From “Subject to Citizen”? History, Identity and Minority Citizenship: The Case of the Mao and Komo of Western Ethiopia

ion of multi-cultural realities. Accordingly, conflicts over ownership of territory have become a noticeable feature of the negotiations of political power in the new regions. The federal boundaries became contested spaces between regional states. As we have seen in the case of the Berta Oromo confrontation over Begi, federal boundaries are the product of majority groups’ claims over territory. The dispersed minorities on the other hand have to negotiate their political agency across these boundaries, which, in fact limits their effective representation in the current federal framework. 44 Paragraph four provides for the right to self-determination up to secession. This is yet another pattern of the interplay of ethnicity and development in Ethiopia and has to be omitted here. 45 Richard Yarwood, Citizenship, Key Ideas in Geography (London – New York: Routledge, 2013), 18. 46 For an interesting critique of the concept of territorial autonomy and ethnic conflicts, see Shaheen Mozaffar and James R. Scarrit, ‘Why Territorial Autonomy Is Not a Viable Option for Managing Ethnic Conflict in African Plural Societies’, in Identity and Territorial Autonomy in Plural Societies, ed. Ramón Máiz and Safran William (New York: Routledge, 2000), 230–53. 47 Yonatan Tesfaye Fessha and van der Beken, ‘Ethnic Federalism and Internal Minorities’.

Thailand was never a colony And an island is surrounded by a lake or a sea These are things we can say with some certainty And you will never escape from your history  […] there are many stateless societies where inaccurate definition has simply been the product of ignorance, illusion, or inattention; but very often the 'definition by illusion' has been a definition of larger scale which became permanently adopted for administrative convenience and ultimately accepted by the people themselves. We may thus say that the problems of illusion have frequently been perpetuated by those of transition and transformation. 1 Social labels in general are best treated as relative terms, which cannot be thought of as having a constant meaning or standard reference, but change in their particular implication and range of reference with the changing social situation. The varying use of these terms through history may be used as an index of changing social and political situations, rather than as an absolute measure of the existence, growth and migration of distinct social bodies. 2 Introduction well as the establishment of the Silte zone within the Southern Nations Nationalities and Peoples Region (SNNPR) through a constitutional referendum in 2001. This "critical test for citizenship-expansion" is a good example for the fascinating proximity of group identity and politics in Ethiopia today. 5 Despite this civil achievement, the parameters that make some groups politically successful while other groups are left fragile and fragmented remain to be analysed. Internal power struggles may contribute to these differences as much as the ever-changing patterns of identity formation in a politically ethnicised society. One way to approach this is to develop an inclusive understanding for minorities' integration in a multi-ethnic state based on the idea of historical minority citizenship: Minorities are not necessarily a cultural given; they are themselves a product of historical as well as socio-political circumstances. An ethnic group can neither be understood without the inherent cultural features and selfunderstanding, nor without considering the outside forces that enable or influence these groups. Hence the interaction between state and individual frames social memories and an inherent understanding of the position a group holds within a society.
The Mao and Komo, small ethnic minorities, are relatively little studied. Both groups have been subjects of various forms of state formation for centuries. Since 1991 they became subjects of citizenship formation in Ethiopia. In the current federal system, the Mao and the Komo have been endowed with formal citizenship rights as minorities. The genesis of their group identities vis-à-vis the historical processes of state encroachment and regional integration, the vertical stratification of the state (institutions) and citizens (locality) as well as the broadcasting of state power between the two are the main concern of this thesis.

Minorities and the State in Ethiopia
The conceptual pair minorities and the state has inspired much scholarship and the question, what holds nations together, moves policy makers, historians and anthropologists worldwide. 6 "Minority" is an ambiguous category. Minorities may live in the centre of the state or at its margins; sometimes they are border people, national or religious, gendered or otherwise distinguished from the mainstream society. They are usually numerical minorities, but they may not always be politically marginalized. There are also politically dominant minorities. 7 Usually though, the minority status does also relate to their possibilities regarding the expansion of citizenship. It affects the way in which they can claim rights or have rights; both positively (building on special treatment, consideration and constitutional provision), or negatively framed by marginalization or exclusion. Much scholarship concerning Ethiopian minorities -a category that gained prominence with the polarization of ethnicity under the current regime -has looked at the question "how does the state accommodate diversity". 8 Less work has focused on the perspective of the minorities: how do groups deal with the assigned status? Which demands and which benefits do they derive from the political framework? Although both perspectives take "minorities" (read any ethnic group in the Ethiopian context) as a cultural given, the latter viewpoint also includes -in a functional perspective -the presence of state and how it shaped groups and affected 6 For an overview on the anthropological arguments, e.g. Thomas Hylland Eriksen, Ethnicity and established power-relations between groups. Accompanied by these changes were high hopes and expectations for political self-determination of historically marginalized groups that have in some places evolved in violent confrontations in the struggle for ownership and control of the regional administrative units. Oromia, on the other hand, the Mao and Komo hold a rather precarious minority status, unacknowledged by the regional constitution (for a rough sketch of the distribution of ethnic groups, s. p…) This thesis is concerned with the often conflictual, highly complex and rarely acknowledged situation of fragile societies living separated across federal territories.
All over Ethiopia federalism has had a huge impact on the peripheries. Political and economic liberalization challenged existing power equilibriums. 14 At the same time the 14 James Ellison, '"Everyone Can Do as He Wants": Economic Liberalization and Emergent Forms of Antipathy in Southern Ethiopia ', American Ethnologist 33, no. 4 (2006): ; Elizabeth E. Watson, state has never had as tight a grip on the peripheries as under the current regime. 15 Overall the local communities may only play a limited role in the process of reconfiguration of politics. 16 Meaningful citizenship 17 is today articulated within the boundaries of the regional federal states. Ethiopia nowadays is divided into nine such states (killil). These administrative units are further subdivided in zones, wärädas and qäbäle. It is not uncommon that each level is assigned to an ethnic group (cp. the Amhara Regional State or Oromia Regional State, etc.), and in some cases the regional constitutions are amended in such a way that ethnic groups can claim zones or special wärädas (like in the given case of the Mao-Komo special wäräda). These districts are usually on the administrative level of the zone, but due to the small numbers of the ethnic groups are only designated as wärädas. 18 Hence on all administrative levels the rights of ethnic groups are potentially territorialized.
But the compartmentalization of political agency falls short in cases where ethnic groups such as the Mao and Komo, hampered by a splintered agency, live as scattered minorities under the constitutions of several regional states. Rarely have studies focused on the politics of fragile minorities with neither the numerical nor the ideological preconditions to struggle for self-determination. Even in the peripheral areas the weak are subjects of the history as told by the strong. Accordingly, highly complex social processes of integration, assimilation and social stratification are simplified in the eyes of current historical and political analysis of the western borderlands, 19 not to mention the relevance for the understanding of such processes for the study of regional or national claims for citizenship. I argue that a continuing epistemological problem in the narration of state formation in Ethiopian history, a lack of immersion in historical and political study, as well as the domestic discourse about ethnicity limited to primordial facts, have led to a rather technocratic understanding of current ethnic politics in Ethiopia. 20 Research Question Political choice in a plural society is a delicate issue. As much as politics are influenced by diverse local structures, local preconditions also affect political practice.
Accordingly, conflicts and contradictions within and between groups, enforced by ethnic federalism, are the result of the up-rooting of pre-existing social relations. Thus, different groups in the same area may both experience political choice and structure differently. These variances make the experience of citizenship expansion essentially difficult to understand, since citizenship promises political equality on a national level. 21 A key to understanding these variances may lie in the socio-historical process of inter-ethnic relations in plural societies. From this hypothesis, the following questions emerge: -What are the antecedents of group formation in the Ethiopian-Sudanese border area and which historical factors affect group distinction? -How did the inter-ethnic, as well as, the intra-ethnic relations between majority and minority groups evolve (e.g. Berta- Mao-Komo or Oromo- Mao-Komo relations)? How did the historical emergence and encroachment of the Ethiopian state since ca. the 1880s influence these relations? -In which way do historical experiences of marginalization among the Mao and Komo, conflate with current approaches of power sharing and political integration? 20 I take "technocratic" here to mean the understanding promoted by the political discourse in Ethiopia today, that takes ethnicity as a stable, cultural fact. 21 Julia Eckert, 'Introduction: Subjects of Citizenship', Citizenship Studies 15, no. 3-4 (2011): 309.

Significance of the Study
As a project in regional policy analysis, this study will fit into the field of comparative regional and ethnic studies. As it takes the perspective of the regional minorities and not of the majorities, it brings the periphery closer to the centre and connects national politics with its local effects. The significance of the study may also be measured by its contribution to the analysis of historical interaction of ethnic groups, and describing the models of coexistence between neighbouring groups on a regional level. Thus, it will contribute to the understanding of the driving forces behind inter-ethnic alliance and conflict generation vis-à-vis political choice.
Little attention has been awarded to the study of the region after all, not to mention the protagonists of this study, the Mao and Komo communities. Accordingly, the importance of the study is also based on the acquisition and presentation of ethnographic data on long neglected groups in western Ethiopia.

Limitations and Delimitations of the Study
The Mao and Komo, footnotes in any work on western Ethiopia, are the main focus of this thesis: specifically, their patterns of adaptation to political changes, alliance making and conflict vis-à-vis their neighbours and the state. The study will neither give a full account of all "Mao" groups in Ethiopia nor be able to shed substantial light on the variously pending linguistic questions on the different "Mao" languages. This study addresses such questions in passing, but focuses chiefly on the socio-political developments around the Mao-Komo special wäräda. Looking into Gambella and Oromia will benefit the comparison, without presenting results from "deep" stationary field work in all locations of the Mao and Komo. The study is based on multi-sited research but it does not deal with the Koma (Gwama and Komo) across the Sudanese border and limits itself to the Ethiopian Mao (mostly Gwama) and Komo, and will only briefly treat the cross-border issues. With regard to the recent arrival (since 2012) of Sudanese Koma in the refugee camps in the Mao-Komo special wäräda the study can neither give substantial information.
"Western Ethiopia " -The Geographical and Demographic Setting Along the north-western border with Sudan extends the regional state of Benishangul-Gumuz of the Federal Democratic Republic of Ethiopia. Together with the Southern Nations, Nationalities and Peoples Region (SNNPR) and Gambella it is one of the constitutionally recognised multi-national states of Ethiopia. In 1991 the region was amalgamated from the two former administrative regions Beni (formerly known as Bela) Shangul, or the Asosa awrajja, later forming parts of Wallaga, as well as the former Mätäkkäl district. Benishangul-Gumuz today comprises three zones (districts) and one special wäräda (sub-district). These are Mätäkkäl, Kamashi, Asosa and the Mao-Komo special wäräda (for the detailed admintrative division of the regional state, map on p. 243). The constitutionally recognized population comprises Gumuz, Berta, Shinasha, Mao and Komo. Additionally, many Oromo and Amhara live in the region.
My research largely focused on the Asosa zone with the Bambasi wäräda as well as on the Mao-Komo  They live across the border in Sudan also (where they are often referred to as Funji.) 23 There is a marked Sudan-Arabic influence recognizable in food habit, dress code, etc. and a strong lexical influence of Arabic on the Berta language (the self-designation of the language is rutan'a, the Arabic word for a (non-Arabic) local language). The Berta originated from the multi-ethnic Funj sultanate (1504multi-ethnic Funj sultanate ( -1821. They migrated into the Ethiopian highlands approximately during the late 17 th century. According to their oral traditions there they first encountered the Gwama (around the river Tumat) which they pushed further south. 24 Their country of origin lies between Fazughli and Roseires in present day Sudan. 22 While I think in parts this thesis would have benefited from more stationary and long-term observation in designated areas of the Tongo wäräda, I console myself with the fact that my research has profited from the coverage of largely separated settlements of the Mao and Komo in different areas. 23 A UNHCR Monthly Statistical Report for Tongo Refugee Camp counted approximately 10,000 Funji/Funj in September 2012. 24 Etymologically reference is made to the Gwama word tubatob ("to drink"), which became Tumat.
For this oral tradition of the Berta see Alessandro Triulzi, 'Myths and Rituals of the Ethiopian Bertha', in Peoples and Cultures of the Ethio-Sudan Borderlands, ed. M.L. Bender (East Lansing, MI: African Studies Center, Michigan State University, 1981), . See also Alfredo González-Ruibal, Once in the region the Bertha did not settle together. Rather each kindred group or clan, led by his own agur, settled in a scattered way mostly on tops of hills and low mountain areas. Traditions refer casually to the scattering of the new migrants and to their settling on different hilltops. This was probably due to several reasons, among which the coming at different stages and the need to settle in hideaway places with natural shelter form external threat. 25 Sudanese Arab traders followed the Berta, which led to the spread of Islam in the region. The influx of these traders, who knew the Berta as jäbällawin (Arab. "mountaineers"), led to mixed marriages from which emerged a feudal ruling class that came to be known as the watawit. When the Ethiopian Empire expanded and incorporated the sheikhdoms of Bela Shangul in 1898, sheikh Khojali al-Hasan became the most influential regional ruler. Khojali was a vassal of Addis Ababa until the 1930s and managed to maintain a semi-autonomous rule over Benishangul.
As the second biggest group in Benishangul, the Berta today occupy many political posts in the capital Asosa. The rural population depends on subsistence agriculture based mainly on sorghum and maize. Christianity generally plays a subordinate role in the region and the Berta are predominantly Muslim.
The main contestant for political power and the numerically most important group in Benishangul-Gumuz are the Gumuz. Their settlements extent north of the Berta, in the area of Mätäkkäl as well as in the Dabus valley. 26 Like the Berta, the Gumuz are also speakers of a Nilo-Saharan language. The Gumuz of Gubba also formed an aristocratic class with historical links to the Funj sultanate. The last leader of Gubba, sheikh Hamdan, also known as däjjazmatch Banja, was, like Khojali, a vassal of the central government in Addis Ababa. 27 His palace was used as a fort during the Italian occupation and was later bombed by British air raids during the War of Liberation An Archaeology of Resistance: Materiality and Time in an African Borderland (Lanham, ML, et al.: Rowman & Littlefield, 2014), 219. 25 Alessandro Triulzi, Salt, Gold, and Legitimacy: Prelude to the History of a No-Man's Land, Belā Shangul, Wallaggā, Ethiopia (Ca. 1800-1898 Lit-Verlag, 2010). 27 Peter Garretson, 'Manjil Hamdan Abu Shok (1898-1938

and the Administration of Gubba', in
Modern Ethiopia: From the Accession of Menelik II to the Present (Rotterdam: Balkema, 1982), 197-210. (1941). 28  discussed. This dispute looms large in the regional politics and is informative not only to understand the old territorial claims between the Oromo leader Jote Tullu and sheikh Khojali, but it also elucidates the distribution of the Mao and Komo as regional minorities across several regional borders. One can say that the migration of the Oromo who reached the area in the 18 th century had a tremendous impact on the demographic situation in the area. The Oromo of western Wallaga remember the Komo, Gwama and "Mao" as the original inhabitants of the area. In Oromia I specifically made research in the Begi wäräda (in the kebeles of Qama and Shonge), of Qellem zone, and I was able to make further small investigations in Gidami, in Dambi Dollo, and Muggi, as well as in Qama Shandi, Guma Gara Arba (cp. map. 247).
From Dambi Dollo one can reach Gambella through Muggi after several hours on the bus. The serpentine road winds down towards the lowlands of Gambella. Behind the federal border between Oromia and Gambella the small town of Bonga 29 is situated. In and around Bonga one may find a scattered community of Gwama speakers, who are referred to as Komo officially; some of the towns-people though 28 Alfredo González-Ruibal, 'Fascist Colonialism: The Archaeology of Italian Outposts in Western Ethiopia (1936-41) ', International Journal of Historical Archaeology 14, no. 4 (2010): 561. 29 Bonga appears as an area of gold panning and was a significant refugee centre during the Southern Sudanese Civil war. Bonga was abandoned as a refugee site in 2005 almost immediately with the signing of the Comprehensive Peace Agreement in Sudan.
have a strong link with the Oromo population from Illubabor and refer to themselves as Oromo (or "black Oromo"). From Bonga it is another hour ride to reach Gambella.
Gambella town has a significant Komo population. In Gambella the Komo count as one of the five indigenous groups, together with the Anywaa, the Nuer, the Majangir and the Opuo. Historically the Anywaa and the Komo lived together in close proximity. The Nuer on the other hand were historic enemies, who were also involved in the slave trade as middlemen. 30 In Gambella town, towards its eastern edge, there is a place called burr komo, "the hole of the Komo", which is slowly being reclaimed as a Komo settlement. From Gambella the road leads to Itang and southeast of Itang one will find the settlement of Pokung. Pokung is primarily a Komo inhabited kebele.

Entering the Field: Methodology and Research Design
The Mao are not one group but many. The term is ascribed to and used by the Gwama, the Hozo, Sezo, the Bambasi Mao, and beyond the area of investigation, also the Mao of Anfillo. 31 Mao is both an ethnic term, as well as a social ascription. As a social label the term is accepted by the political elite, it symbolizes a certain degree of coherence with political territory, and a group idea in the political arena of ethnicity and group definition. The term Mao is also one of the foremost complicating factors for ethnographic work in the region. It is sustained by the governmental discourse, but also by the encounter between different groups. Approaching oral data through such labels obstructs to a certain degree the knowledge we may gain. One may hear, 'the Mao did this and that, lived there and there', instead of referring to Gwama, Komo or else. It made it difficult also, especially in the beginning of the research, to understand exactly who one was talking to. Information such as "we speak Afaan Mao" ("the mouth/tongue of Mao" in Oromo) was difficult to comprehend. Sometimes people would say that they spoke Afaan Mao-Komo, an amalgamation and clear signal that they were referring to the identity discourse of the current administration in the Mao-Komo special wäräda. 30 Douglas Johnson, 'On the Nilotic Frontier: Imperial Ethiopia in the Southern Sudan, 1898-1936', in The Southern Marches of Imperial Ethiopia, ed. Donald Donham and Wendy James (Cambidge: Cambridge University Press, 1986), 219-45. 31 I will attempt to provide a more detailed differentiation of the groups in the second chapter.
When I first started to ask for 'the Komo' in Gambella I was conducting research mostly among the Anywaa and the migrant communities from the highlands. 32 The Anywaa put a relatively strong claim to ownership over the region while the migrant settlers, workers and shop keepers, who settled in Gambella since the 1980s, had a strong sense of ownership of Ethiopia as such. Clearly there collided very different interpretations of citizenship, ownership and belonging. In retrospect, it is not so surprising that neither group seemed to know or care much about the Komo, a small group with few people in the government and the administration, and a rather weak political agency. The Komo lived dispersed among the Anywaa, and for the highlanders they just did not seem to exist. I was able to make a lengthy interview with a Komo elder, on which, together with the review of the available sources and literature, I based my article in the EAe. 33 I was less lucky to obtain any information from, and very little about, the "Kwama" (the Gwama of this thesis). In fact, I was convinced I would be able to interview Gwama who I thought were living in the Bonga refugee camp, not too far from Gambella, when I conducted a series of interviews and group discussions in Bonga with people who referred to themselves as "Burun", who spoke in fact the same language as the "Opuo" in Gambella. In 2006, I had the opportunity to follow a group of Anywaa pastors of the local Mekane Yesus Church for several days to Wanke kebele, where a major settlement of the Opuo was to be found.
Only in 2009 I began to do research in the area south of Asosa, starting from Bambasi where I met Bambasi Mao who referred to their language as mawes aats' tose, 34 to Tongo, the capital of the Mao-Komo Meckelburg) and EAe IV, 'Opuo: Opuo, ethnography' (A. Meckelburg). 34 Cp. Michael Ahland, 'Aspects of Northern Mao Phonology', Linguistic Discovery 7, no. 1 (2009). themselves as Komo or Gwama and rejected the name sit shwala. Back to Gambella I realized that the Komo of Bonga spoke Gwama and understood themselves as black Oromo and felt no connection to the Komo, who they were lumped with. All these encounters, and many more snippets of inquiry into the history and the selfperception of the Mao and Komo, will feed the theme and topics I will present in the second chapter.
In Ethiopia today, much of the local and domestic socio-political discourse is about culture, identity and history. All over the country people are involved in defining their culture; on cultural festivals (Nations and Nationalities Day, etc.) they perform what defines their culture. Bahǝl (ባህል), the Amharic word used in the political discourse on culture, largely comprises primary features, such as dances, dresses, food habits, drinks, material culture, etc. Often it has an historical connotation to it, and speaks of cultural traditions that need to be preserved or revived.
Ethnicized culture has a great revival in Ethiopia. Mao and Komo politicians are involved in standardizing the culture of their groups. Musical instruments, cultural artefacts, dance, and foremost the languages, are developed, created and re-created to identify ethnic groups as such and it is not different with the Mao and Komo. There are attempts to revive the language and make it available for primary education also. It was an interesting time, I suppose, for entering the field because of this interest in identity and history. This was particularly motivating in an environment where ethnicity is traditionally of little importance; all Mao I met were usually multi-lingual with a mixed ancestry while they paid little attention to the politics of identity in present day Ethiopia. 35 The Mao are used to retaining their culture against the neglect of their majority neighbours and live amongst them as historical minorities, a fact that Alfredo Gonzaléz-Ruibal recently referred to as cultural "mimicry". 36 I was working closely with one person who was so important for the undertaking that I want to give his introduction some space here: Andinet Arega Woga is the son of an Oromo teacher and a Komo woman. He lived with his mother in Gambella, and grew up, socially a Komo, while he was also connected to the Oromo. Through the language and as a young modern urban man, he also integrated into the  This is of course a generalization and does not account for people who worked for, or were close to the government; people who, to a certain degree, were able to gain resources from fostering the discourse on a political identity. 36 I am referring to the chapter-title "Of Mimicry and Mao" in González-Ruibal, An Archaeology of Resistance,241. speaking urban milieu of Gambella. And by outsiders he was mostly defined as an Anywaa, who he also identifies with to some degree. A painful episode of identification, he once confirmed to me, was the 2003 massacre in Gambella, when he was lumped with Anywaa men and fled together with other youngsters to Pochalla.
Andinet was very eager to discover the history of the Komo with me and fascinated to meet many culturally related groups in Benishangul as well. Andinet, who worked as a teacher in the Komo kebele, took much interest in the cultural forms and features, instruments and materials we discovered during our travels. Sometimes, I have to say, Andinet was much more eager to learn from people, than I was, especially when my exhaustion from walks between villages was just overwhelming, while Andinet's energy levels seemed not to have suffered at all.
Between 2011 and 2014 Andinet and I travelled between Asosa, Tongo, Begi, Gidami, Gambella and Pokung and many adjacent areas, and often to outlying villages.
In all the years the actual time spent in the field has not exceeded 12 months. In this time, I did over 100 interviews with male elders, youngsters and elderly women. We conducted small censuses in Pokung and Gambella, and we visited 15 villages and kebeles for durations between one day and several weeks.
The study is set in a context of multi-method-approaches, and outlined as a project in "political ethnography". 37 I correlate methods of history and political science with those of anthropology. I contend, that a political science study should indeed "derive inspiration from anthropology's self-interrogation". 38 The study is based on the interpretation of sources and the collection of empirical, qualitative data from the field. The sources are scant, but some materials I gathered on the way are presented for the first time. Sources used in this thesis are historical letters from the regional administration to the government (some kept in the Asosa museum, others in the private archives of families). I was also able to obtain a relatively large body of correspondence between the regional administrations of Benishangul-Gumuz and Oromia. Furthermore, I did archival research in the British Foreign Record Office (in 37 Edward Schatz, ed., Political Ethnography: What Immersion Contributes to the Study of Power Kew in London) focusing mostly on materials related to slavery in the region.
Likewise, I made use of the Arkell papers stored in the SOAS library. 39 I also consulted the existing travel reports, of which Juan Maria Schuver's work was the most helpful. 40 But also Pellegrino Matteucci 41 ,Romolo Gessi 42 and to a certain degree also Enrico Cerulli's works 43 were of importance here. I will compare my own thoughts with the existing literature and briefly review other writings as I go along. A fascinating source of data are the field notes of Professor Alessandro Triulzi, who generously allowed me to make use of his detailed interview transcripts of the 1960s and 1970s. 44 Apart from the review of sources the study draws chiefly upon the methodology of oral history 45 and ethnographic observation. The bulk of the data comes from qualitative interviews, genealogical reconstruction and biographical analysis of the lives and personal history of elders, but also from 'stakeholders' like politicians and administrators. On the settlement level, network analysis was employed to study the economic and social interaction between groups. Additionally, basic linguistic analysis was used to unravel the blanket term 'Mao' and differentiate it into linguistic/ethnic sub-divisions.
The use of oral data is inevitable in an area where most history is preserved in the memory of the people. The Mao and Komo have been and are to a large extent an oral society. History is handed down orally through the generations. 46 Owing to the limited time spent in the field and the obvious language barrier, I was not able to record many oral traditions per se. The majority of data in this thesis is based on the 39 I am taking a slightly different look at these British materials but they have previously been used both by Bahru Zewde and Alessandro Triulzi (Bahru Zewde, 'Relations between Ethiopia and the Sudan on the Western Ethiopian Frontier 1898' (University of London, 1976.; Triulzi,Salt,Gold,and Legitimacy. 40 Wendy James, Douglas Johnson, and Gerd Baumann, eds., Juan Maria Schuver's Travels in North-East oral history of individuals. I make a distinction between oral history and oral tradition based on Jan Vansina, who stated that traditions are based knowledge that is handed down from generation to generation and oral history is based on what a person has seen or heard in his or her lifetime. 47 Despite its subjectivity and bias, oral data is crucial to understand the perceptions they carry for the narrator. 48 My approach was to use semi-structured interviews, in which I would regularly come back to historical questions pending from other preceding interviews, in order to build a web of information which helped me cross check several aspects. A recurring theme was to ask for lifelines: the biography, the place of birth, the parents, personal and group migrations, major events in the clan and the lineage, marriage, children etc. The interview partners were mostly approached because they were recommended by people for their specific historical knowledge, or random discussions developing out of group meetings. My approach was multi-sited. 49 Hence while this approach was crucial in getting a regional overview both on the settlement patterns as well as on the distribution of memories, it also came at the expense of depth and historical accuracy. An interview once done in a village I visited only for a couple of days could not necessarily be followed up the next time. Interesting to note was the respondents' relative disinterest in history. History of the longue durée is something that I found hard to dig out during my several short visits. While the majority groups (such as the Berta) seem to be more influential in their historical imagination, the Mao and Komo have a significantly ahistorical approach. This in itself is a sign of their constant denigration and marginalization. Despite this, I found people to be interested in talking about and sharing with me more general historical occurrences -especially migrations, or even the traumas of slavery -as well as clan distributions and cultural traits (hunting, agriculture, etc.). The accessibility of data raises important problems for the research.
The Mao and Komos' approach to history also reflects the groups' discontent with the state today. Hirsch and Steward indicated, [I]n short, history is a culturally constructed practice. Ethnohistorians have, however, spent more time collecting or piecing together substantive historical narratives and presenting them according to a Western conception of history than they have analyzing local historicity. 50 The conception of history itself poses significant questions to the position the Mao and Komo give themselves in society.
Using Lewellen's conception of the "field" and the "arena", my field is, "the area of political activity" that deals with constitutional politics, memory and identity. 51 The arena consists of pin-points in a wide regional system that ranges from Gambella through western Wallaga to Benishangul. Conflicts related to ethnic identity, ownership, citizenship rights, the re-modelling of established forms of power and the forging of a " Mao-Komoness importance for the regional history and his position in the identity discourse of the watawit descendants (today partly Berta political elites) concern the memories of all groups, the Berta as well as the descendants of slaves. Such border chiefs and rulers can be analysed in terms of Bailey's "middlemen" in a centre -periphery perspective. 53 Another such middleman was mek Kutu Gulja, 54 a lower landlord of the region and 50 Eric Hirsch and Charles Stewart, ' Introduction: Ethnographies of Historicity ' , History and Anthropology 16, no. 3 (2005) Grottanelli (Vinigi Grottanelli, I Mao. Missione etnografica nel Uollega occidentale, Pubblicazioni del Centro studi per l'Africa orientale italiana della Reale Accademia d' Italia 5 (Roma: Reale accademia d'Italia, 1940); s. also Abdussamad H. Ahmad, 'Trading in Slaves in "king of the Mao", 55 who is very important for the invention of a founding figure of some segments of the Mao and Komo today. 56 His grandson Abba Harun is today known as the "father of the Mao-Komo special wäräda": Harun was the lower landlord of the area and administrator of the Mao and Komo, before he had to flee when the socialist government started to imprison the feudal landlords in the 1970s. As an ally of the monarchist Ethiopian Democratic Union (EDU), seeking refuge in Sudan, he raided Begi with a small army of Mao and Komo. By the fame gained through this assault, and with the fall of the Därg government, he could place himself and the "cause" of the Mao and Komo in the political bargaining process of the emerging multi-ethnic region of Benishangul-Gumuz and advocate for an ethnic territory and autonomy of the Mao and Komo. I am interested in the middle-men between centre and periphery who became regional elites and whose history often relates to trajectories of power, and social stratification within groups. Descendants of such elites are often still found in political posts in the regional arena.

