Version classiqueVersion mobile

Poverty and Peace in the Portuguese Speaking African Countries

 | 
Cristina Rodrigues Udelsmann
, 
Ana Bénard da Costa

‘Quantitative Literature’ and the Interpretation of the Armed Conflict in Mozambique (1976-1992)

João Paulo Borges Coelho

Résumé

Recently, a number of reputable academic establishments have observed the fast growth of a new literature about civil wars, based on quantitative methodologies and privileging economic explanations. This literature, mostly rooted on a comparative analysis, seeks to uncover tendencies which allow configuring an universal theory of the occurrence, with the purpose of not only understanding but additionally informing the political decisions of organisms which are supposed to be managing it. The present article aims to bring subsidies for the discussion of the Mozambican case study under the stance of this new quantitative literature, trying to determine in which manner it may be useful to circumscribe and comprehend it.

Texte intégral

Introduction

  • 1 At a global scale, the occurrence of civil wars has risen intensely in the 90s, reaching a peak in (...)

1Throughout the last two decades of the previous century civil conflicts have multiplied, and in some manner as its outcome, so have the academic projects aiming to comprehend them, partly motivated by a need to advance knowledge, partly as a response to the political and humanitarian needs for predicting and preventing them, as well as to improve mechanisms for their resolution and to attenuate their effects.1

  • 2 Collier, Hoeffler & Rohner (2208: 3).

2A large segment of these projects refuses the traditional qualitative stance of research of particular processes, considering these perspectives make generalization difficult and imprint a tendentious character on comparisons (Gates, 2002: 9-10); alternatively, the search for new methodologies based on a “data” compilation plentiful enough to allow, by its statistical treatment, to establish general probabilistic relations which configure theoretical principles. It concerns a perspective that “abstains from any particular conflict and submits the investigator to the discipline of the statistical method”.2 In one of the programmatic texts in this new quantitative perspective, it is stated:

“Statistical patterns are useful in the way they allow to suggest policies which function typically in particular situations. Besides, they defend us from the temptation of illegitimate generalization of personal conflicts, and from the tendency to choose among a multiplicity of possible causes the one that better incorporates into the investigator’s convictions” (Collier et al, 2003: 54).

3Linked to this perspective is the creation of large databases about the conflicts, presuming the general theory developed from its interpretation permits to analyze concrete cases and predict future cases, as well as to inform political decisions.

  • 3 According to Sambanis (2004: 815), “most projects don’t perform an original historical research, be (...)

4The process is not original. It was originated in a database called “Correlates of War” (COW), created by David Singer in the University of Michigan in 1963, with the purpose of accumulating systematically every scientific knowledge concerning war all over the world (http://www.correlatesofwar.org; Suzuki, Krauze & Singer, 2002). Following the steps of COW, several other lines of investigation were developed, in the most part derived from this database, and some new databases that were constituted in the meantime.3 Evidently, this “quantitative methodology” being described here in a simplified manner has been producing diversified results, at least as many as the perspectives which operate the records to subsequently construct a “theory”. While some factions operate closely to the professed Political Science and pay special attention to the “ethnic factor” or to the impact of globalization over conflicts, most see economics, and concurrently econometrics, as the answer to the most important matters. The particular relevance of the last one is predominantly due to the significance of the project “Economics of Civil War; Crime and Violence” produced by Paul Collier and sponsored by the World Bank, which has established a correlation, largely accepted in this field, linking economic inequality and poverty to the civil war (Collier, Hoeffler & Söderbom, 2001).

  • 4 See, for example, Weinstein & Francisco (2005). Sambanis (2003: 106-108) debates the application of (...)

5Although there is an extensive and diverse academic literature about the Mozambican armed conflict, its comparison to this Quantitative Literature (QL) is still quite scarce.4

  • 5 A third set of questions is excluded, regarding the somehow distinctive subject of the ending of th (...)

6For that reason, the purpose is to produce a collection of written reflections, not concerning precisely the whole complex and diverse body of work of the QL, instead basically regarding the way it creates new perspectives intended for the study of war in Mozambique. These notes will be organized according to two or three large nucleus of questions which have been dominating the QL, namely the start of the conflict (causes, nature) and its progression (transformations, inductive conditions).5

Origins and Nature

  • 6 The most recent definition mentions a military conflict taking place inside a national territory be (...)

7What launches a civil war? Probably the reasons are as many as there are civil wars, and hence the large diversity of explanations which, in turn, are very much the result of the difficulty endured by the QL while defining an operative concept of civil war.6

8However, mostly down to the influence of Paul Collier and his perspective, based on the Rational Choice theory, the realm of the QL produced a reasonable consensus around the internal causes, such as those that if exposed would elucidate the origins of the conflict, being economy the most profound reason in this explanation. Civil war would rise from the decision made by a determinate part of society to rebel against the country’s leadership, in a dichotomy logic of a search for material or political gains, or to manifest some grievance (the famous collerian formulation of “greed or grievance”), clearly more motivated by the former than by the later, and therefore guaranteeing an economic rationality for the explanation (Collier & Hoeffler, 2000).

  • 7 The alternative exportation-importation channels from Rhodesia to Mozambique was through South- Afr (...)
  • 8 A United Nations report affirms that the Rhodesian attacks, in the beginning directed against Zimba (...)
  • 9 The fact that the MNR was born with a English acronym is not, in this context, inconsequential. On (...)

9This internal location of the causes for conflict creates some problems since the beginning with reference to the elementary narrative of the Mozambican civil war, such as it is usually constructed from the profusely available documentation. According to these records, the independence of Mozambique is included in a meaningful geopolitical alteration occurred in Austral Africa, during the middle of the decade of 1970, characterized by the disappearance of the colonialism in the region, leaving only the atypical cases of South African apartheid and Rhodesia. For this last country, an independent Mozambique ruled by a regime directly resultant from a revolutionary movement embodied two fundamental perils: on the one hand, was broadening the Rhodesian frontier exposed to the infiltration of the nationalist guerilla, and on the other hand was seriously threatening the vital access to the seashore by the Beira Port and Beira Corridor7. In this manner, the new context demanded immediate responses from Rhodesia, becoming more urgent an open support of the new Mozambican authorities to the ZANLA Zimbabwean guerilla, and the adoption, in March 1976, of the United Nations’ mandatory sanctions against Rhodesia, which in reality implied a radical cut in the relationship with that country. A classic war is swiftly initiated between these two nations.8 In this background emerges the MNR (Mozambican National Resistance), created by the Rhodesian Central Intelligence Office (CIO) as an auxiliary force for their troops.9 During this stage, the MNR has the role of combating the Zimbabwean forces located in Mozambique and destabilizing the central land strip adjacent to the Beira Corridor.

  • 10 For instance, Sambanis (2003: 45-46, note 91), when confronted with evidence of the support given b (...)

10This external origin of the rebel movement does not match the QL’s dominant expectation regarding the causes of civil war that, as mentioned before, is eminently internal. While certainly the presence of external forces is recognized as a variable which can interfere in many ways (financing the conflict, training rebel forces, buying products resulting from pillage, etc.), it is never considered to be a structural element, since this would straightforwardly contradict the principle of “rational choice”, overcomplicating the distinction between conflict among States and civil war. As a consequence, the national frame is for the QL customarily the largest frame, and everything beyond it is “diminished” to be able to squeeze in it as just another secondary factor and consequently subjected to a rudimentary treatment.10

11Evidently, though this interpretation of the Rhodesian origin in this stage of the conflict has been acknowledged as historic evidence, to embrace it exclusively would leave out many important matters. For as intense and militarily efficient the Rhodesian offensive had been, by itself it could never explicate the prompt support given by a considerable number of Mozambicans to the rebel contingent. For that, a few other factors must be taken into account.

12According to Gates (2002), the QL considers as main independent variables related to the risk of occurrence of civil wars: poverty (lack of economic opportunities and low rates of economic development), temporal distance from a previous conflict, ethnic dominance, as well as political instability.

  • 11 For instance, in 1977 there were only 30 communal villages already concluded and 105 in progress (B (...)

