a food of the other side, not of y313§3_Siko.) ‘13333 and ‘y3ggg3_may then be given to correct the insult. Hocart pointed out that relationships of veitabani and veibatiki correspond in many ways to the 13313 (lit. 113 together?) relationship of coastal peoples which involves joking behavior similar to that expected of cross cousins (1913:106). In the case of the veibatiki relationship whatever amounts of_"yourf food the other side sets before you must be eaten. As one informant said formerly such relationships were bib? (weighting, seri— ouS). but today they are mamada (light, of little consequence). 195 In this area, the plant and fish totems (i.e., 1.331311 or 1113fly333) are not forbidden as food nor (in the case of plants) as a source of materials. It is permissible to eat one's own fish, but I was told that if you eat the fish of another matagali in their presence, you may be doused with water, assuming that your relation- ship with members of that matagali permit joking behavior. Hocart states that, if when visiting one's veitabani, one speaks the name of their 1_331311_the women of the host group are likely to respond by giving the guest a beating (1914:738), and Brewster tells of a coastal Fijian, apparently unaware of local custom, who unknowingly spoke the name of a young girl's totem in her presence and received a slight knife wound for his impertinence (1922:106). Today the names of plant and fish totems are freely used, especially between cross cousins. This amounts to a form of banter, often with ribald overtones. For example, one might greet one's cross cousin in the following manner: "Lewa vesi kauta mai g3_133333,“ That is, "Woman of the 1331 tree (the trees of matagali Nakorowaqa) bring the 133333 (the fish of that matagali) here.9 The fish is, of course, a euphenism for the genitals. (Mothers sometimes jokingly refer to the genitals of small children as their 133333, 13, etc.) I have never heard this particular form of banter between male cross cousins. It does occur between females and between male and female cross cousins. And while it is a form of banter, it is also spoken of a kind of praise. It is so for two reasons. It calls attention to 196 the other party's matagali affiliation, and it amounts to a recogni- tion of the relationship which exists between two people (Cf. Hocart 1914:738 for a passing reference to this usage.) As people of different origins, claiming different ancestral deities, and recognizing an association with different natural spe- ' cies, the five descent groups said to be native to the village are clearly people of different "types." One expression of this differ- ence, perhaps its most basic expression, is that each is considered . to have its proper function (33“33313 1 1311) in a ritual division of labor. Matagali Siko provides the chief, the human focal point of this sacerdotal structure. The man who occupies the office, and to a certain extent the senior heads of other chiefly lines, provide political leadership, direct economic activities, and sit as the living representatives of their chiefly ancestors. By accepting the ceremonial prestations offered him, the chief is performing a function vital to the welfare of the community. His more secular duties often involve listening rather than speaking, for he reflects public opinion in directing public affairs. Ideally he is like a father to his people--stern and aloof but with their interests at heart. We will have more to say about rank and the role of chiefs in a later chapter, but here let us note that the sacred and profane aspects of the chiefly role are inseparable. .Indeed, the distinction is not inherent in the system. As the occupant of the office the chief is personally sacred, and as such his role requires a certain amount of separation. Admittedly, on the surface the separation is very slight. The chief 197 eats the same food as others, plants crops as other men do, lives in the same type of house, and interacts freely with others, albeit as a senior kinsman. The separation which exists is a matter of ritual. On formal occasions, and any assembly at which he is present takes on a formal quality, the chief sits above other men, speaks very little, and assumes a somewhat aloof demeanor. As noted earlier, the role of the chief and that of his herald, or m313fl31_13333, are complementary. The chief is approached through his herald and, in turn, the 3313 speaks on the chief's behalf. As already noted the role of 3313.31_13gg3_is filled by elders of_ matagali Nabubuco and Nakorowaqa. It is said that matagali Nakorowaqa are “people of inside the house" (13m313_31 lomani vale). That is, they handle matters like the distribution of food and other arrange- ments indoors while the same sort of functions are performed outdoors (3H1333) by matagali Nabubuco. In practice, however, the two matagali are interchangeable, and this similarity of ritual function, as well as their prior associaton, justifies their classification as matagali 1313 (lit. "matagali together"). Another aspect of the chief's separation lies in his relation- ship with the bati, mataqaji Nawaita. The latter have the responsi- bility of guarding the chief's person in life. Indeed, they are said to form a human fence (331) around him, and at the chief's funeral they stand guard and prevent the passage of any unauthorized person or animal within the temporarily sacred precinct where the body lies.6 Their elder has the privilege of being the chief's "second? (rabe) 198 when y3gg33_is drunk, and it is they who receive the head (31333), the most sacred part, of any animal presented to the chief. They also have the privilege of filling the office of Tui Rara (Lord of the Village Green), today a lay functionary in the Wesleyan Church, but responsible also for the traditional duty of conducting all village-wide divisions (3333) of 33g111 or ceremonial wealth. When the chief requires the services of the 3311 in a fmatter of the land" (33_vakavanua), or when they wish to approach the chief on such a matter, neither party can do so directly. They must first convey their request to the 113113_31_3313, the Fleader of the 3311,? a position held by a specific line within matagali Siko. In the past this functionary was responsible for "safeguarding the dignity of the chief and for enforcing his commands. The 3311 would do the actual clubbing, but they took their orders from him. When the chief dies, he stands guard along with the 3313, Today the eldest male of this line holds the office of village chairman (113333“31_3333). Much of the planning of village work is done by the chief, but most of the responsibility for supervision falls to the chairman. In some places in Fiji, the 3311 or 9borderf are quite liter- ally groups on the physical periphery of the chiefdom, often of dubi- ous allegiance. Indeed, one of the principle strategies of Fijian ' warfare was to subvert the loyalties of the other side's 3311, With respect to matagali Siko, Nawaita constitute what Hocart calls an “internal border“ (1952:33). Members of the same y31333, Siko and Nawaita stand in the relation of older and younger sibling respectively. 199 At this point a few words should be said about the nature of the y31333, We have spoken of matagali as types or categories of people differing in their descent, in their ritual functions, and, we might say, differing in their natures. A y333§3_is a descent cate- gory composed of a variable number of functionally distinct matagali all linked by putative patrilineal descent from a common founding ancestor (33). Up to now I have avoided using standard anthropologi- cal terminology as glosses for terms such as matagali and y31333, Perhaps at this point it is safe to do so. As I hope this discussion has made clear, the term matagali has various referents. In one sense, it refers to a patrilineal descent category dispersed over a number of villages with members perhaps numbering a hundred persons or more. Though in theory they share descent from a common ancestor, in actual practice, it is not possible to stipulate the exact genealogi- cal relationships among all of its members. Indeed, it may not be possible to specify the exact agnatic relationships among all of the members resident in a single village. In this sense, the closest equivalent for the term matagali is "clan." The y31333_is a more inclusive descent category consisting of a number of linked clans. Following Thompson (1940:32) we can refer to it as a phratry, but it is important to keep in mind the nature of the linkage. The constitu- ent matagali recognize descent from a common ancestor and are con: ceived of as the descendants of brothers. But equally important is the fact that constituent matagali acknowledge a single chiefly line, typically conceived of as the descendants of the elder brother. In 200 a sense the y313§3 is a ritual hierarchy with each matagali having its proper sacerdotal function and with the chief at the pinnacle. Dis- persed over several villages and perhaps even districts, the y313§3_ 333_descent category is not to be equated with a political community, but it may provide a core structure for the chiefdom, as does y31333 Siko for the 13333 of Nabena. Geddes describes the Fijian 1733133 as an "endogamous or exoga- mous? group which theoretically owes its origin to one founder, recog- nizes a single ancestor god, respects the same totems, and acknowledges the leadership of the same chief.7 However, he also notes that some y33333 are composites of several "true" y313§3_in which case common origin, cannon god, and comon totems are absent, and the principle source of integration is common allegiance to a chief (1945:36). As noted earlier, it is not clear here whether he is referring to y33333 as recorded in the records of the Lands Commission or to the type of composite groupings Sahlins refers to as "local yavusa.9 If the former is the case, he presumably has in mind a situation such as the Lands Commission's inclusion of matagali Navoka within y33333. Siko on the grounds that the two "stayed together" in Nairukuruku. Sahlins calls attention to the confusion caused by applying a single term, 131333, to dispersed descent categories on the one hand and to higher order local groupings on the other (1968:240). If Fijians themselves use the term thus, the problem would seem to be inescapable, much as the difficulties surrounding the term matagali appear to be. The ethnographer is left with the task of clarifying the 201 meaning of the term in each instance that it is used. Sahlins uses the term "local yavusa" to refer to higher order local groups, pre- sumably classified as yavusa by his informants, and uses the term ”stock" to refer to descent categories. *The local yavusa is a fusion of lines (and their respective groups) around a pivotal chiefly group. . . . The connection of subordinate kin groups to the chief's group may have been originally predicated on common stock (i.e., avusa ua descent category] derivation, cognate descent of core ines, or other acceptable rationals of "staying together." . . . If any two of the present Moalan villages were to fuse, each would become a local avusa within the new combined village. This is precisely what Happened, tradition tells us, in the history of Naroi [the only Moalan village in which such higher level composite local groups are found] (1962:246). Perhaps it is this sort of situation which prompted Geddes to draw a distinction between "true" and composite yavusa. To my knowledge this difficulty does not arise in the area where the present study was conducted. Certainly it does not arise in the village of Nairukuruku, for there the term yavusa refers only to a higher order descent category. However, the fact that Fijians apparently extend the term to composite local groups of the type described by Sahlins suggests that, in their view, the bonds produced by “staying together," and common allegiance to a chief as both pre- requisite for, and product of, that coresidence, are somehow similar to the bonds of common descent. We have briefly examined the functions performed by four of the five matagali considered to be the major structural units in village organization. The fifth matagali, Vusovuso, is a more recent arrival, and while firmly tied by affinity and cognatic kinship to other groups in the village, it appears to be less integrated 202 than they into the village system of ritual specialization. As pre- viously noted, they are said to have been vanua vata (”land together") with matagali Nasautoka at a place in the district of Wainibuka. It will be remembered that Nasautoka are the chiefly group in the neighboring village of Navuniyasi, and whenever Vusovuso requires the services of a chief, they can call upon Nasautoka. In some con- texts they can be said to "listen to“ (vakarorogo) or "follow" (muri) Nasautoka while in others their current association with matagali Siko takes precedence. One informant referred to Vusovuso as "retainers? (133313 31 1313, lit. 9people of [the] house" or'1333 31 1313, "contents of the house") to Nasautoku. What I have been calling ritual specialization is a necessary concomitant of the division of the population into different "types" of people. The ability to perform one's function is partly a matter of the common substance one shares with other individuals of one's type. A chiefly man or a member of the 3311 are appropriate for their roles because they have inherited the substance (blood, 313) of their fathers. Similarly, it was pointed out to me that a man employed by the government as a carpenter-~he was visiting the village at the time--had no formal training in his craft, nor did he use squares, tape measures, levels or other measuring devices. It was said that he exhibited his skill "naturally" because he was a member of a matagali whose traditional function was to act as carpenters (33131) to their chief. But the transmission of natural substance is bilater31.and both parents contribute to the physical and moral make-up 203 of the child. Thus, when a chief dies, if there are too few 3311) to guard his body properly, persons with 13311_blood" (i.e., persons able to trace an uterine connection to matagali Nawaita) may also be called upon. Similarly, even though they do not inherit chiefly rank or titles, the children of chiefly women married to men of commoner status partake in their mothers' nature as well as their fathers. Men of chiefly rank are expected to be in control of themselves, polite, and exemplary. That is, of course, an ideal not realized equally by all, but it is an ideal that is consciously extoled. From what little I know about the activities of youths (cauravou) there is a tendency for young men of chiefly rank to provide leadership and . exert control over situations. Shortly after I arrived in the village, I accompanied a young man of rank to the nearby government station at Vunidawa. There, near an Indian store, we met three young men from the village. They had purchased a case of beer, most of which had already been consumed, and the three were already quite inebriated. When they began breaking the empty bottles beside the path (potentially quite dangerous since most people do not wear shoes), the situation began to get out of hand. But my companion talked to them and calmed them down. Though he too had been drinking, he never lost control. If a member of the chiefly matagali were to drink too much and get into a fight, his behavior would be disgraceful. But if a member of the 3311 were to do the same thing, it would be understandable if not excusable. While the chiefly role emphasizes control, that of the bati emphasizes action. 204 We have seen that it is necessary to draw a distinction between matagali as dispersed descent categories and the local groups, also known as matagali, which are "built up" around localized segments of these categories. We have gone on to discuss various aspects of these social groupings, but most of this discussion has dealt with attributes of matagali as descent categories. It remains to consider the internal structure of local groups and their functions. According to the official model of Fijian society, matagali are internally divided into named units known as 1_tokatoka. The term {1 tokatoka“ is derived from 11333," fto be placed or situated.9 Thus the nomative 1 tokatoka is taken from Bauan dialect and occurs in other dialects of central Fiji as well, but to the east, in Lau, the internal divisions of matagali are sometimes registered as "3311 31 (13339 (lit. “side or edge of [the] earth oven"). At least in northern Lau the term "bati n_i 1219‘" originally referred to a group of house- holds that cooperated in the construction of an oven and contributed as a unit to feasts (Walter 1978a:358). The term 11 tokatoka“ is familiar to the peOple of the Matai- lobau area, though whether it occurred prior to the introduction of the offical model of descent group segmentation I do not know. The term is not normally used in the village. Indeed, the only time I heard it used was when I introduced it into conversation. And if you were to ask a knowledgeable informant about the number and compo- sition of the 1_tokatoka within a given matagali, you may detect some hesitancy. If you press the matter, the informant will list divisions where they exist, but in time you will learn that in some 205 cases these divisions have little functional relevance. For example, in the early weeks of my fieldwork, I was told that the local group known as matagali Nabubuco was internally divided into two 1_tokatoka, known as Solia and Nabubuco. Solia consisted of a single household, while Nabubuco included six households, plus the two attached house- holds of men of matagali Navitilevu. As already noted, I was also told that the local group which normally went by the name of. matagali Nabubuco was actually a y31333, By this my informants meant that the agnatic cores of "Sofia" and "Nabubuco" were v_O_ 13E ("ancestor together") and 133N3313_("tree [i.e., plant totem] together") but constituted two separate lines of descent. Though these two lines are putatively linked by agnation, the exact nature of the connection is not known. Whether these two lines are called 1 tokatoka or matagali seems to be a matter of context. For example, if the divi- sion is being discussed with respect to land ownership or traditional ties to chiefly lines, Solia and Nabubuco will be called matagali, but if the point of discussion is their status as segments of a single local group, they will be referred to as 1_tokatoka. In point of fact, while this difference of descent is relevant in some matters (e.g., land ownership), in terms of most group activities, the single household of Solia combines with all other households in the matagali. It is not only that this single household is too small a unit to function on its own. It is also the case that the solidarity of the local group is such that there is no reason for it to do so, despite the potential for segmentation. 206 Matagali Nakorowaqa presents a somewhat different situation. In the [Q 1I_o_l_a_ 31 1(333 Nakorowaqa is registered as a Ems-.3 composed of several matagali of which only two are extant, matagali Nakorowaqa and matagali Nakoromatua. While the local group "Nakorowaqa" is some- times said to be "really" a y33333, it functions as a single matagali and the divisions within it may be referred to as 1_tokatoka. The diagram on the next page illustrates the plane of segmentation between these divisions (see Figure 4). The local group known as matagali Nakorowaqa consists of two households living on matagali land across the river (i.e., households C and D) and four others in the village proper. Of these, one is headed by a man from another province married to a Nakorowaqa woman (i.e., housebold B). Technically, he is a 1313g1 (stranger, foreigner) in the village, but his children are 1333 (sister's children) to Nakorowaqa, and the members of his household are, therefore, counted as members of the matagali but with dependent status. The line of cleavage in the matagali lies between the descendants of 1 and 2. Three (3) is the eldest male of the matagali, but his hearing is not good, and though he still gardens occasionally, he is enfeebeld by age. The head of household 0 is the eldest fully active male, but he lives across the river and for that reason does not participate fully in daily village affairs. The male head of household A is the next oldest male matagali member, but he has never married, and thus his status is somewhat ambiguous. His age entitles him to the status of junior elder, but the fact that he is unmarried alligns him with the 207 mamzocoxmz rpemeumz yo oczpuzcum Pacemopemcmu--.¢ oc=m_m c.2932 do: p awwa ousgaaaz .auz usuaeocoxaz canxocoxaz 208 youth (cauravou) and younger men. Since the head of household B is a "stranger" his participation in the management of village affairs is limited.8 Thus a major portion of the task of representing the matagali in the village falls to 4 and 5, the heads of households E and F, respectively. These two mean are realted as parallel cousins (FFBSS), and in this kinship system parallel counsins are equated with siblings. There is only a one-year difference in their ages, but even if the difference were less, one would still be 131333_(elder brother) to the other. The relationship between siblings, especially siblings of the same sex, is always hierarchic. The younger siblings should defer to the elder. The relationship between these two classificatory siblings seems especially sensitive. Perhaps there are individual reasons for this, but there are also structural reasons. While the difference between their ages is slight, as members of the same matagal', they are often placed in situations where that difference matters. Thus on several occasions I have seen the younger of the two men begin to speak in connection with ceremonial exchanges only to be interrupted by the other and publicly shamed for preempting the role of the elder. These two men do not often drink y3gg33_together unless the gathering is large and the crowd diverse enough for them to comfort- ably avoid one another. For whatever reason, they both tend to call on classificatory cross counsins whenever they require the assistance of a larger work force than their own households can provide. While it is true that-classificatory cross cousins and affines would be 209 invited to participate in any cooperative work group, other matagali in the village tend to depend on their own members to a much greater degree. Factors which could counter the centrifugal influence of genealogical distance and the tension induced by slight, yet nonethe- less significant, differences in age include matrilateral and affinal ties to the same kin groups, but as a glance at the genealogical dia- gram illustrates, these factors are absent in the case of 4 and 5.’ As is generally the case both men depend heavily on their wives' extended sibling groups for mutual assistance and companionship. In contrast to the built in strain in the relationship between patrilateral ,parallel cousins, 4's relationship to his matrilateral parallel cousins is comparatively relaxed and companionable. Matrilateral parallel cousins are also equivalent to one's own siblings, but in this case, they are not also members of ego's own matagali, and the opportunities for status rivalry are thereby minimized.9 The case of matagali Siko provides another example of the kinds of cleavages which give rise to internal divisions within matagali. The genealogical structure of matagali Siko is depicted in Figures 5, 6, and 7, but before undertaking an analysis of the infor- mation contained there, I should offer a cautionary note. Generally speaking, it is very difficult to establish the precise nature of collateral ties beyond the grandparental generation. ‘This makes it difficult to establish connections between the agnatic lines of two 210 individuals if they do not share a common lineal ancestor in the last three generations. There may be definite statements that their respective ancestors were related to one another as siblings (veitacini), but the genealogical (or other?) basis for classification is unknown. The genealogical chart depicted in Figure 7 is an attempt to relate two of the smaller agnatic lines within matagali Siko to one of the two major divisions. These lines are significant to the dis- cussion that follows because at least one of their members has occupied chiefly office in the past. The dotted lines are meant to indicate generational equivalence. It is not possible to say that every indi- vidual connected by a dotted line classified everyone else of his generation as a sibling, but there is a notion that in the second and third ascending generations (i.e., at the limits of genealogical knowl- edge) individuals of the same generation within the matagali were veitacini. In addition to the agnatic lines represented in Figures 5-7, matagali Siko includes four other lines accounting for an additional 8 households. Three of these lines are untitled, and it is difficult to establish their relationship to the titled lines within the chiefly matagal‘. The fourth, represented by a single household, is a recent arrival from Taulevu village. The principle division within matagali Siko is between two lines tracing descent from Rt. Meli Davetanivalu (see Figures 5 and .10 As noted earlier, Rt. Meli is accredited with the establishment 6) of Christianity in the area, and whether his marriages to these women represents a polygamous union predating his acceptance of the new 211 faith or successive unions is not clear. The descendants of Rt. Meli and Adi Ivamere constitute the lineage of the current occupant of the chiefly office bearing the title 113 1 taukei31waluvu, "the owner of _ the flood." In the last chapter I referred to this group of agnati- cally related households as "lineage A." At this point I should note that all of the houses in the village are named, and these names are inherited patrilineally. Their names are taken from those of the earthern platforms or foundations (y313) of an agnatic ancestor of the household head and his sibling group. These house names may be used to refer to the houses them- selves, to the households who occupy them, to the entire group of agantically related households bearing that name, or to the head of the senior household in the house group. Referring to Figure 5 house- holds 1, 3, and 4 share a single name. As a glance at the diagram shows, the heads of households 1 and 2 are patrilateral parallel cousins. Despite the fact that the name of household 2 is different from that of households 1, 3, and 4, it is considered a member of the "house group" (my term) or lineage which takes its name from that of household 1. The order in which the various households are numbered on the diagram reflects the seniority of the household head and, hence, the rank of the household within the lineage. '(Households bearing the same name are actually numbered in this way. Thus one can speak of "Dakui #2," “Vuravura #1," or "Vunikura #4.") The head of household #1 is the oldest titled male of the Nairukuruku branch of matagali Siko and the current occupant of the chiefly office. 212 The other main branch within Siko consists of Ratu Meli's descendants by the Navitilevu woman, Ro Naibena. One of the things which I wished to learn from the genealogical data depicted in Figures 5 to 7 was the order of succession to chiefly office. I have indicated occupants of the office by an asterisk and have indi- cated the order of succession by numbering. This was fairly difficult to establish, and, of course, what the diagram represents are inform- ants' recollections which may or may not accurately reflect the actual order, but this in itself is an interesting piece of information, one which we will examine in a later chapter. The most prominent ancestor of of the lineage depicted in Figure 6 (#2 on the chart) is reputed to have been Ratu Meli's imme- diate successor and indeed is said to have preempted some of the older man's powers during Ratu Meli's lifetime. As a young man he played a prominent role in the events that led up to the establishment of Cakobau's rule in the area and was the first government chief (3311) of the district of Matailobau under colonial rule. He was the grand- father of the eldest surviving member of this lineage, but informa- tion about this man's father and, hence, his relationship to Ratu Meli, cam from an elder of another lineage. When I first arrived in the village, I was told that there was but one 1_tokatoka within matagali Siko despite the fact that it was the largest matagali in the village. During the Christmas holiday celebrations of that first year virtually the entire matagali ate their meals in common as is, indeed, customary for matagali to do. .A=<= mmamcwov oxvm Iammmmll Eco ucsuoacum Faupmo—mmcmw--. m oczmpm IUII II 0‘ III! II. |I‘\ || . .- J33. . s \n III III! \ ru>mpwuu>mz . - JEJ. 3334 a}. E II \ r \ exouzmmmz. wpmz .pm excusamez m II‘I‘IIII I I II ~ ~ ‘--W 213 =>m_:mh .oxwm =>mpzeh .ox_m . Jo .mca acme 214 \IOI Aim... $85.: omeF ermmmgmo 23323 23338316 9:5: III III .. u m . .J At... an k\ozocov_mz ,C- - Q % fl . 5...... 3.3262 m a: J “4 .N uFo;om=oz < m we wow: 2 a we sc: 3; a x . me F==>ezF nu Q 3. £vaqu , /. .Fwoxoco/f auszaz 33:2. . 252253: «3:62:52 .0me C 6.; 83 Ill. .1 -3 225E: & Q :0: ...E in .m: 33 215 .Aomchpcoov oxvm vpmmauoz $o ogsuusgpm _muFmo_mmcmwuu.n usamwm \‘i" ‘ //Kuy_ . ___ ”w 4 w 3 , ... / 3:03 uz \\ >H F m . \4 ¢ / \. ...%J fl O 25.2. m .. Ahmv/ 3 225 .3— >_:nmA—oz. . .av a: uanuz :pwwmum oz“ Kohhw nwx / laws «37% (.% n_v ..... , - ‘ a. .583% cc «4 Am pm.» oomv :0: .3. 216 But following the death of the father of the heads of households B, D, and E on New Year's Eve and the subsequent events described in the last chapter (i.e., the i sggg to the chief's lineage) it seemed to me that the structural opposition between these two lines of Ratu Meli's descendants became more prominent. I do not wish to imply that there was any overt breech in relations, for cooperation and socializing between the two lines continued unabated. But the follow- ing Christmas the matagali held two separate celebrations, and the line of cleavage was precisely that discussed here--that is, between these two lines of descent from Ratu Meli. The two smaller lines depicted in Figure 7 generally allign themselves with lineage B. The sister of the heads of households I and II (see Sigure 7) was the wife of a member of lineage B and is the mother of the heads of households B, D, and E in lineage B. The house group composed of households I-IV is large enough and its senior men old enough to have an independent existence, but the head of house- hold i is a young man, and his older, unmarried brother is unassertive. The members of this sibling group call the senior male of lineage B BEQEQQQF ("my father"), and household i functions as a unit within lineage B. For example, prior to attending a funeral, it is customary for members of a matagali to meet (veivosaki) and decide how many tabua, head of cattle, etc., they will present there. In part because of its size, the major divisions within matagali Siko tend to meet I separately prior to the coming together of the entire matagali. The men of household i veivosaki along with lineage B. 217 Another way of looking at the dichotomy within matagali Siko is that it involves a structural opposition between the line of the current office holder on the one hand and all of the other lines whose ancestors have held the office in the past. The three untitled lines do not enter into this opposition, but allign themselves on specific occasions according to such things as affinal ties, propinquity, mutual association in past generations, or village of origin.11 It would be wrong, however, to look at this dichotomy as solely a product of status rivalry. Should the chiefship pass out of the current office holder's line, there maybe a realignment, but to the extent that there is an element of status rivalry present in the rela- tionship between lineage A and lineage B, it is a reflection of an underlying structural opposition between the descendants of women of two different matagali (i.e., the two wives of Ratu Meli). The initial marital alliance between lineage A and matagali Nasautoka of Nasautoka village in the district of Nainibuka has been repeated twice, and one of Rt. Meli's sons married a woman of the neighboring village of Naloto in Wainibuka. The present chief spent much of his childhood and youth in Nasautoka, his mother's village, and married his mother's brother's daughter. Thus he, his children, and his father's father have all been gas! to Nasautoka. Similarly, the other line within lineage A has repeatedly married within their own matagal'. The pattern of marriage within lineage B is more diverse, but there has been a repeti- tion of marriage with matagali Nawaita, and the ties to matagali Dreketi of Navuniyasi (the wife's mother of the head of household C is of that matagali which is also his mother's natal matagali) is also 218 salient. This is only partly due to these recent marriages, for there is a relationship of veigalaki, or mutual ministration, between lineage B and Dreketi.’ The point I wish to make is that over the generations the pattern of marriages contracted by the members of these two lineages has increased rather than counteracted the centrifugal force of descent group segmentation. The sharing of uterine and affinal ties to a third group has the tendency of drawing descent lines together. There is one final point that I wish to make with respect to the genealogical data presented hithese three diagrams. Matagali are neither exogamous nor endogamous; they are agamous. That is, matagali membership does not enter into the regulation of marriage. The funda- mental structural principle underlying the kinship system is a prescrip- tive rule of bilateral cross cousin marriage. The fact that most marriages are contracted between members of different matagali is simply a reflection of the fact that, within one's own matagali and certainly within one's own segment of the matagali, cross kin, or dis- tant parallel kin with whom a cross kin relationship can be traced, are simply lacking. If such relatives are found, intermarriage may occur, and given the tendency for alliances to be repeated in subse- quent generations, a pattern of intramarriage may be established. Let us take a closer look at one such marriage. The geneological diagram presented in Figure 8 illustrates five marraiges beween persons of matagali Siko, three of them between members of the Nairukuruku branch of the matagali. The latter are 219 Siko, Taulevu L Siko, NavunEEkBaT"- 233% 5‘6 ' Siko, Taulevu 3 Ego A J)- 37-% Siko,-' Nairuku ku 5 Figure 8.--Genealogical Diagram Illustrating Matagali Endogamy. 220 technically improper unions (i.e., unions between parallel kin) if the relationship between the persons involved is traced partrilineally. Thus when the relationship is traced agnatically ego and his wife (1) are related as veiluveni (parent-child). As his FFFSSSD, 1 is equiva- lent to ego's brother's daughter (80), a relative whom he calls luyggg_("my child"). But ego pointed out to me that his own parents were related as classificatory siblings prior to their marriage. He said that his parents' marriage was an expression of chiefly duty (privilege?), done to prevent "the blood from spreading." The same may have been true of the marriage between ego's FF and FM. Accord- ing to ego, he referred to his wife (1) as watigu_("my spouse," i.e., marriageable woman, cross cousin) prior to their marriage, tracing the relationship to her through his mother who, in turn, could trace a sibling relationship to 1'5 father. Similarly, ego called 2 tavalggu (”my cross cousin") prior to his own marriage and the marriage of his daughter (4) to 2'5 son (5). Ego's mother and 2's wife (3) called each other "sibling," and, thus, while ego calls 2 “cross cousin" he calls 2's wife tinaqu lada (“my little mother"; i.e, mother's younger sister). There are thus two ways in which ego could classify 5, his daughter's husband. If he traces the relationship to his daughter's husband through 3, the former would be his MyZS and, thus, a tagjga_ (younger brother). But, in fact, ego traces the relationship through 2 whom he called cross cousin. Thus prior to his daughter's marriage he called 5 momo ("mother's brother"). 