What others have said
In the secondary ethnographic literature the Mao and Komo exist largely in footnotes.
The Koma (Gwama and Komo) have been subject of some comparative work of Wendy James on Uduk. 57 She also treated the Mao and Komo in a seminal paper, which unfortunately was never officially published. 58 In this paper she has already outlined the general idea of this thesis. She eloquently dealt with the making of Bela-Shangul and Gumuz, Ethiopia: Border Enclaves in History, 1897-1938', The Journal of African History 40, no. 3 (1999: . 55 Abdussamad H. Ahmad, 'Trading in Slaves in Ethiopia'. 56 Center -Cambridge University Press, 1980), 37-67. frontier societies and looked at the Mao and Komo as proponents of two similar but parallel frontier processes.
The first longer treatment of the Koma in Sudan was published by the colonial administrator Frank D. Corfield. 59 But the Koma of Sudan (mostly Gwama and Komo) have been the subject of a more detailed and fascinating study of Joachim Theis. 60 Theis' seminal treatment of the Koma is based on prolonged fieldwork and starts with the idea of the self-reconfiguration of the Koma as an ethnic group. Theis opens his book with a long narrative of the Koma's recent past, their society torn to pieces by slave raids and later the civil war in Sudan. He portrays the life world of the Sudanese Koma, their systems of kinship and marriage, their cultural ecology, etc. Of paramount importance for my own work is not only his historical approach, resembling in many ways the findings of my own research and hence supporting the general aspects of the history of slavery; more important in that regard is the portrayal of the systems of marriage, based on sister-exchange practiced in the Sudan, but lost in the areas of own research. Theis's description of the systems of kinship builds a major source of comparison for my work.
At the beginning of my research I largely felt overburdened by the perceived need to accumulate data portraying two (and more) groups, their past and anthropological present. I had already relied heavily on the works of Alfredo González-Ruibal, but his book came to my rescue. 61 As a state-of-the-art approach, the book has become the main source of comparison for my work, and a point of departure for the thesis presented here. His book and my thesis share similar entry points but most importantly I have found it a reliable source of reference, and a medium to refer the reader to when it comes to the ethnographic past and the material culture. I would not have been able to show these so sensibly as the archaeological anthropology of Alfredo  April 11, 1975, ed. Robert K. Herbert, Working Papers in Linguistics 19 (Columbus, Ohio, 1975,  as well as Harold Fleming, 'The Importance of Mao in Ethiopian History', in Proceedings of the Seventh International Conference of Ethiopian Studies: University of Lund, 26-29 April 1982, ed. Sven Rubenson (Uppsala et al.: Scandinavian Institute of African Studies et al., 1984, 31-38. 65 Bahru Zewde, 'Relations between Ethiopia and the Sudan on the Western Ethiopian Frontier 1898-appendix of original documents attesting for the relations between Asosa and Addis Ababa in the early 20 th century. 69 How this thesis is structured After the introduction, Chapter 1 will give a short overview on the theoretical perspectives of this study. Based on the idea of the African frontier I will introduce the history of the Mao and Komo as a continued frontier process influenced by overlapping and conflicting patterns of state-making, territorialization and ethnization of politics and identity. To capture this, Chapter 1 will also provide the definitions of ethnicity and citizenship, etc. applied. The main theoretical approach is Kopytoff's African frontier thesis: looking at the production of frontiersmen, the movement of groups seems to lend a focal point of entry for the complex social processes that the Mao and Komo underwent. Examples will be provided by the encroachment of the Oromo frontiersmen, the Berta frontiersmen as well as the state frontier, extending into the heartland of the Mao and Komo, which will be analysed in more detail in the following chapters.
Chapter 2 provides a brief overview on the ethnography of the Mao and Komo.
Chapter 3 opens the chronological presentation of the history of the Mao and Komo and inter-ethnic encounters in the region. It offers a short overview of the region before the incorporation of the territory into the Ethiopian state. Here Chapter 4 connects to and presents the making of a periphery in a state-making perspective, focusing on the subject status of the Mao and Komo in the regional integration under the pretext of tributes and taxations.
Chapter 5 gives a more detailed background on the history of slavery as a motive in the centre-periphery relations. Chapter 6 and 7 return to the processes of state formation, modernization and the political changes in the periphery between 1941 and 1974. Chapter 7 also discusses the Civil War and the cross-border insurgencies in the Sudan as a formative as well as traumatic chapter in the history of the Mao and Komo.
Chapter 8 connects the presentation of the historical events to the present political system. First it discusses the practices of memory as a formative act of the inter-ethnic 69 Yirga Tesemma, 'The Process of Political Integration in Asosa-Beni Šangul/Gumuz: A Case Study' (B.A. thesis, Haile Selassie I University, 1973). relations in the post-1991 regional setting. This presentation eventually leads to the appraisal about the political present of the Mao and Komo. Notwithstanding the frontier-history, the post-1991 government has opted for border demarcations to tame the western borderlands. Thus, towards the end of the study I will discuss some structural dilemmas that emerge from these modern federal arrangements in a former frontier zone: territorial arrangements in the federal structure as well as the emphasis on ethnicity create questions for citizenship and identity in terms of regional ownership and belonging; mobility has always been a main feature of the political economy of the region and at the same time elites have always fought for territorialisation and border control.
Notwithstanding the creativity by which the Mao and Komo react to the contestation of belonging and identity, the current trend to formalize territorial control in killil and wäräda and kebeles is the main obstacle in their quest for identity.

Chapter 1
Theoretical Assumptions -The Western Ethiopian Frontier,

Identities and Citizenship
Studying citizenship in rural Africa requires understanding the historical relation between the people and the state. The patterns of state formation will ultimately define the perceptions of belonging of a people. 1 Citizenship, as I will outline further below, for the purpose of this thesis is more than the definition of national belonging in citizenship laws. While such laws are important as "boundary mechanisms", 2 I am more concerned with the emotive aspects of citizenship. Defining the historical formation of local, regional or national society as a term of reference is a precondition for inquiry.
Of interest are the broadcasting of state-power, the local middlemen, the composite of the society and its internal stratification, majority-minority relations and the like. The Mao and Komo have been at the margins of several state formation processes, both Ethiopian and Sudanese. The area of Western Ethiopia has thus been described as a "double periphery". 3 3 Triulzi,Salt,Gold,and Legitimacy,181. the Ethiopian empire. They also lived on the fringes of the watawit sheikdoms as well as of Oromo kingdoms (i.e., the Leeqa kingdom under Jote Tullu). Their historical experience informs the position the Mao and Komo are currently holding in the multiethnic Ethiopian society. These complex social processes will be described in more detail in a later chapter (cp. Chapter 3). First, the theoretical approach to state formation and the inter-ethnic encounter will be briefly outlined in this section.
Seen from the Ethiopian side, the peripheral position of western Ethiopia is rooted in the logic of Ethiopian state formation and the incorporation of the western fringes as a periphery. 4 In the wake of this forceful process of state expansion, the centre coopted local elites or destroyed and substituted local rule. Both processes led to the establishment of regional political rule, the penetration of the Ethiopian state apparatus as well as the diffusion of state culture. But the Ethiopian state did not permeate into uninhabited areas. On the contrary, these areas had a long history of local political formation and inter-ethnic relations. 5 Before the Ethiopian state expansion, local rulers had for long been in competition over the stateless (acephalous, or less hierarchically organized) people surrounding them, as well as over the resources of the region. The encroachment of the predatory state systems aggravated the inter-ethnic relations and power equilibrium between the people. In brief, this meant: regional polities like the Oromo mootidoms or the watawit sheikhdoms had for long competed with each other over control, before they were drawn into a tributary system with the central state. These tributes were in fact handed down to the lower groups of the regional social stratum. 6 These fragile groups were in part reduced to slaves, serfs, tenants, porters, and hunters. Hence the regional society came to be divided into local elites and subordinate people. These situations lead to the lumping of the lower stratum of the society under blanket terms and set in motion a process of "overwriting identities": In 'the old days' various terms were used by central elites of the old states which blanketed fringe communities together under more or less exclusionary names. The most glaring was the Ethiopian shangalla meaning something like the American term 'nigger' (banned in the reform of 1974), and in the case of 4 Donald Donham and Wendy James, eds., The Southern Marches of Imperial Ethiopia: Essays in History and Social Anthropology (Cambridge: Cambridge University Press, 1986). 5 Of course the pre-conquest history of the areas was markedly different in different areas. 6 Abdussamad H. Ahmad, 'Trading in Slaves in Bela-Shangul and Gumuz, Ethiopia'. the Sudan it was (and still is) common to hear the use of 'abid (slave peoples) or zurga (blacks) or regionally inclusive blanket terms such as 'Nuba'. In the case of the Blue Nile, there is an old term, 'Hameg' or Hamaj, applied to people living on the southern fringes of the former kingdom of Sinnar, and another, 'Burun', used of a range of peoples in the foothills and valleys west of the Ethiopian escarpment. Within these blanketed socio-political categories were dozens, if not hundreds, of local indigenous language communities with a strong sense of self as against a variety of neighbours, and increasingly so in concert against the state and its elites. 7 A remnant of this process is the usage of the term Mao for the subaltern people. Today it is not surprising to meet a Mao explain his or her identity/ethnicity as Mao-Komo and the language he or she speaks as Afaan Mao-Komo (afaan being the Oromo word for mouth/tongue/language). The Mao and Komo have adopted the governmental discourse and accept being blanketed together. This concurs with observations made by González-Ruibal: The Mao adopt a humble, silent position in front of the Oromo. In places like Arabi or Egogirmos, the Mao do not even preserve their ethnic selfdenomination: they have fully adopted the term of their dominators and call themselves simply 'Mao'. Even when they maintain their own names, these have often been transformed by the Oromo. 8 The Mao and Komo pose substantial questions to the mainstream ideas of ethnicity, identity, group or social formation and structure. 9 To take an evolutionist stand to their marginality cannot do justice to their experiences. They are dispersed mainly because of the course of history. But, to be sure, they have not been helpless victims of government and the local elites. Any representation of the Mao and Komo would be painfully incomplete if it was to ignore the workings of the frontier, the intertwined and overlapping process of the reproduction of the early peripheral societies till the coming of the tidal federal frontier of today.

The African Frontier Thesis
The study of the formation of African societies has greatly benefited from Igor Kopytoff's rethinking of Turner's American frontier, which built the base of his seminal theoretical approach on the African frontier. 10 Both ethnic and social formations in Ethiopia in general and western Ethiopia in particular, have often been, implicitly and explicitly, described as a frontier: a zone of constant social and cultural transformation. 11 The migrations of people both from the Sudan and from within Ethiopia have led to significant amalgamation of cultures, customs and eventually identities on this Ethiopian-Sudanese frontier. The immigration of Sudanese Arabs, also described in great detail by Alessandro Triulzi, 12 the migration of the Oromo, 13 the immigration of the Busase from Anfillo 14 are all interlocking frontier processes, 10 Hultin, International Development Studies 11 (Roskilde: Roskilde University, 1994), 235-45. 12 Triulzi,Salt,Gold,and Legitimacy. 13  which shaped the regional socio-cultural environment at different points in time. These processes have created memories and narratives, often overlapping and simultaneously challenging the respective perceptions of the past. 15 Considering the moving forces both behind the historical regional formation as well as exploiting this knowledge to predict future processes of identity formation is the merit of Kopytoff's theoretical approach.
The Mao and Komo were affected in two significant ways by this frontier: they were incorporated and became part of the new social formations brought about by the frontier processes, and they were expelled and reduced to refugees and migrants, hiding from the emerging social constructs, and forming new social structures elsewhere. The frontier produces winners and losers. 16 Hence, we might analyse the Mao and Komo as frontiersmen in a double sense: they are in part a remnant society incorporated at the lower end of the new social hierarchy of an emerging regional society. And they are deep rurals, refugees and hideaways, moving to new spaces unoccupied by the majority society. A Gwama elder once made this perception of the historical duality of the Mao- Originating from the work of Fredrick Turner, the frontier thesis has been quite influential both in the description of social and state formation as well as inter-ethnic relations. Turner believed that American institutions were not imported by Europeans and thus a copy of the European institutions but that the process of migration influenced the political culture and the emergence of the new American frontier state: Little by little he transforms the wilderness, but the outcome is not the old Europe, not simply the development of Germanic germs, any more than the 15  first phenomenon was a case of reversion to the Germanic mark. The fact is, that here is a new product that is American. At first, the frontier was the Atlantic coast. It was the frontier of Europe in a very real sense. Moving westward, the frontier became more and more American. As successive terminal moraines result from successive glaciations, so each frontier leaves its traces behind it, and when it becomes a settled area the region still partakes of the frontier characteristics. 18 While Turner "sees the frontier as a natural force of transformation", Kopytoff understands the frontier as "force for cultural-historical continuity and conservatism". For Kopytoff the African frontier is a "local frontier, lying at the fringes of numerous established African societies". 19 Instead of creating a new society, the frontier "provides […] an institutional vacuum for the unfolding social processes".
Accordingly, "a crucial factor in the outcome of the frontier process is the nature of the initial model carried from the metropolitan culture to the frontier". If the communication between metropole and frontier is kept alive, "the frontier may consequently act as a culturally and ideologically conservative force". 20 Donham summarized the "essence of the frontier" accordingly: […] once outsiders have defined an area as a frontier and have intruded into it in order to settle in it, there begins a process of social construction that, if successful, brings into being a new society. The central thesis of this analysis is that most African societies arose out of such a conjuncture of events. And, further, that this process of building new societies, paralleled by the demise of established societies, has been a continuous one in African history. 21 The ideological construction of an institutional vacuum, i.e. the imagination and selfperception of the late-comer as first-comers all my interviews concerning migrations, the respondents initially described the destination of their respective migration as 'uninhabited' by any people, only to admit that there were previous inhabitants, but their status was usually belittled as 'wild', 'backward' and 'uncivilized'. 22 For Kopytoff the frontier is an area of uncertain political rule, in which groups migrate and settle to fill a perceived or real institutional vacuum. Western Ethiopia is defined by interlocking frontier-processes: it provides for "interstitial spaces of different kingdoms and princely polities", 23 into which immigrated various actors to define new spaces. But following the forceful proliferation of state culture and political ideas, it also provides for a "tidal frontier", 24 with the emergence of the Ethiopian state, the imperial (ca. , the socialist (1974)(1975)(1976)(1977)(1978)(1979)(1980)(1981)(1982)(1983)(1984)(1985)(1986)(1987)(1988)(1989)(1990)(1991) and the federal frontier (1991 till today). Today this process is the permeation of the 'developmental state' on the margins, the "centering of the periphery". 25

Deep Rurals
One way of approaching the Mao and Komo is dominated by the idea of the "deep rurals". 26 The 'hills' (i.e. mountain ranges) play a crucial role in the memory of different groups and form a path to a collective identification: some Mao say they come from (Mount) Kiring, the Komo say they originate from Gemi 27 and the Gwama originate in Banga according to their oral traditions. 28 These hills, though, have not been strongholds but places from which groups have been ejected by expanding neighbouring groups. The idea of the deep rurals was introduced to the Sudan- 22 Interviews in Benishangul 2010-14. 23  Institutions of a People of the Sudan-Ethiopia Borderland, Studies of Religion in Africa 13 (Leiden: Brill, 1995); Theis,Nach Der Razzia. 27 The Komo refer to Tullu/Gara Gemi as gewa, cp. also Schuver's Yowé (Yewa) which refers to both the mountain massif and the settlement (James,Johnson,and Baumann,102). 28  Deep rurals are those groups that sought to avoid "subordination by, and cultural assimilation into, the neighbouring, more pervasive culture". 31 These are the cultures that the frontiers process has made peripheral to the emerging groups. Thus, one way to perceive and understand ethnic borders in the poly-ethnic setting of the frontier is the focus on "flight" due to the fear of assimilation and enslavement. 32 This is deeply enshrined in the memory of the fragile minorities on the Ethiopian-Sudanese border. A research note from Joachim Theis is interesting in this regard: The Koma 33 feel surrounded by three powerful nations: "Sudan," to the north; "Janub" (southern Sudan) to the south -which the Koma consider an independent state (I was once asked "Who is the president of the Janub?")and "Shoa" (Ethiopia) to the east. The Koma who live in all three "states" do not feel part of any of them. They fear all three powers and try to avoid 29 Wendy James has given a substantial overview on Jędrej translation of the deep rurals idea into the Nile Valley and its scientific migration from West Africa (Richard Fanthorpe, 'Limba "Deep Rural" Strategies', The Journal of African History 39, no. 1 (1998): ;; Wendy James, 'Charles Jędrej and the "Deep Rurals": A West African Model Moves to the Sudan, Ethiopia, and beyond ', Critical African Studies 8, no. 1 (2015): [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15]. The deep rural idea is also the leitmotive behind González-Ruibal's seminal regional overview (González-Ruibal, An Archaeology of Resistance). 30 [49][50][51][52][53][54][55][56][57][58][59][60][61][62][63][64][65]. 33 Theis uses the word Koma in the Sudanese usage for Gwama and Komo speakers. contact with them whenever possible. They keep clear of the motortracks and tend to settle their disputes among themselves to avoid the involvement of the "hakuma" (government), because "the talk of the hakuma is bad". 34 While the deep rural-paradigm is one useful way to investigate, it neglects the Mao and Komo which did not live in separation of the state, especially on the Ethiopian side of the border. Large parts of the Mao and Komo were integrated, if in a subaltern position, in one way or another into regional systems or the wider Ethiopian polity.
Some were forced to do so, other chose to do so. Thus, this study is about the state in the periphery and less about the "art of not being governed"; 35 it deals with the persistence and maintenance of ethnic boundaries of fragile minority groups in a polyethnic environment.
Today, the Mao and Komo are generally referred to as indigenous groups (this is constitutionally inscribed at least in Benishangul and Gambella). In the western-liberal approach, ethnic groups, once their status is defined, have to be 'accommodated' in the diversity of the multi-cultural state. Accordingly, their minority rights have to be "respected" and constitutional provisions are usually taken to ensure this. But only if we understand how minorities are created, re-created and also understand their own agency, we can identify in how far citizen-rights are negotiated in the complexities of multi-cultural arenas. 36 Understanding the status of an ethnic minority should start with understanding their ethno-genesis.
Ethnicity on the Frontier Social identities are subject to constant redefinition by their bearers and others. Groups can change their composition, or their status, or their name, or their affiliation, or even all these features. 37 34 Joachim Theis, 'Ethnic Identity on the Sudan-Ethiopian Border', Cultural Survival Quaterly Nation, Tribe and Ethnic Groups in Africa (1985). 35  Ethnic affiliation has become an important guiding principle in day-to-day domestic politics in Ethiopia. Since the overthrow of the Därg the current government has experimented with a federal arrangement, emphasizing ethnic identity as the core element for political decentralization. In effect, this meant the creation of linguistically or ethnically defined federal states. The government of Ethiopia promotes a rather primordialist approach to ethnicity and divides the ethnic groups inhabiting Ethiopia into "Nations, Nationalities and Peoples" following largely an essentialist concept: A Nation, Nationality or People for the purpose of this constitution, is a group of people who have or share a large measure of common culture or similar customs, mutual intelligibility of language, belief in a common or related identities, a common psychological make-up, and who inhabit an identifiable, predominately contagious territory. 38 To study politics in Ethiopia today is to deal with the "accommodation of diversity", hence to look at a multi-cultural society drawn together by history, and exposed to forces triggered by the daily politics of ethnicity (i.e. ethnic federalism).
"Accommodation of diversity" has become the main justification of the post-1991's regime's policy of ethnic federalism, the answer offered as a solution to the "national question". 39 Since Barth's approach of the ethnic boundary, functional concepts to ethnicity have gained prominence. 40 Ethnicity exists vis-à-vis the "ethnic other" and is maintained through patterns of self-identification, language, myths of descent, and other forms of in-group identification. I contend here that it is especially necessary to analyse the rationale behind ethnic self-identification, that is to ask where, when and why it is felt necessary to identify for one group against another. Thus, ethnicity has to be understood in a regional, political and economic setting within existing grouprelations. 41 The ratio behind ethnicity should be analyzed in the historical regional setting: 38  [I]ts meaning and practical salience varies for different social groupings according to their position in the social order", and it has its "origins in the asymmetrical incorporation of structurally dissimilar groupings into a single political economy. 42 Ethnicity and the frontier are here seen as two partly overlapping phenomena. The process potentially affects ethnic-affiliation along its courses. This is especially so when states, hierarchical polities or chieftaincies emerge within the frontier. I argue that changing inter-ethnic relations are visible in the boundary mechanisms between groups and thus are better analysed with "constructivist" approaches than on primordial bases. 43 Abbink has argued for an approach to ethnicity that is based on considerations of "political ecology": The assumption here is that groups based on, or acting on the basis of, some ethnic or 'tribal' identity must be seen as located in a wider environment of competing groups of different composition, especially in areas where state influence is relatively weak. The environment is to be seen as composite of ecological-economic conditions and of political conditions in a wider sense: power balances between groups determine the degree of success of groups in maintaining solidarity and achieving results. 44 This approach is useful in the given case because it looks at external forces that forge identity. Ethnicity in itself is not stable and is changing through time; it is prone to socio-environmental conditions both in a synchronic and diachronic perspective. With reference to historical interaction and "ecological-economic and political conditions" (s. above) we are drawn to look at the relations with neighbouring groups like the Oromo and Berta in the given case. Their impact, based on the economic perquisites of any given time, has forged the identity of the Komo and Mao as subordinate groups, either enslaved or administered, or assimilated into emerging structures.
Comaroff's approach sustains this claim because it emphasizes the structural preconditions under which to analyze ethnicity. Ethnicity "has its genesis in specific 42 John L. Comaroff, 'Of Totemism and Ethnicity: Consciousness, Practice and the Signs of Inequality', Ethnos 52, no. 3-4 (1987): . 43 Cp. Barth, Ethnic Groups and Boundaries. 44 Abbink, 'The Deconstruction of "Tribe"', 1-2. historical forces, forces which are simultaneously structural and cultural", 45 it is not unitary and describes both a set of relations and a mode of consciousness. Its "meaning and practical salience varies for different social groupings according to their position in the social order" and it has its "origins in the asymmetrical incorporation of structurally dissimilar groupings into a single political economy". 46 To study political conditions as a force that produces group identity enables us to understand how the ongoing re-orientation of identity in Ethiopia relates to the reinvention of the identity of the Mao and Komo. The salience of history will enable us to understand the changing patterns of self-identification of the Mao and Komo which is instrumental for this study.