13Pertaining to the first variable, the Collier-Hoeffler model (CH) evaluates Mozambique as having a national raw growth “slightly under the average of other countries at war”, within a context further damaged by Frelimo’s economic strategy, in particular the collectivization of agriculture and nationalization of commerce, losses derived from sanctions against Rhodesia and restrictions to the migration of Mozambican miners to South Africa. However, a more accurate chronologic analysis will show that, although the mass departure of Portuguese populations was strong in the beginning having had a direct and profound impact on the economy, the economic measures proposed by the new government can not be associated in this manner to the beginning of the insurgence, since they only had consequences after the Third Frelimo Party Congress in February 1977, at a time when the MNR was already established.11 This same argument is used in favor of the decrease of migratory labour, as opposing to what had been assumed, in 1975 the number of Mozambican miners raised to over 115 thousand, while the price of gold tripled and Mozambique profited from unprecedented revenues of 50 million pounds, which helped to soften the transitional blows. Just since 1976 has South Africa selected a policy of progressive reduction of labour (Hanlon, 1984: 51). Furthermore, the alleged social resentment caused by the economic decline must be analyzed in a background of opposing perceptions, popular enthusiasm about the independence and a high popularity feature, achieved through equalitarian measures announced by the liberation movement. Sambanis himself (2003: 16) is eventually forced to consider the relationship between poverty and the beginning of a conflict as more complicated than the way it is proposed by the CH model, concluding in his later “review” that Renamo’s initial recruits were not unemployed but victims of repression at the hands of Frelimo.

  • 12 As was pointed out by Bertelsen (2005: 7), to Collier, although the colonial heritage may presumabl (...)

14The Sambanis’ statement should be analyzed under the perspective of the second independent variable of the CH model, which holds the less time occurred since the previous conflict, the larger the risk of emergence or reoccurrence of a civil war. This in relation to what can be considered as the only open entrance to a possibility of historic interference.12 More concretely, in the Mozambican case is made a connection between colonial war (ended in 1974) and the new war, launched one or two years afterward. The risk of occurrence of this conflict seemed, to the CH model, to be extremely high.

15Concerning Mozambique, the MNR contingent would have been sustained by Frelimo dissidents, dissatisfied with the institutional dominance from the part of the “southern people” and from victims of the repression started after the independence who had taken refuge in Rhodesia.

  • 13 Knowledgably, the difficulties in the research on this subject are partly derived from the smashing (...)
  • 14 Three months prior to the independence, some important dissents such as Lázaro Nkavandame, Uria Sim (...)
  • 15 Among them, interned religious sects followers, participants in the military upraise of December 19 (...)

16This is quite a sensible matter, undoubtedly deserving further research.13 Here are presented at least three potential groups of actors, namely Frelimo’s historic dissidents, the “dissidents” produced after the independence, and Mozambicans compromised with the colonial regime. Regarding the first group, their ability of constituting an armed movement in opposition to the new regime is less likely, especially if taken into consideration the attempts for the constitution of alternative political movements which, at that time, ended being neutralized by the imprisonments and confinements into internment camps of its main leaders.14 About the second group, there is evidence of some elements joining the rebellion particularly after it had been initiated, that is to say after the attacks and escapes from Frelimo’s internment camps.15 However, is not credible for this component to have had significance and management enough to constitute by itself the rebellion’s core. Sambanis has corrected his interpretation of the CH model and changed slightly his argument, affirming that many of the initial recruits came from the Mozambican diaspora in Rhodesia, and then concluding that “the diaspora may also aggravate the risk of war, by forming the foundation for the first recruits of a rebel organization”. (Sambanis, 2003: 106). But although Rhodesia had a traditionally significant Mozambican community, and being the birthplace of one of the strongest nationalist organizations which have initially integrated Frelimo, it is still unlikely that, in the midst of the Rhodesian diaspora and inside the environment of the governing regime, existed such an immediate capacity to organize an armed action.

  • 16 For this question see, among extensive literature, Borges Coelho (1993), Souto (2003), Bernardo (20 (...)

17Thus, the structural connection to the previous conflict must be found mostly in the context which the last stage of colonial war had created in the middle of the country, such context being characterized by the colonial policy of africanization of war, the complex relations between Mozambique Rhodesia and Malawi, and profound experiences of social engineering, population relocation and creation of ethnic troops. In fact, the nationalist Mozambican guerrilla, around 1972 after having crossed the Zambebe River in Tete, was entering Manica and Sofala in the centre of the country, consequently turning into a direct menace to the Beira Corridor. From that moment until the independence, the centre became the most delicate battlefield, which, in the Portuguese side, had more than 40 thousand Africans representing over 50% of the colonial contingent (Wheeler, 1976; Cann, 1998; Borges Coelho, 2003a). The main military response from the Portuguese consisted in the creation of highly trained and itinerant Special Groups, located in the vicinity of Beira, composed almost exclusively of Africans born in the regions where they operated. In 1974, when happened the coup in Portugal which opened the door to independence, there already existed 83 companies of Special Groups and 12 Parachuting Special Groups, in a total of thousands of men operating in the entire region, primarily in the infiltration route between the south of the Tete Province and the north of Manica and Sofala, which would become the initial operational area for the MNR.16 Without any doubt, the difficulties of integration of these forces in the new regime following the independence have paved the way to their recruitment during the process of formation of the MNR.

  • 17 The relation between ethnicity and violence, much in vogue amid certain currents, constitutes one o (...)
  • 18 This rudimentary application of the variable in the CH model is criticized by Sambanis (2003: 40-42 (...)

18The CH model variable concerning the ethnic dominance is related to the previous question, having manifested large operational difficulties. It is quite a polemic subject among the QL, concerning the ability to revolt of significant ethnic groups without affordance to power17. The CH model finds a positive correlation linking this variable and the beginning of war, derived from the northern group Macua-Lomwe, numerically superior but without access to power, since Frelimo’s leadership was dominated by “southern people”. That “would have caused regional dissidences inside the liberation movement which lead to the emergence of civil war”.18 Amazed by the absence of an armed revolution among the macua-lomwe ethnic, which would demonstrate the effectiveness of this variable, the model then searches for an indirect orientation to link it throughout the history of the liberation movement until the origins of the conflict. It would be much more promising to explore this subject based on the notion – more indistinct and, at the same time, more inclusive – of identity, to account for the regional tensions that without a doubt were being historically constituted integrating ethno- linguistic elements, but also of distinctive paces of development and distinctive distances regarding to power, articulated into the new context of construction of a national identity (Borges Coelho, 2004); and answering to the central question on the origins of the Mozambican conflict being unrelated to ethnicity matters.

19The last QL variable, in some manner associated to the first one, concerns the political instability as a cause for the initiation of civil wars. This variable does not explicitly make part of the CH model, but can be associated to another secondary one, called “democracy”. In essence, this is formulated in the following manner: right away following the independence Mozambique was characterized by a non-democratic regime, and the MNR was originated from factions repressed by this regime, although it was basically more motivated by a search for material benefits than by any resentment. Sambanis believes this interpretation to be inconsistent, claiming that even though the governmental repression may have caused resentments they might not constitute a direct motivation for violence, instead eventually being the prospective improvement of material living conditions. Moreover, he proposes a new variable, not integrated in the CH model but “having possible interest for the launch of a civil war”, which he denominates “State’s incapacity of regulation over the entire territory”. According to this, Frelimo as a liberation movement had hardly penetrated into the territory before taking over power, implying that when they formed a government they had to fill the void left behind by the Portuguese, without nevertheless being capable of managing the territory. This incapacity might have been presumably enhanced by the repression exerted over all of those who had connections to the colonial regime, including the colonial security forces elements who “had they been enrolled, may have helped (to fight the MNR) [and most likely the civil war would not have taken place]” (Sambanis, 2003: 108).

20This is certainly one of the most complex questions to be analyzed. The difficulties in managing the country (including the territory) must be undoubtedly taken into consideration in a background of regional hostility, abrupt removal of the most competent leaders and profound transformation and deterioration of the economy and, for last, repression and incarceration in reeducation camps of many factions compromised with the colonial regime, political and religious dissidents, the unemployed, victims of a persecutory zeal resulting from a perspective of political and ideological purity or just the whim of unprepared guerrilla men. However, this perspective grossly ignores the huge popularity asset in possession of the new regime, freshly out of a liberation war (Egerö, 1987), and in its reductive interpretation disregards the complex role played out in this context by the links between Frelimo and the State. This disregard for the context will lead to the callow and conservative position of uncritically conceive the possibility of inclusion of forces highly trained by the colonial regime to combat the nationalist guerilla in extremely hostile regional surroundings as the solution to avoid a civil war. Finally, there is no doubt that the frailty of the State may be associated to its incapacity to respond in terms of repression to the MNR guerilla, given that during the first years of independence the new regime faced severe tribulations while transforming its guerilla force into an effective conventional army, and all this under attack by the Rhodesian forces (Borges Coelho & Macaringue, 2004). Nevertheless, it is undeniable that in 1979-80 the decease of MNR’s leader André Matsangaísa, the independence of Zimbabwe and the governmental military offensives in Gorongosa and Manica had taken the MNR almost to the point of annihilation.

