Travelling Models of Participation: Global ideas and local translations of water management in Namibia

In recent decades, water management in Namibia has profoundly changed. Beginning in the 1990s the Namibian state has incrementally turned ownership of and the responsibility for its rural water supply to local user groups. While the state withdrew from managing resources directly, it continued to circumscribe the ways in which local communities should govern them. In so doing, a “new commons” was created. Inclusive participation became the leitmotif of the new management scheme and in particular the participation of women was a major political and societal goal. In this article, we use the notion of travelling models as a theoretical guide to explore how the idea of participation emerged in international development discourses and how it was then translated through national legislation into the local context. The results of the analysis show that more than 20 years after the formulation of international conventions the average participation of women in local water committees remains low. However, older women do manage the funds associated with water and thus occupy one of the most important functions. Our explanation takes the wider social and cultural field into account and shows that gender and generational roles provide elder women with autonomy and authority which prepare their ways into these new official roles. We conclude by considering whether and how the travelling model of participation has been changing local social structures in general and the role of elder women in particular.


Introduction
During the past century, water management in arid Namibia has profoundly changed. Political processes and paradigm shifts have shaped the availability of water and altered the life of pastoral communities a number of times. Until the beginning of the twentieth century, most pastoralists obtained water through natural springs, surface water, and hand dug wells. These conditions changed under German colonial occupation, which assigned large areas of land exclusively to German Siedler (settlers). During Apartheid and in line with racist segregation policy the South West Africa administration expanded the settlements of farmers with European descent and began to establish so-called homelands for indigenous groups (Werner 1993). To make those lands habitable, hundreds of boreholes were drilled making available pastures that had only rarely been used in the past (Bollig 2013). 1 Until the 1990s access to these pastures and their wells was ultimately governed by the South West Africa administration under the jurisdiction of the colonial South African state. Boreholes and land were used communally for small-scale pastoral production and household consumption. The infrastructure was owned and maintained by the state while the local administration was largely in the hands of local political authorities, including big-men, headmen, and chiefs. In northwestern Namibia these amounted to a relatively few, elderly and wealthier men .
After Independence (1990), the Namibian state initiated major political changes including a reform of its water supply management. The reform aimed at overcoming the structural inequalities of the past, to democratize water management and to foster social and economic development (Namibia 2004;Falk et al. 2009). Even though drafted immediately after Independence, the legislation is not a response to national history alone. Equally, it is embedded into the larger global development paradigm of the time.
Strongly accelerated by the 1992 Rio Earth Summit, ideas about the appropriate governance of natural resources have significantly changed. It became a central theme that ownership can ensure sustainability and eventually provide an alternative pathway to economic development in the so-called Global South. Participation became a central means to ensure local ownership and the responsibilities for the management of natural resources were frequently handed over to local user groups (Jones and Murphree 2001;Fabricius et al. 2013). In particular, the participation of women became a major political and societal goal. Their participation is thought to enhance the access to vital resources for previously cut-off groups, to transform unequal power relations, and to help overcome poverty more generally (Schreiner et al. 2004;Singh 2008;Masanyiwa et al. 2014). These understandings cumulate in concepts like community-based natural resource management (CBNRM) for which the World Bank summarizes that the "active participation of stakeholders in natural resource decision making and use increases economic and environmental benefits" (Worldbank 2003).
Beginning in the 1990s the implementation of CBNRM led to a drastic reconfiguration of the organizational and institutional landscapes around rural water points (Barnes et al. 2002;Davis 2008;Falk et al. 2009;Silva and Mosimane 2013;Bollig and Menestrey Schwieger 2014;Menestrey Schwieger 2015;Schnegg 2016a). A shift toward self-governance meant turning ownership and responsibility of the borehole infrastructure over to local user groups. This implied finding alternative ways to govern and novel strategies for decision-making surrounding the shared resources. In line with the above mentioned CBNRM principles, the Namibian state insisted on policy guidelines that include (1) the notion of fixed and bounded user groups, (2) the establishment of democratically elected, clearly mandated, and accountable committees, and (3) the specification that the participation of women should be strengthened. Taken together, developments at different scales have contributed to establishing "new" common property regimes in Namibia. Typically, the "new commons" are not managed by local users alone, but with an active engagement of NGOs, hired for implementing the state's resource management programs (Agrawal 2003, 245). In this article, we explore how globally circulating models about community-based natural resource management created water as a new common good in Namibia. By analyzing the emergent structures of water point committees we show how scientific and political models of participation interact with local understandings to produce specific social forms. Moreover, we demonstrate how these patterns of participation can be explained and whether and how these travelling models are changing local social structures in general and the role of elder women in particular.
