Volume 100 1991 > Volume 100, No. 1 > Some reflections on the significance of names in Matailobau, Fiji, by James West Turner, p 7-24
                                                                                             Previous | Next   

- 7

In order to be a member of a human society one must have a name. This is underscored by the fact that, in many cultures, temporary liminality (e.g., of the new-born) is marked by the absence of a name. Incorporation into society, or reincorporation (as in the case of principals in a rite of passage), is marked by the bestowal of names.

Inasmuch as the original incorporation of the individual into society involves the act of naming, names can be seen as creating as well as labelling social identities. Systems of naming and cultural definitions of personhood are necessarily related, sometimes in a very direct way, so that a full understanding of one requires consideration of the other. This paper examines the relationship between names and personhood in Matailobau, a district in the hill country of Viti Levu, the largest of the Fiji Islands. The immediate goal is ethnographic — that is, to examine what names “say” about persons in a particular society — but, in the process, we shall consider in general the role names play in structuring the relationship between the individual and society.

“Individual” is used here to mean a biographically unique social entity. The term “person”, on the other hand, refers to a moral agent who occupies a particular set of statuses in society and who contrasts, at least in some contexts, with all other such agents. While personhood is defined in relation to a position within the social structure, individuality involves the unique experience and interpretation of that position. As the terms are understood here, it is conceivable for two or more individuals to be regarded as in some sense the same person. For example, an ancestral deity, a human ancestor who intercedes between that deity and his human worshippers and a living descendant in his role as priest may all be identified as a single person in the context of ritual.

Naming systems round the world differ with respect to the relative emphasis they place on personhood and individuality. For example, despite the fact that Balinese commoners bear unique names composed of nonsense syllables, these names are seldom used after childhood, and Geertz (1966) convincingly shows that the entire system of personal identification, of which personal names are only a part, has the effect of suppressing individual identity. Who one is, “physically, psychologically, or biographically, is muted in favour of his - 8 assigned place in the continuing . . . pageant that is Balinese life” (Geertz 1984:129). On Truk, too, virtually every individual bears a name that he shares with no one else, living or dead, although the constant use of these unique names on Truk has the effect of calling attention to individuality. Goodenough contrasts the Trukese system, with its emphasis on individual identification, with that of Lakalai (West New Britain), where names “emphasize one's place in a procreational chain or in formally structured kin and social relationships” (Goodenough 1965:271). Goodenough concludes that, in calling attention to individual identity, the naming system on Truk compensates for the general suppression of individuality in most aspects of social life. In both Truk and Lakalai, naming practices and forms of address function as a counterbalance to “the effect that the workings of the social system tend otherwise to give the people's images of themselves and others” (Goodenough 1965:275).

Lévi-Strauss (1966) also discusses the difference between naming systems that emphasise individual identity and those that stress one's position in the social structure. For him, the crucial distinction is between naming and classification. In his usage, to name is to identify a unique individual in much the way we do when we use our fingers to point: “That is John Smith”. In contrast, classifying terms designate the defining features of the categories they label.

Having drawn the distinction between naming and classification, Lévi-Strauss argues that it dissolves in practice. For him, one never names, or never merely names: one always classifies. A personal name either classifies the person who bears it or classifies the person who bestowed it. In the former case “the name is an identifying mark which, by the application of a rule, establishes that the individual who is named is a member of a preordained class (a social group in a system of groups, a status by birth in a system of statuses)” (Lévi-Strauss 1966:181). Very different from this are systems in which names reflect the whims of the individual who gave them rather than the status (personhood) of the name-bearer. Lévi-Strauss emphasises the self-expressive nature of such “free” naming. In such cases the name-giver classifies himself, rather than the name-bearer, in the sense that his choice conveys information about his own tastes, feelings or life experiences.

Even if we acknowledge an underlying similarity between naming systems that permit the subjective classification of self (“free” naming) and those that result in the objective classification of others, the differences are sociologically more interesting. In systems in which names are freely given they differentiate individuals, and they may communicate information about those who bear them or those who bestowed them, but they do not reflect the structure of society in any direct sense. Systems in which names reflect the bearer's position within a framework of statuses are very different. In such systems naming is consciously - 9 used to reproduce society in conformance to a cultural model of its structure. Each new member is given structural relevance through a name; it is personhood rather than individuality that is stressed. The naming system in Matailobau is of this second type. I shall turn now to a description of that system, followed by a discussion of some alternative forms of reference and address that can replace personal names. The paper will then conclude with a section devoted to comparison and interpretation.


As a preliminary to a discussion of naming practices in Matailobau it will be necessary to discuss briefly certain aspects of the social system in which those practices are embedded.

Unlike the practice in some other areas in Fiji, descent in Matailobau is unambiguously patrilineal. By virtue of agnatic descent each person belongs to a named totemic clan (mataqali) dispersed over a number of villages. Although the members of a mataqali claim descent from a common ancestor, it is not possible to specify the exact genealogical relationships among them. Indeed, it may not be possible to specify the exact genealogical relationships among all the members coresident in a single village.

