‘Eating A Country’: The Dynamics of State-Society Encounters in Qellem, Western Ethiopia, 1908–33

Abstract

This article examines state-society encounters in imperial Ethiopia through histories of exploitation and compromise. Focusing on the western province of Qellem, the article investigates Ethiopia's engagement with local rights claims over time, illustrating how the state was imagined, negotiated, and partially legitimated. The inherent incoherence within imperial state structures is traced back to the survival of nodes of indigenous power within territories conquered in the late nineteenth century. Peasant representatives, local elites, and Amhara governors and soldier-colonists engaged with the state to turn it to their benefit, or limit its excesses. Episodes of rebellion, withdrawal, and court arbitration punctuated a cycle of negotiation within which the role of the intermediary was key. Qellem experienced a state-making exercise that was contemporaneous with, and comparable to, the formation of European colonial states elsewhere on the continent. As such, this article provides a radical challenge to dominant historiographical perspectives on imperial Ethiopia.

In February 1921, a group of peasants in alliance with local elites from one of Ethiopia's western provinces, Qellem, took to court a grievance about the imperial state's project of forcefully appropriating and redistributing resources within their home region.¹ Ethiopia had become an empire through the conquest and subjugation of hitherto independent peoples and states. Qellem was conquered around 1886, during a period of Ethiopian territorial expansion to the west and southwest which began in the early 1880s and lasted until the late 1890s. The core state that engineered the pursuit of territorial conquest was Shewa, Abyssinia's southernmost province, under the rule of King Menilek (1865–89), later Emperor of Ethiopia (r. 1889–1913).² Menilek built his empire as a competitor with — and for the same end as — European colonial powers in the Horn of Africa. This article seeks not merely to depict Ethiopia as an active participant in the scramble for Africa, but to argue that understanding the nature of political relationships in post-conquest contexts requires acknowledgement of the compromises inherent in imperial consolidation.

The plaintiffs in 1921 took imperial Ethiopia to its own court. The peasant and elite litigants argued against imperial Ethiopia's severe exploitation of their produce and labour and thus accused the empire-state of breaching its ‘law’ of rewarding allies. The court's decision divided the plaintiffs: the peasants’ complaint was rejected, but the elites’ was partly accepted. Looked at as a snapshot, from the available historical sources, the February 1921 court proceeding in Gaawoo, Qellem, appears to represent an ordinary court argument around a local dispute in an isolated rural territory, which may seem to have no meaning or significance at a national level. But a closer examination of the court debate challenges Ethiopia's dominant historiographical perspectives (the Ethiopianist Great Tradition vs Oromo historiographies), demonstrating instead that imperial expansion into Qellem involved compromise, negotiation, and internal division, as well as exploitation by all parties of the contradictions within Ethiopia's imperial project. The complex history behind the 1921 court argument in Qellem is, therefore, a significant gateway into layers of state-society relations in imperial Ethiopia.

The histories of exploitation and the responses they generated in Qellem connect with colonial Africa's histories of local bargaining which have demonstrated the roles of intermediaries in building colonial states on the continent and served ‘as a window into the complex social worlds of colonial Africa’.³ A number of scholars have examined the role of ‘intermediaries’ and ‘collaborators’ in colonial Africa and studies emphasising the brokering role of colonial chieftaincies are plentiful.⁴ A particularly rich vein of work within African historiography has explored the role of local resistance and bargaining and how the combination of these factors shaped and limited the nature, content, and attitude of African colonial states. Understanding of the Rwandan colonial state has benefited from Catharine Newbury's analysis of its resource extraction methods and the resistance it encountered. More broadly, Benjamin Lawrance, Emily Osborn, and Richard Roberts have highlighted the centrality of intermediaries as engines of colonial states’ interactions with the colonised and as a lens through which to analyse ‘bargains of collaboration’ in colonial Africa. And more recently, Cherry Leonardi and Chris Vaughan examined how mediators in colonial Sudan demanded certain rights and claimed access to the state by employing mechanisms of protest that were both profoundly local yet revealing of wider Condominium politics.⁵ This innovative contribution to Africanist historiography has, so far, had little impact on efforts to examine dynamics of imperial Ethiopia as an indigenous African empire. Like Belgium, Britain, France, Germany, Italy, Portugal, and Spain, Ethiopia was an imperial power, and it built a vast empire in competition with Europeans. In the words of a critical historian, ‘[w]hile rebuffing imperialism successfully in the north, Ethiopia managed to practice it in the south’.⁶ Colonial states’ architecture of rule and extraction, their systems of delegated governance, and the tensions, messy negotiations, and resistance that were observed in European empires in Africa were also present in the Ethiopian empire, albeit with a distinctly Ethiopian flavour.

This article examines state-society encounters in imperial Ethiopia through an analysis of evolving practices of exploitation, negotiation, and resistance in Qellem. It reviews the subject in the two competing — and contradictory — historiographical traditions of Ethiopia, the Great Tradition and the Oromo historiographies, critically examining the perspectives behind the production of historical knowledge that barely recognised African historiography and vice versa. Then, turning to imperial Ethiopia's paradigm of colonialism — the neftegna-gebbar clientelist system — the article investigates the system's key features and its impact on the local population and elites (the balabbat), examining the dynamics of local contestation over resources and access to the state within continental historical contours. It then examines the responses of the local population and their elites, analysing the multiple methods they employed to present their perceived rights in order to protect their means of survival and socioeconomic benefits, respectively. Finally, it analyses the Ethiopian empire-state's structural features in the context of colonial Africa, highlighting how competition over resources and multiple forms of claims over the state by various groups worked together as dynamic tools of state construction. I argue that the state structures of rule and dominance that imperial Ethiopia employed in governing its domains in territories conquered by the close of the nineteenth century were fractured and incoherent. This opened cracks that competing local forces manipulated, which ultimately forced the state interests to become the subject of compromise and negotiation. It was the behaviour of Amhara state actors which revealed disorders within the imperial state and dissonance in the state structure that local Oromo elites used to pursue their agendas, creating paths of self-expression for the weak — the peasants.

