2. COMPARING THE INCOMPARABLE? Placing the issues in a rural development context

The first thing that comes to mind, when reading a comparative study based mainly on fieldwork in Yemen, Mali, and Mozambique, is to ask, is there anything among these very different countries that can be meaningfully compared at all, and, in the affirmative, for what use? The metaphor used in methodological discussions is often that of "comparing apples and oranges". Or can we go to scale, and compare them as "fruit"? I would argue the latter, namely that if the issues are placed in a context of current policy challenges for rural development, we must not shy away from such complex concerns. Rather than the unit of analysis and time frame, what needs to be modified and relaxed is the intent of such a normative output: it cannot pretend to do more than give clues and insights, directions for consideration, reflection on the part of policymakers, and further action-research for validation and consolidation, for made-to-measure solutions. This purpose is refered to in the title of the document as "mapping", which "constitutes a basis for a thorough investigation of a specific research question [...] As such mapping is not neutral, but depends on the particular context. It is more than just description, but stops short of analysis" (Engberg-Pedersen 1998: 95).

Communication is a very important part of implementing the decentralisation process (also discussed in section 3 below), both linguistically (the national language is not normally used at the village level) and contents (the dissimilar awareness outlined above). It has become evident that communication problems, and difficulties concerning consensus-building at community level, will inevitably include disagreement among actors located at different levels of the horizontal networks that are entrusted with the promotion of development^[14].

This is relevant to the present discussion in as far as it will bear much influence in the debate on the envisaged role of traditional structures in decentralisation policy, laying out the parameters and "margins for manoeuvre" officially accorded to such structures. It is the point where, to use J. Ribot’s (1998) words, "the twain shall meet" ("...if ever..."), the interface between community participation and political decentralisation. Both need to be worked on and promoted if this is to happen, but the balance between mobilising existing (possibly "dormant") resources and introducing appropriate new knowledge and technology will depend on what is envisaged in local, village-level development plans and the limitations identified therein.

First and foremost, the participatory evaluation of "indigenous knowledge systems^[15]" would appear a fundamental starting point when analysing local constraints (and opportunities) to the adoption of new technologies, given that these systems have evolved over centuries to adapt to agro-ecological realities, labour conditions, gender roles and socio-cultural circumstances^[16].

It is important to point out that technology transfer does not depend solely on the successful articulation between the two knowledge systems, but that new technologies are evaluated and interpreted locally through indigenous knowledge systems. We must ask the question of whose indigenous knowledge are we looking at, and who will decide how to apply it and what for.

2.1. Modern technocratic versus indigenous knowledge

As different worldviews and knowledge systems collide during this rapprochement, the bottom-up formula of participation will not easily marry the essentially "top-down" framework of decentralisation. It will lead to a confrontation of social actors with a different epistemology, a different way of knowing - particularly when it comes to discussing natural resource management objectives and requirements and their devolution^[17]. The "gap" between what project documents (to take but the most concrete expression of an application of the modern knowledge system of "development") specify as "expected outputs, products, or results", and what has actually taken place, can be measured from a monitoring point of view as an indicator of sustainability - in the sense that this gap may reflect the extent of local appropriation, adaptation, and, eventually, "ownership" and therefore institutional and social sustainability. The dichotomy that often opposes indigenous knowledge to modern scientific knowledge is misleading^[18]; says N. Long: "The encounter between different configurations of knowledge involves a transformation or translation of existing knowledges and a ‘fusion of horizons’"(1992: 274).

The knowledge systems’ institutional settings are described by Usher (1986): "The state system [deriving its legitimacy and authority from the rule of law of the nation-state] rests on a common property concept in which the state assumes exclusive responsibility and capability for managing a resource equally accessible to all citizens. The state manages for certain levels of abundance on a technical basis, and then allocates shares of this abundance to users on an economic and political basis. The system of knowledge is based on a scientific accumulation, organization, and interpretation of data, and management problems are resolved in a technical, ahistorical framework. This system of management is bureaucratic, which is to say, hierarchically organized and vertically compartmentalized. Managers become distinct from harvesters, authority becomes centralized and flows from the top down. The environment is reduced to conceptually discreet components which are managed separately. As these separate management units take on a life of their own, management objectives diverge and become focused on specialized objectives: maximizing fur production, trophy production, or recreational expenditures. Not least, the management of fish and wildlife resources becomes separated from the management of the lands and waters that sustain them" (71).

"The indigenous system [based on self-regulation] rests on communal property arrangements, in which the local harvesting group is responsible for management by consensus. Management and harvesting are conceptually and practically inseparable. Knowledge comes from the experience of every aspect of harvesting itself - travelling, searching, hunting, skinning, butchering and eating. It is accumulated by every individual, and shared intimately and constantly within the household, the family, or whatever is the social unit of production. It is also shared and exchanged within the larger society, and handed down in the form of stories from one generation to the next. In sum, these observations, like those of the state system’s, become coded and organized by a paradigm or a set of paradigms that provide a comprehensive interpretation of them. The knowledge, so produced becomes the cultural heritage of these societies, just as what we call science is part of ours. (...) The indigenous system (...) is therefore intimately linked with their values, ethics and cosmology, which are generally based on an integrated, non-compartmentalized view of the environment" (71^[19]).

By way of the development agencies and the complicity of other partners, the state system penetrates the indigenous system through rural and agricultural development projects. The outcome of a given project is the consequence of a negotiation process, visible and, mostly, invisible, engaging the technical-scientific reasoning of development professionals, with an urban-based educational background in the natural sciences, and the "holistic" reasoning of the project "beneficiaries". Their different worldviews could be succinctly (and simplistically) characterised as anthropocentric versus cosmocentric (without doing justice to ubiquitous syncretic practices) which are anchored in different historical traditions, whereby the former, derived from Judeo-Christian values (but not affirmed as such), have largely dominated the development paradigms, whilst the latter have been classified in the "risks" category of project documents as "cultural" constraints and irrational, exploitative and shortsighted behaviour, implicitly attributed to what are essentially seen as feudal agrarian relations.

2.2. Traditional institutional structures vis-à-vis processes of decentralisation

"The state can designate the administrator and the governor, but our chief comes from our traditions." (From a conversation with a group of villagers in Banga, Tete Province, Tsangano District, Mozambique, January 1999 (Lundin and Alfane 1999)).

