Lessons in nonviolence from the Fiji coups

Gandhi Marg, No. 114 (Vol. 10, No. 6), September 1988, pp. 326-339. An abbreviated, less referenced version of this article appeared as chapter 5 in Social Defence, Social Change.

Brian Martin

Go to

Brian Martin's publications on nonviolence

Brian Martin's publications

Brian Martin's website

On 14 May 1987, the Fiji government was ousted by a military coup led by Lieutenant Colonel Sitiveni Rabuka. The response to Rabuka's regime both within Fiji and overseas provides a useful test of the theory and practice of nonviolent action.

Fiji was taken over by the British as a colony in the 1870s.[1] The native peoples are ethnically Melanesian. The British brought indentured servants from India to work on the sugar plantations. Today the so-called Indo-Fijians - born and bred in Fiji with ancestors from India - make up half the population of 700,000. Melanesians make up 45% and Europeans, part-Europeans and others the remainder.[2]

The Europeans in Fiji have long served their own interests by aligning themselves with the chiefs or aristocracy of the Melanesian Fijians. Fiji gained independence in 1970 under a constitution and electoral system designed around racial divisions. Melanesian Fijians were guaranteed ownership of most of the land, while members of parliament were selected in a complicated fashion in which each voter had four votes, for candidates of different ethnic backgrounds.

From independence until 1987, the Alliance Party held power under Prime Minister Ratu Sir Kamisese Mara. The Alliance was built around and supported by Melanesian Fijians. The opposition National Federation Party (NFP), which was built around and supported by Indo-Fijians, was riven by splits. The 'general electors', associated with the small European population, supported the Alliance. In effect, ethnic divisions were exploited by the chiefs, using the vehicle of the Alliance Party, to mobilise support for a feudal-style hierarchy which put them in a privileged position.

In 1985 the multi-racial Fiji Labour Party was formed.[3] It was an attempt to promote class-based rather than race-based politics. The Labour Party criticised both other parties for serving the rich, and promoted the claims of workers, the unemployed and the poor.

The Labour Party rapidly gained strength and several NFP politicians defected to its ranks. In the 1987 election, the Labour Party joined with the NFP as a coalition and together they won control of parliament. It was this government that only six weeks later was toppled by a military coup.

Any military coup raises a range of questions. For example, who was behind it? What social structural developments opened the possibility of it happening? What could have been done to forestall it or oppose it? Whose interests does it serve?

Here, my concern is with the potential for opposing coups by nonviolent action. I will begin by outlining the possible actions that can be taken against coups, especially by people in other countries. Then I will compare this with the actions actually taken in relation to Fiji. The result will be some lessons for future action.

The events in Fiji are complex. They have included apparent moves after 14 May 1987 toward civilian rule, a second coup on 25 September 1987, a repeat pattern of civilianisation - including introduction of a military-backed civilian government headed by Mara in December 1987 - and the internal security decree of 16 June 1988 which established martial law. No attempt is made here to examine the politics of these and related events.[4] I will refer mainly to the first coup and, in regard to overseas responses, refer mainly to responses in Australia.

Social defence

Nonviolent action is a form of political action which encompasses methods such as demonstrations, boycotts, strikes, sit-ins, parallel government and a host of other techniques.[5] Social defence can be defined as the use of nonviolent action to serve as an alternative to military defence.[6] As a proposed alternative to military force, social defence can be used in two basic ways. First is as defence against a foreign invasion. Here the best example is the spontaneous Czechoslovak resistance to the Soviet invasion in 1968. Another example is the German resistance in 1923 to the French occupation of the Ruhr.

The second focus of social defence is against military coups.[7] An often-cited example is the Kapp Putsch in Germany in 1920, which collapsed after four days in the face of widespread opposition, including strikes, demonstrations and refusals to obey by government officials. In 1961, a revolt by French generals in Algeria, which threatened to lead to an invasion of France, was met by mass opposition in France and by noncooperation of soldiers in Algeria. Also relevant here is the continuing resistance to military rule in Poland by Solidarity.[8]

Nonviolent action can also be used as a strategy for national liberation and social emancipation. Here the best examples are the Gandhian movement in India and the Sarvodaya movement in Sri Lanka.[9] This use of nonviolent action is related to but conceptually distinct from social defence as an alternative to military defence.