To understand how local water management is transformed we use the concept of travelling models as a theoretical guide (Tait and Jensen 2007;Rottenburg 2009;Peck and Theodore 2010;Parnreiter 2011;Peck 2011;Steur 2011;Behrends et al. 2014;Weisser et al. 2014). A travelling model is defined as an abstraction of the social world (model) that is transferred, often through a political process, from one cultural context to another. Many similar, dominant models (including medicine, justice, and development) were initially built in the so-called Global North and involve the application of scientific knowledge. In the national and local contexts where these models are negotiated and applied, they meet existing understandings of the world. Here, a translation process begins in which the model, as an abstraction of the world, is denied, added to, altered, changed, and/or extended. This process involves myriad actors who bargain about the usefulness and truthfulness of the models' assumptions, its interpretations, and its appropriate uses. A travelling model entails an interpretation of the social world, but it also imparts an agency to change the world (Behrends et al. 2014).

Ethnographic context
In the rural Kunene region of Namibia, pastoralism is the dominant subsistence strategy and almost all households own some cattle and small stock. The pastoral economy is constrained by the environment, most notably, a low and unpredictable annual precipitation (Bollig 2006;Bollig et al. 2013). Throughout the research area, cattle, goats, and sheep are the heaviest water consumers. Naturally occurring surface water is extremely scarce and this scarcity structured pastoral mobility patterns until the 1960s (Bollig 2006). During apartheid, the hydrological perforation of the landscape drastically changed its usages. Between 1960 and 1990 the number of drilled boreholes increased almost by a factor of ten (Bollig 2013, 323). The technological infrastructure surrounding these boreholes is heterogeneous: windmills, hand pumps, electronic motors powered by solar panels, and diesel engines. During colonial times up until the 1990s the entire infrastructure was owned and maintained by the South West Africa administration under the jurisdiction of the colonial South African state. Additionally, the state provided diesel for those boreholes that operated with engines. Since the colonial state covered the costs for running and maintaining the infrastructure little local coordination was required. Water was basically free and there was neither a need for establishing strong working institutions at the community level nor to define complex rules of water management. This situation drastically changed in the 1990s with a new water policy turning water into a common good (Namibia 2000). Figure 1 shows the three research areas in northwestern Namibia. The communal settlements in southern Kunene, located near Fransfontein, are inhabited by members of various ethnic groups, mostly by Damara, Nama, and Ovaherero. Under South African colonial rule, the area was part of the so-called Damaraland. In comparison with the other research areas, arable land is relatively scarce. High population densities and the resulting inability to live from the land alone were integral to apartheid politics and forced the integration of people into the colonial labour market. Today, some of the largest herds are owned by absent farmers. These part-time pastoralists live in urban centres and own businesses or are employed by the government (Greiner 2011;Pauli 2011;Schnegg et al. 2013).
The communal settlements in the central Kunene region are located near Otwani and are mainly inhabited by Ovaherero and Ovahimba. Under South African colonial rule, the area was part of the so-called Kaokoland. Households mainly rely on livestock selling as an income strategy. Furthermore, state pensions form a regular cash income for many families. While wealthy households are often headed by elderly, polygamous men, female-headed households and households of young men typically own much less livestock. Young men, especially the sons and nephews of a household head, manage the livestock on a daily basis.
In the northern most region around Okangwati, pastoralism is the dominant economic activity. Equally part of the former Kaokoland, its integration into the South African colonial system was less extensive. Most of the inhabitants of this part of Kunene consider themselves as Ovahimba. They are only marginally linked to labour and sales markets. However, currently the involvement of the state and more importantly of NGOs is highest in this area. Its remoteness and recognized potential for ecotourism explains part of this interest.
