Territorial Conflicts, Agency and the Strategic Appropriation of Interventions in Kenya’s Southern Drylands

A number of scholars have noted that interventions, such as development programmes and climate change adaptation projects, that simplify complex social realities and thus lose sight of the relational dynamics beyond the target or beneficiary group, risk contributing to conflict. This article examines how a series of interventions in a particular dryland area in southern Kenya became embroiled in a long-running territorial conflict between the Loita Maasai (the beneficiary community) and their neighbours, the non-beneficiary Purko Maasai. Based on ethnographic research and by taking a historical perspective, it shows how Loita Maasai leaders systematically appropriated these outside interventions, used and reworked them with the strategic aim of stopping land loss to ongoing Purko encroachment. The analysis reveals two ways in which Loita leaders realized this: (a) by using interventions to stake out spatial claims to land; and (b) by capitalizing on the tendency of interventions to simplify local contexts. This article contributes to the debate on the linkages between intervention and conflict by highlighting the agency of intervention beneficiaries and showing that, through their actions, interventions may unwittingly reproduce and even aggravate existing conflicts.


Introduction
Interventions, such as development programmes and conservation projects, that lose sight of the relational dynamics beyond the target or beneficiary group risk creating new conflicts. Soeters and Zoomers [1] show how climate change adaptation interventions in African drylands may strengthen the resilience of target groups (such as sedentary farmers) by altering natural resource use and management, but simultaneously contribute to conflict as they negatively affect non-beneficiaries, i.e., user groups who are not the target of interventions, like nomadic pastoralists. The problem, they argue, is that adaptation interventions have a tendency to simplify complex social realities and employ a singular perspective that links a single natural resource to a single livelihood (farmers, loggers, pastoralists, etc.) or social group identified along the lines of gender (for example widows), age (the unemployed youth), ethnicity, religion, etc., thereby ignoring competing claims and uses by other groups or communities. Therefore, they argue that for adaptation interventions to be more inclusive and less conflictive it is necessary to look at the wider, social landscape so as to cater for the complex set of relations that exist around natural resource use and access.
Soeters and Zoomers [1] (see also [2]) primarily advance this wider, social landscape view as a tool to improve adaptation programming. This article will use it as an analytical tool and a point of departure to explore linkages between intervention and conflict, but (a) reverses the direction of analysis; and (b) proposes a more historical take of it.
Simplification is not particular to adaptation interventions, but a feature of interventions in general. Simplification happens as intervention planners frame local contexts and define target groups (for the context of community-based natural resource management, see [3]). Scott, in his book Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed [4], argues that 'state simplifications', i.e., the simplifications that state actors make to categorize and bring order to a complex and dynamic socio-environmental reality so that they can manage, govern and implement policies, are integral to the way states 'see' their territories. Many have taken inspiration from Scott's study on how states 'see' to understand how other institutions like for example oil companies 'see' [5] or how international organizations such as the World Trade Organization or the World Bank 'see' [6]. In a way, Soeters and Zoomers [1] also examine how adaptation interventions 'see'. A different kind of reaction to Scott's book has been to turn the sides of the table and rather explore how the state is seen. Corbridge et al. [7], in their book Seeing the State: Governance and Governmentality in India, have sought to understand how different groups of people, both from within and outside the state, view the Indian state. This article draws inspiration from this reversed perspective and positions the (potential) beneficiaries at the centre (as those who 'see') and the intervention on the receiving end (as the 'seen'). Such a perspective draws attention to the agency of those who are the targets of interventions, revealing that beneficiaries do not just 'see' interventions, but also that they interact with them, actively respond to interventions, and integrate and/or transform them to suit their own needs.
The second analytical move is to historicize the intervention-conflict link. An advantage of bringing in historical depth is that this better allows for examining whether a new conflict was triggered, whether an old one was revived, or whether an existing conflict was reproduced. A historical perspective also means historicizing the intervention itself by recognizing that it may be one of the many interventions that have already taken place in a locale, targeting the same beneficiary group. The idea here is that every intervention leaves behind a legacy, as it reshapes local conditions and changes people's behaviour, even if this is not the way that was intended for by the intervention planners, as is often the case. So, in effect, every context in which an intervention unfolds might have actually been partly shaped by the local experience of a previous intervention-and people act on this previous experience. This is precisely where the added value of a combined historical and reverse perspective lies. It allows for an exploration of how beneficiaries negotiate interventions strategically, and how these strategies relate to both an accumulated understanding of how interventions 'see' and operate, as well as ongoing wider relations and struggles of resource use with non-beneficiary groups.
This article contributes to the debate on the linkages between intervention and conflict in African drylands. African drylands are areas where competing claims over land and natural resources have been on the rise, and occasionally erupt in violent conflict, while the flow of funds is increasing, particularly for climate change adaptation but also to address growing political instability. Interventions are, therefore, likely to grow in number and this makes the intervention-conflict nexus terrain worth exploring. Taking dryland Kenya as a case, this article looks at how a series of interventions in southern Kenya became embroiled in a long-running territorial conflict between the beneficiary community, the Loita Maasai, and their non-beneficiary neighbours, the Purko Maasai. The interventions presented include the construction of a primary school by the Catholic Church; the establishment of the local non-governmental organization (NGO) Ilkerin Loita Integral Development Project (henceforth: the Ilkerin development project) funded by foreign donors; the group ranch scheme rolled out by the Kenyan state but funded by the World Bank and other donor agencies; and the forest co-management project implemented by the International Union for Conservation of Nature (IUCN), which was locally known as the IUCN forest project. This article will show how the Loita (or, more specifically, the Loita leaders) strategically appropriated these outside interventions and reworked them to strengthen the Loita Maasai's position in this ongoing territorial conflict.