Ethiopia was by no means unique in developing fractured and inconsistent structures, which local rivals manipulated to challenge policy decisions emanating from the imperial centre. The historiography of colonial Africa is replete with examples of weaknesses in European imperial authority, producing a colonial circumstance that armed local elites and employees of the state with considerable leverage and initiative, which they used to craft the terms of rule. Regarding colonial Nigeria, for example, Moses Ochonu innovatively examined at length how and why the Hausa-Fulani's roles as British imperial agents in the Middle Belt region were too sophisticated, creative, and powerful to be reduced to simple colonial intermediaries. ‘Hausa-Fulani subcolonials were’, Ochonu argues, ‘not mere human tools in the hands of British officials; instead they were essential to the functioning of the colonial enterprise as colonial business essentially devolved to them in most areas of the Middle Belt’.⁷

The article navigates beyond the idea of Ethiopia's exceptionalism in colonial Africa and instead contextualises historical experiences in Ethiopia's south within the broader continental scene. To analyse state-society encounters in imperial Ethiopia and to understand how the Ethiopian state functioned and was perceived at the local level, I focus on imperial Ethiopia's brutal methods of rule and mechanisms of resource exploitation — the neftegna-gebbar system — as well as the varying strategies employed by exploited groups and the role of intermediaries in this tension-ridden environment. The article emphasises historical developments within a single province — Qellem — rather than taking the broader western Ethiopian region as the unit of analysis in order to examine more closely how each rival group in the region, including the state, influenced the process of state-society interaction and how an entire course of developments enabled or limited the way the imperial state wanted to govern its conquered provinces. While the imposition in Qellem of the neftegna-gebbar system in the early twentieth century can be understood as a mark of the expansion of central imperial power in a region that hitherto enjoyed considerable local autonomy, the multifaceted challenges the state encountered tell a more complex story. They reveal the power of the peasants whom the state had viewed as the most powerless, the ability of the intermediaries — local Oromo elites — to manoeuvre local structures on their behalf, and also the autonomy of the empire's own local agents in representing and shaping state interests despite the façade of centralised authority in Addis Ababa. Studying state-society encounters in the early twentieth-century Qellem provides a window into histories of indigenous African state-making projects that the arrival of European colonialism interrupted in the rest of the continent, featuring a unique African system of dominance and the resistance it encountered.

Historiographical context, sources, and significance

Representations and interpretations of Ethiopia's past have been contested for decades. The Great Tradition — nationalist Ethiopian historiography emphasising the achievements of state builders and rulers — is characterised by an overemphasis on Abyssinia, based on its developed literary tradition, the preservation of Christianity in seclusion on a mountainous massif, successful defence of state independence, and a dynastic history running back to classical times. Amateur and professional historiographic narratives in the Great Tradition emphasise these features, avoiding conceptual and theoretical tools that underlie critical historical inquiries and abandoning conventional standards common to African historiography. It is therefore hard to find in the Great Tradition works of historical literature that address the complexities of Ethiopia's past.⁸ However, after the 1974 revolution, when military officers espousing Marxist-Leninist ideology seized power, the Great Tradition narratives could no longer dominate. The revolution closed an extended era of Ethiopian political history and opened a radically disputed chapter in the country's politics and historiography. The revolution opened the door to alternative narratives and exposed the old Ethiopianist historiography to critical inquiry and revision.⁹

Revisionist works, such as The Southern Marches of Imperial Ethiopia, were already underway, and historiographic divergence invited its sequel, Remapping Ethiopia.¹⁰ The result was competing and contradictory perspectives, impassioned debates over specific historical themes, and broader exchanges between scholars. The dichotomy can be understood within the context of contemporary Ethiopia, where fighting over the past means fighting over the present, especially in terms of power and ideology. The debates over Ethiopia's past represent the opposing perspectives the conquered south took compared to Great Tradition historiography. Oromo historiography depicts Abyssinia's conquest as ‘colonial expansion [that] resulted in mass killings, destruction and expropriation of property, plundering [and] enslavement’, presenting the state as an enemy of the peoples in the south.¹¹ While the Great Tradition ignored the slave trade rampant in Ethiopia as a subject of historical inquiry, revisionist Oromo historiography emphasises these themes, although not in terms of highlighting the historical role slavery and the slave trade played in shaping Ethiopia's economy, society, various cultures, the state, and historical studies. Slaving expeditions and slave trade in the late nineteenth century were key instruments of state-building for the founding elites of the empire, but the same practices in Oromo history have been downplayed as mundane historical episodes.¹² Imperial Ethiopia's rule in Oromo territories has been essentialised as domination and subjugation, and the response of the Oromo people are largely presented as coherent resistance.

Closer observation of historical developments in Ethiopia's south may not allow generalised readings of southern histories, which assume the past of the conquered south as essentially continuous, downplaying internal diversity, particularly social divisions, in order to build strategies of counter-discourse. This article considers vital internal divisions within the conquered south, and dissonances and inconsistencies within imperial Ethiopia's state structure. It then mobilises them to unpack and conceptualise the multiple forms of state-society relations that unfolded in twentieth century Ethiopia. Such approaches and careful use of fresh primary sources will help analyse historical developments critically and develop an argument that navigates beyond the controversies inherent in the Great Tradition and Oromo historiographies.

Qellem's historical experiences between 1908 and 1933 are too complex to conform to the underlying assumptions in the Greater Ethiopia and Oromo historiographies. Qellem's dynamics suggest that imperial Ethiopia's structures and functionality were invented through local exploitation, conflict, bargaining, and negotiation as the imperial government imposed its presence through local interaction. Neither Ethiopia's negotiated conquest of Qellem near the close of the nineteenth century nor the imposition of Addis Ababa's direct rule, the neftegna-gebbar system, in 1917, can be explained as efforts to protect the conquered south from warfare and slavery as some of the Great Tradition historical writings attempt to prove.¹³ Imperial Ethiopia's governorship in the conquered south does not appear to exemplify an era of ‘peace and security’ constructed through collaboration with local elites as suggested in the Great Tradition narratives. The idea of the state is fundamentally a process rather than a separate or autonomous entity.

Because relevant contemporary records within Ethiopia's archives are limited in scope, this article draws on a wide range of sources, including memoirs, biographies, intelligence reports, published (Amharic) archival collections, Alessandro Triulzi Field Notes (ATFN) collected in the early 1970s, and interviews conducted in the research region.¹⁴ Primary accounts provide vital material describing local conflicts over the right to resources, the discourses mobilised in the process, imperial interventions, and challenges. YeWellegga Yetarik Senedoch [Documents for Wallagga History] offer critical historical data on Qellem, especially regarding developments between 1908 and 1932. A book composed of correspondence, reports, receipts, and local chronicles covering the 1890s until the early 1930s was collected and edited by Alessandro Triulzi and Tesema Ta'a. The primary focus of this volume is Naqamtee, Qellem's neighbour to the east, but the materials it presents on many other provinces because of the historical roles of its governors make it a fascinating historical account. Paulos Gnogno's Atse Minilik Ager Wust Yetetsatsafuwachew Debdabewoch [Emperor Menilek's Domestic Correspondence], another Amharic volume that presents Emperor Menilek's correspondences with his vassals in all parts of the empire, offers rare primary sources, providing some essential historical data on the early relationships established between Qellem and Addis Ababa. The late Paulos Gnogno was a journalist and amateur historian who got the opportunity, when he was appointed chair of a committee organising the Adwa Victory Centenary, to smuggle out imperial Ethiopia's archives, which the military regime had locked up. Archival records of this early period of imperial Ethiopia are few, so the materials these volumes offer are crucial, particularly on state-society interactions between 1917 and 1933. Intelligence reports, notably the Sudan Intelligence Reports (SIR), travel accounts, and memoirs provide additional sources of historical data. British colonial officers wrote the SIR in Sudan, but they provide considerable information on various historical matters in Ethiopia as well as Sudan and Egypt.¹⁵ SIR files offer access to historical developments between the late nineteenth century and 1936, although they provide limited detail and make mistakes on personal names and locations. Finally, oral data, including information in ATFN, present vital material about court deliberations resolving competition and conflicts.