Two or more clans may be linked by putative descent from a common ancestor to form higher-level groupings called yavusa. The mataqali that make up a yavusa are understood to be the descendants of the sons of the founding ancestor of the yavusa and are ranked in terms of their seniority of descent. Mataqali are, in turn, internally divided into lower-level groups; these are not named as such, but are called after the house name of their most senior members. In most cases it would be proper to call these lower-level groups patrilineages, since their members are able to trace genealogical connections to one another. Although there is no local term for these groups, people are aware of the fact that, in other areas, these subdivisions of mataqali are called i tokatoka.

Each mataqali is considered to have its proper function (i tavi) in a ritual division of labour focused on the institution of chiefship. These ritual functions are ranked and provide a criterion for the ranking of mataqali that are not linked by agnation and, hence, not ordered by seniority of descent. The term “mataqali” is used to refer not only to dispersed descent categories but also to local groups that co-operate in ceremonial exchange (Turner 1987). In the village in which I resided there are five such groups. Local groups form around a core of male agnates, but, in addition to codescent, membership can be justified in terms of affinal ties, cognatic kinship, personal friendship or traditional ties between descent lines. Differences of descent are not neutralised by coresidence. That is, in Matailobau there is no acknowledged process by which coresident cognates - 10 are redefined as agnates as there is, for example, in some societies in the New Guinea Highlands. Given the rule of patrivirilocal residence, most nonagnatic members of these local groups are in-marrying women. 2

The members of a clan are believed to share a bond of common substance — blood — but this bond does not differentiate agnates from cognatic kin. In unilineal systems such as this it is fairly common for theories of conception to include some notion that genitor and genitrix make different contributions to the child's makeup. Such beliefs provide a cultural rationale for the contrast between descent group affiliation and what Fortes (1969:197–9) called complementary filiation. But, despite a rule of patrilineal descent, in Matailobau there is no clearly articulated belief that some component of a person's being is transmitted exclusively in the male line. A child is believed to represent a fusion of the blood (dra) of both genitor and genitrix. These beliefs about the bilateral transmission of substance are associated with a system of kinship and marriage in which bilateral cross-cousin marriage is an important structural feature.

Blood is believed to be coded for certain predispositions of behaviour and character. The ability to perform the proper function of one's mataqali (as chief, priest, warrior or herald) is partly a matter of this coded natural substance that one shares with other persons of the same type. For example, it was pointed out to me that a man visiting in the village was employed by the Government as a carpenter despite the fact that he had no formal training in that craft, nor did he use squares, tape measures, levels or other measuring devices. It was claimed that he exhibited his skill “naturally” because he was a member of a mataqali whose traditional function was to act as carpenters (matai) to their chief. Similarly, a chief or a member of the mataqali bati (i.e., the chief's warriorretainers) is appropriate for his role because he has inherited the blood of his ancestors. But, since these natural abilities are transmitted in the blood, they cannot be confined within the bounds of patrilineal descent groups (since they are not endogamous). Thus, informants told me that, when a former chief died and there were too few bati to properly guard his body, men with “bati blood” (i.e., men able to trace a uterine connection to that mataqali) were also called upon.

Despite the fact that notions about the transmission of natural substance do not draw boundaries around descent groups, mataqali are conceived of as being different kinds of people. Indeed, in the broadest sense, mataqali means ‘type’ or ‘category’. One aspect of the difference among these various ‘types’ of people lies in the names they bear, and it is to the subject of names that I shall now turn.


In Matailobau, as in other areas in Fiji, names are bestowed in pairs. The first name is usually biblical or European in origin, but the second name is always - 11 Fijian. These second names or yaca vakatevoro (‘heathen names’) always have semantic content. They are typically compounds of two or more words (e.g., Tamanivalu, ‘father of war’), and are often poetic or allusive in nature. Although everyone bears two names, they may be customarily known by one or the other.

These pairs of names are gender-specific. First names of biblical or European origin are either male or female, of course, but the semantic content of vakatevoro names may also mark them for gender. Names that include liku (the traditional skirt), tina (‘mother’) or kuro (‘pot’), for example, are marked as feminine. But, in the final analysis, a name is male or female because either men or women have borne it. Men's names are not bestowed on female infants and vice versa. 3

In some areas in Fiji children are named for a senior relative, not by giving the child that relative's name(s) but by bestowing a new Fijian name, one that commemorates some event in the “namesake's” life (Hocart 1929:151). Thus, Sahlins (1962:183) mentions a Moalan name that he translates as ‘Three Days and Three Nights in the Lakeba Jail for Failure to Pay the Tithe’. (The name referred to an experience suffered by the man's maternal uncle.) In some places Fijian names can also refer to events that took place at or about the time the child was born (e.g., Siganisucu, ‘Christmas Day’ for a child born on the 25th of December [Ravuvu 1983:62]).

There are, then, examples of “free” naming in Fiji, but, to the extent that information on naming practices can be gleaned from the literature, it would appear that, in most areas of Fiji, children are typically given the actual name(s) of some relative of either the father or the mother, often on an alternating basis (Williams 1858, Sahlins 1962, Ravuvu 1983). In Matailobau, however, the rule is more restrictive. Except under certain circumstances discussed below, children are typically given a name borne by a senior member, living or dead, of their fathers' mataqali (clan). The name thus “establishes that the individual who is named is a member of a preordained class” (Lévi-Strauss 1966:181).