221 There would seem to be an inconsistency here, for as the diagram in Figure 9 illustrates, the son of a male tavale (cross cousin) is equivalent to a sister's son (vatuvu). A calls C Tavalegu (my cross cousin) A calls B Taci u or Ratukugu (my sister depending on whether she is younger or older than he. .3 A should call D Noqu Vatuvu (my sister's son) Figure 9.--Classification of the M855- Note that in Ego's classification of 5 the term vatuvu (sister's son) and its reciprocal, momo (mother's brother) are inverted. The point is, given a rule of bilateral cross cousin marriage, the relationship between a classificatory vatuvu and mgmg is the same as (tautauvata) that of "child's spouse" to “spouse's father," which in turn fall into a single kinship category, vugg. In that respect the classificatory ‘mgmg (MB) and vatuvu_(ZS) are equivalent; they are both potential affines of opposing generations. Note also that the extension of the term for the mother's brother to MBSS points up an equivalence among ego's male relatives in this very important line, differentiated on the basis of whether they are of a generation equivalent to ego's own or of an opposing generation. 222 Taitai (Makubuna) A'$ Momo Ego Tavale Vatuvu (Momo) -C) Makubuna (Tavale) Ego's MBSS Figure 10.--Classification of Kin in the Maternal Patriline. Ego's M8555 is equivalent to ego's ZSS (= 05 or makubuna), but he might also be classified as a cross cousin (tgyglg).12 The patriline of MF, MB, and M85 are a very important category of relatives to ego; indeed they are his relatives (veiwekani) par excellence. As noted earlier these are ego's true wgka, The term wggg simply means frela- tive," but in the context of death the term takes on the more restricted meaning of "mother's patriline" and by extension includes all of ego's mother's matagali. The marriage of ego's daughter to 5 (see Figure 8), a man related to her most directly as her mother's brother's son (tavale) is considered a completely proper union. He is her mother's brother's 223 son because 1 and 2 are simultaneously matrilateral and patrilateral parallel cousins and thus stand in a sibling relationship to one another. The agnatic tie between 4 (ego's daughter) and her husband is so distant as not to matter. Enough has been said about the intricacies of individual choices among alternate paths of kinship. The general point to be made is that, provided the matagali is large enough to include distant agnates to whom cross relationships can also be traced, marriage with a member of one's own matagali is both possible and proper. A discussion of every context in which matagali membership is relevant or salient would prolong this section unnecessarily. In this and in earlier chapters we have touched on joint ownership of land, the ritual functions appropriate to matagali, the pooling of ritual presta- tions offered at funerals and weddings, cooperation, visiting, and hsopitality among matagali members, and joint ownership of a body of names by matagali. Some of these points will require further discus- sion in later chapters. We began this discussion with a consideration of the etymology of the term "matagal ." Ne suggested that its most basic meaning is I'a group or category of interrelated (intertwined) individuals" and identified the task before us as the description and analysis of that intertwining. We found it useful to make an analytic distinction between two applications of the term matagali. 0n the one hand, it is used to refer to dispersed agnatically constituted descent cate- gories. (It is used :hi a related sense to refer to localized segments 224 of these categories, social units we referred to as localized descent groups.) On the other hand, the term matagali is also used to refer to coresidential, cooperative groups, the members of which participate as a unit in ceremonial exchange and contribute as a group to the general maintenance of the village. Membership in the latter is a matter of "staying together” (tjkgnvata). In making this distinction we implied that descent on the one hand and coresidence plus coopera- tion in the other are distinct principles. They are not only dis- tinguishable as etic categories; the distinction has a basis in Fijian culture as well. Keeping this clearly in mind can help in avoiding one of the difficulties which has confused the understanding of Fijian social organization. But while for the Fijian the sharing of agnatic descent and the sharing of food, place, and ritual obli- gations are distinguishable, they are also similar in that they are both grounds for saying that a group of people are "the same kind" (matagali vata). In the past it may have been possible for the dis- tinction between descent and locality to become blurred over time as the initial bases for particular instances of local group membership (e.g., uterine kinship) underwent reinterpretation in later genera- tions. We will consider this matter more fully in the next chapter. FOOTNOTES--Chapter 6 1In most provinces persons are registered in their father's matagali, conferring upon them immutable legal rights in matagali and But in recognition of variation in traditional principles of recruitment, a degree of choice within a limited range of cognatic descent is permitted in the provinces of Lau, Macuata, and Cakaudrove. 2The ambiguity associated with the term “matagal " seems to be exactly analogous to the structural ambiguity of the Hindi term "jat" or "jati." Like "matagali," "jat" has the general meaning of "type," "species,” or "category, but it also refers to a variety of social divisions variously rendered as "caste,“ “subcaste,” and so on. 3The term "vasu" refers to the position of the sister's child vis-a- vis the mother's mata ali. It involves certain privileges with respect to the property of the latter, but locally this is over- shadowed by another aspect of the relationship. In the context of the vasu 's funeral, the mother's matagali are referred to as the weka, a term which in other contexts re ers to insmen in general. As weka they play a central role in the exchanges associated with the vasu 's burial and mourning. 4As members of the same yavusa, mataqali Siko and Nawaita are co-owners of such a creek. Mata ali which are yavusa vata are said to share a common origin and therefore also share the same ances- tral deity (3g), female spirits, and totems. By treating these as attributes of mata ali, our discussion reflects the way in which the peOple of the area discuss them. 5The word bati (with the possessive suffix -na) means "tooth" and, by extension, the edge or border of a thing. In _this context, it refers to a political and ritual status which we might translate as "people of the border,“ the warrior vanguard of the chief. In the case of veibatiki two political communities are considered bati to one ano er. 6Indeed this suggests one explanation for the use of the term "bati" ("tooth" but more commonly used to refer to the edge of border ofa thing) to refer to the chief' s warrior vanguard. 7The relationship between descent group membership and the regulation of marriage will be discussed later. 225 226 8His position is somewhat different from that of the elder of the two Navitilevu men who "stay with" Nabubuco. While the latter is a sister's son to the senior men of Nabubuco, this man is merely an affine, a sister's husband. Moreover, he is from another province while there has been a long standing association between Navitilevu and Nabubuco. 91n fact, in this case they are members of ego's mother's natal matagali since 4's MZ married a man of her own descent group. 101 should note that the Ai_Vola Di Kawa recognizes but a single 1 tokatoka within matagali Siko. The division discussed here is relevant in some contexts and not others. The term {i tokatoka" is not normally applied to these divisions. They are known by the‘ house names of their senior households. 11In certain contexts their origin in Taulevu village is the operative factor for one line. On such occasions, they follow the titled elder who came to Nairukuruku from that village. 12The kin type MBSSS (which, as noted above, is equivalent to a man's 255) does not appear on any of the genealogies that I collected, but on one of them the term tavale is extended to the FMB who might -' also be calssified as taitai (MF). Now the term tavale is self- reciprocal. That is, if ego calls alter tavale, alter calls ego tavale. Thus in this case alter (FMB) called ego (alter's ZSS) tavalegu ("my tavale"). CHAPTER VII THE WEB OF KINSHIP The preceding chapters have dealt with a variety of topics, among them the use and distribution of land; the production, sharing, and exchange of food; rituals which renew the charter of the politi- cal order; rituals which promote the solidarity of kinsmen while at the same time restating their separation; and descent and coresidence. Each of these can be seen as an integrating factor in community life. But they also define differences between individuals as members of a variety of groups and categories. Understanding the nature of a par- ticular class of these groups and categories is a principal concern of this study. This has required that we unravel the intertwined strands of propinquity and kinship. It remains to consider the structure of the kinship system and its relevance to the problem. A kinship system involves a set of concepts which, singularly or in conjunction, define a number of categories whose members are relationships (e.g., "mother's brother's son') rather than individuals. A kinship system also involves a body of rules which structure the interaction among the members of these categories, and implicit in these rules are a body of values. While concepts and categories and, with a little more difficulty, rules and values can be considered in the abstract, they are realized in the behavior of individual 227 228 actors. In this discussion we will focus primarily on concepts and categories. The behavior of individuals will receive little attention, but paradoxically persons, though admittedly as abstract entities, will be our starting point in the discussion. In a society like this, the problem of understanding what a person is, and how he or she came to be what they are, necessarily involves a consideration of kinship. In articulating the most basic determinants of personhood, we will also be talking about the most fundamental concepts of rela- tedness. - The question of who and what a person is, is inseperable from the facts of group membership. Accordingly, our discussion will begin by returning to a question first considered in the last chapter: what is the nature of the groups known as matagali, and on what basis is one said to be a member? This will lead us to a consideration of Fijian notions of natural substance and the belief that children share in their parents' natures. He will then turn to a consideration of one of the first bestowed and most enduring aspects of personhood, personal names. Finally, we will look very briefly at the process through which the child learns to classify other persons as kinsmen before turning to the system of relationships as a thing in itself. In analyzing the term "matagali" a distinction has been drawn between two types of groupings which overlap in membership, but arenonetheless distinct in terms of the principles on which they are based. The term refers, on the one hand, to local groups as units of village organization, membership in which is based on cooperation and coresidence. But it also refers to social categories defined in 229 terms of patrilineal descent. In the second sense of the term, matagali members may be dispersed over several villages, but within these dispersed categories a number of local segments can be iden- tified. The latter, associated with specific village and lands, may be referred to as localized descent groups, and it is these which constitute the agnatic core around which local groups are formed. Membership in a mataqali qua local group may result from either birth or choices made later in life. With respect to those who are members by birth, patrifiliation is the statistical rule. That is, in cases where a married couple were counted as members of different local groups prior to marriage, it is much more common for them and their children to affiliate with the husband's local group than with the wife's. 0f 44 married men residing in the village, 42 men and their families are counted as members of the same local group to which the fathers of these men were affiliated. Of these, 20 are married to women of other villages and another 17 are married to women who were counted as members of other local groups in the village prior to marriage. Thus, 37 of these men have wives of groups other than‘ their own to which at least the potential for affiliation exists. 0f. the 42 men who are counted as members of the same local group to which their fathers belonged, 40 are also members of the agnatic core of their local group. These last include 1 man of matagali Nabubuco who is cognatically related to the agnatic line that forms the core of that group. While it cannot be demonstrated genealogically, this man and his siblings are also considered to be distant agnates of the core group. 230 The point to be stressed is that, while statistically patri- filiation is the rule, membership in these local groups is mutable for both males and females and not just at marriage, but throughout life. Not only is it commonplace for a woman to become a member of her husband's local group after marriage, but it is also possible for a man to affiliate with his wife's or his mother's brother's local group, or indeed that of any kinsman. Affiliation can entail rights to land for planting and house sites and involves expectations of support and cooperation. But there is this important difference between shifts in affiliation by men and women. For the latter, assuming them to be an outcome of marriage, such shifts conform to the norm of virilocal residence. They are in keeping with cultural expectations, and over the years, a woman's identification with her husband's matagali becomes strengthened by the fact that it is also the matagali of her children. (Assuming that her husband is a member of the agnatic core of the local group, it will be her children's matagali in both senses of the term.) It is usually her own children that a woman turns to for support in her declining years, rather than the children of her male siblings within her natal matagali. Thus women beyond their child-bearing years may think of, and speak of, the matagali of their sons as 'my people." (It would be wrong to overstate this, however. A woman retains rights in her natal matagali, and may return to it for protection and maintenance. Moreover, a woman's natal affiliation becomes salient in certain contexts throughout her life- time.) 231 No comparable shift in identification occurs for men who have affiliated later in life with local groups in which they are not patri- lineally related to the agnatic core. 0f such men and their children it is said that they “stay with" the core members of the local group. "Staying with“ entails the expectation that they will contribute materially to the ceremonial obligations of the local group as well as participate in the network of generalized reciprocity among its mem- bers. It does not entitle them to speak on behalf of the matagali in ceremonial contexts, nor indeed would it be expected that their voice be "big" in other contexts. Such a man may assume a leadership role by virtue of his rank or personality, but he always runs the risk of being reminded that he is, after all, only "staying with" the matagali. Assuming that the host group is that of his wife, the position of this man's son will be more secure, for he will be ya§g_ to that group. It might be argued that this distinction between mataqali qua local group and the descent categories of the same name is unnnecessary --that attached affines and cognates are not full members in groups which are, after all, based on common descent. But in time sister's sons and their descendants may constitute a distinct line within the local group and one which may even be numerically superior to the agnatic core who were originally their hosts. It is not descent from a common ancestor which binds the two lines together, but a relation- ship of connubium. Nonetheless, these two lines, welded together by marriage, with different traditions of origin, different ancestral 232 spirits, and different totems, may continue to function as a single unit for the purposes of ceremonial exchange. Like Nakorowaqa and Nabubuco, they may be spoken of as constituting a single matagali within theéoverall organization of the village despite the retention of different agnatic ties to groups (matagali) outside of the village. In such cases, it seems best to keep in mind the most basic meaning of “matagali"--that is, "kind,9 "type,“ or 'category"--and to understand that social “types" can be based on different criteria. Not only is it possible for individuals and families to become attached to or to separate from local groups, but whole lines may do so as well. But while local group affiliation is subject to change, an individual's membership in a descent group or category is immutable except through adoption. Adoption is infrequent, and the only cases which I encountered effected the transfer of children between closely related agnates or involved illegitimate children who are, in a sense, incomplete persons prior to adoption since they lack paternal kin and are the agnates of no one. While statistical statements are rele- vant to the description of local groups, they are inapplicable to the description of descent groups. Here we are dealing with units whose composition is a matter of ideology, rather than individual choice. All of the children born to a man's wife are considered full members of his descent group, and, except under special conditions to be discussed later, they will be named after other members of that matagali. If a marriage is dissolved through separation, all children old enough to be parted from their mother will remain with the father, 233 for they are of "the same people." The same rule applies in cases where the married couple had been residing uxorilocally. A man's children are not only considered 1 taukei, in the traditional sense, with respect to the lands of their father's matagali, but, assuming they have been registered in Aihygla flihfiéfléa their proprietary status is legally defined as well. Moreover, ritual functions and totemic affiliation are also determined by patrilineal descent, but with respect to the former, an appeal may also be made to complementary affiliation in special circumstances (e.g., if there is an insufficient number of men of mataqali bati, cognatic kin may assist them in guard- ing the body of a deceased chief). In the last chapter we entertained the notion that in the past it may have been possible for the distinction between descent and locality to become blurred over time as the initial bases for particu- lar instances of local group membership underwent reinterpretation in later generations. The question must at least be asked to what extent the tendency for "attached" lines to retain distinct agnatic identities is a product of the g; jug; version of local organization enshrined by the Native Lands Commission. In the past would lines established by sisters' sons have been absorbed by the host group and their relationship to them redefined? Informants doubted that such redefinition would have taken place. It is true that genealogical knowledge is comparatively shallow, but while that may facilitate the absorbtion of nonagantes, the retention of distinct agnatic identities does not require detailed genealogies. What it does require is a 234 commitment to, or concern for, agnation. It is clear that, given the nature of the kinship system, such lines would remain distinct, for their members would stand in a cross-kin relationship to their hosts, and such relationships are likely to be renewed through mar- riages in subsequent generations. Perhaps it would be better to speak of this continued separateness as evidence for a concern with alliance, rather than agantion, but, given the logical structure of the kinship system, the former entails the latter. Still, this pattern of connubium does not rule out the attribution of a distant agnatic relationship between fused lines, for no contradiction is seen to exist in such a situation. Clearly, in this culture the "kind? of person one is, is more a reflection of agnatic descent than it is a matter of local group membership. Now agnation, like descent in general, is a relationship characterized by transitivity. That is, descent is a relationship between a person and his or her ancestor which is first mediated by the relationship between that person and their parent--the male parent in the case of agnatic descent. This mediating relationship, which Fortes calls filiation, results from 'the fact of being the legiti- mate child of one's parentsf (1959:206). Thus filiation would seem to depend upon more fundamental notions of legitimacy and relatedness. With respect to the latter, we would say that filiation is often seen as the outcome of an even more basic relationship which we might call parentage. By 'parentagef I mean a relationship of direct, primary relatedness between persons of opposing generations which is seen as a 235 result of the culturally specific meanings of mating, partuition, and nurturance. Though often related in complex ways, filiation and parentage are not identical, as a consideration of adoption would make clear. The distinction between them is analogous so that between pater/mater and genitor/genitrix. Like filiation, parentage links a person to both parents, though not necessarily in the same way to both nor with the same sig- nificance. It is common for such distinctions to be made in systems of unilineal descent. In such cases they provide a cultural rationale for the contrast between what Fortes has called "filiation? and 'com- plementary filiation' (1969:264). In this local variant of Fijian culture, despite a rule of patrilineal descent, there are no such rationales. That is, there are no clearly articulated beliefs that some component or components of a person's being are transmitted exclusively in the male line though, as we shall see, there are sug- gestive ambiguities in the concept of mana, A child is believed to represent the fusion of the blood (gga) of both genitor and genitrix, but in this area at least, the bilateral transmission of substance is not associated with bilateral descent as Sahlins argues is the case on Moala (1962:168). It is associated, rather, with a system of kinship and marriage in which bilateral cross cousin marriage is a feature of central importance; it is a system whose logical structure is that of direct exchange between partilines. As Levi-Strauss notes, systems of direct exchange, whether manifested in dual organization or bilateral cross cousin marriage, 236 are logically incompatible with ideologies which posit a separate transmission of substance by the paternal and maternal line (1949:393) Levi-Strauss argues that throughout Eurasia there have been systems in which beliefs in a separate transmission of natural substance (e.g., "bone" and ”flesh") are or were associated with systems of indirect or "generalized" exchange. In order to function such systems require at least three groups: Group A gives women to B (but does not take women in return), B gives women to C (but does not, in turn, take women from C), and C gives women to A. In a system in which, hypothetically at least, for generations the males of one line marry the females of the other, and vice versa, the situation will be much different. The diagram below posits the existence of two patrilines linked by direct exchange through bilateral cross cousin marriage (see Figure 11). Assuming that children receive some component of their being-- bone, soul stuff, or whatever--from their fathers, the two patrilines will indeed be differentiated with respect to this factor. But if the mother's contribution to the child's makeup is of a different order from the father's, the two groups will be undifferentiated with respect to this maternally derived component of their natures, for through the exchange of women, they will come to share it equally. But Levi-Strauss maintains that distinctions such as that between relative of bone and flesh are not really based on the contributions of individual parents to their offspring "but on groups, or lineages, the cooperation of which, in and through the marriage alliance, is 237 Figure 11. Direct Exchange. required to form the paired unit which the Gilyak call 2229:: 'those who are born'" (1949:393). In systems ”H1 which marriage is regulated by moiety membership or a rule of bilateral cross cousin marriage, no such distinction between sides is possible. The exchanging lines are indeed a "paried unit," but in one sense, at least, they repre- sent the union of similars, for each'is the source of spouses for the other. As already noted Fijian beliefs concerning conception and the material basis of the parent-child relationship make no consistent, clearly formulated ideological concessions to agnation. Sahlins is correct in maintaining that notions of a bilaterally transmitted 238 natural substance provide an ideological support for bilateral recruitment to the various groups which constitute the cooperative and coresidential units of Fijian society. But it also seems to me that Groves is making a useful distinction when he argues that the groups so constituted are not descent groups, for "descent, however important, is only one among a number of criteria by virture of which membership [in such groups] may be achieved? (1963:288). In a community such as the one in which this study was con- ducted, in which younger members, at least, have been exposed in different degrees to the most fundamental concepts on which the modern theory of inheritance is based, one can expect to encounter various beliefs concerning the determination of the physical makeup of human beings as well as the various behavioral traits generally subsumed under the term "character.§ I once asked an elderly male informant which of a child's parents playsamore determining role in the formation of his or her manners, habits, or disposition. (The Fijian word used, 1_3919, carries all of these meanings.) He thought for a minute and then said that in his opinion the mother plays a greater role because, in the case of male children at least in the early years, she spends much more time with them. In other words, for this informant at least, these aspects of a person's makeup are a matter of nurture rather than- nature. At this point in our conversation there was no reference to blood (gra) or any other natural substance. I then asked a slightly different question. I asked him whether the father or the mother 239 plays a more determining role in shaping the child's physical (ygggggg = body) as well as spiritual and mental (yalgng = mind, soul) makeup. His answer was unexpected. He said that, if the father of the child cares for or "desires" (the word used was vinakata) the mother while she is pregnant, the child will look like him and share his personal qualities. If, on the other hand, the father has not played this nurturing role towards the child in utgrg, the child will resemble the mother. At the same time the informant combined those ideas about the father's influence over the developing child with an appeal to natural substance. He said that the blood which flows in the child's veins is derived from both father and mother. His statements about the mixing of blood led him to observe that,in the past,chiefly men married chiefly women, and their blood remained pure in their descen- dants. Today not only can chiefly men marry low born women, but chiefly women can marry commoners as well. The result is an undesir- able dilution of chiefly mafia, Personal man; (force, power, or efficacy) is thus a matter of the blood, but there are some inconsistencies here. If_children represent a mixing of their parents' blood, it would be reasonable to expect that qualities said to be transmitted by that medium would be inherited from both parents equally. Such is not always the case. One sibling group of chiefly rank, grandchildren of one of the area's first Government trained Native Medical Practitioners, are believed to have special powers (mana) which enable them to cure eye oilments. 240 The utilize herbal extractions in treatment, but it is not these which effect the cure but, rather, the powers of the practitioners. Others using the same herbal remedies would be ineffective, and they themselves have no special powers with respect to other types of ail- ments. A member of this sibling group, a woman married to a commoner, once told my wife and me that her daughter, who was sitting beside her at the time, has not and could not inherit this power, but it will be passed to her brother's children. Along with their chiefly rank, this sibling group's curing powers are transmitted patrilineally.1 Thus beliefs about the transmission of gang, though not systematically formu- lated, would seem to reflect an unilineal ideology. Shortly after a child is born, an elder of the matagali will select a name for it from among the names borne by past and present members of the group. Names are bestowed in pairs. The "first? name is usually biblical; the second name is always Fijian (e.g., fTevita [David] Tabuaf). It is possible for the first name to be taken from one source within the matagali and the second name from another, but in practice the names are treated as a unit. Let me be precise about what fmatagali" means in this context. It is the patrilineal descent group which has the right to name the child regardless of whether the parents of the child fstay with? the father's agnates or a local group based on some other agnatic core. This recycling of names among agnatic kin can make for confusion in genealogical research. The usual way in which Fijians deal with the problem of specifying, for example, which fTaitusi Bici,9 they are 241 referring to, is to affix a number (£393) after the name or to specify whether they are referring to the "big" (i.e., elder) or "little" (i.e., junior) bearer of the name. Or when they wish to distinguish between two living adults sharing the same name, they may simply refer to the names of their first born children, a practice about which we will have more to say presently. In conferring a name upon the newborn, the elder, as the repre- sentative of the matagali, also confers descent group membership and public recognition of the infant's personhood. The child's features will develop and change and its 'soul" (personality) will "harden? (mature), but the name bestowed 'at birth will remain an important attribute of personhood throughout life. At times it seemed to me that there was almost an aspect of personhood about the names them- selves. Several times during interviews informants turned to others present for help in remembering a name. "What is the name of that matagali's Penisoni," they might ask, meaning "what is the full name of all those persons living and dead who have born the name 'Penisoni.'" It was as if they were talking about an identity, a position, that numerous persons have occupied in the past and more will occupy in the future. We are used to titled positions being spoken of in such terms, but not personal names. Once while I was conducting a household survey, an informant asked me if what was being written down would be taken home with me to my own country. When I told him that it would be, he said FAlas! My name is going to America." It was as if a piece of himself were also making the journey, a prospect which both pleased him and made him a trifle uneasy. 242 For it totXLsaid that one knows a person it is not enoguh to recognize their face or even to know something of their character; at a minimum one must also know their name. From time to time during the first months of fieldwork, on occasions that brought the men of several villages together, my hosts would put me through what I came to think of as 'the name game." I would be confronted by some man whom I had met during that undifferentiated whirlwind of experiences and sensa- tions that were the first weeks in the village, and I would be asked by my hosts if I knew his name. The man would smile, I would smile and scratch my head, and my hosts would chuckle. Several things lay behind this little routine. Perhaps one of them was the uncertainty on their parts about the willingness or ability of Europeans to know Fijians as individuals, a product of the colonial experience which brought Europeans to the area but, at the same time, made them struc- turally remote. Perhaps, after all, that was not in question. But what most surely did underlie this was a concern about being known in terms of one's name which, like one's features and carriage, are both public and personal. As a variation on the same theme men would sometimes tease their classificatory cross cousins by telling me their nicknames in their presence. Whether these were "real" or, as I often suspected, made up on the spot, part of their effect was due to the fact that they disguised and even mocked the subject's true identity. Given the recycling of names within the descent group, it is a common occurrence for several persons,. often of approximately the 243 same age, to share the same name. Such persons may refer to their alters as ngguuyaga_(“my name"). There are no special ritual obliga- tions involved, but the relationship with one's namesake, the actual person for whom one is named, tends to be affectively close. Child- ren and young peOple are often referred to and addressed by the term for their namesake's relationship to them. Thus young children may be referred to as tamaga_(“father§), ganeitamana ("father's sister"), tgbgna_("father's father"), or buna_(9father's motherf). Occasionally, children are named for persons who are not mem- bers of their descent group. In order to do this, the child's father should first approach the would-be namesake, present 133%, and ask permission to use the name. In this way cognatically linked lines which 'stay together" may come to share names. Once when visiting a village some distance up river, I was struck by the fact that several persons bore names which also occurred in Nairukuruku. When I asked about this, I was told that ancestors of these people had sought refuge in Nairukuruku during the unsettled period during which Christianity and the sovereignty of the Cakobau government were extended into the area. During their stay they conferred the names of their hosts upon their children. Asking permission before naming a child after a nonagnate is not merely an act of courtesy. It is not only an honor to have a child of a nonagnate named after oneself; there is an expectation that the namesake will be generous towards the child. There are several children and young people in the village who are named after 244 school teachers and ministers, influential strangers who had lived in the village at one time. It is common for their parents and others to refer to them by their namesake's occupations. For example, there are several boys and young men known as '1 talatala" (minister). Young people who are named after agnates may also be known by the titles associated with positions held by their namesakes. There is one young man referred to almost exclusively as ”bull? because his name- sake had occupied the office of district chief in the Colonial admin- istration. Children may even be known by their namesakes' nicknames. There is one young woman who is sometimes referred to as ABakif because that is a nickname acquired by her ggi_(father's sister) for whom she is named. In the local dialect 9251.15 a term for rolling out dough. The young woman's namesake was thus nicknamed because, through choice or her position in the household work force, she fre- quently rolled out dough for £931, a favorite breakfast food. As a child grows older, he or she may acquire a nickname of his or her own. These are often based on personal appearance or some incident in their lives which thus becomes memorialized. But when a person has given birth to, or in the case of males, fathered a legiti- mate child, personal names and nicknames are almost totally supplanted by the practice of tekonymy. Parents are almost always referred to as the father of (tamaj) or mother of (tjgai) their first-born child, their EIEHEEEEE.