Making and Unmaking The State In The Horn Of Africa: A Deep-Time Perspective

2.1. Nomads against infrastructures

While fixation and stability are not an absolute prerequisite for the emergence of the state and nomadic states are known throughout history (Honeychurch 2014), the fact is that states prefer fixing population to the land and most have arisen from settled communities. This makes sense, because mobile communities are, from the perspective of sedentary states, unpredictable, notoriously difficult to police and epistemically illegible (Scott 1998)—and often dangerous. They can also be seen as ambivalent actors, bringing to urban-based states both a new sense of cohesion, when they conquer those states, and factionalism leading to their demise, when they settle in—as theorized by fourteenth-century scholar Ibn Khaldun (Schlee 2002). In the case of the Horn, both perspectives can be found: the sense of danger and unpredictability associated with nomads is well conveyed by written sources of the sixteenth century, a period in which pastoralists wrought havoc to the regional states and collaborated in the collapse of a few (Abir, 1980, Hassen, 1995). The case of the Banadir coast (Somalia) offers a good example of nomads becoming urbanites and eventually bringing factionalism and instability (Cassanelli, 1982). The desire to live outside state structures does not mean that nomads are always outside the state: on the contrary, the most common situation is one of symbiosis, in which pastoralists living in internal or external margins engage in symbiotic relations with settled communities and state actors (Khazanov 1994).

The success of many nomadic communities in escaping effective state control has both to do with their mobility and with the arid environments that they occupy and that are inimical of densely settled populations and the control of one route by a polity can be easily circumvented by the opening of other routes. A nomadic ethos and an arid environment partly explain the difficulties met by the state to take root in the northern part of the Somali peninsula. Meaningfully, long-distance trade, which is one of the elements more commonly associated with the emergence of stratified, urban-based polities, has instead failed to sustain them in the region for any long period time. Thus, the sultanates of Awdal and Ifat, which developed as urban-centered states, deeply enmeshed in long-distance commercial networks, lasted barely three centuries, between 1285 and 1577 (Chekroun and Hirsch 2020). Before the thirteenth century, long-distance connections were sustained by stateless nomadic communities.

Thus, unlike in the rest of the Indian Ocean world, the peoples of the northern Somali seaboard, conducted trade for three millennia mostly outside urban centers and outside state formations. In fact, nomads in the region have consistently refused both infrastructures and external control. They have been aware of the dangers of infrastructure, as noted in the introduction: of their capacity to fix points of trade and have instead opted for ephemeral structures.

In the case of the northern Somali peninsula, instead of towns with port facilities, as it is the case in the rest of the Western Indian Ocean, what prevailed for millennia were beach markets: open, seasonal spaces where nomads from different clans and caravanners sent by inland towns gathered with foreign merchants. Coastal fairs have been documented going from the first century BC/AD to the mid-nineteenth century (Chittick, 1979, Smith and Wright, 1988; González-Ruibal and de Torres, 2018). The archaeological record of such fairs is very different to ports of trade known elsewhere: there is very little in the way of infrastructure (no piers, warehouses, customs buildings). What is found are empty spaces littered with broken pottery, bracelets and beads from China, Southeast Asia, India, Iran, the Arabian Peninsula and Egypt that were part of the commodities traded in the beaches, as well as remains of campsites where nomads and caravanners dwelled during the trade season and that have left flimsy traces (animal bones, hearths, heaps of ashes)—for a detailed discussion of the commodities that were being traded in the medieval Horn see González-Ruibal et al. (2021); Insoll et al. 2021). The contrast could not be greater with other hubs of the Indian Ocean network such as the densely-built, state-controlled island of Suakin (Sudan), the port of Aden or any of the Swahili stone towns in Kenya and Tanzania (Fig. 2). Even during the heyday of the state in the Middle Ages, under the sultanates of Ifat and Awdal, open places without solid infrastructures articulated long-distance commercial interactions.

Beaches provided excellent qualities: they are natural interfaces between land and sea, and the large, flat and open terrain allowed for the presence of a huge number of caravans, merchants and nomads with their herds. The beach fairs could be quite volatile. As there are many potential mooring places along the coast and little infrastructure to anchor trade, markets could move elsewhere in the blink of an eye.

