Natural resource governance in lower Omo, Ethiopia – negotiation processes instead of property rights and rules?

Research on common-pool resources in the last 30 years has hinged on concepts such as rules and property rights for understanding how access and use of natural resources is managed by communities and other actors. However, a small body of literature on mobile pastoralism maintains that resource governance might not always be based on resource-related rules, but instead on negotiation and general norms of reciprocity. Situations conventionally labelled as ‘open access’ might therefore not always be as unregulated and unmanaged as they seem. Here, we examine what the absence of rules for resource access and use means in practice, and how resource users adapt such a governance system to increasing scarcity of pasture land. We conducted interviews, group discussions and participatory mapping exercises in two neighbouring areas, Hamar and Bashada, in the lower Omo area in southern Ethiopia. Both groups are culturally closely related to each other, but showed important differences in their ability and willingness to change their institutions to adapt to resource scarcity. In both Hamar and Bashada, access to grazing was generally non-exclusive. Instead, we found a complex mosaic of ways in which access to grazing was practiced and sanctioned, characterised largely by negotiations and interplay between individual actors rather than by firm rules. Both groups were confronted with increasingly erratic rainfalls and insufficient availability of pasture. Strikingly, while the Bashada had recently established a strictly enforced set-aside area to provide grazing for the end of the dry season, the Hamar rejected such ideas and sought grazing in protected areas, which eventually led to conflict between herders and authorities. Reasons for these diverging strategies might be connected to subtle 446 Degu Tadie and Anke Fischer differences in the degree to which decision-making is individualised and social coordination accepted. These seem to have important implications for community adaptability to changing environmental and societal conditions.


Introduction
Ever since Ostrom's seminal publication "Governing the Commons" in 1990, a large part of scholarly and applied enquiry into resource use by local communities, especially in developing countries, has adopted her and her colleagues' perspective on the governance of common-pool resources. This perspective is manifested, for example, in the Institutional Analysis and Development (IAD) and its successor, the Socio-Ecological Systems (SES) framework (Ostrom 2007), and the design principles for successful community-based resource management (Ostrom 1990(Ostrom , 2009, and has led to an extremely fruitful shift of the academic gaze towards the role of local communities, with probably immeasurable impacts on real-world practices in relation to the empowerment of local actors, too. Two key elements of this perspective are the notions of (i) property rights and (ii) rules, crucial concepts in both analytical frameworks and design principles. Property rights refer to rights of access, withdrawal, management, exclusion and alienation (Schlager and Ostrom 1992) in relation to a resource, or more precisely, a good or service. In addition, three different levels of action can be distinguished, each of them usually guided by sets of rules -operational rules, collective choice rules and constitutional rules. These shape and constrain resource appropriation and use, decision making about operational rules, and modalities of decision making overall, respectively (Schlager and Ostrom 1992). Rules operationalise rights and rights can thus, in fact, be conceptualised as the product of rules (Schlager and Ostrom 1992). And while resource access and withdrawal (e.g. harvest) are situated at the operational level, management, exclusion and alienation require action at the collective-choice level, that is, decisions that might alter the operational level (Schlager and Ostrom 1992).
Operational rules and the specification of related rights to resource access and use tend to be the centrepiece of common-pool resource governance. These rules and rights tend to make clear connections between resources and people, stipulating who is allowed to use (e.g. harvest, collect or abstract) how much of a specific resource (e.g. grass, firewood, water) in a given temporal, spatial and social context (see examples in e.g. Cox et al. 2010;Wakjira et al. 2013). Where such resource-or ecosystem-related, operational rules for resource use exist but the overall use pattern is unsustainable, they are usually seen as a starting point for developing more sustainable resource management approaches (Scoones 1999;Wakjira et al. 2013). Where such rules are absent or exist but are not enforced, an 'open access' situation is conventionally seen to prevail that exposes a resource to overexploitation and degradation (Ostrom 2009). And while the largest part of research on common-pool resources focuses on common property governance arrangements, for which clear, shared and effectively enforced rules are crucial (Ostrom 2009), empirical analyses of open access tend not to go beyond the fatalistic diagnosis of a tragedy of the commons (Hardin 1968) or absence of effective restrictions on resource use (McGinnis 2011, 179). However, a notable exception consists in the literature on transhumance and the management of rangelands by mobile pastoralists, especially related to western and northern Africa (e.g. Niamir-Fuller 1999;Moritz 2016). In this body of literature, there is a strong recognition that grazing lands can be operated as open access, i.e. without clear use rights and resource-related rules, which does not necessarily imply that use is unregulated and unmanaged: Access might not be based on rigid, spatially defined rules but be obtained through "social networks and norms of reciprocity that are characterised by flexibility, porosity, and malleability" (Moritz et al. 2013, 352, see also Mehta et al. 1999;Turner 1999;Galvin 2009).
Here, we draw on this literature to explore what a (seeming) absence of resource-related operational rules for resource access and use might mean for our understanding of resource governance in a situation that, as described by Moritz et al. (2013, 355), is based on a certain "ethos of open access", but where resource scarcity, emerging in recent years (Terefe et al. 2010;Gil-Romera et al. 2011), renders current practices unviable and seems to call for adaptation.