These historical examples are suggestive but far from definitive. In fact, no country has made social defence a significant supplement to military defence, much less begun to convert entirely to social defence. All the historical examples involve relatively spontaneous responses to military threats. Proponents of social defence argue that much preparation and planning is necessary to develop a viable nonviolent defence system. Just as military methods cannot be expected to be very successful without money and resources, the successes of nonviolent methods will depend on luck until they receive similar social and financial support.

In Fiji, the case for social defence seems overwhelming. There is no obvious foreign military threat. Furthermore, the Fiji military forces number only a few thousands, a large fraction of which have been stationed in the Middle East as part of the United Nations peacekeeping force. Any moderate-sized force invading Fiji would receive little resistance. As in many other countries, the major military danger to the Fiji people is from their own military, as events have clearly shown. A social defence system would not pose this danger, and so whatever its weaknesses would certainly provide more 'security'.

The coups in Fiji have been almost entirely 'bloodless'. There was no organised violent resistance. This probably explains why there has been relatively little violence by the Fiji military itself, at least compared to many of the military regimes in South America, Africa, Asia and Europe. Violent resistance tends to legitimise violence by the military as well as to unify it, while nonviolent methods tend to reduce violence by the other side. This at least is the claim by proponents of nonviolent methods, and it seems to be borne out in Fiji.

Nonviolent resistance within Fiji has taken a variety of forms.[10] At the most basic level, numerous people have spoken out against Rabuka's regime, criticising its illegality and violations of human rights. Members of the Labour Party have tried to build grassroots support, travelling to villages and explaining how the 1970 constitution guarantees the rights of Melanesian Fijians.[11] There have been demonstrations and strikes in cities, and many shopkeepers have closed their shops in protest. Even more powerfully, workers in the cane fields stopped work; the threat of failure of the sugar crop, Fiji's major export earner, was a serious one. Of long-term significance, many Fijians have emigrated to escape the repressive political scene, and those leaving are mostly the educated and highly skilled.

The resistance to the Fiji military regime has been explicitly and consistently nonviolent.[12] It is telling that the regime claimed that illicit arms shipments to Fiji, which were revealed by Australian Customs, were destined for coalition members, thereby trying to discredit them as planning violence. (It is said in Suva that a story set in print for the Fiji Times of 22 June 1988, claiming that the arms were imported by the Alliance Party, was censored by the army and all copies of the printed papers destroyed.)

The resistance in Fiji can be analysed readily in terms of the standard concepts of the 'politics of nonviolent action'. But these concepts do not provide a sufficient analysis of one vital part of the struggle: the struggle for allegiance at the level of ideas and cultural beliefs.

At first sight, this criticism of nonviolent action theory seems strange, since the whole theory is based on the struggle for allegiance. Nonviolent action includes an array of methods of direct communication and persuasion, all of which are designed to win over opponents or the uncommitted. Furthermore, one of the great advantages of the use of nonviolent rather than violent methods is that violence can alienate potential supporters whereas violence by the other side can generate ever greater support. This account is fine as far as it goes. What it does not encompass, or includes only with difficulty, is aspects of the struggle for loyalty which involve aspects of culture and politics requiring an analysis of structures and belief systems.

The coups in Fiji succeeded with a minimum of force. There were relatively few soldiers involved. If there had been a concerted nonviolent resistance from the outset, it seems a good possibility that the initial coup could have been thwarted. But the reality was quite different from this hypothetical resistance. A large number of Melanesian Fijians supported the first coup while the Indo-Fijians failed to put up a show of support for the government. The mass rallies during the election campaign in support of the Labour Party failed to materialise in opposition to the coup.[13]

The initial coup succeeded because it drew upon the ethnic divisions in Fiji, mobilising Melanesian Fijians and demoralising Indo-Fijians. The use of ethnic divisions for political purposes has a long history in Fiji. The Labour Party itself represented a challenge to this political use of ethnicity, and the coup represented a reversion to this status quo.

