Three of these are:
- Nandan, Satendra, 1993. 'Postcolonial fictions? Other texts, other lives—an Australian-Fijian experience', Span: Journal of the South Pacific Association for Commonwealth Literature and Language Studies 36. http://kali.murdoch.edu.au/~cntinuum/litserv/SPAN/36/Nandan.html (dead link)
- Ogden, Michael R., n.d. 'Republic of Fiji', in World Encyclopedia of Political Systems, 3rd ed. New York: Facts on File. http://www2.hawaii.edu/~ogden/piir/pacific/fiji.html
- Prasad, Satendra, 1996. 'Limits and possibilities for civil society led re-democratization: the Fijian constitutional debates and dilemma'. London: Conciliation Resources. http://www.c-r.org/occ_papers/occ_fiji.htm (dead link)
 Members of the army were (and are) predominantly indigenous Fijians. Less than 3 per cent of the RFMF were Indians at the time of the coup (West 1988: 223).
 Ratu is the Fijian chiefly title for men of rank; its equivalent for women is Adi.
 A particular observer's explanation will not necessarily fall solely under one factor heading. The predominance of a particular factor within an explanation may determine under which heading it best belongs; but where necessary I have discussed different aspects of an explanation under different factor headings.
 Taukei is the term for a member of a village land-owning clan, but in pre-coup usage had come to mean 'indigenous Fijian'. Since the coup it has gained various political connotations as a result of its use by the Taukei Movement.
 Further discussion about the nature of pre-coup 'democracy' in Fiji, and some of its failings, can be found in Lawson (1988b: 35-47). (Lawson in no way endorses the idea that Fiji democracy's failings somehow justified its overthrow.)
 Notable exceptions were David Robie, a freelance journalist in the Pacific, and Marian Wilkinson, of the ABC's 'Four Corners' programme.
 A likely interpretation of this apparent paradox is that Rabuka was concerned about the damaging consequences of this unrest not for the Coalition, but only for anti-Coalition Fijians. In this case, he would have seen himself as enacting the Taukei Movement's wishes, but in a more efficient manner. This rationale would be consistent with such statements as 'I see how these events could lead to serious situations and threaten law and order and property' (Rabuka, quoted in Robie 1989: 221).
 Dean is an Australian journalist and Ritova a Fijian photo-journalist.
 Stephanie Lawson lectures in Politics at the University of New England. In this paper I have referred to her many articles and papers on the subject, which are noted individually. Her book on the coup, The Failure of Democratic Politics in Fiji, has since been published (Oxford: Clarendon Press, 1991).
 Robert Robertson is an historian at the University of the South Pacific, and Fijian Akosita Tamanisau is a journalist who worked for the Fiji Sun before the coup.
 First, 'tradition' embraces more aspects of a society's culture than its politics, and secondly, 'traditional', in the context of the Pacific, is often taken (rightly or wrongly) to mean 'pre-European contact'. The labels 'custom' and 'customary' carry fewer such connotations, and allow for the evolution of 'traditional' structures through the colonial and post-colonial periods. Compare, for example, these definitions drawn from The Macquarie Dictionary (1981):
tradition, n. 1. the handing down of statements, beliefs, legends, customs, etc., from generation to generation, esp. by word of mouth or by practice.
custom, n. 3. a long-continued habit which is so established that it has the force of law. 4. such habits collectively.
customary, adj. 2. of or established by custom rather than law. 3. Law. defined by long-continued practices.
It must be acknowledged that the distinction between these terms is a fine one (and where others have used the term 'traditional' I have not taken issue with them). The quasi-legal dimension to the term 'custom', however, makes it useful in a political discussion. I also wish to avoid the 'pre-contact' connotations of 'traditional' as far as is possible. Fijians themselves, few of whom have a detailed knowledge of Fiji's pre-colonial history (which regrettably is not taught in most schools), are apt to consider the village practices with which they have grown up as being 'traditional', even though these practices may in fact have developed within the last one or two generations.
 Senior lecturer in Anthropology and Comparative Sociology in the School of Behavioural Science at Macquarie University.