Theoretically, the intervention-conflict nexus discussed here is situated at the overlap of two fields of study: governmentality studies and political ecology. Inspired by Li's book The Will to Improve: Governmentality, Development, and the Practice of Politics [8], this article conceptualizes interventions as 'governmental interventions' [8,9]. Interventions are 'governmental' not because they are driven by 'the government' as in the 'Kenyan government' but because they wish to improve things through planned social change and are manifestations of a process that Foucault called 'governmentality' [10]. Governmentality is about adjusting or transforming-governing-the behaviour of people in a calculated way. Governmental interventions come in the form of policies, programmes, projects and schemes designed in the name of development, nature conservation, climate change adaptation or any other well-intended improvement effort, and are implemented by the state or other institutions such as NGOs, environmental organizations or religious groups. They need to be rationalized and an 'arena of intervention' framed before they are designed and implemented: problems need to be defined, justifications formulated, beneficiaries identified and mechanisms devised to solve the problems and change the beneficiaries' behaviour accordingly [3,8,11]. This is where simplification happens and the 'seeing' of an intervention shapes up, as a 'field of visibility' is constructed, which 'illuminates and defines certain objects . . . [while] it obscures and hides others' [12] (p. 41). This article will look at how the Loita Maasai were identified as the 'beneficiary community', and the exclusion that this implied of other communities, particularly the neighbouring Purko Maasai, as non-beneficiaries.
Although interventions are proposed and implemented to improve things, they 'seldom reform the world according to plan, but they do change things' [9] (p. 276). The focus of this article is on how interventions affect, or are expected to affect, relations of natural resource use and access, even if this domain is not the target of the intervention per se, and how this, in turn, relates to new or old natural resource conflicts. This kind of questions are at the core of political ecology. Political ecology is a field of study that examines how political struggles are implicated in changes at the socio-environmental interface [13][14][15][16][17][18][19][20]. Interventions that affect natural resource relations are thoroughly political because they involve changes in the distribution of access to natural resources. The flip coin of access, defined as the ability to benefit from natural resources [21], is that it excludes others from use and benefit [22] and the reshuffling of these relations clearly has the potential for struggle and conflict as it creates new winners and losers. Access, or rather access control, is at the heart of natural resource conflicts and we shall see how Loita leaders appropriated and reworked interventions in an attempt to consolidate their control over contested lands to secure access for the Loita Maasai-to the exclusion of the Purko.
The research material presented in this article is based on ethnographic fieldwork conducted among the agro-pastoral Loita Maasai in Kenya, during more than 22 months between 2001 and 2010, that investigated the links between interventions and struggles surrounding land and forest access and control. Because fieldwork took place among the Loita Maasai only, the analysis here reflects mostly a Loita Maasai perspective of the Loita-Purko territorial struggle. Data was gathered through informal conversations, interviews, focus group discussions, participant observation and observations during stays and visits to homesteads, temporary cattle camps, markets and manyatas (ceremonial homesteads) and while attending meetings, ceremonies and other important events. A research assistant acted as guide and translator when needed. In addition, a longitudinal in-depth study was conducted of six families with the aim of gaining an understanding of the lives and livelihood strategies of the Loita Maasai in order to place the resource-related conflicts and struggles in their appropriate socio-economic context. A wide range of meetings were attended, including political rallies, barazas (public gatherings like those called by state-appointed chiefs to communicate a new policy), land dispute meetings and so-called 'Loita leaders' meetings'. The latter are important because these are typically held to discuss problems affecting the Loita as a group (like the territorial conflict with the Purko Maasai) or to choose a path of action when faced with a new outside intervention, hence the focus on the practices of leaders in this article. It is important to note that although when facing external threats or developments Loita leaders typically act as a united group, this is not always the case as divisions and factions do exist in Loita leadership and sometimes reflect in their dealings with outside forces (see for example [23]). Three such meetings were attended and these were key for gaining an insight into the collective decision-and strategy-making of Loita leaders. In-depth interviews with key informants and Loita leaders gave further complementary information on this and into specific dealings with the various interventions of interest in this article. Land dispute meetings are relevant because this is where competing claims over land are discussed and addressed. Worthy of particular mention here is a large peace gathering, bringing various Maasai groups together where land and territorial issues were discussed. In total, more than 60 meetings were documented and 62 formal interviews and focus group discussions conducted. The material collected in 2001 was recorded in writing. Afterwards, data was tape-recorded, translated if necessary, and transcribed. All data gathered, including notes on observations and on the more informal interviews and conversations, were thematically coded in preparation for analysis. Finally, project documentation, particularly from the Ilkerin development project and the IUCN forest project, was collected and examined: this gives a glimpse into how these interventions 'see' their arena of intervention.
The article is structured as follows: the next section provides a historical background to the longstanding territorial conflict between the Loita and the Purko Maasai. It will then examine how a number of successive interventions intersected with the territorial conflict through the Loita leaders' strategic appropriation. The subsequent section discusses the findings of the research and provides an analysis of the main strategies. The article then concludes by highlighting the contribution made to the debate on the linkages between intervention and conflict in African drylands.

Territorial Struggles
"The thing is that Loitans don't like to be near foreigners. So, whenever a Purko family settles near to a Loita family, the Loitans will move further away. There is a saying used by the Purko: 'The Loita don't like donkeys'. When a Maasai homestead decides to move, they will move their belongings, old people and very small children on donkeys. Donkeys are the main means of transportation. Thus, whenever a Purko homestead settles next to a Loita family, they arrive with donkeys. So, now, the Purko have encroached onto Loita land, from Narok up to Ilkerin and Morijo. Now Loita land is very small." Interview with a Loita Maasai elder (1 October 2001) The Loita Maasai are one of the many 'sections' in which the Maasai of East Africa are organized. 'Loita' is the anglicized version of Iloitai (sing. Oloitai) in Maa and is used here to refer to both the people and the area where they live. Sections are socio-spatial units: the extent of a territory depends on the ability of a group of people to control the (agro-)pastoral resources it contains. Territories are not fixed, but flexible, and, as we shall see, may become bigger and smaller, and move across the landscape as the people who occupy and control the territory also move. The Loita straddle the Kenya-Tanzania border in the highlands to the west of the Rift Valley (see Figure 1). They are surrounded to the north and west by the Purko, one of the largest Maasai sections. This article focuses on the Loita Maasai of Kenya.