Also involved in the early support for and acquiescence to the coup was the lack of vehement opposition by figures of powerful symbolic importance. Mara, whose party had lost the election, did not exert his influence and reputation to oppose a coup when it was being sounded out just after the election,[14] nor after it occurred. In the following weeks he appeared to serve Rabuka's purposes by being involved in the constitutional commission and the civilian governments that followed Rabuka. The Governor-General, Ratu Sir Penaia Ganilau, played a similar ambiguous role. Other members of the council of chiefs also offered little resistance to the coup. All this made it appear to many that the formal justifications for the coup - that the rights of Melanesian Fijians were threatened by the coalition government - had legitimacy.

This sort of apparently ambiguous situation is ideally designed for nonviolent action, at least in the sense that methods of persuasion, noncooperation and intervention can be used in a straightforward manner; violent attacks on nonviolent activists then would highlight the injustice of the regime. The dilemma was not the implementation of nonviolent action, but rather the difficulty of mobilising people to take the action. Without strong support from key symbolic figures, in the face of longstanding ethnic and other divisions, and lacking leadership, preparation and training in nonviolent action and strategy, a unified response was not made. This negative assessment should not obscure the considerable and powerful resistance that did occur. The point here is that nonviolent action theory devotes much more attention to the consequences of actions that are actually taken than to the structural and ideological obstacles to taking action in the first place.

Nonviolent resistance outside Fiji

Another neglected area in the theory of nonviolent action is the role of groups outside what seems to be the immediate arena of struggle. In the case of Fiji, a large number of people in other countries were startled and disturbed by the coups. Outside Fiji, the stated reasons for the coups sounded hollow, and the ethnic divisions which helped sustain the new regime had little salience. What could people overseas do to support democracy in Fiji? Here I will first outline a range of actions which might be taken by individuals and non-government groups, and then point to the ones which actually were taken up.

For a person in another country, it may at first glance seem difficult to intervene in events far away, but actually there are numerous ways to have an effect. I have already discussed the vital importance for coup leaders to appear to be legitimate. This can be challenged by people openly criticising the new regime and demanding a return to the elected government. Given that Fiji newspapers, radio and television were immediately censored after the coup, the best available outlet for protesters in other countries is their own local media. Letters to newspapers, articles in magazines, programmes on radio, protest meetings and rallies all are effective in making more people aware of the situation. They also have an indirect effect within Fiji by affecting the opinion of people around the world and inhibiting the acceptance of the new regime by other governments.

Another way to support the resistance is to make direct contact. This can include letters to individuals, as long as censorship permits. (Censorship seems not to be have been too extensive after the Fiji coups.) Messages can be passed by visitors, whether tourists to Fiji or Fijians travelling overseas. There is also much routine communication for the purposes of commerce, navigation and weather analysis which could be used for passing political information. For example, computer communication carried out by banks or airlines could be used to transmit information. This could easily be hidden from casual observation by simple coding or putting it in channels designed for engineering checks.

Many other groups make contact between countries, such as diplomats, sporting teams and church officials. These contacts can be used to pass information and advice.

Even more direct is short-wave radio, which provides person-to-person communication over long distances. Because of its geographical dispersion, Fiji has a large number of short-wave receivers which could have been used for obtaining reliable information about the events. Significantly, the Rabuka regime tried to get people to turn in their short-wave sets.

Economic pressure is another potent tool, especially in the case of a small country like Fiji. Trade union bans on shipments to or from the country are one method. Another approach is the consumer boycott. In the case of Fiji, the major 'good' that can be easily boycotted is the tourist trade, since tourism is Fiji's second largest export earner. Refusing to go to Fiji hurts the economy; writing a letter to a newspaper stating that one is refusing to go, and is taking one's tourist trade to more democratic countries, adds symbolic impact to this stand, and is effective even if one had not been planning a trip to Fiji.

Another approach is to provide direct support for nonviolent action within Fiji by offering advice and training. This could be done for Fijians travelling overseas, or done in Fiji by activists ostensibly entering as tourists. If a sufficient fraction of visitors to Fiji were actually nonviolent activists, the regime would be caught in a bind. Allowing the visitors to move unhindered would allow activists to build strength for the opposition, whereas security measures to monitor and arrest suspicious visitors would risk alienating genuine tourists and thus hurting the economy.