 Primarily, these are Eddie Dean and Stan Ritova, Brij Lal, Stephanie Lawson, Robert Norton, Robert Robertson and Akosita Tamanisau, David Robie, and Deryck Scarr. But also discussed to some extent are the views of Kenneth Bain, Michael Danby, Don Dunstan, Christopher Harder, Ralph Premdas, Theo Roy, William Tagupa, Nicholas Thomas, Anthony van Fossen, and Owen Wilkes.
 Fiji's economy has largely recovered (thanks to tax-free zones which act as incentives for overseas investors), although it is doubtful whether ordinary Fiji citizens are seeing the benefits of this recovery. The Fiji dollar has been devalued more than 30 per cent since the coup, and market prices have inflated by over 30 per cent in the same period. 'The government's economic revival', says reporter Ron Harel, 'is a mirage. Most profits are going offshore' (Ron Harel, 'Dateline: Power Play in Paradise', SBS Television 10 November 1990).
 Percentages from North (1987:22). Total from 'Muslims in Crossfire', Pacific Islands Monthly September 1987, p.24. These commonly-quoted figures are not census results; rather, they were drawn from an interim population survey conducted in 1986.
 Specifically, by Rabuka himself. See Dean and Ritova (1988:126).
 When the value of this land is examined, however, a different picture emerges. Less than 10 per cent of Fijians' land is cultivable, and most of the highest-value prime commercial freehold land is owned by Europeans (Premdas 1990:12).
 Lal (1988a:49) notes that by 1986 38.9 per cent of Fijians lived in urban areas.
 The explanations he notes are intervention by the CIA, the manipulation of Rabuka by powerful individuals, and a 'boundless ambition' on Rabuka's part.
 It is ironical that one of those leaders, Apisai Tora, had, until 1982 when he joined the Alliance, been a member of parliament for the predominantly-Indian National Federation Party (Pacific Islands Monthly November 1987, p.15. See also Norton 1977:129).
 The '6 per cent of total voting strength' point stressed by Scarr overlooks this. Scarr implies that the 30 per cent of Fijians who failed to vote would, if they had voted, have voted for the Alliance in similar proportions to those who turned out. This cannot be asserted with certainty given their decision to abstain from voting.
 Kenneth Bain has served as an administrator in colonial Fiji, and as a director at the Commonwealth Secretariat. His book, Treason at Ten (1989) is a personalized account of the coup.
 Under the 1970 Constitution, there were 52 seats in the (lower) House of Representatives. Twenty-seven were communal, filled by candidates of the appropriate race who were elected by voters of the same race. Of these seats, 12 were Fijian, 12 Indian, and 3 General (i.e. other races). The remaining 25 seats were 'national'. Of these, 10 were Fijian, 10 Indian, and 5 General, and they were filled by candidates of those races. Voting for national seats, however, was based on a common roll. Therefore, if the Indian vote for a national seat was outnumbered by the Fijian vote (with General electors usually voting with Fijians), the candidate (of whatever race) favoured by Fijians would win. (The Alliance won many national seats this way; its Indian national MHRs were often elected by Fijian votes.) Under a first-past-the-post system, though, a split in one community's vote had a similar effect. Thus, a split in the Fijian vote led to the NFP election win in 1977 and the Coalition win in 1987 (although a precise explanation is somewhat more complicated). (This note on Fiji's electoral system is based on North 1987:22.)
 The government could have gradually changed the composition of the Supreme Court in its favour, but would have done so at the risk of encountering considerable public hostility.
 It should not automatically be assumed that because Kahan is an Indian, his sympathies lie with the Coalition. His first name, Mohammed, indicates that he is a Muslim. Fiji's Muslim community has not distanced itself from Rabuka's regime as completely as other Indians have. The Fiji Muslim League took advantage of the post-coup constitutional review process to lobby for separate representation for Muslims, a long-sought goal (although not achieved in this instance).
 Harder was one of several lawyers who helped prevent Kahan's extradition from London to Fiji. His book, The Guns of Lautoka (1988) is partly an autobiography and partly a description of his personal experiences in post-coup Fiji.