Citizenship on the Frontier
At the turn of the 21 st century, almost all people belong to a state. 47 Citizenship though is by no means uniform. Its subject differs from place to place and so do its manifestations. 48 Is the overemphasis on a citizenry merely a western imagination of the world? 49 Citizenship is mostly an analytical tool in western political sociology and political science. It is, however, increasingly used to discuss African cases of national integration also. 50 As states largely shape local communities, anthropologists have had their fair share in analyzing the interaction between states and local communities: 45 Comaroff,'Of Totemism and Ethnicity',302. 46 Ibid., 307. 47 Studies 11, no. 1 (2007): 73-82. 50 Samantha Balaton-Chrimes, Ethnicity, Democracy and Citizenship in Africa: Political Marginalisation of Kenya's Nubians, Contemporary African Politics (London-New York: Routledge, 2016); Jean Comaroff and John Comaroff, 'Reflections on Liberalism, Policulturalism, and ID-Ology: Citizenship and Difference in South Africa', Social Identities 9, no. 4 (2003): 445-73;Sara Dorman, Daniel Hammett, and Paul Nugent, Making Nations, Creating Strangers: States and Citizenship in Africa (Leiden-Boston: Brill, 2007); Edmond J. Keller, Identity, Citizenship, and Political Conflict in Africa (Bloomington, IN: Indiana University Press, 2014); Smith, Making Citizens in Africa to name but a few. [...] the new interest in the state arises from a recognition of the central role that states play in shaping "local communities" that have historically constituted the objects of anthropological inquiry; in part, it reflects a new determination to bring an ethnographic gaze to bear on the cultural practices of states themselves. An important theme running through the new literature has been that states are not simply functional bureaucratic apparatuses, but powerful sites of symbolic and cultural production that are themselves always culturally represented and understood in particular ways. 51 Through polices the state affects the relation between ethnic groups; it creates and denies opportunities, blocks or provides access to power and hence willingly or unwillingly affects the internal social equilibrium within groups as well.
Citizenship connects the person to the institutional framework. It is a tool to analyse the interaction of state and people. Citizenship is about belonging to a national community, to feel to belong to it and to gain certain rights and responsibilities from it. Citizenship in traditional approaches was thought in a territorial perspective (the national territory), "therefore", stated Herbst "citizenship laws are critical to examine because these regulations explicitly tie populations to unique, territorially defined polities". These rules thus determine "who's in and who's out". 52 Is citizenship a logical product of post-colonial state-making in Africa? 53 Is it a concept that will enable us to understand the level of integration of peoples into their respective state? Based on a western notion of civil rights and civic education, etc. the concept of citizenship looms large in regards to inclusion, exclusion and belonging. It touches on identity, ethnicity, group-as well as individual rights. What exactly defines the citizen of Ethiopia? According to the constitution of Ethiopia citizenship is defined by birth. Anyone being born in Ethiopia to one Ethiopian parent is an Ethiopian citizen. On the next level it is about belonging to one of the nations, nationalities and peoples that constitute the Ethiopian society. Hence one can be Amhara, Oromo, Tigrean etc. or of shared origin and thus be an Ethiopian citizen.
One should be registered in a kebele and in technical terms it is about having an ID card or even a passport which defines one as a citizen of Ethiopia. It is about paying 51 James Ferguson and Akhil Gupta, ' Spatializing States: Toward an Ethnography of Neoliberal Governmentality', American Ethnologist 29, no. 4 (2002): 981. 52 Herbst,States and Power in Africa,231. 53 Dorman,Hammett,and Nugent,Making Nations,Creating Strangers. taxes and enjoying voting rights. Inherently tied to the imagination of modern democratic states, such formalities of citizenship are not necessarily the overall realities in Ethiopia. The "rights bearing citizen" 54 has to be qualified on the basis of ethnographic facts of lived experiences. Especially at the margins of the state (as opposed to an urban milieu) citizenship in Ethiopia is largely still in the making.
Interestingly though it is exactly this pattern that elevates it to an analytical tool: Precisely by thinking about the shortcomings of citizenship the level of national integration of minorities in the Ethiopian case can be analyzed. To define that gap between citizen and citizenship, is to understand the position of the person as a political being.
One important discrepancy in the often western based quest for analyzing citizens in Africa, is posed by the historical genesis of group identity. Collective group identities are changing, expanding, adapting to various social, and political changes and stimuli. While the western citizen is a collective being vis-à-vis the state, in African contexts the collective being is often bound to many different contexts of which the state might not be the primary pattern of identification. 55 It should hence be interesting to understand how groups are formed, what constitutes them, how do members of these groups align themselves with the state and its institutions. In western Ethiopia, some members of certain groups may be entirely educated in major towns and live in full enactment of their political possibilities, while members of the same group might not entertain the idea that the Ethiopian state is anything they belong to. 56 Such situation asks for an institutional understanding of the link of ethnicity, identity and citizenship. But how are citizenship and belonging experienced by historically marginalized groups?
Citizenship has always entailed aspirations to equality; in fact, citizenship was defined by equality of status, and the integration of social groups into this status meant their formal equality before the law and before the state. Of 54 Smith, Making Citizens in Africa. 55 91-109. course, such formal equality was seldom realised; moreover, most states differentiated citizenship in the allocation of rights and resources. 57 The question, who is a citizen and its manifestations also raise questions on how people come to imagine themselves as legal persons and what it means to them. 58 How is citizenship negotiated? Are the points of reference the state, the region, the group? Does ethnic federalism actually capture the original identity of a group, or is it modelling identity itself? 59 Comaroff and Comaroff noted for the South African case: The generic citizen of postcolonial South Africa may be the rights-bearing individual inscribed in the new Constitution; also, the rights-bearing individual -typically urban, cosmopolitan -presumed in much mass media discourse. By contrast, ethno-polities and traditional leadership [supra note] speak the language of subjects and collective being [supra note]. For most South Africans, it is the coexistence of the two tropes, the citizen and subject that configures the practical terms of national belonging. 60 Citizenship is a helpful category to understand how rights are being defined and distributed and how the internal workings of national affiliation, belonging and autochthony are structured. But belonging or integration are not necessarily a sociopolitical ideal. Not of everyone at least. Integration can be "the battle cry of proclaimed majorities against minorities". 61 The term "integration" evokes the idea of a "defined whole" 62 in which minorities have to integrate, or a periphery that should be integrated in the pre-supposed state. In such perspective, citizenship is an analytical tool for understanding the processes and patterns behind political integration.
Searching for the historical foundations of citizenship or its cultural manifestations is a difficult endeavour though. We have to understand citizenship as more than "the rights to have rights" or national belonging. I am interested in a perspective of cultural citizenship as a form of "subjectification," that is according to Aiwah Ong, "being 57  made by power relations that produce consent through schemes of surveillance, discipline, control, and administration." 63 In order to gain a better understanding of the patterns of subjectification during consecutive governments I will briefly outline the history of citizenship in modern Ethiopia.

A Very Brief History of Citizenship in Ethiopia
The question "who is a citizen of Ethiopia" has to be answered differently depending on the time one is interested in. Before Ethiopia emerged in its present borders, i.e. before the national territory was defined by the interplay of internal expansion and extern (colonial) encroachment, the borders between different Ethiopian kingdoms, polities and other political entities were shifting. Wars of expansion and raids for goods and people loosely defined the shifting boundaries of these political entities, however stable they might appear. Land and the control over it defined the status of the people as well as their relation to the higher, or ruling, administrative body. In the traditional Christian Ethiopian highlands the society was hierarchically structured with the nobility ruling over the peasantry, which was owing it labour, giving parts of their agricultural surplus, and provided services in time of warfare and conflict. 64 At this point western Ethiopia was similarly divided into different polities and chieftaincies. Benishangul was distributed among sheikhs who had co-opted areas that were formerly ruled by meks (of Funj origin), who were themselves tributaries to the kingdom of Sinnar. Like in Christian highland Ethiopia, the labour on the land played a significant role in connecting the people to the wider political structure. In the Oromo region, small kingdoms formed which built on a clear division of nobility that ruled over a peasantry. In between lived people who, building on their own social structures and on less hierarchical systems, got partly co-opted into these emerging political structures or were drawn into subordinate positions. The subjection of Benishangul, and the Leeqa kingdoms, as well as attempts to control the areas of the Baro and Sobat basin, culminated in the colonial project of border demarcation and led 63 Aihwa Ong, 'Cultural Citizenship as Subject-Making: Immigrants Negotiate Racial and Cultural Boundaries in the United States ', Current Anthropology 37, no. 5 (1996): 737. 64 Habtamu Mengistie Tegegne, 'Land Tenure and Agrarian Social Structure in Ethiopia, 1636-1900' (PhD Dissertation, University of Illinois, 2011); Teshale Tibebu, The Making of Modern Ethiopia: 1896-1974(Lawrenceville, NJ-Asmara: The Red Sea Press, 1995. Ethiopia to fix its borders and define a territory and the people she would rule. Sitting at the intersection of millennia of predatory state encroachment the people of western Ethiopia in 1902 could hardly be referred to as citizens of Ethiopia. Nonetheless it is exactly here where the "making of citizens" begins. The presence of the state influenced, at times hardened, existing inter-ethnic relations and enabled certain individuals and groups to climb a "hierarchical" ladder. These individuals became affiliated to the "mainstream" society (i.e. the ruling elite of the centre).
From an Ethiopian state perspective, the people beyond these elites were shankilla or barya, and had no position in the society except for providing resources and labour.
The empire consecutively led people into its service. Soldiers were recruited; slaves and servants were sent from one household to the next and lived in different locations.
Administrators and service men appeared in areas beyond those of their northernhighland origin. It is, nonetheless, not before the Nationality Act of 1930 and with the introduction of the first constitution of 1931 that a serious attempt to define the citizens of Ethiopia as constitutive people was made. Nonetheless citizens were broadly defined as subjects given the monarchic character of the state. Article 1 of the constitution of 1931 reads: The territory of Ethiopia, in its entirety, is, from one end to the other, subject to the government of His Majesty the Emperor. All the natives of Ethiopia, subjects of the empire, form together the Ethiopian Empire. 65 The Nationality Law of 1930 confirms that the subjects of the empire are "any person born in Ethiopia or abroad, whose father or mother is Ethiopian, is an Ethiopian subject". 66 The colonial powers sharing borders with Ethiopia were particularly interested in the question who the Ethiopian citizens were, especially in regard to colonial subjects who would cross the borders of the respective territories for grazing and water. The Nationality Act emphasizes that "Ethiopian subjects having acquired a foreign nationality may always obtain the benefit of Ethiopian nationality when they return to reside in the country and apply to the Imperial Government for re-admission". 68 For a foreigner to become Ethiopian citizen, the Act states among other things, the need to be able to speak and write Amharic fluently. 69 The 1955 constitution did not bring many changes in regard to citizenship rights. 70 The subjects of the empire were rather blanketed in broader terms. A survey of the land tenure in Wallaga province (then combining also the areas of concern here including Benishangul) read that major languages spoken throughout the Governorate General were "Galligna", Arabic spoken by the Shogeles 71 in Asosa and "a dialect spoken by the Shankilas living in the Didessa". 72 Cultural sensitivity was of no concern during the era, which was especially characterized by an Amharization project (fostering Amharic as the national language and Orthodox Christianity as a state religion, Amharization was part of nationbuilding project that envisaged a unified national citizenry within a national territory).
It was exactly this omission of any rights of self-determination and Amharization that brought about regional cultural conflicts. Especially the student movement was concerned with the idea of self-determination of people and the discontent of the people was famously phrased by the student Wallelign Mekonnen who, proclaimed  119-30. To be a "genuine Ethiopian" one has to speak Amharic, to listen to Amharic music, to accept the Amhara-Tigre religion, Orthodox Christianity and to wear the Amhara-Tigre Shamma in international conferences. In some cases to be an "Ethiopian", you will even have to change your name. In short to be an Ethiopian, you will have to wear an Amhara mask (to use Fanon's expression). Start asserting your national identity and you are automatically a tribalist, that is if you are not blessed to be born an Amhara. According to the constitution you will need Amharic to go to school, to get a job, to read books (however few) and even to listen to the news on Radio "Ethiopia" The federal arrangement as of 1995 has built on the idea of nationalities and enforced major steps towards the recognition of the different ethnic groups of Ethiopia, plus the accommodation of diversity. 76 It thus proposes an Ethiopia specific form of cultural citizenship, which is largely territorially defined. Federal regions are theoretically enabled to decide on all sorts of political issues through the regional parliaments and administrative bodies. A salient feature of this is the right to choose the working language of the region and the regional language of instruction. This process has not been without conflicts either which have to be very carefully analysed on a case to case basis.
Looking at this brief overview on citizenship in Ethiopia, we may now make some assumption on the development of citizenship in Ethiopia. Between 1931 and1994 74  several amendments in the understanding of citizenship can be seen. While the citizens of Ethiopia were once the subjects of the crown, certain cultural rights, like the right to language and the entitlement to ethnic history have been enshrined since 1991.
Beyond the formalization of these rights in the constitution, the question of active Towards an Historical Approach to Citizenship Citizenship, as a relationship between the people and a state is a wide, ambitious, and multi-faceted field of inquiry. Citizenship links the subjects of politics to the polity.
As such it speaks about collectives bound by territorial arrangements: Citizenship involves collective, not individual mediations with the respective polity. 78 It is often thought of as a set of rights, duties and performance. Rights, duties and the ability to perform citizenship, refer to the legal relationship between citizens and state: civil rights, or political rights, or social rights, all of which have been changing globally in the past decades. Duties entail e.g. taxation or, in some cases, conscription, etc. questions of "belonging" which has been immensely conflictual in recent years in African states. 80 In a region largely defined by historical marginalization, powerlessness and exploitation, the divergent perceptions of history are equally important to understand the impasses of citizenship performance. Taking inspiration from Mahmood Mamdanis's work I propose to look at the re-production of power relations between the state and the subjects of citizenship from a historical viewpoint. Mamdani's core theme is instrumental here: By locating both the language of rights and that of culture in their historical and institutional context, I hope to underline that part of our institutional legacy that continues to be reproduced through the dialectic of State reform and popular resistance. The core legacy, I will suggest, was forged through the colonial experience. 81 In the last 150 years, large parts of the research area were under political structures designed to gain territorial control and exercise power. Though land was abundant and people were able to easily migrate, at the same time "property rights over people" increasingly developed. 82 Thus began overlapping processes of territorial control: several centres lay in competition over land and people and were eventually bound together by the expanding centre. 83 Ethiopia developed in its western fringes a twotiered system, similar Mamdanis's post-colonial state structures: the peasants were under control of a constellation of locally defined elites in the local states, "supervised" by officials deployed from a pinnacle at the centre. 84 The Oromo kingdoms or the watawit sheikhdoms clearly were bound by frontier-  . 84 Mamdani,Citizen and Subject,287. control of its subjects. Hence it is the legacy of territorial formation and the quest for territorial control from which Ethiopia has formed. These legacies inform the impasses of the centre-periphery divide until today.
There are a couple of important patterns to keep in mind when trying to define the citizen-state relation in Ethiopia from the perspective of minorities or subaltern Designed as a historic approach to citizenship I deem necessary the following outline in order to understand the impasse of citizenship: I will give a brief background to the ethnographic past and present in relation to social change in the next chapter. I will reflect on the historical emergence of the frontier and the fundamentals ambiguities created thereby, which built venue of the historical encroachment of the Ethiopian state (see Chapters 3 and 4). I will give the occurrence of slavery special importance in trying to portray the marginal and powerless position of the Mao and Komo in the regional power architecture. Slavery here is not only seen as a core element in the relations between centre and periphery, but also as a defining pattern of the social memory of the Mao and Komo as well as a problem of Ethiopian political culture. I furthermore portray the provincial administration as a continuation of exploitation of the regional minorities by the regional elites. These patterns are fundamentally shaken during a short period of regional turmoil during the revolution and the Civil War The Frontier: Peoples and Identities The Mao and Komo are not two groups but many. They are often understood to be culturally related peoples. This cultural inter-relation is concerned with their common cultural traditions of e.g. sister exchange marriage, material culture, an egalitarian social organisation, their livelihood strategies of slash and burn agriculture, fishing, hunting and honey production. These cultural traditions and livelihood strategies speak of a common cultural heritage. At the same time this common cultural background, coupled with their rather dark skin complexion -as compared to neighbouring groups -has led to their lumping under various blanket terms, like Amam, Burun, Koma, shankilla. Even the term Mao is rather a social label than an ethnic term. The culture and institutions of these people have been exposed to multiple layers of cultural contact and changes. The "borrowing and lending customs, practices and vocabularies" happened at such a rate that it is illusionary to believe in carving out original cultures. 1 Nonetheless there appears a cultural repertoire. Wendy James remarked, that on a "deeper, perhaps the 'archival' level, there would seem to be in this region elements which have long circulated between the various 'pre-Nilotic' peoples". 2 González-Ruibal refers to this archival level as the archaeology of being Mao and adds to the Nilotic elements, the Omotic and Cushitic traits: "if we excavate the different layers we will find a variety of technologies (social, religious, historical, and material) coming from Koman, Gonga and Oromo backgrounds." 3 Appreciating the cultural traits will give both an idea about the identity of the people in the shatter zone, as well as for the feeling of change that the Mao and Komo endure.
The cultural and political forces that impinge on peripheral groups led to different reactions among different groups. 4 The Mao have chosen forms of "cultural mimicry" and adaptive mechanisms, 5 while the Komo have tended to seek refuge on the frontier stiching their culture together from deep rural elements. 6 In the following I want to first look at the complex existence of xenonyms and selfdesignations and their impact on the identity of the people. This builds the parenthesis for a presentation of shared cultural elements in regards to cultural changes and the practices of memory.

Ethnic terms and ambiguities
The focus of this thesis are the titular groups of the Mao-Komo  by the Italian ethnographer Grottanellli, a term that has later been taken on by Murdock in his general ethnographic overview on African people. 9 This term has recently been criticised by González-Ruibal as implying a form of double primitiveness: as the primitive ancestors ("pre-") of a primitive people ("Nilotes"). 10 After Küspert, 2015 Komo today is an ethnic term that mainly identifies speakers of the Nilo-Saharan language tta komo and is also sometimes used for Gwama, the speaker of twa gwama.
The Komo live in Benishangul-Gumuz Regional State, especially the Mao-Komo special wäräda, as well as Gambella. Both Komo and Gwama are also living in the Sudan.
The term Mao is used in Ethiopia as a social label (which I will discuss further below), as a xenonym for both Nilo-Saharan and Omotic groups, as well as a selfdesignation. As an autonym, it is used by the Anfillo Mao, the Bambasi and Didessa Mao (e.g. Màwés Aas'è, 'the mouth of the Mao' 11 ), which are referred to as northern Mao, as well as the Omotic Hozo (e.g. maw shulojo 12 ) and Sezo (e.g. maw sedjo).
These latter two languages are also often referred to as Begi Mao. Mao is an Omotic term and refers to "man", or "people". "Mao" also appears as self-designation among Gwama of the Mao-Komo special wäräda and the Begi area (in the latter area speakers of twa gwama often refer to themselves as sit shwala, "black people"). In an area in which Oromo is the lingua franca, all groups frequently refer to their language as Afaan Mao.
As such the terms Mao and Komo are both used as ethnic terms as well as social labels and their usage depends on a complex web of historical perceptions and interactions between "indigenous" and immigrating groups as well as between the state and its subjects (table above). The term Mao is also a blanket term, applied by the western Oromo to designate the western border peoples which will be referred to here as Omotic Mao, Koman Mao (Gwama) and Komo. In his seminal article, Bender noted that the term Mao was used by the Oromo "to refer to the very dark-skinned people of the area, much as the general Ethiopian term "Shankilla" is used". 13 I don't think the term Mao is as generic. It rather depends on the social proximity between "black people" and Oromo. Also, Wendy James noted that the prevalence of the term Mao largely depends on the cultural neighbourhood of the border people with the Oromo, and specifically on a historical patron-client relationship between the two. 14 Today the Koman Mao (see map p. 247) use the name Mao regularly, while the lowland Koman (Gwama and Komo, see map p. 247) are usually referred to as Komo and also do rather refer to themselves as Gwama (mostly as Komo though) but never as Mao. A distinctive self-determination is the self-reference "sit shwala" by the Koman Mao who live in Oromia. Almost in an act of "defiance" this section of the people rejects the term Mao. 15 Names are cultural and political practices that can highlight the general features of inter-ethnic relations and patterns of identity formation. Gwama groups do also live in Gambella, but here they are "invisible" among the Komo groups and also use the name Komo for self-reference. But both the 13 Bender, 'The Mao Problem', 128. 14 James, 'From Aboriginal to Frontier Society in Western Ethiopia', 62. 15 González-Ruibal, An Archaeology of Resistance, 262. On a similar point see: Küspert, 'The Mao and Komo Languages in the Begi -Tongo Area', 10. Wendy James noted, for the case of the former slaves, reabsorbed in the valley communities: "the appellation 'Shankalla' has a wide ambivalence which can work to the advantage of the valley people, since a racial category applied by highland society can be read as a social claim, almost a claim of fellow citizenship, on the part of the Gumuz speakers of the valley." (cp. Wendy James, 'Lifelines: Exchange Marriage among the Gumuz', in The Southern Marches of Imperial Ethiopia, ed. Donald Donham and Wendy James (Cambridge: Cambridge University Press, 1986), 144. self-designation Gwama and Komo indicate a social and political distance and independence from majority groups.
Processes of migration and movement have created a complex web of ethnic affiliations, based on assimilation and integration within these different groups and into neighbouring communities. Thus, a dynamic ethnic ambiguity seems to be a feature of these groups as contrary to the attempts of the Ethiopian government to territorialize the groups in specific region, special wäräda or killil (the federal state). The history of the Mao (Gwama) and Komo is framed by migration and dispersal. Samuel Burns in 1947 noted in his sketch grammar on Sudanese Komo that Komo "have occupied roughly the same territory for more than 150 years or longer". 16 But he also indicates a migration from their previous Ethiopian settlements. Caught between "anvil" and "hammer", the "Nilotics to the west" of them and "the Bertas and Gallas to the east", continues Burns, the Komo were dispersed by and subjected to slave raids. 17 For Burns the Koma were, due to dispersal and flight, "broken and scattered", and "fled to live in widely separated sections". 18 16 Samuel Burns, 'Notes toward a Grammar of the Koma Language' (Sudan Interior Mission, 1947). 17 Ibid. 18 Ibid.
The Gwama and Komo often refer to each other as "brothers". They refer to a constant migration between Banga and Gemi, as well as the joint experience of hunting together, marrying and holding celebrations together. 19 But especially the experience of dispersal and flight unites these groups. Garra Gemi is recognized by the Komo as the ancestral homeland, while the Gwama refer to Garra Banga in the Mao-Komo special wäräda as their homeland. The Komo settlements of Garra Gemi are said to be abandoned today, although respondents often told me they would go visit Garra Gemi in the dry season, "because we believe this land is ours". 20 Banga, in the western lowland of the Mao-Komo special wäräda is a kebele known to be inhabited by Komo Saharan as well as Omotic groups. I assume that mostly the sit shwala have adopted the Omotic names. I take this to be an indicator for a long experience of contractual "inter-ethnic clan relations": at a certain point in history, groups with different background may have made a mutual decision to regard each other as brothers. In such cases one may speak of collective adoption. 26 From this "brotherly" interaction, one might assume those who refer to themselves as sit shawla developed their identity based on Nilotic and Omotic cultural traits. They can be regarded as the result of a trans-ethnic history. 24 The most comprehensive list can be found in Grottanelli, I Mao. Missione  The occurrence of the name Yaalo/Yaala/Yaya among all groups is another fascinating feature. The existence of the village of Yala, which was described by Schuver (s. Chapter 3) is particularly noteworthy. 27 Schuver's Yala was situated in the Gwama speaking areas south of Fadasi and a cultural meeting place. I take it as an indicator that both Nilo-Saharan groups and Omotic groups met there and departed from there again taking with them the collective name Yaala. Alliances complement the sphere of inter-ethnic relations.
Also in the case of the Komo and Gwama the primary marker of identity was the patrilineal clan (Komo: m'os), that used to reside in a clearly defined area. 28 The clan base is still an important territorial orientation of origin but makes only but little sense today and is overshadowed by new territorial markers and markers of identity. This patrilineality is nominally emphasized, but there are strong "matrifocal tendencies": 29 the role of the woman is very strong and there is a belief that the mother transfers the identity to the child (a child becomes Mao, Komo or Gwama through the mother, who spends time with it and educates it). This perspective also becomes very important in the discourse on transmitting the language and the "culture" to the children. This matrifocal perspective is also strongly emphasized by Joachim Theis in his work on the Koma groups of the Yabus valley. 30 For Theis, the relation the Koma have with their mothers is of seminal importance, which, according to him, also reflects in the use of the word kum ("his/her mother", by which many words are formed). Samuel Burns noted that the self-identification of the Komo (gokwom) would mean "their mothers". 31 27 James,Johnson,and Baumann,151. 28 Theis,Nach Der Razzia,79. 29 González-Ruibal, An Archaeology of Resistance, [70][71]. 30  were alienated from the forest, and they more and more abandoned hunting and the collection of honey. Where honey can still be collected the Mao and Komo also sell it on the markets. In some areas of the Mao Komo special wäräda, Oromo traders come to visit the villages to buy honey and sell it in the markets. Honey is also, on special occasions, added to the traditional beer (shui and shul) in Komo and Gwama respectively), otherwise it is prepared with sorghum and water. Women earn some money by selling local alcohol and beer during the market days in the villages. Only

