João Paulo Borges Coelho

(text not edited)

Mozambique has been a particular victim of rapid climatic change, regularly enduring the scourge of drought, torrential rains and floods, as well as tropical cyclones. In the 25 years since independence, the country has been hit by two severe droughts of more than two years in both cases, at least 16 tropical cyclones and several floods, two of them of very wide scale. These phenomena are commonly called 'natural calamities'.

After independence, the term 'calamities' acquired a wide meaning: it came to refer to all sorts of evil affecting society in a context of searching for the 'enemy' - forces hostile to the construction of the nation state. Thus, parallel to the calamities caused by regional enemies (above all, apartheid), by extension, there arose a definition of nature as inimical when functioning abnormally. Later, 'calamities' came also to describe, by association, goods donated by the international community to help victims of emergency situations created by war as well as natural calamities, with the Portuguese word 'calamidades' referring to items such as clothing or maize flour. In this text, natural calamities are considered as extreme climatic events - particularly, droughts, floods and cyclones - characterised by accentuated and often very rapid changes in climatic behaviour, with frequently catastrophic consequences for economy and society. Such consequences consist of the threat to security of communities, their culture and goods, whether suddenly and totally (cyclones, floods), or gradually, causing the scarcity of water and food (drought).

Mozambique is obviously not an unusual victim of extreme climatic events. Indeed, nowadays these have become a widespread global problem. Accentuated by various natural or man-made factors, their causes vary from industrialisation to population growth and the intensive utilisation of resources - in many cases, impeding their renovation. Some of these causes have effects which are wholly negative and frequently regional or even global, and are identifiable, although not always consensually or definitively.

In general, the scientific identification of the factors causing extreme climatic events is carried out by the most developed countries, which have greater capacity for scientific research and more mechanisms to enunciate the identification. Normally, these two capacities - identification and enunciation - are found in northern states.

On the other hand, the states of the peripheral south are, correspondingly, not only considered as important agents in the causal processes of 'extreme climatic events' (above all as a result of the poor use of resources), but also find themselves deprived of the social and scientific capacity to identify the factors which give rise to such processes (to 'perceive' them as global factors).

Whatever the causes, global or local, extreme climatic events manifest themselves in a concrete and local way. This text seeks to problematise the Mozambican response to the consequences of these phenomena. What makes the Mozambican case special in this respect in the last few decades is that the intensification of the local effects of abnormal climatic events took place in the context of a particularly long drawn out and destructive civil war. This combination generalised the misery of the rural areas and resulted in the policies adopted to respond to calamities being strongly affected by, but also influencing, this context.

The Mozambican case shows clearly how policies adopted in response to emergency situations caused by natural calamities (and also to prevent such situations) are far from being merely technical operations. On the contrary, they are established in specific historical contexts and depend on the nature of the state that formulates them, as also on the relation between state and society. The conflictuality of this relation is rooted not only in diverse interests but also in the very unbalanced conditions of production and legitimisation of knowledge, which structure the definition of those policies.

The end of the colonial order in Mozambique in 1975 implied a profound breach with the colonial state and the emergence of completely new structures. This resulted from a combination of factors such as the massive exit of the Portuguese population (including the majority of civil servants), the extremely weak colonial heritage in respect of the education of the black population and the posture of the liberation movement, FRELIMO, which advocated such a rupture.

Understandably, the new revolutionary state, 'socialising' in nature, yet weakened by the lack of professionals, means, traditions and procedures, was absorbed by efforts to impede the collapse of the economy, 'control' both population and territory and confront what was considered to be the main enemy - military aggression from the 'white regimes' of Rhodesia and South Africa. It did not pay much attention to questions of security in relation to natural calamities.

Surprised in 1977 by the first major floods of the Limpopo and a little later of the Zambezi, the state established commissions to provide aid to the affected victims. Experience of the Zambezi floods, to combat the effects of which the state established the Inter-provincial Commission for Natural Calamities and Communal Villages (Comissão Inter-Provincial das Calamidades Naturais e Aldeias Comunais - CIPCNAC), revealed the limitations of ad hoc institutions created in times of crisis. CIPCNAC had great difficulty in mobilising resources and, above all, co-ordination, composed as it was of personnel seconded from other organisms on an individual basis. As a result, two years later the government announced the creation of the Co-ordinating Council for the Prevention and Combat of Natural Calamities (Conselho Coordenador de Prevenção e Combate às Calamidades Naturais - CCPCCN), headed by the Prime-Minister. Shortly afterwards, this body created its executive arm, the Department for the Prevention and Combat of Natural Calamities (Departamento de Prevenção e Combate às Calamidades Naturais - DPCCN), headed by a National Director under the authority of the Minister Co-operation.

The first years of the 1980s proved to be very difficult. Alongside the severe drought, which began to make itself felt towards the end of 1981 and which lasted until 1984, civil conflict, which had already existed in embryonic form for two years, now intensified rapidly, with the expansion of RENAMO guerrillas to almost the whole country. The combined effects of war and drought, aggravating the effects of rural socialisation, resulted in the complete destabilisation of rural Mozambique, creating a highly negative structural situation which came to last more than a decade. The consequences for the economy were naturally catastrophic, intensified as they were by the accentuated deterioration in the international terms of trade. As a result, from 1983, for the first time the country became an importer of foodstuffs as well an important recipient of food aid.