Across all communities, the social organization of everyday life is largely determined by kinship. Even though the particular kinship systems differ between communities in the research area, their social organizations show important parallels (Wilmsen 1989). First of all, the Iroquois terminology is common to all communities and defines a wide-spanning kinship network. Second, across all communities the lineage and inheritance systems include women. Even though they are less wealthy than men, women own and inherit cattle, the most important productive resource. Third, among all three communities, women and especially elder women are typically in charge of preparing and providing food and of managing the related finances. Fourth, women are relatively autonomous. In particular, elder women are often respected for their productivity and 20.7% of households are female-headed. Nevertheless, with the exception of few cases in southern Kunene, public debates and community politics are largely dominated by men.
Apart from gender and kinship, age is another important social category which structures daily routines. Elderly persons are met with respect and have the authority to demand help and resources from younger kith and kin. They are the first to talk in public meetings and generally open and close community gatherings with a prayer and advice. Moreover, they provide financial stability to many households. The Namibian state provides an old-age pension to all citizens above the age of 60. At the same time, the elderly have less formal education than the youth. In Namibia, school education was and still is unevenly accessible. Even though schooling is compulsory today, many children still do not attend school. Among the 103 children who live in the seven communities we studied in depth and who are between 8 and 12 years old (the ages when children usually attend primary school), 71% reportedly attended school on a regular basis.  (Schnegg) respectively and are responsible for the overall design and comparative analysis of the data. In the first phase of the current fieldwork, three anthropologists (D. Menestrey, Th. Linke, K. Gradt) stayed for roughly one year between 2010 and 2011 in the southern (Fransfontein), central (Otwani), and northern (Okangwati) parts of the research area to get an in-depth understanding of processes entailed in negotiating and crafting new institutions through daily routines. The qualitative data presented below were collected as part of ethnographic fieldwork and stem from all three research sites.

Materials and methods
After an initial analysis of the ethnographic data collected during 2010 and 2011 our research team returned to the field in late 2012 to conduct the "upscaling" questionnaire we had designed to study the distribution of some of the phenomena found in the community ethnographies. For geographical areas of approximately 250 km 2 stretching around our previously researched locales in Fransfontein, Otwani, and Okangwati, we concentrically sampled 64 communities. Members of all communities sampled were interviewed.
The research protocol contained three different sections. First, rules of water management and the composition of community-based organizational structures for water governance were elicited from committee representatives and the wider community. Each interview took place in public and included both female and male informants, some of whom were active in the water point committee. In most cases, it took a day to finish the interviews in any one community, typically constituted by 5-15 households. In total, we visited 64 communities and all are included in the final analysis. In this article we present information on whether a community-based committee exists for water management and, if so, who serves on it in different roles.

Global ideas
In the beginning of the 1990s, the global discourse shifted away from large-scale to community-based development. This rethinking was justified both scientifically and politically. Scientifically, the late Nobel laureate Elinor Ostrom had demonstrated that the state and privatization are not the only solutions to the 'tragedy of the commons' (Hardin 1968). Comparing 14 cases she showed that many societies have developed complex rules to overcome freeriding and to manage local resources in the long-run. More importantly she defined factors that make institutional success more likely than not which became known as the 'eight design principles' (Ostrom 1990;Ostrom 2009;Cox et al. 2010). Two of these principles relate explicitly to participation. The first of those principles establishes the importance of well-defined boundaries around a community of users and around the resource. The third principle summarizes that groups are more successful in crafting institutions if "most individuals affected by the operational rules can participate in modifying the operational rules" (Ostrom 1990, 90). Although Ostrom has repeatedly called for caution in generally applying these principles to specific cases, the proposed explanations have been taken as solutions to craft institutions of resource governance in the global South (Ostrom 2005(Ostrom , 2009Saunders 2010). In Namibia's water management manuals, all eight principles find corresponding regulations (Linke 2015;Schnegg and Linke 2015;Schnegg 2016a;. At about the same time, neo-liberal ideas proclaiming decentralization and cost-recovery were pushed by the International Monetary Fund (IMF) and the World Bank and became part of the guiding global development agenda (Taylor 2007;Falk et al. 2009;Furlong 2010). Both trends met in a growing concern about the "degradation of nature" and led to the development of concepts like community-based natural resource management (CBNRM) that contain clear premises about the nature of resources and their sustainable uses (Adams and Hulme 2001;Agrawal 2003;Saunders 2010). Among those premises, "participation" became a salient vehicle to foster sustainability and to eradicate poverty (Cleaver 2001;Blaikie 2006;Reed 2008).