 These allegations are mainly drawn from Kahan's own written account, which is reproduced in full in Harder (1988:214-17). A summary of this account (with sics omitted for the sake of brevity) follows:
Mara's initial deal with Colonel Rabuka was that Rabuka hand back power within 30 days from the bloodless coup to Mara controlled interim civilian Government. ... [But Rabuka] started to enjoy the fame and feel of real power. ... He demanded active participation in any new government. ... Rabuka's nagging fear was that ... he could be dispensed of if he gave up real power. ...
The question that worried Mara was that the longer Colonel Rabuka stays in effective control of power and learns from Mara, Ratu Mara will become dispensible. And it would be harder for Mara to get rid of Rabuka. ... We heard that Mara was a very worried man [and] put out feelers and floated an insurance policy plan for Mara. ... He signalled that he can help divide the Army to weaken Rabuka but he needed Indian assistance to obtain arms for this divided Army. ...
Rabuka suspected Mara and he strengthened his secret alliances. ... Taukei started to openly criticise Mara and demanded him removed on the grounds of corruption, etc. The stage was being set for Rabuka's third coup which was planned for April 1988. ... We quickly signalled to Mara. ... Now [Rabuka] was sure that his inner circle had holes. ...
The question is, of course, how much credence does one give to Kahan's words? But then, why believe Rabuka's or Mara's version of events, if both may have a vested interest, for different reasons, in keeping the true story quiet? At the very least, Kahan's story demonstrates that the explanation for the arms shipments could quite possibly be more complicated than Scarr allows.
 An indication of how far Fijian attitudes had come is given by a survey undertaken by A. C. Cato in the mid 1950s, when the Indian population, as a proportion of all people in Fiji, was at its peak. In many discussions with people of both races, Cato found that 63 per cent of Fijians expressed complete rejection of Indians and desired that none should remain in Fiji, and a further 25 per cent were opposed to Indians but willing to let them remain under conditions clearly favourable to Fijians. Furthermore, in 64 conversations with Fijians concerning the possibility of Fijians sharing power with Indians in an independent Fiji, 54 ruled it out completely (Cato 1955:17-19). There is no doubt that Fijians were far more tolerant of Indians by the 1980s, and it seems quite reasonable to expect this trend to have continued had the coup not occurred.
 The terms of the coalition agreement were largely dictated by Labour (Robertson and Tamanisau 1989:212).
 This deficiency has been addressed in more depth by William Sutherland, in his review of Power and Prejudice in The Journal of Commonwealth and Comparative Politics 28 (March 1990): 133.
 Mention should be made here of a new work available at the time of going to press: William Sutherland's Beyond the Politics of Race: An Alternative History of Fiji to 1992 (1992). Sutherland, to a degree not previously seen in similar projects, examines Fiji's history in class terms, making important and original points in the process. His class analysis of Fiji is more clearly argued and presented than those in the older works discussed in this chapter. As an explanation of the coup, however, (admittedly not the book's focus), his analysis contains some of the weaknesses pointed out here. Sutherland, in seeking to draw out the continuities in Fiji's class history, pays the events of May 1987 themselves relatively little attention. The resulting impression given is that the coup was largely a result of economic forces; and, as is the case to varying degrees in other works mentioned here, the traditional/chiefly aspects of Fiji's ruling class are less completely explained than the economic elements. Overall, however, Sutherland's book has much to recommend it, and his detailed chapter on post-coup developments (to May 1992) gives an insightful commentary on often confusing events.
 The historical information which follows is based largely on a reading of two sources: Routledge (1985) and Campbell (1989).
 The colonial administration did, however, attempt to minimize the effect on Fijian land ownership of deals between chiefs and unscrupulous white settlers. A Lands Commission was set up which reviewed all pre-Cession claims and returned about half of the area claimed by Europeans to Fijian ownership (Lawson 1987c:6).
 Routledge describes the extent of this intermarriage in today's Fiji. Mara is 'heir to the Fijian paramount line in the Lau Islands and the Tongan line established by Ma'afu'; both he and his wife, Adi Lady Lala, are 'direct descendants of the first Cakobau, as are also Ratu Sir George Cakobau, present Vunivalu of Bau and former Governor-General, and Ratu Sir Penaia Ganilau,... present Governor-General' (Routledge 1985: 220).
