To understand the current conflict in Cabinda, we should take three essential factors into account. The first is the enclave's colonial history alongside the history of Angola; the second relates to socio-economic and identity issues; and the third is linked to natural resources, namely oil, and its national, regional and international impact.

The historical roots of the Cabinda question, and of the lengthy conflict in this enclave, lie in the colonial occupation sparked off by the scramble for Africa, and more specifically by the so-called 'internationalisation of the Congo question'. Portugal had already included Molembo and Cabinda in its empire in these rivalries, a right recognised by the Portuguese British Alliance Treaty of 1810, and reconfirmed by the Convention signed by both countries in 1815. The Portuguese Constitution of 1826, reconfirmed in 1838, firmly declared that "Angola, Benguela and its dependencies Cabinda and Molembo" belonged to Portugal. More than four decades later, on 1 February 1885, Angola's Governor-General, Ferreira do Amaral, would endorse the famous Simulambuco 'treaty', which is the present-day cornerstone of all independence discourse. Signed in the presence of the Portuguese marines, it was in reality a subtle territorial occupation by the Portuguese, allegedly at the local leaders' request. Furthermore, the concept of 'effective occupation' (which in this context implied military presence) was one of the conditions for recognition of colonial aspirations and territorial claims in the 1884-85 Berlin Conference. It was from this struggle for territory in Central Africa that the enclave of Cabinda was born, situated between the Congo Free State of Leopold II of Belgium (now the Democratic Republic of the Congo, formerly Zaire) and the French Congo (now the Republic of the Congo, or Congo-Brazzaville).

The identity issue evolved essentially from these factors – the Simulambuco 'treaty' and the creation of the enclave as a result of European rivalries in the 'scramble' for Africa. Previously independent and scattered communities (even though culturally and linguistically related) were united within the same territory and under the same colonial authority, with their new borders decided by Europeans.

The geographical factor – combined with decades of almost total neglect by the colonial system of what became the 'Portuguese Congo', and the Cabindan tradition of economic migration to both countries – undoubtedly created for the people of Cabinda points of identification strongly rooted in the cultures of the Congos. The presence of Cabindans in the former French Congo is so significant that since independence a substantial number of Cabinda's elite has participated directly in Congolese political life, occupying high-level positions such as Prime Minister (Alfred Raoul, and Antoine Dacosta) and President (Alfred Raoul as an interim President after the fall of President Massamba-Dedat's regime in 1968), as well as being active in youth movements.

It is from the elite that emigrated to or was born in the Congos that Cabinda's independence movement emerged in the late 1950s, in the form of associations of Cabindans residing in these neighbouring countries. Some of these associations had a regional character, as was the case of the Mayombe Alliance. However, Cabinda's aspiration for independence came clearly to the fore in the early 1960s, with the creation of the Movement for the Liberation of the Enclave of Cabinda (MLEC), led by Ranque Franque. The MLEC joined forces with other separatist groups to create the Front for the Liberation of the Enclave of Cabinda (FLEC) in 1963.

Although no military action was attributed to FLEC until some time after its foundation, the organisation did attempt to enter the political scene in 1963 by establishing a 'Government of Cabinda in Exile' in Kinshasa. Its headquarters in the Cabindan border town of Tschela, did not however undertake any guerrilla activities. It was not until after the collapse of the Salazar regime in Portugal in April 1974 that there was a real 'explosion' of separatist sentiments, emerging from a long period of lethargy. This upsurge of nationalism took place against the backdrop of an extremely complex context in Angola, characterised by confrontation between the three armed nationalist movements (the FNLA, MPLA, and UNITA) on the one hand, and the aspirations of some sectors of colonial society for a Rhodesian (white-dominated) type of independence, on the other.

The MPLA leadership blamed the rebirth of Cabindan separatism in the Kinshasa- and Brazzaville-based media on the oil companies, in particular the French company ELF, and on the local political leaders. The connection between ELF and the Congolese wing of FLEC was clear to Angolans: the separatist leader Alexandre Tchioufou, a highly-qualified Congolese of Cabinda origin, was ELF's administrator for their Congolese operation. With some foundation, they also accused Zaire's President Mobutu of preparing to annex Cabinda, in particular after his meeting with Portuguese leader General António de Spínola in Cape Verde in April 1974.