c. Some Observations on Trans-Ethnic Cultural Elements
There are specific cultural elements that seem to differentiate the groups in the borderlands and their descendants. These elements are internal markers of identity and form a cultural repertoire of the people. The pre-Nilotic cultural inventory is under threat, both by mainstream cultures, or government and development discourses.
While sometimes cultural features seem to have radically changed, in some cases the cultural resistance is outstanding.
Ritual scarification and especially the removal of the incisors, both among women and men, when they reach adolescence are still practiced among the Komo. According to information from interviews, it was not practiced among the Gwama, although I believe that assumption was partially based on the political discourse of "harmful cultures" and shows the divide between lowland and highland Koman as well as the influence of "mainstream culture". 33 Among the Komo and Gwama as a rite of passage the usual explanation is that it is a 'beautification', and neither man nor woman should marry without undergoing the removal. 34 I understand, though, that the removal of the incisors is an optional practice.
The consumption of shui, often acknowledged as a highly important cultural feature, a form of social glue and cultural identification, seems to be practiced more practice. Gwama living with the Komo have similarly practiced tooth removal. But also among the Komo this practice seems to recede. 34 A more "scientific" explanation I got from some elders was that in cases of foot-and-mouth-disease sick people could be fed through the hole in the jaw/teeth with the pinse, the straw that is used to consume the beer. 35 González-Ruibal, An Archaeology of Resistance, 72. 36 Grottanelli, I Mao. Missione etnografica nel Uollega occidentale. 37 Triulzi,Salt,Gold,and Legitimacy,[32][33][34][35][36][37] Berta, known as feda, was traditionally carried out at the end of the harvest season.
The feda was organized among different Berta clans and led by their respective agur.
According to Triulzi the feda was a symbolic manifestation of the settlement patterns and the group identification. […] there is stress on the cooperation, equality, unity, the importance of the forest and its resources, and the ownership of territory. These values were played out through walking the land together. There is also the possibility that the Mao adopted the ceremony from newcomers, but hunting has always played an important role among the indigenous Koman peoples, to which the

Mao belong […] 40
In this regard, the author stresses the fact that also the Komo have a hunting master. 41 I was not able to get to know hunting master in the interviews I undertook, though the forest and the hunt are recurrent themes in the interviews I conducted. Especially also the use of medicines, and spiritual powers deciding over the outcome of the hunting were recurrent topics. The consumption of raw meat either of antelopes or warthogs seemed quite common to me during visits in the villages. The forest, the search for honey and the hunt for animals play an important role in the coming of age of peoples, like the following interview excerpts might show.
A: When I was eleven years, my father said come to me...do you know me...he showed me the honey...he said climb up, and bring the honey down. I was afraid of the bees and wanted to run away. I climbed up the tree and when the bees attacked I fell off the tree. My father was beating me and I went back up on the tree to collect the honey.
Again, another day, my father took me to the bush; we did not have any water with us. We slept near the river, but I didn't know there was water. He was teaching me the rules of the bush. I was crying to him to give me water, but he said, 'I am sorry we don't have water'. Then he gave me honey to eat and I got even thirstier. But he didn't give me water. Then in the morning he showed me the river and I ran into it to drink like a cow.
He also showed me how to hunt and gave me a gun to kill the buffalo. When I shot the buffalo, he was proud of me and said 'you are really my son'. The Komo were known for their hunting skills beyond their territory. As skilful hunters, they were also employed by other groups, as the following story shows. It clearly speaks about the pride related to hunting: A: There is some people called Gindaberet. They also spoke Oromiffa. They used to kill buffalo and elephants to show that they are heroes. After the killing they cut of the tail of the buffalo to take it to their home to show that the have killed the buffalo. I was hunting with them. When they came, they had no gun. When they came, they asked the people who knew the place of the buffalo and the elephant. They asked the akui, 43 "Do you have a gun?" 42 2/2011: Interview Komo elder, Gambella town, 13 August 2011. 43 Akui is a village headmen or local leader.
Then the akui gave them a gun, for the price of five bullets. If they didn't give us the five bullets, we didn't give them a gun. If we give them the gun, they would shoot you and kill you. So, we went with them by two and only gave them the gun the minute we had the buffalo to be killed. We made a trick: because we know how to kill the buffalo we told them that our shot killed the animal, and that he doesn't have a right to dance and sing; because we killed it.
Because the place which their bullet hit, is not bleeding a lot, and the place where I hit, is bleeding a lot. So, the chance is mine to sing and dance. Then we discuss and make them pay the tail (5 birr), the meat (30 birr), the horns (50 birr); then they can take the things home and sing and dance and be a hero. I was hunting with them a lot. I made a lot of money with this. After the Därg came the Gindeberet did not come anymore.
The mother will be happy, because her son is strong; she will put butter on the head and beads around the neck; "you are fighting with buffalo you are strong..." and then they prepare shui for the celebration; he will not do any work for five days, for five days everybody contributes maize and sorghum to make the shui and everybody will sing and dance.
Q: What about elephants?
A: We are looking for his footprint and follow. The elephant is very serious.
When they open and shut their ears it gives some sound. The place where they are you can hear when they are moving their ears. Here it is. When we are strong men, three or four, we say here it is and then we go for it. The person who knows his behaviour is me. When the wind is blowing in the direction of the elephant, then we will not move forward; we have to change the side and come from another side. Elephant seems like a good animal, but it is not; when we are very close, we shoot. We shoot with automatic! The elephant starts shouting, if we run away, the elephant will finish us! We stand and keep shooting. The other elephants are coming to help him and if we run away, they will follow us and kill us.
When he died, we take off his teeth...(whispers) we took it to our home. One is a round 2,000-3,000 birr, the other one we exchange it for the gun; three guns plus bullets. At that time we were afraid of the Nuer, we were fighting with the Nuer.

Sister-exchange marriage
The institution of sister-exchange marriage is one of the central Koman institutions in the research area. Exchange marriage has received some attention and has been studied in the context of the Uduk, Gumuz and Koma. 44 In Ethiopia, González-Ruibal indicated that sister-exchange is still occasionally practiced among sit shwala in remoter places. 45 It also appears that the Gumuz still practice it. Specifically, for a Komo and Gwama perspective, it was most thoroughly studied in the Sudan by Joachim Theis. There are regional and local variances in the system. The essence of the exchange is that the groom gives his sister or cousin to the family of the bride in exchange in order to keep and exact balance between the marrying parties. 46 According to Theis' observation, in its ideal form it guaranties the reproductive capacity of the patrilineal clan, since children from exchange marriages belong to the line of the father, while children from other marriages belong to the line of the mother. 47 There are two things particularly interesting about exchange marriage in the research area. First it was attested to me both by Omotic and Koman groups as cultural history.
Since it belongs to the deeper Koman culture, the Gonga population seems to have taken up this system and hence affirmed the inter-cultural relations that might have been the beginning of the previously mentioned inter-ethnic clan relations. All groups have indicated during interviews that the "original" form of marriage was through sister exchange. But this brings out the second interesting point which indicates that the system has been largely abandoned. Among the highland Koman population a frequent statement is that the tradition of exchange marriage has been changed to a system of bride wealth: "we now do it like the Oromo". In the lowlands it is often 44 Wendy James, 'Why the Uduk Won't Pay Bridewealth', Sudan Notes and Records 51 (1970): 75-84; James, 'Kwanim Pa; James, 'Lifelines: Exchange Marriage among the Gumuz'; Theis, Nach Der Razzia. 45 González-Ruibal, An Archaeology of Resistance, 266. 46 James, 'Kwanim Pa, 241. 47 Theis,Nach Der Razzia,123. related to governmental influences. But the time is not clearly indicated. Some people referred the Därg time, other to the present government. 48 Either way in several interviews the appraisal was, "the government told us to stop". A regular reason for the change was that "people who don't have a sister cannot marry". Bride-wealth today includes money, lifestock (mostly chicken and goats), cloth and alcohol.
The vanishing of sister-exchange, to my understanding, did not significantly alter interethnic marriages. Inter-marriage between e.g. Oromo and Mao is rather rare anyway.
But the changes are surely socially significant, especially since exchange marriage has also been an act of defiance and cultural integrity in the face of the hegemonic neighbours. 49

Religious Traditions
The culture of the Mao and Komo has been changing tremendously. Mainstream society has been imposing majority cultures on the people and the Mao and Komo continue living between adaptation and avoidance. Today mainstream religion is Islam although rigorous attempts to proselytize can be felt both from protestant and catholic groups.
A Komo elder in Pokung once made an eloquent point, when I was asking for the origin of the brick-built church building. He said, that when the missionaries come, "we're either Protestant or Catholic -the rest of the time, we are Muslim. But we drink shui and we pray in the house of god (gubbi waal)." 50 Waal is either a supreme being, or the term for the sorcerer, or the diviner. "Since the missions are here, god's house is empty" was the response of another discussant during the same visit in A contemporary example that illuminates this point further is the existence of the term Yere, a supreme being, that is found in various groups and contexts. Yere can be found as a name for both the Christian God or Allah, 55 as well as in various forms among the Omotic-speaking Triulzi,Salt,Gold,and Legitimacy,[51][52][53][54] Examples from the language development efforts currently undertaken by the Norwegian Mission in Benishangul is the short reader Yan a Yere gi ttwa Gwama, "Worship god in the Gwama language". A Gwama speaker once told me "Yere simply means Allah" (22/2010: Mao elder, Qama, Begi wäräda, 16.10.2010). 56 At a very early stage of the research, when I found evidence for Yere to be known among the 'Mao' in my research area, I was convinced that the people I sat with were Omotic-speaking Mao, only to realize that they were Gwama-speaking Mao. and Mao have adopted the term Yere "the paramount divinity of the Gonga peoples". 57 Although the Busase were nominal Christians, they also had "pagan" gods.
Their supreme deity was the god of the sky god Yére, similar to Yeró of Kaffa. As in Kaffa, there was a superstition of a man-hyena called Kuoro. 58 This syncretism of monotheistic elements with local spirits, demons and the belief in diviners and sacrifices is a prevalent feature also among the Nilo-Saharan Mao and Komo of today. 59 Some observations I made during consecutive occasions regarding religious beliefs and practices might be illustrative at this point. On one occasion a ritual was staged for me. I had bothered my host with several questions so that he decided at one point to ask the local diviner (sit bish) to perform a worship. The ritual beer was water and the chicken that would under real conditions be sacrificed was left alive. During the ceremony in a swal kwama, in the area of Begi, the participants, all wearing Muslim attire (jallabiyyas, scarfs, and prayer caps) were squatting around in the fire place holding their hands in front similar to Muslim prayers. The sit bish lead the ceremony.
He spoke a prayer of which every sentence was answered by the people with the words yere siezi ('God is Great'). The sit bish passed the calabash of water (beer) and everyone who drank, spit some beer in the fire for blessing. The sit bish blessed all attendants by sprying beer on them.
The swal gwama is an interesting feature of the Mao culture. The staged ceremony took place in what Grotanelli had referred to as beehive huts (campanne alveare); it was lying a little outside the main settlement and certainly further away from the Oromo houses. González-Ruibal has documented the use and structure of these houses both among sit shwala and Hozo groups, which he took as proof that these cultural huts were a "pan-Mao phenomenon". 60 57 González-Ruibal, An Archaeology of Resistance, 73. 58 Girma Mengistu, 'The Busase of Anfillo, Qellam, Wallaga (A Historical Study)', 5. 59 Ernesta Cerulli remarked, quoting Grottanelli's observation, that the Komo (our Gwama) pray to a supreme being Yere Siezí on their knees, which, according to her, goes back to Muslim influences (Cerulli, Peoples of 33). 60 González-Ruibal, An Archaeology of Resistance, 296.
A powerful reminder of the fact that the Mao create "spaces of cultural persistence" with these houses where the Mao and Gwama retain their culture is a remark by González-Ruibal: The swal kwama and the Mao cannot be separated. They are the same. It is like the mosques for the Muslim and the church for the Christians. We pray there to our god. The swal kwama has to be perpetuated through generations.
The moment the swal kwama ceases to exist, there will be no more Mao. 61 Despite the importance of what is perceived as deep cultural tradition by the Mao, the influence of other religious concepts is striking. The following is a song, a young informant had texted in the Gwama language. He was a young Christian Mao and went to the Mekane Yesus Church, which is attended by the Oromos of the region:

Gashi man kule. 62
A year after I had met this young man, I had the opportunity to follow a SIL-led mission to record songs from Koman and Omotic Mao in order to use the melodies as a cultural tool to proselytize and give the melodies new Christian wording in the respective language. For this purpose, the SIL had invited an ethno-musicologist who was a respected specialist for such kind of endeavours with long field experience from other parts of Africa. After several discussions and small games and exercises the respective groups of Omotic speakers, and Nilo-Saharan speakers were asked to perform a song which they "found representative of their respective culture". The groups surrounded by the already Christian Mao, the performer of the earlier song, who was invited to the workshop because he was already seen as key figure to access the Mao community, presented no songs (war songs, hunters songs, etc. like the Omotic groups), but staged a dramatic enactment of the visit of a man to the swal gwama and the killing and sacrifice of a chicken, the consumption of beer and the blessing thereafter. All in all, they lengthily performed the visit of a man at the diviner with precisely all the aforementioned ritual aspects. A broom was used as the spear of 61 Ibid., 299. 62 4/2012: Begi, Mao youngster, Begi town, 04.October 2012. Translation: Come my Lord-Shake my hand-It becomes morning-The dark has gone-The sickness has gone-I stand. I am grateful to Klaus-Christian Küspert. It was during joint research, that I was introduced to the informant and that we recorded this song. the sit bish, a piece of cloth was the chicken, an empty basket functioned as the calabash for the beer. All the performers, men and women, were very clear and precise in the ritual steps they presented to the audience.
The workshop leader was puzzled though because there was hardly any melodically aspect and the ritual sacrifice of chicken was not of any help for the ethnomusicological mission.
Both enactments of the sacrifices for Yere show how important this aspect of traditional religion for the Mao (Gwama) still is and how it retains a cultural space of their history.

Five encounters with identity
Citizenship and the idea of belonging are largely defined by the history of the interethnic encounters. How a society opens to new-comers, leaves political space for its minorities, etc. should be discussed before the historical encounter and growth of inter-ethnic relations. We will see in the selected cases below, how ethnic labels, history, and webs of meaning define questions of belonging to the region, to the nation, or to the cultural neighbour. The area of concern is peripheral, and hence a structural challenge to national citizenship. Ethnic groups and individuals often feel neglected by the government and marginalized. These feelings are often being presented on the basis of historical perceptions, as we will see in the course of the thesis.
Apart from cultural relatedness of Mao and Komo that I have presented earlier, there is also a historical relatedness that informs much of the self-identification and the identification by others, and the marginal role they have inherited in their relations with other groups as well within consecutive state systems. Ethnic groups are usually studied in separation. The case of the Mao and Komo provides space for a comparative approach to cultural history. I take it for granted that ethnicity and ethnic groups are hardly stable complexes and that social structure is not a "equilibrium". 63 […] "ethnic groups" and "culture" are cumulative historical constructions, and evolve not in isolation but in interaction, often within a regional or wider political-economic network. This statement implies that there is no primordiality to be ascribed to "ethnicity" except at the danger of reifying difference, and essentializing "culture" (including language). This does not mean that the cultural complexes and ethnic identities referred to by people and elaborated in ritual, world-view, and values have no meaning or are arbitrary bricolages of cultural material. On the contrary, they are a rich source of belonging and group-esteem and show a measure of internal cohesion. But they are dynamic and changing. They cannot be pinned down only to criteria like language, a common territory, a common "psychological make-up," whatever that is, or shared economic life. 64 Ethnicity, as I already indicated earlier, is a dubious identifier. Today, identity discourses and ethnic identification are important due to the political circumstances of politicized ethnicity that we find in all corners of Ethiopia. I also cannot not deny that there is a certain coherence of identification among the Mao and Komo themselves.
Historically, population movements, inter-ethnic encounters, and marginalization have framed the history of these groups. Also, assimilation into other groups and language shift seem to have frequently occurred. Framed by such history, the descendants of these different ethnic formations today live under rather precarious circumstances in three regional states. The following encounters with identity will give a first idea of the complexities that emerge in such a socio-cultural environment.
First encounter: A term among the various Gwama clans is that of the Arab Gwama. I crossed the refugee camp Tongo 1 and Tongo 2 in late September 2012 and headed to Wanga Gitten, where I talked to a respected and widely known Mao elder.
He had been named severally as someone I needed to talk to, and finally we met. The following is an excerpt from an interview conducted with this elder 65  Bergawi. I am speaking the Mao language of those who are living close to the river. I am the speaker of Giten. We call our language Gwama. When I came here we spoke Arabic, but when we came here we learned this language ("goKwama" i.e. twa gwama).
These Mao of Arab-Sudanese descent became the ruling group in the area, who, with the ascent of Kutu Gulja (s. below, chapter 3), managed to defend a certain territory against other forces, such as the Oromo and watawit, and eventually were incorporated into the economic system of the area as lower landlords (balabbat).
Second encounter: In the house of sit shwala elder 66 in Qama not more than 12 km from Begi along the old airfield, people spoke Gwama and insisted their self-name was "sit shwala". At the same time the elder emphasized that they were "Mao". I discussed this Mao-enigma  Mao is a term that comes from Anfillo and was used for black people by the Oromo, for the Kurrio, 68 Gwama and the people they found around Begi and Tongo.
[…] 'sit shwala' ('black people' in twa gwama) is my own language (twa gwama); sit shwala is a reaction to the use of the term Mao. It is a matter of dignity. The Oromo try to assimilate the locals (through marriage) to get the productive land.
Third encounter: In Laki, I talked to a group of elders who had been introduced by the village headman as Komo, during our discussion one elder got angry and said "if you come to ask about the Komo, you also have to ask about the Gwama".
Fourth encounter: Once, asking a local administrator in Gambella for research permission, we discussed my research application, which stated that I was looking for analysis of oral traditions of the Mao and Komo. He told me that there were no Mao in Gambella but I should go to Benishangul and Oromia for that. "But for Komo -no problem". He did not seem much impressed when I told him that there were speakers of Gwama living in Gambella under the name Komo that were called Mao in other parts of the country.
Fifth encounter: A Mao youngster explained his life history to me. 69 He was three years old when his parents died. One of his Oromo neighbours took him and adopted him (guddaficca). Thus he grew up as an "Oromo". He did not know he was a Mao.
But he played with Mao children, and went to the bush with them and for fishing.
At the age of 10 he was sent to the bush to plant trees. He met an old man, and friend of his dead parents. The old man asked him whether he knew he was from Mao.
He laughed at him. But the old man told him the story of his family and where his sister and brother lived. He became interested in his family history: The old man told him, 'if you visit me, come in secret and I will tell you more'. So, he went to visit him frequently. When his gudaficca-family found out about this, they beat him. After which he ran away from this family and escaped to his brother and sister. His brother 68 Kurrio is a blanket terms that appears regularly in the research area. I understand that it designates the Omotic-speaking Mao of the research area (16/2012: Interview Mao administrator, Asosa, 23 October 2012): "Kurrio is similar to Anfillo; they speak an Omotic language; they are the people of Gumma Garra Arba, etc."). Cp. also the Kaffa link described on p.74 of this thesis. Grottanelli  and sister lived with a protestant Oromo family. He was taken to church and baptized.
His elder brother was taken into the Därg military by force. After some years the brother came back and the government gave them land, because they were regarded orphans. The three started to live together until his sister married a Mao and his brother married and Oromo woman. When he was 15, he had an argument about inter-marriage between Mao and Oromo. The argument was taken to the elders. They argued that if Mao girls were to marry Oromos also the Mao should be able to marry Oromo. In this conflict, the Mao elder made an ultimatum, saying, if the Oromos will not accept Mao to marry Oromo girls the Mao girls should come back to their families. After this the Oromo elders accepted and allowed inter-marriage.
He met an Oromo girl at the market. She was looking for help and he helped her carrying coffee. They got married and today they have two children and take care of a nephew.
Fractured or fragile identities?
Comparing these encounters poses many questions but at the same time opens ways to enquire into the Mao-Komo enigma. Obviously, a large population of Gwama speakers (otherwise omitted from the ethno-administrative landscape of western Ethiopia) is currently divided in three groups: the Mao group of Sudanese Arabic descent, the 'Oromized' sit shwala and the lowland Gwama, who are identified and even identify with the term Komo.
Despite the common language, processes of migration and territorial control have largely alienated the groups from each other. It is explicitly at the intersection of patterns of control and the current federal border regime that led to, and will strengthen, the alienation of the groups among each other.
In the following section, I will investigate more closely the forces and factors that contributed to the splintering of Mao and Komo identity, set in a regional comparative approach. It will look at the changes in the society from a regional-economic actor based perspective, mainly considering the historical, socio-political circumstances.
In the later part of the study I will focus on the ethnic territory of the Mao-Komo special wäräda, and the Mao and Komo as the alleged beneficiaries of this territorialized approach to self-determination. Thus, against the historical background of state-encroachment, social-stratification, territorial control and changing patterns of minority-majority relations in western Ethiopia, this study aims at analyzing the historical process that has led to the creation of the special wäräda. As a synchronic and diachronic political ethnography of the Mao and Komo, the study looks at the functions of ethnic federalism in a micro perspective and offers a glance at the changes and continuities of majority-minority relations under the current political framework.
This will contribute to an analysis of the 'accommodation of diversity' with regard to the stated aims of political empowerment and national integration.