The theoretical and practical foundation of traditional authority is generally of a symbolic-religious nature, and is given legitimacy only by the communities themselves (often finally endorsed through bodies such as councils of elders). They are thus "grass-roots institutions" that in effect have to negotiate their power day-by-day, and therefore embody a degree of flexibility that may be extremely useful for the efficient management of natural resources. The physical closeness to their "constituency" allows for the application of a set of rules and norms that will rarely be out of touch with the ecological reality and the management and conservation requirements of the resources in their territory. In West Africa, for example, in the case of negotiating and managing transmigrant cattle corridors, this flexibility is vital if ecological degradation and social conflict are to be avoided. It may also ensure that cultural rights are respected within the complex ethnic mosaic of most rural societies, in Africa including matrilinear as well as patrilinear groups^[20].

In Africa, traditional chiefs act as mediators between a given ethnic group and its environment, and their relationship to the land is determined by the location of the burial places of the ancestors of a given lineage (or tribe). Traditional community leaders are the symbol of an intimate alliance with their territory^[21]. The primary function of such traditional authorities is to ensure peace and harmony in the rural communities within their territory. Thus, a "bad" chief or sheikh would be someone not able to assure this, for example, during celebrations when people consume alcohol and fights may break out. The main tasks of traditional authority revolve around mediating in land conflict resolution and regulating access to land. Solutions are normally reached among the parties involved, often with the mediation of the respective local lineage chief(s) or sheikh. Only when the latter are unable to reach a verdict acceptable to everyone is the traditional chief or the grand sheikh (sheikh daman) approached. He or she is acknowledged to have ultimate knowledge of the customary geographical boundaries and will take a decision, in consultation with his or her counsellors. In performing these services, traditional leaders have obligations and duties without any rights. There is no reward for the work they carry out for the institutions of the government, for organisations or companies, which occupy their time to the detriment of their fields.

To make land claims before the traditional authorities usually takes the form of oral testimony by credible witnesses and state institutions in charge of land administration are rarely approached for conflict resolution^[22], unless one of the parties derives the legitimacy to substantiate their claim from modern legislation rather than customary rights and thus stands to gain from their involvement. There is in fact a broad (and not necessarily consistent) repertoire of norms and laws that the more shrewd and better-informed individuals can draw upon and interpret to their own advantage (discussed in some detail in section 4 below). The resolution of land conflicts, therefore, becomes somewhat unpredictable, and the involvement of traditional authorities does not a priori guarantee that outcomes favour the aggrieved party. Rather, results are determined by a dialectic relationship between traditional norms and individual behaviour, and may be strongly influenced by political considerations. In Mozambique, the vital role of customary institutions in supervising access to land, particularly amidst recent waves of thousands returning after 1992 with the end of the civil war, has been documented (Tanner 1996).

With regard to traditional leadership, the unification of Yemen in 1990 has brought together two very diverse sets of Government policies. In the Republic of Yemen’s administrative and local power structures it is the sheikhs who at village and sub-district level represent both the local population and the state. The government administration is only represented by the district chief, two levels higher up, who exerts authority by confirming or rejecting the choices made at the local level according to customary rules (Mundy 1995). Traditional leadership includes sheikhs, akels (wise men), amins (religious authorities who organise the collection of the zakat Islamic tax), religious leaders (hijra or sada), and, related to water management, mukaddams or almudawels (for details, see Othman and Messer 1999). Although households are directly involved in NRM, traditional leaders assist them in solving community-level problems related to land disputes, the distribution of irrigation water and the practices of the traditional hema rangeland management and conservation system. The influence of such leadership is increasingly re-emerging in southern Yemen.

In rural Mali, the nomination of village chief (chef de village) is for an indefinite period of time and can not be revoked other than in the case of drastic offences against the interests of the village community. In all the villages where fieldwork was undertaken (Mopti and Sikasso regions) the village chieftaincy is inherited, and the chief is named from within the founding family of the village. Like in Yemen, an administrative decision is only taken later to regularise and endorse what is already a de facto situation.

The village chief can be of any age, provided he or she is adult - at the time of fieldwork enquiry, a village chief was active who is only 34 years old, and has held that position for the last eight years. All interaction with other village institutions and partners goes through the village chief or is supervised by him or her. The village chief resolves all conflicts that may arise as the result of the implementation of an administrative decision. In all the research villages in Mopti the village chief is responsible for regulating access to land, however this is not the case in the villages under research in the southwestern part of Sikasso, although the chief is informed of all decisions taken in this respect^[23].

Land conflict is not usually inspired by an intent of appropriation and the notion of private property of land is almost unknown. Given the mode of settlement and the numerous marriage ties, non-autochtonous population groups are not considered «strangers» and enjoy equal rights and duties. The Domain and Land Tenure Code (Code Domanial et Foncier, CDF) states that land on which customary tenure rights are applied has no property value and belongs to the domain of the state^[24], and that the application of customary tenure rights is confirmed as long as the state does not require the land on which these are applied. The code affirms that "customary chiefs who regulate land use on the part of families and individuals according to custom, may in no case use their functions to claim other rights over the soil other than those resulting from their personal use, in conformity with custom"^[25].

Mozambique’s law no. 2/97 is vague towards traditional authorities, and does not specify areas or sectors of collaboration, or the bases for interaction between the administration and traditional authority. But in article 28^[26], para. (2), it makes clear that "the local authority bodies may sound out the opinions and suggestions of the traditional authorities who are recognised as such by the communities, so as to coordinate with them activities which seek to satisfy the specific needs of these communities". Other than for the 33 urban areas that have been municipalised, there is no law, decree or diploma that mentions any institution that could substitute the municipality in "sounding out" the traditional authorities. The situation is even more complicated in the majority of districts, where the extension of local state bodies down to the communities is rare or is absent. In the urban areas, traditional authority is less evident than in rural ones, where institutionally, below the Head of the Post, the physical presence of the state declines and the room for manoeuvre of traditional leaders increases.

Throughout the country the political-administrative figure of the political party secretary persists, a post that was created after independence in order to carry out party political tasks at local level, but also to replace the traditional chief in his role as a link between the population and the state administration. At the community-level, the political secretaries have no legitimacy to look after traditional matters, only the chiefs do. Furthermore, in the perception of rural communities many administrative matters also fall within the domain of the traditional chiefs because of their symbolism, such as in the case of land, and of conflicts without bloodshed.

Throughout the country there persists a coincidence or overlapping between the political secretary and various levels of traditional authority in terms of geographical space and political-administrative functions, which are: mobilising the population for taking part in local development activities; solving minor social disputes that arise in "their" areas; notifying and channelling to the competent institutions problems they cannot solve and those outside their sphere of competence; passing on guidelines from the local administration to "their" population; involving the population in agricultural production; mobilising the population for sanitation work; opening landfills and building latrines; supporting vaccination campaigns; collecting taxes (also a task of régulos); meeting with the population to listen to problems.