As Mozambique could not receive significant help from socialist countries, the government sought aid from the west, particularly from the United States. This was accompanied by the necessary signals that the politics of the country was really changing, with the first steps toward a market economy and a formal request for membership of the World Bank and International Monetary Fund. At the same time, these were paralleled at the political level by the first contacts for peace negotiations with South Africa, the main source of support for anti-government forces in the civil war.

Natural Calamities in Mozambique since Independence

The United States responded positively to the requests for food aid on certain conditions. One of these was that consignments of food aid should reach beneficiaries without passing through state institutions considered suspect, namely the state commercial network and the DPCCN. The Mozambican government responded that the prevention and combat of natural calamities was a matter of national interest. In the end an agreement was reached allowing the DPCCN to carry out the distribution of food aid in partnership with an American NGO, CARE (Concerned Americans for the Reconstruction of Europe). In this way, CARE International came to operate in Mozambique with central authorisation and a very wide mandate, ranging from technical assistance to the training of personnel and organisation of the DPCCN, as well as direct action in the system of transport of food aid.

CARE International helped create a Logistical Support Unit (LSU) in the DPCCN. While the DPCCN defined intervention policy, the LSU functioned as a technical unit, equipped with means of radio communication and a fleet of lorries for the transport of food consignments. The LSU had control over technical co-ordination, statistics, training of personnel and the transport and storage of goods.

The negative impact of drought combined with war forced the DPCCN to widen its range of action. While initially it had focussed its efforts only on Inhambane province, in 1984 it operated in three, and in 1987 in all ten provinces of the country. At this juncture, the Department created provincial branches to co-ordinate its work at this level.

The CCPCCN thus became a weighty institution which, in May 1987, the government sought to transform by creating in its place the National Emergency Executive Commission (CENE - Comissão Executiva Nacional de Emergência), headed by the Vice-Minister of Commerce. Following the appearance of the CENE, Emergency Commissions (CPE) were created in all the provinces, in what was intended to be a more decentralised system: CENE established its list of priority provinces in terms of action, while the CPE established their own priorities and co-ordinated emergency aid operations at the provincial level.

From this time there were two central co-ordinating organisms in the CENE. The first was the Emergency Technical Council (Conselho Técnico da Emergência - CTE), headed by the co-ordinator of CENE. The sectoral emergency units of the ministries involved, namely, health, education, agriculture, construction and water, and transport and commerce, participated in this body. The CTE was responsible for the identification, conception, implementation and control of emergency projects. The other main co-ordinating structure was the Emergency Operations Committee (Comité das Operações de Emergência - COE), also headed by the head of CENE. The aim of this body was to ensure effective relations with international bodies: it was constituted by representatives of individual donor countries, NGOs, UN agencies and government structures (Ratilal, 1989: 77-79; 110-122).

At the end of 1988, with Mozambique having joined the World Bank and showing clear signs of liberalising its economy, the government invited its various partners to reflect upon the efficacy of state-based emergency logistics. At this time, USAID strongly advocated the involvement of private transport operators, while other partners, such as Canada, the Netherlands and the Nordic countries, indicated their reservations about such a decision, arguing that private operators, operating on a commercial basis, would not be inclined to take food aid to the areas considered most difficult.

From the end of the 1980s, the government sought to impede the development of parallel emergency structures which might add to the already very complex organisational situation. An effort was made to integrate existing structures into more permanent ones among state organisms such as the Ministries of Agriculture and Commerce and Transport. At the same time, difficulties created by the war brought the government to invite the NGOs to involve themselves increasingly in the emergency effort. From this time, various tripartite agreements were made between government, NGOs and United Nations agencies, particularly with the World Food Programme (WFP).

The drought of 1991-92, associated with the problems of distribution of food aid caused by the war, created what was perhaps the most difficult situation of the post-independence period. At this time the country requested emergency food aid of some 450 thousand tons a year, a volume that the DPCCN was far from capable of handling. It was at this juncture that the WFP created its own logistical unit (UNILOG). At the same time, as a result of the talks preceding the Peace Agreement of 1992, the United Nations Humanitarian Assistance Co-ordination (UNOHAC) department was created as the humanitarian thrust of the UN peace operation in Mozambique (ONUMOZ).

The severe drought of the early 1990s was also a very important factor behind the pressure exercised by external partners (particularly the NGOs) for the establishment of a peace accord between the government and RENAMO. Only an effective cease-fire would make possible the opening of corridors for emergency aid to reach hundreds of thousands of victims.

In the period of the implementation of the Peace Accord, from 1992 to 1994, there continued to be a vast array of emergency operations. Apart from support for the repatriation of 1.5 million refugees in neighbouring countries and the resettlement of the more than 4.5 million internally displaced citizens, the humanitarian structures were also involved in support for the victims of drought in 1991-2 and 1994-5, and floods in 1996.

However, with the end of the war - without doubt the major cause of emergency situations - the falling away of the original justification for maintaining such a complex structure and the problems which it brought about, as well as international pressure, resulted in considerable restructuring. At the end of 1994, CENE ceased operating and, after a period of discussions with UN agencies, representatives of donor countries and of SADC, a new institutional model, implying an organism much smaller than the DPCCN but capable of ensuring co-ordination, began to emerge. Finally, in June 1999, the National Institute for the Management of Emergencies (Instituto Nacional de Gestão das Calamidades - INGC) was created to substitute the DPCCN.