Names are bestowed shortly after birth by an elder of the mataqali (e.g., the child's father's father or his classificatory equivalent). In giving the names this representative also confers group membership and gives public recognition of the infant's personhood. The child's features will develop and change and its ‘soul’ (yalo) will ‘harden’ or mature, but the names bestowed at birth remain an important element of the person's identify throughout life. They also constitute the child's legal identity. Both names will be recorded in the Ai Vola ni Kawa, the official registry of mataqali membership maintained by the Fijian Administration. The recording of the names ensures a person's legal rights in mataqali land.

Given the recycling of names within the mataqali, it is common for several individuals to share the same name. For example, in 1980 the Ai Vola ni Kawa - 12 recorded the names of 23 males in mataqali Nakorowaqa, Nairukuruku village. These 23 individuals span 5 generations with birth dates ranging from 1885 to 1967. The following are some of the names represented among them:

Jeremaia Cakausese — four individuals (three living); a fifth individual bears the name Cakausese but with a different first name, Ananaiasa.

Taniela Nalaide — three individuals (two living).

Penisoni Waisiga — two individuals (one living). Here we can include a third man, the full brother of the living Penisoni Waisiga, who also bears the name Penisoni but a different vakatevoro name.

Naibuka Nailagoliva — two individuals plus a young boy whose name was not yet recorded in the Ai Vola ni Kawa as of 1980 (two living).

Josese Lautabu — two individuals (one living).

Pita Tiqatabua — two individuals (two living).

Sailasa Nailava — two individuals (one living).

Two other individuals bear names that occur in the other mataqali into which the Ai Vola ni Kawa divides yavusa Nakorowaqa, while a third bears the name of a man who belongs to a branch of mataqali Nokorowaqa in another village. None of these name pairs occurs in any other mataqali in the village.

Persons who share a name refer to one another as noqu yaca (‘my name’), but, when speaking of the name itself, one must say “yacaqu”, using the same possessive suffix (-qu) used for parts of the body and certain kin relationships. A personal name is thus marked as an inalienable part of the self. But it is also something that one shares with others, and conceivably the repetition of names could make for occasional confusion. One way of clarifying which bearer of a name one is referring to is to distinguish one as ‘big’ (levu, i.e., elder) and the other as ‘little’ (lailai, i.e., younger). Teknonymy (see below) also eliminates possible confusion.

The pool of names that a mataqali controls represents a set of positions that a succession of individuals occupy, a fact that people recognise. Once during an interview an informant turned to others for help in remembering the full name of somebody no longer alive. “What is the vakatevoro name of that mataqali's Tevita [David],” he asked, meaning “What is the full name of all members of that group, living and dead, who have borne the name Tevita?”

Names that are not given are positions that are left vacant, and eventually the names may be lost. To ensure that one's name is passed on, a person may formally request that a specific infant be given the name. Thus, a woman married out of the mataqali may present kava (yaqona) to a male agnate and ask that her name be given to his child.

As a form of corporate mataqali property, names are not freely available to outsiders. In order to give a child a name belonging to another mataqali, the father - 13 should first approach the would-be namesake or his representative, present yaqona and formally ask permission to use the name. This formal request is more than an act of courtesy; it is a recognition of ownership of the name. To request the use of a name honours the potential donor, and there is an expectation that the namesake will be generous towards the child. Taken together, these facts explain why there were several children and young people in the village in which I resided who were named after schoolteachers and pastors, influential Fijians from other areas who had lived in the village for a time.

It is possible, through this same process, for names to pass from one local mataqali to another. Although they are comparatively rare, exceptions to the rule of patrivirilocal postmarital residence do occur (see note 2), and in such cases the descendants of male cognates will constitute a distinct patriline within the local group. It will be said that they are ‘staying with’ (tiko vata) the agnatic core, and cognatically linked lines that ‘stay together’ may come to share names. Once, when visiting a village more than an hour's boat ride upriver, I was struck by the fact that several persons bore names that also occurred in one of the mataqali in the village in which I lived. Given the binomial naming system and the very large pool of Fijian or vakatevoro names, this was unlikely to be due to chance. When I asked about it, I was told that the ancestors of these people had sought refuge in my host village during the unsettled period of the 1870s in which Christianity and central government were extended into the hill country of Viti Levu. During their stay, they named some of their children after their hosts.

Finally, on this matter of the diffusion of names between descent lines, it is necessary to consider the relationship between personal names and traditional titles. In Matailobau, chiefly rank is inherited patrilineally and carries with it the right to use the honorifics ratu (‘sir’) or adi (‘lady’) before one's personal name. In any context in which a personal name can be used appropriately as a term of reference or address, the name of a chiefly person will be preceded by the title. In a sense the title is part of the name. Once I heard a young commoner named X refer to an elderly man of chiefly rank as noqu yaca (‘my name [sake]’). Another young man, this one of the chiefly mataqali, said, “What do you mean — ‘your name’? He [the elder man] is Ratu X!” This relationship between titles and names also affects teknonymy; thus, the father of a boy named Ratu Viliame is known as tamai 'tu Vili, ‘father of Ratu Vili’).

