This article explores the challenges faced by state-making in the Horn of Africa over the past two millennia using archaeology’s deep-time perspective. It focuses on state ecology and resistance, avoiding presentist bias in political analyses. The region has some of the oldest state polities, non-linear state trajectories, and persistent small-scale, stateless societies. The article explores themes such as nomadic pastoralists’ anti-infrastructural ethos, internal frontiers, and liminal ecologies defying state control.

Making and unmaking the state in the Horn of Africa: A deep-time perspective


In this article, I intend to use archaeology’s understanding of deep time and infrastructure to explore the ways in which state-making has been challenged in the Horn of Africa during the last two millennia. I will take a long-term approach to state ecology and state resistance so as to eschew the presentist bias that is all too frequent in political analyses, particularly in Africa, and that prevents us from understanding some of the deep undercurrent that explain contemporary phenomena. The Horn is an ideal case for this kind of inquiry because it has some of the oldest state polities south of the Sahara; state trajectories in the region are non-linear and fraught with obstacles, though surprisingly persistent, and small-scale, stateless societies have proved to be extremely persistent, both in the periphery and at the heart of the state. Here I will explore three themes that are illustrative of the relationship between state-building, infrastructures and resistance in the borderlands of the Horn of Africa: the anti-infrastructural ethos of nomadic pastoralists; internal frontiers or zones of difference, and liminal ecologies, such as swamps and escarpments, which defy state control, technologies and imaginaries.



Resistance, Political ecology, Infrastructure, Ethiopia, Somaliland

1. Introduction: infrastructures as traps

As I am writing these lines, the state in the Horn of Africa is, again, challenged. The Ethiopian Federal Army fights the Tigray’s People Liberation Front in what amounts to a full-fledged civil war and international observers are wary of the conflict engulfing the entire region—Sudan, Eritrea, Somalia—and beyond. The impression one gathers from the news is that the state in this part of Africa is fragile —and indeed it is, but also, paradoxically, very resilient. Only that it is a kind of resilience at odds with modern Western political notions. In this article, I would like to look at the state in the Horn of Africa through the lens of archaeology and the notion of infrastructure. This might seem like a strange disciplinary choice, but potential misgivings, as I will try to show, are unfounded. Archaeology has been exploring state ecologies almost since its inception and some very influential theories on state building took ecology and infrastructure seriously and well before it became fashionable in the social sciences (Butzer, 1982): the most famous example is the hydraulic hypothesis, which posited that early states were born out of a necessity of coordinating large infrastructural work related to irrigation (Wittfogel 1956) and that has spurred a plethora of archaeological investigations (Harrower 2009). Americanist archaeology is probably the one that has developed more theoretically-informed work on infrastructures and complex societies and their insights are of use elsewhere (Stanish, 1994, Carballo and Feinman, 2016, Wilkinson, 2019a).

Following archaeologist Daryl Wilkinson (2019a), I understand infrastructure as material assemblages that promote human organization at a large scale, require the input of labor beyond the household, sprawl across the landscape, and do not accommodate people, but things (thus excluding domestic architecture). To this I would add that they tend to be solid and permanent and to enable, channel, or prevent the circulation of humans, non-humans, substances and energy. Stateless societies can develop very complex infrastructures, but many not only do not, but actively refuse to engage in their construction. Instead, states always build infrastructure: as Wilkinson (2019a: 1221), notes, “We can maybe imagine infrastructure without a state, but not the state without infrastructure”.

Infrastructures tend to be portrayed as inherently positive and being cut off from them as a tragedy (Wilkinson 2019b: 29). Indeed, they might be essential for life—providing from clean water to electricity. However, as a state mechanism, they can also be deployed to exclude, oppress and extract resources (Schouten and Bachmann, 2021). Even when as much is admitted, infrastructure is still seen as transparent: it is the environment (nature) that is dark, threatening, resisting domestication, and derailing political programs, be it mud, rain or forests (Gordillo 2018). However, the infrastructure itself can be materially unmanageable and epistemologically obscure (Easterling 2014: 71-76), as things more generally (Olsen 2010). It can work against humans and make their life more difficult. In fact, infrastructures can be usefully conceptualized as traps, physical as well as political.

They are political inasmuch as they have long-lasting political effects and they are traps in that some of these political results are unintended and, once they set in, difficult to change: they fix populations, create inequalities, foster monopolies, redirect and anchor trade, intensify conflict. And they are physical in that their materiality and their material elements are essential to their functioning—be they stones, dirt, concrete or tarmac. Due to their scale, we can describe them as “landscape-traps” (Corsín Jiménez and Nahum-Claudel, 2019: 396–397), something that is no longer a technology but an ecology—a political ecology. Landscape-traps are complex structures that domesticate space and the people that inhabit it. Like any other kind of trap, infrastructures are a modality of relationality, a material entanglement across the boundaries of the human and non-human (Corsín Jiménez and Nahum-Claudel, 2019: 394). And in this entanglement, there is always a political risk.

