Version classiqueVersion mobile
OpenEdition Books


João de Pina Cabral
Antónia Pedroso de Lima

Part II. Choice and Tradition

6. Making The Chief: An Examination of Why Fijian Chiefs Have To Be Elected

Christina Toren

Texte intégral

1The question of why Fijian chiefs have to be elected arises in part out of the recent history of Fiji, where in 1987 a middle-ranking chief from the western side of the main island of Viti Levu, Dr Timoci Bavadra, was ousted in a series of bloodless military coups from the office of Prime Minister, to which he had been elected only a month before. Until this point Fiji, whose first general election had taken place in 1972, had apparently been governed successfully by a parliament based on the Westminster model bequeathed to their ex-colony by the British. What had gone wrong?

  • 1 Population 1988: 342,965 ethnic Fijians, of whom 31.4% lived in urban areas; 341,141 Indo-Fijians, (...)
  • 2 See, for example, Kaplan 1987, Kelly 1988, Norton 1990 and Lal 1992.

2Explanations have centred on relations between ethnic Fijians and Indo-Fijians (the descendants of indentured labourers brought to the country by the British in the late nineteenth and early twentieth centuries), who at the time formed something more than half the population.1 But it is not my intention to address here the politics of ethnicity (which have been ably discussed in depth by many others) or the restoration of democracy in the general election of 1992 on the basis of a revised constitution.2 Rather, I want to show how it is that ethnic Fijians can be committed to the idea and practice of democracy and at the same time insist that only high-ranking Fijian chiefs are fitted to hold the highest offices of state. And further, I want to argue that there is neither contradiction nor paradox here, for this is another manifestation of an old Fijian idea in which the continuity of a properly constituted polity depends on equal and hierarchical relations being made to play against one another in such a way as to produce the peace and plenty that characterize a successful Fijian chiefdom.

Chiefship in the Old Days

3I begin by considering how men born into a chiefly clan come to be accorded the status of paramount. The following remarks are extracted from a conversation I recorded in 1983 with Ratu Sefanaia Laua of the chiefly yavusa Nadawa in the chiefly village of Sawaieke on the island of Gau, central Fiji, where I did fieldwork. We were discussing na gauna makawa, 'the old days', that is to say, the pre-colonial days, and their significance for contemporary chiefly statuses in Sawaieke country (vanua- here referring to a confederation of eight villages).

[The] priest [to the ancestor god/chief] was Tunitoga, if the chief wanted anything he told Tunitoga, it was up to him then to decide about it. There was his herald, Tunimata. The chiefly status then following was the sauturaga, Naraitena [whose prerogative it is to install a high chief]. If the [high] chief was absent then Narai could lake over. Chiefs (Turaga), Raitena [designated owners – itaukei - of the country], Priests, Heralds, Warriors, then the sea people and the landspeople - liga ni magiti, hand of the feast... [W]hen there are seven statuses like that, you have a government (matantitu)...
In the old days there was no country that did not fight. All the villages fought one another... this was the old way of it: the chiefs usually sought to take a girl to marry in a village and thereby perhaps that village would be silent, because the girl was married there. Its pledge was the woman. Yes, you will see this in [the village of] Nukuloa, they have there a yavusa [group of clans connected by ritual ties and intermarriage] which is the fellow of yavusa Navure [in Sawaieke]. They were carried there that they might go as a war-fence. This is the proof of it; there is... mataqali [clan] Sisiwa, tokatoka [sub-clan] Sisiwa. The family of the Retired Methodist Minister... is living in Nukuloa because of this... The Minister is dead, but his lady is living. He was made Tui Nuku [installed chief of Nukuloa] when his strength was going.
CT: Why was no one else installed since as Tui Nuku?
RSL: It's up to the gentlemen in Nukuloa. You see here we have Takalaigau [title of the high chief of the eight villages of Sawaieke country, including Nukuloa]: he is still only called Takalaigau, but he has not yet drunk, the power has not yet been given to him. but that will then just be up to the
sauturaga [to the commoner yavusa who install the high chief]. If they find that he is a good chief, that he cares for the country, they will not delay in making him drink. If his bearing seems to be bad then they will stand back that they might watch for a while. It's possible they might simply cancel [their initial selection] and take another chief.
CT: Is there someone in Nukuloa who can take the title Tui Nuku?
RSL: Yes there is, but the politics of the matter are such that they may consider it and say it's not a good idea. I understand that European leaders such as Prime Ministers of different governments are also installed like that. CT: What about in Navukailagi - Na Ratu?
RSL: Yes, Navukailagi also, if it is correct according to the land that there is a good chief there, it is right that he should run their country.
CT: What do you think-will a chief be installed here as Takalaigau?
RSL: Yes, I believe so, since there is something that they all know-[the putative] Takalaigau and the land chiefs here. The late Takalaigau went around all the villages and strove to ensure that when someone was to be installed as paramount
(Tui), this country should be of one voice. If not, each one would just go its own way. According to custom the status of one who has been installed [is such that] he is promised the people, the earth, the grass that he may be its overall leader. This [the current situation in Sawaieke] is called sauvi - marked out, standing in wait. Here Sawaieke is just marked out. [The installation] is being considered to see if it is right or not. It takes a long time but it is not my task to go and tell [them] that it should be done...
You have been here a long time in the land, [you] know that it is a life that envelops with love. If a chief is good, everything will be his. Look, here is a good proof of it: [suppose] one of these people falls ill, but his behaviour towards the people of the country has not been always good, they will just see that he is lying on his bed [but] they will not be concerned. This is one proof of it that I have already seen here (cf. Appendix 1).

  • 3 The ritual obligations peculiar to certain clans do not constitute castelike distinctions; so, for (...)
  • 4 Observations similar to those in this paragraph have been made by numerous other ethnographers of F (...)