Chapter 3
Serfs, slaves and freemen on the border: Approaches to regional history before the integration into Ethiopia (ca.  Identities are and were in constant flux in the research area. Xenonyms and selfdesignations are overlapping in travel literature, interviews or government reports.
With traders, refugees, conquerors, and eventually researchers, more labels have appeared. But not only the terminological landscape has been changing. The social formations in this area, as elsewhere, resulted from "social alignment" as well as "adoptive responses" due to economic and social circumstances and emerging group integration. 1 This means that groups diffused into stronger or more powerful Oromo country was also increasingly marked by political transformation of the traditional gadaa system towards territorial-based hereditary rule over people and land. 5 In all areas systems of tribute extraction and control were in place, which deeply affected the social systems of the Mao and Komo. Some were reduced to serfs and slave-status and other became middlemen within these political systems. 3 James, Johnson, and Baumann, Travels in North-East Africa. 4

The Gonga Frontier and the 'Mao Problem'
Historical accounts concerning the Mao and Komo often take a south-eastern, which is an Arab, Funj, watawit perspective. 6 While this is certainly an important feature of the groups' past, as we will also see later on, the view from the Sudan has to be supplemented by a view from the east, framed by what happened on the Ethiopian side of the border.
The existence of different Mao groups in western Ethiopia has puzzled various observers. 7 Although it is beyond the scope of this thesis to substantially contribute to this historical problem, a few words seem necessary to understand the connection between the different Mao groups. The main problem for the historical reconstruction of the ethno-genesis of the Mao is the lack of linguistic research to establish the link between the different Omotic Mao languages, their diffusion in western Ethiopia as well as their relation to the  Historical linguistics suggested that the Mao problem is part of the widespread existence of an Omotic Gonga population roughly between the Abbay and Käfa (evidence are pockets of its northern-most extent: the Boro language of the Shinasha).
The Mao languages may be remnants of a once wide-spread Omotic population that has been "split up by Nilo-Saharan incursion from the west, and later Oromo invasion from the east". 9 Recent research has established a strong presence of Gonga memories in the Benishangul area and especially the existence of a Busase past in the region. 10 The Busase are splinters of a Käfa nobility, which migrated into the area of Tullu Walal somehow before the settlement of the Oromo, and subsequently conquered and subjected the areas and population of the Anfillo forest, i.e. Omotic Mao and adjoining Nilo-Saharan groups like the Kwama (i.e. Gwama). 11 This nobility subsequently 6 Mainly framed by the aforementioned travel accounts, as well as modern Sudan based perspectives (e.g. Corfield, 'The Koma'; Theis, Nach der Razzia. 7 Bender, 'The Mao Problem'; Fleming, 'The Importance of Mao in Ethiopian History'. 8 This problem has recently received revived interst; s. Küspert, 'The Mao and Komo Languages in the Begi -Tongo Area'. 9 Bender, 'The Mao Problem', 141. 10 González-Ruibal, An Archaeology of Resistance, 248-49. 11 Girma Mengistu, 'The Busase of Anfillo, Qellam, Wallaga (A Historical Study)', [11][12]. Quoting d'Abbadie, Girma also states that in 1843 the Anfillo were both at war with the Oromo and waged wars against the Massango (i.e the Manjangir and other Nilo-Saharan groups of lowland Illubabor); s. Ibid., 11. established what came to be known as the kingdom of Anfillo. This process proves to have been a most decisive event in regard to later events north of present day Anfillo.
While my own research has brought little new in this regard, a brief overview of what seem established facts is necessary at this point. The most comprehensive theory for the existence of the different Mao groups as well as the establishment of the Busase in Benishangul and Wallaga is another frontier episode: the Busase, overlords of the Mao, migrated to the region of present-day western Wallaga adjoining areas today under Benishangul, and brought with them their patron-client system they exercised over the Mao in Anfillo before. Hence as Fleming put it, the Busase came to rule over their kinsmen. 12 González-Ruibal has substantiated this linguistic argument. Based on pottery excavations he argues that the Nilo-Saharan groups (usually seen as the earliest inhabitants of the region) were already in contact with the Northern Omotic groups two millennia ago, which could "explain many of the cultural similarities". 13 The Busase introduced their patrimonial system of rule to the region; and within this system also Gwama came to be incorporated into a semi-feudal structure, becoming

Mao.
A comprehensive account of the Busase' quasi-feudal production and relation to Due to processes of inter-ethnic contacts, inter-marriage, adoption and cultural amalgamation, the historical Sayoo Oromo may have been quite heterogeneous, as "residence rather than kinship played a prominent role." 27 In effect this heterogeneity and cultural amalgamation could have contributed to the growth of more hierarchically structured societies, both as a sheer necessity to contain conflict as well as due to cultural borrowing, since some of the early kingdoms, like the Mao-Busasse (Anfillo) kingdom, where highly structured and complex societies. 28 22 Ibid., 139. In his descriptions of the Koma (our Gwama, s. further below) Schuver also mentions that the " mulattoes of Leeqa are a result of " crossing Galla with Koma " ; cp. James, Johnson, and Baumann, Travels in 152. 23 González-Ruibal, An Archaeology of Resistance, 243.There exists not only a local Oromo hegemony but also to some degree scholarly reproduction of this hegemony. The expansion of the Oromo and the memory of their conquest build the baseline of the scientific treatment of history of ethnicity in the region. To a certain degree this is unavoidable and largely the result of the historical process, but nonetheless noteworthy to understand our image of the regional past. 24 In some oral accounts the "Busase" and "Ganga" appear as clan names of the Oromo in the Begi area. which they were collected, it is a noteworthy account. First it recalls the integration of "alien" people into the local society and their gradual rise to power. A motif that runs through the local frontier history from the Arab wise stranger, to the näčč taro (the white king) of the Käfa, Busase and Anfillo. Furthermore, it makes the Ganqa visible.
Although they most probably were not a Nilotic group, but, with regard to our previous account, a possible Mao/Busase polity present in the area of Wallaga. The available sources are quite congruent about the heavy impact Jote had on the surrounding Mao and Komo groups during his rise to power. Although it is difficult 32 According to Schuver, Jote in his youth was known as Bula, which is the name Schuver uses to describe him (cp. James, Johnson, and Baumann, Travels in North-East Africa.: xcviii). It might be interesting to mention, but no trace I will follow here, that Paulos Daffa in his account of the strongmen of Wallaga, has contemporaries Jote in Gidami and Bula Djarso of Begi (cp. Paulos Daaffa,Oromo: Beiträge zur Politischen Geschichte Äthiopiens,39). This is obliviously another indicator for the immense complexities of power relations, and actual geographic local ability of territorial claim and rule based on oral accounts. For the argument presented here it wouldn't make a great difference if there was another strong man in the region of Begi with whom Jote was at loggerheads. 33 Tereffe Woldetsadik, 'The Unification of Ethiopia ( : Wälläga', Journal of Ethiopian Studies 6, no. 1 (1961): 74. to ascertain which groups were under his control, Yasin Mohammad notes that the Komos' traditional social structure was destroyed and substituted by the introduction of vassalage in which the local leaders were made representatives (warra gofta) of Jote and responsible for collecting tributes. 34 Also Oromo, loyal to Jote were put in control over the Komo in order to oversee the extraction of produce (this is also referred to as warra gofta). 35 Items used for tribute were honey, ivory and grain.
Jote's general influence and power is seen in the following statement by Schuver, who visited his court in the 1880s: If I call Bula a 'King' it is because a man, who can easily raise 20 000 spearmen, without counting his slave-troops and negro-vassals and who enjoys power and authority comparatively greater than any constitutional monarch, cannot well be called by the same name as the first small village 'chief'.
With the decline of the gadaa system at the expense of territorial control and the diversification of hierarchical structures, several Oromo clans and sub-clans lay in constant fight over territory, property and people with each other. In Western Wollega, Jote Tullu strengthened his position raiding Anfillo, Galaan, and Begi. In Anfillo he subjected the Mao and Anfillo as slaves and in Begi he was able to gain control over the Sudanese trade routes. 36 Jote's success was based on the subjection of territory and people to his rule, and the gradual establishment of a centre of power and capital in Gidami. 37 The highly stratified social structure neatly fit into, and was coopted by the merging Ethiopian state when the region was put under control by the centre.
Based on offices similar to those of the gadaa system, hereditary and institutionalised political offices emerged. Below the mooti were the kooro (cp. abba qooro). They built the connection between the landlords (balabbatota) and the king. 38 They were chosen by a council of landlords but had to be confirmed by the mooti. The kooro were in principle judges and collected tax.  A second wave of immigrants, Funj nobility, followed suit and integrated into the local Berta clans, gradually coming to rule the Berta. The agurs (Berta kings) defined their claims for power around Funj descent and bore the titles meks or manjil, indicating association with Funj royalty. 43 Locally, this new nobility was referred to as jäbelawin (Arabic for mountaineers).
In time, Funj settlers came to absorb the Bertha by forcing their political institutions on them: yet by a process of mutual assimilation and adaptation, they seem to have lost other aspects of their original culture. They spoke the Bertha language, adopted Bertha names and were even said to have become  the watawit, both "seemingly" Muslim claimed supremacy over the "pagan" Berta, who lived as their servants but not, as Marno makes clear, as their slaves. 54 This may indicate instead a form of feudal bondship (s. Chapter 5).
The watawit sheihks were not united. Rather small regional centres emerged around Some of these Amam groups seemed to have successfully resisted slave-raiding. Still a blanket term, it is not easy to identify the groups subsumed beneath, but it may be safe to assume that Grottanelli's 'northern Mao' (Omotic Mao) were among the Amam.
Also, further south in the area of present day Begi, the term seemed to apply to speakers of Gwama. The term Amam has been replaced with the term Mao due to growing influence of the Oromo on the expanse of the Arabs. 58  The reputation of the "fierce Amam" obviously echoed through much of the region, as we learn from a passage of Schuver's travel log; quoted is a statement of the governor of Famaka: Go wherever you choose, but may Allah preserve you from falling in with the Amam. They are so ferocious and hostile against strangers, that they not only kill them, but even their horses, mules and donkeys revenge their intrusion on Amam ground. 61 58 James, Johnson, and Baumann, Travels in North-East Africa.: lxiv, xc. 59 Gessi, Seven Years in the Soudan, 166. 60 Ibid. 61 James,Johnson,and Baumann,49. The "aggressiveness" of the Amam is explained by Schuver by the sheer devastation of the country by the razzias of Turkish soldiers during the Turkish occupation of the Sudan. While all previous attempts to travel beyond Fadasi (by Marno, Gessi and Matteucci) failed, it was actually Schuver who for the first time was able to investigate the westernmost Oromoland and visited the Amam beyond Fadasi. Schuver found the Amam in three areas "situated exactly to the south of Fadasi". The first area Schuver identifies is on the road between Fadasi and Leeqa, where he found three villages. The inhabitants were not able to levy taxes on the trade on this route but engaged in the trade as carriers and cattle drivers. 62 The inhabitants of these villages, Schuver goes on, The population of the Amam is estimated at that time at around 8,000 people and "spread over dozen villages". 68 Schuver's meticulous description of the Amam is interesting to recall in full detail: 62 Ibid., 40. 63 Ibid., 41. 64 The name is interesting because in Gwama 'swal' means house. 65 James,Johnson,and Baumann,40. 66 Ibid., 46. 67 Ibid., 50. I was not able to record any information on Toza in present day Begi. 68 Ibid., 40. This tribe has enjoyed for a long time a much exaggerated reputation for ferocity and inhospitality. The origin of this myth must be sought in the defeats which they inflicted on the vanguard of irregular troops of Mohammed at the time of the Turkish conquest. Happier and above all more obstinate than the Berta, the Amam set up a barrier against Mohammedan invasion, a barrier so invincible that they succeeded in hiding from the covetous Turks the existence of the rich and fertile Galla country of which they kept the keys, so to speak, involuntarily. It is said that the Amam showed themselves so jealous of their independence that they went so far as to massacre without pity the horses and donkeys of the fallen invaders […]. 69 The actual identity of the Amam is hard to establish, and I was not able to evoke any memories with the term. Yet in the 1970s, according to interviews by Alessandro Triulzi, the Amam were remembered to have ben subdivided in 14 clans. The interview eventually only mentions two, the "Yaala" and the "Kuguul." 70 The latter are a sub-group of the Gwama and the first is a trans-ethnic clan name that is mentioned among Gwama, sit shwala, Hozo and Sezo speakers as a clan of their own. 71 The existence of the village of Yaala is particularily noteworthy with regards to the distinction Schuver makes between Amam and Koma. The village of Yala (s. above the possible overlap with the trans-ethnic clan-name) is situated in Koma country. Yala was the residence of a notable Koma who gained influence by marriage to the sister of an Oromo chief. 72 Alliances complement the sphere of inter-ethnic relations. Schuver's description of the Koma is vivid: I saw in the Koma a striking contrast with all the other negro tribes on this side, a people who did not seek to procure slaves; in fact, I did not find any individual of a foreign race, except the Galla princess married to the notable of Yala.
[…] A surplus of goodwill, a general contentment, a lack of strongly marked egoism seemed to me to characterize this race. 73 69 Ibid. 70 Field notes of Alessandro Triulzi (Asosa-Begi, a-bg 4, p.1). 71 Interviews in Benishangul, 2010-2014 72 James, Johnson, andBaumann, Travels in North-East Africa, 151. 73 Ibid., 152. In the course of his further description it turns out clearly that Schuver actually traveled in Gwama territory, as he visits a place called Maganza and the "widely dispersed village of Bosho" 74 , names, today preserved in Gwama clan names (s also pp. 66 and 244 for a list). Hence, we see an example of the division between the highland sit shwala, which are Schuver's Amam and the lowland Gwama (Schuver's Koma) are already laid out. Furthermore, the existence of the Yaala as a clan name in various groups may indicate that Yaala was a permeable boundary for cultural elements of the Amam (Arab, Oromo influences and the Gwama Nilotic) and that the place became a meeting place of different groups hence the diffusion of Yala as a trans-ethnic clan name.
The Gwama of this era were still rather unaffected by the outside influence and also the exposure to slave raids was still little.
He accounts also for the fact that the Amam are most certainly related to the Koma and speak a corrupted dialect of the same language. This recalls the lingusitc variations between highland and lowland Koman. 75 He believed the 'Goma' (Koma) and 'Amam' to be "remains of an aboriginal race", driven into the mountains by successive invasions of the Oromo from the east and the Dinka for the south. 76 So, while the Gwama had already been in contact with the Oromo, their habitat was still rather unaffected by the intruders. The term Mao had not been introduced into the region by the 1880s and the influence of the Oromo was not yet as strong. The history of the Amam is intriguing and opens an interesting way into the ethnohistorical speculations about the Amam identity in relation to the history of the Mao.
Especially the contemporary cultural differentiation of highland and lowland Koman seems to resonate in Schuver's description.
"The House of David": The Making of a Mao Nobility All the above already testifies to a highly complex frontier mosaic, in which indigenous groups, traders, refugees, and rulers competed over resources, and for controlling space and territory. The power was mostly distributed between meks, who 74 Ibid., 154. 75 Küspert, 'The Mao and Komo Languages in the Begi -Tongo Area'. 76 James,Johnson,and Baumann,330.  This migration is directly connected to the history of the Mao clans of the Wärrä Dawd.
An interview excerpt will show some of issues that have to be clarified:  60). 79 Although all these groups belong to the list of Mao clans one can gather today from the people around Begi, the Sättä and Issa are said to be basically oromized and not to have Mao identity anymore. Nonetheless they belong to the mosaic of clans, and hence point into the direction of a Mao history. 80 Probably of Funj descent but the name did not arouse any memories during my investigations on the history of Begi. Instead the enigmatic first "Mao king Bek", "who lived on the mountain besides Begi" is often referred to as the namesake for Begi today. Muslim traders who entered Ethiopia on the hajj. The most prominent example is the highly venerated sheikh Ahamd Umer who eventually settled in Ya'a and whose shrine today is a centre of the Tijjaniya cult in western Ethiopia as well as a centre for ecumenical pilgrimages that attracts thousands of Muslim pilgrims but also Christians.
Pilgrims seek spiritual blessing or are attracted by healing qualities e.g. of the holy water of Ya'a. 82 The account above quite obviously mixes the history of Dawd with that of Kutu Gulja (s. further below) who was according to other oral accounts the 81 12-17 December 1997, vol. 2 (Kyoto, 1997, 391-402. contemporary of Jote Tulu. Nonetheless, the story is consistent in so far, as the Tukrir traders and pilgrims seemed to have settled in the area of Begi and established themselves among the sit shwala who were the Amam of Begi. At this point, according to the local traditions, the exact emergence of these Arab Mao is still quite speculative, but after migration from Kirin, the Mao may have taken political control over the Amam of Begi, where they took over local rule by marriage into the local nobility. This theory recalls similarities of integration and patrimony with the case of the watawit. It is worth mentioning also that "Cutu" is also a place indicated on Italian maps of the 1930s as well as a place mentioned by the Gwynn mission who carried out surveys in the region in the attempt to demarcate the border between Ethiopia and the Sudan (s. further below). As far as this evidence allows us, we might speculate that sections of watawit migrated southward from Mount Kiring to the area of Begi/Tongo/Kutu where they took over the Gwama language and gradually became a ruling linage of the highland Gwama.
Kring itself stands out in the history of the Mao of Begi and some further considerations may substantiate the speculations about the ethnic origin of the Wärra Kutu. Kiring is the name of a mountain near Asosa. It was the important trading post and meeting place of various peoples in the late 19 th century. 83 The first European to visit Kirin was the Dutch traveller Schuver. His description is highly illuminating: Kirin looked a queer place. The huts, in far greater number and far more compactly grouped than we had seen anywhere. They looked more like birds' nests than human dwellings, built as they are on and amid the massive boulders of the steepest and barrennest of murrains, a gigantic avalanche of titanic blocks reaching from the plain till some 400 feet on the mountainslope.
[…] Several of the neighbouring Arab Sheikhs of the Bertas entertain at this important place a kind of consuls, who have to look after the safety of their traders and other subjects, who resort hither during the middle and end of the dry season (February-April) to trade with the Abu-Rôȏ f nomads. These latter convey sea-salt to Kirin on their camels and also domur and in return Gallahorses, iron and Galla-cows; […] The scene in prosperous years was described to me as very lively one, the Arabs, Gallas Amams and jellaba [Bedouin] all 83 It seems to me that today the most regular usage of the term Kirin is a self-identification of Mao elders who refer to their Arab descent.
having their different fenced-in camps of tent or leafy huts in the plain below.
The Kirin people take no part in the market except in so far as their personal necessities are concerned but their Agur takes a slight tribute from the Amams & Gallas and receives presents from the Arab traders. 84 The description of Kirin is vivid and one can easily imagine that at such a meeting place the frontier shows its full socio-cultural aspects. People marry, mix customs and habits, slaves are being brought from everywhere and fill local social strata. The Another assumption about the origin of Kutu Gulja that is quite well-known is that he came from Bornu, which makes him Nigerian immigrants (Fallata/Tukrir). 89 In the following I am compiling a few information that I have collected from Mao elders and the descendants of Kutu. There is usually certain coherence in the oral accounts and certain continuity: the Mao were the first inhabitants of the Begi area, the Arabs ruled the area of Kiring and Kutu became ruler of the Mao. Kutu Gulja, who was later given the title of fitwarari, was also known as the leader of the Araba   These changes had also happened in consequence of the emergence of the Ethiopian empire in the region.

Chapter 4
The Making of a Periphery (ca.  In the attempt to understand citizenship as the legal place a person or a group of people hold in society, the actual patterns of state formation have to be looked at. The entry points are either the way people appropriate or react to the state. Another perspective is to look at the way the state actually governs, administers, and rules. 1 State formation has become a major concern in the social sciences and history 2 and has also been a matter of scrutiny in Ethiopian historical studies. 3 The peculiarities of Ethiopian state formation have led to a very strong narrative of state expansion and enforced a highly influential centre-periphery perspective. In this reading of Ethiopian history, the state culture radiated forcefully into the peripheries which were eventually integrated or are currently being integrated.  (Markakis,The Last Two Frontiers,16). This view though is too So far, we have looked at the early frontier history of the region and its overlapping and intertwined processes of consecutive cultural and political renewal. At the end of the 19 th century, another player emerged on the scene: the imperial Ethiopian state.
There is a problem of simplification here, as the image created is that of a unified periphery that was integrated into the expanding state at once. The process was much more complex. There were notably different forces at work in e.g. the case of the Oromo kingdoms and the sheikhdoms of Benishangul, and it was a gradual process the event of state formation actually was for them. In which way did the patterns of social stratification evolve in the newly forming state? This is the point where the nature of the entity has to be questioned and defined. What is it, at the end of the 19 th century, that the various ethnic groups were incorporated into? A colonial state, a traditional Abyssinian polity, or an empire? "[I] ntegration is a concept that usually refers to the process by which people interact to form some kind of viable political system. The broadness of the term has, however, let to feeling that it may not be an overly useful notion." 5 Were the expansion of the Ethiopian state and the permeation of state culture into the different areas that were forcefully incorporated into the new emerging entity a process of political integration? Eventually this question is difficult to answer and probably has to be understood to be an ongoing process. The very notion of integration is probably ill-suited in the Ethiopian context. It has been assumed by some writers that the concept of integration was an open-ended process to construct political order. 6 Have not rather different centres, with their own peripheries, become connected by economic dependence? A problem that emerges is the almost untameable idea of an integrative centre. The centre, according to this perception, eventually will have incorporated the periphery. Centre and periphery in this perspective are bound to become one. The nature of Ethiopian state expansion has posed many conceptual difficulties. Descriptions and analysis range from unification to national oppression or even colonization. 7 Furthermore it makes the periphery look quite helpless and exposed to the predatory practices of the centre. This in fact is true to a certain degree and of course the periphery is best described in terms of power imbalances. The perspective that subjects the periphery to the mercy of the centre often neglects the inter-group relations and stratification in the periphery itself. 8 The periphery is not a unified whole. This is to say that for a regional noble life in the periphery has been decisively different than for an average peasant. . He states that, "the image of 'radiation' itself denies the capacity of the 'peripheries to innovate (supra note). Moreover, it totally ignores the fact that these 'peripheries' were often former political centres that actively took part in the (re)negotiation of the national political order, among the plural state and non-state actors (supra note)"; cp. Ibid., 281.
Despite all valuable criticism, the interaction between centre and periphery is a matter of great concern and for any problem that arises from the analysis of state formation cannot be neglected. The centre/periphery paradigm has been highly influential in the last decades and has guided much historical analysis since The Southern Marches of Imperial Ethiopia. 9 In his seminal article concerning the fundamental nexus between centre and periphery, Donham calls for "some appreciation" of the centre that dominated the peripheries, i.e. of the structure of the central highland, the Abyssinian society and its institutions, in order to understand the impact these institutions had on the conquered areas. 10 Before the expansion of the core Abyssinian centre, this core had already radiated outwards and expanded into adjacent areas. The Abyssinian core expanded along the granting of gult rights, intermarriage between local women and soldiers who were granted gult rights, "and over time the intermarriage of local elites of mixed parentage." 11 Contrary to the European models of feudalism, though, the lords were not the owners of the land. In European feudalism, the "lord's power extended into the production process". 12 But in the Ethiopian context the rist rights were hereditary and the lord only had rights to the labour of people. Hence, the landlords "sent their agents at harvest time to collect tribute, often as much as they could, but played no role in organizing or directing production." 13 This very much reads like the blue-print to the patterns of rule and dependency that developed in the newly conquered areas.
Although again these patterns cannot be generalized, the famous typology of newly incorporated lands is a noteworthy achievement of Donham's treatise of the "periphery". Donham systemizes three forms of rule after the conquest: these were the semi-independent enclaves, the gäbbar areas and the fringe peripheries. 14 The periphery thus is a highly relative term. The expansion of the Ethiopian state system and its incorporation of the various parts of Ethiopia has been an ambitious and highly ambivalent process. It forged a quasi-state under the pretext of the colonial scramble 9 Donham and James, The Southern Marches of Imperial Ethiopia; Wendy James et al., eds., Remapping Ethiopia: Socialism & after (Oxford -Addis Ababa: James Currey, 2002); Markakis, The Last Two Frontiers. 10 Donham, 'Old Abyssinia and the New Ethiopian Empire', 10. 11 Ibid. 12 Ibid., 14. 13 Ibid. 14 Ibid., 37-44. for Africa, destroyed and annihilated long existing traditional societies, in search for resources to feed the expanding Abyssinian core.
The notion of the centre as the epitome of political power reduces the peripheries to mere recipients of state culture. It negates the fact that the peripheries actually often already had centres themselves with their own peripheries. A potent counter-narrative that the history of Wallaga has in store are the emerging strong men and later mootis of Naqamte and Qellem. 15 An example that also provides a counter-narrative is the palace of Khojali-al-Hassan in the centre of Addis Ababa, which I will treat herein further below. was in this regard that he mentioned that Khojali gave the name Mao to the people ("before that we were sit shwala"). 28 We will see later on that the fight over Begi continued well into the post-1991 period and required a referendum to delineate its position either in the newly created  Anglo-Egyptian Sudan, 1910-1930', African Economic History, no. 17 (1988: 23. 32 Triulzi,Salt,Gold,and Legitimacy,172. 33 This episode is also published in Ibid., 177. Khojali (s. below) had cooperated with the invaders. The country was severally raided,  -1935', 184. 42 Ibid., 185. 43 Ibid., 196. 44 FO 141/571, A.D. Home, Enclosure 1 in Appendix 4, Abyssinian slave raiding, Addis Ababa, (May 2, 1927   ['The Story of What I Saw and Heard: 1896-1922'] (Addis Ababa: Addis Ababa University Press, 2006), 45. The term abigar is a derogative denomination for the Nuer. According to Johnson the term loosely translates as cowboys, originally applied to the Nilotic cattle herders of the Southern Sudan, Dinka and Nuer who were in trade contact with the Oromo, e.g. at the court of Jote Tulu; cp. Johnson, 'On the Nilotic Frontier: Imperial Ethiopia in the Southern Sudan, 1898-1936', 232. sheikh's wealth. 46 The sheikh not only sold slaves but more importantly used unfree labour from his subjects. His regional impact is acknowledged by the fact that people coming from the Sudanese borderlands, especially as court and military slaves, were known in Addis Ababa and beyond as Hojale. 47 Some Berta also referred to themselves as "Xoyalee or Hoyalee". 48 "We were all Tsugel's 49 children" is a statement also found among Komo descendants to describe their relation to Khojali 50 . The was damaged by the dry season [bona] a loss of more than sixty to seventy thousand birr. Work diligently, as usual, so that the money will not remain in vain without serving its purpose implementing the planned gold mining [yäwärq sera = the gold work]. Yekatit (February) 11, 1931. 54 The almost legendary gold of the region (not so much the hardships of those who had to dig it as tribute) has always been a motive in the regional history. In some cases, slavery and gold have inspired the imagination of travellers and observers like in the case of Byron the Prorok, who described the 100-year-old "Mad Sultan Ghogoli": "Tales of his savage cruelties were legend; the mere mention of his name is enough to terrorize the natives. (…) parties which had tried to cross his territory had disappeared and never been heard from again". 55 53 Johnson, 'On the Nilotic Frontier: Imperial Ethiopia in the Southern Sudan, 1898-1936', 229. 54  He was in part responsible for a large number of slaves and workers that were needed in Addis Ababa's modernization process. 59 By the time the palace was built, in ca. 1910, observers estimated a number of ca. 15, 000 "Shankillas" and "Benishangul" in Addis Ababa mostly for household work and slave labour. 60 We might assume that most of them were trafficked to Addis Ababa through Benishangul. Sheikh Khojali obviously set up a regime of plunder in the countryside, that enabled him to please the 56 Avenstrup, Abessinien Kors Och Tvärs, 49; Grühl Max, The Citadel of Ethiopia (London: Jonathan Cape, 1935), 375. 57 The areal on which the compound was built, in a quarter in Addis known as Shogälle midär, is estimated at 340 000 sqm. Only one half of the building still exists; the other half was demolished during the Därg time and a mosque was built on the ground (allegedly to please Muslim sentiments in the area during the revolution). According to information of the family, Armenian, India and Turkish craftsmen were employed to build the palace; the tin-roof shows the logo of a French company. 58 Also Qǝddus Rufa'el (an orthodox Christian church in the proximity of the gǝbbi) was donated by him; Information from the Khojali family, Addis Ababa 28 June 2012. 59 I recorded a statement about Mao men who were traded to Addis Ababa "to work day and night". I had the opportunity to hear some recollection about one person who returned to the Begi/Tongo area. Information from Benishangul 2014. 60 Reminick, Addis Ababa,87. centre and at the same played into his own pocket. Local traditions depict him as cleverer (but also more ruthless) than the other rulers, e.g.: "he made us dig out 300 wäqet of gold, 200 were for his tributes, and the other was for himself".