  • 19 There is another variable in the CH model which deems the low percentage of school attendance among (...)

21Concerning the cycle from the start of war (particularly noticeable in Mozambique’s situation, between 1976 and 1980), it is therefore very difficult to believe the QL, specifically the CH model, has produced stimulating perspectives. The relevant independent variables either ignore any context,19 or are chronologically dyslexic (for instance, the social impact of the economic decline during this stage), are reductive and quite inconsistent (for instance, all analysis made on the performance of the new regime), or practically incomprehensible (the case of ethnic dominance); and when the inference seems to be correct, such as in the case of proximity to the previous conflict, such happens for the wrong reasons.

22Some authors in the field of the QL search to attenuate the most patent difficulties in an application of the CH model to the Mozambican case. Sambanis (2003) suggests the diaspora as the recruitment source for the MNR, reduces Rhodesia’s role to a substitute factor for the internal pillage (paying the price for the serious methodological mistake of a voluntary mix of variables, so to maintain internalized the origin of conflict), and lastly proposes the introduction of new variables (“external actors”, “Cold War”, “disenfranchisement and repression of colonial beneficiaries”, “State’s incapacity to control the territory”). Weinstein & Francisco (2205) reveal themselves more aware of the professed external factor and recognize that, without more rigorous information, it is impossible to examine many questions systematically, although eventually they in the same manner integrate the “external support” into the “pillage of assets” to preserve the endogeny of the CH model.

Duration and Transformation

  • 20 Even though accepting this separation in case the determinants of both processes are clearly distin (...)

23When a group starts an armed rebellion, it does not pose itself the question of when and how it will end. Collier, Hoeffler & Söderborn (2001: 17) claim, based on this fact, that the “duration of a conflict is determined by a set of variables substantially different from the ones determining its beginning”, and the acceptance of this principle has made these two moments to be studied in an increasingly divided manner.20

24Concerning the independent variables prevailing in conflicts, usually the most relevant are the geographic characteristics of the land, social fractionalization (ethnic and religious) and the economic opportunities subsequent to the hostilities. Although less consensual, the interference of external actors has become increasingly relevant (Elbadawi & Sambanis, 2000a).

  • 21 Until then South Africa had performed a rather secondary role, limited to a consolidation of the ec (...)

25This previous variable continues to be crucial to the elementary narrative of the Mozambican case. Indeed, previous to the independence of Zimbabwe the MNR was practically extinct, reduced to only about 300 fighters. Its survival was due to the introduction of a new actor in this process, South Africa,21 which in the course of the “Operation Mila” transferred from the northeast of its territory to the extinct Rhodesia the remaining structure of the MNR, while its fighters concentrated in the mountainous region of Sitatonga. However, just the following year Renamo was once again operating in Mozambique’s inland with a contingent of thousands of men.

  • 22 About the South African support to Renamo, cf. Among extensive literature, Vines (1991) and Martin (...)

26To understand the South African support and this powerful resurgence of the rebel movement it is necessary to regard South Africa not in a statistical manner but as coping with deep changes, in direct confrontation against countries with black majorities which had organized themselves around the Frontline and the SADCC to confront them. Botha replaced Vorster in the leadership, and the military have radicalized their position within the frame of the “Total Strategy” elaborated in 1977 to guarantee the “survival of the white nation” in opposition to what was seen as a communist progression in Austral Africa. During this period, the South African armed forces became dominant and exerted a huge influence over the relationships with neighbouring countries. They expanded their military industry (between 1978 and 1985 their military expenditure more than doubled), created ethnic battalions and scouting regiments – Recces – integrating “dissidents” from all around Austral Africa. With regard to Mozambique, this attitude has provoked a massive “investment” in the recuperation of the MNR/Renamo.22 In turn, to understand the occidental complacency toward this South African radicalization, it is necessary to account for an international context of Cold War, a hardening of the North American positions during the final stage of the Carter Administration, and especially the arrival of Ronald Reagan (Minter, 1994). Together with Iran, Afghanistan and the Horn of Africa, Austral Africa then becomes scenery for the confrontation between these two blocks.

27While Rhodesia’s interest had fallen upon a fairly circumscribed Mozambican area (the Beira Corridor and the bordering areas of Manica and Gaza, where the Zimbabwean nationalists had concentrated), South Africa uses the “new” Renamo to fulfil more ambitious purposes. It certainly aims at the Beira Corridor yet (thus “reassuring” Renamo and guaranteeing at the same time that Zimbabwe keeps depending on the South African deliverance channels), but further than that, it seeks for the substitution of the Mozambican regime, pressuring the entire country its capital and in particular. However, contrary to Rhodesia, South Africa will gradually aspire for more, not just as an auxiliary force but as a progressively autonomous movement with its own political purposes.

  • 23 The Project Social History of War, by the History department of the UEM (Universidade Eduardo Mondl (...)
  • 24 Muaquiwa, for example, is described as an enormous strategically positioned base, with extensive fa (...)

28In 1982-1983 Renamo’s offensive had practically reached every part of the country, with the exception of a few areas in Cape Delgado.23 By infiltrating through the borders or transported by helicopters, the guerrilla men retake the route from the border in Pafúri through the south of Manica and into the interior of the country. In 1982 they were once again standing in the Beira Corridor and Gorongosa from where then proceed to Tete (the first ambushes in the road to Chimoio-Tete took place in June). Their swift establishment in this region presumes the existence of infiltration lines, not only from Manica and Sofala, but also from Malawi. At the same time, throughout 1982 they crossed the Zambebe River and penetrated the occidental part of Zambezia Province, invading several district headquarters and founding important bases (like Mureremba, Eruruni, Alfazema and Muaquiwa24). In the beginning of 1983 they made an impressive entrance into Nampula Province and advanced to Niassa Province. In February 3rd 1984 the railroad connecting Nampula and Lichinga was attacked for the first time, seriously affecting the Nacala Corridor. Simultaneously, in the south of the country, Renamo progressed rapidly from Vilanculos and along the coastline, an action facilitated by the great drought experienced in the beginning of the decade of 1980, which had depopulated large regions in the north of Inhambane Province. In the start of 1982 they were approaching Inhambane, then progressed swiftly toward the south with forces properly equipped with bazookas, mortars and machine-guns, while facing a weak governmental resistance. At the same time, most of the inland region of Gaza becomes the target of attacks that affected the fertile regions in the valley of the Limpopo River and the national road crossing the country from north to south. Also in 1982, Renamo started to perform actions along the entire southern border with South Africa, while the bases in Matsequenha and Kwahla acted as a support to the siege around Maputo, the national capital.

  • 25 For an elaboration on the spatial factor among the QL see for example Tearon & Laitin (2003) and Ch (...)
  • 26 Both the woodland coverage and the mountainous land are measured in percentages in regards to the t (...)
  • 27 Evidence reinforced by the implicit knowledge of the posterior stage of apartheid by the South Afri (...)
  • 28 The CH model deducts a positive correlation, although the mountainous land is small (2.4% of total (...)