4I have quoted from this conversation at length because it illustrates a number of points. Firstly, it suggests that the various clans (mataqali, lit. kinds) owe certain ritual obligations to each other.3 Secondly, it shows that Fijian chiefship is not a matter of predictable lineal succession from father to son. Other high-status men, including the chiefs of commoner clans, in particular those whose task it is to install a high chief, have a significant say in who is to be made paramount. If they are unsure of a man's fitness for office, they will delay his installation - for years if not for ever. This is an indication of the relation of reciprocal obligation that obtains between a high chief and his people - if a chief does not fulfil his obligations then the people withhold their service. Thirdly, chiefs are considered necessary to the continuing autonomy, prosperity and well-being of a country, and are expected, especially in times of war, to provide protection. Fourthly, the specific references to marriage as alliance alert us to the necessity of looking at the nature of the relations it implicates. Finally, Ratu Laua's assertion 'that it is a life that envelops with love' (ni dua na bula e ologa ena loloma) implicates the idea that the fundamental conditions of social life are given by kinship, which ideally encompasses all relations between people, including marriage, and which ordains, as it were, for any particular kinship relation, the precise nature of the duties that are entailed by veilomani, 'caring for each other' or 'loving each other'. The Fijian term for kin, weka, may be extended to include all ethnic Fijians, and kinship terms are routinely used in both reference and address; with the exception of the relation of competitive equality between cross-cousins, all kinship relations are hierarchical.4 I discuss these matters further below. For the moment I return to ethnographic sources.

5In his unpublished The Heart of Fiji, which concerns the chiefdoms or, as Fijians say, countries of Lomaiviti (i.e. central Fiji, of which Sawaieke country, on the island of Gau, is a part), Hocart quotes from his translation of an article in Na Mata of 1906:

  • 5 George Millier, author of the Fijian Grammar and of a forthcoming Fijian Dictionary, informs me (pe (...)

Matanitu, lit. Face of the Chief, or State [CT: 'powers that be']. In the old days men used to seek their chiefs, they used to betrothe ladies from various lands where there was a noble line to marry into their own land so that her children might be noble and their land become a matanitu land thereafter. Some put at their head a successful (yaco) man or strong man in battle and installed him as their big lord or chief, because they wished to become a matanitu. Therefore the leaders of the various clans met together and made over their villages to him whom they had installed as their chief... As regards mata the beginning of the word, this refers to all the people of the land who are to be the mala or messengers of the great lord or chief (tui). The true spelling should be mata-ni-tui... (pp. 327-8 of manuscript).5

6Moreover Hocart records many instances where the offices of high chief (Roko Tui, in his translation 'noble lord') and/or war chief (Vunivalu) are supposed to move between different clans. For example:

[On the island of Batiki] [t]he chief clan is divided into two; Upper Tora ni Bau and Lower Tora ni Bau, by reason of brothers being appointed to live some above and some below. The first chief remembered comes from the lower branch. He was followed by one of the upper branch, then three of the lower branch (ibid., p. 294).

[In Verata] [t]here are two houses from which the chief is chosen, Vunivalu and Sanokonoko. The kava goes from one to the other. If a man from one house is installed the other house 'carries the sovereign' (cola sau). The two together form the clan of Vunivalu. They say two chiefs cannot be appointed successively from the same clan, but in practice it happens. In order of clans Mataniwai was placed first. It was described as 'substance nobles' [turaga lewena (CT: 'substantial chiefs', 'chiefs of the blood')]. They used to hold the chieftainship while the clan of Vunivalu upheld the heraldship and the sovereignty to the saw, but this rule was weak so they gave the chieftainship to the clan of Vunivalu while they upheld the sau (ibid., pp. 400-1).

There are two chiefs in Rewa: the Noble Lord of Dreketi of the clan of Bi House, and the War Chief of Nukunitabua. They came together from Nakauvadra. They divide Rewa and all subject villages and carpenters between them, there is not a village in which they have not both got authority. They are the two cloths of the land... There arc four noble clans besides the two clans to which the chiefs belong... Next comes a group of six clans called sauturaga. They decide who shall be Noble Lord of Dreketi and who War Chief; they instal the War Chief (ibid., pp. 431-2).

  • 6 Hocart lived in Fiji from 1909 to 1912 and did further fieldwork there in 1913.

7Hocart was well aware that historically Fijian chiefship was a matter of ritual precedence, rather than unalloyed hierarchy; nevertheless, he argues in Kings and Councillors that government evolves out of ritual organization in such a way that equality, or a balanced organization of functions, necessarily gives way eventually to centralization and 'a vertical hierarchy' (Hocart 1970: 37) and, in The Northern States of Fiji, that this process was well under way in the Fiji of his time (ibid. 1952: 58).6

  • 7 Sahlins (1985) and (1991), following Hocart, takes a different view of Fijian dualism: in brief, he (...)
  • 8 In respect of my argument in this chapter, George Marcus's comparative analysis of chieftainship in (...)

8As will become plain below, I take a different view. As I see it, today as in the past, the shifting of high offices between clans and the choosing of those men who are to fill them, are structural necessities, expressions of a polity in which instituted hierarchy and competitive equality have to be understood as aspects of one another. This radical opposition between hierarchy and equality is such that it constitutes a single idea of antithetical duality as fundamental to Fijian social life.7 Thus the dualism so superbly described by Hocart in his various works, but which he saw as having given way to a 'vertical hierarchy', far from disappearing, continues to pervade daily life, and informs, for example, sexual relations, kinship, chiefship and ideas of the person.8

9The argument of the present chapter addresses a point that I have not previously considered. It is that Fijian chiefship has to be elective in the sense of being seen to be conferred by the people on the chief, for only thus can it bring about the condition of 'peace and plenty' expressed in the term sautu, of which the base sau may be a cognate for sau meaning 'a high chief', and also 'the commandment or prohibition of such a chief' (see CapeII 1973). If so, sautu may be rendered literally as 'established sau', which itself suggests that 'peace and plenty' is inevitably the condition of a country where a high chief takes precedence over all and everyone fulfils his or her material obligations to others; here the idea is not that a despot commands and a cowed people do his will, but rather that the willing attendance of the people on one who has been chosen from the appropriate chiefly clan by leaders of the appropriate landspeople clan and installed as paramount, in and of itself produces a genuine prosperity for all.