The imposition of the neftegna-gebbar system affected the structure of local society in subtle ways, provoking a series of innovative elite and peasant responses, which were essential in determining the nature of local state construction. When the Ethiopian empire annexed a region, exploiting its resources was often characterised by violence. New systems — first the neftegna-gebbar system, then the creative restoration of the hambifataa system (discussed below) — were responsible for determining the course of local development, in turn shaping Qellem's relations with Addis Ababa. Local people challenged the state in various ways — outmigration, banditry, armed rebellion, court litigation, petitions, and appeals — enabling them to reshape and be shaped by it. Analysing the neftegna-gebbar system in the light of aspects particularly relevant to Qellem helps explain how an African empire's structure of dominance interacted with a stratified local society, revealing underlining tensions in state-society relations.

Qellem in context

The western Ethiopian province of Qellem is located to the west of Gimbii and Najjoo, south of Begii and Benishangul-Gumuz, north of Gambella, and northwest of Iluu Abbaaboor. It is bounded on the west by Sudan (see Fig. 1). Qellem is mostly populated by the Macca Oromo. Dating Oromo settlement in Qellem is difficult, although sources suggest that the Macca Oromo controlled parts of Qellem by the early seventeenth century.¹⁶

Figure 1. Administrative units of Qellem, the 1940s–1974.

Source: Ethio GIS.

Internal transformations within various Macca Oromo groups in Qellem were consolidated in the late nineteenth century, leading to socioeconomic and political innovations, reinforced by a history of conflict between Oromo lords. The gadaa system of consensual governance eventually gave way to emerging petty monarchies. By 1886, Jootee Tulluu (1855–1918), the strongest of the lords in Qellem, defeated some of his local contenders but was unable to beat others. Jootee initially sought the help of Mahdist forces, then in control of Benishangul, a territory to the north of Qellem. The Mahdists did help Jootee but would soon shift their policy towards him from one of alliance to severance. The Mahdists decreed rigid adherence to Islamic laws, imposed poll taxes on people, and demanded higher tribute from Jootee.¹⁷ Jootee responded in a military attack but was overpowered and fled his country. He then sought an alliance with the expanding Ethiopian empire. Jootee met with the commander of the Amhara forces of Ethiopia, Ras Goobanaa Daacee, whose forces, with the support of Jootee, expelled Mahdists from Qellem and all adjacent territories. Jootee became governor of the entire southwestern Wallagga region, which was renamed Qellem after his birthplace in Gidaamii district. Goobanaa bestowed on Jootee the Ethiopian politico-military title, dejjazmach. Thus, Qellem became a province of the Ethiopian empire. Jootee agreed to pay annual tribute to Emperor Menilek, convert to Orthodox Christianity, and allow and support the introduction and expansion of the Ethiopian Orthodox Church.¹⁸ Such was what one might call Ethiopia's negotiated conquest of Qellem and the arrival in the province of the Ethiopian empire-state around 1886.

The neftegna-gebbar system (1908–12, 1917–33)

At different points during the two decades following 1890, various parts of southern Ethiopia were subjected to strict dominion under the neftegna-gebbar system. In the case of Qellem, the system was associated with two separate periods. The first was between 1908, when the province's governor, Jootee Tulluu, was imprisoned, and 1912, when he was released and reinstated. The second round began in 1917, after Jootee was removed, and continued until it was lifted and replaced with a property tax in 1933.

The neftegna-gebbar system, broadly speaking, was a system of rule where regions that violently resisted Ethiopia's territorial conquest in the south were placed under military commanders and soldiers who battered them into submission. The empire would settle its soldiers in large numbers, and the local population would be counted and distributed as vassals of the military settlers according to their ranks. The imperial soldiers’ demand for resources and labour from the subjects handed to them in lieu of salary would be unlimited.¹⁹ A key and favoured Oromo vassal of Menilek before 1908, Jootee Tulluu was able to prevent the introduction of this Amharic system of military clientelism. In 1908 his Amhara rivals in the adjacent provinces accused him of surreptitious dealing to sell part of his province to the British Sudan; they produced a forged document in alliance with his daughter, Birriituu. Exploiting their networks at the Addis Ababa palace they appear to have persuaded Menilek.²⁰ In a letter he wrote following Jootee's imprisonment, apparently to reassure the local elites in Qellem, the emperor concealed the actual accusation levelled against Jootee and instead referred to acts of misgovernment and maladministration, the embezzlement of government money, and the prevalence of injustice in his domain, although all such matters were the hallmark of the empire.²¹ While oral sources support that Jootee was arrears in his annual tribute payments to Addis Ababa, allegations about his harsh punishments were very much familiar to multiple local sources.²² There are two plausible reasons for Jootee's imprisonment. First, his incarceration would secure control of the lucrative trade routes in western Ethiopia for his enemies in the adjacent provinces. Second, for Menilek, Jootee was becoming too powerful, controlling essential resources and trading connections with the outside world such that his actions overstepped his status as a vassal of Addis Ababa. Since Menilek was very cautious about frontier territories, it appears that he was no longer confident in keeping Jootee in power after 1908. Menilek gave Jootee's domain to Negadras Haile-Giyorgis, minister of foreign affairs and commerce, and appointed the Amhara governor of the neighbouring Iluu Abbaaboor, Ras Tesemma Nadew, as overseer. Each peasant family, now gebbar (tribute-paying peasant), was counted, registered, and redistributed under the Amhara colonists. Menilek himself wrote the directive under which his military colonists would be ‘planted’ in Qellem.²³ Jootee's imprisonment, release and reinstatement, and his eventual removal, as discussed below, would turn out to be monumental, marking turning points in the history of the province.