Although never entirely severed after independence, official contacts between the Angolan government and Cabinda's separatist movements resumed during the 1980s. To assist with this, prominent figures of Cabindan origin with high posts in Angola's political and administrative structures were committed to that issue, and the prestigious positions they occupied constituted a real attraction to most Cabindans.

If this situation caused serious embarrassment to the separatist movement, it was not, for that matter, a great success, for it failed to persuade the most important figures to subscribe to Angola's "policy of national reconciliation". In fact everything indicates that the two sides did not share the same spirit or philosophical approach to this process, as illustrated by records of the proceedings of some of these meetings from the 1980s onwards. While the separatist delegates presented the future status of Cabinda as a priority item on the agenda, the government representatives for a long time attempted to deal with the issue solely within the framework of its policy of forgiveness and national reconciliation. While the separatist movements attempted to discuss the future of the enclave, the government tried to conduct the meetings as dialogue between "Angolan brothers", causing the repeated breakdown of rounds of negotiations.

Between 1986 and 1989 Gabon and Congo hosted a series of meetings between the Angolan government and the various Cabindan independence factions, in Pointe-Noire, Brazzaville (Congo), and Libreville (Gabon). The talks were a genuine 'dialogue of the deaf' and dealt mostly with questions of protocol. The Cabindan delegates, who appeared increasingly divided, complained constantly that they were treated with less respect than their Angolan counterparts.

After overcoming these issues, the parties addressed the question of what the first agenda item should be. When the Angolan government finally agreed in the 1990s to discuss the issue of Cabinda's future status first, another problem emerged that the government treated as an obstacle to solving the Cabinda question: the constant divisions between the separatist movements. Their differences ranged from strategic issues to simple questions of personality and leadership struggles. The fact that some of these divisions resulted in desertions and in some separatist leaders allying with the Angolan government (whether through choice or persuasion), seems to corroborate accusations by the independence movements of attempts to "undermine them from within". The Angolan government has since argued, with increasing insistence, that the alleged lack of interlocutors caused by the separatist movements' constant divisions is the main obstacle to dialogue.

Several factors can be identified as the origins of these divisions. Firstly, a close examination of Cabindan separatism reveals tendencies related to the different formative experiences of the Cabindan diaspora in Central Africa. Franque, Tchioufou, Raoul and other figures of Cabindan origin were not educated within the same intellectual or socio-political context as each other. They were too intimately involved in the socio-political processes of the countries where they were formed as people and political activists to remain immune to the conflicting aspirations and strategic interests of the two Congos.

There are also divisions between inhabitants of the Mayombe forest and those of the coastal areas, specifically Cabinda City, with the former accused of denying Cabindan 'legitimacy' to the latter. Still further divisions exist between Francophone and Lusophone communities, creating a particular challenge for the new generation of activists, whose education took place in the post-colonial Angolan context. Their cohabitation with peers whose socio-political and intellectual roots lie in the two Congos is not always peaceful. Finally, there are complex personal interests and aspirations. These divisions have provided the Angolan government significant room for manoeuvre, both in provoking desertions and in simply promoting "understanding" of its position.

Since the end of the war with UNITA, the war in Cabinda has increased in intensity, with the government conducting "mop-up operations". Angolan civil society organisations and opposition parties have traded accusations with the government concerning events in the enclave in recent years. In October 2002, the government sent an estimated 30,000 new troops, including recently incorporated ex-UNITA soldiers, into the province in an effort to repeat their military success against UNITA. Several reports have accused the government of gross human rights abuses, without, however, giving equal attention to the behaviour of the Cabindan factions. At the end of 2003, half a dozen high-level FLEC commanders handed themselves in to the Angolan authorities and were subsequently integrated into the national army, but even this serious blow has not brought the skirmishes to an end.

What is the short and medium term impact of the government's new strategy? Will it mean further radicalisation, completely dismissing the principle of negotiations, even if only for a special status for Cabinda? And what is the strategy of the separatist movement as a whole?