10That Fijians themselves hold this view is suggested in the formulae of ritual speeches, of which the following are examples. They are extracts from speeches made at the mourning ceremonies for a dead chief; here a man of chiefly status is offering a whale's tooth to the people whose task it is to attend on the chiefly dead:

I am touching, sirs, the string of the tabua, that our chiefs may be healthy, that our country be one of peace and plenty, a country of mutual love [or, that the sau he established in our country, that it may be a place of caring for one another]. The word is already heard. It is effective, it is true (cf. Appendix 2).

11The tooth is accepted by a man of one of the commoner clans that compose the yovusa that attends on the funerals of members of the chiefly clans:

I am untying, sirs, the string of the chiefly tooth, a tooth of gratitude, a tooth that is offered to the chiefs Raitena, and all the people of the houses that are gathered together. We accept that it is a good tooth. Well established are our chiefs, well established is everything in our country. If trouble befalls us, may God hear us, let our leader always be Christ. It is effective, it is true... (cf. Appendix 3).

  • 9 He was at the time a young man of 22 years old or so, who had just returned from a tour of duty wit (...)

12The idea that chiefship is a good thing, and that a well-regulated country is one where chiefs are well-established, is generally held; indeed I have, over the years, had only one informant who seemed to think that chiefship should be abolished and the country run on purely democratic lines.9

  • 10 See Note 5 above.

13Turaga is the Fijian term for 'chiefs'; it can equally well be translated differently according to context, e.g. as 'gentlemen', or as 'married men'. Thus 'the chiefs Raitena' referred to above are elderly and/or high-status men of the landspeople's yovusa Raitena-the men who lead that yovusa; they are commoners, not of chiefly birth. Hocart (1913) makes an argument to the effect that 'the word turanga [sic]... leads us back to gerontocracy; it is properly the title of the old men who sit in informal council over feasts and ceremonies'. Be that as it may, the broader application of turaga to all married men is also suggestive - for it is through marriage and parenthood that a man becomes eventually the head of his own household and 'a chief in his own house' - if not elsewhere.10

14Given that I have already analysed different aspects of marriage as process in a variety of contexts and in some detail, I simply point out here that the hierarchical relation between husband and wife requires a transformation on marriage of the equality of relation that obtains between cross-cousins (by definition, all marriages are between cross-cousins), and that over time the relation between husband and wife, while it remains axiomatically hierarchical, in effect transforms yet again into one between equals. In a lifetime in which all one's relationships prescribe either hierarchy or competitive equality, only marriage makes it clear that each form contains the possibility of the other. But what is crucial for my argument here is that marriage effectively removes from any within-clan reckoning of differential status women who may in their natal households be the eldest of their sibling set. Marriage at once makes women adults and transforms them into axiomatically subordinate wives in their husbands' households, which allows men in general to be ranked within clan according to their relative seniority as heads of households (see Toren 1994a).

15In the ritual formulae quoted above, the first speaker, a man of chiefly rank, juxtaposes the terms turaga and sautu: 'that our chiefs [turaga] may be healthy, that our country may be one of peace and plenty [or that the sau be established in our country, sautu tiko noda vanua], a country of mutual love'. It is interesting to note, then, how the second speaker's use of turaga, and his not using sautu, implies a broader reference. Dei vinaka tu na nodatou turaga, dei vinaka tu na veika kece e nodatou vanua. 'Well established are our turaga, well established is everything in our country.'Spoken as they are by a high-status commoner of the clan that installs a high chief, they specifically imply that 'gentlemen' or 'married men' in general, and perhaps especially high-status men of the speaker's clan, have to be regarded as 'chiefs'.

16In his excellent ethnography of Lau, Steven Hooper maintains: 'that decision-making about the succession and the organisation of the installation is not in the hands of the chiefly clans but lies with the landspeople, [is] a crucial factor which militates against the emergence of a chiefly dynasty wielding dictatorial powers counter to the interests of the people'(1982: 157).

17This is true enough. I want to argue further, however, that the elective element in Fijian chiefship is what allows hierarchy and equality at one and the same time to inhere in the relation between people and chiefs; in other words, the position of paramount cannot be a matter of instituted lineal succession, for this would violate the idea of antithetical duality that informs all relations between people. In such a case, people would be faced with a situation where relations were not merely out of balance, but where it would no longer be possible to live in peace and prosperity. 'In Fiji all things go in pairs or the sharks will bite.' But the country where succession is out of the hands of landspeople clans would be a country where 'all things' no longer 'go in pairs'. It could not be described as sautu, as a country where the sau or paramount is 'established', if people in general are not, of their own accord, content to fulfil material ritual obligations to one another.

18The fulfilment of these obligations of mutual service requires, for example, that men, as husbands, garden and build houses, and women, as wives, fish and make mats; that men of 'fishermen' clans provide big fish for feasts and men who are 'the hand of the feast' provide pig; that each clan undertakes particular tasks in relation to other clans in respect of births, marriages, deaths. It is the fulfilment of their prescribed ritual obligations to others that allows the people of a particular yavusa (group of clans), village, or country collectively to outdo another yavusa, village, or country in, for example, the exchanges that form part of life-cycle ceremonies, or in raising money for some community endeavour, or in amassing goods for the grand exchanges that take place on grand occasions-the opening of the Great Council of Chiefs, or a paramount's installation or death. Where, on such an occasion, a country's showing draws the admiration of people in general, then its people may properly be said to have 'shouldered the carrying pole of empowering the chief' (Hooper 1982: 228).