Emperor Menilek now ordered the overseer Tesemma Nadew to distribute economic resources — which in real terms meant peasants — among the imperial soldiers in Qellem. The peasants were to be ‘eaten’ (reflecting imperial Ethiopia's Amharic concept of wealth and labour, and in practice representing the wage the soldiers had to live on). The saying exemplified the system under which they were appointed ‘to eat’ the region. Ager Bella, an Amharic expression which meant ‘eating a country’ — in which one was sent as teklegna, a colonist, beyond the Abyssinian homeland to a specific province in the conquered south — was a common saying among soldiers in Qellem.²⁴ Unlike the region's negotiated autonomous status when it was initially incorporated into the Ethiopian Empire, under which its Oromo governors remained in power and the old socioeconomic structures remained intact, the neftegna-gebbar system demanded a new politico-military structure under which the neftegna emerged as a ruling class of military colonists, reducing the Oromo governors to local co-opted subsidiaries. Most peasants remained on the land they farmed but with a considerably reduced status. They continued to pay taxes, but there was now no limit to the amount or type of produce or labour they had to offer their Amhara overlords. Thus, the peasants became a subjugated class. Although the system reflected power relations in the Ethiopian empire, the process of implementation, local dynamics, and Qellem's exceptional historical experiences produced unique consequences. Local elite and peasant opposition to imperial Ethiopia's exploitation, tensions in state-society relations, and local state constructs’ dynamics had their roots in this historical process whose detailed analysis neither the Great Tradition nor Oromo historiographies wished to represent.

Two men — Abba Shawul and Fitawrari Sahle-Giorgis²⁵ — were assigned to govern the province on behalf of the absentee governor Haile-Giyorgis. However, the arrival of Abba Shawul brought confusion in the division of duties between himself (Abba Shawul) and Fitawrari Sahle-Giorgis, who was appointed to run the civil administration alongside the military occupation that ruled by force of arms. During his stay, Abba Shawul repeatedly disobeyed Addis Ababa's orders that were channelled to him through Sahle-Giorgis.²⁶ This period of administrative confusion overlaps the rebellion of 1909–12 in the province, playing a part in the violence. Sahle-Giorgis and Abba Shawul personified divisions at the political centre. Abba Shawul represented the imperial potentate Lij Iyyasu, and Sahle-Giorgis his brother, Tesemma Nadew. Fragmented, fractured, and personalised hierarchies arose where authorities should have been representing a centralised imperial Ethiopian state. Although such structures often undermined the interests of the imperial government in Addis Ababa, they benefited individual and group forces who acted in the latter's name and helped the imperial system maintain the loyalty of local individual and group forces. The survival of imperial Ethiopia's system of rule in part relied on allowing a degree of autonomy to local socio-economic forces, so the tensions at the state level increased the power ultimately exercised by governors and military commanders in reasserting control over local powerbrokers in the conquered south.

In 1909–12, an armed rebellion broke out in the northern part of Qellem, leading to the disintegration of the Ethiopian administrative system.²⁷ In 1912 Jootee was released and reinstated, and initial arrangements were restored. Nevertheless, in 1917, he was permanently removed, and Qellem was placed under a brutal variation of the neftegna-gebbar system.²⁸ Since the extraction of local wealth was the ultimate purpose of territorial conquests and military colonisation, the peasants were key factors during this interaction. A new Amhara governor, Dejjazmach Berru, arrived in October. Berru's force of seven to eight thousand neftegna occupied the region. This force was an addition to those who had settled there in 1908 and who had remained after Jootee's reinstatement.²⁹ This marked the beginning of the second period of the neftegna-gebbar system. By finally placing the province under direct Amhara rule, this period signalled the termination of Qellem's local autonomy.

The job of reallocating the peasants among the neftegna retinue resumed in the same way as it had in 1908. Soon the true face of Ethiopian colonialism that the Oromo of Qellem had never experienced under Jootee started to unfold. Peasants who had fled to the forests during the first phase rather than face unlimited taxation, corvee labour, and service in the neftegna house, now fled again, this time in greater numbers. Nearly 9,000 of them migrated to the Begii highlands.³⁰ Conceived by the peasants as a continuation of resistance to the system, this mass outmigration was the first reaction to the reintroduction of exploitation. Such strategies were common in early colonial Africa: in colonial northwestern Tanzania, for example, as the government sought to fight sleeping sickness by identifying patients through networks of power, the local population chose to emigrate out of Kiziba as a rejection of royal authority and its cooperation with the colonial authorities.³¹ A detailed description of peasant resistance in Qellem is beyond the scope of this article, but forms of resistance included outmigration, banditry, and armed rebellion. The court litigation discussed below represents only one example of resistance.

Every month an Oromo gebbar was required to pay four Maria Theresa thalers and five qunna of grain, peas, beans, or cowpeas (depending on the season) to the imperial soldier to whom he had been assigned in lieu of salary.³² The Oromo gebbar had to work two days a week for the military colonists, who sometimes demanded cash instead of labour. Additional provisions would be required if a military colonist of a certain rank went on a trip for his master's or his own purposes. Wives were required to grind grain and bring firewood, while children under fifteen worked at their melkegna's (colonist's) residence. The gebbar had to construct fences and build houses when required, and feed horses and mules.³³ All this was in addition to the state tax demanded of every gebbar household.³⁴

The system disrupted social and economic life in Qellem. Since everyone over fifteen was registered to serve as an ‘economic resource to be eaten’ by the colonists, disruption became inevitable. The Macca Oromo families relied on household labour, and cooperative labour played an equally important role. In the pre-Amhara days, the Oromo possessed many social and economic cooperation institutions that survived the Ethiopian conquest. Daboo and daadoo were the most important of these, although still practiced today these institutions were dissolved during imposition of the worst form of the neftegna-gebbar system in 1917–33. Daboo was a form of cooperation in which between 10 and 40 people, and sometimes more, came together for a day of unpaid work when the head of a peasant household asked the neighbourhood for help working on his farm, building a house, or making a road.³⁵ Daadoo usually comprised a smaller group whose members entered into long-term voluntary reciprocal cooperation on equal terms. The members worked for each other in turn and, like with daboo, the labour was unpaid. Both practices were immediately stopped by a system that considered local people economic assets and reallocated them to different neftegna. By so doing, the system undermined the existing culture of hard work that had been fuelled by incentives driven by the labour provided by the associations.