Finally, people overseas can provide refuge to refugees from the regime. Fleeing the regime does not by itself undermine its strength, but many refugees are able to become vocal once they are free from repression inside their home country. The availability of refuge also can encourage dissent from inside, if people realise there are havens if necessary.

While all these measures are quite compatible and indeed predictable within the theory of nonviolent action and social defence, in practice there has been little attention to the issue of acting against a coup from an outside country. Most of the attention in the social defence literature is on nonviolent action within a country against foreign aggression, which is the normal 'threat situation' for which military forces are traditionally justified. There is also considerable attention in this literature to opposition to military coups, but again mainly from within the country where the coup occurs. Yet for nearly everyone in the world, there are many more opportunities to take action against coups elsewhere than in one's own country. It is also much safer for the individual (though moral dilemmas can be severe, since one is intervening in someone else's society).

Perhaps the reason why so little attention has been given to opposing coups in other countries is that the framework of states, including the United Nations and numerous treaties, places great emphasis on the evils of violating the territorial integrity and government prerogatives of other states. The great evil, at least as presented by governments, is attacking or subverting another state. Proponents of social defence may have imbibed this orientation and thus neglected to consider nonviolent action which can offer a potent challenge to foreign governments.

I have purposefully neglected discussing action by foreign governments against military coups. In principle, they could play an enormously influential role in opposing coups. In the case of Fiji, it would have been possible for governments of such countries as Australia and New Zealand to promote Commonwealth and United Nations sanctions, to hinder trade, to block tourists from travelling to Fiji, to cut off economic aid, to withdraw investment, to beam short-wave broadcasts encouraging resistance, and a host of other nonviolent actions. But, as I discuss later, governments are unreliable opponents of repression.


There was no pre-existing organisation or network designed to respond to the initial Fiji coup. Therefore the actual responses outside Fiji were to a large degree ad hoc, just as they were inside Fiji.

The most obvious response was the publication and broadcast of numerous articles and reports about the coup in the mass media. These varied in their analysis and their degree of condemnation of the coup. What is relevant here was the scarcity of information about how people could help oppose the coup. This partly reflects the lack of any authoritative body - of the stature of Amnesty International, for example - which could pronounce on appropriate responses. If such a body had existed, some of the news media undoubtedly would have reported its recommendations as news, even if not endorsing them.

While it is not surprising that the mass media provided little indication of how to oppose the military regime, more disappointing was the response in the 'alternative media'. In Australia, for example, two left-wing weekly newspapers, Tribune published by the Communist Party of Australia and Direct Action published by the Socialist Workers Party, published a large number of articles about the coup, all condemning it. But these articles gave remarkably little attention to how to go about opposing the regime. Aside from direct reportage of the events, continuing attention was devoted to the possible involvement of the United States Central Intelligence Agency in the initial coup. Yet whatever the role of the CIA, the early path of Rabuka's regime did not depend heavily on overt external military support. In any case, the presence or absence of CIA involvement would not make a great deal of difference to practical action against the regime. The attention to the CIA seemed to reflect ideological antagonism to the US government and an attempt to fit the Fiji events into a standard Marxist analysis.

Another problem with the left-wing analysis of the coup was the assumption that capitalism was served by the events. Arguably, the coup, in devastating Fiji's economy, has hurt both local and foreign capitalists. While class issues were certainly important, Marxist analysis tends to elevate them above issues of local hierarchy (the chief system) and ethnicity.

Whatever its deficiencies, the left-wing press provided far more useful material to opponents of the regime than the mainstream press. Tribune and Direct Action offered background political analyses of Fiji and reported on opposition to the regime both within Fiji and overseas, all of which was highly useful to anyone considering their own role.

Among the Australian electronic media, the most valuable function was carried out by Radio Australia, which broadcasts throughout the south Pacific region. Its straightforward reporting of the events could be received loud and clear in Fiji and provided an authoritative counterweight to the censored Fiji media. (The BBC World Service played a similar role.)

One of the major activities by opponents outside Fiji was organising public meetings, rallies, fund-raising and the like. Much of the initiative for this action came from activists drawn from the Fijians living abroad.