Here I start from the idea that stateless peoples are often aware of such political risk and for this reason they tend to develop counterinfrastructural tactics of resistance, which include ephemeral materialities, minor architectures, seasonality, movement and the use of specific topographies and the environments (Criado-Boado et al., 1995; Strother, 2004, Stoner, 2012). Thus, the Pembe of Congo are adamant about their refusal to have chiefs building permanent ritual houses, which they see as a step toward solidifying and extending their restricted powers (Strother 2004); the Karen of Myanmar construct small villages in perishable materials that are virtually invisible and can disperse quickly (Scott 2009: 179-182), many nomadic sites of aggregation are used only seasonally to prevent them from becoming centers of power (González-Ruibal and de Torres, 2018), and some communities reject roads or bridges that may make them more amenable to external control (Saxer and Andersson, 2019: 149; Schouten et al., 2021).

As noted above, I will use an archaeological approach to infrastructures. This means being attuned to their material qualities. It also means adopting a long-term perspective (Wilkinson 2019a: 1218). While materiality has been receiving greater attention with the recent material turn (e.g. Larkin, 2013, Easterling, 2014, Khalili, 2017), deep time is still very much absent in most accounts and theoretical reflections by geographers, political scientists and sociologists. Thus, Larkin (2013: 332) states that infrastructure has its conceptual roots in the Enlightenment—forgetting the central role that infrastructure had in Roman political thinking and practice two millennia before. Mayer and Acuto (2015), in turn, understand large technical systems as developing mostly during the late twentieth century, leaving aside the Inca road network, the Great Wall of China or the Indian Ocean trade system, which were spatially large and logistically complex. As Wilkinson (2019a: 1219) notes, contemporary studies on infrastructure have little to say about “pre-modernity”: “genres of modernity might vary… but the narrative is always about modernity in some form or another” (Wilkinson 2019b: 29). In the case of Africa, the colonial and postcolonial occupy now most of the historical imagination of the humanities and social sciences (Reid, 2011a). References to precolonial (pre-European) history are usually vague and refer to a rather recent period, rarely going beyond the sixteenth century and most commonly starting from 1800 onwards. This is a problem because many contemporary issues cannot be understood properly if we do not take a long-term perspective: this is the case with cycles of state violence in the Horn (Reid, 2022) or the political ecology of road blocking (Schouten, 2019), for instance. Taking a deep-time approach does not mean that the present is determined by the past, that there are direct connections between past and present or that history is repeated. As Mark Twain noted, history is not repeated, but it rhymes. And archaeology is a good way of identifying those rhymes.

The discipline can provide access to times not usually covered by most accounts of African history (Reid, 2011a) and it has indeed contributed decisively to historicizing the remote past, at times in connection with the present—for an excellent example of a long-term archaeological narrative see MacEachern (2018). It can even be argued that it is in the discipline of archaeology that precolonial African history has thrived more vigorously during the last two decades.

In this article, I would like to examine the ecology of the state in the Horn of Africa putting particular emphasis on deep time and materiality. Understanding state ecologies inevitably means exploring borderlands—zones of resistance and friction, where states are made and unmade. While my analysis explores the Horn, with specific emphasis on Ethiopia and the northern Somali territories, some of the conclusions can be extrapolated to other areas in Africa and elsewhere. And although my approach looks at the past, what I describe is far from being irrelevant in the present: the conquest of the frontier and the making and unmaking of the state is still an ongoing process in the Horn (Markakis, 2011, Clapham, 2018) and in many other places of the Global South (Mezzadra and Neilson 2017).

2. A deep ecology of the state in the Horn of Africa

Before talking about state ecologies in the Horn, it is necessary to clarify what is meant by “state” in this article. The concept has been much debated in archaeology, as the discipline has had to deal with a myriad of ambiguous cases, particularly from the beginnings of state formation (Feinman and Marcus, 1998, Lull and Micó, 2011). Recently, Graeber and Wengrow (2021) have even defended doing without the concept altogether, since it is too much imbued with specific notions of the modern nation-state, which is quite unique in history. Instead, they identify three sources of power in inegalitarian polities: control of violence, control of information, and individual charisma, which may or may not appear together and emerge in different sequences (Graeber and Wengrow 2021: 365-369). Here I adopt a broad notion of state as any polity organized around one or more of those three elementary forms of domination and based on the division of functions, socioeconomic inequality and the institutionalization of such division and inequality—what Clastres (1987) defines as “divided societies” and archaeologists sometimes misleadingly as “complex societies” (Stein 1998)—misleadingly inasmuch as stateless societies can be as complex as state ones (Graeber and Wengrow 2021). They include both polities with characteristic state apparatuses (a priestly class, written administration, centralized authority, cities) and other stratified social systems—such as the “chiefdoms” of evolutionary anthropology. This broader notion is helpful in places like Sub-Saharan Africa that do not fit neatly in prevalent models of the state (Southall, 1988, McIntosh, 1999, Markakis et al., 2021). Thus, Meyer Fortes and Evans-Pritchard (1940) have identified several traits that are characteristic of African inegalitarian societies, including the limitation of the monarch’s power, its important ritual function, the idea of the ruler as guarantor of fertility, order of the cosmos and social reproduction, the power balance between different collectives, and the central role of symbols for social cohesion within the state—some of which apply to different polities in the Horn (Haberland 1965). Southall (1988), in turn, developed the concept of “segmentary state” to define many of those sociopolitical organizations that share features with tribal societies, such as the importance of collective actions, lineages and the existence of multiple chiefs that curtail the agency of the king. Archaeology, in turn, has been defining complex societies that challenge established notions of statehood and sovereign power, for example, by documenting large polities with sophisticate material cultures but no trace of sovereign power or strong inequalities. This has been the case particularly in West Africa (McIntosh 1999), yet the Horn provides ample ground for rethinking the state in a deep-time perspective.