Slavery before Abolition
Slavery is a well-known, though highly ambivalent chapter in Ethiopian history. 1 Especially in the wake of the admission of Ethiopia to the League of Nations in 1923, slavery was strongly contested internationally, despite the ongoing Ethiopian attempts to abolish it. 2 Slavery is still a highly contested terrain, as evolves around the reimagining of history in terms of a shameful past, and especially the study of Ethiopia's feudal past has overshadowed the historical analysis of slavery.
In Ethiopia, the academic study of slavery is largely disconnected from the legacies slavery, or as Markakis noted: Slavery has "not merited attention in the historiography of the Great Tradition, save as a trade factor". 3 1898-1940's', Northeast African Studies 12, no. 1 (1990: [5][6][7][8][9][10][11][12][13][14][15][16][17][18][19]. With all appreciation for the idea that the Ethiopian emperors were willing to end the slave trade and expressed their disdain for slavery, I will portrait slavery as deeply rooted in the centre-periphery relations and analyse its devastating effect on the ethnic relations in the research area. For a general appraisal on slavery, the slave trade and the history of abolution in Ethiopia s. Richard Pankhurst, Economic History of Ethiopia,   Slavery has left an indisputable imprint on the social history of western Ethiopia. As a system of subordination and control it was complex and cannot easily be described without elaborating its many facets and nuances.
In order to analyse the impact of slavery on the social system, I will look at its early manifestations in western Ethiopia, its link to modern state formation, especially in the case of the aforementioned Khojali al-Hassan, as well the impact of the international boundary system and the adjoining administrative system in the colonial Sudan.
Exposed to the predatory encroachment of the Ottoman Empire, the Mahdi state and the Ethiopian state, the people of the Blue Nile /Ethiopian escarpment were continuously affected by several forms of subjection: State formation built on state sponsored slavery; there existed household slavery, and the slave trade itself was a lucrative business. The slave trade was taxed both by regional rulers as well as by the central authorities. This created chains of middlemen who were involved in the trade.
These were individual slave hunters or organised bands. And, as we will see, tribute obligations were also paid in people.
One finds certain fluidity between slavery and other forms of unfree labour as well as the payment in humans as tributes and taxes. Many conceptual and theoretical issues arise when one engages with it. James Watson observed a "general reluctance to define slavery", and an "uneasiness" to do so. 6 To make it short, I will here use the term slavery throughout; I will specify the slave trade as a related system, but all forms of unfree labour, the exchange of humans for taxes and the use of un-free labour to large scale produce tributes (e.g. for gold mining), etc. will all here fall under a broad category of slavery.
It is highly revealing to begin with another observation made by Schuver. It shows how complex the analysis of slavery is. The system not only led to the dispersal of Studies 15, no. 1 (2011): 75-92. We will not find such tangible relations between past slavery and the current of citizenship but I will show in the later chapter that the memory of slavery affects intangible aspects of inter-ethnic relations like the trust one has to the neighbour or the state. groups but also changed whole social structures and various groups were uprooted and drawn into the system as middle-men. The remark is an indirect quotation from Schuver's guide, a local watawit trader: 'You know, I just had an idea? You see this young abid, how amiable and hardworking he is. When we get back from Komaland, I will buy him from the Agur. Then: I take him to Agoldi, I give him a cloth, some bracelets, I have him circumcised, I teach him how to fire a gun, I send him two or three times with the hunters against the abid and you will see in two or three years he will have caught a couple of more slaves for me and I will be a fool to pick up farag myself.' […] And Schuver himself goes on to comment on this statement: [A]nd no doubt the young negro, led to Agoldi and well treated by his master, initiated into the formulas of Meccaism and having learned to despise his pagan brothers, would soon make one of those malleable instruments in the hands of the Arabs, who in exchange for a light veneer of external civilization and an absolutely false price, become the persecutors of their own race and the slaves of their enemies. 7 The Background to a Slaving Zone As a "slaving zone" Western Ethiopia was "defined as the geographical area impacted by a given society's demand for slaves." 8 As such western Ethiopia was double slaving zone: both a hub for the Sudanese and Ethiopian slave trade, contributing slaves to the global demand through the Sudan (the Nile route) and Ethiopia (the Red Sea trade).
The hypothesis that I want to offer in the following: at the point of Ethiopian state expansion at the turn of the 20 th century, Ethiopian interest in slaves from the region, met already existing forms of slavery. The slaving system increasingly entered the living space of the Mao and Komo. To better grasp the slaving frontier, it is important to go back to the earlier processes of political formation, to see how the Sudanese Arabic, the Oromo and eventually the highland Ethiopian slaving systems meshed.
The formation of the sheikhdoms of Bela Shangul and the authority that was exercised over the local population had been preceded by the tight levies of tributes on the Funj meks by the Ottoman Empire. With the Turco-Egyptian take-over in the Sudan (1820), more traders, settlers and refugees came to the region of Bela Shangul.
The learned Arabs (faqi) held an ambivalent position in the area, but especially in Bela Shangul they were sought after as religious teachers by the local nobility (Funj and Berta agurs) in which they gradually integrated. 9 Based on their trade links with the Sudan and increasingly with their integration into the local societies, these watawit spurred on the local slave trade. 10 The watawit were "willing to ignore the traditional social structure to the extent of regarding the southern subject class as '"abid"", literally "slaves". A brisk market opened in "orphans", "lost children" and "abandoned wives"". 11 Obsessed with the existence of large gold deposits, the Turkish penetrated farther into the hinterland. 12 Exercising indirect rule, the Ottoman administration took tribute from the Funj rulers, to be paid in gold and slaves. Turkish rule expended and crippled traditional Funj rule "while the southern subjects were stripped of any defence against the enslavement". 13 In 1821 the mek of Fazoghli is said to have paid 2000 male slaves annually to the Ottoman Empire. 14 The most important change in the northern Sudan (north of Khartoum) through the colonial rule of the Turkiyya was the "transition from traditional pre-capitalist systems of land tenure to one based upon the concept of private property, and the 9 Spaulding notes in regard to the wandering watawit: "Although the Funj nobles sometimes allowed a limited number of the more respectable variety of merchants to enter their domains under strict supervision for the conduct of trade, they rigorously excluded the bulk of the watawit by directing their subjects to kill them on sight. The impoverished northern migrants thus tended to congregate in those parts of the south which were not under Funj control -the banks of the White Nile and the gold bearing districts of Bela Shangul" (Spaulding, The Heroic Age in Sinnār, 282). 10 In retrospect, and to explain the history of the emerging watawit, the Governor-General of the Sudan referred to the watawit as "debased Arabs", "who exercise mediaeval feudatory rights over the servile negroid population ("Berta") on the Abyssinian border" (cp. FO 141/571: J.F. Maffey, Governor General of the Sudan, League of Nations, Khartoum, 15 th April 1929). 11    . 19 Triulzi, 'Trade, Islam, and the Mahdia in Northwestern Wallaggā, Ethiopia', 63. ascendancy, and the Berta, reduced in numbers and with their power and unity broken, to have been subjected to slavery (supra note). 20 While the Berta were the first to be absorbed into a system of slavery, other groups to their south were still spared from. In the 1880s the Dutch traveller Schuver found the Koma still relatively isolated. He attests for very sporadic manifestations of the slave trade among the Koma (cp. Chapter 3, Gwama south of Fadasi). It is useful to look at Schuver's account of his guide Wad Bilal, a trader from present-day Asosa: Before leaving home, he had given me, in all sincerity, the following information and instructions.... "The commerce", he had continued, "which we carry out from time to time in Koma country, consists of the exchange of salt, cloth and white beads which we bring to barter for the wild honey and the abandoned women or orphan children of the district […]. It is rare for the Koma to consent to sell us an article against a price decided in advance. In each village they welcome us with demonstrations of joy; […] The next day ... they gather and one brings us a goat, one a skin or a gourd of honey and the chief makes us a gift of some orphan or widow. It is only when we are specially looking for a pretty girl at the request of some Sheikh of the Berta, that we buy her formally for a price agreed with her parents. 21 Wendy James, in her analysis of the same episode remarked the "interesting distinction made in this account, from the provision of slaves as tribute or gift, to their outright purchase". 22 Hence at this point there was still space for an ambivalent practice. I take this as an indicator that the slavery frontier had only started to reach the Gwama areas from the north.
During Schuver's time, we find the system of slavery and tribute already more Jote collected all the income and in turn passed it over to Shäwa. This was after Jote had allied with Shäwa. But before that he kept all the income for himself. He had made it his own property. The poor who were at all incapable of paying taxes were told to do all sort of work for Jote. 28 In the aftermath of the expansion of the Ethiopian state, the slavery frontier would move further inland taking control also of the remoter areas in borderlands. This encroachment of slavery has already been attested by Bahru Zewde: The Berta clearly formed the great bulk of the slaves exported from Ethiopia to the Sudan. Next came the Koma, followed by the Oromo, Amam, and Burun. Although most of them were already slaves or children of slave parents when sold, some of the Koma and most of the Oromo were freeborn and sold into slavery. 29 27 Sidama is a reference to the highland Ethiopian/Amhara; cp. Field notes Alessandro Triulzi (a-bg 1, p.

Slavery in Western Ethiopia
The incorporation of Benishangul accelerated the local systems of slavery. We have already seen that Menelik II requested slaves from Sheikh Khojali (s. previous Chapter). Slaves were used for domestic labour in the household of anyone who could afford them, and were also brought in great number to Addis Ababa where they worked in the houses of nobles. 30 In the research area under the administration of Khojali, subordinated landlords were used to levy a tax in children on their subjects. 31 Child tributes were levied on the local population and the lesser landlords of the region had to give their subjects's children to the courts of the superior lords. 32 The subordinate peasant population was also forced to give their children as tributes and taxes when they were unable to pay in other form.
With the centre demanding all sorts of tributes, Khojali, escalated slave raids on the Mao and Komo. He "revived the slave raiding, hunting and gold mining using slave labour". 33 The effect was that large numbers of Khojali's subjects were forced to mine gold. The use of slave labour for gold mining was already well established. In about 1900 the field mission of Major Gwynn reported from the land of Benishangul, that "all field labour is done by women and slaves." Regarding Dul, a territory tributary to Khomosha and Keili, which was severally raided by the "Abyssinains" from their headquater at Goha, Gwynn reported that Berta slaves do the gold-washing. 34 I think it is important to differentiate between the gäbbar system and slavery. The See also: "Khojele bred slave children, and when they reached maturity, he sent them as tribute to the central government in Addis Ababa" (Abdussamad H. Ahmad, 'Trading in Slaves in Ethiopia',438 was a reservoir for individual slavery, but also for state-sponsored forms of human bondage and exploitation. The gäbbar system, the "unremitting exploitation of the peasantry", "operated as the mechanism by which the ruling class appropriated the surplus of the peasants' produce." 35 Similar to the assessment of Garretson for the Maji area, 36 Bahru Zewde measured the increasingly exploitive forms of this central Ethiopian institution with its distance to the centre: While the system originated in the Abyssinian kingdom, it assumed most iniquitous dimensions in the areas conquered by Minilik in the latter part of the 19th century. As the conquered peoples of southern, western and eastern Ethiopia belonged to nationalities different from that of their conquerors, the mitigating effects of kinship or ethnic ties were lacking, and cultural oppression was coupled with economic exploitation. 37 In this way, the centre created a periphery by proxy in which the centre ruled over the watawit, who ruled over the rural subjects. This pattern can account for the situation of economic inter-dependence that evolved: the watawit rulers with their own peripheries were sustaining the demand of the centre.
Beyond this institutionalised forms of tributes in humans, there was a lively market in humans also, in which state and traffickers worked together. The Ethiopian-Sudanese borderlands have been an area for slave raids for both Ethiopian and  Selassie;1910-1919  The border had an important impact on the situation both in Sudan and Ethiopia. For the Mao and Komo the border was a possibility to evade the slave raids from the Ethiopian side of the border. 45 Paradoxically the border was also a resource for slave traders and hunters, as it enabled evading the more restrictive policies in the Sudan.
Either way, the system of slavery would continue unabated for several more years. The whole of western Ethiopia continued to be politically marginal but clearly economically important for the empire. It was a motor of trade and commerce, tribute and taxes. From an administrative point of view, the centre had a largely formalistic view of the periphery: subjects had to be ruled, the provinces had to be pacified and developed; they provided pay-offs to the central elite and in some cases had to be "mollified by concessions of one kind or another". 3 Also, from a strategic point of view, controlling and securing the borders was an important factor in regional history. Among other incidents, the late 1950s were marked by the beginning of the Sudanese Civil War and the Anya Nya I revolt (1956), which had severe effects on the border population. In its course, the influx of refugees and cross border insurgencies became a recurring pattern in the region that significantly affected the borderland population.
2 Although slavery was one of the main concerns of the new administration of Haile Selassie, the question how slavery was eventually abolished or whether it was abolished at all has to be analyzed on a case to case basis. For the peripheral areas, and from the perception of the interviews I will present later on, it seems, that the gäbbar system resembled slavery and at the same time expanded and was institutionalized in the periphery. This meant an overlap of the bureaucratization and exploitation. 3  lower Oromo chiefs and nobles aimed at raising local support for the Italian rule.
Furthermore, reforming the land tenure system and land policies was an aim of Italians. In order to get the backing of the rural population the gäbbars were given land ownership and the Italian administrators began to tax livestock and land plots. 8 In some areas of Wallaga and Gambella large-scale economic cotton production was started which did not turn out to be very successful. 9 For the Italians who approached western Ethiopia with a Fascist racial mind set, the Mao and Komo were at the bottom of the racial spectrum. While Oromo, Arabs and Amhara were seen as 'civilizable', the Nilotic groups were seen to be 'savage' and  The statement also shows that the Italians were not particularly interested in the local population. Instead they were interested in accessing their labour force and thus kept intact the rural administration through qoros. The local elite were retained in power. A statement by the British forces made on the situation in Benishangul during the War of Liberation praised the local cooperation and described the resentment of the Italian: The local inhabitants and notables have given our forces invaluable assistance by supplying agents, who have brought in accurate information; also they have supplied labour whereby roads have been cleared. 12 That "nobles" also cooperated might be an indicator for the ambivalence of interests.
Seeing the end of the Italians in sight, many nobles might have shifted their alliance again. The same report though, criticised the resistance of parts of the Khojali family against the emerging British forces.
In retrospect, the Italian presence "aggravated the political and ethnic rifts that crisscrossed the local societies and the resulting situation turned out to be more ambivalent and complex in the frontier than in other parts of Ethiopia". 13 Confronted with a new power, the ruling elite had to negotiate their loyalties. Today, in several interviews the Italian time remains relatively obscure. Portraying one's own group as openly pro-Italian is today seen as unpatriotic, but presenting them as Patriots is seen too pro-monarchist (or pro-Ethiopian for that matter), and aligned to a system of regional elitism. The historical account on Khojali can help to illuminate the empirical problems: From his earliest engagement with the Western Oromo Federation we can assume that he was ready to abandon his links with the Ethiopian crown. In 1938 Khojali regained control over Begi, was named sultan and given administrative control over the Benishangul Commissariato. 14 In 1938 he was compelled to mediate between 11 13/2010: Interview with Tongo elder, 29 September 2010. 12 WO 106/2618: "Situation Report 16 th -24 th February 1941Troops Upper Nile Area, Northern Sector". 13 'Fascist Colonialism',568. 14 Adinew Abtew, 'Political and Socio-Economic History of Asossa Wäräda,  In Gambella, according to interviews, the Komo were more inclined to fight the Italians. Nonetheless, there were balabbats like Abdu Rahim (see below for more information), a Kiring, descendant of the Kutu family who had migrated to Bure and later made it to the position of administrator in Gambella. Also, a certain Mukukullu, a balabbat of the Komo, fought against the Italians. After the return of the emperor these men were bestowed with the kabba (the "clothes of leadership") and the "people started farming for them". 20 As Charles McClellan remarked, the Italian occupation bore the chance to reexamine the cohesion of the Ethiopian empire, but he concluded that "the opportunity was not fully used, since in the aftermath of the war, Ethiopians ignored many of the war's fundamental lessons and merely replaced old mythology with new". 21 Instead of re-organizing Ethiopia along more decentralized lines, the lessons from the occupation and the local dissent against imperial rule were largely ignored. The post-war period led to a very strong (re-)centralisation of Ethiopia in the years to come. 15 "Some Berta Songs from Asossa-Bela Shangul, Wallagga", manuscript, collected and compiled by Alessandro Triulzi. 16 12/2014: Berta elder, Kushmangel, Asosa zone, 10 October 2014. 17 González-Ruibal, 'Fascist Colonialism', 568. 18 Margery Perham, The Government of Ethiopia (London: Faber and Faber, 1948), 327. 19 His wife, who nursed him in the hospital, later would claim he was poisoned. This became part of the local Berta lore, collected by A. Triulzi. "Some Berta Songs from Asossa-Bela Shangul, Wallagga", manuscript. According to this information Khojali died on 21st of March 1938 at the age of 113. 20 8/2013: Interview with Komo elder, Gambella town, 11 November 2013. 21 McClellan, 'Observations on the Ethiopian Nation, Its Nationalism, and the 57. Abolition of Slavery Slavery, the most destructive of the exploitative relations between the centre and the periphery -and the regional elite and the border people -was a matter of much international concern. It was a major preoccupation of the domestic organisation of the Ethiopia already before the Italian war. Slavery had been a main obstacle in the negotiation of Ethiopia's access to the League of Nations, and was a leading propaganda issue during the Italian preparation for war. Lastly it became an obstacle in the Emperor's post-war reorganisation attempts. 22 In the 1920s the colonial forces in neighbouring countries used the diplomacy of abolition for their own ends. Britain, which was highly opposed to Ethiopia's admission to the League of Nations, and France, which was supportive of it, lay in squabbles over the issue. France eventually wrote a memorandum to support Ethiopia's admission describing the benign character of the Ethiopian slave systems. 23 For the Ethiopian crown membership in the League, so it was thought, would "ensure Eritrean area. At the beginning of their rule the Italians set up so-called "freedom villages", in which they settled and supported freed slaves. 26 Plenty of propaganda material appeared praising the humanist mission and forecasting the wellbeing that would be bestowed upon the suppressed by the Italian presence. 27 While, as we have seen earlier, the Italians were very interested in exploiting regional discontent, their factual interest in the "natives" of the borderlands was limited and their interest in the gold reserves huge. The Italians relied on forging links with established local rulers, who were able to ensure the supply of gold during the occupation.  27 Baravelli's pamphlet is a good example for how wide-spread the slavery discussion was in that time.
Cp.: G. C. Baravelli, Das Letzte Bollwerk Der Sklaverei: Abessinien (Roma: Società Editrice di Novissima, 1935 30 Fernyhough,Serfs,Slaves and Shifta,223. The claims of the Italians to have abolished it seem oversimplified and somewhat overstated. It might be true that slavery was already heavily declining before the occupation (Cp. Fernyhough,Serfs,Slaves and Shifta,221). been or what other economic factors affected the decline of slavery has been a matter of much debate. 31 In spite of the concerns of the anti-slavery society, that after the restoration of Emperor Haile Selassie the system of slavery would return to Ethiopia, the official end of slavery is usually dated to the year 1942. A decree was issued which proclaimed heavy fines as well as imprisonment for the trade in slaves. Regarding factual slaves, the law provided for a gradual emancipation or voluntary turnover of slavery into wage labour. 32 Despite this, in the given case the deeply entrenched system of slave labour seems to have lived on. In the areas occupied by the Mao and Komo, notwithstanding the increasingly modernized administration and bureaucratic system of taxation, the lines between serfdom and slavery remind blurry, but emancipation was certainly not achieved with the official end of slavery.
The system of slavery died slowly in Ethiopia and even slower in the western parts of the empire. Thus for the Mao and Komo the legacies of slavery, marginalization and exploitation loom large, both in relation to the state and to their political neighbours.
Much of the memories are rooted in the experience of the 1960s as we shall see now, contrary to official historical claims about the end of slavery.