Another important consequence of the neftegna-gebbar system was the blurring of social distinctions that had previously developed among the Macca Oromo. Before the Ethiopian conquest, the Macca Oromo were composed of four social groups of varying status: the Boorana, the Gabaroo, the slaves, and the despised occupational castes. The Boorana were considered few and believed to be of ‘pure’ Oromo descent. They claimed to possess all the best traits of traditional Macca Oromo society: they considered themselves closer to Waaqa, God, for whom they acted as a medium.³⁶ Every qaalluu and qaallittii, male and female ritual leaders in the gadaa age-grade system respectively, were raised only from the Boorana class. The socioeconomic and political power associated with the Boorana group was destroyed under the neftegna-gebbar system, and the Boorana joined the Gabaroo as Oromo gebbar. As peoples of non-Oromo identity absorbed into the Oromo population during the great Oromo expansion in the sixteenth and seventeenth centuries, the Gabaroo occupied lower socioeconomic status. Although the Gabaroo made successful cavalry that helped the pastoral Oromo to conquer territory, they were kept at a reduced status, prompting them to repeatedly rebel. Their status seems to have improved over time, but they were still not equals of the Boorana during the Ethiopian conquest of the nineteenth century.³⁷ The result of neftegna-gebbar was mixed for the slaves, some of whom were classified as gebbar along with their masters, while the colonists reclassified the rest as slaves. The social construct of castes, including tumtuu (smiths) and faaqii (tanners), remained, despite the similarities of their physical appearance to the Gabaroo.³⁸ In short, the neftegna-gebbar system dissolved three classes, subjecting every peasant household to rule by the Amhara colonists, apart from local elites who were co-opted to the new system with significantly reduced status. The age-old Oromo social hierarchy had been dissolved, and the imposition of Amhara domination on Qellem affected social structures, provoking local responses manifested through the mobilisation of alliance and counter-alliance. How this helped the new regime to take root is not well understood in the Great Tradition and Oromo historiographies.

The history of the neftegna-gebbar system in Qellem in one way represents imperial Ethiopia's political centralisation. It illuminates the growth and expansion of state power, through government penetration at the local level and the responses it generated. Amhara, like the Tutsi in colonial Rwanda, were entrusted with substantial governmental authority and ‘used these powers to enhance their control over key resources: land, cattle, and people’.³⁹ Equally important, these transformations linked to the strengthening of the imperial Ethiopian state were vital in the development of political consciousness among the Oromo elites and population in Qellem, and in the formation of a politics in modern Ethiopia that was deeply informed by ethnocultural identities.

The responses

When the Amhara governor of Qellem, Dejjazmach Berru, was eventually recalled to Addis Ababa and arrested (historical records do not indicate why), there was widespread hope for a form of administration better than the neftegna-gebbar system. Addis Ababa entrusted the province to Dejjazmach Gebre-Egrzi'abhér Morodaa, hereditary governor of the neighbouring Leeqaa-Naqamtee, during an interim period lasting from September to December 1920.⁴⁰ The system was maintained, however, and another Amhara governor, Neggadras Yiggezu replaced Gebre-Egrzi'abhér in January 1921. As soon as the new governor arrived, the news spread throughout the province, dashing any hopes the Oromo gebbar who had returned home might have had. Those who had experienced neftegna-gebbar under Berru believed the appointment of a new Amhara governor would be catastrophic, although there was no specific evidence of any visible danger. Yiggezu found a province that the brutally exploitative system had all but destroyed. One report stated that ‘the country is already deserted and mostly abandoned, with a few scattered households left. The people have long abandoned their farm work’.⁴¹ When Yiggezu arrived, there was only a thin line between revolt and acceptance, and there was a pressing need for collective action.

Instead of withdrawing, as in 1909 and 1917, local Oromo elites, the balabbat, along with a few peasant representatives, chose to counter the neftegna-gebbar system by legal means. Their aims were to redress the wrongs they had suffered under the system and to demand the restoration of the old system of hambifataa (from the Oromo word hambifachuu, to retain), where the balabbat benefited from retaining a number of peasant households whom they could tax.⁴²

We now return to a story featured at the outset of the article. In February 1921, a group of Oromo balabbat and peasants presented a case to Neggadras Yiggezu protesting the neftegna-gebbar system. The balabbat served as advocates, representing their regional coordinators. The neftegna also had their own advocates. The case was presented at Gaawoo, Yiggezu's administrative centre. The Oromo balabbat from Gidaamii (the home district of former governor Jootee) employed an Amhara who was knowledgeable in imperial Ethiopia's traditional laws.⁴³ Each neftegna group from various localities of Qellem appointed advocates. One of the neftegna participants, Abba Aberra, summarised the process of deliberation at Yiggezu's court:

There were (Oromo) Balabbat. They rose against us when their people were divided [among the neftegna]. We debated this hotly. The soldiers [neftegna] debated against the [Oromo] Balabbat. The Qelläm people brought an advocate called Towki Shumale and others from Qaqqe, Dambi Dollo. They [the Oromo balabbat and gebbar] said: ‘We will not come under you. Our fathers were also neftegna. When their fathers died, their children lifted up their fathers’ guns and claimed to be Näfţägna’ …

Then we started debating with that Amante Gutama [one of the advocates representing Oromo balabbat and peasants]… In the debate he [Fitawrari Wäldä-Mika'el] held with them, he said: ‘Even if you claim that you have gone to the Näch Abbay [White Nile] with Däjjach Joté [Jootee], we cannot help you since your own master doesn't rule anymore and [the former Amhara governor] Däjjach Birru has been given the land. It is ours by law. So, since we were given your land, you have to serve us and pay tax to us’. So, the case in the court was ended, and a judgement was made in favour of the neftegna.⁴⁴

On the surface, the above text proposes a simple court argument between the entourage of the former hereditary Oromo governor, Dejjazmach Jootee, and the neftegna over status. However, a closer examination of the local socioeconomic and political scene suggests a more complex situation. In 1898, as part of his diplomatic manoeuvres aspiring to further the outer limits of his empire's western frontiers by exploiting the French-British standoff over the Nile and Sudan, Menilek mobilised his forces in the west. The imperial confrontation culminated in the town of Fashoda, Sudan. While the French had a grandiose plan of building an empire that would merge their colonies in eastern and western Africa, the British wanted to extend their hold over Egypt further south to control the Nile confluence and the Mahdist Sudan. Menilek, on his part, pretended to be an ally to all the three powers involved: the Mahdist Sudan, the French, and the British. He assured the Mahdist Sudan of his support based on African brotherhood in case of any incident. For the British, he agreed to impose an arms blockade on the Mahdist Sudan and promised to deny the French a passage across Ethiopia in their ambition to march to the Nile confluence. And to the French, he promised to allow them to cross through his empire and provide them with logistic support as their forces marched to the Nile. Menilek's real ambition was to exploit the situation to conquer territories on his empire's western frontiers. The French mobilised two missions, one from eastern Africa (the Marquis de Bonchamps mission) and the other from western Africa (Jean-Baptiste Marchand mission), scheduled to meet at the Nile confluence in Sudan. In the name of providing logistical support to the Bonchamps mission that was marching from Djibouti across Ethiopia to join with the Marchand mission, Menilek sent a contingent headed by Ras Tesemma Nadew with the French.⁴⁵