The lower Omo case: motivation for this study
We stumbled across this question during our fieldwork on illegal hunting and human-nature relationships in the lower Omo valley in Ethiopia (Tadie and Fischer 2013). While investigating what hunting meant to members of four ethnic groups -the Kara, Hamar, Bashada and Arbore -we found that the hunting of large game had extremely strong social meaning, associated with complex and intricate practices that almost exclusively referred to interactions between people. By contrast, interactions between people and wildlife appeared rather simple, with only very few customs and norms that structured access to and practices related to wild animals. This also seemed to be the case for relations with other non-human elements of the natural environment, for example, grazing land or trees for beekeeping, but our data was too limited to allow us to understand how resource access and use in areas other than hunting was practiced. However, especially among the Hamar, the seeming absence of social norms in relation to natural resources was striking and starkly contrasted with the ubiquitous nature of references to taboos, customs and other informal institutions in the literature on local communities' natural resource use, especially in developing countries (Colding and Folke 2001;Ashenafi and Leader-Williams 2005;Jones et al. 2008). We suspected that this absence of explicit and shared norms related to natural resource use among the Hamar might be connected to their rejection of authority and the individualisation of knowledge and decision-making (Lydall and Strecker 1979a), and result in feelings of helplessness, lack of agency and despair in the face of environmental degradation and increasing scarcity of resources (Tadie and Fischer 2013).
In 2012, we went back to lower Omo to explore these issues in more detail. In particular, we wanted to know how decisions about the use of natural resources, especially pasture for livestock, were being made in a setting where clear resourcerelated rules that could govern access to and use of grazing land and determine norms for people-nature interactions seemed to be lacking. For contrast, we focused on two culturally very closely related groups, the Hamar and Bashada, as these had the same cultural roots but, as we will see, different ways of dealing with resource scarcity.