The role of the customary leadership as regards responsibilities outside of the conventional traditional sphere is thus undefined, due to this overlap and the gap in the existing laws (Lundin and Alfane 1999). "The big chiefs should understand [the importance] of the role of the traditional authorities... to collaborate with them... it is necessary to better understand and know who is the authority at local level"^[27].

2.3. Social capital as linked to mutual support and conflict resolution mechanisms

Social capital, defined here as the context-specific structural interplay, based on social relations and trust, feelings of sympathy and obligation, of certain socially-constructed political, historical, cultural, and communal (non-economic) characteristics that facilitate a social group’s interaction and collective action for the achievement of shared (economic) objectives. The concept has increasingly attracted the attentions of both academia and development agencies. However most, studies on social capital and local-level institutions have taken a rather reductionist approach and focused particularly on the quality of membership in local organisations, expressed above all as attendance of, and active participation in, meetings. We investigate here one particular aspect of what has been called ‘civil society social capital’ (Collier 1998^[28]) and its interrelation with government in the light of its potential for supporting decentralisation measures - namely traditional leaders and the opportunities and constraints of bringing to bear their intimate knowledge of the local natural and human resource base^[29].

In this social capital definition, the scope of what is meant by "social groups" is flexible and context-specific, but has invariably been tied to NRM. Bearing this in mind, the social units responsible for management tasks have been described in the framework of the traditional affiliations in which they are embedded. The importance of understanding the role of identity and its formation (ethnicity, gender, religion, occupation, etc.) as a basis for collective action among the poor, and as a basis for developing more sustainable local institutions generally, has been pointed out by Webster (1998). Institutional development may in turn be facilitated or inhibited by the social trust and local norms and codes of conduct that traditional leaders patronise, nourish and uphold, and by which a social group identifies itself.

Policies and programmes that directly tackle poverty alleviation must therefore make judicious choices in respect to whether and how to associate with traditional leaders to "tap" the social capital they command. This may imply trade-offs, and the overriding objectives must thus be clear and the programmes designed accordingly^[30]. It is important, on the one hand, not to exclude traditional institutions when bypassing traditional norms which from a development perspective appear unjust towards certain segments of society.

On the other hand, the "fallback options" of the poor must not be endangered: by setting up "modern" institutional arrangements in a project context there is the danger that these remain unsustainable without outside support; in the long run, they may undermine or even deplete part of the social capital stock that constituted an imperfect but reliable safety net for the poor in terms of basic food security.

One of the main arguments brought forward against the larger, more active and "official" involvement of traditional institutions in supra-community affairs is that these are often viewed as endorsing and sanctioning a "neo-patrimonial" style of management, largely inconsistent with the emancipation and equal treatment of women, youth, and non-autochthonous groups. This situation may be worse still, that is, their actions and financial transactions more opaque and less accountable, in cases where traditional authority overlaps with important religious authority and functions (FAO 1997c). The clue here is not so much in terms of the social organisation that traditional leaders preside over, including their courts of counsellors, and the respect they command, but in the fact that their ascribed positions (and spiritual authority) may make them inclined to feel that community resources are equivalent to their own. This leads to an almost "natural" inability to distinguish public from private funds (Bayart 1989), "the king from the crown", and has important implications for the fight against corruption and fraudulence.

Decentralisation brings suspicious practices physically closer to the people who will bear the consequences of those same practices, and villagers in a number of countries have been reported to be angry at what they see as the "decentralisation of corruption". In the long run, if accountable leadership is to become a local reality, this may at times be a transitory requisite for successful socio-political transformation. If there exists such a thing as "negative" social capital^[31], of which the Mafia is possibly the most-cited and well-known example, it must be clear that its root cause is not inherently to be sought with traditional, but with any form of authority and power.

Yemen

In Yemen, for instance, it has been found that due to a mix of historical, cultural and sociopolitical circumstances, the social capital embodied in traditional community leaders should be "tapped" with great care, if at all. The study has highlighted a limited number of cases in which such social capital has indeed found its way into local-level institutions geared towards the NRM objectives of a wider subset of the population in the countryside (Othman and Messer 1999). Some issues in the debate on social capital discussed above, namely, the mutual support networks, the common property regimes of rangelands, the collection and redistribution of the zakat religious tax, and, particularly, the protection of hijra (marginal) groups on the part of the rural élite and national polity in Yemen, must be seen in this light. The successful maintenance of most of these practices depends to a certain extent on traditional leaders. Before these functions can be realistically taken over by the state, rural development initiatives must seek to promote the enfranchisement of poor population groups gradually and incrementally, making choices informed by social research at local level.

Mopti (Mali)

In rural Mali, social cohesion is cherished by mutual support mechanisms (such as, for example, the "pari", which is given in kind or cash; Bâ 1999). Yet it appears threatened, first and foremost by problems related to land tenure - as in Niono, Diéma and Konza (Bertrand 1999). In the Samori area, divergences of interest have led to the competing application on the same tracts of land of both customary rules and modern legislation, resulting in a legitimacy crisis at the interface of the two rule systems. Land pressure is exacerbated by return migrants who try to take advantage of the political campaigns, and challenge traditional authority by accusing the latter of not respecting their customary succession rights because of their long absence. In the implementation of Mali’s national communalisation programme (regrouping villages as communities for administrative purposes within the decentralisation agenda), land tenure security will have to be achieved through the redynamisation of rural (state and customary) institutions. To that end, the preoccupation with transparency and social equity must lead the stakeholders in land tenure to draw from their social capital the consensual bases for community-based NRM.

In the Samori area, this means involving the sinangouya (see also section 2.4.(i) below), and the prevention and management of conflicts, the development of communal land charters, and inter-communal cooperation mechanisms. The "sinangouya" or "cousinship" is the result of social pacts that were concluded in order to consolidate friendship and family ties. The sinangouya embodies tolerance and community spirit through the grouping around common meals during traditional festivities and in visits to the weekly markets^[32]. In Samori, it indicates the range of inter-lineage blood pacts consolidated in dafing culture. At its most extreme, the sinangouya formally prohibits marriages between lineages bound by blood pacts. The sinangouya thus contributes to reinforce intra-lineage cohesion and inter-communal cooperation between Baye, Socoura, and Ouenkoro^[33].

The sinangouya, initially, plays a moderating role by facilitating the mediation between the protagonists of conflicts (between individuals, and lineages) over land or other natural resources. In the event of a dispute, the channel of appeal can be at several levels (the village, or founding village) through the customary institutions (the village chiefs) or resource persons (e.g., the blacksmiths) who can contact the village chief and the wise men. At the (individual) user level, mediation passes through dialogue and the offering of excuses in the presence of the customary authority. In this circumstance, the wise men fix the rules of the mediation by making decisions during general meetings. At the village level, social conflicts are managed within each lineage. Grave offences, for example, fall within the domain of competence of the village council, with the channels of appeal being the resource persons (old men, blacksmiths).