Provincial Administration
After the Italian occupation, the reign of Haile Selassie continued uninterrupted from 1941to 1974. Eventually, in 1974, the Emperor was deposed by a military coup.
During his reign the government was marked by a push towards modernization, the inauguration of a first constitution (1931), a revised constitution (1955), the attempt to realize a tax and land reform (1942,1966) as well as the official abolition of slavery (1942/43). The Mao and Komo were connected to the events of the centre by the rural economy and on the lower end of the social spectrum. 31 Jon R. Edwards, 'Slavery, the Slave Trade and the Economic Reorganization of Ethiopia 1916', African Economic History, no. 11 (1982: [3][4][5][6][7][8][9][10][11][12][13][14]; James C. McCann, 'Children of the House: Households and Slavery in Ethiopia, 1900-35', in The End of Slavery in Africa, ed. Suzanne Miers and Richard Roberts (Madision, WI: University of Wisconsin Press, 1988), Serfs,Slaves and Shifta,234. In terms of the development of the idea of citizenship the period signalled little progress. Inheriting a quasi-colonial legacy, the citizens of Ethiopia were subjects of the Emperor and, generally speaking, the idea of citizenship did not gain prominence until the Revolution of 1974 (s. further below). 33 The 1931 constitution was "a modern façade for the absolute monarchy that Haile Selassie was fashioning". 34 We have already seen that in terms of citizenship rights or laws little provision was made.
According to the constitution, the Emperor, "sacred", "inviolable" and "indisputable", ruled over the territory of Ethiopia, from west to east and north to south and all its inhabitants were his subjects. Achievements of modernization during this time, the introduction of a new national currency, a state bank, a postal system and development of a telephone service, etc. clearly overshadowed questions of local discontent and regional development. The imperial project of modernisation was hesitative; economic development was sluggish and unable to deliver the aspired to benefits of modernity such 33 Keller,Identity,Citizenship,and Political Conflict,22. 34  as schools or health care. The school system and the policy of linguistic homogenisation excluded the major part of the population; health care was concentrated in the two major cities. 37 Hence, what is commonly understood as the modernization of the Empire was ambiguous from the perspective of the Mao and Komo, or for any peasant from the communities of western Ethiopia for that matter.
The family of Sheikh Khojali had become successful negotiators of their own political survival and continued to administer in Benishangul between 1941 and1974. 38 Despite the extending centralization and replacing of traditional rulers with new loyal elites, the family retained its position, at least as balabbats of Asosa awrajja.
Traditional elites thus remained in locally confined positions of authority but the centralization process had reduced Benishangul to the jurisdiction of Wallaga province of which it had become a part. 39 Hence a gradual reduction of autonomy between 1941 and 1974 marks the overall centre-periphery relations during this time. 40 Like in other parts of Ethiopia, the government had "liberated itself from dependence on regional support, and in so doing has largely cut itself off from the political forces being generated at the periphery", 41 and in Benishangul and Wallaga the government ruled via a thin layer of regional authority, hardly covering the gap between centre and periphery.
Also, the Mao nobility of the family of Kutu Golja had been able to return to the area of Begi/Tongo in 1941 and continued to administer the Mao and Komo of this region as balabbat. The balabbat of Tongo was then al-mak Soso Kutu, who was later succeeded by his son Harun Soso, who was deposed by the Därg military after the Revolution of 1974 (s. further below). 42 37 Jan Hultin, 'Rebounding Nationalism: State and Ethnicity in Wollega 1968-1976', Africa 73, no. 3 (2003: 405. 38 Ibid. 42 It is worth noting that Soso was using the title al-mak, probably corresponding to mek and indicating some sort of Funj descent. This link, though, was never substantiated by any of the interviews I did with the family.
The balabbat system was in place during much of the post-conquest era. It was based on the traditional land holding system of northern Ethiopia. In the northern context it was "conceived as a collective term of identity signifying inherent right of inheritance to both community and country". 43 In the southern and western areas it was a way of co-opting local elites into the state apparatus. In general, the balabbat were recruited from the local elite, but they were subordinated to the next higher administrative position, e.g. the lowest neftegna. 44 In the periphery the balabbats were serving "… as both objects of central policies of domination and as subjects in their implementation and execution". 45 The balabbat collected taxes, reported and judged crimes, and were "compensated for his services with rights over land and labour service, and retained a share -usually a tenth -of the state tax they collected from his people". 46 The next lower administrative position -below balabbat -in western Ethiopia was known as the abba qoro. This was also introduced during the time of conquest and based on the administrative practices of the mooti or chiefdoms of the Western Oromo. The institution of the abba qoro spread in the region, acknowledging further the traditional forms of authority. It is interesting to note that also in the Mao areas under the rule of the watawit balabbats the office of the abba qorro was employed.
These abba qoro were practically similar to the chiqa shum. The abba qoro were appointed by the balabbats but recognized also a hereditary office. 47 The tasks of the abba qoro consisted of transmitting orders from the balabbat, collect taxes, and monitoring the division of land among the peasants. According to Bekele, the income of the abba qoro depended on the taxes collected. Additionally, peasants were required to do several services and provide the abba qoro with supplies. 48 Fences were prepared and repaired by the peasants, harvest, farming, as well as the cutting fire-wood was done by the people. "In short everything the balabat wants to do is done by the

Abbas Haji Gnamo, 'The Dilemma of Arsi Balabbats: A Study of Socio-Economic Position of Local
Chiefs in Southern Ethiopia, 1886', in Étude Éthiopiennes. Actes de La X Conférence Des Études Éthiopiennes, Paris, 24-28 Aout 1988  He ruled with the help of an abba qoro who was the son of Kore Genda, a certain Hossana Kore from the Makeeso clan. Muhammad al-Amin according to information from Begi ruled through an abba qoro from both a Mao and an Oromo family. He used the labour of "many slaves from the Mao and also had many Oromo slaves. The slaves received only food". 51 Another Mao elder remembered that he was taken to the landlords house (I assume from the approximate age that we talked about the late 1950s) where he had to herd goats, while his mother did grinding and cooking and his father worked as a porter. "Sometimes the landlords used the black people as oxen", he concluded. 52 The same elder substantiated these claims several years later in another interview. In this interview he took the position of an adult and reported about the general treatment of the balabbat without specifying the landlord: We were farming for the landlord. Even if we worked for him, he used our children to work. […] When the maize was ready he came to collect us to harvest for him. Without payment. We worked for them. We built the house.
Our kids looked after the cows. When the house fell down [collapsed] he collected us to rebuilt it. If you were absent he will send the police and you will pay honey or a goat as a punishment. 53 In Gambella also the family of Kutu was able to regain influence. Abdul Rahim from A: Even if we found a lot of gold we could not use it, there is one guy he is called Shami, he is the akui of that area, he was working with grazmatch Abdul Rahim. They used to buy [a] maize and salt and distribute it to us. We were searching the gold. We had to fill the quill of the big black bird with gold and bring it to them. If we didn't fill we had to go back again the next day. If you filled it, you had to take a rest, but only got the maize and salt you got. The akui was collecting in such a way. We understand from the interview that the family worked in continuation of the old feudal order for goods on a gäbbar-base: if they gained what the administrator wanted, that is gold for tax, in return they would get salt and maize. All other goods they would still have to work for themselves. It is worth more than a side remark, but the weighing of gold dust in feather quills was reported by Reginal Koettlitz upon his visit in Wallaga with the Weld Blundell mission in 1898. This confirms the idea of a traditional system of production extending well into Ethiopia's modern era: In these gold districts it is very common to see men carrying a wooden traylike pan about 2 feet long and 15 to 18 inches wide. This pan is used for gold washing. A small goat-skin bag contains quills in which the gold is kept, and other apparatus. They carry also a neat native-made balance, with weights of pebbles or seeds, fitted into a small basket, by means of which they are able to ascertain fairly accurately the value of their washings. These quills filled with gold dust, or small packets of it, or gold rings of different weights, have a known value, and pass as currency throughout these districts, and there are some markets, notably that of Nago, which go by the name of "gold markets," and are frequented by merchants desirous of exchanging their commodities for gold dust. 56 In 2014, I met the elderly woman again and she substantiated several of the claims she had made in the years before. 57 A: During the time there is man called Shame. He came and settled together with grazmatch Abdul Rahim, and sent the people for gold. When he came and settled there, they sent us for gold mining. There was no payment, but we were given one cup of salt and coffee.  ', 1972, 5. 63 Ibid., 10. 64 Ibid. 65 Jenub (also jenubi, i.e. "south" in Arabic) is refereed to both the Anya-Nya of the first Sudanese Civil War as well as to Southern Sudan Liberation Movement (SPLM). 66 Dereje Feyissa,Playing Different Games,195. The government appeared relatively helpless in controlling the arrival of rebels and fighters among the refugees. 67 When, with the emergence of the Eritrean liberation struggle in 1961-62, the Sudan government began its clandestine support of the Muslim cause, the Ethiopian government started to support the Anya-Nya in return. 68 In the course of their fighting the Any-Nya became a largely uncontrollable movement that deeply disturbed local societies. The movement claimed "corvée labour and taxes" from the local population. 69  The Komo who remained protested to the Ethiopian authorities, who gave them arms and set up a police post, together with flag poles and flags so that they could advertise whose protection they came under. 70 During many interviews, the Anya-Nya was referred to as factor of displacement. The Gwama of Laki in the Mao-Komo special wäräda testified about attacks during the time of settlement in Wadessa: The "jenub forced us to work for them. Our wives cooked for them and we were clearing the roads for them". 71

Chapter 7
The Civil War and the making of the Komo (ca. 1970-1991) Since the 1960s ideas about Marxism and socialism spread in student and intellectual circles in Ethiopia, and the liberation of the 'rural masses' from the 'yoke of feudalism' In the aftermath of the revolution in order to quell the emerging opposition, the Därg carried out a violent anti-opposition campaign known as qay shibbir, or Red Terror (1977-79). What followed was a period of violent confrontations between the military regime and emerging ethno-regional opposition movements. Events would eventually lead to the overthrow of the Därg regime in 1991.

The legacies of the Empire and the National Question
Factors leading to the revolution were multiple and cannot be discussed here in details.
The revolution uprooted the monarchy, which gradually was replaced with a military regime of socialist leaning. 1 But a driving force was certainly the student movement. 2 The Mao and Komo, like millions of other Ethiopian peasants throughout the country, were not part of any urban elite, and had no access to the milieu of the student movement, which became the most pronounced "voice of the oppressed masses" -although one might argue that the rural masses were talked for, they did not  Tiruneh, The Ethiopian Revolution 1974 Andargachew Tiruneh, The Ethiopian Revolution 1974; Bahru Zewde, The Quest for Socialist Utopia: The Ethiopian Student Movement, C. 1960-1974(Woodbridge, Suffolk: James Currey, 2014; Messay Kebede, Radicalism and Cultural Dislocation in Ethiopia, 1960-1974(Rochester, NY: University Rochester Press, 2008; Randi Rønning Balsvik, Haile Sellassie's Students: The Intellectual and Social Background to Revolution, 1952(East Lansing, MI: African Studies Center, Michigan State University, 1985. 3  But the proclaimed civil rights largely remained paper work and had no effect on the actual realities of the people. Cultural diversity was strongly suppressed throughout the country, the feudal land order continued largely unabated and the government "allocated special funds for secret agents to lurk in student gatherings and other social activities such as Idir and Senbete, self-help mutual association". 5 The Revolution was ideologically built on the idea of ending the "feudal There is a cultural argument to the demands of this early opposition, which reflects the failure of the monarchy to deal with the cultural diversity of the country. Hultin remarked an atmosphere of ethnic sensibilities led to reversing ethnic denominations such as "Galla" for Oromo, among others. 6 Amharization, similar to what Benedict Anderson labelled "official nationalism", 7 was based on territorial and national unity and integrity, and allowed no attempt to integrate the people of Ethiopia except through "assimilation". To the contrary, "it was basically an elite project, which feared all forms of popular involvement in politics and hence sought to thwart all expressions of political consciousness". 8 The question of nationalities became a larger-than-life issue both the student movement, who almost debated it like "religious sects" 9 But it also inspired the emerging ethno-regional liberation movements. In retrospect, the oversimplification of a quest for self-determination, based on clear cut identities, started a process of alleged liberation of the various oppressed ethnic groups. But in the regional liberation struggle this led to the survival of the fittest (the stronger liberation movement could make the stronger identity claim), which in return led both to the further differentiation as well as superimposition of ethnic identities. As Wendy James rightly remarked, "the struggle on the ground is not always best understood as between ethnic or religious groups, and drawing boundaries can produce as much conflict and misery as it solves […]". 10 The interlocking conflict over self-determination merged with territorial conflicts over control and supremacy, further uprooting the fragile and vulnerable minorities. The movement had its strongest base in Begemder and Gojjam, where it held several districts in 1977. Its main obstacle was its lack of mass popular support, since it was based on feudalist claims and basically an Amhara-Tigray elite movement. 11 Despite this, in western Ethiopia its most prominent member and spearhead of the anti-Därg movement was Abba Harun Soso, the balabbat of Tongo/Begi. The EDU had its base in the southern Blue Nile region in the Sudan and a head-quarter in Khartoum. 12 Harun was the first to lead an attack against the town of Begi in 1976. Despite the small scale of the attack it gained him considerable fame in the region. This fame 10 James,War and Survival in Sudan's Frontierlands,3. 11 Gunnar Hasselblatt, Schreie im Oromoland (Stuttgart: Radius-Verlag, 1980), 23. 12 Interviews with Harun Soso, 2010 basically lasts until today. 13 The Komo, whom he was supporting with weapons were his allies. 14 But the EDU lost ground in the region and the OLF could gain supremacy instead, as the main liberation movement. In the late 1970s the OLF also contacted  Ethiopia as well as those fleeing the Sudan were caught between the liberation movements and the government forces. For refugees from Wallaga, the Yabus site became a major refugee camp. Yabus camp, 700 km south-east of Khartoum, had developed from merely a few houses circled around a giant tree to a town hosting over 5,000 refugees. Especially through the work of the ORA it had a clinic, orphanages and all sorts of logistic infrastructure. 21  Äthiopien (Stuttgart: Radius-Verlag, 1990), 31. 22 Zitelmann, Nation der Oromo, 124. 23 Hasselblatt, Das geheime Lachen im Bambuswald, [48][49]. 24 Zitelmann, Nation der Oromo, 118. 25 Jason W. Clay, Sandra Steingraber, and Peter Niggli, The Spoils of Famine: Ethiopian Famine Policy and Peasant Agriculture, Cultural Survival Report 25 (Cambridge, Mass.: Cultural Survival, 1988); Jason W. Clay and Bonnie K. Holcomb, Politics and the Ethiopian Famine: 1984, Cultural Survival Report 20 (Cambridge, Mass.: Cultural Survival, 1985; Peter Niggli, Äthiopien: Deportationen Und Zwangsarbeitslager (Frankfurt/Main: Gemeinschaftswerk der Evangelischen Publizistik, 1985). villagisation program. 26 For the liberation movement these resettlement sites also posed a military threat, as resettlers were turned into militias in the new localities.
Two types of militia are recruited from among the resettles: those who work inside the camp and those who are sent into the surrounding communities.
[…] The Oromo Liberation Front has long alleged, that the highland settlers organized as militia groups are used by the Dergue as a tool to control and terrorize Oromo communities whose loyalty to the government is suspect.
[…] Militia forces from the resettlement camps seem to be involved in the process of Villagization in three ways. First, they serve as a security force for government officials orchestrating and overseeing the move. Respondents from many different areas mentioned that the party representatives, district administrators, and members of the Ministry of Agriculture arrived in their villages surrounded by armed guards "who did not speak our language". Some recognized these guard as settlers from nearby camps. Second, in some cases, the militia oversee the seizure and collectivization of crops and animals.
[…] Thus, the militia oversee the actual move, including the dismantling and rebuilding of houses. 27 Torture and extra-judicial killings became part of the consequent militarization of the periphery. Oromo, Mao and Komo peasants were all under collective suspicion by the government militia and military to support or fight with the OLF. 28 The Komo were collectively seen as supporters of the OLF and since they were very mobile in the forest and had links to the Sudan, they were most suspect and targeted by the authorities. 29 The Komo were collectively seen as supporters of the OLF and since 26 The resettlement program was designed to relocate and hence rescue victims of the famine in northern Ethiopia in the years 1985 and 1986. These IDPs were then relocated in fertile areas of south-west and south Ethiopia. 27 Sandra Steingraber, Resettlement and villagization in Wollega: Report on Refugee Testimony Collected in Sudan, May-June 1987, paper presented to the conference on Oromo Revolution in Washington, DC, August 15, 1987 28 Sandra Steingraber, Resettlement and villagization in Southwest Ethiopia. A Report Based on Refugee Testimony Collected in Sudan, May-June, 1987, (I) Villagization in a War Zone: Refugee Reports from Western Wollega, (II) What happened to the 800,000 after they got of the Trucks? An update on Resettlement, (III) Integrated Settlements in Gambella: Armed Uprisings and Government Reprisals (The Testimonies of Anuak Refugees in Khartoum); all SOAS archive 29 Sandra Steingraber, Resettlement and villagization in Southwest Ethiopia. A Report based on refugee testimony collected in Sudan, May-June, 1987, (I) Villigization in a War zone: Refugee Reports from they were very mobile in the forest and had links to the Sudan, they were most suspicious and targeted by the authorities. 30 An unjust tax system levied on the local population and the indiscriminate recruitment into the military led to crop failure and poverty in western Ethiopia during the 1980s. A "uniform tax" for the militias and the military of between 12 to 50 birr was imposed. 31 The OLF's public relations organ reported that despite crop failure the government did not send food aid to western Wallaga on allegation that aid would be distributed to the OLF fighters. "Therefore, in order to starve the guerrillas, the regime resorted to starving the whole people". 32 The report concluded that about 500,000 people were affected by the famine in Wallaga and Illubabor with the worst situation in Qellem. 33 Furthermore the government used forced labour both from among the resettles as well as the local population but to construct villages. Also, the forced work of collecting gum arabic killed many people. 34 We were ordered to cut big trees to build them houses. We cut everything down and collected the wood into one place. Those who refused were beaten or killed. They would tie his legs and tell him to walk until he falls into a ditch. Sometimes they put pepper powder in our eyes to torture us as a punishment for refusing to build houses. The whole village suffered. 35 Caught between the Ethiopian and the Sudanese unrest, Mao, Komo and Gwama were targeted by government forces, as well as rebels. Forced labour, and fear of the raids became part of the local social memory. 36 3, No. 4, 1981: p. 11 33 Ibid. 34 Hasselblatt, Das geheime Lachen im Bambuswald. 35 Steingraber (II) What happened to the 800,000 after they got of the Trucks? An update on Resettlement. 36 Gambella, Benishangul interviews 2010-12. A: The SPLA made us work for the road, and made us cut the grass and wood and building houses and making the road. Just they were sitting like monkeys, they were only proud and after we built a house and then they left and made us build another house. As many interviews recall a history of migration, the wish to settle and a life in peace looms large in many of such statements. The example above shows both the events that led to the migration, but also the wish for durable safety (a school for the children, with a metal sheet roof, according to the example above). It is interesting to note though that the respondent avoided answering the question whether he felt an Africa Watch Report (New York, NY: Africa Watch Committee, 1990), 160: "The accounts of hostage-taking and forced labor that the SPLA may be taking captives and civilians in occupied areas in ways that can be degenerated into slavery". the OLF was strongly territorial. The program of the front stated its objective was "national self-determination for the Oromo people", which would ultimately only be achieved "by waging anti-feudal, anti-colonial, anti-imperialist  had an office there). When they took a position all the big weapons were in the hands of the Komo people. And then they said, 'there shall be no death.
[…] You can go and join and make your own group'. My life was saved in this way […].
The "letter" the interviewee referred to in the statement was a pamphlet or manifesto calling for the recognition of a distinct Komo identity. Though I have not seen the document I assume from oral description that it was written in similar vein, supporting the establishment of identity-based local groups in many parts of Ethiopia. The political formations built the nucleus for many ethnic parties that emerged after the end of the civil war. I asked a local administrator of the Komo about the content of the text, and he made the following astute observation: A: Starting from today we don't want to be under anybody. We must create our own group, a community of Komo. We must have our own party. Until now we are walking with OLF, and until when will we continue with them?
We are working for the OLF, we didn't see any work done for the Komo! This is what brought the division between us. They called us black Oromos. He managed to regain his old territory as part of the federal arrangement and thus got Mao-Komo self-administration. The irony is that this territory was actually a formal feudal fiefdom, renamed as the Mao- no. 2-3 (1988): 6-13. Historical claims for territory both from the Berta and the OLF immediately led to a confrontation of the BPLF and OLF. Hence the "stage was set for Benishangul's second war". 46 The OLF's territorial ambitions and claims for an independent Oromia clearly emerged when the OLF dropped out of the transitional conference in 1991 in which the future of post-war Ethiopia was being discussed by the members of various liberation fronts. The OLF pursued its own program in western Ethiopia. The Berta, Mao and Komo call for regional self-determination was neglected, and in Wallaga and Ethiopia like elsewhere has taken a top-down approach. "Territorial division" is used by the centre both "to improve administrative efficiency", and to respond to "pressures of territorial groups seeking self-government". 2 The federal arrangement emphasises the multi-cultural doctrine of unity in diversity as the answer to past social and ethnic inequalities, conflict and inter-ethnic competition. Ethiopia is re-structured into nine federal states, all of which are organised along ethno-linguistic principles.
These new polities are referred to as kǝllǝl (Amharic: 'enclosure', 'boundary' 'confines'; hereafter region). These new regions are vital for the approach taken here.
In these territories, the state offers an unprecedented degree of self-determination. 1 The somewhat blurry terms nation, nationalities and peoples constitutes, according to the Ethiopian constitution, the citizenry of Ethiopia; cp. Article 39 ( §5): "The term "nation, nationality and people" shall mean a community having the following characteristics: People having a common culture reflecting considerable uniformity or similarity of custom, a common language, belief in a common bond and identity, and a common consciousness the majority of whom live within a common territory." 2 Ivo D. Duchacek, 'Antagonistic Cooperation: Territorial and Ethnic Communities', Publius 7, no. 4 (1977): 3-29. Regional parliaments were created with a wide array of institutional rights and responsibilities. 3 In the following I will in particular look at the emotive patterns of citizenship, expressed in inter-ethnic relations influenced by political choice and institutions. First, based on the experiences of inter-ethnic as well as state-subject relations, I will present patterns of peripheral historicity in order to highlight how the Mao and Komo understand their place in the regional social and political sphere. In the second part I will emphasize the federal border regime in order to understand the institutional defects that affect the exercise of citizenship of the Mao and Komo.