The Hamar and Bashada of lower Omo
The lower Omo valley, situated in the very south of the Southern Nations, Nationalities and People's Regional State in Ethiopia, is known for its ethnic and linguistic diversity (Strecker 1976a), and is home to groups such as the Hamar, Mursi, Arbore, Nyangatom and Dassenech. The groups we focus on here, the Hamar and Bashada, much like the nearby Mursi (Gil-Romera et al. 2011), are more closely related to pastoralist than agriculturalist cultures, although nowadays, cultivation of sorghum and maize is widespread (Lydall and Strecker 1979a;Wolde Gossa 1999), and governmental food relief programmes supplement many livelihoods. Both Hamar and Bashada keep cattle and goats, living in small, usually dispersed settlements for most of the year with their families, but, like the Mursi (Gil-Romera et al. 2011), herders will move their livestock, especially their cattle, during the dry season (roughly: December to February) to grazing areas further away once pastures around the hamlets have been depleted (see also Strecker 1976b). As such dry season grazing areas are often located at a substantial distance to the settlement, these trips, undertaken largely by male members of the families, can take days or weeks. Livestock husbandry and pastoralism permeate a large part of Hamar and Bashada culture, from the shaping of social relationships through the exchange of animals (see examples below) to everyday practices and the content of stories, traditions and rituals, such as the boys' initiation rite which involves leaping over a row of cattle (Lydall and Strecker 1979b;Epple 1995). In 2012, people in the villages were still largely independent from larger market economies, with AK-47s, men's clothing and necklaces from plastic beads being more or less the only industrially produced artefacts present in the villages. However, the area is rapidly changing, due to cultural tourism, land degradation and, most recently, the development of large-scale agricultural plantations (see e.g. Turton 2011).
Detailed studies of the communities living in the region began to appear in the late 1960s, mainly by European anthropologists. Since then, scholars from both Ethiopia and abroad have conducted anthropological studies among many of the ethnic groups in the area (Turton 1973;Carr 1977;Almagor 1978;Gebre 1993;Wolde Gossa 1999;Elfmann 2005).
For our focus on the Hamar and Bashada, the work by ethnographers Lydall and Strecker (1979a,b) and Epple (1995Epple ( , 2010 is of highest relevance. In both groups, the coordination of social life seems to be characterised by decentralised collective mechanisms, with slightly higher expectations in relation to collectivism among the Bashada than the Hamar. For example, Epple (2010) describes how a Hamar woman living among the Bashada was criticised for being too independent, not asking for help (in the form of work parties) when it seemed appropriate to the rest of the village (namely, for the construction of her house). Governance of social life in both groups is portrayed as relatively egalitarian, with constant negotiation and debating processes between elders and other community members, until a conflict resolution is accepted. From an individual's perspective, this requires a balancing act between one's own needs and views (which generally seem to be very strongly developed) and the realisation that "all people depend on each other, not only in economic terms, but also socially, to solve conflicts, to re-establish social balance and peace" (ibid., 242). For the Hamar in the 1970s, Lydall and Strecker (1979a, 197-198) describe this as a general striving for "free individual choice in the application of general principles. […] everyone works towards a maximization of choices in any particular social situation" in a context "where hardly any social relationship can be axiomatically trusted" (Strecker 1976a, 591).
Here, we explore how, within these contexts of social coordination, natural resource access is practiced and governed, focusing on the role of operational rules in livestock grazing, as this is culturally the most important livelihood activity of both Hamar and Bashada.

Methods
Our analysis builds on data collected through interviews, group discussions and field visits in 2010 and 2011 (Tadie and Fischer 2013), but focuses on a set of interviews, group discussions and participatory mapping exercises in May and November 2012. Overall, 15 interviews (two of these with two interviewees each, i.e. including 17 interviewees overall) were conducted -eight in three of the four villages comprised by the Bashada area, and seven in Gembella kebele (ward or peasant association, the smallest administrative unit in Ethiopia, usually consisting of several villages) in Hamar District. Interviewees were selected for diversity of viewpoint and included both young men and elders among them, for example, the spiritual leader (bitta) of the Bashada area, a Hamar employee of the district administration, and the Hamar chairman of the Gembella kebele (the chairman of the kebele of the Bashada area participated in the mapping exercise described below).
Interview guidelines were flexible and covered practices of resource use and access, focusing on livestock grazing, beekeeping and farming, resolution of conflicts in relation to resource use and the role of elders, interethnic relationships and the history of different ethnic groups and their interactions, and perceptions of recent changes in resource availability and governance. Interviews tended to start with broader questions on the history of settlement in the area and livelihood activities, and then developed questions on access to grazing, trees and land used for crop cultivation. More specific questions then probed practices related to the negotiation of access among neighbours, among people from different hamlets or with different roles within the same ethnic group, and among people from different ethnic groups.
The interviews were complemented by two group discussions that involved the participatory drawing of a resource use map ( Figure 1) in Gembella, Hamar (20 participants) and with participants from two Bashada villages (n=9). Each group was asked to draw a map of the kebele, locating places of settlement, farming and grazing, and movements throughout the year (Figure 1). The drawing of these maps served as prompts for participants to speak about their resource use activities, emerging conflicts, and ways to address these. As livestock herding and beekeeping are male tasks, usually carried out by young men, and our research interest lay in the actual practices of access taking (although we recognise that gender roles can work in unexpected ways, Lowassa et al. 2012), all participants were male, again ranging in age from youngsters to elders.
All conversations were held in Hamar language, the mother tongue of both Hamar and Bashada, with the help of translators, audio-recorded, and then verbatim transcribed into English. Data was coded in NVivo in a grounded manner, starting from the original, broad research interest (namely to better understand how access to natural resources was governed among the Hamar and Bashada; see above), and refining the analysis in an iterative process.