Besides ensuring the co-ordination of activities related to emergencies, it is INGC's role to involve banks and insurance companies as partners, and to mobilise companies and civil society. The INGC is, like the old DPCCN, under the tutelage of the Ministry of Foreign Affairs and Co-operation.

The development of the response by the Mozambican state to natural calamities was informed by various contextual factors and also by a range of political options. Since independence, there have been two main periods: the socialist and the neo-liberal. In the first, which lasted until the mid-1980s, the dominant context was, without doubt, the extreme hostility with which Mozambique was viewed by its Rhodesian and South African neighbours, a context which, if it did not completely invalidate, made extremely difficult the implementation of coherent and effective policy in the chosen direction. This was not only because it sidelined any policies to respond to the natural calamities in the list of state priorities (or furnished the justification for such) but also because it impeded the development of regional perspectives for the resolution of problems which in large part were regional or global in origin. A clear example of this was the non-existence of agreements on the management of water resources: all of Mozambique's main rivers rise in the hinterland, where there are various storage dams, placing the country at the mercy of the management of these infrastructures upstream during times of drought or flood.

With respect to the initial political options behind the shaping of the response, clearly the most important was the tendency for highly centralised organisation, given formal expression in the Third Congress of Frelimo and developed in the following years. Apart from the need to guarantee control of the country - under attack by Rhodesians and threat from South Africans -, centralised authoritarianism was the result of several factors. These included the example of the socialist states during the Cold War, the authoritarian legacy of the colonial state - of which the new state, somewhat paradoxically, claimed to be the antithesis - and also, for some, the influence of the rigid hierarchy of 'traditional' power. In the economic sphere, this conception resulted in the emergence of a perspective in which progress was conceived in quantitative terms, in a dualist scenario where the state was the prime 'motor', and the peasant masses relegated to the marginal role of backward partner, which should nevertheless ensure its own reproduction. While the state took over the best land and the few installations left by the colonial period - apparently in order to modernise agriculture and sustain the accumulation which would provide the basis for development -, the peasantry would be concentrated in communal villages - in reality labour reserves for state enterprises. In these first years of independence, 90% of investment funds were attributed to the state sector, while the co-operative sector was awarded a mere 2% and peasant families completely neglected.

The first responses of the independent state to the frequent natural calamities affecting the country were not targeted specifically, but rather subordinated to the objectives and efforts stemming from this developmental perspective, which gave great priority to political and administrative control of population and territory. Thus the state pressured and aided the creation of the first 26 communal villages in Gaza, in the aftermath of the floods affecting the Limpopo in 1977. Similarly, the state took advantage of the serious flooding of the Zambezi valley, in 1977-8, to transfer the affected population and promote the creation of new communal villages. For example, in Mutarara district, one of the most heavily affected by floods in this period, in only a few days seven communal villages were created, on the basis of plans brought from Maputo of which the local population was totally ignorant. Despite authoritarian action in the shape of ideological pressure and what was described ambiguously as 'intense persuasive effort' or 'aggressive political mobilisation', the project encountered resolute popular resistance. Clearly, there emerged two conceptions of security in the face of the risk of further floods: the state would impose massive population transfer to higher lands close to roads (thus associating political and administrative control to security from flooding); to ensure their own security, however, communities would only reluctantly leave the river margins where, despite the periodic risk of floods, the soils were much more productive, thus offering much greater security against hunger.

CIPCNAC, the state commission implementing this resettlement, displayed two further tendencies which would come to typify state responses in subsequent years: on the one hand, the increasing centralisation of action (mirroring the working of the state in other areas, particularly, the economy) and, on the other, the search for the development of management capacity in calamity situations rather than their prevention. Rural communities were excluded from decision making here, just as they were in the economic sphere.

The creation in 1980 of a more developed state structure - the CCPCCN, with its executive body, the DPCCN - was part of this 'socialist conception' based on an all-powerful state, which the intensification of war and a severe period of drought in the early 1980s would only reinforce. In this the state sought the capability - here also 'monopolistically' - of protecting society from natural calamities. The intensification of drought and war in the north-central and southern provinces in the following years reinforced this conception by gravely destabilising the rural areas and multiplying the number of highly vulnerable refugees, thus giving space for the DPCCN's action and legitimisation for its uncontrolled growth. The resources were found thanks to increased efforts of the state, which, despite obstacles, maintained social objectives, and to the intensification of international aid.

The adherence of Mozambique to the World Bank and International Monetary Fund, which resulted from economic and social disarticulation affecting the country in the mid 1980s, signified the beginning of a process of profound transformation of the socialist state. It was a process which ended only in the 1990s, when, with the end of the war, the state assumed the clear profile of submission to market dynamics. During the 'grey' period of transition, the DPCCN resisted deeper north-American 'penetration' through CARE International. As the state reluctantly opened up for the humanitarian efforts of international agencies and non-governmental organisations, the DPCCN was transformed into its final shape: an immense and inefficient state body, riddled with corruption and discredited in the eyes of both Mozambicans and the international community.