There are many reasons why the Horn of Africa is a perfect place to study political ecologies. To start with, it is home to some of the oldest state polities south of the Sahara, going back to the early first millennium BC, including the D’MT polity, established around the eighth century BC in the highlands of what are today Eritrea and northern Ethiopia, and the Aksumite state (1st-9th c. AD) (Fattovich 2010). During the last three millennia, the Horn has witnessed the emergence and collapse of a variety of stratified societies, centralized and decentralized, hierarchical and heterarchical, regional and imperial, which cannot be reduced to the simple colonial/precolonial dualism that often divides political history in Africa. Secondly, state efforts and territorial centralization have been challenged many times in the Horn. Instead of a continuous trajectory of centralized, urban polities, the region has witnessed the emergence, collapse and reemergence of often fragile and volatile state formations, which contrast with a more stable history of state formations in the Near East and the Mediterranean basin—after their somewhat shaky beginnings (Scott 2015). Thus, after more than a half a millennium of state history, the Aksumite polity declined between the sixth and ninth century AD and eventually collapsed (Phillipson, 2012, Bard et al., 2014): with it vanished some of the elements associated with state power, including currency, towns and a permanent capital. In the territory now occupied by the de facto state of Somaliland and eastern Ethiopia, short-lived urban-based states existed between the thirteenth and sixteenth centuries, a flash in a long stateless history (Fauvelle et al., 2017, Chekroun and Hirsch, 2020).

From a temporal point of view, we have a state history characterized by both its depth and non-linear character. From a spatial point of view, in turn, we find a mosaic of polities: fragmented or patchwork states, crisscrossed by myriad borderlands and surrounded by wide, and often unruly, stateless space. Here it is essential to change the Western perspective on borders as fixed lines (Barth 2000). In the Horn, the equation wide states/thin borders does not hold. Until very recently, state polities were surrounded by stateless spaces that were several times larger than the polities themselves. Around the third century AD, for instance, the Aksumite kingdom had a maximum extension of 300 km from east to west, whereas the distance from the Aksumite border to the nearest African polity, Meroe, on the Nile Valley, was around 600 km of land inhabited by nomads. Of course, these large borderlands cannot be seen as homogeneous spaces: on the one hand, they were culturally and geographically diverse. On the other, the relationship with the state varied from one place to another. As in West Africa and the Sahel (MacEachern 2018), borderlands in Northeast Africa were, depending on their proximity to state centers, zones open to predation or peripheries with which political alliances could be struck and trade conducted—or not. Distance was categorized both in physical and cultural terms.

Stateless space lay beyond or outside the space claimed as sovereign territory by the state. However, it also lay within such sovereign territory, as we will see. Edwin Ardener (2012) noted that remote places do not need to be located very far away and speaks of a topology (rather than topography) of remoteness. Nowhere is this truer than in the Horn of Africa, where autonomous communities, refusing state rule, can be found at a distance of only a few kilometers from a state center. Thus, during the early second millennium AD, the Somali territories had both settled populations in towns and villages under the control of several polities and nomads that roamed the land in-between quite freely. The state here is better imagined as a constellation of dots, strips of land and thin lines in a wide stateless space. This, in fact, has been the norm rather than the exception with ancient states: Monica L. Smith (2005: 838) has argued that a “spiky, node-and-connector model” of political interactions characterizes ancient states more accurately than the bounded-territory model. The originality of the Horn of Africa is that this network model of state has largely persisted until the twenty-first century.

The heterogeneity of states and borderlands has much to do with the extraordinary complexity and diversity of societies existing in the Horn—dozens of language families, sociopolitical formations, subsistence practices, worldviews, religions—and also with its complex geography, which has deserts, savannas, plains, high mountains with afroalpine environments, rainforests, dry deciduous forests and grasslands, drastic ecosystemic changes occurring at times in only a few kilometers. A fragmented cultural world is thus matched by a fragmented physical geography, which has more often than not become an ally of minority groups across Africa. Thus, mountainous regions such as the Mandara Highlands in Cameroon (MacEachern 2018), the Guera Massif in Chad and the Nuba mountains (Jedrej 2006), the hilly area of the southern Dar Funj (Jedrej 2004) and the Ethiopian-Sudanese escarpment (Grottanelli, 1948, González-Ruibal, 2014, González-Ruibal et al., 2014) became safe havens for a diversity of small-scale communities escaping more powerful groups living in the plains (or the highlands, in the case of Ethiopia). The people of these areas were originally grouped together by anthropologists under the label montagnards paléonigritiques, and considered remnants of a single prehistoric cultural circle extending from Ethiopia to Senegal (Froehlich 1969). Today, it makes more sense to understand them not in terms of a common cultural background, but in terms of similar political ecologies, which are both enmeshed in and refractory to the state.