Two conferences in Dublin and Rio were decisive in transforming these discourses from an intellectual agenda to policies. In 1992, more than 500 experts met in Dublin for the International Conference on Water and the Environment (ICWE) to discuss future uses of water on the globe. The Dublin conference devel-oped four major principles for water management that are still widely applied. Among the four themes they postulate: No. 2 -Water development and management should be based on a participatory approach, involving users, planners, and policy-makers at all levels.
No. 3 -Women play a central part in the provision, management, and safeguarding of water (ICWE 1992).
Even more importantly on a global scale, later in the same year "The Earth Summit" formulated 27 principles for environmental management centering on the ideas of sustainability, poverty eradication, and participation (ICWE 1992;UNDSD 1992;Little 1995). Two of them deal with participation and gender in particular: Principle 10: Environmental issues are best handled with participation of all concerned citizens, at the relevant level (…).
Principle 20: Women have a vital role in environmental management and development. Their full participation is therefore essential to achieve sustainable development. (UNDSD 1992) Both documents establish the salience of "participation", and establish the link to gender in natural resource management. In the following decades, few policies ignored those trends. Namibian delegations took part in the processes and reflected that: "Namibia forms part of the international consensus surrounding these core principles. The Namibian Government regards the 'Dublin principles' as an appropriate basis for the development of national policy" (Namibia 2000, 15).

National politics
Soon after the "The Earth Summit," the Namibian state started to translate the model of participation to the national context. It founded a committee, WASP (Namibia Water Supply and Sanitation Policy), to draft general principles for water supply management. In its final report the committee formulates that "the maximum involvement of the users" should be thought to "promote community based social development taking especially the role of women into account" (Namibia 1993). To implement these political guidelines locally, the responsible Ministry of Agriculture, Water, and Forestry (MAWF) developed a "Water Point Committee Skills Training" handbook to assist extension officers who in turn advise pastoral communities when negotiating the turnover of the existing infrastructure and new institutional regimes. In terms of participation, the guide reads: The fact that in a community some members have more status and power than others, and may be considered more prominent members in the community, does not give them more rights than any other member of the WPA (water point association). In the WPA all members are equally important. (Namibia 2006, 104) Gender issues are treated in an equally specific way: By being actively involved in the management affairs of the water point, women and marginalized groups can become more incorporated in the community activities and decision-making. (Namibia 2006, 7) To exclude women from voting means to exclude half of the community, and will reduce women's commitment and involvement in water point management (Namibia 2006, 104).
The documents underline that gender figured prominently when the Namibian state translated global blueprints for designing new organizational forms. Moreover, "marginalized groups" are the second social category explicitly mentioned. Shortly after overcoming colonial rule this reflects a much broader concern.
The entire handover process spans years and includes three formal steps: Firstly, a phase of strengthening the extension service, sensitizing and mobilizing the communities in community-based water management (CBWM) related matters, and updating the technical infrastructure, which includes fencing the water point and providing clean drinking water. Secondly, the rehabilitation of water points selected for the implementation of CBWM, establishment of an organizational structure, and "training" of local communities to develop and formalize rules of water governance. And thirdly, transfer of ownership of government assets to the community. The entire responsibility for all aspects of managing the water infrastructure is turned over to the community including the replacement of major technical equipment.
The bureaucratic practice of the decentralization process was accompanied by extension officers from the ministry and/or NGOs contracted by the government. They visit the farming communities in their air-conditioned 4×4 cars and typically stay for a day or two. Those trips are highly valued amongst the NGO staff, as they receive what they refer to as a "bush allowance." They organize workshops in which the community is prepared for the handover process. As in many comparable cases, these workshops are a decisive social technology to implement national policies. The process is characterized by stark power inequalities between those who organize it and those who attend, typically symbolized and expressed in gestures and styles. Many of the regulations negotiated in these meetings are pre-drafted, or as one extension officer puts it: "From the government's side, we are having a specific module drafted constitution. But, there are some places, where the community is changing [the module] according to their wish. Then we go back to the offices and type the documents, before we read them again with the community, so that they have to agree, because the chairperson has to sign on those documents." 2 During the second phase, both the organizational and the institutional arrangements are fixed. At the organizational level, two bodies are established. The larger of the two is the water point association (WPA) that usually includes all adult individuals who dwell in a community and want to use the public water point. The WPA chooses a governance board, the water point committee (WPC) that is responsible for managing daily concerns. The WPC includes a number of positions with specific responsibilities: (1) The chairperson, as the one who provides overall direction to meetings and who monitors and supervises the work of the WPA and the WPC, (2) the secretary, who organizes meetings, draws up agendas and takes minutes, (3) the treasurer who collects contributions and "keeps the money safe," and (4) the caretaker, who operates and maintains the equipment of the water point (Namibia 2006, 20). Taken together we begin to see, how a global blueprint of participation travels in space and time to introduce a new way of seeing water and sustainable ways for using it. Community workshops are an essential social tool to communicate modes of governance and at the same time, offer a place for resisting and translating them.