Jootee was involved and helped Tesemma's forces succeed, but not in the way Oromo balabbat and peasant representatives argued in the court. The lowland territories of the White Nile were known for malaria, and most of Jootee's entourage did not join the march. Instead, Jootee recruited hunters from among the local peasants who had experience in the lowland White Nile region, relying on their physical strength and knowledge of the route through the lowlands up to and beyond Gambella (to Qellem's south).⁴⁶ The Ethiopian empire's forces under the command of Tesemma established Menilek's authority in the Baro River region, including in Nasir, just short of the confluence of the Sobat River with White Nile. Menilek's diplomatic machinations gave rise to the Fashoda incident of 1898, leading to London-Paris negotiations to end the standoff. The Mahdist state, meanwhile, suffered a crippling defeat at the hands of British Forces at the Battle of Omdurman.⁴⁷

The Oromo balabbat and peasant representatives at the court claimed neftegna status based on their participation in the 1898 Fashoda expedition that hoisted the Ethiopian empire's flag in the Baro River region. While both parties were aware of the absence of any direct military contribution by Jootee, the balabbat were putting forward their demands based on the recruitment role they had played in their respective jurisdictions. The peasants, now gebbar, made their claims based on sending ‘their fathers and sons’ on military expeditions through which the Ethiopian empire had extended its western borders at the end of the nineteenth century. In the court argument, these ‘fathers and sons’ from the past were presented as gun-bearers of the Ethiopian empire-state, through which both the balabbat and the gebbar had earned equal status with Berru's neftegna, and hence inherited the status themselves. The balabbat, by coordinating the recruitment order that came to them from Addis Ababa via Jootee, and the peasants by contributing their ‘fathers and sons’, should therefore, they argued, be relieved from the neftegna-gebbar system.

Although the Oromo balabbat and peasants allied in going to Yiggezu's court and refusing to submit to the neftegna, they were not wholly united in their aims. By refusing to acknowledge Yiggezu's neftegna and claiming to be neftegna themselves, the balabbat sought to be restored to what they asserted was their old status prior to the imposition of the neftegna-gebbar system in 1917. Accordingly, they asserted their right to have sisso (freehold land), collect taxes, pass state orders to the locals, and be given gebbar (peasant households) on whom they would impose various forms of taxes as their own exclusive revenue. They also demanded labour services as a reward for their decisive intermediary role in the imperial system. The peasants wanted relief from the imperial system that subjected them to exploitative and brutal neftegna lords, and the removal of the Amhara imperial soldiers. Their goal was the restoration of the pre-1917 system where they were required to pay taxes on their land and produce, and might be required occasionally to provide labour services to local Oromo elites. While the balabbat sought to be restored as local elites with high status who would influence local developments, the peasants wanted to return to their former status as imperial subjects.

The arguments featured at Yiggezu's court issued claims extending far beyond local politics. The notion that Oromo balabbat and peasants should have the same status as Menilek's neftegna called for the repudiation not reform of the neftegna-gebbar system. The plaintiffs’ argument sought to persuade the Ethiopian empire to drop the system of rule that was imposed in almost all territories in the conquered south. The argument at Yiggezu's court dealt with a national issue critical for the very survival of the Ethiopian imperial system.

In their legal argument, the Oromo deliberately selected aspects from the past that they believed would be the most powerful in defending their interests, and conveniently forgot other parts of their history. The balabbat's status, for example, evolved twice before the Ethiopian conquest. Between their emergence in the mid-nineteenth century and the century's end they had been de facto local chiefs (then known as abbaa-qabiyyee). Under Jootee they were administrators of several villages subordinated to his authority. As Jootee submitted to the Ethiopian empire through a negotiated conquest, their status survived, but their office was renamed balabbat and their status was reconfigured to fit imperial Ethiopian administrative structures. This reconfiguration introduced a system under which the old abbaa-qabiyyee had to serve the empire in their respective jurisdictions, including the recruitment of soldiers when necessary, in return for the preservation of their old status.⁴⁸ At Yiggezu's court, the balabbat neglected to mention that their pre-Ethiopian sociopolitical status had been reconfigured and co-opted to the empire-state — where providing a type of service to the Ethiopian soldiers who occupied the Baro River region, hoisting the empire's flag, was part of the duties they agreed to when their former overlord Jootee submitted to the Ethiopian empire. The balabbat's representation of their preconquest status argued for reclaiming a better system from the past to displace neftegna-gebbar. This selective reading of the past utilised the litigants’ knowledge of the empire's philosophical foundation, which it conferred upon its colonial forces the power to ‘eat the people as economic resources’. The Oromo elites and cultivators used this argument at Yiggezu's court to protest the neftegna-gebbar system.

The imperial army — the neftegna — had the judges and the entire imperial system on its side, but nonetheless defended its position strongly, choosing advocates for each district of Qellem to represent them at Yiggezu's court. The neftegna group did not attempt to deny the participation by Oromo balabbat and peasants in the imperial Ethiopian forces’ march to the Nile. Instead, they argued that the Oromo gun-bearers of Qellem were involved as soldiers of the deposed Dejjazmach Jootee, whereas they as neftegna were soldiers of the Ethiopian empire-state and therefore qualified for higher socioeconomic and political status. Hence, the argument went, both the balabbat and the local Oromo participants should be subjected to the neftegna-gebbar system with full obligations.

Dejjazmach Yiggezu's ruling dismissed the cultivators’ claim of participating in the Fashoda expedition through their Oromo ‘sons and fathers’, and confirmed their gebbar status.⁴⁹ The balabbat who had not recruited soldiers for imperial expansion were reduced to gebbar status too. However, Yiggezu offered minor concessions to the balabbat, who had demanded neftegna status due to their recruiting soldiers from their respective jurisdictions. They were freed from gebbar status, but required to pay tithes, placing them above gebbar but below the neftegna in the sociopolitical hierarchy. The balabbat's request to be restored to their previous elite status was rejected, but their freedom from gebbar obligations made them candidates for a new form of alliance. Yiggezu's court ruling distinguished the balabbat from the peasants, enabling them to help govern the province. Yiggezu showed his political pragmatism by restoring an Oromo class of co-opted intermediaries in a reduced form. Yiggezu's ruling was an informed verdict that would maintain the socioeconomic and political status of the neftegna but reinvented a local alliance that would help govern the region and exploit its resources. Yiggezu's decision was therefore dictated more by political considerations than justice.

Following Yiggezu's verdict, the neftegna-gebbar system allowed the balabbat to retain some gebbar, reflecting the old Oromo system of hambifataa. Sources determining the specific number of gebbar households they were allowed to retain are scant, but the number varied between localities and was generally less than the number the neftegna were awarded. Under the hambifataa system, the gebbar population paid tax to the state based on their agricultural produce and number of cattle owned. They gave their Oromo overlord a quarter of their produce and paid him one Maria Theresa Dollar and a day or two of labour per annum. The balabbat became the Oromo melkegna (colonists), and occupied an intermediary stratum between the gebbar and the neftegna.⁵⁰ While the hambifataa system replaced the neftegna-gebbar system in the domains of the balabbat favoured by Yiggezu's verdict, the neftegna-gebbar system was sustained in the rest of Qellem until 1933. The hambifataa system, as an alternative to the neftegna-gebbar system, was far less exploitative and served as a refuge for Oromo peasants fleeing the neftegna-gebbar system.