In what follows, I will briefly explore three themes in the political ecology of the state in the Horn of Africa throughout two millennia: nomadic landscapes, internal frontiers, and liminal terrain (Fig. 1 and Fig. 2).

  1. Download: Download high-res image (1020KB)
  2. Download: Download full-size image

Fig. 1. Map of the Horn of Africa with peoples mentioned in the text: 1) Beta Israel; 2) Beta Israel and K’ïmant; 3) Wäyto; 4) K’unfäl; 5) Boro; 6) Mao; 7) Gwama, Ganza, Sith Swala; 8) Chabu, Majangir; 9) Raya Oromo; 10; Sahalla Agäw; 11) Argobba.


  1. Download: Download high-res image (589KB)
  2. Download: Download full-size image

Fig. 2. Map of the Horn showing land cover, historically-dominant modes of subsistence, and places mentioned in the text: 1. Axum; 2. Zeila; 3. Berbera/Ceel Gerdi; 4. Xiis; 5. Ceelayo; 6. Qandala; 7. Ras Hafun; 8. Bel Kurkumu (Benishangul); 9. Ajilak (Gambela).

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.

3. Concluding remarks: escaping the infrastructural trap

In this article I have discussed political ecologies that have historically bypassed or tried to escape infrastructures in the Horn of Africa. I have argued that state infrastructures act as traps that serve to fix and control populations—even when this is not their alleged purpose. Many stateless peoples are aware of this danger and have developed counterinfrastructural tactics. This includes, among others, the refusal to build infrastructures, the construction of ephemeral landscapes that can be assembled and disassembled quickly, mobility and seasonality that prevent specific sites to become centers of power, the avoidance of infrastructure-heavy zones or zones that are liable to be accessed and controlled with infrastructures, and the management of ecologically in-between areas and liminal terrain that are avoided by dominant economies. Through this countereinfrastructural tactics, the communities to which I have referred here—nomads, low-level food producers and internal frontier peoples—have been able to retain their autonomy for centuries or millennia. They have persisted in places that are, in themselves, resistant to the infrastructural drive, places that do not allow themselves easily to be landscape-trapped: deserts, escarpments, mountains and liminal terrain more generally. The political work that infrastructures undertake for states is often carried out by natural phenomena among stateless societies, only that in the opposite direction.

I have adopted a long-term approach, aided by archaeology, to explore long-term phenomena that have shaped historical experience in the Horn of Africa and that still have an influence in the present. I argue that such a long-term perspective provides a different understanding of the contemporary world, one that takes deep time, materiality and space more seriously into account, while avoiding presentim. Yet this approach also has its limits. The formidable challenges that the world faces today—population growth, nationalism, the climate crisis and hypermodern technologies—have no match in the past and may change structures, infrastructures and processes that have been in place for a very long time.

Funding statement

Research for this paper was conducted through the projects HAR2013-48495-C2-1-P and PGC2018-099932-B-I00 funded by MCIN/AEI/ 10.13039/501100011033 and by “ERDF A way of making Europe” and by the Palarq Foundation.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.