29This fast and simultaneous progression of Renamo into such a wide extension as the entire Mozambican territory raises several questions. From the start, it places the geographical factor in a substantially different position than formerly proposed by the QL.25 The variables entailed in matters of geography are restricted to a positive factor for the military evolution of the rebels, which may be represented by the woodland coverage and the rocky configuration of the land.26 However, this ability of the guerrilla to, within the short period of two years, ubiquitously reach everywhere with this much intensity and efficiency poses the possibility of going beyond mere territorial physical configurations into a spatial dimension necessarily connected to the war choreography and its combats, as well as to the South African strategic and tactic outlook, interested in controlling the economic corridors (Beira, Nacala and Limpopo) to hinder the economy in certain costal locations (to supply the guerilla) and in sieges around main cities, particularly the capital. This factor, linked to the intensity and efficiency which had marked the offensive, not only places many nuisances to the possibility of an endogenous creation within the movement (in such case, its evolution would have been much slower and circumscribed to determinate locations in the territory, never reaching the whole country) as they provide for an exogenous explanation in accordance to the demonstrated South African motivation, with the factual evidences of Renamo receiving supplies by helicopters and coastal shipments, the presence of instructors and special forces in the bases, and the sophisticated system of communications with the South African background.27 Even if the land configuration has wielded some influence over the military operations, the geographical factor includes many more elements and certainly cannot be evaluated in such a simplistic and abstract manner.28

  • 29 Between 1978 and 1982 the population involved in the relocation of communal villages has practicall (...)
  • 30 These two explanatory lines are powerfully and simultaneously structuring the same debate over the (...)

30Once more, this perspective does not contradict the argumentation in defence of the internal factors. Quite the opposite, the complementary nature of the fast progression of the guerrilla should be explained by the governmental difficulty to act in response and, now correctly, by the profoundly negative internal social and economic impact of the selected development strategy and the policies put into practice, especially since the beginning of the 1980s. In brief, this strategy was derived from a dualist and highly centralized perspective whose investment efforts were canalized into a public sector seen as modern and full of alleged potential to become a driving force for development, neglecting the bulk of small farmers who formed the smashing majority of population, already weakened by a rupture of the commercial networks and consecutive years of disinvestment. In the beginning of 1983, the national exportations had fallen down to half and the importations had decreased about one third while its value were five times higher than the preceding (Castel-Branco, 1994). The disastrous effects of the development strategy linked to the devastating outcome of war and drought have caused a profound crisis. The rural world is increasingly becoming a paradoxal space, where war violence coexists with an accelerated and compulsory population relocation conducted by Frelimo and the State29, deteriorating the executive and productive structure of the rural society without allowing for the predicted alternative; this world has gradually transformed into a violent space, falling into the fringe.30

  • 31 On the violence of the conflict, and violence as a recruitment method, see among extensive literatu (...)
  • 32 Gastrow (2001), for instance, searches to outline a frame for this kind of transactions in Austral (...)

31In the end of 1983, Renamo had already distinguished itself from the old Rhodesian MNR. With military effectiveness and defiant rhetoric creates social support funds and consolidates its presence around the territory, while growing in number. The recruitment and funding are crucial elements to the QL variable regarding the duration of the conflict, which is still far from being wholly comprehensible. Nevertheless, the available information disengages from the simplicity of explanation offered by the application of the CH model. The economic perspectives progressively blocked from the rural communities, such as the response to this reality and to the modernist discourse from the State that restrains traditional values (Geffray, 1991; McGregor, 1998) are evidently other factor to take into consideration, as long as they remain circumscribed within a frame where violence is an essential mechanism for recruitment, and the increasing possibilities for the guerilla correspond to the enlargement of their operational areas;31 in addition, it is an essential mechanism for financing that beyond having an external component, mostly South African, is also founded on the extortion and extraction of resources by pillaging villages and other locations and ambushing and trading along the border, facts that have yet to be circumscribed and registered in detail.32

  • 33 The accord, described as a non-aggression and goodwill pact, fundamentally prescribed that Mozambiq (...)
  • 34 About the Zimbabwean involvement, started in December 1984 and by far the most significant, see Mla (...)
  • 35 From wandering across a country at war, Finnegan (1992: 71) has observed that “the nature of war is (...)

32The Nkomati Accord33, signed by Mozambique and South Africa in March 1984, may be the symbolic mark for this continuous “internalization” process of Renamo. Following this accord, South Africa was forced to have a more discreet involvement, but remained always present and efficient (Davis, 1985), which produced the diversification of political and material supports to Renamo among the conservative sectors in the United States of America, Europe and some African countries (Vines, 1996) and a larger political autonomy, assisting the movement in attaining its “maturity”. Contrary to what was expected, the conflict escalated in intensity and spread internationally, with the military participation of Zimbabwe (particularly in the Beira Corridor, but also in the Limpopo railroad and Tete railroad corridors), Tanzania (in Zambezia) and Malawi (Nacala Corridor) beside the Mozambican troops.34 Renamo fortifies its position in the territory like the spots in a leopard skin, provoking and feeding off local micro-conflicts, some ancient, others related to rising tensions between tradition and modernity, a few about access to propriety, etc.35 Many of these micro- conflicts returned in the circumstances of a larger conflict and were incorporated into Renamo’s political rhetoric, while others simply gravitate around it. Pertaining to the classic definition, this is the moment when one can really talk about civil war.

33One of the most serious problems of the QL is the intrinsic need to reduce the number of actors, also to immobilize them, so it can be obtained absolute values that provide input to statistic operations. The “population” is transformed into a single category, devoid of any diversity other than a curious ethno-territorial classification, incapable of evolving along the process. This category is reduced to the role of victim (what it precisely is), being as such denied a function of agent in the process. Although it would be more productive to inquiry about the diversified strategies explored as solutions to war and violence, those strategies are not constrained to any ideological choices and are based on matters of survival.

34Likewise the State, in accordance with the CH model, is a simple entity, devoid of complex relations with the Frelimo Party, and its action is regarded as merely repressive. Its contradictory nature – authoritarian and simultaneously populist – falls out of reach, alike to the profound transformation that took place during this period, promoted by measures resulting from the IV Frelimo Congress in 1983, and followed by failed attempts of reconnection with South Africa in 1984, and the stirring internal reforms after an overture to the Occident, a progressive departure from the socialist model (despite the continuity rhetoric) and the adherence to programs of structural adjustment in 1987, at that time in the midst of war. Such transformation would result from it and, at the same time, affects it.

35In the beginning of nineties, civil war was at an impasse, which although real from the military standpoint, in a general perspective was only apparent. Meanwhile, the nature of the actors went through profound changes, consequently leading to the reconfiguration of their relationships at several levels; since then, at a global level of Cold War and at a regional level. Actually, it must be considered that 1989 was the year when the Berlin Wall fell, was also the year when de Klerk rose to power in South Africa and when the apartheid began to collapse. And, for once, these transformations would follow the course of an internal dynamic from which they are a part of, taking place on a fabric exhausted by war, social and humanitarian demands, political motivations, economic needs and corporate interests concerning peace.

36The interpretation of war requires, in this instance, an extreme sensibility and attention to the multiplicity of actors and their respective transformations along the course of time, from which results a permanent reconfiguration of their relations; a diachronic and synchronic interpretation that the paraphernalia of binary correlations, Gini coefficients and “dyadic” connections in the QL aren’t able to provide. The so-called “case” to be studied does not translate into disperse and fragmented data, but into an elementary narrative, a construction in progress and as a result constantly polemic, being part of this construction an accuracy of information and depuration of political and ideological interferences. A narrative pressing forward the collected knowledge and integrating in its body an innate rationality elaborated from the activities and choices made by its agents and by the numerous backgrounds where they perform. This set of elements is better understood by means of categories of process or dynamic, in the sense defined by Kalyvas, Wood & Bell (2007).

Conclusion

37Everything that was here designated as QL forms a complex and much diversified body. However, inside this group of perspectives can be distinguished a few common characteristics. In the first place, the operative rationality, consisting in the reduction of war’s social processes into a set of supposedly objective facts, so that by mean of their statistical combination can be found general probabilistic tendencies from which to construct independent variables able to interpret the meaning of past conflicts and predict future conflicts. A second characteristic, not exclusive but nevertheless dominant, is the primacy of the economic explanation and of econometrics as the preferred technique.

  • 36 If it is certain that war escalated in 1972 with the start of the Manica and Sofala frontlines, the (...)
  • 37 Weinstein & Francisco (2005: 186 and next) suggest a daily collection of information from newspaper (...)
  • 38 In the words of Sambanis (2003: 56), “the studies on the case may help to correct mistakes of measu (...)