19In considering chiefly succession, I am struck yet again by the thoroughgoing logic of Fijian dualism, by the way that sameness and acknowledged and, at the same time, differentiated into the hierarchies of precedence that inform both domestic and chiefly ritual. In a situation in which all men who are married and the fathers of children are turaga, chiefs, where the fundamental unit of social organization is the house, and where balanced reciprocity characterizes exchange relations across houses and more inclusive groups (i.e. sub-clans, clans and yovusa), what requires that certain men be singled out as 'true chiefs' and accorded the status of paramount? The answer lies, I think, in the way that, in Fiji, equal and hierarchical relations invariably implicate one another - so much so that at times they threaten to collapse into one another, and the challenge is to keep them distinct.

Equality and Hierarchy

  • 11 The growth in the eighteenth and nineteenth centuries of the large confederations of Bau, Rewa, and (...)

20The competitive equality of cross-cousins informs relations between people whose countries are veitabani (their respective founding ancestors being represented as cross-cousins) and veitauvu (descendants of a common founding ancestor). Against this we have to set the idea that countries should be and are hierarchically ranked, such that the chief of a particular confederation may take precedence over another. This idea is generally held; but even in respect of the chiefs of the largest and most important confederations, there is no unanimity with respect to what the order of precedence might be.11 So where men of the most powerful confederations are obliged to drink yaqona together, who takes precedence over whom is likely to shift according to context, and by virtue of an unspoken and mutual agreement that each will get his turn in the top central position of paramount authority and be served 'the first cup'. Certainly this is what happens when chiefs of smaller confederations meet, for example, at one of the bi-annual Provincial Councils.

  • 12 My initial work on Fiji is an exploration of how, over time, Fijian villagers constitute an idea of (...)
  • 13 All chiefly rituals entail an initial offering of yaqona root as tribute (i sevusevu, a ceremony th (...)

21Given that Fijian chiefship, and indeed all the minutiae of status differentiation between people in the collectivity at large, is at once expressed and constituted in yaqona drinking, where people are inevitably ranked according to their seating position on the above-below axis that describes any space in which the drinking occurs, it follows that there will, always, be an order of precedence.12 That it cannot be a fixed order is a function of the way that the hierarchy constituted in chiefly rituals is bound to come up against the equal relation between cross-cousins that informs exchange across houses and clans inside a country, and across countries.13

  • 14 For example, the first phase of all life-cycle ceremonies (birth, circumcision, marriage, death) la (...)

22In attempting elsewhere to explain what Hocart's informant meant when he said that, 'in Fiji all things go in pairs or the sharks will bite', I pointed out that 'given that all things go in pairs, a totality consists of a pair of pairs, which explains why, in Fiji, everything really goes in fours'(Toren 1994b).14 This observation is crucial here, too, because it is the antithetical relation between pairs of pairs that makes equal and hierarchical relations play against one another and dependent on one another for their very continuity, as follows:

1. Equality between men and women as cross-cousins is in play against the axiomatic hierarchy between husband and wife.

2. Equality between cross-cousins across houses is in play against the precedence of the firstborn and seniority between siblings within houses.

3. Balanced reciprocal exchange across houses is in play against a within-clan hierarchy where houses are ranked according to seniority.

4. Balanced reciprocal exchange across exogamous clans is in play against ritual precedence where the highest ranking clan is 'chiefly'.

  • 15 The term for chiefs here is na malo, 'the cloths'; it refers not just to men who are, by birth, 'ch (...)

5. Balanced reciprocal exchange across clans and countries that relate to one another as 'land' and 'sea' is in play against the relation of tribute and redistribution that obtains between people and chiefs.15

23Finally, the installing clan's prerogative of choosing a putative paramount and installing him (which implies that clan chiefs are one another's peers, each one a household head), is in play against the idea that succession accords with lineal seniority within the chiefly clan and that all statuses are hierchically ranked as a matter of divine ordinance.

24It is the idea that a totality consists of a pair of pairs that allows equality and hierarchy to be implicated at one and the same time in those key kinship relations where hierarchy inside the house directly conflicts with balanced reciprocal exchange across houses. These key kinship relations are those between husband and wife (outlined briefly above) and between mother's brother and sister's son, who are also ideally father-in-law and son-in-law (veivugoni, a term that denotes the relation between children and their parents' cross-siblings).

25The relation between mother's brother and sister's son is the most formal of respect relations between kin, and it is axiomatically hierarchical; the junior relative-the sister's son-owes unquestioning respect and obedience to the mother's brother. The junior avoids the senior, may not address him, sits at a respectful distance front him in any gathering, in general adopts a humble demeanour when in his presence, and is bound to respect his wishes. At the same time, it is this relation that allows the junior relative, as vasu, to vasuta, or, as I was told, to take without asking (taura ga, sega ni kerea) the property of men of his mother's clan. In cases of chiefly marriages across countries, this relation, traditionally and still today, creates powerful allegiances. Thus, as Ratu Laua remarked, women given in marriage constituted 'a war fence', a means whereby an ambitious village or country 'would be silent, because the girl was married there. Its pledge was the woman.' Traditionally, a powerful chief might exploit his vasu relation to his mother's country; but his own obligations of protection and redistribution of tribute had also to observed, if only because of the respect owed to his mother's brother.

  • 16 A situation that recalls Lévi-Strauss's (1977, 1978) brilliant observations concerning the possible (...)

26It can be argued that the child's right as vasu is in effect a recognition of the out-marrying woman's rights in her natal clan (Toren 1994b). Note however that the vasu relation between mother's brother and sister's son is twofold: it combines the licence that is usually proper only to people or groups who relate to one another as cross-cousins with the extreme respect, avoidance and obedience that is properly shown by a real or classificatory son-in-law to his father-in-law. This makes structural sense, because the children of mother's brother are indeed cross-cousins to the sister's son; and mother's brother's daughter, who outside marriage is the peer of her father's sister's son, in marrying him is bound to become his subordinate.16

  • 17 Toren (1998) analyses pre-colonial sacrificial and consumptive practices in Fiji to reveal their im (...)