Sometimes the title is, indeed, incorporated into the name itself. In this area of Fiji it is possible for some of the lineages within a chiefly clan (mataqali turaga) to be untitled (Nayacakalou 1978:84); that is, their members do not have the right to use the honorifics “ratu” and “adi” before their names. The position of such untitled lines may appear ambiguous, but, in the emic view, there is no contradiction in the fact that they are ‘chiefly’ (turaga) in one sense (i.e., they are - 14 full members of a chiefly clan) but not in another (i.e., they are tamata lailai or commoners rather than turaga, persons of chiefly rank). (See Walter [1978] for a discussion of the contextual relativity of the Fijian concept of turaga.) Children are named after senior members of their own clan, and it is possible for members of commoner lineages within the mataqali turaga to be named after chiefly (that is, titled) members of their clan. In such cases, the title becomes part of the name. Thus, in the village in which I lived, there were two men of the chiefly clan called Ratu Saimone. The senior man was of chiefly rank while his younger namesake was not. In such a case, the children of the commoner are not called “ratu” and “adi” unless they are also named after titled members of the clan. In the case to which I am referring, that has not happened.

As members of their father's descent group, the children of chiefly women married to commoners do not inherit their mother's rank. But the status of illegitimate children can be more ambiguous. An illegitimate child (luve ni sala, ‘child of the path’) is one born to a woman who is not openly cohabiting with a man in a recognised union and whose paternity is not formally acknowledged. Since there is no pater, the child does not inherit his or her rank from the father. Such a child may be adopted by a married couple, in which case he will be included in the adoptive father's mataqali and named accordingly, or he may stay with the mataqali of the mother. The illegitimate child does not inherit his mother's chiefly rank and, properly speaking, does not have chiefly rank to pass on to his own descendants. But he will be named after a member of his mother's clan (as will his children), and he takes the namesake's title as part of his name. In successive generations his illegitimate origins (and those of his line) may be forgotten, disguised by the system of naming.

To a limited extent, then, names diffuse from one patrician to another. Titles also diffuse, but in a more limited way (i.e., from titled to untitled lineages within chiefly clans). But most people are given names borne by other members of their own clan, and most persons whose names are preceded by titles are persons of chiefly rank. In theory and in fact, mataqali (clans) are differentiated by the names of their members. And to identify an individual as a member of a particular clan, as names do, involves expectations about predispositions for behaviour at the same time that it locates him in a hierarchical model of society. In a sense it locates the person in physical as well as social space, for mataqali are associated with particular places, and place-names may be incorporated in the formal names of mataqali (Turner 1988). The individualising function of names is combined with their role in identifying persons as members of social categories. This surely occurs in many other naming systems as well, but, of course, what names “say” about persons in Matailobau is rooted in the specifics of that cultural system.

- 15

A full contextual analysis of the language of personal identification lies beyond the scope of this paper, but it should be noted that, in Fiji, personal names are only one of several kinds of labels that can be used in reference or address. Pronouns (marked for person, number, and inclusion or exclusion of the person spoken to), honorifics (e.g., saka, ‘sir’), kin terms, teknonyms, or nicknames may be more appropriate forms of address in certain contexts, and pronouns, kin terms, teknonyms or nicknames may also replace personal names as terms of reference. For example, one always addresses and refers to a close kinsman of a senior generation by the appropriate kin term rather than by his personal name (cf. Ravuvu 1971). An unmarried kinsman of a junior generation may be addressed or referred to by his given name, a nickname, or a kin term. Mutual use of personal names is a sign of equality among the young, but adults with living children are usually addressed and referred to teknonymously. That is, parents are known as the ‘father of’ (tamai...) or ‘mother of’ (tinai ...) their firstborn child or ulu matua (lit. ‘mature head’). As a person ages he comes to be referred to and addressed in terms of his relationship to his first-born grandchild. Thus, a man once known as tamai Sevanaia is now known as tubu i Vani (‘father's father of Vani’) after his eldest child's own ulu matua.

While personal names in this culture identify persons in terms of their positions within a system of descent groups, teknonymy labels persons in terms of their relationship to other persons. Thus, at the same time that the teknonym designates an individual, it does so in terms of one aspect of his personhood, his status as a parent. What cultural explanation is there for this shift to a relational definition of identity? My informants explained the avoidance of personal names as a form of respect. It calls attention to the fact that, as a parent, one is a completed person. Parenthood is a valued stage in the normal life cycle; children are highly desired, and people attribute envious feelings to those couples unable to have them. But teknonymy not only calls attention to the parent-child relationship. It also calls attention to the relationship between spouses, for parenthood is not something that one achieves alone. Parenthood not only completes the person, but also represents a completion of the conjugal unit. The child of that union is the source of a bond of common substance between persons of different matagali, and, in his role as vasu (sister's child), he represents a politically significant tie between two clans (Turner 1986a).

There is some evidence that the use of teknonymy has increased in this area during the colonial era, probably as a result of increased contact with coastal peoples. A. B. Brewster, an official in the colonial Administration who knew the Matailobau area in the 1880s, noted that, at the time, teknonymy was less common there than in coastal areas. In the hill country of Viti Levu men tended - 16 to be known by their personal names (Brewster 1922:181). But, today, the practice of teknonymy is so prevalent that children may not know the personal names of adults, knowing them (other than as kin of a specific category) only in terms of their children's names.