Results
Our presentation of the results starts with an analysis of the ways in which access to grazing was organised among both the Hamar and the Bashada, and then identifies differences in grazing management between the two groups, notably, the use of a set-aside area. Finally, it points at recent changes in understandings of property and access.

Access to grazing: the role of rules and negotiations
Access to grazing was generally non-exclusive among both groups, Hamar and Bashada: As such, there were no designated user rights for pasture: Where there was a settlement close by to a grazing site of interest to another livestock holder in search of pasture, it seemed that there was an expectation that the local residents would be approached by the newcomer even if there was no livestock present on the pasture at the time of arrival (and the land was therefore not currently in use), but there was a clear potential that these areas could be used at any moment. This expectation seemed to decrease with distance from a settlement. Where access was denied, livestock holders might graze animals without agreement, with a clear risk of physical conflict, especially where groups did not share ancestry or spiritual leaders: However, mechanisms existed that would reduce this risk, and it seemed that usually, exclusion from a grazing area was not an option: No, there won't be a 'no' answer. When he first arrives, he won't say "I came here for grass", he would say "we came here as a guest". Then it would be said "a guest has come" and then a cow hide would be laid out for him, then he would sit down and coffee would be boiled for him, then it would be said "give milk for these guests", milk would be given, "give them food", food would be given, then after they've eaten, they would be asked "you guests, where are you going?". Then, if it is said "we came to you", then other elders from the village would be called, and then they would speak it out. Then we would listen […], in a calm condition, we would talk, then it would be said "ok you settle here [for the moment]"; it would be said like this and they would be given a place. People won't suddenly come and ask, standing on the road. [By., Bashada] This flexible approach to resource use was facilitated by social institutions that helped develop relations between individuals that did not necessarily share family bonds (see Tadie and Fischer 2013 for long-lasting bonds between non-family members created through hunting), and a fluid and complex understanding of livestock ownership that was neither communal nor private. Two types of relationships derived from joint ownership of livestock were described in our interviews (see also Lydall and Strecker 1979b). Beltamo meant that the owner gave a cow or goat to another person, who would take care of it and be allowed to use its milk, blood, offspring and also meat. The recipient would then consider the previous owner a friend. The reverse phenomenon was called siti: A (usually poorer) person would ask to 'borrow' livestock from the (usually better-off) owner, so that they could use the milk, blood or offspring. In both cases, the giver would periodically visit their stock and might demand the original animal, some of the offspring, or other gifts.
When we rear cattle, they [Hamar] would also take our calves and rear them there and they would use for themselves; we also take their cattle and herd them and use them as well; that is why we say "my father's cattle" whenever we see cattle. There is exchanging cattle through bel [tamo]. [G., Bashada] While there was generally a very clear understanding of which cattle belonged to which group, which made intended deceit, mixing one's cattle with others in order to obtain access to a grazing place without explicit consent, very difficult, arrangements such as beltamo and siti facilitated negotiations over access to grazing: There was a widespread perception that the availability of grass had been decreasing in the last 25 years, that the land was degrading -especially in Gembella (Hamar), large areas did not have any grass cover anymore, shrubs were encroaching, and soils laid bare -, and the pressure on the remaining pasture was growing. Some interviewees explained that this was due to increasing cattle numbers in the area, and backed this up with detailed accounts of historical migration movements into the area and increasing cattle ownership (see also Tadie and Fischer 2013).
Within this general, shared understanding of grazing availability and access, the resource use maps from Gembella (Hamar) and Bashada showed two different approaches to the governance of grazing. In both places, livestock owners had fenced off areas (derr) around their homesteads to be able to keep the animals (especially kids and calves, cows and goats for milking, or sick animals) in confined places and away from farmland. However, this approach seemed more widespread in Bashada than in Gembella (Hamar).