Administratively, the Loita of Kenya live in Loita Ward in Narok County, occupying the southern parts of the Loita Hills. In the east of the area, crowning the Loita Hills, lies the Naimina Enkiyio Forest. Moving westwards from the forest, the landscape descends slowly, gradually transforming from 'dry upland forest' to bushland and finally to semi-arid grassland [24,25]. To the north of the forest lies a very steep escarpment, the Narosura Escarpment, that divides the Loita Hills from the Loita Plains. The Loita Plains, though lower in elevation than the Loita Hills are still part of the western highlands. The Loita Plains and the Loita Hills, as the name indicates, used to be occupied by the Loita Maasai. Today, and as a result of the gradual encroachment of the Purko into Loita land, the Loita Plains and the northern part of the Loita Hills are occupied by Purko Maasai. Even some parts along the administrative boundary but within the Loita Ward are inhabited by Purko families. Although in the process of encroachment, Loita territory moved and extended to the south into Tanzania, the total area the Loita controlled shrank considerably, while that of the Purko expanded as they successfully pushed the Loita and other Maasai sections ahead. Despite having been close neighbours for many decades now, the Loita and the Purko rarely mix and continue to live in separate homesteads. Purko intrusion has always been conflictive. During fieldwork in Loita, their territorial struggle came to expression through cattle thefts in the boundary areas, quarrels between leaders of both groups over the boundary between their territories and occasional violent confrontations between Loita and Purko warriors over pastoral resources in the border zones. A rare attempt was made to address this territorial conflict, as well as other issues between other Maasai sections, during a large, traditional peace gathering hosted by the Loita from Kenya and held over two days in 2008 between leaders of the Loita and Purko from both Kenya and Tanzania, as well as leaders from the Kisongo and Salei Maasai sections from Tanzania. This seems to have put a stop to the fighting between warriors but the pressure at the Loita-Purko boundary continues. To understand this ongoing territorial struggle and the process of Purko encroachment, we need to go back in time more than 100 years. Sustainability 2018, 10, 4156 5 of 16 been close neighbours for many decades now, the Loita and the Purko rarely mix and continue to live in separate homesteads. Purko intrusion has always been conflictive. During fieldwork in Loita, their territorial struggle came to expression through cattle thefts in the boundary areas, quarrels between leaders of both groups over the boundary between their territories and occasional violent confrontations between Loita and Purko warriors over pastoral resources in the border zones. A rare attempt was made to address this territorial conflict, as well as other issues between other Maasai sections, during a large, traditional peace gathering hosted by the Loita from Kenya and held over two days in 2008 between leaders of the Loita and Purko from both Kenya and Tanzania, as well as leaders from the Kisongo and Salei Maasai sections from Tanzania. This seems to have put a stop to the fighting between warriors but the pressure at the Loita-Purko boundary continues. To understand this ongoing territorial struggle and the process of Purko encroachment, we need to go back in time more than 100 years.

The First Maasai Move (1905)
The long-running conflict between the Loita and the Purko over territorial control dates back to the early colonial period, when the Purko first arrived in the western highlands after the so-called first Maasai Move in 1905. Prior to the arrival of the British, the Purko had lived elsewhere, far away from the Loita. They were not even neighbours as other Maasai sections occupied the lands between their territories. Purko's territory was around Lake Nakuru and Lake Naivasha in the central part of the Rift Valley, and they actually controlled Maasailand's best pastures. In 1896, the British started the construction of a railway that ran from the coast to Lake Victoria, cutting right across

The First Maasai Move (1905)
The long-running conflict between the Loita and the Purko over territorial control dates back to the early colonial period, when the Purko first arrived in the western highlands after the so-called first Maasai Move in 1905. Prior to the arrival of the British, the Purko had lived elsewhere, far away from the Loita. They were not even neighbours as other Maasai sections occupied the lands between their territories. Purko's territory was around Lake Nakuru and Lake Naivasha in the central part of the Rift Valley, and they actually controlled Maasailand's best pastures. In 1896, the British started the construction of a railway that ran from the coast to Lake Victoria, cutting right across Maasailand. Its construction was an expensive endeavour and to recover some of the costs, the colonial authorities set out to attract white settlers by making large tracts of land available for them, even if they were already occupied by local people [26]. After initially focusing on agricultural land around the upcoming Sustainability 2018, 10, 4156 6 of 16 administrative centre of Nairobi, interest soon shifted to the rich pasture lands of the Rift Valley occupied by the Maasai [26,27]. The British treated the Maasai somewhat differently from the other ethnic groups, in the sense that they decided to negotiate a treaty with them rather than force them out from their land and pay nominal compensation. With the treaty, the British wanted to free up the rich lands in the central Rift Valley by moving the Maasai to a new, exclusive Maasai reserve. The special way in which the Maasai were treated deserves some attention as it is relevant for understanding the hostile relationship between the Purko and the Loita Maasai.