(11t03 "mature headD). Informants referred to this as a form of respect. It calls attention to the fact that, as a married person, one is a sociological adult, but also to the fact that, as a parent, one is a completed person. So prevalent is the 245 practice of tekonomy that children are said sometimes to be ignorant of adults' personal names, knowing them (other than as kin of a spe- cified category) only as the father or mother of so and so. As a person ages, they come to be known in terms of their grandchildren. Thus a man once known as Wtamai_5evanaia9 comes to be called 'tugu i Vani" (father's father of Vani), after his eldest child's .own M m. The welfare of a newborn infant is directly dependent upon both parents. This is true not only in ways that are for us matters of straightforward physical care, but also in the form of certain observances which, in our terms, have more to do with mystical connec- tions between parent and child, though Fijians would not always make this distinction in the same way we do. When a child is born, its soul, intellect, or animating principle (1319) is said to be "wateryA (yalo‘wai), soft, or weak (malumaluma). As a consequence the infant is believed to be in a vulnerable condition and must be guarded. For twenty days after the birth both parents may refrain from combing their hair lest the child be weakened. During the first few months of its life, the child is kept indoors. The mother is careful to avoid foods which are believed to adversely affect her milk Or to other- wise have an harmful affect on the child (e.g., the mother should not eat bananascw1plantains lest the child suffer an attack of diarrhea.) Women who resume garden work too soon are criticized, not only because it means that the infant must either be taken to the garden and thus be exposed to the heat of the sun, or be separated from the mother and thus not be fed on demand, but also because it is believed 246 that resuming work too soon will adversely affect a woman's milk. This means that the father may have to do some things that he would not normally do and makes both parents dependent upon the assistance of their famale kin, including older daughters. When I think back on all those instances when I witnessed one person deliberately instructing another, I am struck by how often the subject matter was a physical skill or technical operation, often something as simple as the proper angle for an axe to strike a log or how many times a digging stick should be thrust into the soil to pre- pare the ground for taro planting. Unlike some cultures in which, we are told, such skills are learned primarily through observation, this kind of verbal instruction is commonplace. In such contexts the flow of information is always from elder to junior. Thus each instance of instruction involves the communication of information, not just about the subject matter at hand, but also about patterns of authority and reSpect among those who are interacting, patterns which are reasserted and reinforced in the process. While the education of children was not a topic which I focused on, it is my impression that, after what might be called technical skills, the most common subject matters of instruction were the categories, values, and behavioral rules associated with kinship. This is to be expected in a society in which all social interaction is structured in terms of kinship. Even after the infant develops into a toddler and then an older, more independent child, its soul or inner essence remains "weak," and therefore, even though increasing 247 demands are made on it, the child is not expected to behave as an adult would. But even before the child begins to speak, an effort is made to teach it to recognize kinsman, and, as among ourselves, some of the first words that a child learns to speak are kin terms for family members. The way in which the child learns to categorize kin outside of the household and to behave appropriately towards the members of those categories, does not reflect genealogical closeness as much as it does propinquity. That is, the child does not neces- sarily learn to apply kin terms first to those who are genealogically closest, but to those who live closest and with whom the child interacts most frequently. The time has come to consider the kinship system in greater detail. Table 1 lists the kin terms used locally, along with their genealogical referents written in a kin type notation which utilizes the following symbols: M = mother P = parent of either sex F = father B = brother . . , Z = sister 53 3 SIPIIDQ of e1ther sex S = son _ . . D daughter CH - ch1ld of e1ther sex W = wife _ , H = husband Sp - spouse of e1ther sex d = male ego 9 = female ego y = alter younger than ego e = alter elder than ego In some cases alternate forms of a kin term are also listed in Table 1. In these cases, though the term and its alternate differ 248 linguistically, the content of the category to which they may be applied is identical. In all cases, the form most commonly used is listed first. Terms which are otherwise indefinite with respect to the sex of alter may be combined with the prefix £27 to indicate a female alter. In one case, the term raivana (a woman's brother's wife), the re: prefix is obligatory. In Fijian, possession with respect to kinship relationships is indicated by a suffix indicating person and number. In all cases the first person singular possesive form egg, or in some cases, the unbound form nogu,are used here. In every case the kin types listed after the kin terms in Table 1 should be considered a partial enumeration of the genealogi- cal relationships to which the term in question can be properly applied. Since the content of any of these categories is theoretically infinite, these listings of kin types are necessarily incomplete. In actual practice, the outer boundaries of these categories correspond to the limits of recognized kin. Perhaps a brief word should be said about the theoretical assumptions which are implied by this method of presenting the data. It is assumed that the relationship terms listed in Table 1 are kin terms rather than labels in some other form of social classification. In order to rescue this assumption from triviality, it is necessary to specify what is meant here by the phrase Akin term." By defini- tion kin terms will be said to refer to social categories whose mem- bers are relationships based upon notions of genealogical connection (Cf. Scheffler and Lounsbury 1971:71). It is assumed that the only 249 Tabla I.-—kinship Torainology . Primary Gonoration Kin Tan Genealogical _ Partial Contant of Catogory Roforult a" a. m} (oSb). FF FF. m and his gmaalogical aquivalants. m.2 FMZH. no. %m%, leW, nan-.13.- balow for additional antrias) b. noda. FM FM, Flu and har ganaalogical amivalants. at. man; wifa of mg nla 239.- ai' tho sacond asconding gonoration ' c. 5% tug], If it, If! and his guloalogical omivalot. m, RI, Fm.- II My PM, out. d. 993, Hi III, m and har g-Ioalogical omnivalaits, If", FFZ. Fllil; a m- .. 21.9.1 G'1 a. m,‘ tag, F F, m. F3. and his gmoological omivalmts. m. Soil. (W. a 0 h. 5 m. M M. PW. H2 and har gonoalogical amivalmts. raw. Sun. a , whSoFM. atc. 6 c. m a l. I. and his 9 logical oouivalonts, FIN”. $95!, 9590’, . 939R. ate.” d. m i. F2 F2 and hor guoalogical amivalonts. III. Sdl. Sill. Shall. mamas ShSdll. wifo of m.- G° a. % o9. FF dihg), “($19 a(I(ECh). a(I(ISW). d”). our“). a(HZH), m C a. .GSU'.QQSHF.¢OM.3WFo “ma dbl!) (soo ahova for a itional ’kin éypo’rai‘mts) 6:4? ) h. m ySb. 60 )9. ll. mam. {(W), atc. (soo holow for additional kin typo Mar-Its 6- ) c. m m. m, ”IS. 0201. fl-‘ZS, m. m. Illi, ll. (SHE. “IS. .3 Mo .2. 4. mm non/m as. «to. sou. vasu. was. 112. sass.“ am. can. . . mo .. m“ w . u- N f. asm.” 1 HIM. III an. I"! 4 9. m “(nil (DH/ll. (OH. “I. SH. IRS. 41:511. M. II. 92M. CM, BRO. Hl/slii m. 1.125;! a" a. 1m. ca ca. coca. czcn. m. mach. m. cm. nasal. 4850159, «coca. We, sFZSCh, lFZSChSp, thSp, m. cm. 380150. W101, M. “20159. ”56. ate. a. m cusp/(soo)“ also, i601. 1121:». sBCh. acnso.” tZChSo. asm. (IZSChSO, 1m, 9m, atc. c. m yoga dCh cZCh and tho childran of a cl's guloalogical omivolonts, 00501.13 rFZSCh atc. ' a" a. m yso. cscn «sch. 35:11.19 nsscn. nzsscnfi aosu. sow. vasu. uzscn. HZSCh, ZSSH. ZSDW. duh. (Sao holo- for additional kin typo rafaronts Go). a. mag. no.2! 0:11. 31:11.23 1125001. rascal. an. «asset. men. m-m21 ’m. QMMQV a (5059. m9. m. m, 'm'o w'a C“. c. mm 1501 iSCh. am. slSCh. am. sDSDM 11 and tubing ara usad 1:1le by units; Log; tutua and: to ba moi-1c“ to ennui-on. Both tukag an tutua aro a so axtondod to tho aldar sibling. For aaso 1n orasontation tom oro arrangad by tho gonoration o? Efioir kin typo rofarants. Thus. additional kin typos ara listod balo- for tuka tutua (soo 6°). Tho taro 'tuou " Mars in othor dialacts to parants' oothor (o.g.. Moala: Cf. Sahlins l :I and parants' parants and no r's brothar (a.g., Bua provinca: Cf. Quain 1948:285). but in this dialoct it rafars only to tho fathor's fathor (Cf. Rivors 1914:277. 282, 285). Unliko tho othor tons for fathar‘s fathor tuoug is not axtandod to tho oldor sibling. 2m: is tho only fun-1. m typo in tho sacond ascending ganaration to which wu- is axtandad. This would aopoar to bo swat anomalous sinco othar astonsions would load ono to oxooct that Bio FFZ would bo classifiod as SEE. (HI). But sho is a woo-n of ono's own partilina, tho sistor of linaal ralativo who- ono calls aldor sib- ling. Thus sho is ono's ratukana (aldar fnlo sibling). As a consoquonco hor childron(FFZCh) can classify ago (moir FISCh)“ mono (B) or tina—(MZ-H). I know of only ono racordod Fijian tanninology in which tho torm for FF/aSb is «tanoa—to tho FFZ. Caoall and Lastor writo that in Nadrau tho tom tuougg refors to "fathor's fathar. his sistor and hor husband" (1545:191). 3! hava listad taitai as tho prionry torn for mothar's fathar and makubuou as an altomativa form avon though tho mority of indiv-iTu'a'Ts whosa gonaologias I collactad rafarrad to chair maiar‘s fathars as mkuougg. My room for doing so is that informants rofarrad to taitai as tho propar taro for nothor‘s father. but said that Vata is 'Just tho sa-o' (tautauvata). In fact. makuou- is tho raciprocal for taitai/bum (HF/m) and is usad oolas and fulas for thoir Eugntor's child. 250 Tabla I Footnotoso-Continuad ‘For all classificatory tan- a distinction is usually oado bati-om thoso aha aro oldor than PM (to- in) and thoso who on youngor (tan lailai). Tata is an affoctionata fona principally usod by childran in addross though it can also ba uso? as a Era of raforanco. all is only raportod by oldor informants. It occurs in othor dialacts as an affoctionata ton of addrass for nothor (Cf. Sahlins 1962:148; Capoll and Lastor 1945: 184. 191. 197). thong Rivars traats it as if it wora a taro of raforanco for oothar in hadron (1914:275). I kno- of no othor rocordad instancas of its uso for fathar though Haltar (1975:186) raoorts it as a torn of addross for fathor's fathar in malavu. f 5A: with - it glcu‘ustmry‘to satcgyiwhoauor : clas:if:catory tina-‘is oldor olu: {oungo‘than ll/FWS Tho or-r am attortnaggaa. 3soquvaontto§§ nusago. ionau saoabovotho torn 'l ' appoars .bo a torn pocu or to Eh? arao. It is an affactionata fom, priurily Faddross, or . ' ‘A porsu's 59?. Si". and thoir gonoological oouivalonts can also bo rafarrad to as m. Soo bolou (6'1). zSCI inforlaon astond tho tor- ‘oouu' to flair D55 and 'tina lailai' to thoir BSD, ralativos who oro. by ono path of rackoning. gonoologically oauivalont to a mala ago's sgstar's childran (vatuvu .and a f-la ago's oan childron (luvo-). Thus sooo malas axtond tho torn mono ('m') to a rolativa who oig a so bo callad by that torl's rociprocal ('25'), and at last ono fools infomnt classifios as cross kin ('B‘) an altar (FOSS) whu oiyut othonuiso bo rogardod as parallol (ness-mom. Tho basis for thoso utonsions lios in tho antansion of tho taro for oldor sibling to tho FFZ (FFZ - oz and, convarsaly, 965a - sySb. Thorafora. 05C)! - linb). As a rosult of tho astonsion of tho tan for 1a to FOSS. tho oalo odors of via author's patrilina (ono's woka dina) ara oithar Ina-o (nalo cross kin of adiacuit ganorations). tavolo (oala cross kin of ono‘s own ganoration). or nkubuna (aala cross kin of altarnato gonorations). horaovor, Infomts saythat tho ukubuna is 'Just liko' ono's va a. and not only‘do nost infomnts uso Makubuna as an altarnato ton for it, but ono fualo infomnt also u t as an altarnato tan for 105. A sio7i'lar pottorn asists for tho fnla odors of duo oothor's patrilino. aAltoiwuativaly oon any rafor to thoir sistors as " salavolo.‘ I bolioao that this tan is usod aoro framautly for a .n's classificatory sistors than for s actual cross-sis siblings. It appoars to ho a co- ooand of solo ('poth') o volo ('21gaag,‘ 'to run asido'). Poi-hops this is a raforulco to tho avoiaaao that should charactariao tho ralationship of cross-son siblings. Tho ton 'tukausu' and 'tacicsu' ara pracaodod by tho prafix 'go' idion appliod to fnla altars. Rivors reords its uso by oalos for thoir oldor (ratukana) and youungor (ratacina sistors in llodrau (1914:276). but in this arao tho orafix is usod by both alas afi foalas in rafarring ir sistors. Ra- is an honorific which also occurs in tho .lo titlo 'ratuu.‘ It is usually droppod in tho caso of a youagar sTItor, but not than rafar- ring to an oldor sistar. ' 9Tho paronthosis aro usod to indicato that altar is oldor(a)or youngor (y) tho ago. Thos 'o(FKh)' is to bo road 'fathor's brothor's child oldor than ago.“ lo'Tlvflf is usod as an altornotiva tar-for a wasn's f-la cross cousin. u‘l'ho rationalo for this astonsion is providod by tho fact that a woun rofars to hor 858 as m. 'youaugor sibling.“ (Soo balow, 6-2). Tho solo aoplios to tho oat-usion of mum. l'zl'houugh mstcunly usod in raforanca to ono's own soouso. tho tar- wati- can also bo utodod to HZJU. and thoir rociprocals. or to any gvalg of duo opoosito son. 1'3ltar1ua noans 'socond.‘ and in tho Bauan dialoct it as usad to rafar to socondary wivos in a pal housohola. Tho usago hora is a rafloction of tho unity of solo-sax siblings (Honda; H and 16), but a so of tho possibility of loviritic and sororal unions. In tho past ono‘s spousa's sans-sax sibling's spousa my litarally havo boon ono‘s ‘socond" and vico vorsa. Both tho lovirata and sororato wora practicod in tho past. and thougi at ona tioo thoy wora discouragod by sao colonial aainistrators in tho arao bocausa than wora at variants with canon law, tho lovirato and sororato still occur occasionally. “a distinction is drawn batwoon tho ralationship of ms to «you. which is a rospoct ralationship. and that of Hyl to radii, which is lika that of opposita-saa cross cousins. i.o., am and fuiliar with Joking allowod. Indood. tavalo is an altarnato taro for all tho ralationships which my ba classifiod as daku-, and sacondory nrriagos Ethan porsons who stand in a daku ralationship ara ponissiblo. M appliod E a f-la altar. tho taro daku nay bo minad with tho postfis 'fl.‘ ”mainfomntaxtandodthotorl'dak'tohisFMonthogI-unudsthatshousthowifoofa-n(i.a..FF) ww- ho callad 'w oldor sibling“ (tuka . 1“l'ho ton ya can bo usad to rafor to ono‘s soousa's parants and thoir gonoological aauivalonts. Thus, it could ba appl to tho raciorocals of all tho ralationship listod hora. but aach of thoso may also ba callad % ('8') or 291 ('FZ'). To consam spaco only tho kin typo rafarants in tho first dasconding ganoration ara stad. 17lloto that dospito tho fact that iBCh and sChSp aro vu oo, tho kintypo «201 is not includod hora. A man's sistor's child is his m. but infonants acauato vugg- m3 vatuvu. .yugo- is an altarnato taro for «ZCh. 13 but so tho footnota accmonying 61o ontry for mono. 19As a consawonca of tho axtonsion of tho tor- for oldor sibling (tuka-) to FFZ, foolas, as wall as males antand tho tan for youngor sibling to thoir BSCh. 2%“. I do havo gonoological avidonco for tho axtonsion of tho taro 'tacio_u' by a :1an ago to his FBSSCh. I do not hava cmrablo avidanca for a funk ago. But a mn‘s FBS is oouatod with har brothor and. honco, har 3350 should bo oouivalont to hor BSCh. Horaovor, both a foala ago and har FBSSCh would bo motors of tho sano oartilino and, as flora of altarnato ganorations. thoy could only ba ralatod as ratukana/tacina. Again. I do not havo gonoological avidonca for tho axtonsion of tho tons for youngar sibling to a woaan's m. but it is a logical possibility. 251 Tabla I.-Footnotaso-Contimod 21‘I’agi aaans 'to woop or cry.‘ Poi-hops tho significanco hora is that 1291 will aourn ono's doath. 22Qinggg-canalsobousodtorofartolf(.3131). SooG’zabovo. 23It is to bo axpoctad that a oolo ago's 3001 would ba aguatod wiui aga‘s own BCh. but tho fact that a f-ala ago also ant-ids tho taro for 'DCh' to hor BDCh is a dapartura fru tho upoctod pattorn. Sinco a woun rafars to har so as man; ('my child's spousa“) and to bar Bibi as luvm ('oy child'), it would so. to follow tilt a wuan would oauata hor oocn with hor SCh (n diva). Howovar. a waan's soon is vag to ogo's patri- lina and is thus nakubuna to hor. Ho will havo wora to say about this latar. 2‘0». infamnt rafarrod to his FZS. tho child of his 5,; (F2) and bar husband (uno- ho callad vu , as m. In turn, that altar( 's F25) rafarrad to ago as his oakubuna. Ho havo alroady ducouunta tho ant-usonofthattaatothalls soothafootnotoaccwanyiflqthoantryfora). 252 valid reason for using kin type notation in the translation of rela- tionship terms is that the meanings of those terms are ultimately based upon concepts of genealogical relatedness. The suitability of this rather than other forms of translation has to be determined by a consideration of those aspects of the belief system which provide the basis for the relationships so classified. In interpreting the mean- ing of Fijian kin terms in the way implied by their presentation in Table 1 (i.e., as a system of categories whose most basic meanings are genealogical), I am appealing to a set of beliefs which are external to the terminology itself. The beliefs to which I refer are those concerning the nature of relatedness. In the most basic sense rela- . tives (veiwekani) are those who share common substance (353, "blood"). In an earlier chapter we noted how, through analogy, those who live together, and thus derive their substance from the same land, can come to share a similar bond. Certain linguistic usages also provide supporting evidence for the view that the terms listed below are extended from their primary genealogical referents to more distance classificatory kin. Fijians often find it necessary to specify that a particular relation- ship is "true" (dina sara) rather than classificatory. In the case of some terms of reference, this distinction is obligatory. Thus a classificatory father (tama) is either a tamana levu ("big father," i.e., father's elder brother, whether true or classificatory) or tamana lailai ("little father,“ i.e., father's younger brother). In most cases only one's true father would be referred to as 33_tamagu (Amy father"). The common article '33? acts as a linguistic marker. 253 It is true that an analysis limited to genealogical consid- erations results in minimal definitions of kin terms. Because they are so basic to social life, kin terms are especially "resonant." The richness of meaning conveyed through their affective and metaphoric associations will necessarily elude us if we restrict ourselves to genealogical data, but I would maintain that the genealogical frame- work provides thelogic of the system. This does not mean that one can predict on the basis of genealogical data alone which kin term ego will apply to a particular alter. Marriage patterns are such that two persons may be connected by several paths of relatedness, each entail- ing the application of a different kin terms. Genealogical considera- tions do not always determine which of these alternative relationships will be stressed. It will not necessarily be the genealogically clos- est relationship which is selected, although if one is much more direct than the others, it will usually be salient (Cf. Sahlins 1962: 163). This absence of full predictability does not, of course, entail that the kinship categories themselves are not genealogically struc- tured. The kinship terms listed in Table 1 are organized into a number of reciprocal sets as shown in Figure 12. In Fijian these reciprocal sets are denoted by a special form constructed by the addition of the prefix gel: to the kin term followed by the postfix :31, Thus the relationship between mother (3133:) and child (1333;) can be rendered as either veitinani or veiluveni. Each of the terms listed in Table 1 can be thus transformed into a reciprocal term, but, in fact, some of 254 .mumm pauocq_uom ma voupcamco usage sex--.- mczmvm .c:xau.o> a “N3.0 mmzmv axon :wuu3w0> . Asz zmbv axon cFuazpm> n nay -.uaz _co—u>uu.m>1 «xv lean: :n>_ocpo> n “oNKmv -e>»az fi+=opa>aapo> ammzmv ia>_a¢ Fem—o>au_o> u Ammzm + gumzm~iimpa>ah x ........... ------iiuiiiiuiliiiuu--- i- AmNKm + gu~ubv . opa>ah caacaxvm> a ascax Fcpoaupo> oasmx cpum m> n Amway -.uah _ . up Aamov -axsp \\ ow .com=> o> n :u. lama>i+ Agu~bv :>:aa> x ---- .......... ------l--- . :3 MW; LN: 32 + 3.; 98: g A ---mmmmmmmmm---.------i---------- ..... ------uiimmwmu-mwmw-i.-------- wea:.u_o> ..eaEau—o> n i i v i g -2: .. E \\ wcznaxg—m> h mummy m>mn + Chum—um .18v 1 35332 x -- ..... ---------------------------------.-----mm- was: +H~zzv zazmu +i~azv .au_ae aha p: on» o> n Asum .mm» leach . _ A“: as -33 t «we copaapom —a:»=: so» scab mama pauocaeuoz .wawmnua copuacocoa 255 these forms occur more frequently than others. For example, the term veikaruani is grammatically correct, but the relationship between a man and his WZH or a woman and her HBW is more likely to be described as a kind of siblingship. Similarly, the relationship between a woman and her female cross cousins (QFZD/MBD) can be referred to as 1317 raivani but may also be described as cross cousinhood (veitavaleni). In the same manner the relationship between dBW/QHB and 9ZH/dWZ (131: 1333331) might also be referred to as a realtionship of veiwatini. Indeed, all cross relationships of one's own generation may be referred to as veitavaleni or, if one wants to emphasize that the relationship obtains between persons of opposite sex, veiwatini. and contrasted with the relationships between parallel kin of that gen- eration (veitacini). This contrast between cross and parallel recip- rocal sets is made in the first ascending and descending generations as well. While the terms for kin of the second ascending and descend- ing generations are marked for crossness, the contrast between cross ~and parallel is partially submerged when viewed from the perspective of a female ego. .131131 (MF),.3333 (FM), and 3133_(25Ch) are clearly cross relatives, and despite the fact that 333333: does not discrim- inate between cross and parallel kin (i.e., it not only includes dDCh and QBDCh who are cross kin, but also a woman's DCh who are not) informants speak as if it does. They say that veimakubuni are "the same as“ veitavaleni (i.e., they are cross kin) and contrast them with veitacini. But the relationship between a person's MM and a 256 woman's DCh, parallel kin to one another, also falls within the cate- gory of veimakubuni. We will examine this pehnomenon more closely a bit later. The terms listed in Table 1 discriminate along the following dimensions: 1. ‘3337-The terms which distinquish on the basis of the sex of the person referred to are comparatively few in number and are limited to the first and second ascending generations (6+1: ‘13337, tina-, momo, and 331; 6+2: taitai, bubu, and nada). One other term, raiva-, is marked not only for the sex of alter, but also for the sex of the speaker since it refers only to a woman's female cross cousin. Other terms may be combined with the prefix "ra-" (i.e., tuka-, taci-, 13337,.1333r, and 3333337) to indicate a female alter though, in prac- tice, the prefix is consistently employed only with 13537. The suffix "~ni" combined with 3333 is another optional marker for the sex of alter. 2. Generation--With the exception of 13337 (6+1, 6'1) and . '13337[1331¢(G°, G+2/G'2) each of the terms in Table 1 "places? the .person referred to in one of five generations, or so it appears if one limits investigation to primary genealogical referents. In fact, the terms 133339 and 533333379 can be extended to alters of generations other than that of their primary genealogical referents in ways analo- gous to the extension of the terms tuka-/taci-. We will look at this phenomenon more closely in a moment, but here we might consider one implication of this extension of kin terms to kin type relationships 257 of more than one generation. For example, the fact that the term "momo" (MB) can be extended to the M855 suggests an equivalence of the first ascending and descending generations. The notational sys- +1, Go, etc.) tem which we have used to indicate generation (6+2, G has treated as a unitary phenomenon what are in reality two separable dimensions, generational distance (D, 1, 2) and generational direc- tion or seniority in relation to ego (+/-). The fact that the latter distinction is neutralized with respect to certain terms (i.e., 3333) points up the fact that generational distance is the more fun- damental notion.2 3. Relative Age--The terminology bifurcates the universe of kinsman of ego's own generation into 33333 kin relationships which may be lumped under the reciprocal form veitavaleni, and parallel kin or veitacini. Veitacini proper are in turn bifurcated into those who are elder than ego (1313:) and those who are younger-(1331:). This terminological division on the basis of relative age parallels the normative distinction between those same-sex siblings to whom ego owes respect (1313:) and those over whom ego has authority (1331:). This terminological divsion among parallel kin of ego's own generation occurs also among ego's parallel kin of the parental generation, though in the first ascending generation the distinction is not inherent in the terms themselves, but is made through the use of the modifiers 113131 ("bigi or Helder") and "131131? or 513331 (Alittle," “short,9 or "youngeri). Of course, the dimension of real- tive age which obtains in the first ascending generation is distinct 258 from, though similar to, the dimension operative in ego's own genera- tion. In the parental generation parallel kin are classified as elder or younger in relation to ego's parent with whom they are lumped. 4. Cross vs. Parallel--As the arrangement of the chart in Figure 12 implies, this distinction is fundamental to the classifi- cation of kin in every generation. Indeed, the fundamental quality of this dimension makes it difficult to define it succinctly. As Trautmann has observed with respect to Dravidian terminologies (with which this Fijian terminology shares many formal properties), an explanation of crossness famounts to an explanation of the Dravidian terminology itself? (1981:47). At the heart of the matter is the distinction within ego's own generation between those kinsmen who are related to ego as siblings or the children of his or her parents' same-sex siblings (parallel cousins, i.e., MZCh and FBCh) and their genealogical equivalents, and those who are related to ego as the children of his or her parents' opposite-sex siblings (cross cousins, i.e., MBCh and FZCh) and their equivalents. The latter are potential affines; the former may not be. Of course, the problem lies in spe- cifying what these genealogical equivalences are, but such an analy- sis goes beyond the goals of this discussion. 5. ‘Qggrees of Affinity--This dimension is given expression in four of the terms listed in Table 1, wati-, vugo-, 3313, and 13333, Three degrees of affinity are distinguished: first degree (i.e., wati:, H/W); second degree (i.e., 33g_-, SpP/ChSp and 3313, dBW/QHB and QZH/dWZ); and third degree (i.e., karua WZH and HBW). 259 This dimension implies the existence of a more basic contrast, that between consanquinity and affinity. But if there is a relation- ship between the terminological system and the prescriptive rule of‘ bilateral cross cousin marriage, the essence of the system would appear to lie in the negation of his contrast.3 Strictly speaking, this is not true. It would be more correct to say that the opposition 'between consanguinity and affinity is expressed in terms of the con- trast between cross and parallel kin. If we consider the more inclu- sive categories labeled by the reciprocal forms listed in Figure 12, we see that the terms for affinal relationships are identified with those for cross kin. Yet the existence of special terms for fspouse,“ "spouse's parent"/"child's spouse,“ 95pouse's same-sex sibling's spouse," and so on, suggest that, even though the concept of affinity can be expressed in terms of the opposition between cross and parallel kin, it is a seperable dimension of semantic contrast. Before offering a structural overview of the terminology, I want to take a brief look at a phenomenon referred to earlier: the extension of certain kin terms to kin type relationships belonging to more than one generation. Most of my informants had never known their great-grand- parents, did not know their names, and some were unsure as to how they should classify them. Some said that their kin of the third ascending generation would simply be 33_("ancestor"). The knowledge of those who did know the identity of their great grandparental kin was limited to their father's grandparents. In classifying those kin, 260 my informants employed terms whose primary genealogical referents belonged to the parental generation. The diagram in Figure 13 illus- trates the pattern which emerged. With respect to the relationships illustrated in Figure 13, the terminology recognizes only three generations: ego's own (Go), the first ascending (6+1), and the first descending (G'l) generations. If we were to conceive of generational distance in spatial terms, we would say that the terminology recognizes only two positions: 0 units removed from ego (i.e., ego's own generation) and 1 unit removed from ego (see Figure 14). Among those relationships 0 units removed from ego some are classified as senior (1313:) and others as junior (1333:). Similarly among those relationships which are 1 unit removed from ego in generational space, some are classified as senior (tama-ltina-), others as junior'(1333:). Of course, Figure 13 is incomplete since, with the exception of ego's FM, FMF, and FMM, it does not illustrate ego's classification of cross kin. That deficiency is remedied by Figure 15. The pattern of extending terms beyond the generation of their primary genealogical referents is not as well developed in the case of cross kin. Nonetheless, informants' tendency to extend the term f331333:" (“DCh") to their MF and the stated equivalence of 331333: and 133313: can be taken as an indication that cross kin of the second ascending and descending generations are viewed as equiva- lent to the cross kin of ego's own generation. Similarly, the use of momo as an alternate term for the M835 and the predicted extension 261 ( -‘-~_— m-‘m— _—_ .... C 7? ID I 2 m. G. OI I 1 l I" : l l 5___ _ l _ ___.___ i i.-. Tama- fi) Tina- fl , :71 E90 [:LjTuka-lTaci r°r ‘ l L a"; Luve- (E L. ..----“ 3— ..g 8 undo I o— -. A g Figure 13.--Terminological Equivalences of Alternate Generations. P--- Tama-lTina- Tuka- 262 -1 Luve- Taci- ‘ Junior Senior Figure 14.--Generational Distance. 263 """"W I O E s 2. —l> —o ’2‘ £3 1... ____________________________ .--. 3 (Makubu-) Bubu E i Taita1 5 i E :- ----------------- 4 ——————————— -- I I E l 0 1’. ("t Momo Nei 3 : Nei : s i ;_ ........... _-_-_---_--_..__.----- i E i i ‘ 2L in. g ' 1:1 Ego (I)? é Tavale- : l 1------ ————————————————— .—r . . (Momo) (T1na-) : : Vatuvu Vatuvu I : . : .... 1: (I) -i Vatuvu (g " a) ' . . I I I l I L-o - Makubu- (BA Makubu- A) A Tavale- E Figure 15.--Equivalence of Cross Kin of Alternate Generations. 264 of the terms 3333 (MB) and 333_(§1) to MFF and MFM respectively (in keeping with the extension of those terms to FMF and FMM) suggest the structural equivalence of cross kin of third and first ascending and first descending generations. As striking as this equivalence of alternate generations is, the most important structural feature of the terminology is the con- trast between cross and parallel kin. This aspect of the terminologi- cal system is paralleled by contrasting rules for behavior towards cross and parallel kin of the three central generations (i.e., first ascending, ego's own, and first descending generations). The most important of these rules is that which prescribes bilateral cross cousin marriage. People say “we marry our 133313." Indeed, opposite sex cross cousins may refer to one another as 331133, “my (classifi- catory) spouse" whether or not they ever marry. People speak of the appropriate nature of a mother's request for the hand of a brother's daughter on her son's behalf. In fact, however, I know of only one instance in which a man married his actual MBD and one other case in which a man married his FFBDD. Since a man and his FBD are considered "true" siblings, veitacini dina, informants equated these two marri- ages as examples of marriages between veitavaleni dina.4 The fact that very few people marry the child of their mother's brother or father's sister does not lessen the validity of their.claim that they marry their 133313, for marriage with a more distantly related cross cousin satisfies the rule equally well. Nor is their generalization (which is really a normative statement and 265 not a statistical generalization at all) invalidated by the existence of occasional marriages between distantly related parallel kin, persons related as veiluveni or veitacini. In most such cases persons are able to trace an alternate, cross kin relationship which justi- fied their marriage, and marriages involving such manipulations may be considered testimony to the strength of the rule, rather than breaches of it. Nor do the small number of marriages between persons who considered themselves nonkin prior to marriage constitute excep- tions to the rule, for in such cases spouses are considered cross kin by marriage. The terminology is an integral part of this system of kinship and marriage structured by the rule of bilateral cross cousin marriage. When reviewed from the perspective of a male ego, it functions 33_11_ the universe of kinsmen consisted of just two exogamous patrilines bound in connubuim (see Figure 16 below). Figure 16 represents a highly idealized view of the system. Neither patrilineal nor matrilineal moieties exist in this area of Fiji, and ego's kin will never fall into just his patriline or one other. But there is a strong tendency toward dualism in Fijian social organization, a characteristic manifested in the ceremonial exchanges which mark the major transitions in ego's life cycle. For ego the most important opposition is that between his or her own patriline and the patriline of ego's mother, an opposition which is given full- est expression in funeary rites. But while Figure 16 represents a highly simplified model of the system, the pattern of direct exchange 266 Tamaqu A Nei Momo Tinaqu Nada Tamaqu 3| Nei Momo Tinaqu Rataciqui Tavalequ Ego (l Tavalequ/Watiqu Luvequ ‘1 Vatuvu/Vugoqu Taci qu I ‘ Makubuqu Luvequ + Vatuvu/Vugoqu Figure 16.