 Once the eventual independence of Fiji was seen as inevitable (given the world's changing political climate), the colonial government carefully groomed a number of chiefs to assume the mantle of power. Even the political system adopted, however, may not have totally satisfied the eastern chiefly elite. Lal has observed: 'when Ratu George Cakobau asked that Fiji be returned to Fijians, he meant, without a shadow of a doubt, Fijian chiefs. In fact, he himself along with Ratu Sir Kamisese Mara, E. Vito Qiolevu and others reportedly petitioned the Queen in 1969 to do precisely that' (Lal 1988a:64).
 Lawson has noted that the paradoxes of Fiji's land system go back to the earliest days of the colony: 'So complex and various were existing land tenure practices that it was not until 1918 that the Native Lands Commission finally agreed to the designation of the mataqali as the unit through which ownership would be vested' (Lawson 1988a:7). By the 1940s, 'the practical universalization of the mataqali had been enforced so effectively by the administration that younger Fijians had inadequate concepts of their own genuinely traditional divisions' (ibid.: 8).
 The concentration of economic activity in the west (and the resulting need for access by foreign companies to land) may also have served to aggravate western commoners' grievances against their chiefs, as the following 1978 quotation from a Fijian sociologist demonstrates:
In the Western division we see many partnerships between leading chiefs and overseas companies. This is primarily because the chiefs aren't allowed to sell land, so the easy way for an overseas company to acquire land is through partnership with a chief, who usually becomes richer as a result. (Bolabola 1978:156).
Thus, commoner grievances, which as noted earlier are seen throughout Fiji, are probably particularly evident in the west.
 Sakeasi Butadroka is, on first examination, a perplexing character. He has been widely quoted as saying, on the day of the coup, 'Where is Kamisese Mara? Don't blame Bavadra, don't blame anybody, blame Kamisese Mara who sold Fiji. ... Mara, the bloody Judas Iscariot!' (Robertson and Tamanisau 1988: 70). This was interpreted by some journalists at the time as evidence of Fijian outrage at Rabuka's action. Yet Butadroka came out in support of the coup, telling his party in July that 'All that I've been fighting for in the past seventeen years Rabuka has now won for Fijians' (Scarr 1988a:90). He was also, as previously mentioned, the one who moved in Parliament in 1977 that all Indians be expelled from Fiji. Butadroka has himself said, however, that 'telling Indians to get out of Fiji was just a political tactic to stir Fijian feelings and so force Ratu Mara to do something about the poverty in Fijian settlements here' (Robertson and Tamanisau 1988:47). He is consistently pro-Fijian commoners in his behaviour, and quite wary of chiefs. He has criticized post-coup developments which strengthen chiefly power for that reason, his FNP going so far as to call for the abolition of the chiefly system altogether (Lawson 1990:16). His outburst on 14 May 1987 has been amply explained by Scarr: Butadroka 'was attacking his aversion, Ratu Sir Kamisese Mara, for creating the 1970 Constitution, which made the coup necessary, in Butadroka's terms, by failing to hand the country back to the Fijians' (Scarr 1988a:90).
 This remark is especially perplexing when one notes that it was made by an Australian: Australia's states have very few linguistic and cultural differences, yet they vigorously pursue (and perpetuate) their separate interests.
 Here it is particularly relevant to recall the earlier-noted degree of intermarriage between chiefly families throughout Fiji (particularly post-Cession), which resembles that of the royal families of nineteenth century Europe. A western chief may in fact be more closely related to eastern chiefs than to his own subjects.
 The main reason for this move seems to be an agreement between the three existing confederacies that the presidency of the Republic will circulate between their paramount chiefs every five years. Among western chiefs there has been 'a strong feeling that the proposal would further marginalize the western traditional leadership' (Thomas 1990:137). Although the move for a fourth confederacy is a chiefly one, the Coalition has expressed support for the concept, as it would represent a more equitable distribution of power than exists at present (Sharma 1989:18).