However, this volatility is not necessarily obvious in historical texts. In them, points of trade along the coast tend to appear associated with a few specific places, often the same for millennia. In the Periplus of the Erythraean Sea, for instance, a travelogue of the mid-first century AD (Casson 1989), the author mentions six ports along the Somali coast. Yet several more have been located: they can be easily spotted as concentrations of tumuli, built by pastoralists to bury their dead in the places where they traded. In some sites, such as the region of Xiis (usually identified with ancient Mundu), the nomadic necropolises are not limited to a single spot, but extend for dozens of kilometers along the coast, thus showing that places of trade shifted over the years, probably depending on the availability of local resources and clan politics (González-Ruibal et al. 2022).

The situation was similar in the Middle Ages and later. Berbera offers good illustration of how they operated. The town is today a commercial node in the Red Sea and one of growing regional and suprarregional importance (Stepputat and Hagmann 2019). Most authors have assumed that it was a town in the past, as it is mentioned in medieval and earlier sources. However, Berbera originally referred to a region (Barbaria) and to a people (Barbaroi), not to a specific settlement or port. It was still described as a region and a people as late as the fourteenth century: for Arab traveler Ibn Battuta, Berberah was a “population of Blacks” whose capital was Zeila, not Berbera (Fauvelle-Aymar et al., 2011: 65). And even after Berbera solidified as a town at some point in the fifteenth century, trade was still conducted outside it and at times at some distance, which shows that the state probably did not have control over the transactions. Proof of that are the eight sites that we have documented in a radius of 30 km around Berbera and that were used for long-distance exchanges between the mid-first millennium AD and the nineteenth century (González-Ruibal and de Torres, 2018; González-Ruibal et al. 2021). Some of them were very small, an artefact scatter covering a few hundred square meters and a cemetery with some tombs, meaning that nomads were able to attract merchants away from the main centers of trade, possibly avoiding competitors, state control, taxation or other clans. One of these small sites in the vicinities of Berbera is Ceel Gerdi, a shell midden of less than 0.3 ha, dated around 500–700 CE, where a tiny nomadic community spent time tending their herds, fishing, hunting, collecting mollusks… and trading with the Byzantine, Sasanian and Aksumite empires, either directly or indirectly (González-Ruibal et al. 2022). Archaeology shows that Berbera itself did not enjoy a stable commercial monopoly and that fairs shifted from place to place. Lacking permanent infrastructures, there was nothing that tied nomads to a specific locale: as long as there was a beach and water, there could be a trading place (Fig. 3).

1. Download: Download high-res image (764KB)
2. Download: Download full-size image

Fig. 3. The beach market of Siyaara, some 30 km from Berbera. Despite being used for almost 800 years (eleventh-nineteenth centuries), it developed virtually no infrastructures.

---

Nomadic fissiparous politics explain this shifting geography. By way of example, in Berbera, in 1847 a clan dispute occurred that changed commercial networks in the region (Cruttenden 1849: 51): one clan moved to Bulhar, a small village some 30 km west of Berbera, and some traders followed them. This would have been impossible had a large investment in infrastructure been made in the other town or had Berbera been the only available roadstead in the region. When we excavated Bulhar we found that the place had already being used for trade before 1847, as shown by Asian market wares characteristic of the late eighteenth and early nineteenth centuries. The site revealed repeated seasonal occupations and numerous imports, evincing that nomads had been successful attracting trade away from Berbera for over a century (Torres et al., 2017). An attempt at fixing these growing exchanges was undertaken by colonial authorities since 1874–1875, first the Turco-Egyptians and then the British (Akalin 2014: 13), who created several infrastructures, including a lighthouse, a fort, a customs building and port facilities. The initiative was initially successful and the town became a port of trade absorbing a steady volume of goods (Pankhurst 1965), but commerce declined in the 1930s, the town was eventually abandoned in the 1950s and today is a heap of ruins—another proof, if need be, that heavy investment in the built environment does not necessarily lead to increased stability (Bachmann and Schouten, 2018). Nomadic occupations resumed as if nothing had happened and in very much the same way.