Hildred and Clifford Geertz (1964) have argued for a causal connection between the extensive use of teknonymy and “genealogical amnesia” among Balinese commoners, pointing out that genealogies depend upon knowledge of the name of one's ancestors and their collaterals, and teknonymy suppresses such knowledge. In Matailobau, too, genealogical knowledge is very shallow. Most persons do not know the identity of ancestors more remote than their grandparents and have little knowledge of collateral relationships in the grandparental generation. 4 I think that, in Matailobau, however, “genealogical amnesia” simply reflects a lack of interest in pedigrees rather than the effects of teknonymy. While young children may not know the personal names of the parents of some of their playmates, they will, as they mature, learn the personal names of all of the adult members of their village and of other villages in the immediate area.

If the use of teknonymy does not affect genealogical discourse, the same cannot be said for the personal names that they replace in everyday use. On the one hand, I found that the repetition of names within generations and between successive generations could create confusions in genealogical research. (One is reminded of Evans-Pritchard's [1956:164] observation that the Nuer do not name sons after their fathers lest they become merged in the recitation of genealogies.) On the other hand, I was aware that my informants sometimes used this feature of their naming system (i.e., the recycling of names) as a mnemonic device when trying to reconstruct genealogical relationships. Knowing, for example, that a limited number of names are used again and again in a given mataqali helps in remembering the name of a woman married to an ancestor. How this affects the accuracy of what is “remembered” I cannot say. But, if genealogical knowledge is not extensive, it should be remembered that, in a system such as this in which names represent positions in the social life of the community that a succession of individuals occupy, the names themselves perform one of the functions of genealogy. For, if one purpose of genealogies is to allow individual identities to escape oblivion, at least for a time, the extensive recycling of names represents another way of “keeping a name alive”.

While names identify persons in Matailobau as members of particular categories, and teknonyms identify them in terms of their relationships to others, nicknames have the potential for expressing the uniqueness of individuals. No adult man in the village in which I resided was habitually referred to by nickname, though some had been known by them earlier in life. The only context - 17 in which I heard men's nicknames used was in the form of teasing behaviour where I was told the nickname in the person's presence by someone in a joking (cross-cousin) relationship to the victim. Nicknames are used for the young and unmarried, however. Nicknames may refer to the individual's appearance or habits or to particular events in his life. (For example, one small boy was frequently referred to as Karasini [‘Kerosene’] because he had once mistakenly drunk from a container of that fuel, thinking it to be water.) Of more interest to us here, however, is the fact that, in Matailobau, nicknames often refer, not to the individuality of the bearers, but to the persons for whom they are named. For example, children and young, unmarried people are often referred to and addressed by the term for their namesake's kinship relationship to them. Thus, a child named after his father is often called Tamana (‘His Father’), and his father, in turn, can be referred to as tamai Tamana (‘father of His Father’). In the same way, children are also referred to as Ganeitamana (‘Her Father's Sister’), Tubuna (‘His Father's Father’), or Buna (‘Her Mother's Mother’, a term that can also be extended to a person's father's father's sister [FFZ]). For some individuals these nicknames based on kin relationships become the principal label by which they are known.

Young people may also be known by the titles associated with positions held by their namesakes. Thus, there is one young man in the village who is addressed and referred to almost exclusively as Buli because the man for whom he was named, his father's father's brother's son (FFBS), had occupied the office of district chief (buli) in the colonial Fijian Administration. In another case, a young girl is known almost exclusively as Nasi (‘Nurse’) because that is the occupation of her namesake, her father's father's brother's daughter (FFBD). Similarly, there are several boys and young men in the village known as i Talatala (‘Pastor’), named after pastors who had been assigned to the local church.

The child's identification with the individuality of the namesake (as opposed to a status, an aspect of the namesake as a person,) is apparent in those cases in which young people are known by their namesakes' nicknames. There is one young woman in the village who is often referred to as Baki because that was a nickname acquired by her nei (father's sister), for whom she is named. In the local dialect baki is 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 frequently rolled out dough for roti, Indian flatbread, a favourite breakfast food. (See Goodenough [1965] on the inheritance of nicknames in Lakalai.)

It bears emphasising that, while nicknaming has the potential for stressing the uniqueness of the individual, in this culture it frequently has the effect of subsuming it within the identity of a person's namesake. This amounts to an - 18 acknowledgement of the fact that

a personal name can never stand independent of the social personalities with whom it has been associated. Extrinsically [the name] connects the experience of a living individual with a socially significant past (Ellen 1983:41).