While the media releases and public meetings of opponents helped to generate awareness and concern, there was no apparent overall strategy for promoting direct action. Rather, various actions working in different directions were taken.

One key activity was to lobby governments to take action against the illegal regime. This approach was supported by officials from the deposed Bavadra government and was eagerly adopted by many supporters overseas, who in turn hosted various visitors from the Bavadra government. Numerous letters were written and delegations organised to appeal to presidents, prime ministers and, not least, the Queen.

By my assessment, this approach was largely fruitless from the beginning. Governments are guided much less by legalities and justice than by practical strategic assessments. The Fiji Labour Party government promised a foreign policy more independent of the strategic interests of the United States and, for example, had a platform of banning visits of nuclear ships. Therefore it was easy to predict that the United States government, while mouthing platitudes about democracy, would provide little support for opponents of the coup.

The Australian government, which has long been subservient to the US government when strategic military concerns are at stake, seemed bound to follow the US lead. Every ideological factor should have led the Australian Labor Party government to exert major pressure against the coup, remembering that the ALP had been thrown out of office in 1975 in a 'constitutional coup' with certain similarities to the Fiji events (but an absence of any military involvement).

As noted earlier, Australian government action against the coup could have been devastating. But in practice effective nonviolent action was avoided. After a period of verbal condemnation and little effective action, the Australian government changed its practice of recognising foreign governments to one of recognising states. Thus it could recognise the Fiji state although it might supposedly disapprove of the new government. This semantic subterfuge served to obscure the double standards that would have been even more blatant had the new Fiji government been recognised while other military regimes remained in diplomatic opprobrium. Even so, the Australian government did not move to recognise the 'states' of Afghanistan and Kampuchea.

The large amounts of energy put towards lobbying governments, trying to obtain an audience with the Queen and the like were a waste and a diversion. Governments are the least likely bodies to take action against the crimes of other governments, as shown for example by the abysmal record of governments in failing to act against genocide in other countries.[15]

The statements and actions of governments are important, undoubtedly. The question for activists is how to best use their energies to oppose the foreign military regime. Arguably, it is more effective to generate concern and action at the grassroots, which then will act as a pressure on governments as well. After all, governments are occasionally responsive to popular concerns. But without obvious grassroots support, lobbying has little hope of success if the lobbyists are not saying exactly what the government wants to hear.

A much more effective channel for action against the Fiji regime was through trade unions. Bans on trade with Fiji were instituted by trade unions in Australia and New Zealand shortly after the coup, and these were a highly effective form of pressure.

The trade union bans were lifted after two months when it appeared that trade unionists in Fiji were no longer being repressed. The struggle for loyalty within Fiji certainly encompassed trade unions, and both rewards and threats induced some Fijian trade unionists to reduce their opposition to the regime. This in turn allowed some foreign trade union officials to argue against the bans. They were encouraged in this by pressures from governments and corporations to leave the issue to 'proper diplomatic channels'. Bans were reimposed after the second coup in September 1987, and again lifted by top Australian trade union officials, in spite of rank-and-file support for their continuation, after dubious claims that Fijian trade union rights had been restored.

The story of trade union opposition has many complications, but the basic points are clear. The bans were a highly effective form of nonviolent action, but the maintenance of the bans depended on a struggle over the status and actions of the new regime as well as the degree of public support for trade union action. Once again the theory of nonviolent action gives a good account of the power of nonviolent methods but gives less direction on how to maintain the action in the face of the struggle for legitimacy.

As mentioned before, tourism is a major economic activity in Fiji. After the coup, the number of tourists visiting Fiji dropped drastically: the country essentially received the wrong sort of publicity, and no longer appeared to be an idyllic haven from tension and strife. Tourism has suffered ever since, though it has been helped towards normalcy by cut-price tour packages and by various governments' tacit or overt acceptance of the new regime.

The overseas opponents of the coup could have, but did not, mount a major campaign around a boycott by tourists. For example, leaflets could have been distributed to all people visiting tourist agents, letters written to newspapers and a formal committee to promote 'ethical tourism' could have made pronouncements against going to Fiji. (Some critics argue that virtually all tourism to Third World countries is part of the wider exploitative relationships between the rich and poor parts of the world, so whether is would be advisable to recommend any tourism as 'ethical' is a debatable point.)