Similar failed attempts at fixing trade are more recent: Stepputat and Hagmann (2019: 803-804) refer to the construction of a modern market in the border town of Tog’wajale to both foster and better control cross-border exchanges. The project was largely unsuccessful, as traders simply moved to the other side of the border in Somaliland. Even during the colonial period, the ports supported by colonial authorities and where the greater infrastructural investment was made (Zeila, Bulhar and Berbera) had to compete with many lesser ports that operated with no infrastructure at all: Karin, Cankhor, Maydh, Waqderia, Laas Qooray, Durdureh, Bandar Meraya, Bandar Siyaada, Bandar Qasim, Bandar Khor, Ras Caluula, and Ras Asir (Pankhurst 1965: 61). The situation in the early twentieth century was thus not so different from the first. These ephemeral, shifting, anti-infrastructural ports can be seen as a form of long-standing “subversive logistics” (Shell 2019), which is part of a larger “subversive mobility” (Shell 2015; Schouten 2019) that includes nomadic routes and pirates (Dua, 2019). That this subversive mobility and anti-infrastructural ethos is not just unconscious or the product of external interpretation is demonstrated by an event told by traveler Richard Burton who visited the Somali peninsula in 1854: the Ayyal Ahmed clan had started building a small stone tower near Berbera, which the builders intended to use for storing grain, “but the suspicious savages, the Eesa Musa, and Mikahil [other clans], who hold the land, saw in it an attempt to threaten their liberties” (Burton 1910: 280). Even the most innocent form of permanent architecture is seen as a potential political menace.

The nomadic ecology of the northern Somali territories cannot be reduced to the two variables of mobility and anti-infrastructural ethos. If nomads could prevail for so long and develop their politics against infrastructure it was because they were allied to a specific topography, precipitation regime and winds. The northern Horn is divided into three zones: a coastal plain (Guban), a mountain range parallel to the coast descending into a high plateau (Gollis/Oogo) and an inland plateau (Haud); the different precipitation regimes makes grazing possible at different times of the year in different zones, but not sustained grazing in a single one (Lewis 1999: 32-42). It therefore dictates nomadic movements and prevents the fixation of herds: attempts at interrupting traditional mobilities, like those carried out by the British in the twentieth century, ended up in disaster, as the concentration of animals in certain areas depleted their scarce resources, intensified soil erosion and increased aridity (Mohamed 2004). Both precipitations and winds, in turn, depend on the monsoon, whose winds brought merchants to the coast and made long-distance exchanges possible. This complex ecology involving ships, herds, deserts, luxury commodities, local raw materials, distant cities, rains and winds worked in the Horn of Africa for over two thousand years, most of the time without state intervention or very little. This resonates with the present situation (Little 2003). Interestingly, the most developed infrastructural sector is telecommunications (Feldman 2007), which is the less materially heavy of all infrastructures. As in the past, Somalis today are able to partake in a globalized world with little resort to large-scale infrastructures.

2.2. Internal frontiers

The American West has deeply colored our understanding of frontiers: we imagine them as fronts of expansion, which, after a period of predation and unruliness, are finally incorporated into the State and “pacified”. Africa—as Igor Kopytoff (1987) demonstrated—offers different examples of frontiers, which are culturally and politically internal, not peripheral, to existing societies. However, in Kopytoff’s model, internal frontiers still imply the displacement of members of the already established polity to its spatial margins, where they create a new community, distinct, but symbolically tied to the original entity. The internal frontiers to which I refer here are of a different kind, inasmuch as they were not created by the fission and displacement of a group. Rather, they were created when a group invaded a space that was culturally and politically distinct, but such invasion did not lead to complete homogenization and incorporation of the indigenous communities. Instead, what I would call “zones of difference” were created that remained outside the dominant order—but linked to it—and that were inhabited by people not fully assimilated, but subjected to the state from a jurisdictional point of view, for example, via the payment of tribute. These zones of difference are in many ways akin to proper frontiers, in that they were often perceived by dominant society as unruly and wild.