Grazing management among the Bashada: Coordination and a set-aside area
In terms of use of the wider landscape, participants in Bashada described how they used specific places for grazing, farming, settlement and other activities, for example, as salt licks and for beekeeping. Their grazing grounds were situated in the northwestern and western part of their territory, bordering Mago National Park, while their settlement areas were established in the central, eastern and south central part (Figure 1). Most agricultural plots were situated in the fertile areas around the settlements and adjacent to dried-out river beds (Figure 1). The scrubland and small woodlands in the southern and eastern parts of Bashada were mainly used for beekeeping and occasional grazing by small numbers of cattle and also as saltlicks. Cattle herding was carried out as an informally communal activity (dashed arrows in Figure 1A show direction of movement): Those livestock owners who could send several youth to take care of the livestock would send their herds first, followed by those with fewer or no herders, so that their livestock could be watched by others.

Degu Tadie and Anke Fischer
There are always cattle that follow the first cattle to be released. They can't go ahead of the first one; they would follow them. They would stay 100 or 200 metres away; they would make a row and graze in one direction. [K., Bashada] This approach was seen as beneficial in terms of defence against both predators and human enemies, such as the Mursi, although it was widely recognised that cattle raiding and related conflicts with the Dassenech, Mursi and Nyangatom had substantially decreased. 1 It also allowed a more efficient use of the available grass, avoiding unnecessary trampling.
Interestingly, a few years ago, the people of Bashada had started to set-aside a part of their traditional grazing land as a reserve for dry season grazing. This measure had been suggested by the district administration, and been negotiated and agreed after ten days of community deliberations. It appeared to be strictly enforced -no-one was allowed to enter the set-aside area during its closed-season, not even for beekeeping, as there were concerns that carelessness during the harvest of the honey could cause bushfires. Elders would jointly decide on the start of the open season, in which all Bashada livestock owners, independent of wealth, livestock numbers or status, were allowed to move their animals there for grazing. As soon as the rains started, livestock would have to leave the area again to allow the vegetation to recover for the next season.

Grazing management in Gembella (Hamar): less coordination, no set-aside but use of a national park
This was somewhat different in Gembella (Hamar) ( Figure 1B). Agricultural plots were, again, situated along the sandy river beds, and joint herding was seen as beneficial to protect oneself from cattle raids, but participants described how people sent their livestock all over the kebele, simply dependent on availability of grass and water, even to areas traditionally known for conflicts with the Dassenech (red X, Figure 1B).