The advantage of a campaign around tourism is that it would affect, potentially, a large fraction of the population in countries such as Australia and New Zealand. Because holidays in Fiji are affordable by a sizable proportion of people in these countries, the message that Fiji had become an undesirable destination would be a potent one. Tourists and potential tourists could also be encouraged to write to the Fiji government or Fiji Embassy saying that they plan to travel elsewhere until democracy is restored in Fiji. These actions are something that anyone can do. By contrast, government actions and even trade union bans only involve a limited number of people who make the key decisions; others can only lobby or promote discussion.

As mentioned, the tourism factor has been a potent one even without concerted action to deter people from becoming tourists. With a plan of action mapped out in advance for such a situation, a tourist boycott could become a significant method of nonviolent action.

In summary, foreign government response to the Fiji has mainly been rhetorical, and numerous governmental nonviolent actions which could have been made have not even been mooted. The continuing efforts by overseas opponents of the coup to lobby governments had predictably poor results: most governments have been much more interested in their immediate political and economic interests than in making stands for justice and democracy and in supporting grassroots opposition to the military regime. On the other hand, several other approaches have been more effective. The large number of articles, letters and newsletters have spread information; trade union bans were very potent economically and symbolically, while they lasted; and the tourism factor has been important even though it was not pursued systematically.


The responses to the Fiji coups highlight an area of nonviolent action theory and practice which needs development: how to foster action against aggression and oppression when there are potent symbolic supports for it. Nonviolent action theory gives extensive guidance for taking action when it is clear to everyone who the aggressors are. It also explains why people decline to take action.[16] But it is less helpful in showing how to mobilise people in an ambiguous situation in which the aggressor is able to use key symbols, such as ethnicity and nationalism, to nullify opposition.

The study of nonviolent action against military coups normally focuses on opposition within the country. Yet in many cases non-government opposition from other parts of the world can play a major role. The overseas opposition to the Fiji coup has been vitally important. Yet there were no organisations with plans to confront such a situation. Advanced planning could include establishment of decision-making procedures, liaison with trade unions, plans for boycotts, networks involving a wide range of organisations, communications including short-wave radio, and regular training in nonviolent action. Since military coups regularly occur around the world, such planning unfortunately would not suffer for lack of events for application.

The Fiji coup has stimulated the beginning of some planning for similar threats in the South Pacific. The various Fiji support groups, the Nuclear-Free and Independent Pacific organisation and others are now in a position to take prompt and more organised action should coups occur elsewhere in the South Pacific.

One of the biggest problem facing activists is loss of interest in the topic by the public and the media. The outrage over the first Fiji coup kept the events in the news for quite a few months, and the second takeover by Rabuka in October 1987 rekindled interest. But the passing of time, the apparent legitimation of the regime through recognition by foreign governments, the dropping of trade union bans and the general difficulties associated with the lack of stimulating breakthroughs, have made it very difficult to muster new initiatives against the regime. This has been made all the more difficult by the various negotiations towards a new constitution and civilian government, in which Mara and other established politicians have participated. Outrage is difficult to mobilise against a regime which is cautious about appearing too overtly repressive.

I have not dealt in this article with a vitally important topic which has received relatively little attention: the process by which military regimes are gradually pushed towards representative democracy. There are quite a number of examples in which this has taken place, such as Argentina and Greece. While there are some studies which examine the inherent problems faced by militaries which take over government functions,[17] and studies which examine the process of moving from military government to civilian government,[18] there is little on the role of nonviolent action in this process. It is an important area for study, but it requires a careful analysis of the political, economic and social system as well as a host of other factors including the role of foreign governments, which are not often enough integrated into studies of nonviolent action.