Compromise, bargain, and negotiation as tools of state construction

Imperial Ethiopia's governance structures and its ability to rule resulted from dynamic local relations involving negotiation and renegotiation as much as conflict and violence. Despite the imperial government's claim of an all-powerful presence and the Great Tradition's narratives displaying a smooth, harmonious, and locally-accepted state structure, the empire's rule had been applied at the local level through indigenous societal dynamics, and bargaining with local forces. This resonates to some extent with experiences in French West Africa, where ‘the intersections of local relations of power with colonial ones … show that the colonial state functioned quite differently day to day than the French officials often knew or wished to acknowledge’, suggesting that the ‘French hierarchy of rule was not as straightforward and orderly as it appeared from above’.⁵¹

Recognising that the local Oromo elites, balabbat, played a key part in the local negotiations and bargains, benefiting considerably in the process, and eventually helping the Ethiopian empire build state structures in conquered territory should not lead us to conclude that these were the only intermediary roles they played between the state, governors, and the local population. Emily Lynn Osborn, Benjamin Lawrance, Mari Webel, Cherry Leonardi, and Chris Vaughan have examined how employment in other colonial settings served as an engine of social and economic mobility, leading to the birth in colonial Africa of a new class of state functionaries during the twentieth century.⁵² These important studies have revealed the part various African intermediaries played in making colonial states in Africa. However, the intermediary roles of African state employees in other colonial settings differed substantially from the roles balabbat and peasant representatives played in Qellem — and, perhaps, throughout Ethiopia's conquered south. While the social, economic, and political power of the balabbat were reinforced by the state's recognition of their status and possession of large tracts of land, they differed from many other intermediaries in colonial Africa by remaining outside the formal state structure and not being salaried. The same was true of the peasant representatives who spoke and acted on behalf of their own class. These representatives were either peasants, or people with a heavy peasant background, with no recognition from the state; but their role was crucial in presenting peasant petitions to the balabbat, governors, or other state actors, and bargaining on their behalf. The balabbat and peasant representatives illuminate how new alliances, negotiation, and networks of power determined the methods and processes through which Ethiopian imperial rule functioned at the local level.

In terms of what the Oromo gebbar wanted to achieve from Yiggezu's administration, their desires may not have been as rigidly constructed as their court rhetoric implied. During Gebre-Egrzi'abhér's brief tenure (September-December 1920), thousands abandoned their refugee lives and returned home. Those who had stayed to endure the system accepted Gebre-Egrzi'abhér's authority, although he did not lift the neftegna-gebbar system. There was still dissatisfaction, but Gebre-Egrzi'abhér controlled extreme and irregular abuses, silencing the peasants, and keeping them in their homesteads.⁵³ The gebbar apparently hoped to obtain confirmation from Yiggezu of the improved conditions Gebre-Egrzi'abhér had put in place. When news of Yiggezu's judicial verdict spread, many of the Oromo gebbar who had experienced the worst aspects of the neftegna-gebbar system saw it as a harbinger of more bad times ahead.⁵⁴ The continuation of the neftegna-gebbar system meant the addition of more imperial soldiers over each peasant household because the peasants were an economic resource for the soldiers to live on. As an important representative at the local level, Yiggezu had to negotiate and reconstruct authority as a dynamic tool in a way that could engage local actors, rather than continue the methods of his predecessor.

By taking their case to Yiggezu's court, the balabbat and peasant representatives tried to turn political and socioeconomic changes to their advantage. Taking the neftegna-gebbar system to judicial review suggests that people imagined that by virtue of being under the rule of imperial Ethiopia, they were entitled to at least some rights both individually and collectively. Pursuing recognition by imperial Ethiopia demonstrates a commitment to engaging with and claiming state power. Importantly, the balabbat and the peasants recognised the state apparatus imperial Ethiopia had installed after removing Jootee. The case of Qellem's attempt to obtain rights and recognition through litigation at Yiggezu's court represents the earliest available record of how local people chose to communicate with the state. Local plaintiffs’ arguments against the colonists demonstrate the process of negotiating subjecthood and, even, claiming citizenship within the Ethiopian empire.

The local people's cause was, broadly speaking, rejected, but set a precedent that would influence state-society encounters over the following decades. The encounter at Yiggezu's court demonstrated the expectation that the state should recognise people's demands and address their grievances. The court proceeding and decision can be understood as early examples of local pushback against state exploitation and subjugation, the central government's agents’ craft in handling such claims, and actors’ ability and willingness to negotiate and renegotiate the terms of rule. Moreover, it is an indication of how the processes of state construction were embedded in local contexts in imperial Ethiopia. Overall, the process became a local tool that helped imperial Ethiopia's project of state construction, not in the way it was designed in Addis Ababa but as it was negotiated between agents of the state and local forces. As Jocelyn Alexander has argued in relation to Zimbabwe, ‘[i]t is in the local struggles over the power and authority that states must take root’.⁵⁵ In line with suggestions by recent Africanist literature, these local interactions were pivotal to the process of state formation. The encounters at Yiggezu's court went beyond stories of subjugation, conquest, and exploitation, and demonstrate the place of the state in society. In this sense, the state is not a fixed entity but instead represents ‘ideas, discourses and imaginaries, as well as institutions, actors and processes’.⁵⁶ Arguments that state construction has not been limited to its core fit well with the example of Qellem, where a political system was shaped, reshaped, contested, and resisted as much at the local level as it was at the centre.