The village chief is the ultimate channel of appeal, as a last resort. The "land institutions" of the founding village are the channels of appeal to manage inter-village conflicts at that level. In case of persistent conflict, the paramount chief is called upon. In the event of further persistence of the conflict, the protagonists may approach the administration (the head of the district, chef d'arrondissement, of Baye) to manage the conflict. Government authorities are only approached for specific types of conflict (theft, divorce, hunting accidents, etc.). The involvement of the Traditional Female Groupings in the management of social conflicts can take place at two levels, depending on the causes and sources of these conflicts. At the household level, three female leaders are selected by the sinangouya from within the affected lineages, to resolve the conflict from within the households.

The sinangouya is also a regulator of inter-community social capital and is always present as a factor acting upon the regulation of land tenure conflicts. The Bankass Initiative conceals unquestionable assets from the point of view of the implementation of the communalisation process. The positive dynamic at the interface between human and natural resources has been reinforced by the maintenance of a framework for dialogue between the customary and modern land institutions, due to the practice of standardising the customary acts of mediation. The latter, signed in turn by the customary chiefs (of the village hamlets and founding villages), the local representatives of the chamber of agriculture, the local administrative authority (the district head), the representatives of the Ministry of Rural Education (Encadrement rural) and NGOs, the head of the administrative circle (Commandant de Cercle), and the legal authority (the judge), indicate the importance of coding of the habits and customs in the process of elaborating the communal land charter. Initiatives such as Bankass, offer an opportunity to codify the customs and norms, for the development of conventions between customary and modern institutions, and to work out the legal status of administrative entities (villages, communities, circles, and regions). They are likely to legitimate endogenous land tenure institutions and to make NRM laws more efficient in Mali, whilst local development agencies aim at reinforcing the institutional capacities of customary authorities in the process of participative development. The consensual bases that they generate (participation-responsabilisation, prevention or management of conflicts, land tenure education, mediation skills) symbolise the socio-cultural values of local population groups, a vision that is all the more justified in the context of the communalisation process.

Sikasso (Mali)

In Sandougoula (Sikasso region), the griots, blacksmiths and Touré (marabouts) regulate almost all conflicts together. After reconciliation, the payment of a lump sum is asked from each party (the amount is left to discretion). If the conflict is solved by the village council, these sums are used to buy symbolic cola nuts for the conciliators. If it is solved at the level of the village chief, they are put in a case and entrusted to the oldest of the council members and will be used in the interest of the village (e.g., for a reception or ceremony). If the conflict is solved at the level of the gouatigui, the sums paid by the parties are put in a case kept by the respective gouatigui - thus the peul population group have their case and the blacksmiths theirs. The final payment of a litigation is the occasion for a celebration proportional to the conflict, and may be a great festival to which all nearby villages are invited.

Men of caste play a significant role in conflict management. A noble person who refuses the reconciliation proposed by a griot (the caste of musicians) is looked down upon. The sinangouya that exists between the peuls and blacksmiths obliges them to accept reconciliations suggested by one another. The one between the Sidibé and Sangaré lineages (including the 3^rd and 4^th sons of the ancestor of the peuls) obliges them to support each other, to reconcile themselves in all circumstances, and prohibits them to keep resentment. This value is taken advantage of in Koussan, where the population includes mainly Sidibé and Sangaré lineages. In the event of major conflicts, a village calls upon another, depending on whether they are of the same clan or maintain particular friendship relations, and the mediation team is chaired by the village’s coordinating gouatigui.

Mozambique

The social capital of communities in Mozambique^[34], mainly in rural ones, is a "capital that originates, grows and develops in a productive system exogenous to the laws of the modern state. When the state is absent, or is kept out, it is this capital that sustains the system of production and reproduction of social groups, and of the communities. It circulates in a stock exchange of social contacts and rests on group loyalties based on kinship, ethnicity, religion or, to a lesser extent, political affiliation"^[35]. The exercise of power of traditional chiefs attending to their personal and to the interests of the community, is based on the process by which (hereditary) political power is constantly legitimised through actions and practices that satisfy community interests, which the community retributes in turn with a counter-provision of services. The latter refer to the respect, prestige and economic assistance (gifts and help in agricultural work, for instance) that are granted to the chief by members of the community, the route by which the customary leader attains a higher status, and receives social standing. Within a system of inter-dependence this constitutes part of a community’s "social capital" in the form of symbolic capital, according to Bourdieu (1977: 171-183).

Vylder (1996) widens Putnam’s (1993) definition to "include processes and structures [...] as different institutions relate to each other, power structures, social norms and networks within and between different organisations... The cultural setting, including the family and kinship structure, the extension of ethnic diversity, the role of customary and traditional laws, define the wider context in which social capital is accumulated (or depreciates). Even the simplest agricultural production system requires a minimum of social capital." For example, "all traditionally agricultural cultures have established strict social norms preventing the theft of agricultural products and cattle; when cattle theft is frequent the social capital of an agricultural community has become severely eroded" (13). Part of the wealth accumulated "trickles down" in the form of benefits to the needy, whether they are relatives (part of the lineage, or a wider kinship unit) or not, following a principle traditional wealth redistribution and mutual assistance.

In rural northern Mozambique, this ensures that the community does not accuse the better-off members of the community of having achieved their economic gains through witchcraft^[36]. An accusation of witchcraft works as a mechanism for maintaining social order, imposing a certain top-down vertical solidarity^[37], within the hierarchies and unequal status positions of rural society. Social cohesion and harmony are a product of the relatively effective and efficient functioning of social, political, economic, educational, judicial and religious institutions. Jointly and in integrated ways, these institutions interrelate to enact in individuals the search for the solutions to their problems in their more immediate surroundings (in the family group, neighbours, or better-off members of the community), by transforming them into social beings. For example, in order to meet basic needs without using money, which is in short supply in most rural communities of Mozambique, individuals (their households and lineages) have established relationships of interdependence that operate on the basis of kinship links, spatial and spiritual identity around which mutual support groups are formed in agriculture and various community undertakings^[38].