    1. Abir, 1980 M. Abir
      Ethiopia and the Red Sea. The rise and decline of the Solomonic dynasty and Muslim-European rivalry in the region
      Frank Cass, Abingdon (1980)
    2. Ahmad, 1999 A.H. Ahmad
      Trading in slaves in Bela-Shangul and Gumuz, Ethiopia: border enclaves in history, 1897–1938
      J. African History, 40 (3) (1999), pp. 433-446
    3. Akalin, 2014 D. Akalin
      Somali’de Berbera Limanı ve Osmanlı Devleti’nin Bölge Üzerindeki İddiaları (1839–1894)
      Tarih İncelemeleri Dergisi, 29 (1) (2014), pp. 1-35
    4. Ardener, 2012 E. Ardener
      “Remote areas” Some theoretical considerations
      HAU: J. Ethnogr. Theory, 2 (1) (2012), pp. 519-533
    5. Bachmann and Schouten, 2018 J. Bachmann, P. Schouten
      Concrete approaches to peace: infrastructure as peacebuilding
      Int. Affairs, 94 (2) (2018), pp. 381-398
    6. Bard et al., 2014 K.A. Bard, R. Fattovich, A. Manzo, C. Perlingieri
      The chronology of Aksum (Tigrai, Ethiopia): a view from Bieta Giyorgis
      Azania: Archaeol. Res. Africa, 49 (2014), pp. 285-316
    7. Barth, 2000 F. Barth
      Boundaries and connections
      A.P. Cohen (Ed.), Signifying identities, Routledge, London (2000), pp. 25-36
    8. Butzer, 1982 Butzer, K.W., 1982. Archaeology as human ecology: method and theory for a contextual approach. Cambridge: Cambridge University Press.
    9. Burton, 1910 R. Burton
      First Footsteps in East Africa
      John Dent. First edition 1856, London (1910)
    10. Cassanelli, 1982 L.V. Cassanelli
      The Shaping of Somali Society: Reconstructing the History of a Pastoral People, 1600–1900
      University of Pennsylvania Press, Philadelphia (1982)
    11. Carballo and Feinman, 2016 D.M. Carballo, G.M. Feinman
      Cooperation, collective action, and the archeology of large-scale societies
      Evolut. Anthropol.: Issues News Rev., 25 (6) (2016), pp. 288-296
    12. Casson, 1989 L. Casson
      Periplus Maris Erythraei
      Princeton University Press (1989)
    13. Caulk, 1984 R. Caulk
      Bad men of the Borders: shum and shefta in Northern Ethiopia in the 19th century
      The International Journal of African Historical Studies, 17 (2) (1984), pp. 201-227
    14. Chekroun and Hirsch, 2020 Chekroun, A., Hirsch, B., 2020. The Sultanates of Medieval Ethiopia. In S. Kelly (ed.): A Companion to Medieval Ethiopia and Eritrea, 86-112. Amsterdam: Brill.
    15. Cheesman, 1936 R.E. Cheesman
      Late Tana and Blue Nile
      Frank Cass, London (1936)
    16. Chittick, 1979 N. Chittick
      Early ports in the Horn of Africa
      Int. J. Nautical Archaeol., 8 (4) (1979), pp. 273-277
    17. Clapham, 2018 C. Clapham
      The Ethiopian developmental state
      Third World Quart., 39 (6) (2018), pp. 1151-1165
    18. Clastres, 1987 P. Clastres
      Society Against the State: Essays in Political Anthropology
      Zone, New York (1987)
    19. Corsín Jiménez and Nahum-Claudel, 2019 A. Corsín Jiménez, C. Nahum-Claudel
      The anthropology of traps: concrete technologies and theoretical interfaces
      J. Mater. Cult., 24 (4) (2019), pp. 383-400
    20. Criado-Boado et al., 1995 F. Criado-Boado
      The visibility of the archaeological record and the interpretation of social reality
      I. Hodder, M. Shanks (Eds.), Interpreting Archaeology. Finding meaning in the past, Routledge, London (1995), pp. 194-204
    21. Crummey, 1986 D. Crummey
      Banditry and resistance: noble and peasant in nineteenth-century Ethiopia
      Banditry, Rebellion and Social Protest in Africa (1986), pp. 133-150
    22. Cruttenden, 1849 C.J. Cruttenden
      Memoir on the western or Edoor tribes, inhabiting the Somali coast of N.-E. Africa, with the southern branches of the family of Darrood, resident on the banks of the Webbe Shebeyli, commonly called the River Webbe
      J. Roy. Geogr. Soc. Lond., 19 (1849), pp. 49-76
    23. David et al., 1981 N. David, P. Harvey, C.J. Goudie
      Excavations in the Southern Sudan, 1979
      Azania: Archaeol. Res. Africa, 16 (1) (1981), pp. 7-54
    24. Dimmendaal, 2008 G.J. Dimmendaal
      Language ecology and linguistic diversity on the African continent
      Lang Linguist. Compass, 2 (5) (2008), pp. 840-858
    25. Dua, 2019 Dua, J., 2019. Hijacked: Piracy and economies of protection in the Western Indian Ocean. Comparative Studies in Society and History (61(3): 479-507.
    26. Dira and Hewlett, 2018a S.J. Dira, B.S. Hewlett
      The Chabu hunter-gatherers of the highland forests of Southwestern Ethiopia
      Hunter Gatherer Res., 3 (2) (2018), pp. 323-353
    27. Dira and Hewlett, 2018b S.J. Dira, B.S. Hewlett
      Cultural resilience among the Chabu foragers in southwestern Ethiopia
      African Study Monogr., 39 (3) (2018), pp. 97-120
    28. Endalew, 2006 T. Endalew
      Inter-ethnic relations on a frontier: Mätakkäl (Ethiopia), 1898–1991
      Otto Harrassowitz, Wiesbaden (2006)
    29. Easterling, 2014 K. Easterling
      Extrastecraft The power of infrastructure
      Verso, London (2014)
    30. Fortes and Evans-Pritchard, 1940