27That antithetical duality also informs Fijian concepts of the person, especially in respect of existential effectiveness, is evident in the transformations that have occurred in this domain from the pre-colonial period to the present day (Toren 1998).17 In brief, Fijians hold to a theory of the person that rests on recognition of one's relations to others and the obligations that reside in these relations. One distinguishes oneself to the degree that one demonstrates who one is as a function of what one is given to be in relation to others. And in each moment of relationship-be it hierarchical or equal-one strives to assert oneself as the subject of one's own acts, one strives for oneself and against becoming merely the object of the other's manipulation. So in any particular relation with kin-be it between cross-cousins, siblings, parents and children, spouses, or in-laws-ego volunteers the behaviour appropriate to that relationship, and in so doing compels an appropriate response from the other, that is to say a response appropriate to 'a life that envelops with love'. The willing respect and obedience of ego to his or her superior is vakaveiwekani, 'in the manner of kinship', and, by the same token, the evident assertion of another's superiority in giving food, clothing, money, or whatever is undermined by ego's asserting control as at once the implicit initiator and consumer of the gift. Thus one often hears someone say of others, of their parents, their older siblings, their mother's brothers - i.e. of persons to whom they are subordinate - 'they give me everything I want', and if one asks why, one is told 'because they love/pity me'.

Electing the Chief

28Given the foregoing, it follows that if a particular Fijian chief is to be paramount of a country, then both his choosing and his installation have to be initiated by those who, because they take precedence in landspeople clans as 'owners' (original inhabitants of a land), are in a position to accord a paramount precedence in that vanua, country. The people lewe ni vanua, literally, the substance of the country), by virtue of fulfilling their own obligations, at once empower the paramount and compel his fulfilment of obligations towards them all. Only if they have chosen the paramount can the people's initial allegiance to him be known, and only in this case can they effectively withdraw their allegiance should they find him inadequate.

  • 18 The malign power of the ancestor Gods may be unleashed through witchcraft; see Toren 1995.

29A formally installed high chief is a living instantiation of the immanent power of the ancestors, which today comes under the aegis of the transcendent Christian God; as such the mana of a paramount inevitably and properly harms any one of his people who refuses him due service. But if a paramount fails in his duties towards the people, if he holds on to goods and valuables that are given him in tribute and is not seen to redistribute them, then his word is evidently no longer mana, effective, and the people have nothing to fear if they in turn withhold their services. In other words, Fijians know that it is they who, by virtue of their service and willing tribute, empower their chiefs. This idea extends both to the ancestor Gods and to the Christian God: the ancestor Gods no longer have malign power because 'no one attends on them any more' and by the same token the Christian God is all-powerful because of his many millions of worshippers.18

30It seems fitting to quote here the words of Ratu Sir Kamisese Mara, President of the Republic of Fiji, on his own selection and installation in 1969 as Tui Nayau, paramount chief of Lau.

  • 19 Steven Hooper prefaces his extract from this article with the observation that Ratu Mara 'was clear (...)

'The concept of primogeniture - the eldest son automatically inheriting the title and possessions of his father - is not Fijian. It is a European import. In Fijian tradition it was the most able and suitable of the line who received such honours... he was made chief by the people, and by their acclamation and permission only. If he proved unworthy, he did not last long... It was the British era, with its concept of inheritance as a right, instead of a privilege, which froze (or fossilized) the chiefly system...' He went on to relate that he had at first been undecided whether to accept the titles himself. He finally agreed-as the most expedient means of changing the present concept. 'Actually it is not a change, it is a return to the past idea which was much more democratic. The rigidity of the chiefly class is only a hundred years old. It is not what was meant at all' [Nation, Alliance Party magazine, vol.1 (5), July 1969, quoted in Hooper 1982: 158].19

31It will be apparent to the reader that I cannot agree with Ratu Mara's remarks concerning the 'rigidity of the chiefly class', for people who are by birth chiefly do not now, and never did, constitute a class in the analytical sense of the term; they are mataqali turaga - of chiefly kind, and as such, form one of the seven kinds of people (mataqali, clans) who together, as in Sawaieke country, constituted a traditional Fijian vanua: 'Chiefs (Turaga), Raitena (the owners - itaukei-of the land), Priests, Heralds, Warriors, then the sea people and the landspeople... [W]hen there are seven statuses like that, then you have a government.' The material in this chapter has obvious implications for the military coups of 1987 and the subsequent recent history of Fiji. One can argue that, so far as ethnic Fijians are concerned, Dr Timoci Bavadra would not, despite his election by a majority of the electorate and however excellent his personal qualities, be allowed to continue as Prime Minister of Fiji. Here I should note that Fijian villagers do not consider the post of Prime Minister to be political (vakapolitiki) in the sense that it is purely a matter of election; rather, they consider it to be a form of high-chiefly status (that is to say, only those accorded this status should be nominated for election as Prime Minister), and this is apparent, at the outset of this chapter, in Ratu Laua's explicit analogy (in 1983) between the installation of a high chief and a European Prime Minister. In a Fijian perspective, were Dr Bavadra to be accorded precedence over all, then the government could no longer be seen to be properly constituted. This cannot be the perspective of Indo-Fijians; thus Brij Lal's analysis suggests that Fijians were led astray by their chiefs, whose own interests required that the people be kept in ignorance:

For the first time in the modern history of Fiji, a paramount chief... was not at the head of national Fijian leadership. Instead, a middle-ranking chief from traditionally neglected western Viti Levu had been elected to the highest office in the land. In a society where ascribed status, hierarchy and protocol are accorded great weight, and where traditional ideology holds that the business of government is the prerogative of chiefs, and chiefs alone, Bavadra's elevation posed troubling questions... a successful performance on his part would discomfort chiefs who already felt themselves under siege from the corrosive effects of modern life and were struggling to reassert their authority over their people. Bavadra's pledge to educate his people to recognize the difference between their traditional and their political obligations aggravated their fears... at long last, after nearly a century of struggle, the western Fijians were within reach of gaining a measure of the power they felt was their due (Lall992:301-2).