2.4. Roles and status: Existing dimensions of traditional leadership

Traditional community leaders have status (legitimacy), based on the lineage (or tribe) as a social unit and on religious grounds, as mediators between a given ethnic group and its natural and social environment. Their prime function is to ensure peace and harmony in the rural communities within their territory. Given the vast heterogeneity of traditional community institutions, which reaches well beyond the simplistic dichotomies taken as parameters within the confines of the definition employed in the present context^[39], the present study focuses on the existing dimensions of leadership in a given rural area, and on the membership dynamic of their attributes. In many traditional agricultural and pastoral societies there is the custom of not passing decisions unless there is an absolute consensus among the local élite that in effect acts as watchman over natural resource management at the village level.

This unanimity rule presupposes important qualities in conflict resolution, mediation, and, especially, in negotiation and conciliation (Swift 1991). These skills will much depend on experience, charisma, prestige and status, and are thus embodied by traditional leaders, and are the very mechanism by which one can aspire to traditional leadership. Although status may be ascribed at birth through the lines of inheritance, it will often be challenged over time, and, unless a traditional leader lives up to expectations, he or she may be relieved of their duty.

In Hadran (in the Directorate of Bani Matar, Sana'a Governorate), Yemen, the grand sheikh (sheikh daman) represents the whole community within official circles, deals with tribal and rangeland disputes at the sub-district level (uzlah), and is responsible for solving problems at the village-level. The rank of sheikh at the local (uzlat) level comes after that of the grand sheikh. The "wise man" (akel), is found at the third level, tackles land and irrigation disputes, and collects the religious tax (zakat) from the farmers. The amin, who prepares the list of farmers for the payment of the zakat for the akel, is sometimes ranked at the fourth level. Other important persons at the community level are the notables (alian), who speak on behalf of their lineage (ruba), attending the meetings arranged by the sheikh or the akel. In Alhami, the akel of the fishermen has a basic role in regulating the dates, quantities and the designation of the areas for fishing. The sheikhs are responsible for supervising the rangeland activities related to hema, and the mukaddam (guardian, usually the head of an extended family) of the mian (water source), who is selected by the farmers, is responsible for distributing irrigation water, and the cleaning and maintenance of the irrigation system.

In the Samori area of Mali (circle of Bankass, Mopti region), there exist customary supra-village resource management institutions that function under the aegis of the land tenure institutions of the founding villages, whereby the lineage chiefs assist the customary land tenure institutions. Being embedded in a structure of age groups, the former decide on the management modes by fixing the codes of conduct (social, economic and environmental) applicable to all users (autochthonous and allogenous). Under the supervision of the village chief, the general meeting of the council of wise men proposes the management rules related to agricultural production and to resource conservation in the reserved (protected) areas.

Three women are delegated to this assembly by the village chief and invited to analyse the debates in order to revert to him or her with modified suggestions. The decisions to be taken are approved by the chief of cults and the village judge. The neighbourhood representatives (the lineage chiefs) inform the members of their respective lineages. Under customary rules, once decisions are taken by the land tenure institutions of the village, the council of wise men designates the youths of the village for the firm application of the management rules to all users of natural resources, without distinction of sex or ethnic group.

At village hamlet (villages-hameaux) level, decisions relating to agricultural activities are taken by the council of users under the aegis of the chief of the village hamlet. They are informed by the village chief of the sanctions and authorisation rules to be applied. Again, three women are delegated to the council by the chief of the village hamlet and invited to analyse the debates to revert to him or her with their suggestions. Through the council, the users inform the members of the household of the decisions taken and of the modalities for the implementation of the agricultural calendar. Junior age groups follow up on the decisions related to breeding and gathering activities.

In social terms, the legitimacy of traditional authority rests with the quality of their management as reflected in the equity of both autochthonous and allogenous users regarding their rights of access to community resources. Customary authority is legitimated through lineage members’ adhesion to the management processes of village-held land and the levels of integration between the lineages in the reinforcement of social cohesion. It restricts the application of resource use rules by deciding on principles of crop rotation - leaving certain plots fallow - and by determining the boundaries within which village-held land can be used. This leads to unused and unclaimed tracts of land between the villages, acting as "buffer zones". Although the land tenure authorities know the limits of their land well, they do not speak about them explicitly in order to avoid conflicts, leaving these fringes unused^[40].

In the Koussan (circle in Yanfolila, Sikasso region) area, all the population practises Islam, of which the imam is the charismatic leader. He closes the debates (in particular at the time of dispute reconciliation) with blessings. As a spiritual guide of the community, the imam plays a significant role in decision-making, because he is well-placed to declare an act licit or not. The imam is named by the council of wise men according to his morality and educational level, and can be autochthonous or not, noble, or a man of caste. He is entitled to gifts in nature or kind, and, moreover, people voluntarily bring part of their zakat (Islamic tax) to him.

As an armed body, the local hunters are assigned the role of protecting the population and their property, the territory and agricultural fields, and women going into the bush (to collect shea tree nuts), and also to transmit urgent messages. In Ouassoulou, these men are called "lanaya mogo", the "persons worthy of confidence", and have by far the greatest capacity of mobilising the local population^[41]. This is also due to the fact that the hunters dispose of the only purely traditional organisation that is structured from the local, village level up to the national level, including also the sub-local, circle (cercle) and district (arrondissement) levels. It is as important to take into account that the hunters are famous for their level of both objective and mystical knowledge. The "mandé mori" or " kontoro" oath reinforces the solidarity among hunters. The decisions taken by the hunters can be questioned only by themselves or with their agreement. Given the continuous reduction of wildlife, the hunters of Koussan have recently decided to prohibit forest access to allogenous hunters.

At the socio-cultural level, customary authority in the Samori area has suffered from a number of invasions (by the dina, the toucouleur, the bambara of Ségou, and the French), which have accelerated their decline. In the Islamised villages, the hogon has come to occupy the functions of the aamirou (the village chief). Following this arrangement each Muslim member of the alamodiou and kaana NRM institutions was replaced by their sons. These have become almost passive observers, given their only limited knowledge of natural resource management. Over the course of time, the alamodiou and kaana have become increasingly similar to the religious worship institutions solicited to bless the exploitation of natural resources, and less effective at protecting natural resources. As for the tons, they have not suffered from as many negative influences owing to the fact that they do not have a religious connotation. However, the colonial administration reduced their functioning by establishing the forest guards to replace them, which left them without a mandate in NRM^[42].

According to Lundin (1998), traditional authority in Mozambique includes: persons who hold local traditional power - the traditional chiefs, lower (lineage) chiefs^[43], and chiefs of social groups^[44]; those who hold spiritual power - spirit mediums and traditional doctors, and, in Islamic areas, the machehe; those with knowledge of how to cure psychological, physical and social ills - traditional doctors; those who know about plants, leaves, roots, fruits, etc. - herbalists; those who know the skills for the physical survival of the agricultural community - traditional midwifes, rain-makers, ironsmiths, clay and straw workers, those who know about animal habits, the movement of water, plant cycles, etc., and those who, to some extent, hold economic power; those who know and can work with social control mechanisms and those who control cultural transmission - chiefs, those officiating rituals, judges, elders, spirit mediums, traditional doctors, and so on. These skills or tasks can, and often do, overlap.