Local translations
In rural communities, this implementation started in the mid-1990s and continues today. In the beginning of this process, the water point committee is set up as a central organizational body where local water users can participate in establishing rules to manage water. Our research revealed that in 2012 fifty-four (84%) of the 64 communities we studied had an active WPC. Active means, that the members meet on a regular basis to discuss and solve water related problems. In theory, all positions are elected. However, the election process is much better described as a negotiation and consensus-seeking effort to determine who may serve in which office. The following analysis traces the details of the committee composition that emerged and asks how the models entailed started to shape and change the social world. Table 1 gives the gender distribution across the four recognized WPC roles. 3 The table shows a clear pattern. Two of the four roles are male dominated. These are the head of the organization, or the chairperson, and the caretaker. In contrast, the gender difference diminishes for the secretary, a position filled to 36.8% with females. The ratio completely turns for the treasurer. Significantly more females than males are elected to keep the money safe. Over all roles, males dominate and only 26.9% of the positions are filled with women. This is less than was intended by the state which encourages "a quota for male and female members of the WPC" at close to fifty-fifty (Namibia 2006, 105).
As discussed, age is a second, salient category in the regional cultural context. Table 2 gives the age distribution of those in office. On average, people are 40.8 years old. Among the 899 people reported in the census interviews, 412 were 18 years of age and above and thus qualify for positions in the WPC. Their mean age is 38.1 (Stdv. 18.01) and thus about the same as that of most office holders. Between the four positions, we find again a clear patterning supported by an ANOVA analysis. The difference between the positions is considerable (eta=0.378) and statistically highly significant (p=0.000). We find both the chairperson and the treasurer to be significantly older than the caretaker and especially the secretary.
The organizational forms introduced are strongly linked to rationalities of organizing and accounting and go hand in hand with formal education. As we mentioned before, education is unevenly distributed and generally younger generations are more highly qualified. Since schooling is often discontinuous, many people could not give schooling years. Asked about schooling, many simply replied that they "know to read and write." We thus distinguish in Table 3 only between people who have a basic formal school education and thus literacy and those who do not. Again, we can compare those numbers to the overall average. In the entire population among those who are above 18 years of age, 61.4% reported knowing how to read and write. Thus, and not surprisingly, we find a slight bias toward more formal education among the committee members. Within the committee, there are again stark differences. For good reason, all secretaries have formal education. Interestingly, the pattern is highly homogeneous among the other office holders. Most surprisingly at first glance, being a treasurer does not require literacy skills. As we show below, the elderly women in this role qualify through other means.

Discussion
Beginning in the mid-1990s, the globally circulating development model of participation was translated into Namibia's pastoral communities. As a result, new  organizational structures with specific patterns of office holders emerged. These patterns diverge from the ideal of (gender) equality laid down in national legislation and training manuals. The following discussion uses the social and cultural context to explain the translation process at the local scale. We begin with the simplest tasks: caretaker and secretary. The caretaker is in charge of pumping the water and arranging or undertaking necessary repairs. These tasks not only require technical knowledge but also some physical strength. Comparable activities, including vaccinating and castrating cattle, are commonly performed by younger men. Hence, the choice of caretakers adopts a general cultural model about the division of pastoral labor that is organized along the lines of gender and age. Given the fact that formal education is much more common among the young and required to become secretary, it is not a surprise how these positions are filled.