Such zones exist across the Ethiopian highlands. Not surprisingly, they are particularly common among the Amhara and the Oromo, the two ethnic groups that have historically expanded and dominated Ethiopia (Levine 1974). The Amhara spread through the northern plateau from the thirteenth century onwards (Tamrat 1988); the Oromo conquered most of the southern plateau between the sixteenth and nineteenth (Hassen 1995). In both cases, internal frontiers were created that were never assimilated and that survive to our days, although under increasing pressure.

The way these zones of difference work is very similar: the people of the zones are the remnants of conquered populations which have retained elements of their social organization; their religion (or a religion other than the dominant one), and elements of their culture (including language), but subjected to strong hybridization with the dominant society and other hegemonic cultures. The minorities of the zone develop economic practices that are complementary with those of dominant society and often despised, such as blacksmithing, pottery-making, circumcision, healing rituals, etc. (Freeman 2003). Zones of difference include the Mao and Boro pockets among the Oromo (González-Ruibal, 2014, González-Ruibal et al., 2014) and the K’unfäl, Beta Israel (Ethiopian Jews), Wäyto and K’ïmant’ among the Amhara (Gamst, 1969, Quirin, 1998, Freeman, 2003). Whereas Mao territory lies on the edge of Oromo land, the groups living among the Amhara often occupy spatially central positions, despite having been incorporated six or seven centuries ago (Tamrat 1988), proof of the tremendous resilience of those zones of difference.

It comes as no surprise that many of these groups exist in challenging environments, including swamps (the Hozo and Seze Mao), high mountains (Beta Israel), escarpments, and forests. The difference of the zones is thus not only cultural, but also physical. It is indeed impossible to understand the political ecology of zones of difference without taking into account their environment and topography. As Scott (2009: 166) notes: “locations of very high friction—swamps, marshes, ravines, rugged mountains, heaths, deserts—even though they may be quite close to the state core as the crow flies, are likely to remain relatively inaccessible”. This is the case from Southeast Asia to the maroon communities of the Americas, which were often very close to state centers, but in inhospitable surroundings (Price 1979).

In the case of the Horn, the K’unfäl live in the isolated and steep Belaya Mountain, in the middle of the Metekel lowlands, and no all-weather roads still exist leading to the top, but they are still very close to the state core in the Amhara highlands—about 50 km in a straight line. The situation of the Boro, in Wämbära Mountain, is similar, although the road here was finally built in the late 2000s. While many Beta Israel and K’ïmant live interspersed among Amhara communities, a sizeable number used to inhabit the Qwara plateau¹, a massif detached from the northern Ethiopian plateau that still today can only be reached by walking over 30 km from the nearest gravel road and climbing 1300 m along an extremely steep slope. The Wäyto, in turn, have survived on banks of Lake T’ana and the Hozo Mao persist, surrounded by the dominant Oromo, in the impenetrable swamps of the Dabus River. Difficult environments have provided shelter for these communities, either because members of the dominant group were not interested in occupying them or because they are difficult to access—at times both. Simultaneously, they have provided ample economic opportunities for the minority groups. The Wäyto and Hozo specialized in exploiting fluvial and lake resources, including hippopotamus and elephant hunting, fishing, and papyrus and reed gathering (Cheesman, 1936, González-Ruibal, 2014, González-Ruibal et al., 2014) (Fig. 4). This was in the interest of members of the dominant society, who acquired those products from the people living in internal frontiers, but at the same time those zones were also open to predation, if need be. The Beta Israel, for instance, were victims of repeated attacks by Ethiopian kings, mostly during the sixteenth and seventeenth centuries which resulted in looting and capture of slaves (Kaplan 1992: 59-64), whereas the Mao were a source of forced labor for the Oromo until the twentieth.

1. Download: Download high-res image (836KB)
2. Download: Download full-size image

Fig. 4. Wäyto women making baskets. A typical example of internal frontier people, the Wäyto no longer hunt hippos, but engage in other economic activities that are complementary with the dominant Amhara society.