32Despite my general admiration for Lal's history of twentieth-century Fiji, I cannot find that he has any real understanding of the political sensibilities of Fijians. For while it is true that 'traditional ideology holds that the business of government is the prerogative of chiefs, and chiefs alone', in a situation in which all married men are turaga, chiefs, able commoners should also be prominent in government (as Ministers, top civil servants and so on), and so they are; it is notable, too, that the present government includes a number of women in the top posts. My main point here, however, is that Lal's analysis by and large focuses on class and regional interests that, I would argue, pale into insignificance as far as the majority of ethnic Fijians are concerned, and here I include the 'urban-based professionals... people of achieved status and essentially middle class values... [whose] social and material world had widened considerably, beyond the comprehension of their kin who remained close to the communal culture in the isolated villages' (ibid.: 302). My knowledge of such people is limited; nevertheless, those I do know well enough to have discussed these matters with them at once abhorred any violence against Indo-Fijians, asserted themselves to be anti-nationalist and pro-democracy, and rationalized the coups as understandable and their outcome as justified. Lal himself acknowledges this when he points out, without any further comment, that:

although the importance of class and regional factors in Fijian politics should be recognized, they should not be pushed too far. Many urban, middle class Fijians still supported the Fijian establishment and were active opponents of the Coalition government. Prominent western Fijians were among the leading members of the Taukei Movement [the extreme nationalist wing of Fijian politics] even though they privately supported Bavadra's public criticism of eastern dominance in Fijian affairs (ibid.: 302).

33Marcus's view that contemporary Fijian chiefship was shaped in part by the deliberate efforts of colonial administrators concerned to establish efficient indirect rule, and in part by the politics of ethnicity, is undoubtedly correct. Thus I agree with him that: 'the plural nature of Fijian society and the quite explicit atmosphere of ethnic politics cannot be underestimated as factors accounting for the greater vitality of chieftainship, top to bottom, in Fiji than in Tonga. Cultural boundary maintenance is thus an immediate concern among Fijians for whom, as elsewhere, the chief is a central figure' (Marcus 1989: 200).

  • 20 The course of the relationships between Colonel Rabuka, who of his own accord instigated the coup, (...)

34I end on this note in order to suggest that the antithetical logic of Fijian chiefship is necessarily projected on to the institutions of the state. The success of parliamentary democracy from the first general election in 1972, the inevitability of the coups of 1987, and the restoration of democracy in the general election of 1992 on the basis of a revised constitution can be explained by this same logic. Democracy works well in Fiji when it is seen to be under the aegis of Fijian chiefs; but, given that hierarchy itself has never been allowed to become an encompassing value, neither can the ethic of equality be allowed to predominate-as it would if men of high chiefly status were no longer seen to take precedence in government.20

35I have argued in this chapter that the elective element in Fijian chiefship is a structural necessity, an expression of a polity in which instituted hierarchy and competitive equality are mutually constituted and, as such, have to be understood as aspects of one another. It follows that there has to be, as it were, a chiefly element in Fijian elections.


Appendix 6.1

1. RSL:... Mai na gauna oya me vaka au sa tukuna vei kemuni na kena Bete o Tunitoga, sa vinakata na turaga e dua na ka qai tukuna vei Tunitoga, o koya sa qai tu vua na kena lexua. Sa tiko na matanivanua o Tuinimata, na i tutu vakaturaga qai tarava na sauturaga qo na Raitena. Kevaka sa yali tiko na turaga sa na rawa ni 'takeover' ko Narai. Turaga – Raitena – Bete – Matanivanua – Bati - qai ko ira na kai wai vata kei ira na kai vanua-na liga ni magiti. Tautauvata oya ni vitu nai tutu vaka oya qua dua na matanitu... E liu sa sega ni dua na vanua e sega ni vala. Veivala kei ira kecega na veikoro... Ia qo na gaunisala e liu, era dau vakaqara na turaga me ra kau na yalewa ki vakawati ki na dua na koro me rawa ni na galu kina na koro oya ni sa vakawati kina na yalewa. Kenai vakadei na yalewa. Io kevaka o na raica oqo mai Nukuloa, e dua na yavusa tiko e kea o iratou vata sara ga na yavusa qo nei Mitiele o Navure. O ira ya sa kau me ra laki ba ni valu. Ya na kenai vakadinadina. Oya na... mataqali o Sisiwa, tokatoka o Sisiiua. Mai Nukuloa sa toka kina oqo ko ratou ko Naitalatala vakacegu... Naitalatala sa mate, O Radini Italatala ga e sa hula tu. Sa hull me Tui Nuku ni sa lako na nona kaukaua.

CT: E na vuku ni cava sa sega ni buli tale e dua me Tui Nuku?

RSL: Vakatau sara ga vei ira na turaga ni Nukuloa. O ni sa raica tiko oqo eke na Takalaigau, sa kacivi wale tu ga e na Takalaigau ia sa sega ni se gunu, sega ni se soli vua na kaukauwa. la oya sa na qai vakatau vei iratou na Sauturaga (Raitena). Kevaka o ratou na raica ni turaga vinaka ni na lomana na vanua, sega ni bera e ratou sa na vagunuva. Kevaka ena via ca toka nona i vakarau ena via vakasuka toka me ratou na raica toka mada. Rawa in ratou na bokoca tale ka taura tale e dua tale na turaga.