38This rationality raises some relevant methodological questions. Since the start it entails meticulousness and quality of the alleged empirical data, many times coming out filled with gross mistakes or codified according to doubtful standards. There are numerous examples. Fearon (2002, Appendix 2) believes the Mozambican conflict has occurred between 1976 and 1995 (ignoring the ceasefire of October 1992) and the colonial war taking place between 1965 and 1969 as a Portuguese internal conflict. The same standard is followed in the basic classification of the UCDP/PRIO, which considers the colonial war as a minor conflict in 1964-1965, an intermediate conflict in 1966-1971, and a proper war in 1972-1973 (Strand, Wilhelmsen & Gleditsch, 2002).36 At last, Collier, Hoeffler & Söderbom (2001) demarcate the colonial war to a period ranging from 1964 to 1975 (ignoring the 1974 ceasefire). Some authors manifest a certain restlessness concerning these problems, suggesting new research processes37 and more careful reviews of the designated case studies (Sambanis 2003; Gates 2002; Blattman & Miguel 2009; Mack, Humphreys & Weinstein 2004), recognizing the binary quantitative variables would not allow to comprehend the contexts, discussing conceptual problems of codification and simultaneously rejecting variables that “did not deserve to be taken seriously” (Hegre & Sambanis 2005).38

39Even disregarding the rigor of these “facts”, there is still a need to question the more general meaning of the dismemberment of real social processes and their reduction to a sequence of data. The contextual problems are multiple, both at temporal and spatial levels. The attempts to solve them by means of establishing temporal sequences (quinquennial, in several instances) are insufficiently convincing, since the cuts do not correspond to periodization criteria. In this situation of application of the CH model to Mozambique, the only defined temporal distinction is the indistinguishable separation between start and duration of the conflict.

  • 39 There are already radical cases of application of variables in a virtual world, for example to demo (...)
  • 40 The “triumph” of economy comes up in splendour in the words of de Jack Hirschleifer (quoted in Cram (...)
  • 41 About the general critic of the collerian perspective, see Ballentine & Sherman (2003), Malone & Ni (...)

40These are the facts that, once grouped into sequences, compose the “fast food” of this theory. These facts are related to each “unit” (every conflict or every country represents a “unit”) which, linked among themselves through complex formulas of probabilistic calculation, allow to validate the independent variables, which is to say, the sentences within the theoretical text. Subsequently, this text is not directly originated from the reality of such processes but from this intermediation, during which its intrinsic rationality, pre-existing before the explanative operation, is lost. The significance of this interaction between occurrences and actors disappears and is replaced by the new significance the investigator wants to imprint it with.39 This new significance is mostly achieved through the QL, in the exclusive realm of a hypothetical triumph of economy over every other subject,40 not of economy in general but of a neoclassic economy anchored in a singular individual who rationally chooses war as a way to reach his purposes. As it was declared by Cramer (2000: 11), we are witnessing changes in the explanation of conflicts, from the romantic perspective of grand motivations to the cynicism of interests, while in practice the causes become reduced to “functional and unhistorical abstractions”.41

  • 42 It is suspected that in many occasions the facts are “massaged” to fit into the theory.
  • 43 About the new perspectives of integration of the external factor, see for example Lockyer (2008); G (...)

41In the Mozambican conflict case, beyond a lack of rigor of the information42, the problem is mostly structural, given that the orthodoxy of the endogenous economic explanation prevents the apprehension of an extremely important external dynamic which is created and presented throughout the entire conflict;43 the dyadic perspective prevents the apprehension of the number of actors involved; their “congealment”prevents the perception of their transformation during the conflict and which manner this dictates a permanent reconfiguration of the links among them; the perspective of a country as an homogeneous unit prevent from encompassing the complex territorial variations. In whole, the outcome is like the one referred by Kalyvas, Wood & Bell (2007: 1-2) when they declared that “these works result in essentially formal and highly abstract models, whose main problem is in this abstraction often being obtained at the cost of unrealistic assumptions, frequently derived from economic analogies whose empirical operationalization appears to be hard to carry out.” It is akin to expecting a movie and instead finding a disarticulated sequence of bad images barely connected to reality.

  • 44 Some examples are the use of the term “dyad” (a loan from cytology, meaning each one of the double (...)

42By looking for the objectivity of numbers as an evasion from the beliefs that swarm the research labour, the QL becomes ironically attached to them, transmitting a strange and contradictory sensation of “high-tech” technology prompted by a neoclassic conception inspired in a 19th century colonial flavor. Such an example is the perception of ethnic or religious groups, evaluated by head tallies and placed into clearly demarked territories, collective movements acting under a single infuriated and ambitious individual, backed by a State repressing everything – all this transmitted by a terminology worthy of semiotic research.44

43The detachment from reality and the extremely conservative perspective in these studies conducted under the supervision of prestigious universities and institutions as significant as the World Bank are not that surprising, if taken into account the genesis of the QL itself. Seemingly new, it nevertheless corresponds to an old northern outlook over the atavistic southern barbarianism, an effort to explain it by means of the old confliction between their social organization and modern greed, and whose interest regarding the violence of conflicts in distant countries equates to a fear of these same conflicts also affecting the North. In truth, most of QL is a deviation dedicated to a laborious exercise which however has given up on reaching true explanations. It misses out on everything that Theda Scokpol called “historical imagination”, and this wouldn’t be a reason for worry if it wasn’t in this instance informing many important political decisions related to the occurrence of conflicts; decisions which affect its course and its outcome.

Bibliographie

Africa Watch (1992) Conspicuous Destruction: War, famine & the reform process in Mozambique, New York: Human Rights Watch

Armon J , D Hendrickson and A Vines (eds.) (1998) “The Mozambique Peace Process in Perspective”, London: Accord – An International Review of Peace Initiatives, 3

Ballentine K & J Sherman, (eds.) (2003) The Political Economy of Armed Conflict, Londres/Boulder: Lynne Rienner

Bernardo, M.A. (2003) Combater em Moçambique. Guerra e Descolonização (1964-1975), Lisbon: Prefácio

Bertelsen B.E. (2005) “War, Peace and Development in Mozambique: A critical assessment”, Peace Building and Post-War Aid Workshop, Bergen: CMI, June

Blattman C. & Miguel, E. (2009) “Civil War”, Centre for Global Development, Working Paper Nº 166, March

Borges Coelho J.P. & Macaringue, P. (2004) “The Role of Mozambique’s Armed Forces in a Changing Security Context”, in Peter Batchelor, Kees Kingma & Guy Lamb (eds.), Demilitarization and Peace-Building in Southern Africa. Vol III: The Role of the Military in State Formation and Nation-Building, Hampshire UK: Ashgate

Borges Coelho J.P. (1993) “Protected Villages and Communal Villages in the Mozambican Province of Tete (1968-1982). A History of State Resettlement Policies, Development and War”, Bradford: Bradford University (doctorate thesis)

Borges Coelho J.P. (2003a) “Da Violência Colonial Ordenada à Ordem Pós-Colonial Violenta. Sobre um Legado das Guerras Coloniais nas Ex-Colónias Portuguesas”, Lusotopie

Borges Coelho, J.P. (2003b) “Antigos Soldados, Novos Cidadãos: Uma Avaliação da Reintegração dos Ex-Combatentes de Maputo”, in B. Sousa Santos & J.C. Trindade (eds.), Conflito e Transformação Social: Uma Paisagem das Justiças em Moçambique, Porto: Afrontamento

Borges Coelho J.P. (2004) “Violent Possibilities: state identity and conflict in Mozambique”, 2nd Workshop of the Research Network “From Inequality to Insecurity? The Place of Crime and Violence in Development Thinking and Action”, Lisegaarden (Denmark), 23-25th of September

Cahen M. (1989) “Is Renamo a Popular Movement in Mozambique?”, Southern African Review of Books, December

Cann J.P. (1998) Contra-Insurreição em África (1961-1974): o modo português de fazer a guerra. Estoril: Atena

Castel-Branco C.N. (1994) “Problemas Estruturais do Desenvolvimento Agrário”, in C.N. Castel-Branco, Moçambique – Perspectivas Económicas, Maputo: UEM/FFE

Cederman L.-E., L. Girardin & A. Wimmer (2006) “Getting Ethnicity Right. An Expert Survey on Power Distributions Among Ethnic Groups”, http://www.sscnet.ucla.edu/soc/faculty/wimmer/ESEG_APSA_Paper%20final.pdf

Chojnacki S & N Metternich (2008) “Where to fight?: the spatial dynamics of violence in areas of limited statehood’, San Francisco: 49th Annual ISA Convention, 26-30th of March

Clarence-Smith G. (1989) “The Roots of the Mozambican Counter-Revolution”, Southern African Review of Books, April