All human societies face the problem of incorporating impermanent individuals within a persistent (though transformable) structure. Systems of naming play a role in this incorporation. Since each individual is physically and biographically unique while at the same time occupying a stereotyped position in society, Strathem (1970:59) points out that it would be appropriate if names simultaneously designated unique identities and classified persons in terms of social structure. In fact, systems of naming often stress one or the other of these; that is, they either emphasise the uniqueness of individuals or, conversely, categorise persons in terms of statuses that they occupy. The question of whether each of these types of naming systems is associated with particular features of social structure has not been tested cross-culturally so far as I know (cf. Kuschel 1988), but Goodenough (1965), who compared two Oceanic naming systems differing with respect to this dimension of contrast, offers a specific hypothesis. On Truk, where individuals are consistently subordinated to the groups to which they belong, names call attention to the unique identity of every individual. In Lakalai, by contrast, individuals are free to compete for renown, and naming customs and modes of address there “function to offset competition by serving as continual reminders that people are, after all, part of a social order . . .” (Goodenough 1965:274–5).

Goodenough, then, sees the key to understanding naming practices in these two societies to be a tension between the desire for individual expression and the need to assert group values. But he is careful to point out that naming practices need not always reflect this tension, and I do not think that the naming system of Matailobau should be interpreted in these terms. The people of Matailobau are very conscious of one another's foibles and failings as individuals, and, indeed, these provide a constant source of material for discussion. On the one hand, the social system, structured in terms of rank, seniority and gender, limits individual expression in many contexts, but, on the other hand, Fijians value individual choice within the limits set by those parameters. Indeed, the word “veitalia” (‘suit yourself’) is the watchword in many social situations. Despite this, individual Fijians are also keenly aware of the fact that their desires for self-determination sometimes conflict with obligations entailed by the social structure. But, rather than emphasising the uniqueness of individual identity and, - 19 thus, “compensating” for strictures imposed by the social structure, naming practices and modes of address emphasise one's position within society.

Goodenough's interpretation of Trukese and Lakalai naming practices is based on the recognition that names are a form of communication. Whether one uses a personal name or some other mode of address can convey a specific attitude (e.g., respect, familiarity or affection). The choice of which of several names or forms of a name one uses can also express shades of meaning. Names themselves convey meaning irrespective of any semantic content they may have, for the system in which they are embedded will be structured in terms of “something about which people are concerned, something about their own identities or the identities of others that they want to emphasize” (Goodenough 1965:275). In the case of Matailobau, that emphasis or concern is social continuity — that is, a concern with the individual's place within the enduring framework of society and the chain of descent linking the ancestors to future generations.

It is important to remember that the naming system that has been described here is specific to a particular period of time as well as to a particular region. The binomial aspect of the system is no older than the 1870s, being a product of the conversion to Christianity. The way that names are used has also undergone changes; we have noted that, in this general area, the practice of teknonomy was less common in the 19th century than it is today. Brewster (1922:178–9) also notes that, among the hill tribes, men were often referred to, not by the ancestral name formally given to them by their fathers' clans, but by nicknames given to them at birth by the women attending their mothers. These nicknames were suggested by the new-born infant's appearance or behaviour. Brewster (pp.179–81) also mentions the fact that, in at least one hill district, Noikoro, initiates were given new names after undergoing circumcision. Unfortunately, he gives no information on how these names were selected or how the acquisition of new names affected a person's designation in everyday discourse. He makes no mention of this practice of renaming occurring in Matailobau.

Clearly, there have been some changes in naming practices and modes of address since European contact, but there has been continuity as well. In the past, as in the present, mataqali owned names and bestowed them on each generation of new members. But the system as it exists now crystallised during a period of change that involved conversion to Christianity, rapid depopulation, and the instituting of colonial administration. In its blend of new features and specific elements of tradition, the naming system that evolved asserted the continuity of society in the face of these dramatic changes. Later, long after the present system of naming had already taken shape, the colonial Government created a system of land tenure that involved an imposed systematisation of Fijian social - 20 organisation and the formal registration of descent group membership. This registration preserves a written record of names that can have an effect on name bestowal. 5

It is useful to compare the naming practices found in Matailobau and the system that Lindstrom (1985) describes for the island of Tanna in Vanuatu, for, despite differences between the two societies, the naming practices in both cases exhibit a concern with social continuity. Lindstrom emphasises the contrast between the principle of rank found in Fiji and the egalitarian nature of Tannese society, but shows how this is related to the equally significant difference in their respective principles of group recruitment. On Tanna a man's local group membership is determined by naming rather than descent. Lindstrom refers to these local groups as name-sets. Each name-set controls a stock of names to confer on new members, a body of garden land, and other forms of property. Each name owned by a name-set carries with it rights to specific plots of land and, thus, it is through naming that men acquire rights in land. Several name-sets share a single kava-drinking ground that serves as a site for nightly kava drinking, ritual, and as a burial ground. If a name-set becomes defunct, the men of other name-sets localised at the same kava-drinking ground have the right to bestow the names of the defunct group and, thus, recreate it.

Lindstrom points out that a system of groups based on naming is better able to assert itself against the vicissitudes of history than systems based on descent. The survival of descent groups depends on the accidents of birth, and, once a descent group disappears, it cannot be recreated. Unilineal descent groups can survive in the face of demographic perturbations, but this may require institutionalised fictions such as incorporating the children of female agnates in a patrilineal system. The survival of the group depends on a temporary suspension of the rule of recruitment. In a system based on naming this flexibility is built into the principle of recruitment. The point that Lindstrom stresses is that the system of naming on Tanna allows for the constant reproduction of the past in the present (i.e., a particular configuration of groups in which successive generations of members occupy named positions).