---

As noted above, these internal frontiers have in common with external ones their unruly quality. Despite the strengthening of the grip of the state during the last half a century, this unruly quality has far from vanished. This persistence does not have to be seen as a direct connection between past and present or the return of the past (demands have changed, as have political actors and technologies), but rather as long-term trends shaped by a combination of ideologies, concepts of the state and material affordances and limitations. Northwest Ethiopia provides good examples of enduring resistance in internal frontiers.

Thus, the K’ïmant’ have risen against the Amhara several times since 2015, bringing havoc to a large area by blocking roads and ambushing vehicles: in turn, Amhara peasants have killed dozens of K’ïmant and displaced thousands. What the K’ïmant reclaim is that their zone of difference is legally recognized as such (as a special administrative zone). The K’ïmant live in a strategic transit zone between Sudan and Ethiopia and their road blocking can have an effect disproportionate to their numbers and indeed their struggle has been enmeshed in the larger conflict between the federal state, the Tigray’s People Liberation Front and the tensions between Amhara and Tigrayans—a good example of how internal frontiers can contribute to larger state-wide destabilization.

Nevertheless, those who make the internal frontier ‘wild’ are more often than not members external to the minority groups. Zones of difference have been historically the home of brigands (shïfta), mystics, political refugees and aristocrats trying to build up power in the margins—the different characters are not always easy to distinguish (Crummey, 1986, Reid, 2011a, Reid, 2011b). All of them come from the state core, politically and culturally speaking, unlike the minority groups. The most notable example is that of Kassa Haylu, later Emperor Tewodros, who started his career as a highwayman in the Qwara mountains, which, as I have mention, was the historical homeland of a large Beta Israel community (Reid, 2011a, Reid, 2011b).

Brigands and disgruntled nobles can be regarded as a sort of state debris that is deposited in frontiers (both internal and external) only to come back with a vengeance (Caulk, 1984, Crummey, 1986). There are not many warring aristocrats today in the zones, but we have militias and terrorist groups fulfilling similar roles. In fact, if we are to find spaces in which the rule of the state is threatened, we should look for these zones of difference, which are not necessarily in the periphery nor inhabited by culturally-distinct groups, but by members of the dominant society and rather close to the heart of the state. Thus, Islamist armed groups thrive today in the Gollis mountains, in the northern Somali region, very near the main towns (Hohne 2014), where they threaten the stability of the regional states—Somaliland and Puntland. Although the struggle has taken the robes of Islamist terrorism, it emerged originally as a typical internal frontier conflict, with the state trying to exploit resources in the territory of a minority group, the Warsangeli—in this case a clan, not an ethnic group (Hoehne, 2014, Hoehne, 2015).

In the case of Ethiopia, some mountain pockets have successfully resisted the recent infrastructural onslaught of the developmental Ethiopian state, with its strong emphasis on road-building (Clapham 2018). They are so rugged and building roads there so expensive—and unprofitable in political and economic terms—that they remain largely unconnected. In many ways, the situation does not differ from the one documented centuries ago. Local lords still hold all the effective power, shifting between tactic compliance to the state, brigandage and all-out rebellion, and the regional administration is powerless to intervene. One of the few successful state interventions has been the establishment of cellphone towers (for which mountains are perfect places), but cellphones are used by local guerrillas and brigands to actually subvert state control, as they enable a better organization of members living in a disperse environment. Regions like Tach’ Armach’ïho or the Qwara highlands are not significantly more controlled by the state than they were two centuries ago. The situation in Tigray since November 2020, with the federal army stuck in a guerrilla war with the TPLF in remote mountain areas of the region, again demonstrates that the core of the state can become a core of resistance, that the rugged geography of the Horn can always be mobilized against attempts of state control, and that frontiers there are still as much internal as they are external.