Some of the Maasai sections in the northern part of Maasailand, particularly the Purko, already had a history of collaboration with the British before the treaty negotiations took place [28]. When the British first arrived in the area, they invited Maasai warriors from these northern sections as mercenaries to participate in punitive expeditions against other peoples, rewarding them with livestock [28,29]. The alliance that ensued between the British and the northern Maasai sections, particularly the Purko, was mutually beneficial: 'The Maasai, badly hit by the human and animal plagues of the 1880s and early 1890s, needed time to recover their stock and to reorganize their society. The British, hampered by lack of money and troops, and in a weak position, could not afford to antagonize the Maasai who controlled their lines of communications', i.e., the railway [28] (p. 553). The British developed a close relationship with a particular Maasai leader, Olonana, who as a laibon ('seer', prophet, diviner) played a key role in ritually sanctioning the warriors' cattle-raids. The British often approached the laibon Olonana to recruit warriors for the punitive expeditions [28]. In 1901, they conveniently made him Maasai 'paramount chief' [27][28][29][30]. Among the Maasai however, Olonana's authority was contested, not only as paramount chief but also as laibon. Olonana had been involved in a feud with his half-brother Senteu, also an important laibon. They both claimed to have inherited the position of overall Maasai laibon from their famous father Mbatian after he passed away [28,30,31]. This fraternal feud happened at a time when drought, and human and cattle epidemics triggered intersectional cattle-raiding and warfare and the two brothers, in their functions as laibon, became important leading figures [28,[30][31][32][33][34][35][36]. Senteu started to sanction the raids of the Loita against the Purko and other sections and became their chief laibon, while Olonana did the same for the Purko and others sections from the north. The competition between Olonana and Senteu resulted in the Purko and Loita becoming archenemies and this hostility dominated the intersectional Maasai war in the late nineteenth century. Initially, the Loita had been on the winning side. By 1895 however, Olonana convinced the British to join him in his fight against Senteu and his raiding Loita [28], and in 1902, under pressure from this powerful alliance, the Loita finally surrendered [28,33]. The intra-Maasai war ended with this surrender and widened Olonana's sphere of influence within Maasailand [28]. For the Loita, their defeat meant losing control over large parts of their territory in the western highlands: they lost the grazing lands on the Mara Plains [34,35] and remained with a much smaller territory comprising the Loita Hills and Loita Plains only.
So, when the British wanted to move the Maasai out of the central Rift Valley a few years later, they approached their former ally Olonana. Thanks to Olonana's support and his influence among the Maasai, the negotiations over the first Maasai Move went through relatively easily [27] and they reached agreement in August 1904 [26]. It is unclear, however, whether and to what degree the Loita were included in the negotiations: '[t]he Loitai were said to be represented, but none actually put their mark to this document' [27] (p. 34). The treaty created two Maasai reserves. The plan was for the Purko, the Loita and other Maasai sections to move north to a reserve on the Laikipia Plateau and the rest of the Maasai sections to a reserve south of the railway [26,27]. By creating the reserves outside the central Rift Valley [26,27], the Maasai effectively lost control of their best grazing lands. Despite the subsequent mass movement of Maasai to the new reserves, some stayed put and refused to move [27,32]. The Loita, for instance, remained where they were in the Loita Plains and the Loita Hills, an area that was not part of the newly created reserves [27,34,35]. Their presence there was tacitly accepted [34,35]. However, and even though they did not experience the distress of the imposed move themselves, they did experience the indirect consequences of the treaty when Purko Maasai started to arrive in the region. The first Maasai Move was particularly bitter for the Purko who lost superb pasture lands in their homeland around lakes Naivasha and Nakuru [26,27]. A portion of the Purko community refused to move north and between a third and a quarter (see Sandford 1919 in [27]) embarked on a southward trek across Mau Forest to uninhabited grazing lands between Narok and Mosiro [34,35]. This area abutted the Ewaso Ngiro river, and the Loita Plains lay just across it. Soon, Purko herders started to venture into the Loita Plains too. With the intersectional warfare fresh in their minds, these Purko families wanted 'to follow up the Loitai defeat and to replace them south of the Mau' [34] (p. 95). The Loita, still weak and recovering from the intra-Maasai war, started to lose ground, and by 1909, about 4000 Loita shared the Loita Plains with about 2000 Purko [34,35]. The archenemy of the Loita had become their closest neighbour-and their biggest threat.

The Second Maasai Move (1910-1913)
About five years after the first Maasai Move, and under pressure of incoming land-hungry settlers, the British once again reached out to Olonana with the aim of nullifying the northern reserve and moving all Maasai there to the south to an extended southern reserve so that the Laikipia Plateau could be opened up for white settlement [26]. To the surprise of the British, Olonana consented: he actually wanted to reassert his diminishing control over the spatially-divided Maasai [26,27,29]. Even before signing a second treaty, a big trek southward commenced in April 1910 marking the start of the second Maasai Move, which only ended three years later, after considerable resistance from some Maasai leaders [26,27,34]. Olonana passed away in March 1911 and his last words on his deathbed were that 'the (Laikipia) Maasai had to obey the Government and should move with their cattle to the Loita Plains in the south' [26] (p. 180). This last message was used to pressurize Olonana's successors to sign the treaty [26,27]. It was signed in April 1911 and exactly two years later, the last Maasai that moved from Laikipia arrived in the southern reserve [26].
For the Loita, the second treaty meant that their territory now officially belonged to the new extended Maasai reserve in the south. The extension not only included the Loita Hills and Loita Plains, but also the Mara Plains and later it would incorporate the Trans-Mara further west too [32]. The extensions were made to accommodate the Maasai and their herds from the northern reserve [26]. By 1910, 10,000 head of cattle arrived to the Narok area [27]. A number of Keekonyokie Maasai that had stayed behind in the Lake Naivasha area also arrived and settled on the Loita Plains (Sandford 1919 in [27]). They were joined by 10,000 Maasai from Laikipia, mainly Purko, whom arrived with 200,000 heads of cattle and more than double the number of sheep (Sandford 1919 in [26,27]). The Purko soon started to take over the highlands west of the Rift Valley: '[t]he Purko, spreading outwards from where they had been brought by the Moves, occupied the largest area, stretching south and west from Narok to the Mara River and the Loita Hills and across the border into Tanganyika' [34] (p. 96). The arrival of the second wave of Purko forced the Loita to leave the Loita Plains for good: they retreated into the relatively inaccessible Loita Hills.