--Classification of Kin by a Male Ego. 267 which it depicts is sometimes approximated. There is a tendency for marriages between patrilines to be repeated in subsequent generations, though the initial alliance will not be repeated in every marriage and perhaps not even in every generation. Still the bond of connubium can provide the justification for the "staying together“ of two patrilines. The result may be a "composite" local group such as that formed by matagali Nakorawaiwai, Nabubuco (Solia), and Nakorowaqa discussed in the last chapter. Such groups are bound together by both sharing (as cognates) and exchange (as affines). When viewed from the perspective of a female ego, however, the classification of kin is more complex than the situation depicted in Figure 16. The difference arises in the classification of kin in the descending generations. It is as if a woman classified her brother's descendants in the second descending generation from her brother's perspective and at least some of her own descendants of that generation from the perspective of her husband. The diagram in Figure 17 illustrates the facts of the matter. Ego refers to 1 and 2 (her BSCh) as 1331g3_("my younger sib- ling"); they, in turn, refer to ego as ratukagu (e2). Ego equates 1's wife with her brother's wife (her raivana) but classifies the husband of 2 as "my son's son“ (3333_3133). Ego could also extend the same term to the wife of 1, for given the rule of bilateral cross cousin marriage, her BSChSp is equivalent to her SCh (3333 3133). Ego refers to 3 and 4 (her BDCh) as makubugu, a term which we have glossed as "daughter's child." They, in turn, refer to ego, 268 their MFZ, as 3333. There would appear to be an inconsistency here. A woman's daughter's children, also 331333:, are parallel kin to her, but in the case of 3 and 4 ego is extending the same term she uses for a category of parallel kin to alters (her BDCh) who stand in a cross kin relationship to her. They, in turn, refer to her with a term (3333) reserved for female cross kin of the second ascending generation. The inconsistency is only apparent, however. Informants consider makubuna to be a category of cross kin. It can be used to refer not only to a male ego's DCh and a female ego's BDCh, but also to a person's MF (131131). All of these are cross kin, and informants contrast them with parallel kin of the second ascending and descending generation (tuka-ltaci-). Ego's classification of 3 and 4 as makubugu is consistent with informants' understanding of the term as 'a label for a category of cross kin. Moreover, her classification of 3 and 4 parallels her brother's classification, for he also refers to them as makubugu. It might be expected that ego would extend the term for fson's child" (3333:3133) to 3 and 4, for she refers to their father as "my child? (33 133333), and ego's own son would refer to 3 and 4 as 133333. Like makubuqu, noqu diva is a term for cross kin of the second descending generation. What seems to be operative here is ego's status as a member of her brother's, brother's son's, and brother's son's son's patriline.. As such, 3 and 4 are 3333 ("sister's child") to her just as they are to the male members of the descent group. I should note that my information concerning a female ego's classification of her BDCh (3 and 4 in Figure 17) was obtained, not 269 .omm m_asmu a an :_x we copuau_ermma_uii.fifl meampm -saaxaz -sasxa: as a m . .M. fl io>34 -om=> +1omz>+ 1m>24 +o>=4+ uwumh @w We a>wo ianzxmz- m swumh +10>34+ m>po a: : . - swoop. . ......>=..m_V -o :> cam imxah Mu ease iwuah. +-om=>+ 10>:4 .im>wmm 270 through the use of the genealogical method, since that particular kin type relationship does not appear on the genealogies of any of my female informants, but by eliciting kin terms for kin type rela- tionships on a hypothetical genealogy. This classification is sup- ported by an observation made by my wife. Having heard the term {131131.("MF“) applied to a female alter on several occasions, she inquired as to the basis for this extension. She was told that the woman in question was the sister of ego's 1311315 normally alter would be referred to as 3333_("father's mother'), but as 131131fs sister, her relationship to ego is similar to his. Conversely, ego's relationship to her MFZ is similar to ego's relationship to that woman's brother (MF), a man who calls ego makubugu. This extension of 131131 to MFZ is analogous to the extension of 131333 (eSb, FF) to FFZ. It is ego's classification of her own daughter's children that appears problematic. As previously noted, a woman's DCh are parallel kin to her, and yet she refers to them by the same term as her hus- band, for whom they are cross kin. From the perspective of a male ego, the term makubuna is consistently applied to cross kin in the second descending generation and contrasts with 133133, which is con- sistently applied to parallel kin. For a female ego, however, the term makubugu does not discriminate between cross and parallel kin in the second descending generation. That is, it is applied to both cross (e.g., BDCh) and parallel (e.g., DCh) kin of that generation. For a female ego, the terms "nogu diva" and “ acigu" do discriminate 271 between cross and parallel kin, but unlike the terms employed by a male ego, for whom all kinsmen of the second descending generation are either 1331g3_or makubggu, noqu divu and 1331g3_are not exhaus- tive. As Figure 17 illustrates, the pivotal position of women is expressed through the kinship terminology. In classifying kinsmen of the second descending generation, it is as if a woman were poised between the partriline into which she is born and the patriline into which she is married. In classifying her brother's children's children, it is her patrilineal affiliation which is salient; alter is either 133133 (parallel) or makubugu (cross). In classifying her own child- ren's children, it is her affiliation with her husband's line which is salient; alter is either 3333 3133_(cross) or makubugu (parallel).5 Thus, for a female ego, a makubuna of the second descending generation is 33§3_to either her natal or affinal patrilines or their structural equivalents. For a male ego no such dual perspective exists. As mediators between the patrilines of their brothers and their husbands, women, and the relationships traced through them, are referred to as paths (1313). Relationships through men establish lines of vertical cleavage separating the line of the older brother from that of the younger with implications which we will discuss in the next chapter. Relationships traced through women, on the other hand, establish horizontal bonds between patrilines. The former are expressed through sharing and cooperation, but also through status rivalry and competition. The latter are expressed through exchange but also involve affective warmth and support. 272 In discussing relationships traced through women, it is neces- sary to define several terms which are used locally to refer to cate- gories of cognatic kin. While they take their referents form the domain of kinship, they are not, strictly speaking, kinship terms. In this respect they are comparable to such English constructions as, say, "relative on my mother's side,“ but unlike such phrases, two of them, 1333 and _v_a_s_u_ , may be used in address as well as reference so that it would perhaps be best not to press the matter of their exclusion from the list of kin terms. The 3333 relationship is familiar to anyone acquainted with the literature on Fiji. Indeed, its more dramatic expressions have been the object of speculation since they first came to the attention of European observors. The 3333 is a man's sister's child (331333) and, more particularly, his sister's son, but the term can be extended to the children of any women who are natal members of a man's matagali. It may also be extended to the matagali.of the sister's child as a unit or to matagali into which women of previous generations have married. I have even heard it used to refer to men of villages into which women of a matagali have married, irrespective of those men's matagali affiliations. Thus, in the most restricted sense, the 3333_ relationship exists between mother's brother and sister's son, but is extended to include the relationship between a woman's child and her entire natal matagali and is subject to other extensions as well. Much has been written about the privilege of the Fijian sis- ter's son with respect to the property of the mother's brother or, in 273 the case of chiefly 3333, with respect to entire villages. In-some areas the right to seize the property of the mother's brother is limited to the context of ceremonial exchange, a context in which the property offered to the mother's brother's group is, ultimately, being offered to their ancestral deity. Thus, the sister's son is in effect stealing the items of scarifice and may pay for his depredations with a beating at the hands of his mother's brother's sons. Such seizures are certainly the most dramatic expression of the 3333 relationship, but the extent of the 3333fs privileges, and the contexts in which they may be exercised, differ from area to area in Fiji. Less dra- matic, but perhaps more basic, are the patterns of respect which must.bbtain between mother's brother and sister's son. Informants view this as an outgrowth of the respect which must obtain between the cross-sex siblings through whom the relationship is traced. Simi- larly, just as a request made by his sister is difficult for a man to refuse. so also is a request made by her child. The respect between 331333 (a man's sister's child) and 3333 (mother's brother) is mutual, but in the Matailobau area at least, the emphasis is very much on the respect which the sister's child owes to the mother's brother. Though the 3333_ relationship is traced through a person with whom both parties share the bond of common substance, a woman who is mother to one and sister to the other, it necessarily joins two distinct patrilines. As his mother's son, the 33§3_shares something of her brother's nature, but as his father's son, the vasu remains 274 an "other," a representative of a line with whom one stands in an exchange relationship. This relationship was begun, or at least renewed and reinforced, by the marriage of the sister. . Here as elsewhere in Fiji the; term 13313? has the general meaning of "kinsman" and can be used to refer to anyone whom one recognizes as a relative, regardless of the nature of the relation- ship. But in the Matailobau area it also has a more restricted meaning; it refers to the matagali to which one stands in a 3333 relationship. This more restricted meaning of the term has parallels elsewhere in Fiji. In Moala (Sahlins 1962:150), the Exploring Islands of Northern Lau (Walter 1975:186), and elsehere in Lau (Hocart 1929: 33-4), 331333 refers not only to relatives in general, but also to the relationship between cross-sex siblings. In these areas 3313: is analogous to the Bauan 9333:. In Matailobau there is no special 6 term for cross-sex siblings; they are simply tuka-ltaci- to one another, but the partiline of the brother are 3313 to the children of the sister. The patriline of one's true mother's father, mother's brother, and mother's brother's son can be referred to as one's 3313_ 3133 or ."true 3313." One's 3313_are people to whom one looks for support through- out life. They may extend cultivation rights to matagali land. They may provide ego's own or his children's spouses. But it is with respect to the death of their 3333 that the role of the 3313 assumes its most dramatic expression. Indeed, it is really only in the con- text of death that the term T3313? takes on its more restricted mean- ing. 275 Funerals are organized around a series of exchanges between the matagali of the deceased (ng_jhtgu§gj_nj mate; lit.: "the owner's of the death") and his or her wgga, In order to secure their partici- pation, the wg§g_must be notified of their vasufs death, and the ‘ manner in which they are informed is a significant factor in determ- ining the level of their involvement. As a public service, Radio Fiji regularly broadcasts death notices (1_tukutuku nj_lggg) during Fijian language programming, but unless there is some extenuating circumstance, the Egg; must be notified vakavanua, "in the way of the land" (i.e., personally by special messenger.) Upon receiving news of the death the wg5a_meet (veivosaki) to assemble the wealth (j ygu) which will constitute their initial prestation to the tauggj_gj mate. The size of this offering depends on many factors, including the age, sex, and rank of the deceased, but it should include many mats, perhaps as many as ten small tabua, and at least one larger 3993. When they arrive at the vflgflimfig. the 525g make their initial presentation and receive yaggng_and per- haps one or two small tabua in exchange. Prior to their arrival, the tagtgiflniflmgtg_have placed a large tabua_under the head of the deceased. This is called the ulu kali (ulu - "head"; 5211_= "wooden headrest"). The wg§a_in the person of a makubuna of the deceased (ideally a mother's father but any person of an alternate or ego's own generation) removes the Egg gali_and replaces it with another large ‘tgbga, The wgga then lift the body of the deceased and remove the pile of mats on which it rests replacing them with other mats which 276 they have brought. Thus, in effecting the exchange of both the large tabug_and the mats, the wgkg move or lift the body of the deceased. When the time comes, they will claim the body itself and carry it first to the church and then to the grave. When the burial party arrives at the graveside, two more medium sized tabua are presented to the weka, The first of these, .ng.1gvg_gjhvua, is given when the body is lowered into the grave. In this context, lgyu_means "to sink or lower." yua_can refer to any fruit or produce, but in the local dialect is used more specifi- cally for yams. In this context it also carries the connotation of "offspring" or "descedant." At the same time or shortly after the ‘lgvgugi.!ug is given, a second tabga_called kena tirovi is presented to the weka, .Iéggvi means "to hinder or prevent," for the kena tarovi is given to prevent the wgka from following the deceased into the grave. At some time after the funeral there are two other exchanges which may, and often do, occur at the same time. The first of these is known as the gum (lit.: "his/her remains"). The yfinipgna is not to be confused with the personal possessions of the deceased, some of which may also be given to the weka, It consists, rather, of mats and perhaps a few small Egbug_given by the taukgj_nj_mg§§ to the weka, The second named transaction consists of a single large tabua, called gg_kggg.glg, presented to the weka_in exchange for which a magiti is given to the £2!££i.fli.fl飧~ There is obviously much, much more that could be said about funeralssin general and funeary exchanges in particular, but the 277 point that I wish to make here is that the major exchanges between the matagali of the deceased and the wgkg constitute a recognition by the former of the connection between the deceased and the matagali of his or her mother. The nature of this connection is emphasized; it is not merely an important affective tie, but a bond of sub- stance. The taukei ni mate give back the body to the people from whom, at least in part, the living person had been derived. Thus the wgka receive at death that which they were for the most part denied during the life of the deceased: the right to claim their kinsman as their own. As may be expected, two persons who stand in a relationship of veiwekani to the same patriline stand in a kind of equivalence to one another. Such persons are said to be rau vasu vata, "they two (who are) yg§u_together." Though it may be overshadowed by some other relationship, persons who are va§g_yatg_are joined by a kind of extended siblingship. In cases where the mothers of such persons call one another V"siste_r," those persons will indeed be siblings to one another, for one's mother's sister's child is equivalent to one's own sibling. I once attended a funeral some distance upriver from Nairuku- ruku. When our party arrived and entered the vgl§_gjhmg§§, I was surprised to see an old man from our village already present.and, moreover, sitting at the upper end of the house. When I asked about this, I was told that he and the deceased were veitacini (sib- lings), for the mothers of both men had been women of matagali Nakorowaqa (see Figure 18). 278 = Mataqal i ‘ . Nakorowaqa -"-"' Indicates classifi- catory siblingship Figure 18.--A Relationship of Vasu Vata Referring to Figure 18, the mother's of A and B stood in a sibling relationship to one another because their fathers had been siblings before them. Because their mothers stood in a sibling rela- tionship to one another, A and B were not only rau vasu vata to matagali Nakorowaqa but also classificatory siblings as well. It may be recalled from the last chapter that matagali Nakorowaqa and Nabubuco had constituted a single local group in the past and are today considered fbrothers." One basis for this sibling- ship lies in the fact that in the generation immediately preceeding that of their oldest living members, men of both groups married women of a third patriline which today constitues a distinct segment within Nabubuco. The sons of these men were not only va§g_vg§g to this third line, but also stood in a sibling relationship to one another. Thus, 279 not only may local groups (matagali) consist of two affinally and cognatically related patrilines, but may include others to which one or both of these lines are similarly related. In this as in other Fijian dialects, the term fkgwgfl conveys the general meaning of "offspring" or "line of descent," but like wgkg_it also conveys a more restricted meaning as well. In Matailobau cognatic kin one step removed from the veiwekani relationship are said to be kawa_to one another. That is, a person and the children of his or her XEEE are vékgw§_to one another or are said to stand in a relationship of veivakawai. The following examples will suffice to illustrate the concept. Case 1: Ego stands in a relationship of veiwekani to alter, his MZDS, but stands in a relationship of veivakawai to alter's children (see Figure 19). Ego 4—’ Veiwekani <----> Veivakawai Figure 19.--A Relationship of Veivakawai (Case 1) 280 As Ego's MZD alter's mother was related as a younger sibling to ego, and, thus, Ego was the equivalent of a mother's brother to alter. Ego calls alter vatuvg_(dZS) and, in turn, is called mgmg_ (MB). Thus Ego and alter (who stands in a vasgglike relationship to Ego) are related as veiwekani. But Ego's relationship to alter's children (Ego's MZDSCh) is quite different. They are vg§u_to their own mother's patriline, not to Ego's. Ego and they stand in a rela- tionship of veivakawai. Case 2: Ego claims to stand in a relationship of veivakawai to Nabubuco giving as his reason the fact that his father's father (whom he calls £u_b_ugu_) was 3931; to that matagali ("El—99.39.393.99. vasu vei eratou"). A=é ,///””////, AL Q? = Mataqali Nabubuco % 4-——-v Veiwekani .’ 25 *---** Veivakawai 9 Figure 20.--A Relationship of Veivakawai (Case 2). 281 Case 3: The same Ego referred to in Case 2 says that when they are born, he and his ZChCh will stand in a relationship of veivakawai (see Figure 21). a———9 Veiwekani ¢----+ Veivakawai Figure 21.--A Relationship of Vivakawai (Case 3). Ego's ZChCh will stand in a vg§u_relationship (= veiwekani) to the patriline of their own mother's father's and mother's brothers. One's own weka, or mother's patriline, constitute the most important category of cognatic kin, both affectively and structurally. But the wgkg_of one's parents, groups to whom one stands in a rela- tionship of veivakawai,_constitute an important extension of one's kin network. Earlier we noted that relationships traced through men (agnatic ties) establish lines of vertical cleavage between groups. Moreover, when coupled with the principle of seniority of descent, agnation results in a system of hierarchy in which the line of the elder brother outranks the descendants of the younger brother, an aspect of the kinship system which we will explore more fully in the 282 next chapter. Thus, agnatic descent both unites and divides people. It not only constitutes an important reason for saying that people are of the same "type" (matagali), but it also entails divisions within such categories. 0n therother hand, while strictly speaking they are are a matter of individual kinship networks, cognatic ties, and especially ties of veiwekani and veivakawai can unite groups establishing, if you will, horizontal bonds. As previously noted, such ties can serve as a justification for the "staying together“ of two or more patrilines. Over time, what were distinct groups externally bounded by agnatic descent and relating to one another in terms of exchange come to be regarded as people of the same “type.' Internal realtions are then a matter of cooperation and sharing (of land, food, personnel, and valuables) as well as exchange. At that point, it can become difficult to unravel the various principles of organization which bind the group together. FOOTNOTES 1There is no necessary link between chiefly rank and curing powers. Other chiefly lines have no such powers while some commoner lines are also believed to have the power to cure specific types of ai ments. 2I am indebted to Thomas Trautmann's comprehensive study, Dravidian Kinship, for this observation (1981:43). 3In considering the possibility of such a relationship, we would want to consider such factors as informants' tendencies to equate SpF with M8, SpM with F2 and the extension of the term "Sp" to any cross cousin of the opposite sex. 4Actual siblings are veitacini dina sara, or "very true“ siblings, a phrase which may be further clarified by stating that two persons share one father and mother. 5Note that while a woman's diva is cross kin to her, he or she is parallel kin (taci-) to that woman's husband. While a woman classifies her daugher's children in the same way that her husband does, thus neutralizing the contrast between cross and parallel kin, that contrast is preserved in the classification of her son's children. 6This is not completely correct. Salavolo (sala = path; volo = to go aside) is an alternative term for a man's sister, an allusion, perhaps, to either the avoidance pattern which should obtain between cross sex siblings or to the fact that their descendents represent diverging lines. 283 CHAPTER VIII PRINCIPLES OF HIERARCHY In Chapter VI we considered the principles of descent and coresidenCe, as manifested in several units of Fijian social organi- zation. The unifying theme of that chapter was the attempt to under- stand what Fijians mean when they say that people, organized into groups or categories, are of "the same type." While still falling under the general heading of social organization, the subject matter of the next two chapters is more diverse. Among other things, they will consider the ways in which rank and neighborhood affect social relations. In a sense these constitute a residue of topics not dealt with in the preceding chapters, but a unifying theme does exist. While Chapter VI considered what it means to say that peOple are simi- lar, this chapter and the next will look at some ways in which people of different "types" (matagali) relate to one another. Hierarchy, specifically differentiation on the basis of rank, is perhaps the most fundamental feature of Fijian society at all levels. It is, of course, possible to look at this hierarchic aspect of Fijian society as something which divides it, and in the context of contemporary party politics there have been attempts to make the chiefly hierarchy an issue (Cf. Norton 1977). But it is also possible to view ranking as a unifying principle, for basic to it is the notion _284 285 of ritual interdependence and mutual ministration. It is foreign to that secular bias of the modern world which allows us to speak of "politics" and "the political" as something separable from other aspects of social life, notably the sacred. Throughout this discussion, "ranking" will be understood to mean the division of society into social categories which are ordered in terms of prestige.1 Often these categories are themselves char- acterized by an ordering of their members. Such is the case in Fiji. Basic to ranked societies is the notion that human beings are funda- mentally unequal. While the differences in social value associated with ranking tend to be translated into differences in power, inequal- ity of rank is not necessarily accompanied by differential access to material resources. This definition of rank differs significantly from that adopted by Walter in his analysis of chiefly succession in the Exploring Group of northern Lau. According to Walter: Rank in Mualevu [an administrative district in the Exploring Group] is a hereditary status derived from a parent and ascribed by order of birth, hence it is a medium of stratifi- ' .oation only between those persons acknowledging descent from some common ancestor or ancestors. There can be no relative rank between persons lacking common descent. For example, a chief is of superior ascribed status to a commoner by reason of the hereditary occupational status ofhis descent group as chiefs (as opposed to that of the commoner's roup as, say, warriors) and not by any criterion of rank Walter 1974: 302). ' What Walter calls "rank" is only one of several principles operating in Fijian society to effect the distribution of status positions ordered in terms of prestige. This principle will be referred to 286 hereafter as seniority of descent, a term whose meaning will be clari- 2 fied in the course of this discussion. The benefits of defining ranking in more general terms, as we have done, is that it then subsumes various manifestations of hier- archy in Fijian society under a single term. This is advantageous because, as we shall see, Fijians themselves speak of these various instances of hierarchy as if they were similar in principle. In the Fijian view ranking is a reflection of the hierarchic aspect of the sibling relationship. In tum, hierarchy among siblings 3 Among siblings of the is the outcome of two factors--sex and age. same sex, the elder outranks and has authority over the younger, but the relationship among cross-sex siblings is more problematic. Gen- erally speaking, brothers have authority over sisters, though it is perhaps better to think of this relationship in terms of responsibil- ities. The brother has a responsibility to attend to his sister's welfare, a duty which does not fully terminate with her marriage. He also has a responsibility to monitor her behavior, but both of these responsibilities are matters of some delicacy, for brothers and sisters must avoid discussion of, interference in, and even overt awareness of each others intimate affairs. For her part, the sister has the responsibility to consider her brother's judgments, to be sensitive to his position as guardian of her morals, and to provide, for his comfort and will being, such personal services as he might require. Throughout life, brothers and sisters are under considerable obligation to honor one another's requests. A brother may request the 287 services of his sister and her husband in gardening, housebuilding, or in meeting his ritual obligations. A sister's husband may make similar requests of his wife's brother, requests which are based on the brother-sister relationship. A woman's husband may request the use of garden land on her behalf and that of her children, or a woman may turn to her brother for refuge in a dispute with her husband. Indeed, informants have spoken to me of cross-cousin marriage as an extension of this obligation of brothers and sisters to honor each other's requests, in this case the sister's request for a son's wife. Given the rule of patrilineal descent, there is a cultural emphasis placed on the relationships among brothers, but it is the relation- ships within and between sets of cross sex siblings, structured by proscription and exchange respectively, that cement the system. As noted above, with respect to hierarchy, the cross-sex sibling relationship is problematic. Unlike some other areas in Fiji, in central Viti Levu sex takes precedence over primogeniture with respect to both rank and succession to office. First born women do not outrank their younger brothers nor are they eligible to hold traditional public office, even in name. But while it is true that brothers outrank and have authority over sisters, it is also true that during their own childhood and adolescence older sisters shoulder much of the responsibility of caring for younger brothers so that, with respect to authority within the sibling group,the principles of age and sex come into conflict for a time. This effects their rela- tionships throughout their lives. If the brother is elder, a married sister cannot ask for anything directly. Her requests may be relayed 288 by her children, not directly to their mother's brother, for as their m9 2 they owe him respect, but their gaggle, their mother's brother's children. If, on the other hand, the sister is elder, she is able to make her reqests directly. ' When Fijians explain ranking, they do so with reference to the relationship among same-sex siblings--specifically, brothers and their offspring. In the cultural model, rank is determined by senior- ity of descent. Just as the elder brother outranks the younger, so do his descendants outrank those of the junior sibling. Thus, refer- ring to the diagram below (see Figure 22), the descendants of the elder brother (A) outrank the junior lines of B and C. Similarly, ‘within line A, the descendants of 1 outrank those of 2. Moreover, as descendants of the first born line, the descendants of 1 outrank 2, A Senior in terms ;. Junior in terms of descent of descent Figure 22: The Principle of Seniority of Descent. 289 It is In! reference to the principle of seniority of descent that Fijians explain the ranking and functions of the various matagali within .the y‘a_vusa_ and of various lines within the matagali. The idiom used to express these hierarchic relationships is that line (or matagali) X is gage ("elder') and line y gong ("junior"). Now when an informant says of someone that they are "older" then someone else, they could be claiming one of two things. They could be saying that the first person is a descendant of a line senior to that of the second person, or they could be referring to personal seniority. That is, they may be saying that the first person is chronologically older than the second. It is not always immediately clear from the context which of these two statements is being made. It is important to keep in mind that the principle of senior- ity of descent really only provides a model for expressing hierarchic relationships among descent groups or categories. It is not possible to substantiate genealogically that one matagali within a yayu§a_is senior to another in terms of descent. Indeed, it may not be possible to provide genealogical evidence for the claim that, within a single matagal , one line is senior to others. This is, of course, not to say that such claims are without a genealogical foundation, but only that the genealogical basis is posited rather than demonstrated. As the Fijian anthropologist Nayacakalou has observed: A basic incongruity in this notion of seniority of descent is that, although the principle is precisely stated, it lacks the most elementary foundation on which it can be based: a firm genealogical framework. Except occasionally among chiefs, genealogical knowledge is very poorly developed among Fijians. To remember four generations is, for most, 290 a great achievement. In these circumstances, it is obvious that seniority of descent may be disputed, and, in fact, it is. . . . Although the [Native Lands] Commission was given genealogies, and these seemed authentic on the basis of their internal evidence, . . . many of them were only skeletal, giving a line of ancestors ending in the contem— porary chief and rarely with the elaborations necessary to show relationships of existing families (1975:33). Note that the principle of seniority of descent presupposed an ordering on the basis of personal seniority within the sibling group. While seniority of descent is used as a model of, or an idiom for expressing, the hierarchic relationships among descent categories and groups, hierarchy among individuals of the same descent group tends to be structured in terms of personal seniority. That is, within the extended sibling group, ideally composed of all the men of a single generation in a given matagali, individuals are ranked according to their birth order. For example, while it may not be possible to establish the genealogical basis for, or reach any consensus about, the relative seniority of each patriline within the Nairukuruku branch of matagali Siko, it is a simple matter to estab- lish relative seniority of individuals within a given generation, and interaction is structured accordingly (Cf. Sahlins 1962:263). Simi- larly, while the line known as matagali Solia within the local group known as "matagali Nabubuco" is said to be gase (elder in terms of descent) and the line recorded in the records of the Native Lands Commission as "matagali Nakorowaiwai" is said to gone, leadership within the local group and the ranking of its members is determined by seniority. Thus the leader (liuliu) of the local group is its most senior male, despite the fact that he is not a member of Solia. 291 The same pattern obtains among the other local groups in the village. Precedence (e.g., in yaqona drinking) is accorded, and positions of leadership filled, on the basis of personal seniority, rather than seniority of descent. We can say that, on the level of groups and categories, ranking necessarily involves a focus on differention through descent, but with respect to ranking within groups, the emphasis is on the unity of the extended sibling group and seniority within it. While it would be correct to say that the various yavusa repre- sented in the area are ranked with respect to one another, it is not possible to say that this ranking reflects seniority of descent. Even though such hierarchy may be spoken of in the idiom of seniority relations, thererr is no tradition of conmon descent among the different yayusa, The hierarchic relations among yayu§a_are a reflection of political realities. Thus, Nakorowaqa and Nabubuco (referred to in some contexts as yaygga) rank below yayuga Siko because their ancestors took the latter in and accepted their senior matagali as chiefs. Simi- larly, locally the senior matagali within ygyusa Siko ranks higher than the two chiefly matagali, Navitilevu and Nasautoka, whose posi- tion they undercut through their own rise to political dominance. Navitilevu had been an especially powerful group, but they were defeated in warfare by an alliance of local tribes in the late 18605 or early 1870s and were scattered over villages in the Wainimala, Wainibuka. and Waidina valleys (Brewster N.D.:53). Today they are the dominant matagali in the village of Saumakiya upriver from Nairukuruku, and 292 their prior association with Nakorowaqa and Nabubuco becomes salient in some contexts. Though locally they rank below'yayu§3.5iko's senior matagali,Navitilevu and Nasautoka are nonetheless of chiefly status themselves, and thus rank higher than all nonchiefly matagali in the area. It would be misleading to think of this ranking of descent categories as an all-inclusive system. It would be difficult to say, for example, whether Nakorowaqa ranks higher than Vusovuso or vice versa, though in terms of the village itself, the former's status as j taukgi_might be relevant to the matter. What is clear is that both rank below Siko and other chiefly matagali and below matagali Nawaita and Nadovunisili, the bgtj_wihin yayuga Siko. While it is true that the ranking of the various matagali within a yayg§a_is considered to be a reflection of seniority of. descent, it is also possible to view it as an outcome of a preference ordering within the ritual division of labor. These points of view are not mutually exclusive, for in the cultural model the ritual divi- sion of labor among the constituent matagali reflects the birth order of their founders, sons of the founding ancestor of the ygxusa, The specifics of this ritual division of labor differ from place to place in Fiji, but as a consequence of Bau's rise to political ascendancy and the adoption of a version of Bauan social structure as a model for Fijian organization throughout the group, familiarity with the ideal structure of Bauan yayg§a_is widespread. According to the model, a Bauan yavusa is comprised of five matagali purported to have been 293 founded by the first five male descendants of the yavusa's founder. The first four of these matagali had specific functions as shown in the diagram below (after Geddes 1945:36). Founder of Yavusa 1 2 3ZL. - ‘4.JS .Jk Mataqali Turaga (Chiefs) Mataqali Sauturaga (Chief's Henchman) Mataqali Bete (Priests) Mataqali Matanivanua wat-o II Elder a“? Younger Figure 23.