Collier P. & A. Hoeffler (2000) “Greed and Grievance in Civil War”, The World Bank, The World Bank Research Group, Policy Research Working Paper No. 2355, May

Collier P. & A. Hoeffler (2002) “On the Incidence of Civil War in Africa”, Journal of Conflict Resolution, Vol. 46, No. 1. DOI: 10.1177/0022002702046001002

Collier P. & A. Hoeffler (2006) “Civil War”, draft chapter for the Handbook of Defense Economics, http://users.ox.ac.uk/%7Eeconpco/research/Civil-War.pdf

Collier P., A. Hoeffler & M. Söderbom (2001) “On the Duration of Civil War”, The World Bank, The World Bank Research Group, Policy Research Working Paper No. 2681, September

Collier P., A. Hoeffler & D. Rohner (2008) “Beyond Greed and Grievance: Feasibility and Civil War”, http://oep.oxfordjournals.org/cgi/content/short/gpn029v1

Collier P., L. Elliot, H. Hegre, A. Hoeffler, M. Reynal-Querol & N. Sambanis (2003) “Breaking the Conflict Trap. Civil War and Development Policy”, Washington: World Bank/Oxford University Press. http://hdl.handle.net/10986/13938

Cramer C. (1997) “Civil War is not a Stupid Thing: Exploring Growth, Distribution and Conflict Linkages”, http://www.soas.ac.uk/economics/research/workingpapers/file28886.pdf

Cramer C. (2000) “War, Famine and the Limits of Economics”, Conference on Global and Local Dimensions of Food Security: Threats, Challenges and Responses, University College Cork, 13-15th of April, http://www.ucc.ie/famine/conference/ papers/Cramer.doc

Davis R. (1985) “South African Strategy Towards Mozambique in the Post-Nkomati Period: A Critical Analysis of Effects and Implications”, Uppsala: The Scandinavian Institute of African Studies, Research Report No. 73

Egerö B. (1987) A Dream Undone: The Political Economy of Democracy 1975-1984, Uppsala: Nordiska afrikainstitutet

Elbadawi I. & N. Sambanis (2000a) “External Interventions and the Duration of Civil Wars”, The World Bank, The World Bank Research Group, Policy Research Working Paper No.2433, September

Elbadawi I. & N. Sambanis (2000b) “‘Why Are There So Many Civil Wars in Africa? Understanding and Preventing Violent Conflict”, Journal of African Economies, Vol 9, No. 3. DOI: 10.1093/jae/9.3.244

Elbadawi I. & N. Sambanis (2002) “How Much War Will We See?”, Journal of Conflict Resolution, Vol 46, No.1. DOI: 10.1177/0022002702046003001

Fearon J. & D. Laitin (2003) “Ethnicity, Insurgency and Civil War”, American Political Science Review, 97(1): 75-90. DOI: 10.1017/S0003055403000534

Fearon J. (2002) “Why do Some Civil Wars Last so Much Longer than Others?”, July 12th http://www.stanford.edu/group/ethnic/workingpapers/dur3.pdf

Fearon J. (2003) “Ethnic Structure and Cultural Diversity Around the World: A Cross-National Data Set on Ethnic Groups”, http://www.stanford.edu/group/ethnic/workingpaper/egroups.pdf

Finnegan W. (1992) A Complicated War. The Harrowing of Mozambique, Berkeley: University of California Press

Flower K. (1987) Serving Secretly, London: John Murray

Freire Antunes J. (1995) A Guerra de África (Vol. 1), Lisbon: Círculo de Leitores

Freire Antunes J. (1996) Jorge Jardim, Agente Secreto, Lisbon: Bertrand

Garfinkel M., S. Skaperdas & C. Syropoulos (2005) “Globalization and Domestic Conflict”, CESifo Working Paper No. 1510, http://econpapers.repec.org/paper/cesceswps/_5f1510.htm

Gastrow P. (2001) “Organized Crime in the SADC Region: Police Perceptions”, Institute for Security Studies, Monograph 60

Gates S. (2002) “Empirically Assessing the Causes of Civil War”, Annual meeting of the International Studies Association, New Orleans, 24-27th of March

Geffray C. (1991) A Causa das Armas. Antropologia da Guerra Contemporânea em Moçambique, Porto: Afrontamento

Gersony R. (1988) Summary of Mozambican Refugee Accounts of Principally Conflict-Related Experience in Mozambique, Washington DC: Department of State/Bureau for Refugee Programs

Gleditsch N. P. (2002) “The Future of Armed Conflict”, The Madame Madeleine Feher European Scholar-in-Residence Lecture, The Begin-Sadat Centre for Strategic Studies, Bar-Ilan University, https://besacenter.org

Gleditsch N. P., A. Knudsen, A. Suhrke & H. Urdal (2003) “Conflict and Development: Framework for as Proposed Research Area”, Oslo: International Peace Research Institute / Christen Michelsen Institute

Gleditsch N. P., P. Wallensteen, M. Eriksson, M. Sollenberg & H. Strand (2002) “Armed Conflict 1946-2001: A New Dataset”, Journal of Peace Research, Vol 39, No. 5. DOI: 10.1177/0022343302039005007

Hall M. & T. Young (1997) Confronting Leviathan: Mozambique since independence, London: C. Hurst & Co

Hanlon J. (1984) Mozambique: The Revolution Under Fire, London: Zed Books

Hegre H. & N. Sambanis (2005) “Sensitivity Analysis of the Empirical Literature on Civil War Onset”, http://journals.sagepub.com/doi/abs/10.1177/0022002706289303

Hoile D. (ed.) (1994) Mozambique 1962-1993: A Political Chronology, London: Mozambican Institute

Humphreys M. (2003) “Economics and Violent Conflict”, https://www.unicef.org/socialpolicy/files/Economics_and_Violent_Conflict.pdf

Johnson P. & D. Martin (1986) Destructive Engagement: Southern Africa at War, Harare: Zimbabwe Publishing House

Kalyvas S., L. Wood & J. Bell (2007) “The Dynamics of Civil War”, report presented during the Annual Meeting of American Political Science Association, Chicago, August 30th, http://www.allacademic. com/meta/p228675_index.html

Laitin D. & D. Posner (2001) “The Implications of Constructivism for Constructing Ethic Fractionalization Indexes”, in K Chandra, ‘Symposium: Cumulative Findings in the Study of Ethnic Politics’, APSA-CP Newsletter, Vol 12(1), http://www.nyu.edu/gsas/dept/politics/faculty/chandra/apsa-cp2001.pdf

Lockyer A. (2008) “The Levels of Analysis Approach to the Study of Foreign Intervention in Civil War”, Brisbane: Australian Political Studies Association Conference, 6-9th of July

Mack A., M. Humphreys & J. Weinstein (eds.) (2004) “Understanding Civil War: Quantity versus Quality?: Toward more effective collaboration between the Quantitative and Qualitative conflict research communities”, conference report, Bellagio: Human Security Centre/Liu Institute for Global Issues, 5-7th of April, http://www.hsrgroup.org/docs/Publications/Additional-Publications/Human-Security_ReporProject_Understanding-Civil-War-QuantityOverQuality.pdf

Malone D. & H. Nitzschke (2005) “Economic Agendas in Civil Wars. What We Know, What We Need to Know”, United Nations University/World Institute for Development Economics Research, Discussion Paper No. 2005/07. DOI: 10.1057/9780230595194_3

McGregor J. (1998) “Violence and Social Change in a Border Economy: War in the Maputo Hinterland 1984-1992”, Journal of Southern African Studies, Vol 24, No. 1. DOI: 10.1080/03057079808708566

Minter W. (1989) “Clarence-Smith on Mozambique”, Southern African Review of Books, June Minter W. (1994) Apartheid’s Contras: An Inquiry into the Roots of War in Angola and Mozambique, London/New Jersey: Zed Books

Mkandawire T. (2002) “The Terrible Toll of Post-Colonial Rebel Movements in Africa: Towards an Explanation of the Violence Against the Peasantry”, Journal of Modern African Studies, 40, 2. DOI: https://doi.org/10.1017/S0022278X02003889

Mlambo N., s/d, “Raids on Gorongosa: Zimbabwe’s military involvement in Mozambique 1982-1992”, SACDI Defense Digest, Working Paper no. 3

Nordstrom C. (1997) A Different Kind of War Story, Philadelphia: University of Pennsylvania Press