Chairperson and treasurer are the most influential roles. Both office holders are older, indicating that they require some amount of authority and respect. The chairperson provides direction during meetings and is responsible for overall coordination and for handling conflicts (Namibia 2006, 6-7). In the larger cultural context, elderly men usually dominate the public domain and are often active as mediators when conflicts arise. Interestingly, education plays a less significant role in committee leadership than we expected and both roles include a significant number of people who cannot read and write. This observation is especially interesting for the treasurer and may hint at a specific form of institutional bricolage (Cleaver 2012;De Koning and Cleaver 2015). While formal education facilitates collecting fees and keeping books, it is not a necessary precondition for becoming a treasurer. As we show in detail below, being trusted and relatively autonomous often substitutes for accounting skills. In general, the role of the treasurer is the most interesting and the characteristics of those who fill the role is least easily explained.
For Sub-Saharan Africa few studies of "gendered participation" exist that would allow a one-by-one comparison. As Masanyiwa et al. (2014) have pointed out, most analyses focus on political institutions at the national and regional level and rarely reach into the local domain (Masanyiwa et al. 2014, 96). These authors contribute to overcoming this gap by examining participation in water and health services in rural Tanzania. Comparing 10 villages they conclude that although reforms created some space for women to participate, e.g. in local meetings, women rarely speak up and thus have low influence (Masanyiwa et al. 2014, 118). On an average 34% of the positions in village councils were filled with women while very few women had leadership positions allowing them to shape decision making processes. Most importantly, patriarchy, household responsibilities, complicated election procedures, and assumed personal qualities of women hinder female participation (Masanyiwa et al. 2014, 110-114). Studies from other world regions point in a similar direction: women's participation in community management projects is generally much lower than intended by national reforms. The factors explaining this include a lack of experience in a specific field, low levels of education, specific rules that exclude women, and institutional factors ranging from norms and values to local social and political structures (Meinzen-Dick and Zwarteveen 1998;Agarwal 2001;Zwarteveen and Meinzen-Dick 2001;Schreiner et al. 2004;Singh 2008, 936). For example, Meinzen-Dick and Zwarteveen show that female participation in Asian water users' associations is much lower and ranges between zero and 15%. They identified explicit formal criteria that excluded women, e.g. membership restricted to the head-of-household who is often male. In addition, informal regulations including the association of women with the family and the house and recruiting of positions through male networks have a decisive effect comparable to some of the processes observed in Namibia (Meinzen-Dick and Zwarteveen 1998, 340).
Dealing with a different resource domain, Agarwal's study of community forests in India and Nepal shows that national decentralization programs initiated in the early 1990s failed to implement equal gender participation. Even on the most general level of nominal participation, which would correspond to WPA membership, less than 10% of women were active (Agarwal 2001(Agarwal , 1626. At the level of representation and decision making, the number decreases significantly. On the one side a rule, that only one member can represent a household makes it difficult for women to enter the debates. On the other side public roles are largely withheld from women. Theses causes partly overlap and partly contrast with the case presented here. As shown above, public and central leadership roles (chairperson) are typically held by men; nevertheless, women outnumber men as treasurers, a position which is fundamental for water supply management and does require some leadership skills. Complementary to the literature cited above we do not examine factors that hinder participation but explore, when and why women participate.
By way of example, we focus on Grootvlakte; a typical farming community numbering about 10 households. The houses, most of them built from mopane sticks (Colophospermum mopane) and plastered with a mix of cow-dung and soil, are grouped around the water point. A diesel engine pumps groundwater and regular contributions for water are necessary, most importantly in the dry season. The first water-point committee was elected in 2003 as part of the handover process described above.
In Grootvlakte, the committee consists of a chairman, a treasurer, and a caretaker. Although the chairman does not live in the village permanently, he comes into play when the community wants to apply for assistance from the government or when a serious conflict arises. The role of the treasurer and the caretaker are much more critical on a daily basis. Both roles are necessary to guarantee an everyday water supply. Although the positions of secretaries, advisors, and even the chairperson can stay vacant for some time, this does not apply to the treasurer and the caretaker.
While the role of the caretaker is limited to tangible activities, the role of the treasurer is more complicated and requires diplomacy and trustworthiness. Managing money is a sensitive issue since in most households cash is scarce and available only for a few days a month. The treasurer must not only manage the community's account, but also remind poor neighbors of their outstanding debts, resist the temptation to "borrow" some of the money to buy food when the household runs out of cash, and stand up to frequent rumors and allegations of embezzlement. Community members who meet these requirements are rare, as the case of Regina reveals.