In the second main period we have defined, after the changes which followed the Peace Accord of 1992 and the institution of formal multi-party democracy, the state took on the markedly neo-liberal role of mediation between market and society. Consequently, the hitherto sustained 'structural vocation' of social protection disappeared. In this period, the evolution of the response to natural calamities was also informed by factors of context and of political option. With regard to the former, there were indeed great changes. In the first place a climate of regional and internal peace and security was established, allowing policies related to emergency aid to victims of natural calamities to be clearly distinguished from the context of war. In the absence of a direct threat, demilitarisation of society went ahead rapidly, and the prime need of the state constantly to reassert its political and administrative control in the country fell away (the struggle for political control took other forms). In the second place, détente in southern Africa enabled the widening of regional and international co-operation, having an impact in meteorological and food security forecasting.

This new dynamic, no doubt very positive, did not imply however a profound change of state attitude toward coping with calamities. In fact, research on forecasting continued to occupy a secondary place on the list of the state's priorities, which were confined more and more to the area of management. Following the prevailing logic of the market, the state focussed its action in management of emergency situations. The INGC, created in 1999 to replace the DPCCN, mentions management in its title but omits any reference to prevention. Unlike forecasting, management enables the collection of much more substantial funding, more direct commercialisation of emergency action (set out in the precepts of the INGC), besides creating the appearance of a more effective reply to calamity situations.

Policies formulated in this new context have not substantially altered the relation between state and communities in the sense of enabling greater participation of the latter. On the contrary, in some ways they have exacerbated even further the marginalisation of local communities, as the previous socialist state, despite everything, undoubtedly had more social purpose than its successor.

After almost thirty years of independence, policies of response to natural calamities continue to limit rural communities to a very marginal role in the process. This is not surprising in that it corresponds to the marginal position they occupy in other sectors of Mozambican life, despite the post-independence popular government having emerged from armed struggle.

We can however indicate several reasons why the situation should be very different in this respect, particularly taking into account the fact that Mozambican rural communities constitute a clear majority of the population (around 80%), and are perchance most exposed to the negative effects of extreme climatic events.

But there are also economic reasons in the sense that, as Mozambique is an eminently agricultural country, rural communities carry out as their main activity the domestic agriculture which produces most of the food crops.

Domestic agriculture is, in general terms, dryland agriculture reliant on burning, and uses the hoe as its main tool in a system of cultivated and fallow fields. Because it depends to a great extent on rainfall and soil fertility, it is an agriculture that maintains a delicate and vital equilibrium with the environment. Its viability depends, by definition, on the close attention its practitioners give to climatic variability.

We can consider local knowledge relating to these climatic variabilities at two levels. Firstly, the integration of a body of knowledge socially transmitted over time and constructed on the basis of observation, repetition and the pattern of phenomena, supporting actions of local prevention. Historically, this knowledge structured criteria for the sitting and building of villages (above all, close to water) and verification - not only of the fertility of soils, but also of the comportment of rivers, of the threat or benefit presented by various types of rainfall, of the hidden signs in, for example, the development of particular plagues of insects or particular types of winds - so as to construct forecasts (forecasting being the most effective form of guaranteeing security).

Upper Zambezian communities are exemplary in the construction and use of this knowledge of resource utilisation and forecasting. The Tawara language, on the right bank of the river, distinguishes with great precision soil types such as those directly irrigated by the rivers (gombe, consequently the most disputed lands), and the interior lands (kunja), occupied when there is no room in the former. The choice of land depends also on knowledge of specific indications of fertility or suitability for particular cultures, such as the presence or otherwise of certain types of trees or grasses. Besides soil suitability, the choice of crops to develop depends also on the always centrally important forecast of success of such cultures particularly in relation to water. Thus, for example, in certain zones considered very suitable for the cultivation of sorghum, in years where drought is foreseen, this may be substituted by millet, which is much less demanding of moisture and has a shorter maturation cycle; or by maize, because experience shows that sorghum is much more vulnerable to certain local predatory birds. (Oliveira, 1976: 3 passim).

Evidently, this store of knowledge is constructed and used in the context of real power relations, in which those possessing it maintain predominance over the majority who are supposed to benefit from it. Apart from 'common' knowledge there is also ritualised knowledge, the construction of which is also historical (through accumulation or adaptation), and whose availability and manipulation depends on specific rules and codes. The Tawara are part of a complex of Shona peoples whose structure of knowledge is based on the belief that, after death, the spirit of particular individuals is embodied in an animal, the most preponderant being that embodied in the lion (mphondoro). The embodied spirit returns to contact the community through a medium, the mvula, whose exclusive status is socially recognised. It is through this medium of the mphondoro that the dominant spirit advises the community on difficult decisions or those of vital importance, particularly where agricultural crops are concerned and in everything connected with water. It is not by chance that the term mvula signifies either spirit medium or rain.

On the north bank of the river, peoples of the Marave complex hold regular propitiatory rain ceremonies at the end of the dry season, in a cosmogenic context where nature, society and the cult form part of an indivisible triad. The ceremonies are performed in small village sanctuaries, and in larger and more central shrines in times of serious drought. Here, the spirit may lie dormant for long periods, appearing through its medium only in situations of crisis caused by natural calamity or by grave offences against social order or against the cult itself (Schoffeleers 1992: 61, 80). Rain is considered a social good that can only be obtained by a (political) territorial chief who, for the purpose, calls on the spirit medium. Thus, according to Schoffeleers (1997: 64-5), rain ceremonies include recognition of the powers of the chief and the dependence of the people on him. Only the chief and no one else can call for rain; if it does not fall, he or one of his subjects is to blame, and only he can remedy the problem.