Oliveira P. (2006) Renamo: Uma Descida ao Coração das Trevas, Lisbon: Europress

Rodrigues R. (ed.) (1995) Os Últimos Guerreiros do Império, Amadora: Editora Erasmus

Roesch O. (1989) “Is Renamo a Popular Movement in Mozambique?”, Southern African Review of Books, December

Sambanis N. (2002) “A Review of Recent Advances and Future Directions in the Quantitative Literature on Civil War”, Defense and Peace Economics, Vol 13(3), pp. 215-243. DOI: 10.1080/10242690210976

Sambanis N. (2003) “Using Case Studies to Expand the Theory of Civil War”, CPR Working Papers, no. 5, May

Sambanis N. (2004) “What is Civil War: Conceptual and Empirical Complexities of an Operational Definition”, Journal of Conflict Resolution, Vol 48, No. 6, December. DOI: 10.1177/0022002704269355

Santos B. O. (2008) “Histórias Secretas da PIDE/DGS: Entrevista com Óscar Cardoso”, http://group.xiconhoca.com/2008/12/05/historias-secretas-da-pide-dgs

Souto A. N. (2003) “A Administração Colonial Portuguesa em Moçambique no Período de

Marcello Caetano (1968-1974): Mecanismos e Relações de Poder”, Lisbon: Universidade Nova (doctorate thesis)

Stiff P. (1999) The Silent War. South African Recce Operations 1969-1994. Alberton: Galago

Strand H., L. Wilhelmsen & N.P. Gleditsch (2002) Armed Conflict Dataset Codebook (version 1.1.), Oslo: International Peace Research Institute

Suzuki S., V. Krause & D. Singer (2002) “The Correlates of War Project: A Bibliographic History of the Scientific Study of War and Peace, 1964-2000”, Conflict Management and Peace Science, Vol 19, No. 2: 69-107. DOI: 10.1177/073889420201900204

Vines A. (1996) Renamo: From terrorism to democracy in Mozambique?, London: James Currey

Vines A. (1998) “The Business of Peace: Tiny Rowland, Financial Incentives and the Mozambican Settlement”, in J. Armon, D. Hendrickson & A. Vines, The Mozambican Peace Process in Perspective, Accord – An International Review of Peace Initiatives, No. 3

Weinstein J. & L. Francisco (2005) “The Civil War in Mozambique: The Balance Between Internal & External Factors”, in Paul Collier & Nicholas Sambanis (eds.), Understanding Civil War: Evidence and Analysis (Volume 1: Africa), Washington DC: The World Bank

Wheeler D. (1976) “African Elements in Portugal’s Armies in Africa, 1961-1974”, Armed Forces and Society, 2(2). DOI: 10.1177/0095327X7600200206

Wimmer A. & B. Min (2006) “From Empire to Nation-State: Explaining Wars in the Modern World, 1816-2001”, American Sociological Review, Vol. 71, December. DOI: 10.1177/000312240607100601

Wimmer A., L.-E. Cederman & B Min (2009) “Ethnic Politics and Armed Conflict: A Configurational Analysis of a New Global Data Set”, American Sociological Review (Online Supplement), Vol 74, April, pp. 316-337. DOI: 10.1177/000312240907400208

Woodward S. L., s/d: “The Inequality of Violence: On the Discovery of Civil War as a Threat to ‘the North’ in the 1990s and the Debate over Causes and Solutions”, APSA Task Force on Inequality and Difference in the Developing World, http://paperroom.ipsa.org/papers/paper_5344.pdf

Notes

1 At a global scale, the occurrence of civil wars has risen intensely in the 90s, reaching a peak in 1994-5, and since then has been consistently decreasing (Sambanis, 2002: 216; Fearon, 2002; Gleditsch, 2002). Some have found the causes for this tendency in the long cycle of general transition from empires to nation-states (Wimmer & Min, 2006). The region most affected by the phenomenon is the Sub-Saharan Africa, and the reasons for this prevalence are discussed, among other, by Collier & Hoeffler (2002), Elbadawi & Sambanis (2000b) and Fearon & Laitin (2003). All websites were consulted in June 2009.

2 Collier, Hoeffler & Rohner (2208: 3).

3 According to Sambanis (2004: 815), “most projects don’t perform an original historical research, being mostly founded on COW”. Among the most important databases the emphasis is put on the “Minority at Risk” by Ted Gurr, the “State Failure Project” by the George Mason University, the “Uppsala Conflict Data Programme (UCDP) related to the “Armed Conflict Dataset” by the International Peace Institute in Oslo (PRIO) (Gleditsch et. al., 2002; Gates, 2002), the “Ethnic Power Relations (EPR) by the University of California (Wimmer, Cederman & Min, 2009; Cederman, Girardin & Wimmer, 2006), and the “‘Cross-National Data Set on Ethnic Groups’ by James Fearon (Fearon, 2003). Many more are in progress. For a global outlook on quantitative literature, see Sambanis (2002), Humphreys (2002).

4 See, for example, Weinstein & Francisco (2005). Sambanis (2003: 106-108) debates the application of the Collier-Hoeffler model (CH) to Mozambique, but does it based on an only source, specifically an earlier version of the works by Weinstein & Francisco. For an analytical vision of the CH model applied to Mozambique, see Cramer (2000) and Bertelsen (2005).

5 A third set of questions is excluded, regarding the somehow distinctive subject of the ending of the conflict (found solutions, post-conflict effects, etc.).

6 The most recent definition mentions a military conflict taking place inside a national territory between a State and one or more groups, from which results a tally of at least 1000 direct deceased (according to the COW criterion) or 25 direct deceased every month (according to the UCP/PRIO criterion). See, among extensive literature on this topic, Collier & Hoeffler (2006), Blattman & Miguel (2009).

7 The alternative exportation-importation channels from Rhodesia to Mozambique was through South- Africa, entailing in average a distance four times longer while doubling the costs (Mlambo s/d).

8 A United Nations report affirms that the Rhodesian attacks, in the beginning directed against Zimbabwean camps in inner Mozambique, rapidly evolved into large-scale operations aiming not only civil population as well as economic and social infrastructures. “Since May 1977, the Rhodesian forces have been using Mirage fighter-bomber aircrafts and 100 kg fragmentation bombs.” (see Borges Coelho, 1993: 370-371).

9 The fact that the MNR was born with a English acronym is not, in this context, inconsequential. On the creation of the MNR/Renamo and his action during this stage, see for example Johnson & Martin (1986), Minter (1994), and Vines (1996). The role played by Rhodesia in creating the movement is eloquently referred by Ken Flower, its instigator: “(…) The CIO proceeded to the recruitment of Mozambicans who were encouraged to carry on their own actions without having to depend on the support of Rhodesia. The surprising easiness in the development of the MNR indicated that we were following the right path, particularly since we kept it clandestinely small and manageable during the first years and, at the same time, it served as eyes and ears for our Intelligence in Mozambique” (Flower, 1987: 302).

10 For instance, Sambanis (2003: 45-46, note 91), when confronted with evidence of the support given by Rhodesia to the MNR, rushed to declare that it wouldn’t have worked if were not for the failed agricultural policies, intense political repression and “southern dominancy”.

11 For instance, in 1977 there were only 30 communal villages already concluded and 105 in progress (Borges Coelho, 1993: 334). In effect, this program has only become significant since 1977, following the floods in the Zambebe and Limpopo Rivers in 1977 and the systematic relocation policies.

12 As was pointed out by Bertelsen (2005: 7), to Collier, although the colonial heritage may presumably influence the conflict in a certain manner, such influence is much less important than the economic features. This way, he fights the “pollution” within the context (Collier et. al., 2003: 66).

13 Knowledgably, the difficulties in the research on this subject are partly derived from the smashing dominance of an official narrative destitute of problematisation.

14 Three months prior to the independence, some important dissents such as Lázaro Nkavandame, Uria Simango, Joana Simeão and Paulo Gumane were already under detention in Nachingwea, Tanzania, among hundred other. See Hoile (1994: 26 and passim).

15 Among them, interned religious sects followers, participants in the military upraise of December 1975 and mostly victims of the purifying wave in the midst of the defence and security forces. André Matsangaísa, the first leader of the MNR himself had escaped from a re-education camp.

16 For this question see, among extensive literature, Borges Coelho (1993), Souto (2003), Bernardo (2003), Cann (1998), Rodrigues (1995), Freire Antunes (1995 and 1996), Santos (2008).