-~Division of Labor within the Yavusa. Several points should be noted. Firstly, a similarity of ritual function in different villages or chiefdoms is sometimes a justifica- tion for saying that two matagali are equivalent whether or not they are yavusa vata. Secondly, the ritual functions appropriate to matagali of different yayusa affect their ranking vis-a-vis one another. For example, as bati to the chiefs of matagali Siko, Nawaita and Nadovunisili rank higher than matagali Nakorowaqa whose proper function is that of mata_nj_yanua, With respect to the former, Nakorowaqa are 1193 gj_k§kgna ("the bearers/recipients of food").4 The bati are gase (elder), the liga nj_kakana, gone (junior). Another 294 way of saying this, is that the nature of their relationship to chiefly descent groups affects the rank of nonchiefly matagali vis-a-vis one another. And thirdly, just as it is possible for a matagali to have an enduring relationship with more than one chiefly group, so is it possible for them to have different functions, and hence, different ranks, in the contexts of these relationships. With respect to descent groups and categories, it is possible to view the ranking system as a consequence of the fundamental division between chiefs and commoners. Chiefly rank is inherited patrilineally and along with it, the right to use honorifies before one's name-- ‘ratu ('sir") for males and 3g: (“lady") for females. Not all members of chiefly matagali are so entitled. In Nairukuruku there are three patrilines within matagali Siko which have been without titles for at least three generations, and there are several plausible explanations for this. One informant offered an explanation not unlike a textbook exposition of the internal structure of ranked descent groups. He said that in the past, if a chief had several children, the title 'ratu' would only be passed on to some of them. Only the eldest would be a "big ratu," others 'little.ratu." Some younger siblings were "cast aside? (gigs Ejkg) with respect to titles and could not pass them on to the next generation. Thus over a number of generations some members of the matagali would have titles, and others would not. According to this informant, today and in the recent past there has been a proliferation of titles. He pointed out that the ancestors had been ygku (wise), because if there are many titled lines, there'flscompetition for supremacy. If two men 295 of an extended sibling set are fratu," and one takes precedence in yaqona drinking, the one who has been passed over is angry. In talking about this status competition between titled men, the inform- ant took the index fingers of both hands and raised them in a climbing motion, first one higher and then the other. Another informant, himself untitled, claimed that an ancestor had renounced his claim to a title. The informant traces the origin of his line to Taulevu village. (knanother occasion he referred to himself as the ulgnmatua_(first born), a reference to the fact that the Taulevu branch of Rokowai's descendants were elder and the Nairu- kuruku branch younger. It is said that at some time during the life of Rt. Meli Davetanivalu (see Chapter VI), his older kinsman, the chief of Taulevu, was seriously ill, and asked Ratu Meli to strangle him and put an end to his suffering. In return for this service, the elder kinsman is said to have renounced his descendants' claims to the paramouncy which, frbm that time on, passed to Ratu Meli's line. The same account was given to me by several individuals. Indeed, it had become an issue in discussions concerning chiefly succession. But if this particular informant was linking his untitled status to this event, he was the only one to make this connection. Since those who bear the titles firatu“ and "adj' are considered to be the descendants of former occupants of chiefly office, the possi- bility arises that untitled lines within chiefly matagali have separate origins, perhaps as the descendants of sisters' sons. After genera- tions of coreisdence and coactivity the details othhese separate origins may have been forgotten, but the tradition of their commoner 296 status retained. These attached lines may have functioned more or less as retainers to the titled families. Indeed, there is an ele- ment of ministration in the relationships between some of these house- holds and that of the current paramount, though it would be difficult to say how much of this is due to propinquity and their status as junior kinsmen. The head of one such household serves as village chairman, and in this capactiy he is regularly called upon to translate the paramount's directives into action. The head of another is a former turagaugi.kgrg ("village chief") an office whose occupant is today little more than a town crier. The position of these untitled lines might appear ambiguous. On the one hand they are considered full members of the chiefly des- cent group (matagali turaga) while on the other hand they themselves are not of chiefly status. By this I mean that neither they nor their descendants are‘eligible for chiefly office for a long as their untitled status is remembered. Another way of saying this is that they are members of the chiefly descent group, but not members of the chiefly class (Cf. Walter 1978b:4,10). But in the emic view there is no ambiguity in their being "chiefly? (turaga) in one sense (i.e., full members of a chiefly descent group) but not in another (i.e., being tamata lailai or commoners rather than turaga, persons of chiefly status). The content of the Fijian category "turaga,' a term which I have consistently translated as "chief" or 'chiefly," is highly contextual, a fact that will have to be dealt with presently. In considering the position of untitled lines within chiefly descent groups, there is another explanation which has to be considered. 297 That is the possibility that they descend from the illegitimate sons of chiefly women. I hasten to add that no informant ever suggested this to me. Even if it were known to be the case, or even suspected, it would not be openly discussed, for to impute illegitimacy to another is a serious matter. Even cases which I knew about, even after it was known that the facts of the matter were familiar to me, were not openly discussed. Like the English "bastard,“ the term for illegiti- mate offspring, lgyg_nj gala ("child of the path9)'is a form of vituperation. Illegitimacy is defined locally to include children born to women who are not openly cohabiting with a man in a recog- nized union and whose paternity is not publicly acknowledged.5 Since paternity is not acknowledged (even though attributed in gossip and accusation) the child does not inherit his or her status from the biological father. Such children may be adopted by a married couple, in which case they are included in the adoptive father's matagali, or stay with their natural mother's matagali. Should the mother enter into a subsequent union, the child may go to live with her but is more likely to be cared for by her parents or shunted from household to household in the mother's natal matagali. Illegitimate children of titled women do not inherit their mother's title. Thus illegiti- mate sons would have no title to pass on to their descendants who may beomce an established line within the host matagali. Children are usually named after a specific member of their own matagali so that the same names are recycled generation after generation. It is possible to think of a corpus of names as the 298 exclusive property of a particular matagali. Indeed, as we have seen, names are almost thought of as entities or persons specific to this or that matagal‘. The issue of naming is relevant here because today it is pos- sible to acquire the namesake's title as part of one's own name. A member of one of the untilted lines within matagali Siko is named after an older, titled man, and the younger man's full name includes the hororific 'ratu.f By selecting names from among the titled mem- bers of the mother's matagali, it is possible to superficially dis- guise the fact of illegitimacy, though such was not the motive in the case alluded to here. But, in such an event, a man's children do not inherit the right to attach titles to their names, and the same strategy would have to be repeated in each generation until the facts of the matter are no longer remembered. The meaning of the Fijian word lturaga,9 translated here as "chief" or 'chiefly,' has a contextual or relational aspect which the English word, of course, shares.6 While in the village I was fre- quently told of impending events which the fituragafl were to attend. In the early months of field work, I found such statements to be l ambiguous, but gradually began to be able to predict the referent of fturagafl from the nature of the event under discussion. In the con- text of these statements, the "chiefs" being referred to might include all of the men (and possibly women) of the local group known as mataqali turaga--the "chiefly mataqali,9 mataqali Siko; the liuliu or leaders plus other elders of all of the local groups in the village; 299 or all of the men (turaga) of the village as opposed to women (marama; Cf. Walter 1978bz4). In the context of other statements, the word Ttgragafl might refer to the occupant of a particular chiefly office, to persons in other leadership positions, or to persons of chiefly rank in general as opposed to commoners. Locally fturggaf in the last sense would include all the titled member of matagali Siko, Navitilevu, Nasautoka, and so on. All of these usages have in conmon the notion of hierarchy which is, of course, a relational concept. Something or somebody is only dominant in relation to another. In the examples cited above, the somebodies who are "higher" are turaga, but the terms applied to the categories in relation to which their position is defined differs from case to case (e.g., members of the mataqali [levu gj_matagalil as opposed to its leader, or women as opposed to men.) In keeping with our earlier, very general definition,I will refer to each of these instances of hierarchy as a manifestation of ranking. The fact that Fijians use the term 'turagafl in these various contexts would seem to indicate the recognition of an underlying similarity. Such is indeed the case, for in discussing any of these contrasts, Fijians rely on two basic idioms which tend to be used interchangeably. They may refer to the dominant position in a hierarchic relation as elder (gage) and the subordinate position as junior (gong). By this they may mean that the dominant party is senior in terms of descent, or chronologically older, or simply that the relationship between domi- nant and subordinant is structured as if the former were elder, in age or descent, and the latter junior. They may also discuss such 300 relationships using a spatial--that is, linear-:metaphor, a tendency which they share with many peoples around the world, but one which, in Fijian culture, is elaborated in the conceptualization of social space in many contexts (e.g., seating order at meals). They often speak of the dominant or senior party in any hierarchic relation as "above" or "higher than" (3 cake, _e_ 9919) the subordinant/junior party whose complementary position may be described as 'below" (e_ra). In its broadest sense £9£292.i$ applicable to any status position which may be described as senior to, or higher than, some other, but it could be argued that it is the relationships among lineal kinsmen, structured by patterns of authority and respect, reflecting generation, relative age and seniority of descent, and given spatial expression in various social contexts, which provide the cultural model for all hierarchic relationships. Despite the fact that the English 'chief' is equally relational and hierarchic, its connotations (i.e., with respect to a particular form of political office) make it an inappropriate translation for the Fijian 'turaga? in some of the instances discussed above. In his ". . . Examination of Hierarchical Notions in Fijian Society . . ." (1979b), Walter examines the difficulties of translation which surround the word and suggests certain conventions. I refer the reader to his discussion of theaproblem, but have found it convenient to adopt other usages in some cases. Throughout this and other chapters the word "chief? has been used to refer exclusively to the titled occupants of traditional public office. As the i taukei n1 waluvu (Powner of 301 the flood"), the chief of Nairukuruku is the highest ranking office holder in the gang; (chiefdom) of Nabena. Thus from time to time the occupant of that office has been referred to here as the "paramount chief" or, simply, "the paramount." The senior men who function as the ggufagtg_heads of commoner matagali are referred to here as "leaders," a direct translation of the Fijian term (liuliu) describ- ing their function. And, finally, the term "chiefly" has been used to refer to any person, or descent group whose members include such persons, who have inherited the privilege of using the titles "ratu" and "adi" before thier personal names. "Chiefly matagali" are matagali whose proper function it is to provide the occupants of public office. As members of such matagali men who bear the title "ratu" are at least potentially eligible for such office. The term "chiefly office" will be understood to refer to only those offices which must be filled by a man of chiefly rank (e.g., the paramount) as opposed to those offices which are filled by men of commoner status (e.g., the Tui Rara). A feature common to those societies characterized by what we have termed ranking is the tendency for all social relationships to be tinged with, if not structured by, hierarchy. This penetration into the totality of social life may be symbolized in various ways, but foremost among them are patterns of etiquette, those rules dictat- ing acceptable behavior between the occupants of different statuses. We have already said something about how such rules structure inter- action within the domestic unit (e.g., at meal times). But here I 302 want to focus on the way in which notions of hierarchy affect the relationship between husband and wife. Implicit in the discussion of food procurement in Chapter III was the understanding that the married couple constitute a partner- ship. Each has specific economic tasks for which they have the pri- mary responsibility, and together with other household members, they share the duties of providing for the domestic unit. For as long as he is fully active, the senior male makes decisions (lewa) concerning the male labor force, and his wife does the same for the female members of the household. The married couple also function as a unit in meet- ing the ceremonial obligations entailed by their respective kin net- works, each contributing the form of property or labor appropriate to their sex. While it would be correct to describe the relationship between husband and wife as a partnership, it is not a partnership of equals. Indeed, the tone was set in the ceremonial exchanges which played . as prominent a role in the traditional wedding. Though there was an element of competition involved, the prestations made by the groom's side, and hence their prestige, were expected to be greater. With respect to the household in general and his wife in particular, a man is turaga. The English adage "A man's house is his castle," also applies to Fiji. Assuming that the married couple reside in the husband's natal village (which may also be the wife's natal village), a married women's rank within her husband's local group (matagali) is derived from his 303 own position within it. This, in turn, is a matter of seniority. When outsiders are present or whenever meals are being eaten, a wife sits below her husband. She should defer to his word, at least publicly, though heshould also respect her judgment in matters within her domain. On most public occasions at which both sexes are repre- sented, it is the husband who speaks on behalf of the household (e.g., in ceremonial exchange). A woman who preempts the male role is wag§g§_(proud, boastful) and is resented by men and women alike for being viavia‘ tagane ("wanting to be a man"). Given the element of hierarchy in the husband-wife relation- ship, hypogamy (i.e., the marriage of a chiefly woman toga commoner) results in an anomalous situation. I was told that in the past such marriages would have been irregular and probably not permitted, but there are currently eight hypogamous unions in the village. In those cases where the wives are also members of the Nairukuruku branch of matagali Siko, they tend to play a prominent role in women's affairs, not only in their husband's matagali, but in the village as a whole. In each case they are middle-aged women, not only of high rank, but also of intelligence and strong personality. In any hypogamous union, there is a potential for conflict between a wife's conscious- ness of her chiefly status and the deference which she owes her hus- band. The problem can be resolved by compartmentalizing roles (e.g., in some contexts a chiefly woman may be served yaggga_before her husband while in others she conforms to the role expected of her as a wife.) But it is felt that this incongruity of statuses might 304 lead in some cases to outbursts of temper or, worse, suppressed resentment that may be detrimental to a woman's health. Earlier in this discussion contrast was drawn among several principles of hierarchy. Among matagali acknoWledging a common ancestor the order of precedence is said to be a reflection of the relative seniority of their descent. The same principle is appealed to in discussing the order of precedence among the various localized segments of a single matagali as well as the various patrilines within these segments. At none of these levels is there likely to be the necessary genealogical knowledge to demonstrate the basis for the ordering. And there is an apparent paradox in the fact that at the level of greatest genealogical depth (i.e., among the various matagali of a single yayusa) there is total agreement about this ordering, while at the level of least genealogical depth (i.e., among the various patrilines of the same matagali within a single village) there may be considerably less agreement. The order of precedence among different ygyusa, categories unrelated by agnatic descent, is recognized to be a reflection of the political relations among them, and these, in turn, are a product of various historical processes. Discussion of the hierarchic relations between matagali of different yayg§a_involves an appeal to the ritual functions specific to each. Finally, among individuals of the same sex and members of a single mataqali, rank is largely a matter of seniority. That is not 305 to say that seniority of descent is suspended at the individual level, merely superseded. A man forced to defer to a slightly elder member of his own matagali may harbor in his heart knowledge of his own line's seniority, but defer he must. Of course, ranking in terms of personal seniority is occasionally complicated by incongruities of generation so that, for example, a man may owe respect behavior to a classificatory "little father" (tamana lada; i.e., father's younger brother) or mother's brother who is, nonetheless, younger than he. (Patterns of respect are more likely to be emphasized in the case of a classificatory mother's brother.) Personal seniority establishes the order of precedence in many contexts: it determines where one sits at meals, in church, and in any public assembly including yaqona sessions; it determines in what order one drinks yaqona and whether one serves others or is served by them; in public assemblies it determines whether one speaks or only listens and, if they speak, in some cases it may determine the order in which they do so; and it determines the rank of the household one heads vis-a-vis other households in their 1 tokatoka. This list could be expanded indefinitely, for there are no interactions between mem- bers of the same sex which are not affected by seniority. This means, of course, that it also affects the interaction of members of differ- ent matagali, but in such cases hierarchy is also a reflection of other factors--for example, ritual function. The point which I wish to stress is that, within the extended sibling group composed of members of the same generation within a 306 single matagali, personal seniority is the dominant principle of rank. One context in which it supersedes seniority of descent is in determining the order of succession to office, as the following sec- tion will illustrate. In it] attempt to reconstruct the order of succession to the paramouncy, and in doing so interrelate historical materials with the oral accounts given to me by informants. To my knowledge, the earliest description of Nairukuruku is contained in a brief entry in the journal of Commodore Goodenough who visited the area in 1874. Goodenough was in Fiji to assist the newly appointed British Counsul, E. L. Layard, to determine whether the islands should be annexed by Her Majesty's government. He led the boats of the H:H.S. Pearl as far upriver as the village of Nakorovatu. The water level was low (it had taken the ship's cutter five hours to cover three and one-half miles), and so the party con- tinued overland the few miles to Nairukuruku, the last Christian outpost. It may be recalled form the first chapter that in 1872 repre- sentatives from Matailobau and the neighboring district of Soloira had met with W. S. Carew and, along with other Naitasiri tribes, had pledged their support to the Cakobau government. In turn, they were promised Government protection in the event that they were attacked by the non-Christian tribes further upriver. Despite their pledge, in the following year the Waikalou of Serea village, one of the principal groups of Soloira, participated in an attack on a village in Naitasiri district. Moreover, they and their allies threatened retaliation 307 against the European planters along the upper Rewa in the event that they or the Cakobau government became involved. Despite the threat, the government proceeded against the "rebels," put down the disturb- ance, imprisoned their chiefs, and confiscated lands.7 During the troubles of 1873 Nairukuruku and Matailobau dis- trict as a whole remained firm in their support of Cakobau despite yaygsa Siko's veibatiki relationship with Serea. But even after the government's success against the Waikalou and their allies, unrest continued in the interior, and later in the same year the residents of Nairukuruku, fearing attack from the district of Noimalu to the west, felt it necessary to build a war fence around the village. As a Christian outpost, it became a potential target for any of its non-Christian neighbors who wished to pursue their own quarrels without government interference, but in this case the roots of the dispute seem to lie in the burning of a village which the Noimalu were attempt- ing to rebuild. Goodenough mentions that the burning was motivated by the Noimalu people.'s‘failure to obtain the permission of the Matailobau before rebuilding (Goodenough 1876:218), probably indicating that the village had been destroyed by the latter in an earlier round of fighting. Brewster reports that the Matailobau suffered a defeat at the hands of a Noimalu force and that peace was restored when the Matailobau agreed to pay an indemnity. The last installment of this "war debt" was handed over at Taulevu village in 1886 (Brewster N.D.: 54). Brewster gives no dates for this conflict though it is likely to have occurred between 1873 and 1875. 308 The area seems to have been reasonably quiet at the time of Goodenough's visit in January of 1874. He makes no mention of forti- fications around Nairukuruku, but trouble between Matailobau and Noimalu was obviously simmering. Goodenough mentions that a young man from the Noimalu village of Nadawarau resided in Nairukuruku at the time. He was being prevented from returning to his native village, but it is not clear from Goodenough's account whether the obstruction came from the Matailobau or his own people. The chief of Nairukuruku expressed amazement (or dismay?) when Goodenough reported his inten- tions of traveling on to Nadawarau. The chief at the time of Goodenough's visit was clearly Ratu Meli Davetanivalu. Local tradition attributes the introduction of Christianity to Rt. Meli, and Goodenough reports that the chief of Nairukuruku was the first in the area to become Christian, and, in the Fijian view, proclaim his allegiance to Bau in the process. He told Goodenough that his conversion had taken place seven years earlier in 1867. At that time the Reverend Thomas Baker had been stationed at Davuilevu on the upper Rewa and was responsible for the conversion of the villages of Naitasiri, Viria, and Nadawarau, as well as villages along the Wainimala. It is likely that Ratu Meli was converted by Baker or one of his catechists. Baker, the only missionary to be killed in Fiji, died in the same year that the Matailobau chief claimed to have been converted. In his journal Goodenough refers to Ratu Meli as "the governor of the twelve tribes of Matailobau" (Goodenough 1876:218). It is not 309 clear what Goodenough means by "tribes," whether settlements, mata: .9111, or some other type of division, nor is the physical size of the area known then as Matailobau clear from his map which is limited, really, to the courses of the Rewa and Wainimala rivers. It does seem that by this time the chief of Nairukuruku was the paramount of the yanga of Nabena, the political core of the old district of Matailobau. This would seem to be the most reasonable interpretation of Goodenough's reference to this personage as the "governor" of Matailobau. The only other interpretation to place on the phrase is that the chief referred to was a functionary in the nascent native administration. I am not familiar with the details of how the dis- trict was being administered by the Cakobau government, if indeed it was being administred on a routine basis. Most of the references to the area in the papers of the Cakobau government are found in the reports of W.S. Carew, the Secretary for Naitasiri, to the Minister of Native Affairs. This was a period of flux. By March of 1874 the Deeds of Cession to Great Britain were signed, and by April an Interim Government was installed. It is not clear whether a native magistrate had already been appointed at the time of Goodenough's visit, but by May of 1874 that position was held by a younger kinsman of Ratu Meli (Harding 1916:69). This younger man was to play a central role in the administration of the district for many years, and, as we shall see, the effects which his career had on succession to the paramouncy are still being felt. Since the day of Ratu Meli the paramount chief of the ygnua_ has been drawn from one of the chiefly lines of the Nairukuruku 310 branch of matagali Siko. This was not always the case. It will be recalled that according to local tradition, the paramouncy passed to Ratu Meli and his descendants in accordance with the wishes of an ailing chief of Taulevu, an elder kinsman. Seriously ill, the older man asked his younger agnate, and some say namesake, to strangle him and end his suffering. The story does not mention any reaction on the part of the Taulevu chief's children or agnatic nephews to the dying man's bequest, but according to some reports, on the occap sion of.of the deliberations leading to the selection of the current office holder, his descendants played down the story and put forward a claim as representatives of the line of the El! matug (first born). On several occasions when the matter under discussion was village history in general and not chiefly succession, I was told that Ratu Meli had been the youngest son of Rokowai. The older sons remained in Taulevu. Apparently, this is the basis of the Taulevu claim to descent from the ..«Lflu _ma_tga_. As a younger sibling, if. that were indeed his relationship to the Taulevu chief, Ratu Meli's succession to the chiefship conforms to the pattern revealed by the order of succession since that time, but as to why the chiefly lines of Taulevu had been excluded from the office, I have no explanation other than this anecdote. What part the Nairukuruku chiefs' relationship to the Cakobau and, later, the Colonial government played in their dominance within the yangg_l do not know. Sometime in 1873 warriors from the Noimalu village of Nadawa- rau attacked a Christian village in Nabubuco near the headwaters of 311 the Wainimala and killed forty of its inhabitants. In response the government sent a punitive expedition under Major James Harding into the interior. In April, 1874, Harding led his force of 220 men south from Viti Levu Bay through Nalawa district. Knowing that the people of Nakorosuli village, less than ten miles upriver from Nairuku- ruku, were allies of Nadawarau, he waited for word about their inten- tions before proceeding up the Wainimala.8 It was necessary for them to declare whether they were friends or foes before an attack was made on Nadawarau because Nakorosuli could block retreat down the Wainmala and Rewa. On April 30, Harding received word from the chiefs of Nakorosuli, Navuca, and Nakorowaiwai refusing to aid the Government forces. On the next day messengers from Matailobau, possibly from Nairukuruku, told Harding that the Nakorosuli people intended to attack the Government force from the rear while they marched upriver to Nadowarau. Early on the morning of May 4, Harding attacked Nakorosuli driving off the majority of its inhabitants. A few managed to escape to Nadawarau, but most fled down river to Matailobau or across country‘to Nasautoka on the Wainibuka. On the day after the attack on Nakorosuli the native magistrate of Matailobau brought a force of 150 men to give a 5393 (formal disply of military support) for Harding. Whether any of this force participated in the attack on Nadawarau I do not know, but given the state of hostility between Mataildbau and the Noimalu village, it is likely that they did. Two days later the principal fugitives from the attack on Nakorosuli were captured in Matailobau. Two of them, Rokotuinaceva, 312 chief of Navuca, and Taunadi, chief of Nakorowaiwai were found guilty of the lighter charge of speaking evil about the Government and sen- tenced to six months hard labor in Levuka jail.9 Harding writes that "as we agreed with the Rewa and Navosa people" (his Fijian troops?), the people of Nokorosuli, Navuca, and Nakorowaiwai were told that they formed a matagali under the new bull (government appointed district chief) of Matailobau.10 The identity of the latter is not given, and it is unclear whether Harding is referring to Ratu Meli or, more likely, the younger kinsman whom Harding refers to earlier as the native magistrate. The fugitives were told that in the future they were "to live within the boundaries of Matailobau and to plant there . . . and the lands which they originally occupied were forfeited to the Crown" (Harding 1916:19). The inhabitants of the defeated settlement of Nakorowaiwai probably included the ancestors of the matagali of the same name who today form the core of the local group known as "Nabubuco." The final pacification of the highlands of Viti Levu did not come unitl 1876, but by that time the principal centers of resistance were well to the west of the Matailobau area. Still when Carew visited the Wainimala tribes to confirm their support of the govern- ment, he found them torn between the pledges of support they had given him at Navuso in 1872 and their ties to the resistors, the Nuyakoro people of the upper Sigatoka valley. Not only did they have ties of kinship to Nuyakoro, but the latter were also trading partners, sup- pliers of barkcloth and pigs to the peoples of the Wainimala 313 (Gordon N.D.:278). Nonetheless they remained firm in their support of the government and remained out of the conflict. Throughout the troubles of the 18705, the Matailobau were especially steadfast in their support of the government. Years later the chiefly families of Nairukuruku were to point to this in support of their request for a native governor or 3959, a Matailobau man, to replace the European Resident Commissioner in what was then the province of Colo East (C.S.0. 1904-3046). Following the subjugation of the Noimalu in 1874, leadership within the district of Matailobau and the oral record of succession to the office of paramount become complicated by the fact that there were now two competing systems of administration. 0n the one hand there were traditional chiefly offices, succession to which was determined by customary principles, and on the other hand, there were new positions of leadership within the Fijian administration of the Colonial government. The most important of these on the district level was the office of Buli Matailobau, or District Chief of Matailo- bau. Access to offices in the new administation depended upon govern- ment appointment, and the authority of the office holder was supported by the power of the government. The first bull, as well as some sub- sequent holders of that office, were drawn from the chiefly families of Nairukuruku. This no doubt bolstered the prestige of the office, but it also had the potential for engendering rivalry between the occupants of the traditional office and their younger kinsmen in their role as government chiefs. 314 Figure 24 illustrates the order of succession to the office of paramount chief, the office bearing the title of "Owner of the Flood." I am uncertain of the accuracy of Figure 24 with respect to the relationships among the earlier occupants of the office. Specifi- cally, questions surround the relationship between Ratu Meli and 2 on the one hand and between 2 and 3 on the other. As noted earlier, the necessary information was supplied by an informant who was not himself a direct descendant of 2. The only surviving grandchild of 2 was unsure of the relationship between her grandfather and Ratu (Meli, as were others whom I questioned. But several things are clear. The first is that 2 was playing an active role in local affairs even before the area was completely pacified. It was he who led the con- tingent of men from Matailobau who pledged their support to Major Harding, and it was he who was first appointed as bull of Matailobau. He was.to play a prominent part in local affairs for many years, resigning from government service in 1896, but reamining on the payroll until his death, probably in 1908. 11The second fact which clearly emerges is that people today interpret the relationship between 2 and 3 as being characterized by rivalry. It will be remembered that when the descendants of 2 were troubled by a series of sudden deaths, the adept whom they consulted interpreted the problem as retribution for 2's usurpation, as government chief, of the primacy of position due his elder kinsman, 3. It is unclear whether, as indicated here, 2 held the traditional office before 3 or whether, in his position as 315 . , .avgmmmwgu cu :ovmmwouzm mo emueoii.e~ meamwm moan .p mom“ .9 es a a a e----,,--a . a sieée e W :3. .x. m H 316 government chief, he preempted the position of the traditional chief, 3, succeeding to the paramouncy after 3's death. The first Bulj_ Matailobau, sometimes referred to as the turaga levu (the great chief) looms so much larger in the collective memory than the other man that it is difficult to reconstruct the order in which they occupied the office of paramount. Nor does Figure 24 represent the only account of their relationship which I was given. Another informant, himself not a member of the chiefly matagali, interpreted their relationship as one of siblingship (veitacini), and two informants referred to 2 as the elder of the two. For the subsequent office holders reconstruction of the order in which they occupied. the. paramouncy is a little more reliable, especially from 5 onwards. While the precise genealogical links are not known in every case, ties of reputed classificatory siblingship are indicated. The year of birth is known for 5 through the present office holder, and what these indicate substantiates informants' statements concerning the dominant principle of succession: the office of paramount is transmitted in order of seniority within the 12 Not since titled lines of the Nairukuruku branch of matagali Siko. the successor to Rt. Meli (1) has the office passed from father to son (or son's son if the account represented in Figure 24 is accepted). Given the uncertainties surrounding the genealogical relationships among the various office holders, not to mention differences in opin- ion concerning seniority within ancestral sibling groups, consensus about seniroity of descent among the various chiefly lives is elusive. 