In this way, a structure of knowledge (and its social organisation) which is considered from the outside to be constituted by magic-religious elements of no material efficiency for real action is, on the contrary, for upper Zambezi communities, absolutely fundamental in their understanding of material reality and the maintenance of social order (Oliveira, 1976: 99). And the significance of this knowledge rests at a level, albeit relatively minimal, of efficacy, whether of objective knowledge gained from experience, or of ritualised knowledge in which the community objectively believes.

It is also important to mention two central characteristics of this body of popular knowledge. One is its extreme diversity from region to region - corresponding, no doubt, to the adaptability resulting from the close attention invested in climatic phenomena which a material culture dependent on agriculture needs in order to reproduce itself. The second is its flexibility and adaptability in the face of change, whether in 'common' or ritualised knowledge.

Without doubt, the ability of these sets of knowledge to foresee and defend local communities from natural calamities is very relative and would have been much greater in the past, when ecological conditions were more favourable, population density was much lower, and when they may have been deployed in their totality.

However, in the last hundred years, various factors have contributed to limit the capacity of response of communities to natural calamities, a large number of them the result of the unequal confrontation of interests between the state and these communities, particularly during the period of submission to the colonial order, but also after independence in 1975.

In respect of the first, during the process of establishing its dominion, the colonial state promoted the political dismemberment of existing communities, removing their state of unity and radically altering relations of power and family ties, which regulated not only the norms of access to and direction of natural resources but also structured all productive activity. In doing so, it determined not only what the communities could no longer be but also what they would have to become, in a dynamic in which dismemberment was accompanied by a double breach for rural communities: on the one hand, in respect of their land (knowing the behaviour of climate means knowing it on the basis of specific land and conditions) and, on the other, in respect of their internal organisational networks, eroding the corpus of community knowledge of the types already mentioned, empirical and ritualised knowledge.

The colonial state exerted this effect in various ways. In the first place, it implemented in a frequently violent way policies of politico-administrative control and of forced rural labour, which implied massive shifts of population, the transfer of entire villages to neighbouring, but also frequently distant and unknown, territories. Cotton concentrations, 'compounds' for workers in the colonial agricultural undertakings, compulsory villages in time of war or even migration to the poor suburbs on the periphery of the cities - all these processes consisted of the transfer and concentration of populations in larger units, which dismembered communities and removed from them their most important strategic weapon in the response to climatic disaster: the relationship to the land, and dispersed settlement.

The concentration of population in larger agglomerations, and located according to criteria external to the communities themselves, resulted in an induced syndrome of over population, meaning that land was more scarce and access to it more difficult and distant (making more uniform, and reducing the size of, cultivation plots), making water supply more difficult, reducing and overstocking the pastures. This phenomenon of concentration also resulted in larger numbers of victims on the passage of a cyclone or when the level of a river rose suddenly.

Still in the realm of vulnerabilities induced by the colonial factor, it is important to consider the various forms of forced labour and cultivation. The first because they systematically removed from peasant families adult males, disturbing the community at its roots, weakening its capacity for foresight in its response to the variabilities of climate. The second accentuated this dynamic in a particularly perverse way, to the extent that, for example, the introduction of obligatory cotton cultivation removed in a very direct way space and time for food crops, thus drastically reducing decentralised food reserves and paving the way for cyclical and chronic famine. In a society still free of the model of accumulation, food reserves were important in guaranteeing security in times of insect plagues and drought. The generalised and structural reduction of these reserves made such society clearly much more vulnerable to natural calamities, particularly the most devastating of all - drought.

The break brought by independence did not alter the process of progressive and critical disarming of rural communities in coping with crisis situations caused by natural calamities. On the contrary, it accentuated the situation by introducing new elements, particularly the development of the new state's agrarian policy, which intensified the resettlement of population (that is, removal and concentration) in an unprecedented manner through its planning and systematic application to the whole country, and the irruption of the civil war.

The new communal villages reproduced many of the difficulties that the colonial strategic hamlets had imposed on rural communities some years before, making them worse as a result of new impositions. Apart from the intensification of concentration, the most important without doubt was the introduction of collective forms of production by the state. Somewhat paradoxically - since they emanated from a state which called itself popular - these forms of production, which had a politico-ideological rather than economic meaning, signified powerful and unequal competition with the already weakened peasant agriculture, in the sense that they occupied the best land (marginalising peasant production to poorer and more distant soils) and sought to withdraw labour from family cultivation for state agricultural enterprises. Thus, the policies of the socialist state raised the levels of concentration of rural settlement (resulting in such concentrations being more vulnerable to floods) and made family agriculture still more difficult (reducing food reserves and thus making such settlements more vulnerable to drought).

Finally, both states - colonial and socialist - fought with equal energy and intolerance the ritualised knowledge of local communities. Because it constituted an obstacle to Christian penetration, the former made it illegal and criminal if there was suspicion of consolidating or enabling forms of resistance to the colonial order, or simply neutralised it through its subsumption into colonialist space reserved for folklore and the exotic. The attitude of the socialist state was based on positivist rationalism, which stigmatised as 'obscurantist' - and therefore strongly criminalised and repressed - all forms of ritual explanation of natural behaviour which were locally based and which fell outside their criteria. This attitude was reinforced by the identification the socialist state made between such knowledge and so-called traditional power, which was regarded with great suspicion owing to its history of close intimacy with the power of the colonial state at the lower levels.