CT: Tiko e dua na turaga mai Nukuloa sa taura rawa na i tutu oqo?

RSL: Io e tiko, ia na vakasama ga vakapolitiki e ra na raica ka kaya ni sega ni vinaka. Au kila ni turaga ni vavalagi talega era dau buli me vaka na Prime Minister ni veimatanitu vaka kina.

CT: Vakacava mai Navukailagi - Na Ratu?

RSL: Io, Navukailagi tale ga, ke sa dodonu nai lakolako vakavanua, e dua na turaga vinaka e tiko e dodonu me cicivaka nodratou vanua.

CT:A cava na nomuni vakasama? Eratou na bulia e dua na turaga me Takalaigau?

RSL: Io, au vakabauta. Ni sa tiko e dua na ka e ratou sa kila vata tiko na Takalaigau kei iratou na turaga ni vanua eke. Na turaga Takalaigau sa bale oti sa wavoki oti e na veikorokaro ka saga o koya ni sa buli e dua me Tui, na vanua oqo esa na dua ga na vosa. Kevaka e sega, sa na dui lako ga. Nai tutu vakavanua ni dua esa buli, sa yalataki na tamata, na qele, na co me sa kenai liua liu vakadua ko koya. Oqo e vakatokai na Sauvi, tu vakawawa. Qo o Sawaieke sa sauvi tu ga. Sa nanumi tiko, sa qui vakaraici e vinaka se sega. Na kena berabera, ia sega ni noqui tavi meu laki tukuna me caka... Qo na ka kece sa saga tiko na matanitu me vakasuaka tale mai nai vakarau vakavanua. Kemuni sa mai dede ena vanua, kila ni dua na bula e ologa ena loloma. Ni dua na turaga vinaka ena tu kece vua na ka taucoko. Raica mada qo nai vakadinadina na vinaka. E so era na tauvimate-ia sega soti ni dau vinaka na nonai tovo ki vei ira na lewe ni vanua. E ra na raica ga ni davo tu, era sega ni kauwaitaka. Dua na kena vakadinadina au sa raica oti eke.

2. Au tara saka tu mada ga na wa ni tabua, ni bula vinaka na turaga e nodatou vanua, sautu tiko noda vanua me vanua veilomani, a rogoci tu mada ga na vosa. Mana e i dina [cobo].

3. Au luvata saka na wa ni kamunaga vakaturaga, kamunaga nai vakavinavinaka, kamunaga tale ga e cabo tiko yani vei iratou na turaga Raitena, kei na veimatabure kece e soqo tiko. Keitou ciqoma me kamunaga vinaka. Dei vinaka tu na nodatou turaga, dei vinaka tu na veika kece e nodatou vanua, soli tiko vei kedatou na leqa sa tiko me rogoci keda tiko na kalou, a noda i liuliu tu mada ga ko Karisito. Mana e i dina. Taba saka mai na kamunaga. I i [cobo], A kamunaga saka e levu. Ka liu, ka liu, ka liu. E duo dosa, amuduo [cobo] muduo [cobo] amuduo [cobo] muduo [cobo], Vinaka saka.


1 Population 1988: 342,965 ethnic Fijians, of whom 31.4% lived in urban areas; 341,141 Indo-Fijians, of whom 42.4% lived urban areas (Lal 1992:217). Note that by 1988 Indo-Fijians had emigrated in large numbers and were now outnumbered by ethnic Fijians.

2 See, for example, Kaplan 1987, Kelly 1988, Norton 1990 and Lal 1992.

3 The ritual obligations peculiar to certain clans do not constitute castelike distinctions; so, for example, traditionally any fit young man might fight for his vanua, and it was not only men called the 'hand of the feast' who gardened, nor only 'fishermen' who fished; today, as in the past, all men as husbands are obliged to produce kakana dina (true food, i.e. root crops such as taro and yams) for their families, and men of chiefly clans, like all village men, garden and fish. Note that two of the statuses mentioned by Ratu Laua-those of 'priest' and 'warrior' - no longer exist as such, nevertheless, their clans continue to be recognized.

4 Observations similar to those in this paragraph have been made by numerous other ethnographers of Fiji. See, for example, Hocart (1952, 1970), Hooper (1982), Turner (1986), Walter (1978).

5 George Millier, author of the Fijian Grammar and of a forthcoming Fijian Dictionary, informs me (personal communication) that 'Matanitu is a mysterious word, especially the tu part... Turaga may itself be a "fossilised" equivalent of Samoan tulaga "status, rank, position", a noun derivative from tu ("stand")... "state" is a good non-committing gloss, but tu also has sexual connotations of dominance etc. as might be expected.'

6 Hocart lived in Fiji from 1909 to 1912 and did further fieldwork there in 1913.

7 Sahlins (1985) and (1991), following Hocart, takes a different view of Fijian dualism: in brief, he takes hierarchy to be an encompassing value, and places little emphasis on balanced reciprocity in exchange and the equality of cross-cousins. This may be because his historical essays primarily concern the period of the partial hegemony of the great chiefdoms of Ban, Rewa and Cakaudrove, whose chiefs seem almost certainly to have been concerned to establish, if they could, a vertical hierarchy. I have no fault to find with Sahlins's data, but I do disagree in part with his interpretation of them. My historical-cumethnographic analysis of the history of chiefship in the vanua of Sawaieke shows that Fijian chiefship - past or present - cannot, as 'value', encompass the pervasive antithesis between hierarchy and equality. Rather, its efficacy and its continuity require that hierarchy and equality remain in tension with one another as opposing, and equally important, aspects of social relations. The 'General History of the Country of Sawaieke' (Na i tukutuku raraba ni vanua ko Sawaieke), recorded by Lands Commission officials in 1916, covers a time-span that begins some 150 years or so earlier and ends at the period that Fiji was ceded to Britain in 1874. It includes an account of the origin of the title Takalaigau that is in many respects a classic tale of 'the stranger chief', but one whose implications are, I argue, rather different from those understood by Sahlins in respect of Lau (see Toren 1994b).