In summary, the Angolan government's strategy is based on two active pillars, and a further absent one. The limitations of the two active pillars of this strategy have already started to become apparent. The first pillar consists of the continued 'search' for a 'valid interlocutor', who is understood in advance to be non-existent. The government is supported – or supports itself – in this endeavour by the separatist movement's constant internal divisions

The second and more prominent pillar is the military defeat of the separatist movement. Despite the government's justifications of its right to defend any part of its national territory on the basis of sovereignty, the fact remains that military violence not only increases resentment among the population, but also puts the government in a difficult position internationally due to the flood of reports about human rights abuses. In addition, experience illustrates that fighting against movements whose bases are located over the border has high-risk regional geopolitical implications. Furthermore, this violence provides greater media exposure for the the separatist movement and their cause, until now little known abroad.

The absent pillar of the government's strategy is the socio-economic re-investment of the petrol-dollars in Cabinda, which was expected to be the first 'trump' to be utilised in such psycho-social war games. Some Angolans claim that if the post-independence government had invested oil revenue in socio-economic and infrastructural improvements, this would have decreased the grievances of the more moderate protagonists of the Cabindan cause (including those accepting only limited autonomy) and minimised the impact of separatist discourse. Although the government decided in the early 1990s to allocate 10 per cent of the annual oil revenue to the province's budget, this measure was not part of a consistent policy framework. There are already several groups in Cabinda – including some not in favour of independence – protesting against the profound degradation of the province, the absence of basic infrastructure, the pollution of the sea, and the increasingly precarious livelihoods of local fishing communities.

However, even the intelligent use of these reserves as a socio-economic and psycho-social pillar of the war against the Cabindan separatists will be insufficient if the fundamental issue – the assertion of Cabindan identity, sharpened by socio-economic frustrations and interests – is not resolved.

While the war and the militarisation of large parts of Cabinda's interior continue, the Angolan government keeps publicly reaffirming its willingness to work towards dialogue and a negotiated settlement, or even to hold a referendum on self-determination. Yet the endless postponement of broad dialogue with the separatist movements surely coheres with the strategy of maximising their fragmentation and minimising the small chance of meeting the so-called valid interlocutor the government keeps on 'searching' for. The Angolan authorities have failed to appreciate the recent, growing role of the Cabindan Catholic clergy in this matter and risk further alienating an interlocutor and potential moderator – the Church in Cabinda.

In early 2004, the Angolan authorities twice prohibited the launch of the civic association Mpalabanda, finally launched in March 2004 under the auspices of the Catholic Church in Cabinda. These events further radicalised public opinion in the enclave and demonstrated again that the position of the Church hierarchy had shifted from purely humanitarian concerns to a much more openly political stance.

The situation of the separatist movement is much less clear-cut and it is hardly possible to formulate joint strategies within such a divided universe. Everything indicates, however, that the main strategy is the internationalisation of the Cabinda question. To achieve this, there have been a number of efforts to involve Portugal in the process again, with the declared aim of persuading the Portuguese state to resume its supervisory role in accordance with the famous Simulambuco Treaty. The second component of this strategy is to get the UN involved. This would mean that Portugal, as the 'supervising power', would act as it did in Timor. Meanwhile, the most radical factions attempt to maintain the military pressure on the ground, seeming to accept the resulting deterioration in the current situation and the excesses perpetrated by both sides.

The combined components of this strategy have little chance of success. Whoever may be in power in Lisbon, Portugal cannot afford to openly confront Angola on this issue given the increasing Portuguese commercial interests at stake in the country and both countries' membership of the Community of Portuguese Speaking Countries (CPLP). Furthermore, Portugal is bound by the Alvor Accords of January 1975, even if their validity changed after the MPLA unilaterally seized power on 11 November 1975.

The government strategy of forever looking for the non-existent interlocutor may eventually prove to be a double-edged sword; all this 'looking' gives increasing exposure to the separatist movement, while it pushes the Cabindan clergy to take a radical political position. This may lead to a situation in which a compromise solution based on a broad consensus (far-reaching autonomy) becomes less likely. It also perpetuates the risks of instability for the region as a whole.