 The west is at present split between Kubuna and Burebasaga. Tovata is a wholly eastern confederacy. (See map for approximate borders.) The east/west divide does not, however, represent a divide between Tovata on the east and Kubuna and Burebasaga on the west: the latter confederacies are both politically dominated by the east coast of Viti Levu, and their paramount chiefs are easterners. Hence the move by some western chiefs for a wholly western confederacy.
 To again use the Australian states analogy, the parochial Tasmanian 'north-south divide' does not prevent a common interest from prevailing when Tasmania is dealing with the Commonwealth government or competing with mainland states.
 The village was Viseisei, 15 kilometres from Nadi airport on the west coast of Viti Levu (Bain 1989:115).
 Titles such as 'Tui' and 'Vunivalu' (which translates as 'warlord') are reserved for the highest chiefs in the hierarchical chiefly system. For example, the Tui Cakau, Ratu Ganilau, and the Tui Nayau, Ratu Mara, are both paramount chiefs of their regions.
 Ganilau once asserted that the Vunivalu of Bau would be obeyed everywhere in Fiji without question (Keith-Reid 1982:16).
 This was reported by Taniela Veitata, who did not identify the chief concerned (cited, ibid.: 139). Veitata was the Alliance MP who appeared in the description of the coup given in the Introduction.
 That Scarr presents information useful in a custom framework, and then interprets it in a racial manner, must be attributed to a belief in what Lawson calls the 'Myth of Cultural Homogeneity', which is that all Fijians have identical interests. My discussion of this myth has been implicit throughout sections 2-4.
 The Council was and is the highest of the Fijian community's deliberative bodies. But even it is not 'traditional' but 'customary': it had not existed before Cession, but 'apparently arose almost by accident when, at Gordon's installation as "supreme chief" of Fiji, he had used the occasion to consult with the Rokos (chiefs appointed as governors) who had assembled there from all over Fiji' (Lawson, 1987c:3).
 During the early stages of the review process, the Constitutional Review Committee, 'weighted in favour of the Taukei Movement, ... did not bother to breakdown the [800 written and 161 verbal submissions] or even to take note of majority support for the old constitution among Fijian submissions. Instead it focused entirely on the last submission, that from the chiefs' (Robertson and Tamanisau 1988:132). Its report of 21 August 1987 endorsed all of the chiefs' proposals except for the 'consensus' system of election (ibid.). Since Rabuka's second intervention, 'discussions regarding constitutional changes have taken place under a non-elected government without public consultation' (Prasad 1988:114).
 Former South Australian premier Don Dunstan was born in Fiji and is chairperson of the Fiji Independent News Service.
 The Great Council of Chiefs, prior to the coup, consisted of all 22 Fijian members of the (lower) House of Representatives, 15 members (8 chiefs and 7 others) appointed by the Minister for Fijian Affairs, and 2 or 3 representatives chosen by each of Fiji's 14 provincial councils. (These councils are Fijian bodies. The 14 provinces were set up by the colonial administration last century, and range widely in Fijian population from 4,462 in Namosi to 55,343 in Ba. The smaller councils are staunch supporters of the post-coup regime.) Since the coup the MHRs have been excluded from the Council of Chiefs. The Council determined on 9 June 1990 that in future it will consist of 42 members chosen by the provincial councils (i.e., 3 each) and a dozen drawn from and nominated by the government (Dunstan 1990).
 The remaining senator is chosen by the Council of Rotuma. Rotuma is a tiny Polynesian dependency of Fiji.
 The pre-coup electoral system is discussed in section 2 of this paper.
 This is an important point which has not featured to date in many analyses of the 1990 Constitution. Section 41(2) of Government of the Republic of Fiji, Constitution of the Sovereign Democratic Republic of Fiji (1990), states that:
For the purpose of electing the members of the House, voters shall be registered on one of four separate rolls, that is to say -
(a) a roll of voters who are Fijians;
(b) a roll of voters who are Indians;
(c) a roll of voters who are Rotumans; and
(d) a roll of voters who are neither Fijians, Indians nor Rotumans.
Seats are assigned to each roll as previously noted. An important distinction between the 1970 and 1990 constitutions is that 'Fijians' are no longer defined as all Fijians, but rather, as all Fijians 'registered or eligible to be registered in the Vola ni Kawa Bula' (Section 156[a]). Those Fijians remaining will therefore go onto the fourth roll (see also Pacific Islands Monthly August 1990, p.13).