8 In respect of my argument in this chapter, George Marcus's comparative analysis of chieftainship in Polynesia demands to be addressed virtually line by line-especially, so far as I am concerned, in respect of its consideration of Fiji, where I find myself often in agreement with his analysis-agreement that is often nevertheless of the 'yes, but...' variety. To take just one example, Marcus views the Fijian polity as being more similar to the Tongan than to the Samoan, whereas (as will become increasingly plain) I would argue that, in different ways and in equal measure, it is remarkably similar to both (see Marcus 1989).

9 He was at the time a young man of 22 years old or so, who had just returned from a tour of duty with the United Nations forces in the Middle East. His views were made plain as a group of young men and women and myself were drinking yaqona - i.e. they were expressed in front of his peers. No one else present ventured to agree with him, and, according to my experience, he was highly unusual; most young men, while they may refuse 'to attend on' a chief who does not personally have their respect, hold entirely conventional views about the desirability of everything that is properly vakaturaga, chiefly.

10 See Note 5 above.

11 The growth in the eighteenth and nineteenth centuries of the large confederations of Bau, Rewa, and Cakaudrove was achieved in part through actual warfare and in part through the way that smaller chiefs sought the help of bigger ones in order to avert war (see Sayes 1984). In the contemporary polity, the chiefs of important confederations are likely to fill high offices of state such as Governor-General (when Fiji was still part of the Commonwealth) or, today, President of the Republic. It seems, however, that such titles are bound to rotate, as it were, across chiefdoms. So, for example, Ratu Sir George Gakoban of Bau (great-grandson of the famous Cakobau, called Tui Viti, King of Fiji, by Europeans) was the first Governor-General, and he was followed by Ratu Sir Penaia Ganilau (Tui Cakau, paramount of Cakaudrove, who later became the first President of the Republic). The current President is the former Prime Minister and high chief of Lau (Tui Nayau), Ratu Sir Kamisese Mara. Election of men to these offices of State is in the hands of the Council of Chiefs.

12 My initial work on Fiji is an exploration of how, over time, Fijian villagers constitute an idea of hierarchy as a fundamental principle of social organization, via their understanding of people's disposition in space, relative to one another, whenever they are gathered together to eat. to drink yaqona, to worship, to observe life-cycle ceremonies and so on (Toren 1990). For further analyses of yaqona-drinking and some of its manifold implications, see, for example, Toren 1989 and 1994b.

13 All chiefly rituals entail an initial offering of yaqona root as tribute (i sevusevu, a ceremony that requests, as it were, chiefly acknowledgement of one's presence) and the subsequent redistribution of yaqona as drink to all those who are present. Yaqona-drinking is at once the most everyday and the most sacred of Fijian rituals; it is by virtue of drinking the cup offered by the chief of the installing clan that a man is made paramount. Fijian chiefship is constituted, on a day-to-day basis, in a struggle to transform in ritual (paradigmatically, in yaqona-drinking) balanced reciprocity and equality between cross-cousins into tribute and hierarchy between people and chiefs. That this struggle is in principle unending is a product of the fact that all dynamic, fertile and affective processes are founded in the relation between cross-cousins.

14 For example, the first phase of all life-cycle ceremonies (birth, circumcision, marriage, death) lasts for 'four nights', and a herbal remedy prescribed by a healer has usually to be followed for 'four nights'.

15 The term for chiefs here is na malo, 'the cloths'; it refers not just to men who are, by birth, 'chiefly', but to the recognized leaders of all the leading clans of a vanua (country), including the paramount.

16 A situation that recalls Lévi-Strauss's (1977, 1978) brilliant observations concerning the possible array of complementary oppositions that is inevitably entailed in 'the atom of kinship'.

17 Toren (1998) analyses pre-colonial sacrificial and consumptive practices in Fiji to reveal their implications for ideas of the person, together with historical transformations in these ideas as a function of the impact of conversion to Christianity. The antithesis between hierarchy and competitive equality pervades production, exchange and even tribute. But consumption is at once the end and the beginning of these processes, and in pre-colonial Fiji was the privileged site of existential agency.

18 The malign power of the ancestor Gods may be unleashed through witchcraft; see Toren 1995.

19 Steven Hooper prefaces his extract from this article with the observation that Ratu Mara 'was clearly the heir apparent and was immediately recognised as Paramount Chief on the death of his father. However, Ratu Mara himself was well aware that sucession was a matter of selection and not of right' (Hooper 1982: 157). Note that Hooper also records that neither Ratu Mara's father, nor his grandfather, 'had been formally installed in the proper manner as Sau, though they were accorded the respect due to the holder of that office'. Toren (1994b) includes a discussion of why it is that a putative Sau is often named and never installed.

20 The course of the relationships between Colonel Rabuka, who of his own accord instigated the coup, Ratu Sir Kamisese Mara, the Prime Minister whose government had been ousted by Dr Bavadra's Coalition, and Ratu Sir Penaia Ganilau, at that time Governor-General, can also be understood to accord with Fijian ideas concerning the responsibilities of warriors and chiefs (see Mara 1997, Milner 1997, and, for an alternative view, Lal 1992).


Christina Toren has a B.Sc. (Hons) in Psychology from University College London and a Ph.D. in Social Anthropology from London School of Economics. She is currently Senior Lecturer in Anthropology and Psychology at Brunei University. She is the author of numerous papers and two books: Making Sense of Hierarchy: Cognition as Social Process in Fiji (London, 1990) and Mind, Materiality and History: Explorations in Fijian Ethnography (London, 1999).

© Etnográfica Press, 2000

Conditions d’utilisation :