      Fortes, M., Evans-Pritchard, E. E. 1940. Introduction. In M. Fortes and E.E. Evans-Pritchard, eds.: African political systems, 1-24. London: KPI.
    31. Fattovich, 2010 R. Fattovich
      The Development of Ancient States in the Northern Horn of Africa, c. 3000 BC–AD 1000: An Archaeological Outline
      J. World Prehist., 23 (3) (2010), pp. 145-175
    32. Fauvelle et al., 2017 F.-X. Fauvelle, B. Hirsch, A. Chekroun
      Le sultanat de l’Awfāt, sa capitale et la nécropole des Walasmaʿ: Quinze années d’enquêtes archéologiques et historiques sur l’Islam médiéval éthiopien
      anisl (51) (2017), pp. 239-295
    33. Fauvelle-Aymar et al., 2011 F. Fauvelle-Aymar, B. Hirsch, R. Bernard, F. Champagne, F.X. Fauvelle-Aymar, B. Hirsch
      Le port de Zeyla et son arrière-pays au Moyen Âge: Investigations archéologiques et retour aux sources écrites
      F. Fauvelle-Aymar, B. Hirsch (Eds.), Espaces musulmans de la Corne de l’Afrique au Moyen Âge, DeBoccard, Paris (2011), pp. 27-74
    34. Feinman and Marcus, 1998 G.M. Feinman, J. Marcus (Eds.), Archaic States, James Currey, Oxford (1998)
    35. Feldman, 2007 B. Feldman
      Somalia: amidst the rubble, a vibrant telecommunications infrastructure
      Rev. African Polit. Econ., 34 (113) (2007), pp. 565-572
    36. Fernández, 2003 V.M. Fernández
      Four thousand years in the Blue Nile: Paths to inequality and ways of resistance. Cuatro mil años en el Nilo Azul: Caminos a la desigualdad y vías de resistencia
      Complutum, 14 (2003), pp. 409-425
    37. Fernández et al., 2007 V.M. Fernández, I. de la Torre, L. Luque, A. González-Ruibal, J.A. López-Sáez
      A Late Stone Age sequence from west Ethiopia: The sites of K’aaba and Bel K’urk’umu (Assosa, Benishangul-Gumuz regional state)
      J. African Archaeol., 5 (1) (2007), pp. 91-126
    38. Feyissa and Hoehne, 2010 Feyissa, Dereje, Hoehne, Markus V., 2010. State Borders and Borderlands as Resources. In Borders and Borderlands as Resources in the Horn of Africa, eds. Dereje Feyissa and Markus V. Hoehne, 1–25. Eastern African Series. Suffolk and Rochester: James Currey.
    39. Freeman, 2003 D. Freeman
      Understanding marginalisation in Ethiopia
      D. Freeman, A. Pankhurst (Eds.), Peripheral people: the excluded minorities in Ethiopia, Addis Ababa University Press, Addis Ababa (2003)
    40. Froehlich, 1969 J.C. Froehlich
      Les montagnards paleonigritiques
      ORSTOM, Paris (1969)
    41. Gamst, 1969 F.C. Gamst
      The Qemant: a pagan-Hebraic peasantry of Ethiopia
      Holt, New York (1969)
    42. González-Ruibal, 2014 A. González-Ruibal
      An archaeology of resistance. Materiality and time in an African borderland
      Rowman & Littlefield, Lanham, MD (2014)
    43. González-Ruibal and de Torres, 2018 A. González-Ruibal, J. de Torres
      The fair and the sanctuary: gathering places in a nomadic landscape (Somaliland, 1000–1850 AD)
      World Archaeol., 50 (1) (2018), pp. 23-40
    44. González-Ruibal et al., 2014 González-Ruibal, A., Marín Suárez, C., Sánchez-Elipe, M., Lesur, J., Martinez Barrio, C. 2014. Late hunters of western Ethiopia: the sites of Ajilak (Gambela), c. AD 1000–1200. Azania: Archaeological Research in Africa 49(1), 64-101.
    45. González-Ruibal et al., 2021 A. González-Ruibal, J. de Torres, M.A. Franco Fernández, C. Martínez Barrio, P. Gutiérrez de León Juberías
      Asia in the Horn. The Indian Ocean trade in Somaliland
      Archaeol. Res. Asia, 27 (2021), p. 100289
    46. González-Ruibal et al., 2022 A. González-Ruibal, J. de Torres, C. Martínez Barrio, M.A. Franco Fernández, A. Fernández Fernández, G. de León, P. Juberías, J. Yravedra, M. Gaudiello, A.J. Dualeh
      Nomads trading with empires: intercultural trade in ancient Somaliland in the first to seventh centuries CE
      Am. J. Archaeol., 126 (1) (2022), pp. 103-138
    47. Gordillo, 2018 G. Gordillo
      Terrain as insurgent weapon: An affective geometry of warfare in the mountains of Afghanistan
      Polit. Geogr., 64 (2018), pp. 53-62
    48. Graeber and Wengrow, 2021 D. Graeber, D. Wengrow
      The dawn of everything: A new history of humanity
      Allen Lane, London (2021)
    49. Grottanelli, 1948 V.L. Grottanelli
      I pre-Niloti: un arcaica provincia culturale in Africa
      Annali Lateranensi, 12 (1948), pp. 281-326
    50. Haberland, 1965 E. Haberland
      Untersuchungen zum äthiopischen Königtum
      Franz Steiner, Wiesbaden (1965)
    51. Harrower, 2009 M.J. Harrower
      Is the hydraulic hypothesis dead yet? Irrigation and social change in ancient Yemen
      World Archaeol., 41 (1) (2009), pp. 58-72
    52. Hassen, 1995 M. Hassen
      The Oromo of Ethiopia, 1570–1860