Again, the Tawara communities of the southern bank of the upper Zambezi exemplify how this long process of confrontation between the logic of states and the logic of communities resulted in the critical weakening of the latter. Like many other communities, the Tawara began to migrate at the beginning of the twentieth century, particularly to Southern Rhodesia, as a result of a combination of centrifugal factors within Mozambique, and the attraction exercised by emergent agro-industries and mining in the neighbouring territories. This tendency, intensified from the 1930s by Salazar's New State, transformed the region into a simple labour reserve, in which foreign interests competed with the plantation economy of the colony. Several decades later, exhausted and impoverished, with the beginning of the nationalist struggle for independence, the region acquired high strategic value for its importance as a corridor for the passage of Frelimo guerrillas to the south of the Zambezi (where the Cahora Bassa dam was being constructed) and thus to the centre of Mozambique, as well as a corridor for Zimbabwean guerrillas fighting the colonial order in Rhodesia. As a result the Tawara communities were subjected to an intensive colonial programme of resettlement, according to strategic military precepts of counter insurgency, but also linked to the filling of the shallow lake of the dam.

During the implementation of this programme, there was great tension between the two main colonial agencies. On the one hand, the Office of the Zambezi Plan (Gabinete do Plano do Zambeze - GPZ), concerned with the filling of the Cahora Bassa dam, conducted studies with a view to resettling people whose lands would be submerged - detailed studies which included the identification of fertile areas, with access to water, and also cultural matters. On the other hand, given the proximity of guerrillas, the Tete provincial government and the security and defence forces urgently needed to complete the establishment of defensible villages to impede contacts between guerrillas and populace, and were not interested in the delays which the GPZ studies implied. The second criteria prevailed and the target population - between 15 and 25 thousand inhabitants - was resettled in villages whose location was decided on solely military criteria, a process which imposed the doubly negative effect of concentration and marginalisation, in a context of rapidly spreading intense and internationalised warfare.

The communities formed by the tens of thousands of people dislocated by the filling of the Cahora Bassa dam were not interlocutors in this process, which occurred in 1974 and caused profound ecological alteration in the region. This alteration created an entirely new context, requiring a radical alteration of the body of empirical knowledge. Formerly, the regime of the river was 'alive'; drought and flood followed one another in a pattern of relative regularity which successive generations of riverside communities learnt to recognise and in their own way interpret and forecast. This was now succeeded by a dam and lake with a 'dead' regime, without a regular pattern, in which the variations had no relation to the annual seasons and in which the policy of maintenance of the water level had nothing to do with the inhabitants, but only with the needs of producing electricity. Consequently, there took place the progressive 'unlearning' of behaviour particularly in relation to the small and periodic floods and to the community emergency precautions against unusual floods. This erosion of empirical knowledge was accompanied by a corresponding erosion of ritualised knowledge, no doubt prefigured by Cahora Bassa's submergence of Kanyemba's grave in Malima, seriously affecting the cultic ceremonies and their actual impact in the forecasting of the weather.

Finally, this process of degeneration of the capacity of the community to respond to natural calamities reached its highest point with the intensification of civil war, which spread to practically the whole country from the first half of the 1980s and lasted more than a decade. It resulted in the most gigantic process of demographic dislocation seen in Mozambique, affecting about 40% of the population who became refugees inside the country - in cities or safer regions - or in neighbouring countries, completing the brusque estrangement of communities from their lands. As a result of the destabilisation thus caused, popular knowledge (which does not long survive to removal from context and absence of practice) were even further eroded and rural communities lost almost completely the capacity to act as interlocutors in national action in response to natural calamities.

As a result of a transition process that combined international pressure, the ending of the civil war and also the critique within the former socialist state, the state radically changed in nature and policies, drastically reducing the domains of its intervention (in so far as it privatised large areas of its operations) and abandoning its former social vocation. Now it sought to assume a new vocation of facilitator of the market function, in a word, it assumed a new neo-liberal character. In the countryside, this new order created a situation still far from stable, whose main dynamics are unequal competition for the occupation of land between commercial interests and rural communities, which are attempting to re-establish themselves after the immense upheaval of dislocation caused by the war.

Among the strategies of response to natural calamities, the present state dismantled the DPCCN, an apparatus which, owing to centralisation, had grown beyond the 'manageable' and was not only ineffective and costly but also riddled with corruption. In its place, in 1999, the INGC was created. This was an institution created from a combination of internal consensus and the intervention of external agencies (of the United Nations, bi-lateral co-operation, such as USAID, and also non-governmental organisations).

However, the internal consensus is one obtained within the state itself or sectors close to it. It is a consensus in which there was certainly no space at all for tangible participation of rural communities. As a result, the INGC emerged as a structure which, far from seeking an original strategy for response to natural calamities based on partnership between the state and the various organisations of civil society with rural communities, favours rather the combination of international management techniques in disaster situations - to the detriment of prevention - with the residual elements of the previous strategy, that is, reliance on direction still far too centralised.

Certainly, great steps have been taken in recent years in the development of regional co-operation relevant to systems of prevention, particularly in respect of meteorology and hydrometrics. However, the significance of such advances, translated into very frequent, detailed and reliable forecasting, has been offset by the fragility of the Mozambican meteorological service, which is relatively marginal in the structure of the state apparatus, and consequently lacks personnel and means of participation in these systems.