Finally, concern about opposing military coups should not be at the expense of general action and strategies for promoting justice and equality which, arguably, are what are required to help prevent coups. It is almost always easier to prevent a coup than to reverse it. Indeed, preparations to oppose coups could possibly be more useful as a deterrent than as a treatment. It is too late to stop the Fiji coups, but the lessons from Fiji should be used to help prevent similar events elsewhere.[19]


1 For background information on Fiji see for example Brij V. Lal (ed.), Politics in Fiji: Studies in Contemporary History (Allen and Unwin, Sydney, 1986); Michael Taylor (ed.), Fiji: Future Imperfect (Allen and Unwin, Sydney, 1987).

2 In Fiji, the different ethnic groups are called Fijians, Indians and Europeans. However, most of the 'Indians' long ago lost contact with India and are 'Fijians' in the sense of being citizens. Therefore I prefer the clumsier terminology of Melanesian Fijians and Indo-Fijians, referring to both as Fijians, which does not confuse ethnicity with citizenship.

3 Stephanie Hagan, "The party system, the Labour Party and the plural society syndrome in Fiji", Journal of Commonwealth and Comparative Politics, Vol. 25, No. 2, July 1987, pp. 126-140; Brij V. Lal, "Emergence of the Fiji Labour Party", in Lal, op. cit. note 1, pp. 139-157.

4 See for example Stephanie Hagan, "Race, politics, and the coup in Fiji", Bulletin of Concerned Asian Scholars, Vol. 19, No. 4, Oct-Dec 1987, pp. 2-18; Anthony B. van Fossen, "Two military coups in Fiji", Bulletin of Concerned Asian Scholars, Vol. 19, No. 4, Oct-Dec 1987, pp. 19-31. Several books are forthcoming about the Fiji coups, including Robbie Robertson, Shattered Coups (Pluto Press, Sydney, 1988).

5 Gene Sharp, The Politics of Nonviolent Action (Porter Sargent, Boston, 1973).

6 Anders Boserup and Andrew Mack, War Without Weapons: Non-violence in National Defence (Frances Pinter, London, 1974); Johan Galtung, Peace, War and Defense: Essays in Peace Research, Volume Two (Christian Ejlers, Copenhagen, 1976); Gustaaf Geeaerts (ed.), Possibilities of Civilian Defence in Western Europe (Swets and Zeitlinger, Amsterdam, 1977); Stephen King-Hall, Defence in the Nuclear Age (Victor Gollancz, London, 1958); Adam Roberts (ed.), The Strategy of Civilian Defence: Non-violent Resistance to Aggression (Faber and Faber, London, 1967).

7 Adam Roberts, "Civil resistance to military coups", Journal of Peace Research, Vol. 12, 1975, pp. 19-36.

8 Jan Zielonka, "Strengths and weaknesses of nonviolent action: the Polish case", Orbis, Vol. 30, No. 1, Spring 1986, pp. 91-110.

9 Detlef Kantowsky, Saryodava: The Other Development (Vikas, New Delhi, 1980); Geoffrey Ostergaard, Nonviolent Revolution in India (Gandhi Peace Foundation, New Delhi, 1985).

10 I have relied especially on the journal Fiji Voice (Fiji Independent News Service, PO Box 106, Roseville NSW 2069, Australia).

11 "Sunrise in the West", Fiji Voice, No. 2, October 1987, p. 5.

12 "Sydney group calls for civil disobedience in Fiji", Fiji Voice, No. 2, October 1987, p. 14.

13 Karin von Strokirch, "Coup d'etat: the Fijian response", Arena, No. 80, 1987, pp. 35-41.

14 Karen Mangnall, "Blueprint for supremacy", Sunday Star (Auckland), 15 May 1988, p. A12.

15 Leo Kuper, Genocide (Penguin, Harmondsworth, 1981).

16 Sharp, op. cit. note 5.

17 Christopher Clapham and George Philip (eds.), The Political Dilemmas of Military Regimes (Croom Helm, London, 1985).

18 Guillermo O'Donnell, Philippe C. Schmitter and Laurence Whitehead (eds.), Transitions from Authoritarian Rule: Prospects for Democracy (Johns Hopkins University Press, Baltimore, 1986).

19 I thank Terry Darling, Lisa Schofield, William Sutherland and others who prefer to remain anonymous for useful discussions, and Stephanie Hagan and Ralph Summy for valuable comments on a first draft.