17 The relation between ethnicity and violence, much in vogue amid certain currents, constitutes one of the most confusing, polemic and conceptually fragile features in the QL. The theory is founded on the concepts of “ethnic dominance” (relationship of power and ethnicity dimension) and “ethnic fractionalization” (heterogeneity). The main source of information for the databases, in this specific topic, has been the soviet Atlas of the Peoples of the World (Atlas Narodov Mira) from 1964, and in some instances the State Failure Project. The confusion instated in this field, here only superficially explored, is well expressed in the harsh criticism patent in the words of Laitin & Posner 2001: 13-15: “the users of the ELF index (Ethno-linguistic Fractionalization, with data from 129 countries) assume that the ethnic fractionalization of a determinate country is fixed, like its topography or the distance separating it from the equator. Just like the borders of a country don’t change, it is assumed that its ELF values also remain constant (...). The value of an ELF measure ignores the social reality crossing several dimensions of the ethnic identity in all countries, and a political life characterized by different levels of fracture in several different dimensions (…)”, etc. The authors have concluded that this supposed measurement is equivalent to taking a single figure, for example the inflation rate of a country in 1945, and from it deduce the prosperity level of this same country in 1990. For research on the ethnic factor in the QL, see Woodward (s/d).

18 This rudimentary application of the variable in the CH model is criticized by Sambanis (2003: 40-42, 107), for being superficial and focusing exclusively on the larger group, for not always being well measured, for omitting other components of ethnic affiliation (such as race and religion) which might be used to support ethno-political action, for not revealing anything about the transversality of ethnic, religious, racial and identitary ruptures in general, etc.

19 There is another variable in the CH model which deems the low percentage of school attendance among young men as a dissatisfaction factor or indigence as having a positive correlation in the Moza mbican case right after the independence, this variable completely ignoring the context of transition. In fact, the decrease in percentage corresponds to the departure of European populations, and since the African youths practically had no access to education during the colonial times, for them the situation only could have improved.

20 Even though accepting this separation in case the determinants of both processes are clearly distinctive, Eldabawi & Sambanis (2002: 307-308) present a group of technical and practical reasons for which they should be regarded jointly. Collier et. al. (2003: 80) considers the length of civil war harder to investigate statistically than its beginning, as while comparing countries at war and at peace, such occurs in a much limited variations among countries at war.

21 Until then South Africa had performed a rather secondary role, limited to a consolidation of the economic connections to Mozambique and to a discreet collaboration in the Rhodesian warfare. About the South African support to Mozambique, cf., among extensive literature, Vines (1996), Martin & Johnson (1986), Minter (1994), Davis (1985).

22 About the South African support to Renamo, cf. Among extensive literature, Vines (1991) and Martin & Johnson (1986), Minter (1994), Davis (1985), Oliveira (2006), Africa Watch (1992). About the military operation of the ‘Total Strategy’, cf. Stiff (1999).

23 The Project Social History of War, by the History department of the UEM (Universidade Eduardo Mondlane) gives some steps in the sense of making a meticulous cartography of the progression of Renamo inside the territory, based on regional and local testimonies.

24 Muaquiwa, for example, is described as an enormous strategically positioned base, with extensive farmland cultivated by the population. On occasion, South African planes would land there with supplies. There were some South African instructors in this base.

25 For an elaboration on the spatial factor among the QL see for example Tearon & Laitin (2003) and Chojnacki & Metternich (2008).

26 Both the woodland coverage and the mountainous land are measured in percentages in regards to the total surface of the country. For a definition, go see appendix in Collier, Hoeffler and Söderbom (2001).

27 Evidence reinforced by the implicit knowledge of the posterior stage of apartheid by the South African authorities themselves. See, for example, Davis (1985) and Stiff (1999).

28 The CH model deducts a positive correlation, although the mountainous land is small (2.4% of total area), when compared to the average in countries at war (24.9%) and at peace (15.2%). Sambanis (2003).

29 Between 1978 and 1982 the population involved in the relocation of communal villages has practically doubled, going from 1.160.437 to 1.808.693 individuals , and the purpose was to embark about 6,500,000 until 1985 (Borges & Coelho, 1993: 345; 361).

30 These two explanatory lines are powerfully and simultaneously structuring the same debate over the Mozambican conflict, opposing the supporters of war as aggression and destabilization to those who give more importance to the negative impact of State policies. Regarding this, see a famous debate in Southern African Review of Books between Clarence-Smith (1989), Roesch (1989), Cahen (1989), Minter (1989), etc.

31 On the violence of the conflict, and violence as a recruitment method, see among extensive literature, Minter (1994), Gersony (1988), Vines (1996), Geffray (1991), Nordstrom (1997), Finnegan (1992), Borges Coelho (2003b). The “adherence”, although entailing perspectives of access to resources, is always candidly and smashingly described by the “adherents” themselves as forced.

32 Gastrow (2001), for instance, searches to outline a frame for this kind of transactions in Austral Africa, observing that since the middle of the 1980s was established a new link between groups of neighbouring countries and South Africa based on trafficking merchandises such as ivory, drugs, minerals, rhinoceros horns and others, in exchange for several products. This trade would have been pursued with other countries bordering Mozambique, like Malawi. About extortion, see for example Vines (1998).

33 The accord, described as a non-aggression and goodwill pact, fundamentally prescribed that Mozambique was to stop giving support and shelter to the South African ANC and South Africa would do the same in regards to Renamo. See Armon, Hendrickson & Vines (1998).

34 About the Zimbabwean involvement, started in December 1984 and by far the most significant, see Mlambo (s/d). Some authors give especial attention to the internationalization of conflicts (Elbadawi 1999; Elbadawi and Sambanis 2000). However, they always approach it as an endogenous variable that affects neighboring countries through an effect of “contagion” or “diffusion” (Sambanis 2003: 43-44; Kalyvas, Wood & Bell 2007).

35 From wandering across a country at war, Finnegan (1992: 71) has observed that “the nature of war is so different from one region to another that can be deceiving to only talk about Renamo.”

36 If it is certain that war escalated in 1972 with the start of the Manica and Sofala frontlines, then it must be taken into account that in 1871 it was causing over 1000 deaths every year (see Borges Coelho, 1993: 183).

37 Weinstein & Francisco (2005: 186 and next) suggest a daily collection of information from newspapers during the duration of war. The problems caused by such kind of solutions are noticeable.

38 In the words of Sambanis (2003: 56), “the studies on the case may help to correct mistakes of measurement and to improve the conceptualization and evaluation of the variables utilized to test the proposed hypothesis from the theory of quantitative research. The case studies may also help to build better theoretical models”.

39 There are already radical cases of application of variables in a virtual world, for example to demonstrate hoe ethnicity and nationalism play an important role in the making of conflicts. Cederman & Gerardin (2007: 12) have even developed a computational program that facilitates the replication, simulation, analysis and validation of complex social processes, giving special attention to civil wars. Besides, it allows inscribing empirical facts into the model, to calibrate environments and mechanisms to appropriate levels of realism. We are already entering the aseptic world of a virtual country and videogames, a world free from the pollution of reality.

40 The “triumph” of economy comes up in splendour in the words of de Jack Hirschleifer (quoted in Cramer 2000): “as long as this continent (of conflicts) is being explored, economists will cross ways with a certain quantity of native tribes – historians, sociologists, psychologists, philosophers, etc. – who, in the course of their several intellectually primitive ways, have preceded us in acknowledging the dark side of human activity. But every time we, the economists, get involved, we evidently sweep away these atheoretical aborigines.”

41 About the general critic of the collerian perspective, see Ballentine & Sherman (2003), Malone & Nitzschke (2005), Mkandawire (2002), Cramer (1997, 2000) and Woodward (s/d).

42 It is suspected that in many occasions the facts are “massaged” to fit into the theory.

43 About the new perspectives of integration of the external factor, see for example Lockyer (2008); Garfinkel, Skaperdas & Syropoulos (2005).

44 Some examples are the use of the term “dyad” (a loan from cytology, meaning each one of the double chromosomes in the reduction division of the cellular core) applied to the relation between the State and the rebels; viewing of the process as a “puzzle”; the atavistic violence of African societies; and the numbers, always the numbers and formulas to lend an appearance of scientificity to these interpretations.

Acheter

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search