Between the Moves and Independence (1963)
After the second Maasai Move, the British embarked on integrating the new Maasai reserve in the emerging administrative structure of the colonial state. They started by drawing administrative district and location boundaries. In 1913, the Maasai reserve was divided into two districts, Narok and Ngong (the latter was later renamed Kajiado District) [26]. Locations often followed the territorial occupation of Maasai sections [34]. The Loita were caught in a weak and vulnerable position. They continued to face territorial pressure from the numerous and politically well-connected Purko Maasai and little by little they also lost control of the northern Loita Hills. It was at about this time, in 1928, when their territory, covering the southern Loita Hills, became a separate location called Loita Location [37]. The creation of Loita Location had two effects. One the one hand, it confined the Loita to the southern part of the Loita Hills and weakened their claim to land that before had belonged to their territory, such as the Loita Plains and the northern Loita Hills, which were now occupied by Purko families. On the other hand, it somehow slowed down Purko encroachment and the territorial boundary between the two groups seems to have stabilized, at least until around independence in 1963.

The Strategic Appropriation of Interventions
It was around independence that outside development interventions and initiatives started to make their presence felt in Loita. The arrival of improvement interventions reflected a much earlier refocus of colonial state policy that started in the second half of the 1940s, away from the pursuit of political control and administrative order towards the 'development' of the country [26,38]. But the state was not the only body driving development interventions. Others, such as missionaries, and later NGOs and environmental organizations, became important development actors in Loita. A pattern emerged whereby development actors would approach the Loita leaders, believed to be the representatives of the Loita Maasai, with their proposals and projects seeking acceptance and support. This created space for Loita leaders to deliberate on these proposals, and act on them. All this coincided with a revival of the Loita-Purko territorial struggle. In 1961, the Maasai Mara National Reserve was gazetted and this resulted in the eviction of the Purko who occupied the Mara Plains, whom slowly started to push the western Purko-Loita boundary eastwards [39]. Purko encroachment became an acute problem again and the Loita realized that they needed to defend their territorial boundaries against further dispossession and appropriation by the Purko to secure the natural resource base for their semi-nomadic pastoral livelihood. Faced at the same time with increasing visits from outside actors to implement development projects and programmes, the Loita leaders (particularly chiefs and councillors) responded by trying to use these interventions to address the problem of Purko encroachment. This section describes how intervention after intervention became entangled with the ongoing territorial struggle, as leaders negotiated and tried to rework each intervention with the aim of getting the upper hand in this struggle.

The Leshuta School (1967)
Missionaries were important developmental players in the field of education [38,40]. Around independence there was only one school in Loita and the teachers were African Inland Church missionaries [41]. The school had few pupils and closed during World War II [42]. King [43] (p. 406), who surveyed the state of education in Narok District, described Loita as 'the most remote and least-schooled sector of the Maasai'. This is in line with the way the Loita, at the time, thought about sending a child to school. It was not considered to be a benefit or something useful or wanted. Instead, it was generally considered to be an insurmountable loss to the family as if the child had died.
After independence, the Catholic Church proposed to construct a second school in Loita. Loita leaders only consented after it was agreed that they would choose the place where it would be constructed: "In 1967 the Loita allowed the Catholic Church to build a second primary school. The Loita councilors decided that Leshuta would be its location. It was situated near the border zone with the Purko. It was a strategic choice. The Loita school was to stop the further intrusion of Purko into Loita" [41] (p. 102).
Loita leaders agreed to the school, not because they wanted schooling for their children but more because they saw it could be used as a way to deal with the problem of Purko encroachment. They hoped that a permanent structure in the form of a 'Loita' school-and not a 'Purko' school-would function as a physical barrier. However, their strategy failed in an ironic twist of fate. Unlike among the Loita, schooling was popular among the Purko, and Purko families were drawn to settle in the vicinity of Leshuta to allow their children to attend the school [41]. The area became more populated with Purko than Loita and eventually became Purko territory. Today, the Leshuta school lies in Purko land (see Figure 1). Although their plan to stake out a claim to land by using a development intervention had failed, the experience did set the stage for future action. Using interventions (whether in the name of development, nature conservation or any other well-intended goal) would become a recurring strategy employed by Loita leaders to prevent further land loss.

The Ilkerin Development Project (1968)
The first local NGO in Loita was the Ilkerin development project. It was initiated in 1968 with Dutch donor money through development agency Cebemo (renamed Bilance and then again as Cordaid). The Ilkerin development project originated out of the friendship between the Loita chief at the time, who was also a prominent laibon, and a Dutch missionary from the Catholic Mill Hill Society [41,44]. When the missionary visited the chief in the late 1960s, he told the chief that the Purko had refused to accept a development project that he had proposed because they did not want to set aside land for it [41]. The chief must have seen a window of opportunity to address the problem of Purko encroachment, and suggested that the Loita would take over this project. This is interesting because the chief actually had a reputation of being a 'stumbling block' for development in Loita [44] (p. 17). But the choice of the project site was telling. Together with the rest of the Loita leaders it was decided to locate it in the west of Loita, on the slopes of the Ilkerin Hills, hence the name 'Ilkerin development project' [44]. The missionary had actually advocated for a place in the centre of Loita called Entasekera, which is next to the Naimina Enkiyio Forest and, therefore, with permanent availability of water [45]. The Ilkerin Hills are in the drier lowlands, in the same border area where the Leshuta school had been constructed, an area, as we have seen, that was actively being encroached on by Purko. The lowlands are good for wet-season herding but in the dry season they lack water [44]. Loita Maasai's use of it was, thus, seasonal and settlement temporary and therefore they had weak claims to the land [44]. They were afraid that the neigbouring Purko would slowly take over these valuable wet-season pastures.