317 Informants told me that when the present office holder was selected, a meeting (base) was convened of the members of matagali Siko from all four villages in which it is represented.13 Apparently, in the discussion (veivosaki) that followed, one of the matters which was brought up was the transfer Of the paramouncy from Taulevu to Nairuku- ruku. The present office holder cannot claim to be descended from the ulumatua, the first born son of Rokowai, nor is he the eldest male of matagali Siko residing in Nairukuruku, but the only man older than he is originally from Taulevu, and is ineligible for the chiefship on those grounds. At one point in a conversation, the present paramount tapped himselfiuithe chest and said that he was the eldest "chief" in the village. By this he meant not only that he was the eldest member of his matagali (with the exception of the Taulevu elder referred to above), but also that he was the most senior in terms of descent. If the diagram in Figure 24 is accurate with respect to the relation- ship between the two lines referred to earlier as lineages A and B (the descendants of 3's younger brother and 2 respectively), then he is correct on both counts. Walter reports that in northern Lau, while there is an ideal of sibling succession to office, there is also a notion that the office should eventually return to the seniormost line of descent (1974:306-7). When sibling succession restores the title to the senior line, the Matailobau consider the outcome "pr0per" or fitting (dodonu), other things being equal. The difficulty lies in achieving a consensus about what constitutes satisfaction of the ceteres paribus clause. 318 Despite what was said earlier about the importance of senior- ity within the extended sibling group, choosing a new paramount is not simply a matter of comparing birth dates. If it were, there would be more agreement on the identity of the successor. Informants told me that the formal installation ceremony (veibuli) was not held for the last several paramounts, including the present 1 tagkei g1 ‘waluyg, In effect they have been ‘ge.fagtg but not denigre_occupants of the office. Informants attributed this omission to dissatisfaction and covert resentment about the eventual outcome of the selection process.14 In each case relative age seems to have been the deciding factor, but other criteria would have been appealed to as well. Seni- ority of descent is one such principle, but others include the personal leadership qaulities of the candidates and the related factor of support for their candidacy, both within and outside of the matagali (Cf. Nayacakalou 1975:42). I suspect that during the deliberations which led to the selection of the present paramount, one of the factors which was held against him was the fact that up until middle age, he had spent much of his time in his mother's (and wife's) natal village in the district of Wainibuka. While junior in terms of descent (at least in the version presented in Figure 24), for generations lineage B has con- tracted marriage alliances in the immediate area. If has provided two .9211 (district chiefs), and its members pride themselves on their tradition of active leadership in matters both governmental (yakamaf tanitu) and traditional (vakavanua). I do not wish to imply the 319 existence of a two-sided competition between lineages A and B, for in the immediate past other lines have sought and held the office as well. Nor do I wish to paint a picture of bitter partisanship. Quite the contrary. Though it constitutes thelargest such group in the village, matagali Siko is also characterized by solidarity and c00p- eration. Its joint undertakings are a model of effective organiza- tion. Nonetheless, while the sentiments of common descent are openly expressed ("we are all brothers, and he [the parmount] is our father") status rivalry exists below the surface. That is the nature of the system. On a number of occasions in the preceeding chapters, there have been references to what were called "traditional ties between chiefly and non-chiefly lines." The local term for the relationships thus ambiguously described is "veigalaki." I have not encountered this term in the literature on other areas of Fiji, though the institution itself is similar to the relationship between chiefly and nonchiefly lines which Nayacakalou describes for Nakorosuli village a short distance up river (1978:80-81). In Nakorosuli the commoner lines attached to chiefly families are refererd to as their companisons (i _l_a). , Despite the fact that the term "veigalaki" is clearly a compound of root plus prefix and suffix, so specific is its meaning that informants viewed the term as indivisible. Normally prone to etymological musings,in this case they offered no suggestions con- cerning the meaning of the root from which the term is constructed. 320 The root in question, "gala" appears to be equivalent to the Bauan 'gyyga," meaning "to minister to or attend." In the local dialect a phonological shift from "r" to "l" sometimes occurs so that, for example, "vakarokoroko" (respectfully) can be rendered as "vakalokoloko." Though my informants were certainly aware of this shift, no one sug- gested that it occurred in this instance: When asked about the meaning of "393,". informants chuckled. In the Bauan dialect "gala" means "scrotum." If this analysis is correct, then the local term is related in meaning to "veigaravi," whose primary referent is a formal welcoming ceremony, but which is also used in the general sense of "waiting upon" or "serving." The reciprocal formulation "yai_¥ gala [qara] + .51" conveys the notion of mutual ministration rather than one-sided servitude and is used only in reference to the type of relationship under discussion here. The outside observor's initial reaction to the role of the chief in this area of Fiji is to view him as primer inter pares. He is accorded the premier position in any public assembly. When he finishes eating a meal in public, all others clap respectfully. His presence guarantees that a yagaaa session will be conducted with decorum. Unless a guest of even higher rank is present, he drinks first and receives "aaba tgja" (three handclaps) when he drains his cup. When a mag_1'_t_i_ is offered to the community asawhole, a portion, aa 21;; aim (lit. "the head of the feast") is reserved for him 15 personally. Still, no great aura attaches to the chief. Though 321 they should address him as "aaka" (sir), younger men and women often neglect to do so. His position is rather that of a senior and respected kinsman. In many contexts the weight which his word carries depends not on his office, but on his personal prestige. But if from all of this, the outside observor concludes that the chief is really just the first among his equals, the leaders of the comwnity, the function and significance of chiefship have eluded him. In the deep- est sense, the chief is the keystone of village life, for the welfare of the conmunity depends upon the fulfillment of obligations of nutual ministration by him and his followers. When they offer and he accepts in the context of ritual prestation, each has fulfilled their role in maintaining the well being of the conmunity. Unless this is grasped, those instances in which chiefly groups have spread and their domi- nance has been accepted without conquest or armed agression, indeed through respectful invitation, remain incomprehensible. Such, accord- ing to tradition, was the nature of matagali Siko's coming to Nairukuruku, and the relationships referred to as veigalaki date from the earliest period after their decision to "stay with" the people of the area. The village of Nairukuruku was not always situated where it is today, nor was it the only settlement in the area. One of these other hamlets, Ulira, was situated across the river on Nakorowaqa land. Another concentration of population lay further upriver in the yaaaa_known as Naikakau. I have heard matagali Nakorowaqa referred to as "591 Ulira" (people of Ulira), and a Navitilevu man 322 referred to his own matagali and Nabubuco as 591_Naikakau. According to an informant, one of Ratu Meli's sons became what he called 19.12.9911. Ulira and another son aaigalgfl Naikakau. It is not clear whether they resided in these hamlets or whether they were merely the chiefly ljalja_(leaders) for the people living on or asso- ciated with those lands. What is clear is that matagali Solia, one of the patrilines within the local group known as matagali Nabubuco, ministered to (gaaayj) 22.1“322521 Naikakau, and matagali Nakorowaqa served in a similar capacity to aa_j.taakaj Ulira. These relation- ships have persisted over the years. The single household of matagali Solia continues to serve as leaa a1 gala ("contents of the household") or taaatahaj_yala ("persons of the hosuehold") to lineage A of matagali Siko and especially to its senior house, that of the para- mount. The households of matagali Nakorowaqa and of the other patri- line within Nabubuco serve in a similar capacity to the households of lineage B. Similarly, the ljalia of the aati, an untitled man of matagali Siko, refers to matagali Nawaita ‘asflewe _n_i_ vale to him and his line. Given the nature of Fijian society, it would be reasonable to expect that relationships of veigalaki would have been based upon intermarriage, specifically upon relationships established by the 3 marriage of commoner women and chiefly men. However, the genealogical data do not support this interpretation. The women of matagali who stood as l£!£“21.!212 to them would have served in chiefly households from time to time, but there is no evidence of formal unions between 323 them and chiefly men. It is interesting, however, that of the eight hypogamous unions mentioned earlier, two are between members of patrilines which stand in a veigalaki relationship to one another.16 Additionally, there have been two other marriages between partilines related as veigalaki, those between chiefly men and women of commoner matagali. The relationship between the jjaaawaj.yala_and their "chiefs," the chiefly households to which they are attached, is one of mutual obligation. The 1_e_w_e_p_j_y_al_e are obligated to assist (garayi) the chiefly line on ritual occasions. This may involve the preparation and serving of yagaaa_by the men and the cooking and serving of food by the women. The men of the laaauaj_yala_may also assist in the distribution of goods such as mats and taaaa_or‘aaaiti. For their part the chiefly households should provide taaaa, cattle, and even money should the lewe aj_vale formally reqest them to meet traditional (vaka vanua, "of the land") obligations. This obligation should proba- bly be viewed as mutual, but the emphasis is on the role of the chief as a provider. The households of the chiefly line are also under an obligation to fulfill the ritual role of the chief on behalf of the .laaa a1 gala (i.e., in the giving or receiving of magjti, tabua, yagaaa, etc.) When an elder of lineage B within matagali Siko died, some of the first persons on the scene were his laaa_aj_yala, members of matagali Nakorowaqa and Nabubuco. They washed and prepared the body, and throughout the four days of the funeral proper and the ten days 324 of initial mouring they were present in the house of the deceased and in a second reception house for the visitors who flocked to the funeral. Young men prepared yagaaa, middle aged men served it, and an elder sat as ma‘Ster of ceremonies in the house of the deceased. They also assisted in the preparation of aagjtj, both in the actual slaughtering of the animals and in the divisions that followed. Simi— larly, they assisted in the various "divisions of wealth" (aaaauai ‘yaa) in the form 0f.£9229: mats, and cloth which occurred during the mourning period. They also played an active role in many of the exchanges of yagaaa_and £29!E.that took place during the funeral. Hundreds of gallons of yagaaa_were consumed during the funeral and ten days of initial mourning, and the preparation of it was a constant chore. Much of this work was done by youths and men of matagali Dreketi of Navuniyasi village who also stand in a veigalaki relation- ship to lineage 8. (A highly educated informant said of matagali Dreketi that they are "liege" to lineage 8.) Detailing all of the services which the laae a1 gala performed during the funeral, burial, and mouring period would take us far afield, but it should be noted that their contribution did not end there. Exactly one year after the death of his father, the deceased's eldest son, who works in an urban center on the other side of the island, arrived with cement blocks, bags of premixed cement, and, I believe, a small stone marker. The next morning another son of the deceased came into a Nakorowaqa house as the family was sitting down to breakfast. He formally requested, in recognition of the relation- ship of veigalaki between them, that Nakorowaqa assist the deadman's 325 lineage in the construction of a grave cover and marker. The work turned out to be a community project involving the men and boys of most of the village's households, and the village was compensated by the presentation of a aagiti. But what I want to stress is that the formal request made to Nakorowaqa was phrased in terms of the veigalaki relationship. Another funeral, that of an old woman, provides an example of the reciprocal obligations of the chiefly households to their laaa_ aj_yala, The woman had been married to a man of matagali Nabubuco, but her natal matagali was Dreketi of Navuniyasi. It will be recalled that there is a relationship of mutual ministration between Dreketi . and lineage B of Siko. Moreover, the dead woman had been brought to the village by a man of lineage B for the expressed purpose of being the wife of the Nabubuco man whom she married. (I encountered several of these "arranged" marriages in past generations in which chiefly men more or less forced the issue upon commoners.) Throughout the funeral a senior woman of lineage 8, the brother's dauther of the chiefly man who had arranged the marriage, kept a vigil in the upper end of the _v_a_l_a _n_i_ __te_. When yaggya was served, she was given a prominent position in the drinking order. On several occasions when visiting groups presented their j_ regurequ (regu = kiss; a farewell prestation of goods given to the family of the deceased), they were received on behalf of the i taukei ai_mate ("the owners of the death," i.e., the matagali of the deceased) by this woman's brother's son. People stressed that her vigil was not just an expression of the 326 veigalaki relationship between her lineage and the dead woman's natal matagali, but more importantly it was in recognition of the fact that her "father" (i.e., father's brother) had brought the woman to the village in the first place. gage aj.yala_may refer to the senior male of the chiefly .L lineage to which they are attached as their lialia_(leader), a role which he may fill in a literal sense when both groups attend ritual functions in other villages. When a death occurs in any village in the area, ties of kinship and marriage are bound to exist between the deceased and someone in the village, and the death usually becomes a concern of that person's entire matagali. The ties of kinship and marriage between Nairukuruku and some other villages in the area, and with specific matagali within those other villages, are so manifold that when a death occurs, it becomes a concern of the entire village. Such is the case with respect to matagali Dreketi in Navuniyasi, Nairukuruku's nearest neighbor upriver. When word arrived that a young woman of matagali Dreketi had died, matagali Nakorowaqa and Nabubuco met jointly to discuss the property which they would present. When the number of taaaa had been decided upon, and the women of the two matagali had presented the mats and cloths they would be giving, the two matagali carried these to their chiefly ljalja, the senior house of lineage 8.. The senior man of matagali Nabubuco formally presented the gifts to the senior male of lineage 8. This man was suffering from the after-effects of a severe stroke, and so the goods were accepted on his behalf by his wife's brother, also an elder of matagali 327 Siko. 'Later the property amassed by lineage B was presented to the chief, the senior member of the Nairukuruku branch of the matagali. Thus when the contingent drawn from these three matagali--Nakorowaqa, Nabubuco, and Siko--attended the funeral, their joint offering was presented on the chief's behalf by a senior member of matagali Siko and accepted on behalf of the family of the deceased by the chief of Navuniyasi. In a way the relationship between laaa_aj_yala_and the chiefly lines to which they are attached are like those relationships between matagali of different villages referred to locally as "paths" (gala). The latter will be discussed in the next chapter. Like them, the veigalaki relationship provides or defines channels for formal inter- action between people of different matagali. But unlike the institu- tion of "paths" which provides linkages between villages, the veigalaki relationship is one of the things which structures and makes possible the "staying together" in a single village by people of. disparate "types" (matagali). FO0TNOTES--CHAPTER VIII 1Note that as it stands, this definition does not eliminate such universal features of social differentiation as categories based on age and sex. While it would be a simple matter to word the defi- nition so as to eliminate this difficulty, to do so would be to ignore the fact that in societies characterized by ranking, one or both of these factors may be stressed and, combined with other fea- tures, may be a principal determinant of rank. 2According to the position adopted here, Walter's analysis equates rank, an ordering of statuses on the basis of prestige, with one of the criteria upon which such orderings are frequently based (i. e. , seniority of descent.) 3Prior to Christianity the existence of polygyny complicated the picture. In addition to age and sex, the ranking of half-siblings was affected by the order in which their mothers were married. 4The mata ni vanua is closely associated with the presentation and distribution _of—food and in many places had the exclusive right of eating the chief's leftovers (Cf. Hocart 1913: 109). 5There are questions inherent in the wording of this definition. What is meant by "openly cohabiting with a man in a recognized union?" Clarification of this point will have to be deferred. 6Among the meanings listed in the Second Edition of Webster's New International Dictionary are "the head, top, or uppermost part; "the head or leader of any body of men;" and "principal part; most valuable portion." 7Cakobau had no reason to treat the Soloira people leniently, for it is believed that a Soloira village had given refuge to his arch rival, Qaraniqio, after the latter had fled Rewa during one stage in the epic war between that kingdom and Bau (Derick 1946:88n). In fact, the participation of the Waikalou in the attack on the Naitasiri village may have been related to this, for at one point Cakobau had persuaded warriors from Naitasiri to attack Qaraniqio in his refuge. They came under the pretence of presenting yams to the Rewan leader, but the next day at dawn, they attacked and killed 100 of their hosts. I do not know the identity of the Soloira village that harbored Qaraniqio nor of the Naitasiri villages that attacked it, but perhaps 328 329 the involvement of the Waikalou in the attack on Nakorovou was moti- vated in part by a desire to revenge this earlier treachery. 8Nakorosuli is the :highest ranking village along the upper Wainimala as far upriver as Laselevu and Wairuarua. Thus it consti- tuted a major rival to Nairukuruku, the dominant village of the lower Wainimala basin. Following pacification, Nakorosuli and its satellite villages became the district of Nagoneicolo. This and the old dis- trict of Matailobau, essentially the vanua of Nabena, were later combined in a single district, also known as Matailobau. 9Levuka on the island of Ovalau was then the capital of what was soon to become the Colony of Fiji. 10I believe that Nakorowaiwai was a village on the other side of the river a mile or so upstream from the present village of Nairu- kuruku. 11This is the year in which he ceases to be listed as the Buli Matailobau, retired, in the Blue Book of Fiji. 12Information concerning dates of birth and death is given in the 51 Vola £1 Kawa, the official registration of matagali membership. 13They are Nairukuruku, Taulevu, Navunidakua, and Nakorovatu. 14It is difficult to imagine :that': these reactions would have been absent in the past. Prior to pacification, when open expression of such sentiments might have led to armed conflict, overt expressions of unaminity could avert bloodshed. Today such considerations are less crucial. 15In cases where the magiti includes a slaughtered animal aa_ 21! 131 ma iti is not the actua ead of the beast. That is the portion reserved or the bati. 16Another three are marriages within avusa Siko between men of mata ali Nawaita and women of Siko. One 5 between a man of mataga i Nabubuco and a woman of Navitilevu with whom Nabubuco share a special relationship to be discussed in the next chapter. Though technically it is not a hypogamous union, it should be mentioned that a woman of the untitled line of the liuliu aj_bati is married to a Nawaita man. CHAPTER IX NEIGHBORS It is probably a common experience among ethnographers: for the first few months of one's stay in the community whose life one came to study, there is a reoccurring feeling of being the only person who does not know what is happening or is about to happen. This is often accompanied by a mental, and perhaps even physical, running to catch up lest one miss some unique, though crucially important, event. This is undoubtedly one of the major contributors to that deep fatigue the ethnographer experiences after the initial elation of arriving in a place one has waited so long to see. I certainly felt this to be the case, and it was not just a problem of acquiring a working knowledge of the language. It was partly due to the sheer number of community events that took place. Perhaps it was not that hectic to live through (though it did seem that sometimes a community-wide fatigue would set in), but it was sometimes difficult to record and analyze one event before being swept away by the next. Funerals, weddings, meetings, and competitive fund raising drives (gall) followed closely, one upon the other. It was only later, after pattern and predictability began to emerge, that I became aware of the fact that everyone did not know what was happening all of the time. In fact, knowledge of local events 330 331 was eagerly sought after, as if it were a scarce and valued commodity. This was necessarily a matter of knowing who was, or would be, doing what, when, and where, matters for frequent speculation and comment. Often this was not a matter of gossip, for it dealt with publicly observable comings and goings, actions, and interactions, rather than the intimate affairs of life. There is much to be learned about a society from the subject matter of its small talk, those brief, formalized interactions that fill the interstices of social life and make it run smoothly. For example, does it focus on impersonal topics like weather or more highly charged matters such as food and dreams? In a Fijian village the most common topics for such exchanges are sources and destinations, or comings and goings. A man sitting in the doorway of his house sees another passing by carrying a machete and wearing work clothes. The question "Bau vei kei T'("Where are you going?") might elicit a terse reply ("To our gardens" or, simply, "To the other side [of the river]"), or the second man may stop and chat awhile and in the process tell his questioner what he intends to do in the garden and why, and, perhaps, who will be helping him. Or men sitting near the village path may see a newcomer walking up from the river carrying a bag or bundle. "Lesu mai vei?" ("Where are you returning from?") The newcomer might answer the question briefly and hurry on, provoking speculation as to what he had in the bundle and why he was in a hurry. But he is more likely to say, not only where he has been, but also who he saw there, what he and they were doing, and so on. 332 Sometimes these questions (Where are you going? Where are you coming from?) are meant to elicit specific information, but more often than not, they are simply a way of politely initiating an inter- action that may be briefly concluded or expanded into a chat. When these questions about comings and goings are asked within the confines of the village, the response may simply be "the other side." The side of the river on which the village stands, léééflénflé (this side), is contrasted with the other side where the principal gardens are found. Implicit in this is the contrast between the domestic sphere of the village and the uninhabited, yet cultivated (i.e., culturally trans- formed) sphere of the gardens. To say that one is going to the other side is tantamount to saying that they are going to their veiwere (fields). There is another spatial dichotomy which is also frequently utilized, and, again, it is made in reference to the river. By con- vention the village is bisected into an upper and a lower half, and these disignations are assigned in reference to the river. The gala or "upper" end of the village is upstream; the ra_or "lower" end is downstream. I want to consider now the implications of this spatial division for village organization. The map in Figure 25 illustrates the location of houses and other features within the village. It will be noted that there is a clustering of houses by matagali affiliation. If one looks at the houses "below" (i.e., downriver from) that imaginary line which bisects the village, it will be seen that they belong to members of 333 33203 9331!...» \‘H? I. s r! \ AV «55530660.. is): u as igeoocnwuuanfilaemunc _ 5.3%: e... e. .emep_a> we» to aeg--.m~ deam.a 8.53.: :83... m . 3.53:» 283.! m.» 35833. Inseam: I y 3.5... :Meuwez I 9.3. :38: D 2&1 5., . WW; wkbfivrvo Hug e! . ... . o .. i... as ...... m \\Illl l oi 05.00339qu 3» O r c \ mo 2; e _ chained .05! s so» W -\ lmfl_.%w.k§w , 4M. .9! .Uh/ h 0 \~ . *. “fl . as... ......3 man, .m. mm... 4% . W. W a cod-u. a 25. and 1 m in: o 75 w ” influcfiibdauwuu-nvifl (fog m D I n”: *‘rlo & gig «a "C“ _ w , (a 9 within. own-.5311. wank Bur-r1 1111111111 1111 1 11 I “ohm-$3.. ... 9.... " hm Wee®e ..- e . 4 mm _ m s U D .. . e . .. m. a” U D $ 5.? n— _ I . an... m . a so/ I a ..- . w... ..Upon/ m. .883 5.3:. . _ . . . 3:33.3- .a¢u.33h (m - 01—90 .1119 .5 a flag TII . 48353:?» n m 8 .unfluuflo/ouo . m u . 334 matagali Siko and Nakorowaqa exclusively. One Nakorowaqa house more or less straddles the line, as do two houses owned by members of matagali Nabubuco. All of the houses of matagali Nawaita and Vusovuso and a majority of the houses of Nabubuco lie above that line. The four houses of matagali Siko near the outskirts of the upper end of the village were built there because of lack of space in the lower half of the village. . There is a tendency for the village to segment into lower and upper halves which are, of course, largely a reflection of matagali membership. One context in which this division emerges is in visiting patterns. For example, one seldom sees men from the upper half of the village at ygggga sessions down at the lower end. Given the fact that the houses of matagali members are clustered, this tendency to visit within one's own half of the village might appear to be a reflection of matagali affiliation, perhaps in conjunction with uterine and affinal ties. But locality is definitely a factor also, for there are men of matagali Siko living at the upper end of the village who seldom drink at the lower end, and there are men of matagali Nawaita with affinal ties to Siko who are also seldom seen there. Another context in which the distinction between iupperi and "lowerf occasionally assumes significance is in the organization of communal work. Some tasks are divided between the two halves of the village. For example, on one occasion, the men of the village were to build new cookhouses and latrines for the families of the headmaster and one of the school teachers. The work was to be accomplished over 335. a two-day period. On the first day the matagali g 22: matagali Siko and Nakorowaqa, were to build the outbuildings for the school teacher, and on the second day the matagali g 9919 were to do the same for the headmaster. As it turned out, men showed up on which ever day suited them--some on both days and, no doubt, some on neither. But the case does illustrate that the territorial division is a latent principle of organization. For several months during our first year in the village, this dual division assumed primary importance. It did so in connection with an effort to raise money to build a new district school. A competitive fundraisjng drive was planned which was to involve the selection of several Adi Dauloloma or Pcharity queens,f each representing a village, or division of a village, served by the district school. Such fund- raising events doubtless occur elsewhere in the:world, but the ."Ldi festival" has a peculiarly Fijian character. Each of the participating communities or groups nominate an fiadi? to represent them, usually a young girl. On a predetermined day the groups meet and give competi- tively on behalf of their queen. As their supporters file forward to make their contributions, the girls sit "in state? wearing special out- fits purchased for the occasion. The contributions are totaled and, from time to time, announced to the crowd. At this particular 391 festival a certain amount of strategy was involved; large sums were withheld until just the right moment in order to produce the maximum effect (i.e., dishearten competitors and rally supporters). On this particular occasion six agj_and their supporters were to be involved, one each for the villages of Taulevu and Navuniyasi, 336 two for Nairukuruku proper, and two representing Nairukuruku's "expatriates,“ persons residing in Suva and in urban centers on the western side of the island, respectively. As it turned out, two weeks before the festival was to be held, the country was hit by a hurricane. Heavy rains triggered mud and rock slides which disrupted traffic on the main road linking Suva with the western side of the island, and so on the appointed day those living and working in Nadi and Lautoka were unable to attend. Even without the attendance of some of the tamata cakacaka (wage earners) around F$5000 were raised on that day. Those who had been unable to attend did so gg_m§§§g_at a later date and presented the village with their contribution and were presented a magiti_in return. In anticipation of this event it had been decided in a public meeting more than three months earlier that the residents of Nairuku- ruku were to be divided into two groups,each of which would select an adj, This was deemed necessary because Nairukuruku is so much larger than either Taulevu or Navuniyasi that the outcome of the competition would otherwise be a forgone conclusion, and without the element of competition, contributions would suffer. Now there are several ways in which the community might have been divided. It might, for example, have been divided along yavusa lines. That is, the two matagali of yavu§g_5iko, Siko and Nawaita, might have united in opposition to the other local groups in the village. Instead, however, it was decided to divide the community on the basis of the opposition between £919 and £2: That is, matagali Siko and Nakorowaqa were united iri opposition to matagali Nabubuco, 337 Nawaita, and Vusovuso, though it should be stressed that this oppo- sition was spoken of in territorial terms. Following this decision there was a noticeable split between the two halves with respect to group activities. Indeed, two days later, matagali Siko and Nakorowaqa met to select a committee to guide their fundraising efforts and to discuss joint activities for generating income. It was agreed that each house of the two matagali should be assessed F$25 as an initial contribution, and it was publicly decided how this assessment was to be divided among the various house- hold members.1 Discussion then turned to other ways in which money might be raised. The most important of these turned out to be coopera- tive garden work for hire, both for members of the two matagali and for outsiders. This took place during February and March. Even though taro is not a seasonal cultigen, this is a busy period for planting after the heavy consumption and light in-puts associated with the holiday season. In addition to their joint money-making ventures, there was an intensification of socializing among the men of the two matagali at the 53 end of the village and a corresponding drop-off in visiting at the upper end of the village. Since I was affiliated with matagali Nakorowaqa and my wife with matagali Siko, I was expected to conform to this pattern. 0n the couple of occasions when I drank yaggng with men at the upper end of the village, it provoked joking comment, but the message was clear. I was expected to show solidarity with the people I fstayed with." 338 The frequency of interaction among men at the 53 end of the village was partly due to the fact that each instance of joint work was concluded by a yaqona, but it was manifested in other contexts as well. For example, the two matagali ate and drank together on various "tenth nights? (twentieth night, thirtieth night, etc.) following the death of a prominent elder of matagali Siko. Moreover,in memorium, both matagali were observing tabu for the duration of the hundred-day mourning period. With the exception of elders, who wore black ribbons pinned to their shirts, men were forbidden to shave or drink alcohol, and women were required to wear a black ribbon pinned to their blouses. 