2.3. Liminal ecologies

Some of the zones of difference mentioned above occupy what can be described as ambiguous terrain: spaces that are spatially liminal and ontologically undetermined—sharing aspects of different environments, while being irreducible to any of them. Liminality is not only a cultural construct, but a physical phenomenon—geological, geographic and biological. Liminal ecologies have often become a place of resistance: consider the case of the pirate shore: pirates benefit from the affordances of a space in-between land and sea, where they can make the most of both and that is not easily policed (Beloff 2013). Liminal terrain can be an obstacle for state expansion, but at the same time it also offers manifold opportunities to the people who live in it. Feyissa and Hoehne (2010) rightly note that borders are not only zones of constraint and conflict: they can also be seen as resources (political, cultural and economic) that can be tapped. What is true of political borderlands, also holds for natural ones. Liminal ecologies are rich in resources, as they provide access to a diversity of ecotopes. At the same time, liminal ecologies are difficult to grasp epistemically and hard to control physically. Zones where a variety of subsistence practices are possible—hunting, fishing, gathering—in a diversity of ecological settings, reminds James C. Scott, 2015, pose “insurmountable obstacles to the imposition of a single political authority”. It is not a coincidence that several of the case studies presented in this issue deal with this ambiguous terrain: lakes, swamps and wetlands (Schouten, Verweijen, Kubuya and Murairi; Bachman, Moro and Pendleton; Marijnen).

Excellent examples of liminal ecology are the long escarpments that flank the Ethiopian plateau to the west and east. The transition between the plateau and the lowlands is often short and dramatic: in Gambela, on the border between Ethiopia and South Sudan, the world changes physically and culturally in just twelve kilometers, which is the distance it takes for the mountains to collapse from 2000 m of altitude into the plain, 600 m above sea level. Some ethnic groups have ended up occupying this niche and making the most of its affordances: it is the case of the Raya and Yeju Oromo, the Argobba and the Agäw in the eastern escarpment (Trimingham, 1952: 223; Simoons 1960), or the Majangir, Chabu and Mao in the west (see below). People in the escarpment have used their geographical position for centuries to avoid being ruled by dominant societies, while at the same time interacting with them for their own profit: the Argobba, for instance, were key in facilitating the trade between the eastern lowlands and the Ethiopian plateau. The lowlands in general have been avoided whenever possible by members of dominant societies living in the highlands, whose bodies were not adapted to the extreme heat and the manifold diseases of the plains. Weather, insects and bacteria, thus, were as much part of the tactics of resistance of the local communities as their bows and arrows, spears and—today—Kalashnikovs. In most cases, thus, regional states opted for a political economy of predation, consisting in seasonal raids, instead of the permanent appropriation and management of the peripheral lands (Ahmad, 1999, Endalew, 2006). Meaningfully, the first attempts at effective, permanent control by the state, which in some places are as recent as 2004, materialized in roads (to enable access to the border to tax collectors, traders and soldiers), schools (to transform “primitives” into citizens) and clinics (to battle disease).

There is some archaeological evidence to prove that the ambiguous terrain of the escarpment has been the zone of equally ambiguous ecologies for thousands of years. Archaeological research in Benishangul, a region south of the Blue Nile bordering Sudan, identified a site, Bel Kurkumu, radiocarbon-dated to between 6000 and 4000 years ago, which yielded abundant pottery of the type found in Mesolithic and Neolithic sites around Khartoum, in central Sudan (Wavy Line, Dotted Wavy Line and Rocker), but associated with locally-made knapped lithic tools in quartz and with no clear evidence of food production (Fernández et al. 2007). The location of the site in the rugged escarpment, the lithic artefacts and a net sinker found at the site are more consistent with foraging practices, including fishing. Small bands of hunter-gatherers were probably moving along and across the escarpment making the most of the ecological diversity of the different niches, connecting two worlds (that of the early Sudanese agriculturalists and the Ethiopian foragers), while avoiding the trap of sedentarization and food production. This was probably already a tactic of resistance (Fernández 2003).