"Around 1970 the process of encroachment by the Purko Maasai into the areas where traditionally the Loita Maasai grazed their livestock had reached not far West from the Ilkerin hill. By staking out the boundary with a project, the Loita leaders hoped to prevent the loss of more grazing grounds to the Purko" [44], see also [39] (p. 18).
The reasons for choosing the Ilkerin Hills as the location of the missionary's new project were exactly the same as for the choice of the site for the new school. A strong claim to land can be made by building permanent structures [46]. And indeed, during a meeting between the missionary and the chief and other leaders, it was explicitly agreed to construct offices, staff houses, classrooms and a cattle dip [41,44]. The Ilkerin development project was eventually constructed according to plan, and has since succeeded in its unofficial goal of stopping Purko appropriation of the western side of Loita territory (see Figure 1). The construction of dams, both at the Ilkerin development project and in the vicinity of the Leshuta school, solved the issue of water in the area.

The Group Ranch Scheme (1970s-1980s)
Towards the end of the 1960s, the government of Kenya introduced a World Bank-funded land adjudication programme that was designed to create group ranches in Kenya's pastoral lands such as Maasailand. Land adjudication involved identifying and demarcating the boundaries of tracts of land to become group ranches and registering the resident people who would then become the group ranch members. Boundaries and membership would eventually be legalized. Policymakers hoped that the tenure security that group ranches afforded would incentivize group ranch members to invest in land and livestock. The Maasai, however, saw them mainly as a way to avoid losing more land. The promise of tenure security appealed to the Maasai, who, facing threats of land appropriation from within and from outside their communities [26,47,48], hoped that group ranches would secure their collective access to land. External threats included the creation of protected areas by the state and the steady infiltration by the agricultural Kikuyu and Kamba. Internal threats involved the appropriation of the best grazing lands by some wealthy and powerful Maasai seeking to establish individually-owned ranches. The first Maasai section that accepted the group ranch scheme are the Kaputiei of Kajiado District, and it seems that they did this primarily because they wanted to stop losing more land to their southern neighbours, the politically well-connected Kisongo Maasai [26]. The Loita turned out to be very similar to the Kaputiei in this regard.
The Loita accepted the group ranch scheme in the 1970s. The motives were twofold: they feared more land loss to their Purko neighbours and they were also worried about losing prime land to powerful Maasai individuals (from within and from outside the Loita community) seeking private title, something that was occurring in other parts of Maasailand (see also [41,44]). Acceptance was a 'defensive strategy' [48,49], i.e., a way of legally securing the boundaries with the Purko in order to prevent future land losses to them. Accordingly, it was decided to form one group ranch encompassing the whole Loita territory [44]. However, when district authorities rejected this proposal because they considered that the area was too large to be managed effectively, Loita leaders discussed an alternative of three group ranches [41,44]. But the process of land adjudication halted at this stage because the Purko and the Loita could not agree where the perimeter boundary would pass [44]. Loita interlocutors accuse the Purko of boycotting this process because demarcating the Loita-Purko boundary would put a final stop to their ongoing encroachment on Loitaland. They claim that, ultimately, the Purko wanted to push the Loita out of Kenya into Tanzania so that they could appropriate the valuable forested Loita highlands. One Loita interviewee recalled how he was told by a Purko politician: 'We have pushed you from the Ewaso Ngiro river to Narosura, and then we pushed you over the escarpment, and now we are still pushing you [all the way to Tanzania]'.
A renewed attempt to adjudicate the Loita into group ranches was made in the 1980s only that this time it was decided to form two group ranches. Once more, land adjudication failed, but this time it was not due to disputes with the Purko but due to quarrels within the Loita leadership about where the internal boundary should lie [23].
Loita territory was never cut up into group ranches. In fact, Loita is quite exceptional in this regard as it is the only Maasai section that has not (yet) formalized land tenure, either in group or individual holdings. It is often popularly asserted that this is due to Loita Maasai resistance to changing their traditional way of life [25,42,[50][51][52][53]. However, my data reveals a willingness to demarcate Loita land formally but recurrent failure to do so because of unresolved Purko-Loita and intra-Loita boundary disputes (see also [41,44]).

The IUCN Forest Project (1998-2005)
The Leshuta school, the Ilkerin development project and the group ranch scheme had all been proposed in the name of development. But when attention shifted to the Naimina Enkiyio Forest in the east of Loita territory, a different language started to be used by intervention actors to justify their plans and projects, namely that of biodiversity conservation and sustainable forest management.
In 1998-1999 preparatory work was carried out by the environmental organization IUCN to develop a community-based forest management plan for the Naimina Enkiyio Forest (information on the IUCN forest project comes from [24,46,52,[54][55][56]). The initiative for this project had actually come from a group of Loita leaders organized around the Ilkerin development project, who invited IUCN in reaction to a plan by the local district authority, the Narok County Council, a Purko-dominated body, to establish a nature reserve [52,57]. The Loita feared that they would lose access to this valuable resource if the local authority would assume control of the forest, and vehemently opposed this plan. By bringing in a strong international player such as IUCN and developing a project backed by foreign donor money (first by the same Dutch donor of the Ilkerin development project, and later by the European Commission), they hoped the local authority would be deterred from moving ahead with its nature reserve plan.