0n the hundreth day after the death, the two matagali were to meet for the lifting of the tabu, On that occasion each man was to bring , one tabua or one drum of kerosene, plus one bottle of rum or case of beer, and each woman was to bring one mat or a bolt of cloth. When I asked people why the village had been divided as it was, obstensibly on territorial lines, they pointed out that the com- bined population of matagali Siko and Nakorowaqa balanced that of the combined population of the other three matagali. But the same approxi- mate balance would have been achieved had Siko combined with Nawaita or, for that matter, Vusovuso. There is, of course, the fact that there are ties of marriage and uterine kinship between Siko and Nakorowaqa. In fact, the chairperson selected for the Nakorowaqa com- mittee was a Siko woman married to a Nakorowaqa man, and it was their daughter, £53 to matagali Siko, who was selected to represent the two matagali as their "adj," But there are also affinal and cognatic ties between Siko and Nawaita, and, moreover, the two are yavusa vata. 339 It seems reasonable to conclude that it was territorial proximity which was salient in the opposition that emerged.2 In the past, the localization of matagali within the village was more systematic than it is today. The blurring of matagali "wardf boundaries which has occurred is primarily a response to population growth and the consequent scarcity of suitable house sites in the heart of the village. This localization of matagali can be seen as an extension of what is thought to have been an earlier settlement pattern, one of dispersed fclan hamlets.9 We have discussed the major territorial divisions in terms of an opposition between upper and lower because these are the terms which village residents use to describe it, but it might also be thought of as an opposition between “inner" and fouterf when viewed from the persepctive of the chief's household. In a sense, the lower end of the village is simultaneously public and sacred. It is no accident that the church is located there close to the paramount's house and the houses of other chiefly families whose past members have served as chiefs. The raga, or village green, is also located . there, and while it functions as a playground for school children, it is also transformed from time to time into the site for public cere- monial. One exhibits decorum when walking at the extreme lower end of the village. Men often walk with one arm behind their back, their hand grasping the other arm at their side, for it is disrespectful to swing one's arems while walking near the chief's house. As already noted, the houses of matagali Siko are clustered in the lower end of the village and have been expanding fupwardsf and 340 outwards as the population of the matagali has increased. In close proximity to them and to the inner sanctum, as it were, are the houses of their tgmgtg,nfl_vale, mataqali Nakorowaqa and, beyond them, matagali Nabubuco. (It is interesting that people speak of the locus for the ritual function of matagali Nakorowaqa as lying inside the chief's house and that of Nabubuco as lying outside the house.) On the outer edge of the village lie the houses of the bati, the fteeth," "edge," or "border" (all of these are meanings of the term), matagali Nawaita and, with them, matagali Vusovuso who had Fstayed withf Nawaita in the past and have continued to intermarry with them. Viewed in this way space within the village reflects not only a dual opposi- tion, but a layering or progression from inner to outer, from sacred to profane, and from passive to active. It is possible, of course, to look with an even finer focus at the structuring of space within the village. For example, if one notes the placement of the various watertaps and keeps track of the households which utilize them, it will become clear.that these clus- ters correspond closely to networks of maximum interaction. The taps provide water not only for cooking and bathing, but also for the wash- ing of pots and pans and, often, for the washing of clothes as well. (From time to time the chief or Village chairman will have the turaga .niwkgrg call for women to do their laundry in the river, but after observing the directive for a time, women will do their wash in one place or the other as it suits them.) Thus the taps are a meeting place for women and girls especially. 341 Of course, the placement of the taps is not fortuitous, and the clusters of heuseholds that they serve are bound together by more than a common water source and by more, too, than sheer proximity. Despite the title of this chapter, relationships based on propinquity alone do not constitute a distinct and meaningful category. One does not relate to another person as a neighbor or fellow villager only,‘ but as a kinsman of some specific category with whom, or near whom, one lives. This is not to say that propinquity, and especially village membership, are unimportant, quite the contrary. But relationships that are based on neighborhood are also structured by kinship. The tie may be distant or even fictive, but it is the domain of kinship which provides the moral content of the relationship. Given the previously noted tendency for clustering on the basis of matagali affiliation, a tendency for propinquity among male agnates is to be expected. What might not be expected is the fre- quency with which affinally linked households are directly adjacent to one another. The latter tendency can be illustrated by reference to two clusters of households. Cluster 1 (see Figure 25) consists of four households, two of which are headed by men of matagali Siko and two by mean of matagali Nakorowaqa. The head of household 18 has never married. He lives with his aged and nearly blind mother, and their house functions as a huge or sleeping house for the young men of the village, including those of household 4. (In keeping with the pattern of brother-sister avoidance adolescent males should not sleep in the same house with 342 their sisters.) The head of household 18 frequently visits and eats his meals with household 4. He and the head of household 4 are related as FBS, and as parallel cousins they stand in a classificatory sibling relationship to one another. Additionally, the fathers of these two men were close and named their sons after each other. Thus the head of household 4 is named after the other man's father and vice versa. In addition to household 4, the head of household 18 interacts regularly with the other households of the cluster, but especially with household 7. The head of household 4 married a sister of the head of house- hold 5. Thus the children of household 4 are vg§u_to household 5, and their mother is the true gel (FZ) of 5's children. Because their relationship to the head of household 5 is one of respect, the child- ren of 4 spend less time in 5's house than his children spend in theirs, for as their true father's sister, the wife of 4 is a warm and indulgent figure in the lives of 5's children. In addition to daily visiting by adults and children, there is a good deal of economic cooperation between the two households ranging from mutual exchanges of child care to occasional assistance in the gardens and in house- building and repair. Small gifts of cooked food are also frequently exchanged. One of 4's sisters is married to the head of household 7. The. interaction beweeen households 4 and.7 is perhaps less intense than that between 4 and 5. Nonetheless, there is interaction on a daily basis, though largely on the part of women and children. Still, the men of the two households are frequent drinking companions. 343 The heads of households 5 and 7 are classificatory siblings, and though his sibling group form a distinct line within matagali Siko, 5 functions as a member of 7's tokatoka for the purposes of ceremonial exchange. There is a good deal of cooperation between the women of these two households, and at one time they cooked their families' meals together for several weeks. I do not wish to imply that this cluster of households forms a discretely bounded unit in any sense. Of course, each household in the cluster has close ties to other households in the village. For example, 18 frequently visits at the house of his married sister, household 33. Similarly, 4 interacts on a fairly regular basis with the husband of another married sister (17), and the children of that sister by an earlier union are frequent visitors in his household. Finally, the head of household 7 and his wife interact on a daily basis with the household of his unmarried younger brother (8) and their widowed mother. But I do want to stress that it is proximity which, in addition to kinship and affinity, provides the basis for daily interaction among the houses of the cluster. I would like to stress also that the relative positions of households 4, 5, and 7 predate the affinal ties which presently exist among them. It should be pointed out, too, that the patriline of household 4 is 1242213143. to the lineage with which both 5 and 7 are affiliated. The second cluster to be analyzed reveals much the same pat- tern, but in this case it is composed of households of three different matagali. The head of household 10 (see Figure 25) is the senior 344 member of the patriline known locally as matagali Solia, one of the two lines, linked by cognatic kinship, which compose the local group, matagali Nabubuco. The head of household 10 is a man of, middle age, too young to function as an elder. He is a frequent visitor in household 26, that of the senior male of matagali Nabubuco, a man whom 10 calls tgxglg_(cross cousin). But the members of his household interact even more frequently with the households . of two of his married sisters, households 11 and 14. There is the same pattern of visiting between his children and his sisters' house- holds that were encountered in cluster 1, as well as the same patterns of cooperation. Indeed the head of household 11 depends more heavily on his wife's brother and his other tavale in matagali Nabubuco than he does on other households of his own matagali. It is worth considering the position of household 14, for as a member of matagali Navitilevu, the head of that household is not .i.£é!§§i in this village. When questioned about his affiliation, he pointed out that his tutua (father's father) had married a woman of matagali Siko from Taulevu village. Thus his father had been va_su to matagali Siko. Moreover, his father had been born in Nairukuruku, as had the informant himself. His (the informant's) mother was the sister of the head of household 26. The latter is his true memo, and the informant is lagu .to matagali Nabubuco. He had planted on Nabubuco land for almost thirty years, but recently his gardens there had to be abandoned because of the depredations of cattle, and today he plants on Nakorowaqa land. His ties to matagali Nabubuco are manifold; 345 not only did he and his father marry Nabubuco women, but his sister is also married to a Nabubuco man, the head of household 16, her MFBSS (=MBS). As was the case with cluster 1, the relative positions of households 10, 11, 14, and 16 predate the current affinal ties among their members. But in this case the association among the four patri- lines involved is based upon affinal ties in previous generations (see Figure 3, Chapter VI). The same pattern of affinal ties among adjacent or nearby houses is repeated elsewhere 'hi the village. Of the 22 households in which there are unions that are endogamous with respect to the village , 10 (z 20% of all households) are adjacent to, or no more than one house away from, the household of the wife's married brother (9) or the wife's father (1). In all ten cases, the wife's agnates are affiliated with awn different from that of the hus- band. Another 4 households are immediately adjacent to the houses of the husband's mother's brother, and in two cases the mothers' brothers's matagali is different from that of the sister's son. There are two conclusions to be drawn from this: not only does the frequent interaction encouraged by propinquity seem to promote affinal ties between households, but, when possible, there is also a tendency for affines to build homes near one another. Given the tendency for affinal ties to 'be repeated in subsequent generations, the association between propinquity and affinity may also be repeated, giving spatial expression to the importance placed on ties through women. On one occasion while returning from a funeral in a village. some distance upriver, the boatman, himself a resident of Nairukuruku 346 though not born there, compared the village we had just visited to the one to which we were returning. He said of both of them that they were 5259 lgsa, contented or peaceful villages, and contrasted them with others in the area. One of the things which produces, as well as reflects, Nairukuruku's contentment is the relatively smooth admin- istration of village affairs. Nairukuruku is a community which works. One manifestation of this are the various material improvements which the village residents have made in their community. There is a large church of cement block construction; there is a cement walkway down the middle of the village and cement steps leading down to the water's edge at both ends of the village, not an insignificant amenity in such a wet climate; there is a well-run cooperative store which supplies sugar, flour, tea, rice, tinned fish, kerosene, and other items at low prices while consistently paying its members a dividend at the end of the year; there is a small diesel generator, financed partly by a government grant and partly by the villagers themselves, which supplies cheap electricity from sundown until ten o'clock each evening; and thre is piped water supplied by a small stream in the hills above the village. Most of the labor involved in the construction and mainte- nance of this infrastructure was, and is, supplied by the villagers themselves. Village affairs are overseen by a set of five committees: the village committee (kgmfijj_njflkg§g); the school committee (kgmiti El koro vuli); the cooperative store committee (komiti n1 sitoa); the church committee (komiti ni lotu); and the committee responsible for 347 the financing and maintenance of the electrical system (kgmfitj_nj_ livaliva). The most important of these is the village committee whose chairman is referred to as the village chairman (jjgmanhnj_kgrg). The village committee deals with any matter which effects the running of the village as a whole. It meets infrequently and much of the day- to-day work of village management (e.g., the scheduling of communal labor) falls on the village chairman. In effect he is often carrying out decisions made by theparamount. When some decision enacted by the committee, its chairman, or the chief has to be communicated to the entire village, either the chairman or the paramount tells the ‘turagguniukgrg (literally, 'village chier) who calls it out in a formalized manner at various points in the village. During the hotter months from November through March people awaken early, and often the ‘tgragauniukggg began his rounds, in full voice, at about six o'clock in the morning. In exchange for his services the turaga_nj_kgrg_is entitled to a small monetary compensation or to community assistance in the planting of a garden. During our stay in the village, the men of the community supplied the suckers and labor for a garden of 300 to 400 mounds (huge) of tavioka for a new turgga_njflkgrg_who had held the position for about three months at the time. (The monetary value of the labor and suckers plus the market or subsistence value of the garden's projected yield was far greater than the cash stipend of F$6 per month.) It is interesting that the post of turagaflnj‘kgrg, so impor- tant in the management of village affairs under the colonial 348 administration, has been reduced to little more than a town crier. Prior to independence, the tgrgga_nj_kgrg, as the government's repre- sentative on the village level, was responsible for organizing communal labor (e.g., housebuilding and repair) and had the option of reporting malingerers to the magistrate. Since they interferred most directly with their everyday affairs, it is not surprising that the regulations concerning communal work were one of the least tolerable aspects of the colonial system. And it is perhaps understandable that the role of the turaga £1.52E2 has been redefined and his former duties absorbed by the village committee and its chairman. The chairman, as the representative of the village committee and the chief, must rely on moral persuasion rather than the threat of legal action to secure participation in communal work. Though the job can involve time- consuming, unscheduled work of a trouble-shooting nature, the village chairman does not receive compensation nor do the chairman of the other four committees. According to the village chairman, it is his responsibility to tell the turagaflnj_kgrg_what to call out in matters that deal with the business of running the village while the paramount determines what will be called out with respect to matters Vof the landV--that is, matters regulated by custom. In fact, however, I would see the tgraga_ n1 kgrg emerging from the paramount's house more often than the chair- man's just prior to calling out the message (i tukutuku) regardless of the nature of the matter at hand. The village chairman would appear to have a certain amount of discretion with respect to the question of what is, and what is not, 349 to be a communal project. When a Nawaita man married to a Siko woman decided to rebuild his house, he asked the village chairman to have it called out by the turaga.nidkgrg:-in a word, to be treated as a village project. This was not a formal request (kerekere) "in the manner of the land" (vakavanua). and no yaqona_was given. The rebuild- ing of a house is a major undertaking involving the replacement of the bamboo walls and the replacement of any of the hardwood posts which have become weakened. In exchange for village participation, the man promised the village a magiti in the form of one cow, and it was understood that yaqona_would be provided after each day's work. The turaga_nfi_kgrg called the work, and the Nawaita man carried out his end of the agreement. This provoked a certain amount of grumbling by another. man who had recently rebuilt his house. The village chair- man had not treated that as a village matter. This probably made very little difference because a large percentage of the households in the village participated in the work anyway, and in exchange for their services they received a meal.and yaqona each day and a magiti of one cow when the job was completed. Perhaps the difference in the way in which the chairman handled the two cases is to be found in his relationship to the first man and his matagali. Not only is the chairman the liglju of the bati, mataqali Nawaita, but his daughter was also. married to the Nawaita man's younger brother. Moreover, the wife of the Nawaita man was the daughter of chairman's namesake. I would say that, in the light of these relationships, the second man's grumbling, which was very mild, had a decided air of philosophi- cal acceptance. 350 Among his other duties the village chairman claims the responsi- bility of keeping the peace within the village. On this point the convergence between his modern role as chairman and his traditional role as the leader of thenbati is interesting. There were no very serious breaches of the peace while we were in the village, but, from time to time, young men or boys were publicly disciplined for misbehavior. On the occasions which I witnessed, such disciplinary proceedings, they occurred in the context of public gatherings ini— tially devoted to other matters. One such instance was a village meeting held on the day after our first Christmas in the village. The meeting was being held to discuss preparations for the flag: Dauloloma“ fundraising drive already discussed. On Christmas day there had been a 5911, a competitive fund- raising effort on behalf of the electrical system. It pitted the women who live in the village against those who had been born there, but who live and work elsewhere. Prizes were awarded to the winning side and to the highest individual contributor. The matter was hotly contested, and the fun continued with an impromptu dance that con- tinued into the night. At one point a disturbance was created by four young men who had apparently been drinking beer. On the following day, toward the end of the meeting, the matter of the disturbance was brought up by a prominent woman of chiefly rank, one of the organizers of the 5911. Presently the para- mount's son went out and, making a circuit of the village, publicly called for the young men to appear before the assembly. One by one 351 the young men walked in, silent and downcast. One was accompanied by his father and his father's elder brother.- The fathers of two of the others were already present as was the elder brother of the fourth youth. The young man sat quietly with their heads hanging as the secretary read the minutes of the meeting back to the assembly. When the formal business of the meeting was concluded, the young men were called before the assembled elders sitting in a ring at the upper end of the house. The offenders sat there, bent forward and looking at the floor, as a series of elders lectured them on their wrong doing in front of the entire assembly. The paramount was the first to speak. He told them that their parents, as well as everyone else in the village knew about their drunkeness and their foolish behavior. The other speakers reiterated the point, some with greater sterness, others more quietly. When the elders had finished, one of the young men left, two others retired to the very back of the house, and the fourth began to serve as cupbearer to the elders who resumed the drinking of yaqona, (The next day the same young man served yaggna_to the guests assembled in the house of the woman who had first voiced the complaint.) Most of the other occasions on which the forces of social control had to be brought publicly to bear were similar, and the actions which provoked them were also of like kind. In each case, representatives of the community lectured the offenders, detailing the nature of their wrongdoing, pointing out the shamefulness of their actions, and exhorting them not to repeat their offences. It 352 is acutely embarrassing to be made the object of public censure in this way, but more important than feelings of shame, which are not publicly observable, are the outward manifestations of shame, coupled with efforts to avoid repeating the offense. I did witness one case in which the offenders, young boys, were threatened with a whipping. Their offense had been to encourage a fight between two horses in which one of the animals had been wounded. The owner of the injured animal, a young married man, was asked if he wanted the boys to be punished in this way, and he quietly rejected the suggestion. I suspect that the whipping was really just a threat meant to drive home the message of the stern lectures which they received. Such threats of public punishment would be inappropriate in the case of older youths or men. As for the behavior of young women, the problem of social control is largely a familial matter, and, in fact, it is frequently their brothers who take responsibility for their actions. Perhaps the most significant reason for the success of the committee system is that an attempt is made to include repesentatives of each of the five local groups or matagali among which the various households in the village are divided. This means, in effect, that the two halves of the village, £919 and [1, are also represented on each committee. Additionally, with respect to the school committee, members are drawn from each of the three villages which the district school serves (i.e., Nairukuruku, Taulevu, and Navuniyasi). The effort to distribute the responsibility for managing village affairs 353 equally extends even to the unloading of supplies for the cooperative store; the five matagali do this in rotation. Also significant is the fact that, in the case of three committees--the village committee, the school committee, and the store committee--the chairman are members of matagali Siko, not only the largest of the five mataqali, but also the chiefly matagali. The committee which manages the electrical system is headed by a man of matagali Nawaita, and thus a member of ygyu§g_Siko, married to a woman of matagali Siko. The church comv.. mittee, which should probably be treated as separate from the com- mittees which handle the secular affairs of the village, is headed by the minister (1 talatala). All of the committees which manage village affairs are exclusively male. The members of the various secular committees are selected at public meetings open to all the adults of the village. For example, the school committee was reappointed at a general meeting held in conjunction with the convocation ceremony at the end of the school term. When I say that representatives are "selected,f I mean just that. There are no competing nominations and no show of hands for or against an appointee. A senior man presents the nominations as a motion, and those present voice their assent. Doubtless, if there were strong public opinion against a nominee, it would be voiced, but I never witnessed such an occurrence. On one occasion, village repre- sentatives were to be chosen to attend a national meeting of the Fijian Association. A senior man of chiefly rank nominated two men, one from matagali Nakorowaqa and another from matagali Siko. The 354 households of the three men are roughly adjacent to one another, and there is frequent interaction among their members. The money needed to send the two men was to be supplied by the half-brother of the Nakorowaqa man in return for village assistance in the fencing of his gardens. The assembly, in this case all men, voiced their assent. I believe that their approval was given, not so much in deference to the senior man who had made the nominations as it was in recognition of the way in which the trip was to be financed and because there was no strong feeling against the men going. It is not really necessary here to consider the workings of the various committees in detail, for this is not a political study. But it should be pointed out that the affairs managed by each of them also involve the efforts of various functionaries, some of whom are compensated for their labor. In addition to the turagaflniflkggg and village chairman, whose roles have already been discussed, the management of village business also involves a village treasurer (gag EIMXEEJ who is responsible for depositing funds under the juris- diction of the village committee as well as church funds. But it is the store committee which routinely handles the largest amounts of, money. In addition to a chairman, the store committee also includes a treasurer responsible for collecting and tallying the daily receipts and depositing them in the cooperative's account. But the major responsibility for the functioning of the store falls on the store keeper. In my opinion, his is the most demanding of all public posi- tions (though others are unwilling to admit this) because it requires 355 his presence for two hours in the morning and two hours in the even- ing everyday throughout the year. People tend to buy things in small quantities as they need them, in effect, on a daily basis. It is not the morning hours which are most onerous, even though they begin a 6 o'clock, an unpleasant time to rise in the cooler months. It is the evening hours which are the most disruptive since they require that the stdrekeeper leave ygggng sessions just as they begin to liven up and return two to two and a half hours later with a lot of ground to make up. In exchange for his services, he is entitled to F36 per month or a garden. From time to time issues arise which require a general village meeting. The planning session for the ”adi" fundraising drive discussed earlier is a case in point. Unless the topic under discussion relates to a specifically male domain, such as the meet- ings devoted to discussion of the agricultural development scheme, women not only attend, but sometimes voice, their opinions in strong terms. But women sit towards the batk or Plower' end of the assembly, and men sit towards the front, their position determined by relative age. One of the matters which regularly requires village-wide discussions is the financing of the electrical system. The electrifie cation scheme is the pride of the village. The generator and the building in which it is housed are spotlessly maintained by a man who is also responsible for starting and shutting down the generator each day. Another man does all of the household wiring. Depending on the 356 size of the house this entails one or two light fixtures and outlets. In exchange for their labor, both men are entitled to a small monthly stipend. A The installation of the system was expensive, and there are also the regular expenditures for fuel to run the generator, as well as the occasional costs associated with maintenance and the installa- tion or replacement of wiring and fixtures in houses. Each household is assessed F$2 per month toward meeting these costs. In 1980 a drum of deisel fuel cost about F$66, and the generator requires slightly_ less than two drums per month. There are 49 houses in the village, but a number of them are vacant, their owners having either died or are living elsewhere. Thus, there is a monthly deficit in excess of F$30. This is aggravated by the fact that not everybody pays the monthly rate on time. A three-month grace period is allowed, after which the power to delinquent households is shut off. The operating deficit, as well as the village's share of the initial cost of the system are made up through periodic s911_and other fund raising efforts. One such sell, which netted over F$500, pitted the various matagali against one another. When all the money had been collected, the committee members in charge of the electrical system were called forward and were formally presented with the money. They, in turn, made a counterprestation of yaqona to the assembly. As yaggna_was drunk, the 5911 gave way to a general discussion of the financing of the electrical system and other issues of concern. It was suggested that the cash flow problem could be amelior- ated by shortening the grace period to one month. Households would 357 have until the middle of the next month to pay the rate before losing their power. The motion was brought to a vote and was passed by a show of hands. Discussion then turned to other matters. The question was raised as to whether the minister should be required to pay the rate for his residence. It was decided to exempt his household if he would pay the rate for the church for one week in every month. The village chairman then raised the matter of the poor turn-out for village work such as the cutting of grass in the church yard. A prominent woman of chiefly rank pointed out that the church itself was in need of a cleaning and suggested that the women do something about it. This provoked one wag to comment that the women probably fished too much anyway. (The church was cleaned the following day, and the woman who raised the matter supplied refreshments for the work party.) One point which provoked total agreement concerned one of the vacant houses for which the rate was never paid. From time to time the house was used by the owner's younger brother and his family as well as the owner's aged mother. These people thus had the use of the electricity without paying for it. The matter was brought up at another gathering the next day, and the younger brother was publicly lectured by several chiefly elders, including the paramount. At one point he tried to say something in his own defense, but was shamed into silence. He then served several rounds of yaqona, strain and humiliation showing on his face. 358 Still to be dealt with was the main issue of how the continued deficit was to be met. The price of fuel had doubled since the gen- erator was installed and could be expected to continue to rise. One man suggested that the monthly rate be raised to F$3. His elder half-brother argued against this and favored the retention of the current rate with the deficit to be made up through periodic sell, ' The younger half-brother pointed out two arguments in favor of the higher rate. The warranty period for the generator would soon expire, and the village would then have to bear the full cost of any repairs. Moreover, there was the matter of periodic repairs to fuse boxes and wiring in individual houses. IThe lights in the vakatunuloa where the meeting was being held were hooked up by an extension cord to a nearby house. Almost on cue the circuit breakers were activated, and smoke could be seen pouring out of the house. But in the end, the elder brother's motion carried the day; the F$2 rate would be retained. The issue had generated some heat, and the meeting was concluded by long benedictions offered by three different people. More ygggna_was sent for, and the meeting was transformed into a "grog? session that lasted several more hours. The real issue here--one which reoccurred throughout the even- ing--was the manner in which the costs of the system should be dis- tributed.. Should it be shouldered in equal measure by all, or should there be allowances made for differences in ability and willingness to pay? The latter alternative, represented by the periodic §_Q_l_i_, (was preferred because it conforms to the cultural pattern in which those 359 who can afford to give more do so and are awarded with prestige. Often that prestige is a thing to be won on behalf of a social unit through competition with others like it. There was also the considera- tion that periodic s911_would draw upon the earning power of tgmggg; cakacaka (wage earners) living elsewhere, permitting them to demon- strate allegiance to their natal village. To me, the easy-going efficiency with which the village was run was most noticeable on those few occasions when large number of visitors were present. Of course, the village swells with visitors at Christmas time when family members living elsewhere return home, but such visitors are the responsibility of individual households. But on several occasions during our stay, the population of the village was more than doubled, and the feeding and housing of the visitors was a community concern. On one instance there was no ques- tion of planning beforehand, for the visitors had come for the funeral of a chiefly elder who had died suddenly. On another occasion there was a major meeting of the Wesleyan church drawing people from all of the villages in the taba cakacaka (diocese) of which Nairukuruku is the seat. On that occasion several hundred visitors were fed and housed for several days. A large arbor (vakatunuloa), a kitchen, and a latrine were constructed. Money was collected to buy bread, butter, flour, sugar, tea, and yaggng, and each household was tithed two bundles of large taro corms. Cattle were killed and meals cooked and served, all with a minimum of fuss and, seemingly, without bene- fit of sleep for days on end. 360 One of the things which makes such events go so smoothly is the existence of special relationships between visitors and one or more households in the village. Visitors know where to go; the designated hosts are spoken of as their I'doors" (kgtuba). The event described above was too large, and some of the visitors from too far afield, for reliance on these traditional relationships alone. The village chairman drew up a list assigning visitors from specific villages to certain households, but the logic of the pairing was similar to that on which these traditional relationships are based. For example, one household hosted youths from several villages in Muira because the deceased father of the household head had recog- nized a relationship of classificatory siblingship to the Muira man who accommpanied them. This example illustrates an important point. For any individual,the web of kinship extends beyond the confines of their own village relating them to persons in other places. Some of these other persons will be agnates, but to a large extent these extra- community networks are built upon cognatic and affinal ties, the result of intermarriage in present and past generations.3 If we were to plot affinal ties on a map, we would find a dense criss-crossing of relationships within the village, a slightly less dense concentra- tion of relationships between the village and its nearest neighbors (i.e., Taulevu, Navuniyasi, and Saumakiya), and beyond this the frequency of intermarriage is inversely proportional to the distance between communities. With few exceptions, marriages beyond the Wainimala basin are a comparatively recent phenomenon. 