The system of name-sets is able to adjust the distribution of population among local groups through adoption and renaming. But, under certain demographic conditions, the relationship between names, group membership and rights in land undergoes a transformation.

Steady population increase causes local groups to run short of names with entitled land rights. Neighbouring name-sets, over-populated themselves, can no longer absorb the over-production of individuals through adoption (Lindstrom 1985:39).

Thus, in areas of high population density, naming no longer serves as the basis

- 21

for group membership, and, instead, patrilineal descent determines group affiliation and rights in land. This suggests that perhaps patrilineal descent was already present in Tannese consciousness as an alternative structuring principle. One wonders to what extent the system of name-sets was itself a response to a demographic trend: depopulation. Lindstrom (1985:43) acknowledges that 19th century epidemics and the resulting population decline may have had a transforming effect on local group structure.

Lindstrom is correct in noting that a system based on nomination is easier to manipulate (so as to reproduce it) than are systems based on descent, but the latter, too, can use naming as a means of reproducing the past in the present (cf. Leenhardt 1979). Such is the case in Matailobau, where naming is consciously involved in the reproduction of society by conferring personhood on new individuals, while, at the same time, asserting the continuity and discreteness of descent groups. Namesakes are almost always agnates, frequently deceased agnates, and it should be noted that the relationship between the living and their ancestors is one of “participation” in Lévy-Bruhl's (1966) sense of that term. The ancestors have an interest in the actions of their descendants and may punish them for a failure to honour their obligations to living kinsmen. Similarly, they look out for the interests of the living and may cause misfortune to befall those who offend or anger their descendants. But, just as the actions of the living are of interest to the ancestors, so do the past actions of the ancestors affect the lives of the living. The impact of the ancestors' actions is not just a matter of empirical cause and effect. The living may also suffer misfortune, arising from mystical causes, because of the misdeeds of the ancestors. (See Turner [1986b] for a case that illustrates this.) Like most Melanesian peoples, the Fijians of Matailobau believe that their health and well-being depend on maintaining good relationships with their ancestors. This can be ensured by honouring their customs, especially exchange obligations, for the very act of ceremonial exchange is considered pleasing to the ancestors (Turner 1987).

Not only do the actions of the ancestors affect the living, and vice versa, but in certain contexts the identity of the living descendant merges with that of the ancestor. That is true, for example, when a chief accepts an offering in the context of ceremonial exchange, or when the elders of those clans considered to be the owners of the land accept the first-fruits of the yam crop (Turner 1984). This merging of identities occurs not only in the context of ritual but also when descendants suffer from spirit possession by an ancestor. Kava (yaqona) or other goods offered to the victim are meant to propitiate the ancestor and bring an end to the state of possession. This identification between the living and dead can also be seen in the tendency to refer to events in the life of an ancestor in the first person: “I did such-and-such” or “After leaving X, I went to Y” (cf. Sahlins - 22 1985:47 and 1962:254; see also Lindstrom 1985 for a description of the same phenomenon on Tanna).

While the living members of the mataqali are identified with their ancestors as a general category, a name is a special bond with specific ancestors. Despite the fundamental social transformations that were occurring during the period in which it took shape, the current system of naming in Matailobau asserts the link to the past and to the ancestors who play a role in providing or withholding prosperity and well-being. The ancestors are made present in their names and in the persons of their descendants who bear them. The teknonyms that frequently replace names also assert continuity by identifying persons in relation to their children and grandchildren—that is, with respect to specific descendants rather than ancestors. Individual identity is socially important but muted by a naming system that stresses social position. To paraphrase Sahlins (1985:146), by encompassing the existentially unique (i.e., the individual) within structural constants, the people of Matailobau embed their present in their past.