 See an earlier footnote for an explanation of why Scarr presents this custom information in a racial framework.
 The leasing system has produced a similar result. Van Fossen notes that 'the Fijian Administration leases land to Indians only for short periods, on the understanding that it will eventually revert to supporting the traditional [chiefly] conception of rural life'. But that same policy is destroying Fijians' land: 'Since Indians ... have little incentive to develop the land under these conditions, there is severe erosion and poor conservation' (Van Fossen 1987:21).
 Owen Wilkes is a member of the Peace Movement Aotearoa. His article, 'U.S. Involvement in the Fiji Coup', appeared in Peacelink and Wellington Pacific Report.
 This suspension was not, however, complete: the American Peace Corps programme continued, as did funding of some other projects (Howard 1990:18).
 Premdas's discussion of internal factors is probably best categorized as racial. While Premdas does point to the likely role played by important Alliance figures, and does demonstrate that many Fijian fears were unfounded, he seems largely to accept the 'myth of cultural homogeneity', and stresses the clash of Fijians and Indians somewhat more than class or custom. He discusses many conflicting explanations, however, and is not as readily pigeonholed as Scarr, Roy, or West.
 Lieutenant-Colonel Jim Sanday was second-in-command of the RFMF at the time of the coup. He now resides in Australia.
 Fijian soldiers stationed on UN peace-keeping duty in the Middle East seem to have been less strongly socialized in Alliance and chiefly values; over half of them voted for the Coalition in the election (Dean and Ritova 1988:34).
 Robie, however, relays 'persistent stories' that several officers opposed the coup, and says 'some soldiers were reportedly detained because of their refusal to follow Rabuka's leadership' (1989:229).
 See chapter 4 for a discussion of the second intervention, and Ganilau's and Mara's roles.
 Cameron's letter (not actually published by The Bulletin until after the second intervention) is reproduced in Robie (1989:243-44).
 The opinion of New Zealand's ex prime minister is reported in Robertson and Tamanisau (1988:94).
 Allegedly corrupt Alliance members included Apisai Tora (see Saffu, 1989:16) and Peter Stinson (Jim Anthony, speaking on Judy Fasher, 'Tuesday Despatch: The Fiji Military Coup', ABC Radio, 19 May 1987, transcript, Parliamentary Library, Parliament of the Commonwealth of Australia, p.10). Tora was an ex-minister, a leader of the Taukei Movement, and later minister in the Republic. Stinson, also an ex-minister, stood prominently next to Rabuka at the first post-coup press conference (Alley 1987:491), and joined Rabuka's 15 May Council of Ministers (Robertson and Tamanisau 1988:71).
 The correct spelling is 'Marela'.
 It is also interesting to note that the army's corruption inquiry preceded a Taukei Movement campaign for the removal of Mara on corruption grounds. Rabuka allegedly instigated this campaign to put pressure on Mara as part of a power struggle between the two. (See Mohammed Kahan's allegations in section 2, this paper.)
 For a discussion of distinctive aspects of Pacific political systems, see Fry (1983).
 A starting point in this literature is Hobsbawm and Ranger (1983). Other key discussions are Babadzan (1988); Keesing (1989); Linnekin and Poyer (1990); Jolly (1992); Thomas (forthcoming).
 One reason why I have used the term 'custom' through most of this paper is to avoid the ambiguity surrounding the term 'tradition' (see section 1).
 Such systems do leave parts of society under- or un-represented; but viewed from a Lockean 'social contract' perspective, they may have validity. The society may be considered to have developed as if, at some stage, its people had unanimously agreed to adopt the system favoured by the majority, that majority then choosing a custom-style system. This view does overlook the phenomenon of such systems being forcibly adapted by, or incorporated into liberal democratic systems by, colonial powers. But even in these cases, a demonstration by a society of support for its new hybrid system (for example, through a referendum at independence) can, from a Lockean perspective, bestow validity upon it.
 Despite the indentured labour system, they did choose to stay at the end of their indenture.
Page created 3 September 1998