Regina occupies the role of treasurer and has been responsible for keeping the community's funds in Grootvlakte since the middle of the 1990s. At that point in time, the government began to prepare some communities in the Kunene area for CBNRM. During the past 20 years, Regina has repeatedly resigned from her position "being tired of rumors and accusations" as she explains. Nevertheless, the community member who took the job soon handed the responsibility back to her. As one of them explains, "you cannot work with money while you are hungry yourself." Regina is a female household head in her fifties. Like the majority of other women, Regina is responsible for managing the household budget and supplying family members with sufficient food throughout the month. She is admired and respected within the village for being a successful farmer. Her herd contains about 20 female cows and about 50 head of goats and sheep while the majority of households in Grootvlakte possess small stock but no cattle. Regina is usually present in the village as she wants to have an eye on her cattle. She is not only responsible for her own animals but is also entrusted to keep livestock for her relatives. Because of her active role in farming, Regina has taken care of water issues since she moved into the community. Indeed, her role as treasurer is often expanded to that of chairperson, as she calls in and leads meetings in his absence, communicates with the government, and is usually contacted first in case of any conflict or trouble.
The particular case is singular in our observations in Kunene. Nonetheless its shows how, by selecting an older woman as treasurer a community establishes a complementary authority to the older man who serves as a chairperson. Even though gender roles rarely allow older women to challenge the chairperson in public debates, female elders are most trusted to manage a scarce resource and often have the relative financial and social autonomy for doing so. Even if older women do not own cattle, most of them receive an old age pension and financial support from their children. Therefore, they are typically more secure than households with younger heads. Asking other community members to pay their fees or reminding them of their outstanding debts requires authority and respect.
The social system provides them the needed autonomy and authority.
Throughout the research area, women are typically in charge of preparing food and providing as well as safeguarding the cash for supplying it (Schnegg 2016b). Therefore, they are more easily entrusted to save money for diesel fuel and eventually to provide water. It is a common and often confirmed fear, that men, especially younger men, will use the community's money for other purposes. Experiencing a certain monotony in daily life, many younger adults are tempted to spend time and money in local bars, where friends await to be invited for a drink. To be able to invite them for a beer, the "theft/borrowing" of small sums of family money is a common practice. In this situation women and especially elder women are perceived to be very trustworthy individuals and least subject to such temptations. Selecting a woman as treasurer thus translates a model practiced in a different field, household finances, to the new context, water governance (Cleaver 2012).
In Grootvlakte, like in other communities with a diesel engine, the position of the treasurer is most important to secure the availability of water for the community. If communities run out of diesel fuel for the pumps, the water will be finished soon after. In comparison to the studies mentioned above, where women usually just fill minor roles or their de facto influence is very low, some women in northern Kunene do have a 'leadership' position. For assessing the consequences this might have, we distinguish two levels: individuals and the groups.
At the group level, we do not find any difference between the sharing arrangements that exist in communities in which women play an active role and those in which they do not. Neither the rules of access nor the distribution arrangements differ significantly. Unlike in many other cultural contexts, the relations women and men have towards water are mostly the same. Both have a similar interest in clean drinking water. As water is not piped, it needs to be carried to the homestead where it is used for human consumption. Unlike in other parts of Africa, carrying water to the house is not entirely a female domain. If the water point is too far to be reached by foot, it is mostly younger men who are responsible to transport water with the donkey carts to the homestead. As we have seen, both women and men possess livestock and women actively engage in livestock farming. Moreover, if gardens exist they are entirely rain fed.
On the individual level, the analysis points into two distinct directions. On the one hand, the role of treasurer can be conflictive. Most people do not have cash savings at hand and the money they receive from salaries, pension payments or sales are spent within a few days. In those few days, school fees are paid, groceries are purchased, and cloth and cosmetics for the entire month are bought. As the following ethnographic vignette reveals, the lack of cash can fuel conflicts, envy, and distrust.
Sarah, is treasurer in one of the communities surrounding Fransfontein. One day in 2013, she came home from town carrying bags with groceries and cloths.