The politics and layered nature of the conflicts surrounding these two forest interventions are complex and have been discussed elsewhere [23]. Of relevance here is that a group of Loita leaders mobilized the IUCN forest project not only to ward off the local authority's forest plan but also to address the problem of ongoing Purko intrusion. The failure to demarcate and formalize the perimeter boundary of Loita territory during the group ranch discussions meant that steady Purko Maasai encroachment had continued. The Narosura Escarpment to the north of Loita had been the natural boundary between the Loita in the Loita Hills and the Purko on the Loita Plains when the administrative Loita Location (which was later upgraded to 'Loita Division', and again, after the promulgation of the new 2010 constitution renamed 'Loita Ward') was demarcated in 1928. However, by the time the IUCN forest project took shape, a number of Purko families had climbed the Narosura Escarpment and settled in the area surrounding the northern tip of the forest. So, although these Purko families officially lived in Loita Division, they had in fact pushed the territorial boundary between the Loita and the Purko sections further to the south and to the disadvantage of the Loita. These Purko families exploited their strategic location (i.e., living at the northern edge of the forest and next to one of the two roads leading into the Loita Hills) and came to monopolize the little forest-based tourism taking place in Loita. Loita tourist guides resented this, because Purko guides would take tourists deep into the forest into forestland claimed by the Loita. Eventually, the Purko entered a lucrative agreement with a tour company, and at one point they benefited from 60% of the area's total tourist revenues, while they only constituted 1% of the population in Loita Division [25,55,58]. This inequitable distribution of benefits has been another source of contention between the Purko and Loita [25], and only added to their territorial hostility. It also worried the Loita that the Purko were now targeting their precious forest (see also [57]).
With these issues in mind, the Loita leaders, when working out the forest project together with IUCN, made sure that the forest was being referred to as the 'Loita Forest' rather than using the Maa name Entim e Naimina Enkiyio (Forest of the Lost Child) after a well-known Maasai legend. Consequently, the project was officially called the 'Loita Forest Integrated Conservation and Management Project', and in the early project documentation the forest was consistently referred to as the 'Loita Forest' (see for example [55]). In this way, Loita leaders ensured that the Loita (and not the Purko) were defined as the beneficiaries of the project. Moreover, calling the forest 'the Loita Forest' clearly dismissed the Purko claim over the forest's northern tip. This strategy was successful at first, but backfired in the run-up to the implementation phase of the IUCN forest project. The Purko families in the north protested against their exclusion claiming they were forest users too, and IUCN, whose policy was to promote equitable, participatory and inclusive forest management, agreed to incorporate them as beneficiaries. These Purko families (not the whole Purko section) thus became 'stakeholders' in the forest project. This change was evident in the way IUCN staff and consultants started to write about the forest and the project. The forest was now being referred to with the more impartial 'Loita/Purko Naimina Enkiyio Forest' (see for example [25]) and the project was renamed the 'Loita/Purko Naimina Enkiyio Forest Integrated Conservation and Development Project' [56,59]. So, even if the Loita leaders succeeded in averting the local authority's forest reserve plan by initiating the IUCN forest project, they failed to address the issue of Purko encroachment.
In the end, however, the implementation phase of the IUCN forest project never went ahead. For reasons that are beyond the scope of this article (but see [23]), Purko resistance to the IUCN forest project increased and also in Loita, people started to oppose the IUCN forest project. Opposition became aggressive and violent. In 2004, an angry crowd of mostly Purko (but also some Loita Maasai) headed to the Ilkerin development project where the IUCN headquarters were housed. Eyewitnesses insist that the mob carried bows and poisoned arrows and also brought containers with petrol to burn the buildings. It is also said that Ilkerin development project staff, who had been warned about the approaching group, had called the police in anticipation of clashes. Policemen fired when the crowd ran towards the project buildings, killing one person and injuring another [57,[60][61][62]. This averted a major clash but the gravity of the situation forced IUCN to halt the project. In 2005, IUCN wanted to resume the forest project, but conflicts broke out again and they decided to pull out permanently and, as one interviewee lamented, they took with them the US$2.56 million set aside for the project.

Discussion
This article has considered how Loita leaders have cautiously negotiated and creatively reworked-even initiated-governmental interventions with the strategic aim of keeping Loita land in Loita hands. The ongoing and long-running territorial struggle with their neighbours, the Purko Maasai, is crucial for understanding this. We have seen how this territorial struggle has its roots in past hostilities that had more to do with cattle-raiding and laibon rivalries, and less with territory and land. It was the British who gave this existing animosity a territorial twist, by bringing the Purko closer to the Loita through the enforced Maasai Moves during the early colonial period. The Purko, themselves having lost their homelands in the central Rift Valley and later also the Laikipia Plateau and the Mara Plains, slowly started encroaching on Loita territory and eventually took over large areas of land that used to be controlled by them. Purko pressure at the fringes of Loita land has continued ever since. For the Loita, Purko encroachment effectively means land loss and a shrinking territory, while the Loita population is growing and far-reaching land-use changes have taken place (from pastoralism to agro-pastoralism) [63]. It is in this context that maintaining collective access to land and securing the territorial boundaries becomes a critical concern. After all, the Loita semi-mobile agro-pastoral livelihood depends on access to the natural resources that their territory affords.
To address the problem of Purko encroachment, Loita leaders mobilized and used interventions in clever ways. Two interlinked strategies can be discerned. First, leaders used interventions to stake out spatial claims to land. In the case of the Leshuta school and the Ilkerin development project, Loita leaders clearly negotiated the location of these interventions by demanding that the buildings be constructed near the territorial boundaries with the Purko. The locations did not make sense from an interventionist point of view, simply because these were areas with no water during the dry season. Yet, this was precisely the reason for the Loita leaders to have them there: the Loita had weak claims on those lands because no permanent settlement was possible there due to the seasonal lack of water. In this context, the school and the project buildings were meant to function as 'territorial markers' [64] and Loita 'flag posts', i.e., visible and permanent claims to the land demarcating the Purko-Loita boundary. Similarly, the state-led group ranch scheme was accepted because it offered the opportunity to formalize the perimeter boundary of Loita territory, and in this way, it was hoped, to secure it from Purko encroachment. The Narok County Council plan to convert the Naimina Enkiyio Forest into a nature reserve was rejected outright by the Loita majority, an underlying reason being that it was seen as a Purko-orchestrated attempt at taking over this valuable resource. In response, (a group of) Loita leaders invited IUCN to develop a forest project, not necessarily because they wanted to conserve the forest's nature, but to strengthen the Loita Maasai claim to the forest and avert both the Narok County Council's forest plan and Purko appropriation of the forest, something that had already happened with the northern tip of the forest.