… yesterday, my goats went there, now they went this way. They would go as they wish to go. Today this direction, tomorrow that direction. [M., Hamar]
Towards the end of the dry season, this regularly also included the area designated as Mago National Park, which, albeit illegal, tended to be tolerated by park staff. In February 2011, at the end of a particularly hard dry season, herders left the 1 It could be argued that decreasing resources would increase the potential for conflicts. Indeed, our findings here of reduced conflict seems to contradict Gebre's (2012) observation that conflicts, in particular those with fatal outcomes due to the use of firearms, between Dassenech and Hamar had increased in recent years. However, as Gebre (2012) points out, causal effects of resource scarcity on incidence of conflicts are difficult to identify, and there are numerous other socio-cultural, political (e.g. the role of governmental mediation efforts, or of law enforcement across country borders) and technological factors (e.g. availability of semi-automatic rifles) that interact with availability of resources in triggering or preventing conflictive action. national park well before the rain started, leaving their cattle behind with the aim to retrieve them later, after the message spread that two herders had shot two of the very few remaining giraffes, and that the district government therefore would now strictly enforce the non-grazing policy in the park and expel all livestock herders.
Although the people of Gembella kebele had received advice similar to that given to the Bashada villages, they had not established a set-aside area, and were therefore reliant on other grazing land when their own resources were exhausted towards the end of the dry season or in times of draught. Grazing was generally scarce in Gembella in comparison to the area of the Bashada villages or Kara, which was seen as much more fertile and less prone to shrub encroachment. Our participants from Gembella described how they, in the event of shortage of water and pasture for livestock, would search for places with available resources, and then negotiate access with the people present in that area. Usually, this was done by vividly describing the difficulties that the livestock were facing, and by picturing the possible consequences for the livestock and their owner if they were denied access. Our Bashada participants illustrated how these negotiations worked: First the guest would come to the house and would see the host's cattle condition. Since he is pastoralist he would observe the cattle when they come back from grazing and whether they are full or not. Then he would say "your cattle are in good condition and their hair shines. But since there is no pasture in our area, our cattle are not well, their hair became shaggy and they lost weight". Then when the host asks "don't you have a pasture for the cattle?" then he would answer, "There is no grass in our area and all the cattle are failing and are about to die because of hunger". Then again he would say "how come your cattle are so well, their hair is so shiny, do they graze in a good place?" The host would say "they are using the demarcated area for the dry season". Then the guest would ask "please save about two of my cattle". He won't tell exactly the number of his cattle, he would mention only two or three. Then the host would answer "for the time being there is grass, no problem, just bring them". Then after the guest returned to his house, he would send ten, fifteen or twenty of them but he mentioned only two or three when he first asked. Once they are mixed with his cattle, then the host would take them to his pen. Then after some time the guest would come again to see the conditions of his cattle. [P., Bashada] Overall, thus, traditionally, clear resource-related, operational rules for grazing seemed to be absent in both Bashada and Gembella (Hamar) and access to pasture was largely negotiation-based, with the Bashada giving more importance to the collective coordination of grazing activities than the Hamar. The recently established set-aside area in Bashada, which did require the collective agreement of very clear resource-related rules, seemed to work well, whereas a similar initiative in Gembella had not been supported by the community. This was seen to be due to fundamental objections to property-related rules: M: You can't protect it […]. The only choice is to be together and starve together. Here, there is no culture to protect this.

Recent changes: understandings of property
However, some Bashada interviewees also described how the idea of private property, slowly being introduced through agricultural extension and governmental education measures, started to undermine the social balance in their community. For example, acting on recent government advice, people had started to fatten up selected bulls in their derr for sale on the market. This speaker criticised the increasing privatisation of land and other resources: Making an area closure in a faraway place is good. But this thing near me which we're being advised to do "making a closure using fences here near my house" is difficult. It would create conflict among us; it would get us into quarrels with my neighbours. This education which is being given [from the government] to make a grazing closure in the village in between the people may get us into conflict.