In northwestern Mozambique, the political structure of the vangoni in Banga (Tsangano district, Tete central province), uses ngoni terminology, and includes: The abambo - head of the family unit (banja); the ankhoswe - head of the lineage^[45] (limana); the nyakwawa - head of the lineage territory (mudzi, or group of lineages); the mfumo - head of a group of lineage territories, and of his own banja, limana, and mudzi. Within this territory, called "regulado" by the Portuguese colonial administration, the mfumo was appointed as its head, as régulo (a derogatory term, meaning "little king").

The mfumo is the head of the hierarchically highest lineage which first occupied a larger territory, and from where the occupation of relatively smaller plots by lower lineages took place. In the case of Banga and the mfumo Gimo, local history can be traced back to the conquest of the territory of Angônia by the vangoni. For the case of the vangoni, there exists a nkhoswe wa makossi - a paramount chief - who at present is Mr. Sebastião M. Dahama who lives in Ulongwe in Angónia. This is a title of nobility of a formerly centralised political structure, which has mostly lost the deferences to the great chiefs, but with much symbolism by way of the respect it commands.

Normally, as in the case of Banga, the nyakwawa is a member of the dominant lineage in the lineage territory where the population resides, with a history dating back to the period of the occupation of the territory. The nyakwawa is the uppermost representative of the population. In north-eastern Mozambique, the nyakwawa is the paramount representative of the population groups. It is the basic traditional political authority in the village, responsible for the well-being of the village^[46]. When the territory is divided into two or more neighbourhoods, in reality two or more lineage territories, two azinyakwawa^[47] also exist, one hierarchically above the other^[48].

The ankhoswe is a small "nyakwawa" within the village (aldeia) whose tasks are restricted to his lineage. Delegated by the nyakwawa, he is in charge of all tasks to be carried out in the community. The ankhoswe is the first instance appealed to in case of land conflict, being the person who knows the traditional forms of resolution of these conflicts and the boundaries of the lineage-held lands best.

The social organisation of the populations of the administrative post of Netia is also based on the lineage as a social unit (called nloko), and the lineage chief is called nanloko or humu. Normally, the name of the nloko is the same as that of its chief, and is the name that will be taken on by its offspring. For example, if the humu is called "mweri", his or her nloko will be called nloko na mweri (the lineage of mweri); its members will be called ananloko-a-mweri (the members of the lineage of mweri), or an-anloko-a-humu mweri (the members of the lineage of the humu of mweri)^[49].

Below the lineage level we find the erukulu (or irukulu, in the plural), which in Emakhuwa means "belly"), which refers to the family unit or to the group consisting of all the children of the same mother, a feature of matrilinear social organisation. The father is the head of the erukulu, but his authority over the children is limited by the brother of the mother of the children, because by custom it is he who is responsible for most of the problems of the children of his sister. For the children of the couple it is the atata (or azitata, in the plural). The makhulupale, the elders, of the nloko constitute the council of elders, with the basic role of advising the humu and of deliberating on his decisions. Each elder in the council is generally representative of the members of his erukulu, within which he takes decisions and may decide on minor social quarrels without previously consult the nanloko.

This means that in the nloko some elders have the status of humu (or mahumu, plural) at the level of their erukulu, and therefore play an important role. Other elders hold the right to succeed the nanloko, but may prefer not to and leave the place to their younger brother, while they take on the role of counsellors. This is interesting because it allows for the existence of young and formally educated chiefs, under the tutelage of the elders, who hold more symbolic authority within the community, because of their "proximity" to the ancestors.

Some azitata hold the status of humu. It may happen that, through their chief, a numerous erukulu splits off the founding nloko to constitute a new nloko, in which case a segmentation takes place. However, the force of kinship relations will ensure that the new lineage and its new leader will continue to depend on the humu of the founding lineage. This influence in more than one lineage leads to a situation whereby the status of humu increases or changes to mwene - the chief of the lineage territory (the equivalent of the nyakwawa in Banga). The enthroning ceremony (mukelampa em emakhuwa) is another way of raising the status of the head of the lineage from humu to mwene. It is a moment of learning, qualification and civic education on what from now on will be the mwene’s behaviour and socio-political functions. Each humu or mwene has a female apiyamwene, normally the niece of the chief, with the role of counsellor.

2.4.(i) West African land management institutions: The example of Mali^[50]

Broadly speaking, land management is the task of the lineage chiefs. If land disputes fall within the domain of the sharia under Islamic law, the imam is in charge of resolving the case, and may choose to be assisted or not, and normally presides over the closing (conciliation) ceremonies that sign the resolution of a dispute (Messer 1992, 1993). The customary village chief is the manager of the land within the village territory, according to traditional rules. The chief of cults (chef de culte) is accountable to the founding family of the founding village.

Among the dogon population group, this function is assigned to the hogon, among the dafing to the zora, and among the agro-pastoralist peuls to the maoudo leeynol - they are always nominated among the lineage chiefs that claim descent from the founding ancestor. In their exercise of customary authority, they do not, given their social status, intervene directly in land management, but delegate their functions to each lineage chief, who will manage lineage-held land under supervision of the customary village chief, the village judge, and the council of wise men^[51].

With the upsurge of democratisation, and rice cultivation, new village hamlets are being created in the Sourou valley. In order to turn areas under cultivation into village hamlets, the public administration has approached the traditional authorities of the founding village to carry out exhaustive inquiries. This has resulted in the hamlet-creation process being carried out basically along customary lines, as follows: in respect to customary authority, the claim for settlement of the new user colony is transmitted to the massa. The massa then consults the lineage chiefs with regard to the relevance of the claim and the land tenure regime of the land parcel in question. The district government official, the chef d’arrondissement, is then informed of the designation of the new village hamlet, and informs the new colony of the necessary administrative formalities^[52]. This is done by the massade, who are persons chosen among the lineages of the founding village to carry out the functions of the massa, and their assignments in turn vary from one area to another depending on the degree of Islamisation of the founding village, and the character of their relations (marriage, friendship, and socio-professional ties) with the massa of the founding village. In administrative and juridical terms, they report to both the local government and to the massa of the founding village. Concerning land tenure, the village hamlet disposes only of a temporary management autonomy over the customary lands of the founding village. In village hamlets in which rice cultivation is the dominant agricultural activity, customary loans of land parcels take place rather differently.