Sarah works in a nearby hunting lodge and had just received her monthly pay. When one of her fellow committee members saw Sarah, he initiated a complaint that Sarah was stealing the committee's money to spend it in town. When it came to the episode in our interview, Sarah reflects: "I cannot even bathe myself, because I am carrying money, then I am apparently bathing with committee money. The people used to gossip about me, while they were sitting at the water point." 4 In the small communities, rumors circulate fast and can have lasting damage to one's social reputation. Participation thus does not necessarily lead to empowerment and a number of former treasurers state that their position within the communities worsened due to conflicts about money. As a consequence, they may not want to engage in water management anymore (Harris 2009;Cleaver and Hamada 2010).
On the other hand, participation can strengthen the position of women and open doors to the political domain. If women stay active in local governance institutions over longer periods of time, they usually get opportunities to expand their activities. The participation in further organizations or committees (land, grazing, managing donor fees) and the Traditional Authorities Board are among the most common paths. If this becomes the case, women gain further reputation and respect and become less vulnerable to rumors and the other factors that challenge their social position as community representatives.
The work of Singh (2008) and Cleaver and Hamada (2010) also demonstrate that the role of women and their scope of action within natural resource management is often oversimplified. Singh contrasts a gender-based participatory paradigm introduced by development agents with community perspectives on water governance. Comparing 27 villages in rural India, she shows how the introduction of a formal participatory strategy into specific sociocultural contexts may lead to unintended results (2008,938). One example refers to the homogenization of women within the participatory paradigm; Singh's empirical evidence reveals that the participation of high caste women in local water governance institutions can actually lead to the exclusion of poor women who lose access to the resource (2008,935). In a similar vein, Cleaver and Hamada (2010) show how gender is intertwined with other dimensions of inequality such as wealth or caste (2010, 30) and call for a more holistic view of gendered resource governance, participation, and exclusion.

Conclusion
Starting in the 1990s the Namibian government turned ownership of and the responsibility for its rural water supply to local users. While the state withdrew from managing resources directly, it formulated the ways communities should govern resources and in doing so created water as "new commons." A leitmotif in this process is the notion of participation. In this article, we have applied the concept of travelling models as a theoretical guide to analyze how the understanding of participation is translated between contexts and scales. Moreover, as an abstraction of the world, the model entails and provides agency to change local social worlds. The concluding remarks cover both steps.
First, after Independence Namibia started redefining its land and water rights. In this process, the Namibian state drew on globally circulating models of natural resource governance, most importantly CBNRM. Its argument was common place at the time: resource ownership encourages sustainable use, which eventually fosters economic development. Participation -mainly expressed as the idea of the general equality of all adult community members -became the key principle to creating ownership and thus to attaining sustainability and development success. Since then, participatory organizational structures have been developed in northwestern Namibia and most communities have an active water point committee today.
In rural communities, these travelling models of participation bring an abstraction of the social world as understood in the Global North into conjunction with a social network that is dominated by kinship as well as gender practices and roles. At the intersection of global ideals and local norms specific forms of governance emerge. Secretary and caretaker roles are relatively clearly defined and the choice and criteria of office holders can largely be justified through the skills required for performing them. For the casting of the two remaining positions two social categories play a major role: gender and generation. While the chairperson is typically an elderly man, the treasurer is an elderly woman. Here, our results contrast many existing studies that have shown, that women's participation in community management projects is generally low and that few women have leadership positions. While the role of the chairperson is officially defined as a leadership position, the role of the treasurer is not. Only a closer look at daily praxis reveals that a treasurer position requires authority, respect, and trust. This trust goes hand in hand with the relative social and financial autonomy of many elder women and their experience in safeguarding and managing finances for supplying food. Their autonomy is backed by the kinship system that endows elderhood with authority and respect.
Second, the specific translation of the participation model that emerged began to transform social reality itself. While we could not find any consequences at the level of institutional regimes or in delivery of the water supply, the public positions mattered for the individuals themselves. Interestingly the autonomy of elder women does not vanish with new organizational forms that are built on civic governance and formal education. As we have shown, a social role that used to be carried out in private domains is transferred to the newly created public sphere. However, the role of treasurer brings distinct social fields into conflict creating ambiguity for women who hold that position. Many women have been accused of stealing money which questioned about their social reputation at large. The responsibility thus potentially increases their vulnerability. At the same time, if successfully performed, the position can also open doors to the larger public and political domain. The interesting question after this research remains, whether and how the developments on the individual level will change the role of elder women in particular and the social structure in general.