A more recent example is to be found in the Ajilak foothills of the Ethiopian plateau in Gambela. This is ambiguous terrain at its best: definitely not the mountains, but neither the proper South Sudanese plain. It is a topographic shatter zone where the escarpment is fractured in myriad rocky hills. In these hills, several campsites were found that were seasonally occupied around 1000–1200 CE (González-Ruibal et al., 2014, González-Ruibal, 2014). All faunal remains retrieved from the archaeological sites belonged to wild animals, including antelopes, warthogs, snakes, monkeys and fish. However, the people in the foothills were also in contact with agro-pastoral communities and were using pottery identical to that documented in South Sudan (David et al. 1981), which they exchanged with their neighbors, perhaps for game and other forest resources. These people were moving between the South Sudanese swamps and the Ethiopian escarpment, exploiting different ecotopes, dealing with different societies (including Ethiopian highlanders, as shown by obsidian only found in the plateau), while avoiding being trapped, like their forebears several millennia before, in the sphere of influence of any neighboring society.

Almost one millennium later, the Chabu follow similar patterns. They live in the forested escarpment of southern Gambela and, although often described as hunter-gatherers (Dira and Hewlett 2018a), they are actually people in-between—like the space they inhabit. The Chabu practice hunting, plant-gathering, fishing, beekeeping and slash-and-burn agriculture and are better described as “low-level food producers” (B.D. Smith 2001), a label that comprises a variety of subsistence strategies. These should not be seeing merely as the survival of an ancient mode of production, but actually as an adaptation to the ambiguous ecology of the escarpment. Honey-gathering, for instance, which is very characteristic of many of the groups inhabiting the Ethiopian escarpments, both west and south (Simoons 1960: 203-204), is part of a multiethnic economy tying the dominant societies in the plateau and the minorities of the escarpment, as most of the honey is sold in highland markets (Fig. 5). The same mixed practices that we see in the livelihoods of the Chabu and other neighboring groups can be observed in their culture. In the case of the Chabu they speak a linguistic isolate (Dimmendaal 2008), which incorporates elements from many and very different languages (Nilo-Saharan and Omotic). Similarly hybrid are their traditions and their material culture, which borrow heavily both from the highland farmers and from other minority groups. The liminal ecology of the Gambela escarpment, which is characterized by a belt of thick forest, provides both economic opportunities and shelter. Ren’ya Sato (2017: 212-213), referring to the Majangir, neighbors of the Chabu, talks of “forest tactics”, which includes seeking shelter in the forest in the case of attack and dispersed versus concentrated settlement pattern..

1. Download: Download high-res image (1MB)
2. Download  Download full-size image

Fig. 5. Chabu beehives and hut in the middle of the forest in the escarpment of Gambela. A typical liminal ecology exploited by the Chabu, who sell honey in the highland markets.

---

In fact, living in the escarpment has its risks, because, unlike the lowlands, they are close to the centers of power in the plateau. Therefore, many groups have been faced with the dilemma of accepting powerful and menacing neighbors—the Chabu have been progressively chased away by industrial coffee plantations (Dira and Hewlett 2018b)—or escaping to the lowlands. This has in some cases produced split identities. The Gwama people, for instance, divided into two during the last two or three centuries: those who remained in the limit of the plateau became Sith Swala (“Black People”) and those who fled to the base of the escarpment remained Gwama. Although they share the language and many customs, the Sith Swala have been progressively incorporating elements from dominant societies (particularly Oromo), with which they have symbiotic (but largely asymmetrical) interactions (González-Ruibal, 2014, González-Ruibal et al., 2014). This splitting has occurred again as recently as 2004–2005. As mentioned above, the Ethiopian government constructed infrastructures in the escarpment with a “civilizing” purpose: to attract local communities, to better control them and to provide services. Offered with the opportunity to resettle, some of the people living at the base of the escarpment and belonging to the Gwama, Komo and Ganza groups decided to move to the new villages and take advantage of the services provided by the government. Others, however, decided to remain in their homeland or move further away into the Sudanese lowlands (González-Ruibal, 2014, González-Ruibal et al., 2014. It is possible that this new split will lead to processes of ethnogenesis similar to those that had already taken place in the past.

However, the process of villeagization is far from being straightforward. During my visits to the area (2007, 2009), I was told that many of those who had opted for resettling were still abandoning the state-sponsored villages to hunt, fish and gather honey in the lowlands, often disappearing with their entire families for weeks—following patterns of mobility that are centuries old. While the escarpment is being progressively incorporated into state structures, it is still very much ambiguous terrain that defies control and categorization.