The difficulties which forecasting has in achieving a central place in strategy shows up the lack of local assimilation of present day perspectives: while forecasting requires relatively high levels of investment, which are planned yet have no direct return in financial terms, the management of disaster situations can be supported by a much easier international (and national - within civil society) recruitment of resources, since their greater visibility arouses humanitarian sentiments with a much stronger sense of urgency.

Recent history has shown consistent increase in intensity and frequency of natural calamities in Mozambique, with a corresponding increase of loss of life and destruction of goods, to the point that it has seriously affected the country's economic, social and cultural growth. There are two main causes of this situation: on the one hand, climatic alterations which are occurring at a global level and particularly in southern Africa; on the other, the marked weakening of the capacity of response of Mozambicans, particularly of those living in the countryside, as a result of the negative effects of colonial and post-colonial policies.

Although originating in states apparently of contrary or radically different orientation (the colonial as opposed to the socialist state, or the latter as against the present neo-liberal state), these policies have a common denominator, based on the principle of extreme centralisation in which the state assesses and seeks to resolve problems, while the majority of the population fulfils the role of spectator and executor of the proposed lines of action.

This centralisation presupposes that the state is itself capable of confronting and resolving problems which, in the case of the response to natural calamities, recent events have shown to be increasingly doubtful. In fact, despite the efforts to improve policies of state response, namely with the creation of the INGC as a result of analysis of the errors of the DPCCN, and the greater effort of co-ordination, the floods of 2000 and 2001 have revealed the serious weaknesses of the state in this area. Indeed, it could not have been otherwise, considering that much more stable and older states, more powerful and better endowed with human and material resources, have also shown themselves incapable of handling catastrophes of these dimensions.

On the contrary, the improvement of capacity for response can only emerge from an entirely new perspective, based on balanced principles of decentralisation (as opposed to the tradition of centralisation) and of prevention (as opposed to the tradition of concentrating on the capacity to react).

To the process of decentralisation the present government programme dedicates two pages (Governo de Moçambique 2000a: 91-3), in which decentralisation appears confused with the construction of the state apparatus at the local level. Yet it is necessary to consider the concept at a wider level, which includes not only the construction of the state, at a local level and on the basis of local forces, but also the involvement of social sectors and forces outside the state.

In the first place, this is because such involvement creates conditions for the strengthening of a perspective of prevention, which does not depend simply on the mobilisation of formal knowledge circulating in official channels, but also relies on knowledge produced outside officialdom, including popular knowledge. This implies that the state creates space enabling itself to integrate the former (produced by its own organs or channelled through regional or international networks) with the results produced in research centres, private universities and other sectors, as well as with popular knowledge. The incorporation of the latter, besides ensuring popular participation in prevention, complements formal academic knowledge which, in the field of forecasting, is far from being infallible or absolute, and thus needs to get away from positivist and exclusivist posture, so as to be able to welcome new perspectives and interpretations.

But there is a further dimension to the importance of popular knowledge. Where it exists, its efficacy is based not only on the capacity to read and decodify empirically the signs of nature and its social transmission but also on the fact that, apart from daily experience, its ritualisation confers on it a very high level of credibility in the eyes of the community, which makes it a powerful element of popular mobilisation for actions of prevention of calamity situations.

A mobilised community is not merely a repository of warnings of imminence of calamity situations, but one which reacts as a subject in contexts requiring prevention, searching out refuges when it knows there are floods, or increasing foodstocks in the face of approaching drought.

The new position that communities should occupy in the strategy for responding to natural calamities implies the establishment of advance warning mechanisms and the management of crisis situations at the local level, including the establishment of local civil defence, controlled locally and co-ordinated vertically with provincial and national authorities. Such mechanisms, including systems for monitoring floods and the most secure escape roots, enable communities to carry out a more active role. They can thus become more effective partners of the state, perceiving with greater clarity warnings of imminent calamities issued from the centre and, conversely, managing the stations installed locally to feed the meteorological forecasting system.

The errors of the past show that population settlement imposed from above according to criteria and interests outside the target communities brings profoundly negative results. Apart from this, removing populations from richer and better watered areas in the name of security in the face of flood situations is impractical. The valley bottoms and other lowlands, economically more productive and supporting concentrations of population, cannot be permanently abandoned owing to the episodic threat of floods. On the other hand, civil engineering construction guaranteeing the security of the population in these areas is also insupportable, owing to the dimensions required. The only solution, therefore, lies in an effective system of forewarning with full local participation.

It is also true that co-ordination does not only mean the exercise of good neighbourliness or interchange of information between departments at the highest level of the state apparatus. It includes, rather, the search for coherent functioning, which means, on the part of the state, the highest efficacy combined with an open and integrative vision. Thus, for example, the definition of agrarian policies and of land settlement must take into account, and seek to counter, the process which, historically, has made communities more vulnerable to calamity situations. Similarly, as the Constitution stipulates that the security of citizens is part of defence policy and the concept of security can legitimately be extended to include all forms of threat - including the non-military -, the mission of the armed services should explicitly include the availability of its medical, logistical and engineering wings for the protection of the people in times of calamity, in co-ordination with the civil defence forces.

The principles of decentralisation and co-ordination constitute a great challenge to the true nature of the existing state. This means that the state will have to abdicate important elements of its power of decision and restrict the range of its intervention, while correspondingly reinforcing the level of local power and intervention.