Like European colonial states in Africa, the ambiguity, incoherence, and often contradictory character of structures and processes are key characteristics of the imperial state of Ethiopia. This demonstrates ‘contradictory social forces’ working in imperial Ethiopia's conquered south, and these social forces had on one side a considerable degree of influence in the evolution, consolidation, and eventually survival of the imperial state, and on the other possessed inroads through which they were themselves reshaped and altered by the imperial state.⁵⁷ The imperial state in Qellem was thus always fragmented, fractured, incoherent, inconsistent, and tension-ridden. In most cases, state officials were participants in rather than arbiters of the contestation, bargaining, and negotiation which characterised the politics in Qellem.⁵⁸

The neftegna-gebbar system in Qellem ended in 1932 when it bred more violence that made the region almost ungovernable again. Emperor Haile-Sellasse replaced the last Amhara governor, Dejjazmach Mäkonnen, with Dejjazmach Habte-Mariam, son of an Oromo governor of the neighbouring Naqamtee province. Habte-Mariam reorganised Qellem's administrative structure, introducing a property tax system after that of his home province of Naqamtee, and ruled in relative stability.⁵⁹

Building on this, local Oromo elites, presumably under the guidance of Habte-Mariam, resurrected the legal case they had lost in 1921, and petitioned for the abolition of the neftegna-gebbar system and the introduction of the qelad system, a system of land appropriation where all agricultural land would be measured and redistributed among state, civil, and military authorities, the Ethiopian Orthodox Church, and the local elites (balabbat), turning the peasants into landless tenants.⁶⁰ In the same petition, their old Oromo overlord, Jootee Tulluu, was presented as a man whose mistakes had reduced them to gebbar status.⁶¹ Although the Oromo balabbat did not represent the interests of the common people, the demand they presented to the empire — the abolition of the neftegna-gebbar system — was a significant one. The emperor had to compromise. He agreed that the gebbar must be relieved, but the qelad system — which would require agricultural land measurement across the entire province — was deferred, and would not be realised until 1944.⁶² Instead, Habte-Mariam introduced a property tax that was more liberal than the neftegna-gebbar system.

The introduction of the property tax benefitted the Oromo balabbat and the peasants in different ways. The balabbat enjoyed lighter tax demands until Habte-Mariam's departure. The peasants were freed from the unlimited exploitation by the imperial soldiers to whom they were assigned as resources, and were instead allowed to pay taxes on their produce and livestock directly to the state. In August 1933, Qellem came under the direct control of the emperor as part of his centralisation policy designed to secure the revenue from the region, and Fitawrari Ashenafi, the region's first salaried official, replaced Habte-Mariam.⁶³

By terminating the neftegna-gebbar system and introducing a property tax instead, Dejjazmach Habte-Mariam achieved substantial success in engineering a greater degree of local compliance with state rule. He removed the brutal neftegna soldiers and governors, winning him immediate peasant support. He promoted local Oromo elites who were already recognised as intermediaries through Yiggezu's verdict and brought those who had lost such status under the neftegna-gebbar system into the same class, reorganising them and reconnecting them to the imperial Ethiopian state structure. In so doing, Habte-Mariam expanded and consolidated the state's embeddedness in society whilst simultaneously enabling society to reshape the state and expanding how one could be shaped by the other. The socioeconomic and political identity of local Oromo elites in Qellem resembles those of African intermediaries within European colonial states, in what has been referred to as a ‘bargain of collaboration’.⁶⁴ Local Oromo elites helped mediate interactions between local customary laws and state laws, crafting old rights in new ways, negotiating access to resources, and helping the state's construction. The office of balabbat became vital in imperial Ethiopia's state architecture and in the state's interaction with its subjects, especially in Qellem.

Erecting and maintaining structures of direct administrative control in rural Qellem, as well as keeping peace and order, were contingent upon the implicit acceptance of central rule by the local population and a ‘bargain of collaboration’ with their elites. In this political system, Amhara agents of imperial Ethiopia, despite their monopoly of the formal corridors of power at local and national levels, were often only one of the participants, and were not the ultimate arbiters. Thus, ruling and maintaining domination in the provinces, as was the case in many European colonial states in Africa, required much more than the employment of brute force. ‘Colonial domination’, as Bruce Berman notes, ‘turns out … to have been an extraordinarily complex social [and economic] process involving far more than the use of force’.⁶⁵ The case of Qellem represents how an African empire contemporaneous to European empires on the continent regenerated itself in the conquered south and how it negotiated with indigenous peoples and their elites.

The case study examined in this article demonstrates that problems between the central imperial government and local socio-economic forces define the state's contradictory features of centralisation and fragmentation. The frictions, tensions, and inconsistencies inherent in the channels of state apparatus, and links between the indigenous elites and the local population, allowed local agents of the imperial state and indigenous co-opted classes to enjoy autonomy and discretion until the 1974 revolution. In other words, imperial policies and decisions that were passed from Addis Ababa to the local level were compromised and negotiated — and occasionally even reversed.

Conclusion

Several challenges undermine analysis of imperial Ethiopia's state-society dynamics at the local level, but the most important one emanates from Ethiopia's two major historiographical traditions — the Ethiopianist Great Tradition and the Oromo one. Both historiographies present contending and clashing interpretations, emphasising histories that can make sense in the national politics of Ethiopia by neglecting the more complex historical interactions that can be read from the analysis of dynamic local interactions contextualised in the broader African historiography. The Ethiopianist Great Tradition has been confined to a celebration of Abyssinia-cum-Ethiopia's developed literary culture as exceptional in sub-Saharan Africa, analyses of the Abyssinian state and its central institutions, defence of state independence, and a history of tradition running back to classical times. The Oromo historiography rejected the Ethiopianist interpretation as an imperial narrative bent on celebrating colonial institutions and history, spending all of its energies on constructing an alternative historical narrative that could be weaponised in promoting Oromo culture and the national struggle for self-determination. The problem is that both historiographies have kept the debate isolated from consideration of African historical contexts and scholarship, and made little attempt to examine the roles of local state and non-state actors whose competition and compromise have significantly shaped the nature and content of the state in Ethiopia.

The histories of exploitation in Qellem and the responses it generated connects with the role of intermediaries and resistance in colonial Africa. The history of Qellem illustrates the shifting roles of intermediaries based on the social and economic benefits they obtained from conflicts or negotiations, and demonstrates how competition over resources, and the bargaining which resulted, ultimately shaped the state-building project itself. As Ethiopia's central imperial government moved to end Qellem's local autonomy based on its military and administrative capacities, it was able to install a brutal exploitative apparatus, the neftegna-gebbar system, but governing the region became precarious. The early responses were peasant mass withdrawal to the neighbouring province and armed rebellion, giving way to court argument, and later banditry, the cumulative effects of which forced Addis Ababa to change governors frequently until it finally dropped the neftegna-gebbar system in 1933. The neftegna-gebbar system, despite its severe exploitation and subjugation, was vital in creating a local opposition that would become a force to be reckoned with. However, the state also offered minor concessions to local Oromo elites who would serve as intermediaries between the state and the local population, creating new forms of alliance and structures of power. The history of this interaction illuminates a project of local state construction whose design was generated not according to a blueprint created in Addis Ababa but through creative local contestation. The combined role of intermediaries, peasants, and state actors in Qellem reveals that the imperial Ethiopian state functioned at a local level in ways which were quite different from the perception of imperial officials in Addis Ababa or what they wished to recognise. This historical process resonates with the role of African intermediaries and colonial employees across colonial Africa. Overall, the history of Qellem reveals a vital African state-making exercise that was contemporaneous with European colonial states on the continent, a historical process that gave birth to modern states on the continent.