The creation of village hamlets in the valley of Sourou is therefore based on a fragile balance between customary authority and that of the state. It can certainly be postulated that customary leadership is relatively more important in the negotiations of land loans to allogenous groups (farmers and pastoralists) for rice cultivation. Traditional authorities usually take a careful stance with regard to the village hamlets’ leaning towards autonomy. As a result of the massive return migration of population groups originally from the Sourou valley, these increasingly consider the creation of village hamlets a kind of deliberate negligence of their succession rights over customary lands, justified by their prolonged absence. This leads to some frustration in relation to the reclaiming of land on loan, or conflicts concerning inter-village boundaries. The flexibility of customary authority mechanisms can do much to prevent these, by regulating access to land under rice. There is thus the mounting tendency among traditonal leaders to insist on an annual renewal of customary land loans.

The land tenure institutions of the village of Libbé-Peul include the massa (the chief of the massake, who, among the dafing, is also the customary village chief) assumes a political-military authority and his advice is sought on all decisions at the village level. His nomination is based on the deontological rules of succession, and he is chosen among the descendants of the massa (the massaden). Eligibility is based on a good knowledge of the lineages’ genealogies in the different villages, governance capacity, and qualities serving to reinforce social cohesion at the lineage, village- and inter-village level. NRM is the task of the massa and the zora, who determine the management and user rules concerning crop harvesting, hunting, fishing, and wood cutting. Their watchman functions (acting as social, economic and environmental police) are delegated to the tons, which are made up of age groups with the task of patrolling the forest area and of applying local land tenure rules. Every ton (or tondéni as they are often called in the Samori area^[53]) nominates its own chief, and there thus exists the chief of the traditional ton-(déni) of the hunters, the chief of the traditional ton(déni) of collective hunting, the chief of the traditional ton(déni) of the wells, the chief of the traditional ton(déni) of the waters of the Sourou valley, and so on.

The alamodiou in the village of Pissa and the hooré gollé in the village of Libbé-Peul supervise the mobility of local users, both in the areas devoted to agricultural production and in the reserved areas. In Libbé-Habbé, Pissa, Baye, and Oula villages, the hooré gollé collaborates with the ton inwe. The supra-territoriality of the alamodiou and the importance of its police functions (social and environmental) equip it with a resistence to an unaccomodating national legislation, and characterises its persistence and sustainability. The tasks confered upon its executive members^[54] allow this institution to serve as an intermediary between natural resource users and land tenure institutions. The richness and smooth functioning of the traditional and customary structures, integrated horizontally and vertically, in these two rural areas of Mali, at the extreme of the country’s national borders and at a distance from the capital Bamako, testify to the necessity and logic of decentralising NRM.

2.5. Traditional institutions’ support to local livelihoods at the state-civil society interface

In lineage-based, hierarchical societies, mechanisms for redistribution are an expression of the moral obligations of the senior members of the community who are responsible for ensuring social and spiritual reproduction, as well as its junior members, responsible for ensuring economic welfare. They are essentially networks of mutual support and solidarity, as well as clientelistic relations among kinship groups of unequal social status, nested in historical power relations and priority rights for some lineages which are largely the consequence of who occupied the land first. The descendants of the first to settle and cultivate the land will claim superior social status, as will those who are members of a lineage who came to dominate others by battle and conquest. This leads to the configuration of local property relations and differentiated rules of access to land and other natural resources, in turn leading to the subordination and poverty of certain groups^[55].

The privileges of one particular lineage and clan, therefore, have their roots in the ancestral domain, legitimised through religious ideology and ritual by means of the symbolic capital embodied in traditional authorities (Bourdieu 1977). Arguments are made in favour of bypassing traditional structures and authority in NRM programme design precisely on the grounds of their not including such population groups adequately into decision-making, as working within these structures would amount to an inadmissible formal sanctioning and "institutionalising" customary male chauvinism.

In Yemen, common property of rangelands may be viewed as a form of mutual assistance (Othman and Messer 1999). Legally, agricultural land is privately owned, and in practice it is managed by the extended family (alakel). Hence, the basic unit for NRM is the household, and the main social unit for rangeland management is the bait - a group of households that constitutes a segment of the tribe. Hijra (marginal) groups, including butchers, barbers, and other poor groups, provide several services to the villagers, especially during the occasions of marriages, deaths and national festivals. In return, these groups are protected by the tribe, and are also exempted from other duties, such as fines (gharamat, used as contributions for the common benefit of the villages), membership fees of charitable organisations (algamiat alkhiriah), and tribal warfare. In Hadran village (Bani Matar Directorate, Sana'a Governorate) in Yemen, decision-making of traditional leaders goes through several steps. First, the imam of the mosque calls a meeting of the leaders and the sheikhs, akels and elders. When these meet (Fridays), the sheikh and the akel present a problem or a case for approval to the community. If there is opposition, a discussion is initiated by the alian, and changes are introduced, until a consensus is reached. Sometimes the members of the community (that is to say, those who are not regarded as traditional leaders) participate in the discussion, especially the more senior villagers, whose opinions are widely respected. Cooperation exists, for example, between traditional leaders (sada and sheikhs) and the farmers in terms of sharecropping (ishrak), and between the akels (who meet and establish rules for fishing), and the fishermen cooperative (which enforces them).

In the study areas of Mali, the relations between farmer and pastoralist leaders are indicative of a high degree of integration of crop and animal production systems. They concern, among other things, the sphere of animal traction, the feeding of draught animals, the contracts regulating the provision of manure, and the custody of agricultural surplus including forestry and livestock products.

Likewise, several cooperation mechanisms are at work in the relations between farmers, blacksmiths and pastoralists (concerning, for example, the fertilisation of land, the security of productive activities, and the lending of inputs), and those between users of land and its customary owners, autochthonous and allogenous. Many of the farmers - both holders of land rights under the customary regime and sharecroppers - have recently turned to rice cultivation and are increasingly engaged in the rearing of draught animals (bulls) and the "conversion" of surplus livestock to draught animals. These animals are given in trust to the transmigrant pastoralists, who engage in barter arrangements with the sedentary villagers (rice against custody and against milk). The pastoralists also enjoy access to some plots for rice cultivation themselves, and lend draught bulls to the dafing rice farmers.

In their role as «economic police», the ton(déni) (see section 2.4.(i) above) in the Mopti region in the northeast of Mali support village solidarity and mutual support networks. Their sphere of action reaches from works of collective interest, in the fields, to the management of the rice (and cereal) storage banks and emergency stock facilities to ensure food security at the village level.