Discussion
We set out to explore the role of property rights and operational rules in relation to resource use among the Hamar and Bashada of the lower Omo valley; thereby unpacking what the presence, absence or simply the character of such rights and rules might mean for open access vis-à-vis common property regimes. We found a complex mosaic of ways in which access to grazing was practiced and sanctioned, characterised largely by negotiations and interplay between individual actors rather than by firm resource-related rules that just needed to be enforced. This was in contrast to the arrangements described for other traditionally pastoralist communities in the wider region and elsewhere in the world, which often take the existence of indigenous resource-related rules as a given, even where these are challenged by governmental interventions or the increasing heterogeneity of users (Berger 2003;Axelby 2007;Gilbert 2013;Conte and Tilt 2014;Yembilah and Grant 2014). However, it was in line with and provided further and more detailed insights into a line of thought in the transhumance literature (mainly from western and northern Africa) that maintains that open access regimes might not be governed through clear resource-related institutions, whether formal or informal, but through intricate mechanisms based on "procedural rules that allow resource access to be malleable to political negotiation/bargaining" (Turner 1999, 643-645;see also Niamir-Fuller and Turner 1999) and an "ethos of open access" (Moritz et al. 2013, 355). These studies show that open access cannot simply be equated to resource degradation and a tragedy of the commons (Moritz et al. 2013), as these more procedural and flexible arrangements might fit well with both the social and ecological conditions (Cole et al. 2014;Moritz et al. 2015).
The prominent role of human-human relations in grazing management -as opposed to rules that govern human-nature interactions -mirrored our previous findings in relation to hunting (Tadie and Fischer 2013) that emphasised that human-nature relations cannot be understood in isolation from human-human relations. The present study expands on the nature of these social relations as processes rather than resource-related rules as usually understood in the common property literature. Thereby, it adds nuance to the anthropological perspective on property as relations between people (as opposed to relations between people and things; Hann 1998). Some of our observations, for example, the common practice to ask other resource users who might have a claim on the grazing area for their agreement -although it is understood that access will almost never be deniedmight evoke expressions of territoriality and social boundary defense as described by Myers (1982) and Cashdan et al. (1983) for several hunter-gatherer societies. However, a closer look at the Hamar and Bashada cases suggests important differences. For example, there was neither an explicit expectation of reciprocity of access nor the perception that the people local to the grazing land had a right to be asked, thus implying some form of ownership (as described by Myers 1982 for the Pintupi Aborigines); rather, the negotiations carried out before letting cattle graze close to someone else's livestock seemed to be a way to avoid potential conflicts. There also did not seem to be a clear distinction between in-and outgroups as highlighted by Myers (1982): Negotiations took place within and between ethnic groups, with gradual differences rather than a clear-cut distinction. These negotiations also depended on locality, with no clear 'territories' defined but instead a spectrum from grazing land close to a settlement to areas further away. Lastly, negotiations did not seem to be used to plan and allocate grazing areas in any systematic way as proposed by Cashdan et al. (1983) for foraging bushmen -again, in our case, the process of agreeing access largely seemed to have a social function, namely to avoid conflict.
In addition, the differences we found between resource access arrangements in the Bashada and Gembella villages seemed to reflect larger differences in social arrangements between the two groups. In Bashada, decision making tended to be more strongly influenced by groups of elders and the spiritual leader than in