361 A Fijian would not live in the village for any length of time before being fit into the web of interrelationships which bind it together. When visiting another village, an attempt is made to establish the existence of some connection, perhaps even through third and fourth parties, which will allow the hosts and guest to place one another within their respective kinship networks. Within clusters of communities among whom intermarriage is frequent, no difficulty exists. The guest's ties to the host community will be well known. If he needs to stay in such a village for any length of time, he will take up reisdence in a household with which his patriline has an established tie, either agnatic or cognatic. He may refer to this customary host as his "doorf in the village. For example, when Tomasi of Taulevu village, a member of matagali Nawaita, visits for any length of time in Nairukuruku, he stays in the house of Saimone, also of "4 matagali Nawaita, a man whom he calls tykggu, me elder brother. In this case the connection is agnatic, but when Pita of Naqara village visits Nairukuruku, he stays in the house of a cognatic kinsman, Semi of matagali Nakorowaqa. Semi's father's sister had married a man of Pita's line, in fact Pita's namesake. Pita calls Semi kawa, though whether the marriage of Semi's father's sister was the basis for this I do not know. Pita considers Semi his Vdoor' in Nairukuruku despite the fact that his own matagali is also represented there. Conversely, Semi refers to himself as the gaunisala or Vroadf to Naqara, and it was he who led the contingent of Nairukuruku men as they entered that village to attend a funeral. He also figured prominently in the 362 formal speeches of excahnge on that occasion, as did Pita on behalf of the taukei_nj mate, the other side in those transactions. Such ties between individuals and patrilines of different villages are not to be confused with the institution of envoys, offi- cial channels of communication among villages and chiefdoms, also referred to as "paths" or "doors.9 Another term applied to these official envoys is taukei Hi sala_or Fowners of the path.f In some other villages in the area (e.g., Naivucini and Nakorosuli) envoys are known as _m_a_ta_, a term foundi elsewhere in Fiji. When Pita. refers to Semi as his Vdoor" (katgba), he is talking about something which is a personal relationship and a tie between patrilines, but such rela- tionships are purely a kinship matter (kénni veiwekani). They may be activated in matters concerning individual families and their matagali, but when communication is an affair between the paramount and the chiefs of other villages 0V.!2£!§: it is transmitted by the official envoys. Not all villages can be contacted directly in this way. In some cases the messages must be transmitted by second or third parties. Thus, should it be necessary to contact Naqara village concerning a matter which relates to the village of Nairukuruku as a whole, a communication which is understood to emanate from its chief, it is carried first to Taulevu by the tgukej_nj_§gla_to Nabena. From Nabena the message is finally relayed to Naqara. Messages are trans— mitted from Nairukuruku to villages in Nabubuco district by a simi- larly indirect path. Should matagali Nabubuco of Nairukuruku wish to 363 contact their distant agnates in that district, they may go directly, for they are, after all, yavusa vata.’ But in matters “of the land“ (vakavanua) and the affairs of chiefs (vakaturaga) which relate to the village of Nairukuruku as a whole, the information is conveyed first to Nakorosuli by a representative of the patriline who are taukeu El sale to that place. From there the message is conveyed to the village of Nasalia and from thence to the desired village in Nabubuco dis- trict. Through these sometimes indirect channels, messages were traditionally conveyed to places as far away as the Wainibuka and Waidina villeys and to Namosi province. An examination of the channels through which messages would have to pass in order to reach villages up and down river from Nairuku- ruku reveals two nodal points in the network of communication, the village of Taulevu downriver and the village of Nakorosuli upriver. The political and historical importance of these villages is relevant here. As we have already seen, the office of paramount was formerly based in taulevu. Perhaps the importance of Taulevu in the network of official envoys stems from this former poligical prominence. With respect to Nakorosuli, the significant factor lies in its position as the paramount village in the ganga which includes the villages of Naiwaisomo, Wairuarua, Waibalavu, Nasalia, and Laselevu. The envoy system is not entirely a thing of the past. When questioned about the instances in which he carried messages to Nako- rosuli, one of the villages to which he is taukei n1 gala, an inform- ant mentioned two occasions on which he relayed news concerning the 364 death of chiefly men of matagali Siko, as well as the time he carried a message concerning an impending sglj_lexu_(major fundraising effort) to help pay for a monument to commenorate Ratu Meli Davetanivalu, the chief who broght Christianity to Matailobau. On all three occasions he carried the messages to his counterpart in Nakorosuli , the we to Matailobau, a member of matagali Nabena, yavgsa Loma (Cf. Nayaca- kalou 1978:85). Another exampleof the continued relevance of the relationships which the envoy system reflects can be seen in the asso- ciation between matagali Nakorowaqa and Navitilevu. Nakorowaqa are envoys to Navitilevu, and when members of matagali Navitilevu visit Nairukuruku in pursuance of some "matter of the land" (e.g., to attend a funeral) they come to a Nakorowaga house even though there are mem- bers of their own matagali residing in the village. As already noted these resident Navitilevu households are attached to matagali Nabubuco, but it is with Nakorowaqa that they "stayV on such occa- sions, along with members of their matagali who reside elsewhere. When asked about the basis for his position as tagkej_nj_§§lg_ to two villages, one informant, a member of an untitled line within matagali Siko, shrugged his shoulders and said it W35.9!é.flé.£§. .makawausgra, "a very old thing.i He suspected, however, that it was based upon ties through women between his line and those of his counter- parts in the other villages. The informant's hypothesis refleets the Fijian tendency to describe relationships through women as Vpathsf (sale) which connect groups, He reasoned that the envoy's role was simply an outgrowth or formalization of the type of relationship 365 described above (i.e., that which exists between Pita and Semi). Hocart offers the same explanation for the institution. He mentions that one imformant from Ba province suggested a comparison between the relationship of cross cousins (veitavaleni) and a counterpart of the envoy system found in western Viti Levu (1913:116). There are also indications that the vasu relationship may be pertinent to an understanding of the envoy's role and vice versa. You will recall that a man's vasu is his sister's child (1am) and, more particularly, the sister's son though the term is extended to include the relationship between a woman's child and her entire natal matagali and is subject to other extensions as well. There have been various attempts to explain or unravel the signifi- cance of the vasu relationship, but perhaps at this point we should limit ourselves to a consideration of Fijians' own explanation of this special relationship, for that explanation bears directly on the institution of envoys, those formal birdges or, to use the Fijian idiom, "paths” between communities. Marshall Sahlins reports that Fijians on the island of Moala recognize that A man's sister gives his "blood" (gra), his "descent" (kawa). to her children. A man'S'father's sister carries his "Eloo " also and gives it to her children. A man, again, carries the "blood" of his mother's brother, carries the “descent" (kawa) of that man through his own mother. . ... One's cross- relatives represent a new line created by one's own stock and therefore linked to it forever. In a way, this is a keystone statement for it allows us to fit together a number of aspects of family life, kinship terminology, and kindred composition that had heretofore presented themselves spearately. The strong bonds a woman maintains with her family of birth is one of these features, as is the preference for cross-relative 366 connections over parallel where alternative reckoning of kin- ship is possible. The "seriousness" of the brother-sister relation, the preference for continued intermarriage with certain groups, the significant frequency of residence with or near the wife's family . . . all of these are intercon- nected social facts (1962:168). I agree with the importance which Sahlins places on the Moalan view that the sister's child represents a new line linked to one's own. More importantly, my informants would agree, but his ultimate conclusion--that "descent is bilaterally reckoned and honoredV--how- ever true of Moala, is inapplicable to the Wainimala area. In his review of Sahlins' book, Murray Groves seizes upon the statement that Moalans describe cross relatives as persons of "sacred blood" (drg tabu). He concludes that What . . . this amounts to is that Moalans attach importance to complementary filiation, and regard the relationships that result as "serious." Why should a complementary filia- tion give rise to a relationship more "serious“ than a primary filiation. How does "sacred blood" link Ego with male, but not female cross-cousins so as to set up a "serious" relation- ship with the former but not with the latter? And what blood of any kind, "sacred" or otherwise, establishes the formal "serious" relationship that exists beween Ego and Parent-in- law? (1963:281). Groves accuses Sahlins of overlooking the obvious fact that the Moalan kinship system is a Dravidian or Vtwo section system . . . that . . . places kin in those categories of relationship in which they would necessarily stand if the society in question consisted of only two lineages exchanging women prescriptivelyf (1963:281). Accord- ing to Groves, the "seriousness" of cross relationships is to be under- stood as a reflection of the fact that structurally these are, or are equivalent to, ego's affines. |I 367 It seems to me that Groves trivializes the statement of Sahlins' informant. I do not think that the point which Groves accuses Sahlins of overlooking is lost on the Moalan: namely, that the impor- tance of the sister's son stems from his membership in a line with whom one's own line exchanges persons and valuables. The informant does not state the matter directly. .What is important to him is that the sister's son shares a common substance with him while at the same time representing a Vnew linef indissoluably linked to his own. Brothers are only brothers [veitacini s2 veitacini 93], but the sister's child is a new path. . . . Y65F‘33565er or your sister makes a new line [kawa]. Brothers are only in the house; they have been there—from the past to today. But the line of my sister is a new line (Sahlins 1962:168). The "path? represented by the sister's son is one which future genera- tions of one's own patriline may travel for many purposes, and one which can be strengthened by future intermarriage. Like the envoy, the sister and her offspring can transcend the boundaries between groups. And, in turn, like the sister's son, the envoy has a sacred quality. His person and his mission must be respected. Indeed, the attack which led to the scattering of the Navitilevu (see Chapter VIII) was in retaliation for their failure to do so (Brewster N.D.: 55). .I was drinking yaqona one afternoon in the house of a Navitilevu man. When we were joined by a man from Navuniyasi village, a member of matagali Nasautoka, the subject of "paths? somehow arose. I was told that matagali Nabubuco were the fowners of the path? to Nasautoka. Today, members of matagali Navitilevu and Nasautoka live El 368 directly across the river from one another in the villages of Saumakiya and Navuniyasi respectively. But in the past it would not have been possible for a Nasautoka man to visit Navitilevu territory unescorted. Should he do so, he would have been summarily clubbed. If a member of matagali Nasautoka wished to visit in Navitilevu territory, he must first approach the Nabubuco people who would then convey his request to the Navitilevu. If the request were granted, members of matagali Nabubuco would accompany him and insure his safety. An appreciation of the importance which Fijians place on relationships through women as paths of communication between groups, as well as their tendency to identify the institution of envoys with such relationships, is enhanced by a consideration of the nature of Fijian society as it was prior to pacification. In the past such relationships were.quite literally paths. In the absence of agnatic ties, communication beween groups was only possible where such rela- tionships existed. CHAPTER IX--FO0TNOTES 1At the time the official exchange rate was approximately F$1.00 = U.S. $1.20. 2Another factor which might have been relevant to the lines of opposition that were drawn is the fact that, in subtle ways, the planning of the event was guided by a married couple of national prominence. The husband is a member of mata ali Nawaita, and his wife's natal matagali is Siko. Perhaps the way in which the village was divided a owe them to influence both "sides" while favoring neither. They reside in Suva and contributed on behalf of the Suva adj, their daughter, and so were obstensibly competing against their own mata ali. While, from their point of view, all of this might have Been a consideration, I suspect that the latent opposition between "upper" and “lower" would have emerged in any case. 3The terms "affinal" and "cognatic" are used here to refer to etic categories. It should be kept in mind that some of the relation- ships which the anthropologist would consider affinal are classified as consanquineal in this system (e.g., MZH and WMB are equated with F), and some cognates are classified along with agnates (e.g., MMB is equated with FF). 4All of the personal names used in this section are pseudonyms. 369 CHAPTER X CONCLUSION Though they have focused on a number of specific issues, taken in conjunction, the preceeding chapters have illustrated a funda- mental point. Social units whose respective members are regarded as people of different "types," who claim different origins, and who acknowledge different ancestral deities and totemic assoCiations, nonetheless regard themselves as a community. As we have seen, the bonds which unite them include mutual dependence on village lands; propinquity and the resultant coactivity; common allegiance to a chief; ritual interdependence; and ties of uterine kinship and marriage. But we have also noted that some of the very principles which trans- cend the boundaries between groups (e.g., propinquity, affinity, and cognation) may, in turn, be involved in the definition of those bound- aries. Thus, in the past Nakorowaqa and Nabubuco were said to be a single matagali because their members were joined by affinal and uterine ties, "stayed together," and c00perated as a unit in ceremonial exchange. Taken together, these factors defined the two of them as a unit in opposition to other like units. While ties through women also linked each of them to other groups in their own and neighboring communities, these did not establish similar relationships of matagali vata . 370 371 In the course of our discussion, we insisted upon the distinc- tion between social units which are defined in terms of descent and those which are primarily a matter of coresidence and cooperation. (Though the latter almost always involve affinal and cognatic ties between descent lines, it is in terms of coresidence and coactivity that they are defined.) We insisted on this distinction despite the fact that Fijians refer to groups of both types with the same set of terms and apply the same names to specific groups of either type. The distinction between agnation and other principles of recruitment becomes salient in many contexts, but Fijians also see a fundamental similarity between people who share patrilineal descent and those who, on the strength of other sorts of ties, merely "stay together.f As one informant put it, "if people live together, work together, and eat together, they are mataqali vata ('matagali together'),f despite any other factors which may separate them (e.g., agnation). In an earlier chapter we noted that, unlike descent group membership, local group affiliation is mutable. Thus the composition of any such group reflects the individual choices of its members and can therefore be described in statistical terms. While this is true, it is also necessary to understand that Vcomposite" local groups (such as Nakorowaqa and Nabubuco once constituted) reflect certain structural features of the system. To put it another way, the con- trast between agnation and local group composition is not one of structural principle (descent) versus empirical event but, rather, one between different structural principles which are, themselves, 'I 372 encompassed by a more comprehensive structure, that of affinal exchange. For invariably in cases in which local groups represent a fusion of lines unrelated by agnatic ties, the link between them is traced through women.’ Even in cases where one group seeks refuge with another on the basis of some other principle, their coresidence is likely to be further cemented by intermarriage. As noted in Chapter 7, the structure of the system of rela- tionships is such that, for any individual, it is as if the universe of kinsmen were divided into just two agnatic lines linked through connubiumt.. In fact, this ideal structure is never realized, but there is a tendency for affinal links between patrilines to be repeated in later generations. In the context of subsequent exchanges such lines function as “sides' in relation to one another. They may or may not be coresident segments of a single local group, but if they are, the group's internal relations will involve exchange as well as sharing and cooperation. These exchange relations between affinally linked descent lines are literally the source of new persons. In the absence of publicly acknowledged paternity, a child lacks true agnates, and unless the paternal line fulfills its initial exchange.obligations to the mother's people, the child will lack the full support of his mother's brothers. In a sense, the person himself is an object of exchange. The maternal patriline loses their daughters' children in life, but the latter are returned to them in death, for it is their gg$a_who bury them. Having said that the composition of local groups (i.e., matagali) reflects certain structural features of the kinship system, 373 let me be clear that it is primarily in the context of exchange that such groups are defined vis-a vis others of the same order. The pooling of ritual goods, food, and other forms of wealth for joint presentation in the context of funerals, weddings, sell, and other events sets each matagali apart from others. Individuals may have ties to both "sides" in an exchange, and thus may experience con- flicting obligations, but on most occasions, should an individual wish to contribute, he or she does so as a member of his or her own local group. Internally, of course, local group membership is expressed through the reciprocal exchange of food, labor, and support and through commensality on ritual occasions. As noted in an earlier chapter, the ritual division of labor, which is itself an exchange of services among groups, is a matter regulated by descent. Therefore, the members of “composite? local groups may be differentiated on the basis of their occupational specialization (e.g., chief, warrior, or herald) just as they are differentiated on the basis of agnation. In some contexts, these differences may be honored; in others ignored. Thus on some occasions the two men of matagali Navitilevu who "stay with? Nabubuco are asked to assume the role associated with their . chiefly rank. On other occasions, however, their chiefly rank is submerged, and it is their status as attached cognates which is salient. As elsewhere in Fiji, those who are united in one context may be differentiated in another, and it bears repeating that the contexts in which boundaries are defined usually involve exchange. 374 In the preceeding chapters we have seen that the sharing or exchange of food, yaqona, ceremonial wealth, names, and work (i.e., ritual occupation) are central to the question of what separates “us" from "others.“ What I refer to as "sharing" and "exchange" are, of course, aspects of a single phenomenon, reciprocity. They are alike in that they necessarily entail a distinction between fyou“ and VI,V but among other things, they differ in the emphases with which this distinction is stated. In sharing, the emphasis is on conjunction, similarity, or community. Vis-a-vis outsiders, acts of sharing estab- lish our identity. In exchange, the emphasis is on disjunction or differentiation, but the differences stated are never so extreme that they cannot be transcended by the same activity which calls attention to them. I It would be wrong to suggest that sharing is a less important aspect of Fijian culture than exchange.) As among all human groups, it is the very basis of society. But it is more understated, and understandably so, for calling too much attention to the conjunction which joins 'I“ and "you? may call into question that which should remain unquestioned. Exchange, on the other hand, is often elaborated in such a way as to underscore the very separation which it would bridge. In the process economic transactions may become a medium for various forms of aesthetic expression--music, dance, oratory, or bodily ornamentation. One consequence of this expressive quality is an expansion of the messages which can be transmitted through the medium of exchange. The Melanesian interest in the objects of exchange could be misinterpreted as an obsession with objects as things in ‘I 375 themselves. But as Annette Weiner points out in her restudy of Trobriand exchange, the objects given and the mode of their presenta- tion are also valued as statements of the donor's inner feelings. In order for Trobrianders to read specific exchanges as reflections of the thoughts and feelings of the partici- pants, they must possess a system of interpretation. Ethno- graphers have assumed that this system can be understood by eliciting the rules for specific exchange events. But the system described ih this way allows only two messages; either participants adhere to the rules and "all is well," or they do not follow the rules, in which case "something is wrong." Among the Trobrianders, however, [Weiner] observed a much “ wider range of communication through exchange acts. Trobrianders evaluate each other's thoughts and feelings in the light of the value of specific exchanges, and they take care that the dispositions they wish imputed to themselves will be correctly read by others (Weiner 1976:212). In Fiji, too, each instance of ceremonial exchange is poten- tially a public statement of inner feelings. In the quantity and quality of objects given and in the speeches which accompany their giving, they become material expressions of sentiment. At weddings and funerals, written accounts are kept of all that has been given. What has been received will later be redistributed, and through these counterprestations, the original recipient also has a channel for expression. In addition to the major exchange sequences which accom- pany marriage and death, there are many other occasions, both public and private, on which similar messages can be conveyed through giving or withholding.. In a culture in which exchange activities are at the focal point of social organization, One may expect a concern for balance in transactions of all kinds. Such is the case in Fiji. This is mani- fested not only in the recording of all that is given and received 376 at major exchanges and the mental bookkeeping that accompanies less formal transactiOns; it also extends to other aspects of social rela- tions. There is, for instance, a feeling that wrongs should be balanced and injuries repaid. As with all such imbalances, there is no statue of limitations in this regard. As noted earlier, the split between matagali Nakorowaqa and Nabubuco had been precipitated by an assault’ committed many years ago. -Despite the passage of time and the restoration long ago of good relations between the two groups, I was specifically told that an imbalance remained. To restore that balance the services of an adept (y§_nj 331) were retained to over- see an i sore, or ceremony of reconciliation, the most prominent fea— ture of which was the presentation of yaggna by the matagali of the attacker to the patriline of the victim. Ceremonial obligations and debts of all kinds are a matter.of concern not only to the living, but also to the ancestors. I was told that during that period of unrest during which the power of the Cakobau government was extended into the area, the ancestors of a group living in a village upriver sought refuge in Nairukuruku. Nearly a century later, their descendants, having long since returned to their own lands, experienced a number of misfortunes in quick suc- cession, and they sought the services of an adept to learn the reason for their difficulties. The adept suggested that the trouble may lie in their unpaid debt to the descendants of the Nairukuruku people who had sheltered their forefathers. This imbalance in the relations between the two groups was evidently distressful to the ancestors. The 377 other group presented a magiti and yaggna_to the Nairukuruku people, and their troubles came to an end. Both of these examples exhibit how the maintenance of balance in exchange relations (and here we are speaking of exchange in the broadest sense) can influence general well-being. This need to achieve balance provides a means for effecting the behavior of others. Among kinsmen any formal request (kerekere) requires consideration, but when accompanied by giving, whether of yaqona, tabua, food, or service, the request becomes difficult to refuse. This coercive aspect of giving prompts Weiner to make a general criticism of the anthropological approach to exchange. She points out that not only can exchange serve as a more subtle medium of expression than is generally appreciated, but it also provides a way for obtaining and exercising power. Weiner feels that this is an aspect of exchange relations which anthropologists have, if not ignored, at least relegated to secondary importance. She believes this bias is a result of their fcontinually emphasizing the primacy of exchange as social rather than economic' and argues that Vwe must push exchange beyond the level of our view of the social world and seek to understand exchange as the means, however limited, of gaining power over people and control over resources in the widest sensef (1976:220). Self-interest is not merely an element present in exchange; it is the primary motivating force behind it. I do not wish to become involved in a discourse concerning first principles, but the characterization of exchange as an attempt to gain power in the widest sense strikes me as being very true in the case of Fiji. But in order to appreciate that fact, I am afraid that 378 we have to emphasize "the primacy of exchange as social rather than economic." For I am not speaking here of all the calculations which X makes in order to influence the behavior of Y (Weiner refers to these as a kind of risk-taking), even though these are certainly present and may well be the mainspring of individual actions. I am speaking here of the belief that exchange, in an of itself, is crucial to the mainte- nance of the social order, for through it power is obtained, not over persons merely, but over resources and, ultimately, the forces which control not only the production of those resources, but also life and death, sickness and health. The mutual ministration which should obtain between the living and their ancestors, between commoners and their chiefs, and between the affinally linked descendants of cross sex siblings can only be effected through exchange, and the very act of exchange makes the land prosper. APPENDIX 379 APPENDIX A BRIEF NOTE ON FIJIAN ORTHOGRAPHY In Fijian the consonants “b," "d," and "g“ are always nasalized regardless of whether they occur initially or medially. Missonaries found that when these were rendered as "Mb," "nd," and 'ng,9 Fijian readers tended to insert an extra vowel between the consonants. The written alphabet was modified to avoid this confusion. The other major departure from standard English othography is the rendering of the voiced “th” as "c." Thus, in written Fijian "b“ is pronounced as "mb" as in "number" "d" is pronounced as "nd" as in Ffind" "g" is pronounces as "ngi as in "singer" "q“ is pronounced as "ng" as in ”linger? "c“ is pronounced as "th" as in “there“ (but.not"thin") 380 BIBLIOGRAPHY 381 BIBLIOGRAPHY Adams, Richard M. 1975 Energyand Structure: A Theory of Social Power. Austin: University of Texas Press. Arno, Andrew 1974 “Ritual of Reconciliation and Village Conflict Management in Fiji." Oceania 47:49-65. Barrau, Jacques 1965 “L'Humide et le Sec: An Essay on Ethnobiological Adaptation to Contrastive Environments in the Indo-Pacific Area." Journal of the Polynesian Society 74:329-346. Belshaw, Cyril S. 1964 Under the Ivi Tree: Society and Economic Growth in Rural _— Fiji. London: Routledge & Kegan Paul: 1965 "The Effect of Limited Anthropological Theory on Problems of Fijian Administration. In Induced Political_Change in the" Pacific: A Symposium, edited by Roland W. Force. Honolulu: Bishop Museum Press. Brewster, A.B. 1922 Thewliirl Tribes of Fiji. Philadelphia: J. B. Lippincott Cfimpany. N.D. "Genealogies and Histories of the Matanitu or Tribal Govern- ments ofCentral Viti Levu." Mimeo in the National Archives, Suva, Fiji. ‘ Buckley, Joseph P.; Furgiuele, Angelo R.; and O'Hara, Maureen J. 1979 "Pharmacology of Kava." In Ethnopharmacglogjc Search for Psychoactive Drugs, edited by Daniel H. Efron, et al. New York: Raven Press. Burrows, Edwin G. 1939 "Breed and Border in Polynesia." American Anthropologist (N.S.).41:1-21. ‘ Capell, A. 1941 A New Fijian Dictionary. Suva, Fiji: Government Printer. 382 383 Capell, A. and Lester, R.H. 1941 "Local Divisions and Movements in Fiji." Oceania 12: 21-48. 1945 "Kinship in Fiji." Oceania 15:171-200. 1946 Oceania 16:109-143; 2343253; 297-318. Derrick, R.A. 1946 A History of Fiji. Volume One. Suva, Fiji: The Govern- ment Press. ' Firth, Raymond 1965 Primitive Polyneasian Economy. New York: W. W. Norton & Company Inc. 1967 The Work of the Gods in Tikopia. London School of Econ- omics Monographs on Social Anthropology Nos. 1 and 2. New York: Humanities Press Inc. Ford, Clellan S. 1979 “Ethnographical Aspects of Kava." In Ethnopharmacologic Search for Psychoactive Drugg. edited by Daniel H. Efron, et’TT. New York: Raven Press. Fortes, Meyer ~ 1959 "Descent, Filiation and Affinity: A Rejoinder to Dr. Leach." .flgg 59:193-197; 206-212. 1969 Kinship and the Social Order: The Le acy of Lewis Henry Morgan. Chicago: Aldine Publishing ompany. France, Peter 1969 The Charter of the Land: Custom and Colonization in Fiji. Melbourne: Oxford UniVersity Press. Gajdusek, D. Carleton 1979 "Recent Observations on the Use of Kava in the New Hebrides." In Ethno harmacolo ic Search for P5 ch04 activeDrugs, edited Ey Oaniel H. Efron, et al. New York: Raven Press. Geddes, W. R. 1945 Deuba: A Study of a Fijian Village. Memoirs of the Polynesian Society, Vol. 22. Wellington: The Polynesian Society Incorporated. ' Goldman, Irving 1970 Ancient Polynesian Society. Chicago: University of Chicago Press. 384 Goodenough, James Graham 1876 Journal of Commodore Goodenou h, During his Last Command as Senior Officer on the Australian Station, 1873-75. London: Henry S. King & Co. Gordon, Sir A. H. N.D. Letters and Notes Written_during the Disturbances in the Hichlands (Known as the "DeviTCountry")of Viti Levu [ng, Edinburgh: PriVately printed by R. & R. Clark. Groves, Murray - 1963 "The Nature of Fijian Society.“ Journal of the Polynesian Society, Wellington 72:272-291. Handy, E. S. C. 1940 The Hawaiian Planter, Vol. 1. His Plants, Methods, and Areas of Cultivation. Bernice P. Bishop Museum Bulletin 161. Honolulu: Bishop Museum. ‘ . Harding, Major James 1916 "Diary of the Nadawarau Campaign, 1874." Transactions of the Fiji Society, 1916. Hecht, Julia 1977 “The Culture of Gender in Pukapuka: Male, Female, and the Mayakitanga 'Sacred Maid.'" Journal of the Polynesian Society 86:183-206. Hocart, A. M. 1913 "The Fijian Custom of Tauvu." Journal of the Royal Anthropological Institute 42:4371449. 1914 "Notes on Fijian Totemism.f Anthropos 9:737-739. 1929 Lau Islands, Fiji. Bernice P. Bisohp Museum Bulletin 62. Honolulu: Bishop Museum.. A 1936 Kings and Councillors: An Essay in the Comparative Anatomyiof Human Societ . Cairo: Printing Office,Paul Barbsy. [Chicago: T e University of Chicago Press, 1970 . 1952 The Northern States of Fiji. Royal Anthropological Insti- tute, Occasional Publication No. 11. London. Holmes, Lowell D. 1979 “The Function of Kava in Modern Samoan Culture." In Ethnopharmacologic Search for Psychoactive Drugs, edited by DanieTTH. Efron, et al. New York: Raven Press.‘ 385 Leenhardt, M. 1937 Gens de la Grande Terre. Paris: Gallimard. 1947 Do Kamo: Person and Myth in the‘Melanesian World. [1979] Chicago: University of Chicago Press. Legge, J. D. 1958 Erétain in Fiji, 1858-1880. London: Macmillan & Co.. t . Lester, R. H. 1941 "Kava Drinking in Viti Levu, Fiji." Oceania 12:97-124. Levi-Strauss, Claude 1949 The Elementary Structures of Kinship. Boston: Beacon [1969] Press. Malinowski, Bronislaw 1935 Coral Gardens and Their Magic, Volume II. The Language of Magic and Gardening, New YfirkE AmeriCan Book Co. [Bloomington: Indiana University Press, 1965.]‘ Meyer, Hans J. 1979 "Pharmacology of Kava." In Ethno harmacolo ic Search for Psychoactive Drugs, edited by OanEel R. Efron, at ET. New York: Raven Press. Nayacakalou, R. R. 1975 Leadership in Fiji. Melbourne: Oxford University Press. 1978 Tradition and Change in the Fijian Village. Suva: South Pacific Social Sciences Association. Norton, Robert 1977 Race and Politics in Fiji. New York: St. Martin's Press. Pfeiffer, Carl C.; MUrphree, Henry B.; and Goldstein, Leonide 1979 "Effects of Kava in Normal Subjects and Patients." In Ethnopharmacologic Search for Psychoactive Drugs, edited by Daniel H. Efron, et al. New York: Raven Press. Quain, Buell 1948 ‘fjjian Village. Chicago: The University of Chicago Press. Rappaport, Roy A. ' 1968 Pigs for the Ancestors: Ritual in the Ecology of a New Guinea People. New Haven: Yale University Press. 386 Rivers, W. H. R. 1914 The History of Melanesian Society. Cambridge: ,Cambridge University Press. Rutz, Henry J. 1977 "Individual Decisions and Functional Systems: Economic Rationality and Environmental Adaptation." American Ethnologist 4:156-174. 1978 "Fijian Land Tenure and Agricultural Growth." Oceania 49 3 20-34 0 Sahlins, Marshall D. 1962 ‘Mgala: Culture and Nature on a Fijian Island. Ann Arbor: The University of Michigan Press. Scarr, Deryck 1972 I,,the very Bayonet. Volume 1 of The Majesty of Colour: A Life of Sir John Bates Thurston. Canberra: Australian National University Press. Scheffler, Harold W. and Lounsburg, Floyd G. 1971 A StudyinStructuralSemantics: _Ihe Siriono Kinship System: Englewood Cliffs, N.J.: Prentice-Hall Inc. Silverman, Martin H. 1971 Disconcerting Issue: Meaning and Struggle in a Resettled Pacific Community. Chicago: The University of Chicago Press. Spencer, Dorothy M. 1941 Disease, Religion, and Society in the Fiji Islands. Monographs of the American Ethnological Society 11. New York: J. J. Augustin Publisher. Steinmetz, E. 1960 Piper methysticum. Amsterdam: Published by the author. St. Johnston, Alfred 1889 Camping;amongCannibals. London: Macmillan and Co. Thompson, Laura 1940 Southern Lau, Fijj: An Ethnography. Bernice P. Bishop Museum Bulletin 162. Honolulu: Bishop Museum. Trautmann, Thomas R. 1981 Dravidian Kinship. London: Cambridge University Press. Twyford, I. T. and Wright, A.C.S. 1965 The Soil Resogrges of the Fiji Islands. Suva, Fiji: Government of Fiji. ,387 Walter, Michael A. H. B. 1974 "Succession in East Fiji." Oceania 44:301-322. 1975 "Kinship and Marriage in Mualevu: A Dravidian Vaviant in Fiji." Ethnolggy 14:173-180. 1978a "Analysis of Fijian Traditional Social Organization: The Confusion of Local and Descent Grouping." Ethnology 17:351-366. 2978b "An Examination of Hierarchical Notions in Fijian Society: A Test Case for the Applicability of the Term 'Chief.'" Oceania 49:1-19. Weiner, Annette B. 1976 .Women of Value, Men of Renown: New Persepctjyes ia Trobriand Exchange. Austin: University of Texas Press. Newspapers Fiji Times. Microfilm. National Archives, Suva, Fiji. Papers of the Cakobau Government Outwards Correspondence, Provincial Secretaries Office [Naitasiri]. National Archives, Suva, Fiji. Other Government Documents Colonial Secretary's Office, Papers. National Archives, Suva, Fiji.