      The Red Sea Press, Asmara (1995)
    53. Hoehne, 2014 M.V. Hoehne
      Resource conflict and militant Islamism in the Golis Mountains in northern Somalia (2006–2013)
      Rev. African Polit. Econ., 41 (141) (2014), pp. 358-373
    54. Hoehne, 2015 M.V. Hoehne
      Between Somaliland and Puntland: Marginalization, militarization and conflicting political visions
      Rift Valley Institute, London/Nairobi (2015)
    55. Honeychurch, 2014 W. Honeychurch
      Alternative complexities: the archaeology of pastoral nomadic states
      J. Archaeol. Res., 22 (4) (2014), pp. 277-326
    56. Insoll et al., 2021 T. Insoll, N. Khalaf, R. MacLean, H. Parsons-Morgan, N. Tait, J. Gaastra, A. Beldados, A.J.E. Pryor, L. Evis, L. Dussubieux
      Material cosmopolitanism: the entrepot of Harlaa as an Islamic gateway to eastern Ethiopia
      Antiquity, 95 (380) (2021), pp. 487-507
    57. Jedrej, 2004 M.C. Jedrej
      The Southern Funj of the Sudan as a frontier society, 1820–1980
      Comparat. Stud. Soc. Hist., 46 (4) (2004), pp. 709-729
    58. Jedrej, 2006 M.C. Jedrej
      Were Nuba and Hadjeray stateless societies? Ethnohistorical problems in the eastern Sudan region of Africa
      Paideuma, 52 (2006), pp. 205-225
    59. Kaplan, 1992 S.B. Kaplan
      The Beta Israel: Falasha in Ethiopia: From Earliest Times to the Twentieth
      NYU Press, New York (1992)
    60. Khalili, 2017 L. Khalili
      The roads to power: The infrastructure of counterinsurgency
      World Policy Journal, 34 (1) (2017), pp. 93-99
    61. Khazanov, 1994 A.M. Khazanov
      Nomads and the outside world
      Cambridge University Press, Cambridge (1994)
    62. Kopytoff, 1987 I. Kopytoff
      The internal African frontier: the making of African political culture
      I. Kopytoff (Ed.), The African frontier: the reproduction of traditional African societies, Indiana University Press, Bloomington (1987), pp. 3-84
    63. Larkin, 2013 B. Larkin
      The politics and poetics of infrastructure
      Ann. Rev. Anthropol., 42 (2013), pp. 327-343
    64. Levine, 1974 D. Levine
      Greater Ethiopia: The evolution of a multiethnic society
      University of Chicago Press, Chicago (1974)
    65. Lewis, 1999 I.M. Lewis
      A Pastoral Democracy: a study of pastoralism and politics among the northern Somali of the Horn of Africa
      James Currey, Oxford (1999)
    66. Little, 2003 P.D. Little
      Somalia: Economy without state
      Indiana University Press, Bloomington (2003)
    67. Lull and Micó, 2011 V. Lull, R. Micó
      Archaeology of the origin of the state: the theories
      Oxford University Press, Oxford (2011)
    68. Markakis, 2011  J. Markakis
      Ethiopia: The last two frontiers
      James Currey, Oxford (2011)
    69. Markakis et al., 2021 J. Markakis, G. Schlee, J. Young
      The nation state: a wrong model for the Horn of Africa
      Max-Planck-Gesellschaft zur Förderung der Wissenschaften, Berlin (2021)
    70. McIntosh, 1999 S.K. McIntosh (Ed.), Pathways to complexity in Africa, Cambridge University Press, Cambridge (1999)
    71. Mayer and Acuto, 2015 M. Mayer, M. Acuto
      The global governance of large technical systems
      Millennium, 43 (2) (2015), pp. 660-683
    72. MacEachern, 2018 S. MacEachern
      Searching for Boko Haram: a history of violence in Central Africa
      Oxford University Press, Oxford (2018)
    73. Mezzadra and Neilson, 2017 S. Mezzadra, B. Neilson
      On the multiple frontiers of extraction: Excavating contemporary capitalism
      Cult. Stud., 31 (2–3) (2017), pp. 185-204
    74. Mohamed, 2004 J. Mohamed
      The political ecology of colonial Somaliland
      Africa, 74 (4) (2004), pp. 534-566
    75. Olsen, 2010 B. Olsen
      In defense of things. Archaeology and the ontology of objets
      AltaMira, Lanham, MD (2010)
    76. Pankhurst, 1965 R. Pankhurst
      The Trade of the Gulf of Aden Ports of Africa in the Nineteeth and Early Twentieth Centuries
      J. Ethiopian Stud., 3 (1) (1965), pp. 36-81
    77. Phillipson, 2012 D.W. Phillipson
      Foundations of an African civilisation. Aksum & the northern
      James Currey, Suffolk (2012)
    78. Price, 1979 R. Price (Ed.), Maroon societies: Rebel slave communities in the Americas, Johns Hopkins University Press, Baltimore (1979)
    79. Quirin, 1998 J. Quirin
      Caste and class in historical north-west Ethiopia: The Beta Israel (Falasha) and Kemant, 1300–1900
      Journal of African History, 39 (2) (1998), pp. 195-220
    80. Reid, 2011b Reid, R. J. 2011b. Frontiers of violence in North-East Africa: genealogies of conflict since c. 1800. Oxford: Oxford University Press.
    81. Reid, 2011a R.J. Reid