The second strategy relates to the inherent tendency of governmental interventions to simplify complex realities, and particularly the inclination towards identifying a clear and apparently homogeneous beneficiary community [1]. Once this happens, the purview of the intervention narrows down, and in the process, loses sight of neighbouring non-beneficiary communities with whom beneficiaries may compete with over use and access to resources. Loita leaders, keenly aware of this tendency after years of dealing with successive intervention actors, have strategically capitalized on interventions' simplifications and singular perspectives. The trick was to ensure that interventions became 'Loita' projects-not 'Purko' projects. As discussed in the introduction of this article, the identification of beneficiary communities is typically based on singular livelihoods or social groups identified along the lines of gender, generation, ethnicity, etc. Although the Loita shared the same livelihood with the Purko, i.e., (agro-)pastoralism, they were clearly distinguished from them along socio-spatial lines through their sectional identity. Thus, it was easy for Loita leaders to ensure that 'the Loita Maasai community' became the beneficiary of intervention.
The experience with IUCN is instructive. Initially, Loita leaders succeeded in ensuring that the Loita were identified as the target group. But after Purko protest, the IUCN forest project was forced to include the Purko families living around the northern edge of the forest, a change most visibly clear with the renaming of the project. These families had been initially non-beneficiaries. By including them into the project, the IUCN forest project employed, in effect, a 'social landscape approach', i.e., an approach that recognizes the multiplicity of user groups even if the relations between them are conflictive. This is the approach advanced by Soeters and Zoomers [1] (see also see [2]) as a way of mitigating the risk of conflict, but what the case of the IUCN forest project shows is that, on the one hand, this came too late: the conflictual relation between the Loita and the Purko had already been aggravated by the intervention's initial approach, leading eventually to deadly violence. On the other hand, no matter what approach the IUCN project would take, its very existence was entangled with the Loita-Purko territorial hostility from the start through the strategic actions of the Loita leaders.
This final point became clear during the violent clash at the offices of the Ilkerin development project. The Purko-dominated group did not only attack the Ilkerin development project because IUCN was housed there and they wanted the IUCN forest project to stop. Their intention, it seems, was to burn down the whole Ilkerin development project infrastructure. This aggression by a Purko-dominated mob towards a Loita-based NGO is expressive of the century-old Loita-Purko struggle over land and territory. The Ilkerin development project, as we have seen, unofficially represented Loita Maasai success in stopping Purko encroachment on that side of their common boundary. Loita leaders had achieved this by ensuring that the Loita became the beneficiaries of the Ilkerin development project and, indeed, for many years the Loita benefited greatly from its programmes-to the exclusion of the Purko even if they lived practically next to the project site. Now with the IUCN forest project, the Loita, under the leadership of the Ilkerin development project, had tried to do the same. Reminiscent of their Ilkerin development project strategy, they had tried to make the IUCN forest project an exclusive Loita Maasai project and even though it now included the Purko families along the northern edge of the forest, it still excluded the remainder of the Purko Maasai and continued to be a heavily Loita Maasai-influenced project. If it is true what my Loita interviewees say, i.e., that it is the Purko's wish to displace them from the valuable Loita Hills, then the 2004 attack of the Ilkerin development project and the IUCN forest project, the former an established and the latter a potential barrier to their encroachment, can be seen as an attempt at paving the way to achieve this.

Conclusions
Loita Maasai leaders used successive interventions to stake out territorial claims, by strategically capitalizing on interventions' inherent tendency to simplify. They did this in an attempt at strengthening their position in a long-running territorial conflict with their neighbours the Purko Maasai. Even if this was not always successful (the Leshuta school for example was not able to stop Purko encroachment), the interventions did become deeply implicated in this struggle. Therefore, they at least perpetuated and maintained an old conflict, and at some point, even contributed to its violent escalation.
The contribution of this article to the topic of this special issue (the linkages between (adaptation) interventions and conflict in African drylands) is that it highlights the agency of local people and particularly that of beneficiaries. This was done by building on the social landscape approach, which recognizes wider relational dynamics of (competing) resource use, claims and access, but by reversing the perspective of analysis and taking local dynamics and local people's challenges and experiences (including experiences with previous interventions) as the point of departure. Such an analytical angle helped to disclose that beneficiaries are neither passive nor submissive recipients of interventions. They actively engage with them, and, if necessary, change them to suit their own interests. Such interests may not have anything to do with the situation that the intervention seeks to improve. In fact, interventions may be strategically used and reworked to address a completely different problem from the perspective of the beneficiaries, such as an ongoing conflict with another social group. In situations of conflict and struggle, interventions may be mobilized to strengthen the beneficiaries' position. Thus, interventions may unwittingly reproduce and even aggravate existing conflicts, or help the beneficiaries to move to the winning side. In this way, interventions may weaken, marginalize and jeopardize non-beneficiaries' livelihoods.
Finally, the analysis presented here adds historical depth, particularly by recognizing that any intervention is part of a long list of interventions that have already been proposed or implemented in a locality. Local people integrate passed experiences with interventions and their response to any new intervention is shaped by this legacy. If the aim of this special issue is to understand how and why interventions (re-)trigger conflicts, this needs to be taken into consideration. The case of the Loita Maasai confirms that it is indeed important to look 'beyond beneficiaries'. But it also shows that a social landscape approach could greatly benefit when it also looks back in time beyond any particular intervention.