At present, the neo-liberal state seems to be doing so not in favour of the majority of the population (which does not guarantee tangible returns except, potentially, in higher levels of legitimacy), rather the space it concedes always comes to be occupied by market forces. This means also that the state, although acceding to the elaboration of circumstantial management policies of crisis situations, will with difficulty be disposed to heighten its co-ordination at the level, for example, of altering its agrarian policies so that communities can be protected from commercial interests which expel them from the best soils to marginal areas where they are more vulnerable to calamities.

The strengthening of local power presupposes the search for cohesive partnerships between local forces which, from below, can pressure the higher levels of the state in favour of these changes. It introduces a dimension which is much beyond the question of the search for a new strategy of response to natural calamities. There is perhaps here - in the tension and dialogue between communities, civil society and the basic structures of the state - the beginning of the uphill path to what Sousa Santos (1998: 34) calls the 'the State as new social movement' (Estado-novíssimo-movimento-social), a construct which envisages the building of a state in which the present stress on the elitist accumulation of wealth and the lack of concern for the future will be substituted by preoccupations of solidarity, social welfare and security for all Mozambicans. A security which enables it to foresee with operational planning (to attenuate its effects) and manage with coherence and efficiency the so very often catastrophic results of natural calamities.

Abrahamson, H.; Nilsson, A. (1993), Moçambique em Transição. Um estudo da história de desenvolvimento durante o período 1974-1992. Gothenburg: PADRIGU - Peace and Development Research Institute.

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: University of Bradford, Ph.D thesis.

Borges Coelho, J.P. (1987) «Historical notes on the conservation of soil and natural resources in Mozambique"». SADCC (ed.), History of Soil Conservation in the SADCC Region (Report no. 8). Maseru: SADCC Soil and Water Conservation and Land Utilization Programme.

Borges Coelho, J. P.; Macaringue, P. (1999), From negative peace to positive peace: An historical approach to the role of Mozambique’s armed forces in a changing security context. Bonn: BICC/CCR.

Borges Coelho, J. P., Littlejohn, G. (2000), Mozambique case study (the case of the 1997-98 El Niño). Boulder: UNEP/NCAR.

Brito, Luís de (1991) Le Frelimo et la construction de l’etat national au Mozambique: Le sens de la référence au marxisme (1962-1983). Paris: University of Paris VIII, Ph.D thesis.

Carmo Vaz, Álvaro (1999), «Água, Desenvolvimento e Conflito" (oração de sapiência por ocasião da abertura do ano lectivo de 1997-98)». Direcção Científica (ed), Oração de Sapiência. Maputo: UEM, 37-51.

Casal, A. Y. (1988), «A crise da produção familiar e as aldeias comunais em Moçambique», Revista Internacional de Estudos Africanos, 8/9: 157-191.

Chenje, M.;Johnson, P. (eds) (1994), The State of the Environment in Southern Africa: a Report by the Southern Africa Research and Documentation Centre. Harare: SARDC.

Chenje, M.; Johnson, P. (eds) (1996), Water in Southern Africa. Maseru and Harare: SADC/IUCN/SARDC.

Christie, F.; Hanlon, J (2001), Mozambique and the great flood of 2000. Bloomington, Indiana University Press.

Governo de Moçambique (2000^a), Programa do Governo, 2000-2004. Maputo: s/e, February.

Governo de Moçambique (2000b) Programa de Reconstrução Pós-Emergência. Maputo: República de Moçambique.

Hanlon, Joseph (1991), Mozambique: Who Calls the Shots? London: James Currey.

Isaacman, Allen (1992), «Coercion, Paternalism and the Labour Process: The Mozambican Cotton Regime, 1938-1961», Journal of Southern African Studies, 18 (3).

Medeiros, Eduardo (1991), «Formas de cooperação e ajuda mútua nas comunidades nyungwe de Tete», Arquivo, 10.

Minter, William (1998), Os Contras do Apartheid. As Raízes da Guerra em Angola e Moçambique. Maputo: Arquivo Histórico de Moçambique.

Oliveira, Carlos Ramos de (1976), Os Tauaras do Vale do Zambeze, Lisboa: Junta de Investigações Científicas do Ultramar (Grupo de Missões Científicas do Zambeze).

Raikes, Philip (1988), Modernising Hunger. London: James Currey.

Ratilal, Prakash (1989), Mozambique: Using Aid to End Emergency. New York: UNDP, World Development.

Sachs, Wolfgang (ed) 1995, The Development Dictionary: A Guide to Knowledge as Power. London: Zed Books.

Schoffeleers, J. Matthew (1992), River of Blood. The Genesis of a Martyr Cult in Southern Malawi, c. A.D. 1600. Madison: The University of Wisconsin Press.

Schoffeleers, J. Matthew (1997) Religion and the Dramatisation of Life. Blantyre: CLAIM.

Sousa Santos, B. (1998) «A reinvenção solidária e participativa do Estado». Texto apresentado no Congresso sobre A Sociedade e a Reforma do Estado. São Paulo.

Vellez Grilo (1972), Aspectos Socio-Políticos do Distrito de Tete. Lourenço Marques: Serviços de Centralização e Coordenação de Informações.

Wuyts, Marc (1978), Peasants and rural economy in Mozambique. Maputo: CEA.

Wuyts, Marc (1989), Money and Planning for Socialist Transition: The Mozambican Experience. Aldershot and Brookfield: Gower.