      Past and presentism: The ‘precolonial’and the foreshortening of African history
      J. African History, 52 (2) (2011), pp. 135-155
    82. Reid, 2022 R. Reid
      Atrocity in Ethiopian History
      J. Genocide Res., 24 (1) (2022), pp. 98-108, 10.1080/14623528.2021.1992924
    83. Sato, 2017 R. Sato
      Sedentarization of Nomadic Shifting Cultivators: the Majangir of Lowland Ethiopia
      Senri Ethnol. Stud., 95 (2017), pp. 193-231
    84. Saxer and Andersson, 2019 M. Saxer, R. Andersson
      The return of remoteness: insecurity, isolation and connectivity in the new world disorder
      Soc. Anthropol., 27 (2) (2019), pp. 140-155
    85. Schlee, 2002 Schlee, G. 2002. Régularités dans le chaos. Traits récurrents dans l’organisation politico-religieuse et militaire des Somali L’Homme 161 : 17-50.
    86. Schouten, 2019 P. Schouten
      Roadblock politics in Central Africa
      Environ. Plan. D: Soc. Space, 37 (5) (2019), pp. 924-941
    87. Schouten and Bachmann, 2021 P. Schouten, J. Bachmann
      Buffering State-making: Geopolitics in the Sudd Marshlands of South Sudan
      Geopolitics (2021), pp. 1-19
    88. Scott, 1998 J.C. Scott
      Seeing like a State: How certain schemes to improve the human condition have failed
      Yale University Press, New Haven (1998)
    89. Scott, 2009 J.C. Scott
      The art of not being governed. An anarchist history of upland Southeast Asia
      Yale University Press, New Haven (2009)
    90. Scott, 2015 J.C. Scott
      Against the grain. A deep history of the earliest states
      Yale University Press, New Haven (2015)
    91. Shell, 2019 J. Shell
      Elephant convoys beyond the state: Animal-based transport as subversive logistics
      Environ. Plan. D: Soc. Space, 37 (5) (2019), pp. 905-923
    92. Smith and Wright, 1988 M.C. Smith, H.T. Wright
      The ceramics from Ras Hafun in Somalia: notes on a classical maritime site
      Azania: J. Brit. Instit. Eastern Africa, 23 (1) (1988), pp. 115-141
    93. Simoons, 1960 F.J. Simoons
      Northwest Ethiopia: peoples and economy
      University of Wisconsin Press, Madison (1960)
    94. Smith, 2001 B.D. Smith
      Low-level food production
      J. Archaeol. Res., 9 (1) (2001), pp. 1-43
    95. Smith, 2005 M.L. Smith
      Networks, territories, and the cartography of ancient states
      Ann. Assoc. Am. Geogr., 95 (4) (2005), pp. 832-849
    96. Southall, 1988 A. Southall
      The segmentary state in Africa and Asia
      Comparat. Stud. Soc. Hist., 30 (1) (1988), pp. 52-82
    97. Stanish, 1994 C. Stanish
      The hydraulic hypothesis revisited: Lake Titicaca Basin raised fields in theoretical perspective
      Latin Am. Antiquity, 5 (4) (1994), pp. 312-332
    98. Stein, 1998 G.J. Stein
      Heterogeneity, power, and political economy: Some current research issues in the archaeology of Old World complex societies
      J. Archaeol. Res., 6 (1) (1998), pp. 1-44
    99. Stepputat and Hagmann, 2019 F. Stepputat, T. Hagmann
      Politics of circulation: The makings of the Berbera corridor in Somali East Africa
      Environ. Plan. D: Soc. Space, 37 (5) (2019), pp. 794-813
    100. Stoner, 2012 J. Stoner
      Toward a minor architecture
      MIT Press, Harvard, MA (2012)
    101. Schouten et al., 2021 Schouten, P., Verweijen, J., Murairi, J., Batundi, S.K., 2021. Paths of authority, roads of resistance: Ambiguous rural infrastructure and slippery stabilization in eastern DR Congo. Geoforum.
    102. Strother, 2004 Z.S. Strother
      Architecture against the state: the virtues of impermanence in the Kibulu of Eastern Pende chiefs in central Africa
      J. Soc. Architect. Historians, 63 (3) (2004), pp. 272-295
    103. Tamrat, 1988 T. Tamrat
      Process of ethnic interaction and integration in Ethiopian history: the case of the Agaw
      J. African Hist., 29 (1) (1988), pp. 5-18
    104. Trimingham, 1952

      Trimingham, J.S., 1952.Islam in Ethiopia. Oxford: Oxford University Press.
    105. Torres et al., 2017 J. de Torres, A. González-Ruibal, M.A. Franco, C. Martínez, K.A. Aideed, A.M. Shabelle
      Bulhar: A Colonial Town in Somaliland. Report from the 2017 Excavations
      Nyame Akuma, 87 (2017), pp. 55-61
    106. Wilkinson, 2019a D. Wilkinson
      Towards an archaeological theory of infrastructure
      J. Archaeol. Method Theory, 26 (3) (2019), pp. 1216-1241
    107. Wilkinson, 2019b D. Wilkinson
      Infrastructure and inequality: an archaeology of the Inka road through the Amaybamba cloud forests
      J. Soc. Archaeol., 19 (1) (2019), pp. 27-46
    108. Wittfogel, 1956 Wittfogel, K., 1956. The hydraulic civilization. In W. Thomas (ed.): Man’s role in changing the face of the earth, 152-164. Chicago: Wenner-Gren.