- 23
  • Brewster, A. B., 1922. The Hill Tribes of Fiji. London: Seeley, Service.
  • Ellen, R. F., 1983. Semantic Anarchy and Ordered Social Practice in Nuaulu Personal Naming. Bijdragen tot de Taal-, Land- en Volkenkunde, 139:18–45.
  • Evans-Pritchard, E. E., 1956. Nuer Religion. Oxford: Clarendon Press.
  • Fortes, Meyer, 1969. Kinship and the Social Order. Chicago: Aldine.
  • Geertz, Clifford, 1966. Person, Time, and Conduct in Bali: An Essay in Cultural Analysis. Cultural Report Series No. 14. Southeast Asia Studies. New Haven: Yale University Press.
  • ——1984. ‘From the Native's Point of View’: On the Nature of Anthropological Understanding, in Richard A. Shweder and Robert L. Levine (eds), Culture Theory: Essays on Mind, Self, and Emotion. Cambridge: Cambridge University Press, pp.123–36.
  • Geertz, Hildred and Clifford Geertz, 1964. Teknonymy in Bali: Parenthood, Age-grading and Genealogical Amnesia. Journal of the Royal Anthropological Institute, 94:94–102.
  • Goodenough, Ward H., 1965. Personal Names and Modes of Address in Two Oceanic Societies, in Melford E. Spiro (ed.), Context and Meaning in Cultural Anthropology. New York: Free Press, pp.265–76.
  • Hocart, A. M. 1929. Lau Islands, Fiji. Bulletin 62. Honolulu: Bernice P. Bishop Museum.
  • Kuschel, Rolf, 1988. Cultural Reflections in Bellonese Personal Names. Journal of the Polynesian Society, 97:49–70.
  • Leenhardt, Maurice, 1979. Do Kamo. Chicago: University of Chicago Press.
  • Lévi-Strauss, Claude, 1966. The Savage Mind. (La Pensée Sauvage). London: Weidenfeld and Nicolson.
  • Lévy-Bruhl, Lucien, 1966. How Natives Think. New York: Washington Square Press.
  • Lindstrom, Lamont, 1985. Personal Names and Social Reproduction on Tanna, Vanuatu. Journal of the Polynesian Society, 94:27–45.
  • Nayacakalou, Rusiate, 1978. Tradition and Change in the Fijian Village. Suva: Fiji Times & Herald.
  • Ravuvu, Asesela, 1971. Dependency as a Determinant of Kinship Terminology. Journal of the Polynesian Society, 80:480–4.
  • ——1983. Vaka i Taukei: The Fijian Way of Life. Suva: Institute of Pacific Studies, University of the South Pacific.
  • Sahlins, Marshall D., 1962. Moala: Culture and Nature on a Fijian Island. Ann Arbor: University of Michigan Press.
  • ——1985. Islands of History. Chicago: University of Chicago Press.
  • Strathern, Andrew, 1970. Wiru Penthonyms. Bijdragen tot de Taal-, Land- en Volkenkunde, 126:59–74.
  • Turner, James W., 1984. ‘True Food’ and First Fruits: Rituals of Increase in Fiji. Ethnology, 23:133–42.
  • ——1986a. ‘Owners of the Path’: Cognatic Kinship Categories in Matailobau, Fiji. Oceania, 56:294–303.
- 24
  • ——1986b. ‘The Sins of the Father’: Rank and Succession in a Fijian Chiefdom. Oceania, 57:128–41.
  • ——1987. Blessed to Give and Receive: Ceremonial Exchange in Fiji. Ethnology, 26:209–19.
  • ——1988. A Sense of Place: Locus and Identity in Matailobau, Fiji. Anthropos, 83:421–31.
  • Walter, Michael A. H. B., 1978. An Examination of Hierarchical Notions in Fijian Society: A Test Case for the Applicability of the Term ‘Chief.’ Oceania, 49:1–19.
  • Williams, Thomas, 1858. Fiji and the Fijians. 2 Vols. London: Haylin.
The New Zealand Journal of History

Vol. 24, No. 2, October 1990 contains:

Graeme Davidson Cities and Ceremonies
Jock Phillips Of Verandahs and Fish and Chips and Footie on Saturday Afternoon
M.P.K. Sorrenson Giving Better Effect to the Treaty
Alan Ward History and Historians before the Waitangi Tribunal
Marcia Stenson History in New Zealand Schools
James Belich Hobson's Choice (Review Article)

The New Zealand Journal of History is published twice yearly, in April and October, by the University of Auckland.

Subscription rates for 1991, payable in advance, post free: Domestic $24.00 (incl. GST); Overseas $NZ30.00. Back numbers available: $8.00 (incl. GST) per issue in New Zealand and $NZ10.00 overseas.

Subscriptions and all business correspondence should be addressed to the Business Manager, New Zealand Journal of History. History Department, University of Auckland, Private Bag, Auckland, New Zealand.

1   This paper is based on field work supported by the National Science Foudation and conducted from November 1979 to March 1981. I wish to thank the people of Nairukuruku village for their hospitality and tolerance during our stay. I also wish to thank Monty Lindstrom for reading and commenting on an earlier version of this paper.
2   Of 44 married men in the village in which I resided, there is only one case in which a man and his children have affiliated with the wife's natal mataqali, and, in that case, the man is from another area in Fiji. Two other men reside with local groups in which their mother's agnates constitute the patrilineal cores. The paternal grandfather of one of these men was a European, and his position is necessarily anomalous. The other man ‘stays with’ his mother's brother and has been joined by a classificatory son of his own mataqali and the younger man's wife and child. The younger man's wife is from the village, as is the wife of his classificatory father, and it is significant that neither man stays with his wife's mataqali
3   The official registry of mataqali membership (Ai Vola ni Kawa) maintained by the Fijian Administration lists the names of 167 individuals in the chiefly mataqali in the village in which I resided. Of that figure of 167, there are only two instances in which persons have been given vakatevoronames that are usually given to members of the opposite sex. In both cases they are females who have been given names typically used for males.
4   While people's knowledge of relationships in their own generation is often very extensive, it is frequently not based on knowledge of genealogical relationships in ascending generations. In order to know what kin term to apply to a particular alter, it is sufficient to know how some other relative classifies that person. It is not necessary to know what genealogical basis, if any, there is for this classification.
5   One informant told me of an instance in which two children of her mataqali living in a neighbouring village were given names that “not even our fathers knew.” The informant explained that an old man of the mataqali had the names written down in a book that I suspect was a copy of the Ai Vola ni Kawa. The old man realised that the names of these ancestors had been ‘lying dormant’ and so bestowed them on the two children.