In rural Mozambique, the relationships between traditional and state-appointed local leaders vary across regions, and there is continuing dispute over power and space. The major traditional chief in the area considers himself by tradition the "lord of the land", according to a legitimacy that is related to the symbolic power flowing from the diverse traditional norms of each region. However, with the armed conflict (1984-1992), a lot of land in Mozambique was abandoned when populations were dislocated to other areas within the country or took refuge in the neighbouring countries, and in some cases the collective memory of a given area, in respect to who has customary rights over which land, was "lost". The political secretary considers him- or herself the person sanctioned by political power, claiming to have taken over leadership ever since independence in order to involve the population in development activities, to consolidate national unity and independence.

Generally, problems between individual members of the different lineages (alimana), depending on the issue in question, are discussed in an assembly, the azinkhoswe (lineage chiefs), the azinyakwawa (village chiefs at the aldeia level) and the azimulumuzana (elders) of the lineage in conflict, or else at the level of the community court (coti), where the population of each village is represented.

In rural Mozambique, the forms of interaction between community leaders and development managers vary considerably between regions and local settings, according to the manner in which each institution involves the communities to establish programmes objectives. Since it is not regulated, the relationship between traditional leaderships and the state depends heavily on personal attitudes. In Tsangano district, for example, the head of the Mtengo-wa-Mbalame administrative post interacts with the mfumo Gimo, but complains of the lack of definition at a higher level of governance, which creates problems also at the local level. It would seem as if relationships between traditional leaders and non-government organisations (NGOs) are good - at least in the study districts of Tsangano (Tete Province), Monapo (Nampula Province) and Matutuíne (Maputo Province). For the most part, harmonious interactions can be traced back to the fact that a majority of NGOs have deliberately chosen to associate traditional leaders to their activities. This goes usually beyond using the participation of traditional authority to call for meetings with the local populations, and includes not only a more permanent role as intermediary, but also, beyond that, their being assigned concrete implementing responsibilities in a number of domains upon which NGOs intervene, notably land use planning and common property resource management.

In the communities of Banga and Djavula (Matutuíne district, Maputo province), since the state is de facto "absent", the traditional chief functions to the full extent of his or her customary range of responsibilities. In Banga, there is a chairman of the locality^[56], but not in Djavula. In Netia, a political-administrative chief (the chefe de posto) is present, and there is an overlap in the same geographical space between the traditional chiefs and the political secretaries, and a duplication of tasks both in the rural areas and in the administrative post headquarters. The party secretaries filled a major political space up until mid-1980, since they were linked to the village structure of the communal villages policy, an attempt to group dispersed farmers into centralised and previously demarcated areas, where there would be a school, health units and other state services. With the disintegration of the communal (collective) villages (aldeias), and the return of the population to their former habitat by occupying geographical space according to traditional patterns, the sphere of influence of the political secretaries diminished, or even disappeared, along with their basis of authority^[57]. At the political level, the deputies of the provincial, district and local "People’s Assembly", elected in 1986 in the last general elections held under the one-party system, have as of 1999 still not been formally "demobilised".

2.5.(i) Traditional institutions and gender

In rural Yemen, women do not in fact participate significantly in traditional, male-dominated decision-making mechanisms, although their role in agriculture and natural resource management is considerable, particularly in rangeland-based activities (Othman and Messer 1999), and even though a significant number of households are female-headed (often due to male labour out-migration to Kuwait in the past - before the Gulf War in the early 1990s). Historically, land ownership in many Yemeni communities has been associated with tribal kinship relations, but the socially marginal hijra population groups are not viewed as part of the kinship unit, and are thus not entitled to land ownership under customary rules. Participation in decision-making related to the affairs of the local community in general, and NRM in particular, depends on individual and group relations to agricultural land ownership. The landless do not usually participate in the decisions concerning natural resource management. This does not mean that members of marginal groups can not own land at all, but that they can do so only on as individuals, not as groups.

In rural Mali, women usually designate three female leaders to represent the interests of all women during meetings. Under certain circumstances, the village chief delegates one of his counsellors, to join them in order to analyse together the meeting agenda, as no amendments can be made to decisions already taken. Women are increasingly represented within village associations and play an ever more important role in decision-making during general assemblies. In Ouassoulou village there are secret associations among widowed women called "mousso koma", one of the protective institutions of the people and their property, and its help is appealed to whenever a danger threatens the community: caterpillars, drought, epidemics, etc. The mousso koma exert their functions on a voluntary basis, but the gouatigui often present gifts to their members. In northeastern Mali, in the Mopti region, there exist village-based women’s associations called ganodins, which focus on mutual aid and solidarity and include both married and widowed women. The traditional female groups, GFT (groupements féminins traditionnels) of the rural commune of Baye (circle of Bankass), are centred around ritual, marriage and friendship ties. In Seeno they serve, although rather discreetly, as a vehicle of expression for women’s opinions regarding NRM. The GFTs contribute to fund the village treasuries, and to fight crime^[58].

The matrilinear areas in northern Mozambique present a picture whereby women are more integrated into community affairs - even though in many patrilinear areas women are dynamic and increasingly important actors in the informal market in the cities including Maputo, the national capital^[59]. Matrilinear descent is not the same as matriarchy, but the matrilocal factor among the amakhuwa, for example, places women in a more favourable position in society than does the patrilocal nature of the patrilinear societies in the south and centre of the country. It is important to note in addition that the greater or lesser participation by women in domestic decision-making is also frequently conditioned by the internal hierarchy of an age group, and the consequently lower status of recently married women compared to the high social status confered by motherhood of adult children.

In the matrilinear areas of Banga and Netia (in the Monapo district of Nampula province), one notes the important role played by women in the households, particularly in the management of agricultural production with the decision-making power they hold regarding the amounts to offer and/or to sell, and regarding the granaries, which are linked to the responsibility women have to prepare food for the children throughout the year, an activity that presupposes the availability of basic foodstuffs. But there is inequality between women and men even in matrilinear areas, for example during the rites of widowhood.

Within the largely traditional role of women in Netia, that of the apiyamwene of the lineage deserves to be highlighted. She is a socio-political leader of prominence and always by the side of the lineage chief (the humu), in her quality of "mother of mothers", whereby she is the symbol of the origin of the family group. As "mother of mothers", the apiyamwene is the person whom other women of the lineage consult and approach with their concerns. She is the intermediary through which others from the lineage, particularly its younger members and other women, pass their concerns on to the lineage chief, including, if applicable, criticisms toward the improper behaviour of the chief. Because of her role and social status, the apiyamwene is always invited to provide feedback in the meetings between the chiefs and the council of elders, where the decisions that affect the lineage are taken.