Degu Tadie and Anke Fischer
Gembella. There, adult men as the owners of the livestock, or their older sons as the most experienced herders, would individually take decisions where to take the livestock for feeding and watering, with little or no interference being expected from other users. Adult men appeared neither willing to accept the authority of individual leaders nor to co-operate in collective action. In property rights terms, no-one, not even a collective, was seen to have the rights to exclude others from the use of grazing land. As a consequence, there were also no clearly specified rights at the operational level, and access and withdrawal (i.e. grazing) had to be negotiated on an ad-hoc basis. While our interviews touched on such arrangements only in relation to natural resource governance, our findings align with those by Epple (2010) and Lydall and Strecker (1979a) on the organisation of social life more generally (see also Girke 2011).
Juxtaposing the governance of grazing access among the Bashada vis-àvis the Hamar of Gembella allowed us to move beyond a mere diagnosis of an absence of resource-related rules as the reason for open access and ultimately, resource degradation. In fact, we do not make claims about causal relationships between presence of such rules and the current state of the land (see Turner 1999 and Gilbert 2013 for a critical discussion on the assumed links between pastoralism and "overgrazing") -and indeed, rules might only be meaningful where some degree of (potential) resource scarcity is perceived (see examples in Scoones 1999). Our interviewees seemed to share the view that scarcity of pasture land had only recently become an issue, possibly due to increased human and livestock populations, which might have reduced resilience in the face of variations in rainfall. This view appeared to concur with that of the Mursi (as identified by Gil-Romera et al. 2011) who suggest that overgrazing and subsequent woody encroachment caused by growing livestock numbers has been one of the main reasons for the decreasing availability of pasture. Interestingly, a reduction in rangeland because of the designation of protected areas (such as Murulle Controlled Hunting Area or Mago National Park; Turton 1987) was not mentioned as a reason for a lack of grazing, possibly because at least Mago, but to some degree also Murulle were de facto used, in spite of their formal designation (see also Turton 2011).
In the face of this scarcity, our two study groups reacted very differently. While the Hamar of Gembella saw the independence of individuals' decision making (Lydall and Strecker 1979a;Girke 2011) as paramount and rejected governmental suggestions for set-aside areas, the Bashada, whilst only slightly less egalitarian and individualistic than the Hamar (Epple 2010), appropriated and adapted these suggestions to their own needs, but at the same time discarded and criticised governmental advances that would have undermined their understanding of common property.
Our analysis highlights that cultural acceptance of rights and rules in resource governance should not be considered as a given, and that the level of such acceptance can vary even between neighbouring and culturally very closely related groups. Where arrangements for resource governance are based on negotiation rather than on rules, such negotiation might follow, of course, its own set of (procedural) rules, which, as illustrated here, might be guided more by social than by environmental considerations.
We argue here that this can have two types of implications. First, it can hamper adaptation to environmental change, for example, to changes in resource availability. With the effects of climate change and decreasing availability of rangelands, adaptability will be crucial, but our understanding of adaptation of community-based governance of natural resources over time is still quite rudimentary (Wakjira et al. 2013). Our findings seemed to suggest that there are differences in adaptability even between two otherwise very similar cultural groups: In both groups, grazing was governed largely by process-based and social rather than resource-related rules. But while the Bashada had accommodated some resource-related rules (e.g. concerning the set-aside area) into their governance, the Hamar did not, and explicitly objected to these; a finding that seems hugely intriguing from a cultural psychology perspective (Moritz 2008). In practical terms, the use of scenarios in participatory workshops and group discussions that elicit reactions to and evaluations of different ecological and social scenarios together with governance options could be a first step to explore these differences. In more conceptual-analytical terms, future studies could also investigate how rules that guide access and use of resources are embedded in collective choice and constitutional arrangements. Epple (2010) and Girke (2011) describe such higher level governance structures among the Bashada and Hamar for social life in general, but little is known about this in relation to resource use. Such work could shed light on the ways in which such rules can realistically be altered.
Second, where communication with other actors is involved, implicit assumptions about the acceptability of property rights and resource use rules might render interaction between resource users (such as the Hamar) and external actors (such as agricultural extension advisers) ineffective. There needs to be, therefore, an even greater awareness and appreciation of the role of negotiation-based governance systems among both practitioners and academics.
More generally, and looking ahead, our analysis illustrates the potential usefulness of a more psychological and sociological perspective on governance: Future work could investigate the factors and dynamics underpinning communities' acceptance and interpretations of governance arrangements. Places such as Hamar and Bashada would provide excellent cases for a comparative exploration of the role of everyday activities and the resulting socialisation (Moritz 2008) in producing and reproducing a local governance culture.

Literature Cited
Almagor, U. 1978. Pastoral Partners: Affinity and Bond Partnership among the Dassanetch of South-West Ethiopia. Manchester, UK: Manchester University Press.