Marginalization as Lived Experience
Marginalized groups, a term usually employed to describe specialised caste-like groups in various hierarchically structured societies in Ethiopia, are the springboards here to reflect on the question: are the Mao and Komo marginalized groups? The Mao and Komo are not a caste of a given society, but rather here portrayed in a wider regional social context as citizens of Ethiopia. Still some of the aspects about marginality that have been so eloquently elaborated by Dena Freeman and Alula Pankhurst, 4 build an important super structure for reflecting on the place of the Mao and Komo in the given social context. In their study Freeman and Pankhurst concluded, that marginalisation can be analysed in "five interrelated dimensions": spatial, economic, political, social and cultural. 5 To begin with I want to compare the findings from this thesis with the concepts elaborated in their book Peripheral People in order to approach the question of marginality. 3 For a thorough analysis of the institutional set up see van der Beken, Unity in Diversity. Nonetheless I argue that there is a great gap between regional autonomy provided on paper and the actual shortcoming of decentralization. Land administration, which has become salient in recent years is undergoing constant meddling of the central government. On the problems of decentralization s. Teferi Abate Adem, '"Decentralised There, Centralised Here"' as well as Alexander The spatial dimension accounts for the shared living space between and within groups. Marginalised castes sometimes live outside the settlements or on its margins.
Sometimes they are and sometimes they are not allowed to enter the market places and are forced to display their products in a segregated area. 6 From an Oromo perspective, the Mao of Oromia and Benishangul with whom they live together are subaltern minority groups. In areas where the Mao and Komo live together with more related groups like the Berta, no patterns of segregation are to be discerned; rather a scattered and dispersed homestead pattern can be observed. 7 Where Mao live in majority Oromo villages it is quite obvious that they live rather segregated, at the margins of settlements; or even at a greater distance to the Oromo settlements: "[t]he limited surface and marginal space that they occupy is an eloquent metaphor of the Mao's place in the dominant order. 8 Especially ritual places, like the swal Gwama, are placed at a distance from the villages.
An interesting example is the Mao settlement of Ya'a. The site of Ya'a is a famous site of ecumenical worship, a Tijaniyya shrine dedicated to Sheikh Ahmad Umer. 9 The Mao (Gwama) of the region live separated several kilometres away from the main settlement, which is largely inhabited by the Oromo. On the eastern edges of Ya'a is a settlement of Ganza. The Ganza, speakers of an Omotic language, are also subsumed under the Komo lable in the region. Most Ganza actually live in the Sudan. A young Oromo who grew up in Ya'a near the Ganza settlement once told me that "we used to give the Mao our dead cows". 10 This derogatory remark evoked a very strong social distance, portraying the Ganza as scavengers.
Spatial distance nonetheless is also important from the reverse perspective: It is employed by the Mao themselves in the term Komo, which usually identifies those Mao who are living in a greater social and spatial distance to the mainstream society.
The term is usually used with some form of admiration. The Komo were "those who were lucky" and ran away. 11 6 Ibid., 2-3. 7 González-Ruibal, An Archaeology of Resistance, 339-40. 8 Ibid., 340. 9 Ishihara, 'The Life History of a Muslim Holyman: Al-Faki Ahmad Umar'. 10 This statement was made in a casual discussion in one of the local coffee houses, which are mostly frequented by town administration staff and teachers. Rarely do Mao or Komo outside the administration frequent these places. 11 14/2010: Interview with Mao elders in Ya'a, Tongo special wäräda, 2 October 2010.
The economic dimension speaks to difficulties of the minorities to possess land, and the limited access to livestock, etc. 12 Access to cattle in the lowlands is practically absent. Only in the highlands and in cases were Mao groups are living with Oromo cattle is accessible and sometimes shared for ploughing. "Our grandfathers did not have cattle; some wise Mao men could make money from honey and could buy some cattle and goat; when the Oromo came, they trained us in animal husbandry, farming and ploughing and alike; […]." 13 There are also several animal diseases like gendi (= trypanosomiasis) which affect the cattle so improvement of the farming conditions and a call for government support for rental tractors is often made.
There is no recognisable differentiation today in terms of access to land. Despite this, it seems that the Mao in Benishangul follow further economic specialisations like honey production and bamboo harvesting. They also do not, like the other groups, seem to employ sharecroppers, which might indicate an economic imbalance with other groups. In western Ethiopia land remains a contested issue in the face of land distribution of settlers and re-settlers since the Därg Socialist era, and also in light of governmental policies of land evictions, villagisation and resettlement (s. further below).
The political dimension accounts for the exclusion of the minority groups from political decision making, their rights to attend assemblies or their acceptance in courts, etc. 14 The Mao and Komo have relatively few people in key positions in the government or parliament today. Old prejudices and stereotypes, relating to their former slave status, are being reproduced in contested political campaigns. While it is often difficult for Mao children to get promoted for higher education in Oromia, the reverse is true for Oromo kids in Benishangul. There also Oromo complaining that they have to claim Mao alliance to have access to resources. This is especially so in the Mao-Komo special wäräda, where the Oromo are a titular minority. But since the Oromo are more strongly connected to the economic centre in Tongo the Mao and Komo mostly frequent the lower schools in the surrounding regions.
Social marginalization "is expressed in restrictions on social interaction, commensality, joint labour, membership of associations, burial practices and, most 12 Freeman and Pankhurst, Peripheral People, 5. 13 14/2010: Interview with Mao elders in Ya'a, Tongo special wäräda, 2 October 2010. 14 Freeman and Pankhurst, Peripheral People, 5-6. profoundly, intermarriage." 15 The most pervasive social taboo between peripheral people and the mainstream society is indeed the marriage taboo. Marriage relations follow a very complex pattern in the research region. They are directly related to local majority-minority relations. Intermarriages between the Mao, especially the Gwama and Komo are regular. From an Oromo perspective relations between Mao and Oromo are often not welcomed but they happen. Especially wealthy Mao men often do have second wives from the Oromo. Hence the intermarriage of Mao men (especially the descendants of the former Mao/Arab nobility) is not unusual. 16 But there are many stories according to which marriage was not allowed between Mao -Komo and Oromo, due to social distance and cultural prejudices.

The Inheritance of Inequality: History and Memory
The previous examples indicate a perpetuation of historical factors of inter-ethnic stratification. In the following I want to make more sense of the narratives of the Mao and Komo in relation to their place in the society they feel they inhabit. I understand this "place in the society" as the social sphere which "refers to a societal selforganization to create a common cultural landscape on which various forms of performance and public drama are staged, and through which a social bond among strangers is created and public life maintained." 17 In order to understand the social sphere I will first describe the cultural landscape based on memories displaying the trajectories of marginality of the Mao and Komo. 18 Majority-minority relations are being defined by historical experience. The are subjective and change the object matter according to the present. Memories, as a matter of fact, cannot exist without oblivion. 19 Elites and politicians need to gain control over the imagination of the citizens they represent. Elites compete with the ordinary members of groups in defining the political discourse. Social memories become contested in the attempt to build a common historical narrative.
History is a contested field in Ethiopia. It is a project of the state, or the majorities, and the subaltern hardly have a voice in the making of their own historiography. 20 In the rearranging of history, also the former masters have to re-invent the perceptions of the historical relationship with the subaltern population; often they style themselves as benevolent masters, or re-invent historical events: When slaves were captured to work on fields, one finds people today saying that these slaves 'were brought together to be narrate. An Anywaa man once told me that, "the Komo are interesting. We thought they would be gone by now. But they are still here". His point was that from an Anywaa perspective the Komo were so few they could hardly play a role in the establishment of the federal state of Ethiopia in 1991. And also as a numerical minority, many deemed them unfit for cultural survival. Intermarriage would have them dissolved in the last century is the assumption here.
The imagination of a people's place in the society depends on the perceptions of history. For parts of the Oromo, the memory of subjection, marginalization and exploitation has helped frame nationalistic discourses. Nonetheless, the politicization of a cultural identity mainly based on the common language has been the single most important pretext for forging a common Oromo identity and has overshadowed  . 23 Vansina, Oral Tradition as History, 'Ideology and Oral Traditions'. 25 For an overview of social memory studies, see Jan Assmann and John Czaplicka, 'Collective Memory and Cultural Identity', New German Critique, no. 65 (1995): 125-33;Jeffrey K. Olick, '"Collective Memory": A Memoir and Prospect', Memory Studies 1, no. 1 (2008): [23][24][25][26][27][28][29]. 26 Assmann and Czaplicka,'Collective Memory and Cultural Identity',129. the manifestation of the other, which is ethnicity. In the first of his five famous propositions about ethnicity, John Comaroff 27 conceptualized "ethnicity" as (having) "its genesis in specific historical forces, forces which are simultaneously structural and cultural". Slavery, I would argue, is exactly one such force that helped forge modern conception of the self and the other in modern Ethiopia. Its medium is the social memory. Crumley (2002) summarizes social memory as […] the means by which information is transmitted among individuals and groups and from one generation to another. Not necessarily aware that they are doing so, individuals pass on their behaviours and attitudes to others in various contexts but especially through emotional and practical ties and in relationships among generations [...] 28 Aware of the danger of over-emphasising the "cultural storage of the past", 29 I think that memories, subjective and distorted as they may be, can help understand the emergence of collective "we-group"-ideas. In this regard oral data can contribute to our understanding of the institution of slavery in the region and the awareness of the Mao and Komo about their place in current society. But the way memory influences the interaction of people needs further scrutiny and the methodological complexities of this approach should be discussed in more detail. 30 The data explored throughout the thesis introduce two broad tropes, both in need of more research: memories as a tool for a deeper understanding of the institution of slavery in western Ethiopia and second the subjective projections and emic conceptualizations of the present built on past experience. In the following I will briefly refer to the latter pattern: From a Mao and Komo perspective, this current geographical position stretching over several federal borders and the international border in Sudan is often portrayed as a lamentable obstacle to political participation. In many cases, the history of 27 Comaroff, 'Of Totemism and Ethnicity'. 28 Carole Crumley, ' Exploring Venues of Social Memory ' , in Social Memory and History: Anthropological Perspectives, ed. Maria Cattell and Jacob Climo (Walnut Creek, CA: Altamira Press, 2002), 39-52. 29  From this observation, the informants drew a couple of conclusions: They related the scattered situation of their homesteads as the main reason for a present lack of political participation. 34 Today, they said, all positions were taken by the Berta and Oromo and especially to the Mao-Komo special wäräda they do not feel they have any link. A story that was presented with much concern was this: three years back a child of their 31 2/2011: Interview with Komo elder, Gambella town, 13 August 2011. Aterfa Mustafa is the political leader of the Komo section in Gambella and the chairperson of the Gambella Peoples Unity Party. The interesting point here is how the scattered living situation is connected to the resource person, Aterfa, who connects the different Komo section today. 32 3/2010: Group interview with Bambasi-Mao, Mutsa, Bambasi wäräda, 17 September 2010. 33 Abba mooti was the horse name of Muhammad al-Hasan the last independent ruler of Fadasi. The mountain-chain that is the background for the Bambasi town today is still locally referred to as abba mooti. 34 Although at the same time the informants acknowledged that during the time of Hailie Selassie and the Därg, they had forms of political representation (I guess the reference here is the system of abba qoro and then the kebele and PA administration). community got lost and was nowhere to be found. The group tried to get assistance from the government and the police. But the search was soon to be abandoned, without any success. The group members organised a rally in Bambäsi and demanded from Bambasi to the Didessa, "was an attempt to escape a 'land tax' which had been imposed on the Bambassi area. According to the Diddesa account, those who had no money to satisfy the tax were told they they had to 'give a child' for the land they occupied. It is said that some fled to Diddesa to keep from giving up their children while others fled out of fear because they had no money nor children" (Ahland,'A Grammar of Northern Mao (Màwés Aas'è)',21. Q: Why did you not refuse?
A: How can you refuse?
There was an utter silence that resonated in the room after this answer. 36  During group discussions with Gwama speakers in Ya'a the general theme of the "backwardness" of the Mao and Komo often came up. Backwardness was usually explained via the lack of oxen for ploughing, the distance to schools, and the lack of political participation: The problem was, we are very backward; only a few wise people are breeding cattle, since we have been ignorant for long; we are still ignorant; The landlords themselves used the black people as an ox; I [the discussant himself] was taken to somebody's house to herd the goats; my mother worked and grained in the house, while my father was working as a carrier; that work was forced and not paid; The Komo are now deep in the bush, because they ran away from slavery; those who are lucky ran away; this also led to a scattered village system; the villages are scattered because the people were hiding in the bush […]. 38 But there is a certain admiration among the Komo for those who ran away and lived in the wilderness, free, untamed. The circle of marginalization was powerfully captured in the testimony of one Mao elder: We were farming for the landlord. Even if we work for him, he also used our children to work for him. Only his sons will go to school. This is why we didn't change ourselves until now. [...] when his kids finished school they came to administer us.
From this brief excursion into oral history, we can assume that the Mao and Komo lived in a continuum of subjection, enslavement and flight. With its decline, regional power relations were, again, expressed through servitude, e.g., by impose tribute and tax collection and political incapacity. Federalism (Leiden-Boston: Brill, 2011), 127. 41 Barth,Ethnic Groups and Boundaries,3. 42 Abbink,'Ethnicity and Constitutionalism in Contemporary Ethiopia',160. and most widely discussed result of this territorial arrangement is Article 39( §4) of the Ethiopian constitution, which grants the nations, nationalities and peoples of Ethiopia the "right to self-determination up to secession". 43 Federalism (since 1994) was meant to redraw the map of Ethiopia along ethnolinguistic lines creating a quasi -(ethnic) federation. This was seen as the final answer to the national question, the emancipation of the peoples of Ethiopia from former feudalimperial overrule (ca. , as well as more immediate forms of national oppression during the Socialist Därg period .
The territorial re-organization of Ethiopia led to the demarcation of federal boundaries framing citizenship along ethno-linguistic criteria. Civil rights in Ethiopia are tied to group rights. The belonging to a specific group is emphasised in the constitution. The preamble of the federal constitution opens with the following remark: "We, the Nations, Nationalities and Peoples of Ethiopia: strongly committed, in full and free exercise of our right to self-determination, to building a political community founded on the rule of law and capable of ensuring a lasting peace, guaranteeing a democratic order, and advancing our economic and social development." Article 39 is the main article concerned with group rights: Rights of Nations, Nationalities, and Peoples 1) Every Nation, Nationality and People in Ethiopia has an unconditional right to self-determination, including the right to secession.
2) Every Nation, Nationality and People in Ethiopia has the right to speak, to write and to develop its own language; to express, to develop and to promote its culture; and to preserve its history. 3) Every Nation, Nationality and People in Ethiopia has the right to a full measure of self-government which includes the right to establish institutions of government in the territory that it inhabits and to equitable representation in state and Federal governments. 44 The emphasis in defining groups is on language and territory. Hence a specific Ethiopian form of regional citizenship seems to be highlighted in the political framework. Membership in the political territory is further defined by the regional constitutions. They define both who belongs to the indigenous political community of the federal state and define the constitutional capacity to accommodate minorities.
Theoretically, in the federal regions (i.e. killil), like in national territories (states), the connection between territory and citizenship also applies. "Citizenship" according to its spatial understanding, "is territorial and bound to the dimensions of a particular geographical unit". 45 But there are complex challenges to territorial autonomy, 46 and the successful convergence of the "territorial matrix of the federation into separate ethnically defined territorial units […] creating ethnically pure sub-national units" is a difficult, if not an impractical undertaking. 47 In Ethiopia as elsewhere, the congruent overlap of territory and (ethnic) identity, in a federal arrangement, is always an abstraction of multi-cultural realities. Accordingly, conflicts over ownership of territory have become a noticeable feature of the negotiations of political power in the new regions. The federal boundaries became contested spaces between regional states.
As we have seen in the case of the Berta Oromo confrontation over Begi, federal boundaries are the product of majority groups' claims over territory. The dispersed minorities on the other hand have to negotiate their political agency across these boundaries, which, in fact limits their effective representation in the current federal framework. 44 Paragraph four provides for the right to self-determination up to secession. This is yet another pattern of the interplay of ethnicity and development in Ethiopia and has to be omitted here. 45 Richard Yarwood, Citizenship, Key Ideas in Geography (London -New York: Routledge, 2013), 18. 46 For an interesting critique of the concept of territorial autonomy and ethnic conflicts, see Shaheen Mozaffar and James R. Scarrit, 'Why Territorial Autonomy Is Not a Viable Option for Managing Ethnic Conflict in African Plural Societies ', in Identity and Territorial Autonomy in Plural Societies, ed. Ramón Máiz and Safran William (New York: Routledge, 2000), 230-53. 47 Yonatan Tesfaye Fessha and van der Beken, 'Ethnic Federalism and Internal Minorities'.
The federal approach has compartmentalized political agency, which has brought out enormous competition among groups, leading to further sub-division of territory.
This approach creates the illusion of mono-ethnic territories and it sparked continuous demands for separate administrations, which -in some case -led to the splitting of territories into ethnically defined zones, wärädas (districts) and liyyu wärädas (special districts). 48 Regional integration is presented as the key to national unity. Federalism has undoubtedly changed the integration of groups into the Ethiopian state, opening ways for notoriously marginalised groups to claim and gain citizenship rights.
Regional structures of self-rule have, in some cases, led to more inclusive policies.
Federal borders formalise decentralisation and are expected to foster regional autonomy. These borders are a creation of the central and regional political elites, in order to organise, systematise and administer; constructed by the political planners, they are enforced by police and town staff or, at times, displayed in border control, checkpoints and sometimes in violent clashes between groups of various actors over territory, and their maintenance requires administrative posts. The exercise of border control is diametrically opposed to the frontier process. The idea of unsettled space, so central to the frontier, is always a construction which inevitably leads to the dichotomy of "first comers" and "late comers" and the polarization of autochthony. 49 The territorialisation of group and kin, the demarcation of territory and the making of borders and boundaries, contradicts the internal frontier. Boundaries, international or federal, of course challenge movement, migration and cross-border settlement; hence challenge citizenship and territorial belonging. "First comers" and "late comers" become constitutionally enshrined, encircled by the regional borders. Their demarcation via the federal states has put a halt to an ongoing process of ethnic encounters and changing relations and instead institutionalizes a majority-minority situation at a given time, creating competition among groups for institutional resources often based on autochthony claims or demographic numbers. How is the demographic and ethnic situation for the Mao and Komo, who live across several federal and inter-regional boundaries? 48 Abbink, ' 604. 49 Donham,'On Being First';Kopytoff,The African Frontier. The population numbers as given in the "Summary and Statistical Report of 1994Population and Housing Census" (Addis Ababa, 1996   Territorial control over the new federal states has become vital for the elites administering these territories. Simply put, the more territory these regions encompass, the more administrative positions they materialise. Asnake analysed the federal conflict scenarios based on territorial arguments. 55 He claims that the federal border regime leads to conflicts over administrative structures and ethnic territories, 51 Southall, 'The Illusion of Tribe'. 52 Asnake Kefale, 'Federal Restructuring in Ethiopia: Renegotiating Identity and Borders along the Oromo-Somali Ethnic Frontiers', Development and Change 41, no. 4 (2010): 615-35. 53 Michiel Baud and Willem Van Schendel, 'Toward a Comparative History of Borderlands', Journal of World History 8, no. 2 (1997): 214-15. 54 Joel S. Migdal, 'Mental Maps and Virtual Checkpoints. Struggles to Construct and Maintain State and Social Boundaries ' , in Boundaries and Belonging. States and Societies in the Struggle to Shape Identities and Local Practice, ed. Joel S. Migdal (Cambridge: Cambridge University Press, 2004), 3-26. 55 Asnake Kefale, 'Federalism: Some Trends of Ethnic Conflict and Their Management in Ethiopia'. border conflicts as well as to settler conflicts. Abbink also argued, that "majority of conflicts now dubbed 'ethnic' in Ethiopia are about boundaries between territorialized ethnic groups. Fights about identity are being waged in order to establish the borders of districts and zones, and the 'identity' professed by local people is the deciding element." 56 The federal borders create new borderlands between the neighbouring regions.
These new domestic borderlands have become vital venues of conflicting identity discourses. Conflicts occur in the shadow of population movements, seasonal migration as well as pastoral mobility. 57 While boundary conflicts are arguably a salient feature of the new federal arrangement, in some cases, due to provisions of tenure security, the boundaries have also helped to tone down some conflicts. 58 Whether boundaries are used to stir political conflict or ease tension is a matter of political manipulation by political elites and the exploitation of the territorial demands of the population.
While a higher political level -the state -is certainly needed to decentralize power, the reality of the federal border is an exercise of the local/regional administrative power. A border is best visualized by border checkpoints. In Ethiopia, these domestic border controls look more or less the same: military and federal police check the passengers of public transport, who commute between towns across borders. Another way to ascertain the border is by administrative action: in areas of mobility for important places, like wells, or the blocking of roads, the expelling of settler groups, etc., is carried out by local authorities, police and militia. Such incidences, too, are a display of boundary control. 59 Two types of new minorities have resulted from demarcating the regional states.
The first category would comprise recent groups of settlers and re-settled people (e.g., from the Därg time) now encircled by new boundaries. Also labour migrants, who moved to other areas than their region of origin. There might also be spontaneous 56 Abbink,'Ethnicity and Conflict Generation in Ethiopia',397. 57 Ibid., . The appendix of this paper lists numerous conflict cases (up to 2006)  Office, […] it was decided that they come to be administered under the authority of our regional administration.
Although, […] in these kebeles remarkable development activities were carried out, the maps prepared and distributed by your office as well as by other [relevant government offices] indicate that the four kebeles are included 60 There have been several conflicts around new settler minorities, which have experienced political problems and entered into violent conflicts, in western Ethiopia Gambella, Wollaga, Benishangul, In such conflicts sentiments over autochthony are activated by the regional elites. (cp. Abbink, ' 604;also Dereje Feyissa,'The Experience of Gambella Regional State'. under the authority of the Oromia National Regional State administration.

This situation is not tolerable. 61
Based on the settlement pattern of the Komo, the authorities in Benishangul made a claim to administer the mentioned localities. The administration of Oromia ignored the claim and as a result the Mapping Authority issued maps showing the concerned kebeles as belonging to Oromia. Locally this case is often cited as an example for side lining the territorial claims of the Komo as a minority group.
Boundary making in the region has been a long process and it is far from being finalised. This is directly related to the problem of movement and territorialization in Ethiopia. The letter translated above shows vividly how difficult this process is. The antagonistic administrations use the settlement patterns of people to claim territory, but at the same time, real-time migration as well as shifting identification, makes it almost impossible to claim a specific territory at a given time.
We have already seen how the region of Benishangul was drawn into consecutive territorial conflicts after the end of the civil war in 1991. Especially the area of Begi became a fighting ground for Berta and Oromo over territory. Based on their respective historical relocation both Berta and Oromo claimed the territory of Begi as part of the newly established federal states. The territorial conflict and the consecutive referendum over the status of Begi show the inherent problems of the quest for territorial expansion in the federal system. The EPRDF intervened and contained the OLF in favour of the claims of the BPLF. After the OLF was defeated and replaced by the Oromo Peoples Democratic Organization (OPDO), the conflict over Begi continued. A letter of complaint that was sent by the regional administrator of Benishangul (at that time Abderrahim Hassen, a Berta from Bambäsi) shows the level of discontent on the side of Benishangul and also condemns the OPDO push for the integration of Begi into the territory of Oromia. 61 The president of Benishangul Gumuz to the Mapping Authority (4/2/1999) ጉዳዩ፥ ቤኒሻንጉል ጉሙዝ ብሄራዊ ክልል ዉስጥ የሚገኑት 4 ዎረዳዎች በካርታ ሥራ ያለኣግባብ ወደ ኦሮሚያ ብሄርራዊ ክልል ስለመጠቃለላቸው በተመለከተ (Subject: Concerning the wrong delamination of the 4 districts of Benishangul Gumuz National Regional State into Oromia Region National Regional State); translated from the Amharic original.
To solve the issue of border demarcation, democratically and peacefully, between the two neighbouring regions (Region 6 and Region 4), the local elders and representatives from the two regions have made repeated efforts to come together for a common discussion. It however was unsuccessful to reach in agreement due to the pressure that was exerted by Region 4.
Region 4 has taken steps and decisions of integrating districts of Begi, Manasibu, Gidami, Anfillo and Bäle and also deny the identity of nations and nationalities (of these regions) and claim they do not inhabit the aforementioned districts (…) Among those undemocratic acts, which are taken against our people, (…) are the denial of their democratic rights by the armed OPDO governing organs, restricting the movement of the people, without the identity card of Oromia, in the region , restricting our cadres in Begi wäräda to carryout successfully activities, as well as efforts to ban the organization of Mao-Komo. (…).
Eventually a referendum was called in which the people were asked to vote for the Benishangul (the voters' symbol was the ostrich) or for Oromia (the symbol for it was the odaa the sycamore tree, a mythical symbol for the Oromo identity). Involved administrators told me that they are very sure they were cheated out of the referendum and that the 'ostrich would have won'. 62 But that eventually the office of the prime minister sent a decision which stated that Begi would become part of Oromia. Hence a border was established, cutting through a common cultural territory of the Mao and Komo. Oromia and Benishangul-Gumuz undertook opposing measures to accommodate the Mao and Komo minorities in this area. Thus, the border became a decisive factor in the evolving majority-minority relations.
Citizenship and group rights are expressed along ethnic lines. Multi-culturalism is uttered by emphasizing the cultural differences of groups to idealize "unity in diversity" and portray federalism as a "safeguard to stability". Despite this, citizenship rights are largely defined by territorial arrangements. The Mao and Komo as a result of these constitutional facts were awarded an ethnic territory, the Mao-Komo special wäräda. This ethnic territory is also known as Tongo special wäräda, after its capital. Both the capital and the whole special wäräda are peripheral and marginalized.
Infrastructure is still relatively poor and much of the rural economy is based on traditional subsistence and a web of local markets throughout the wäräda. The economy of the population is largely based on traditional rain-fed agriculture, via hoe and shifting cultivation. Oromo and Amhara use ox plough agriculture in some areas, and the subletting of cattle for agriculture seems to become more and more common.
Furthermore, coffee-growing, bamboo-harvesting and to a lesser degree hunting and fishing, mostly carried out by the Mao and Komo as well as honey production, are important additions to the economy. The main market is in Tongo. The population of 63 Though it should be said that I was not able to do any interviews in the camp. Access had been denied to me in the first instance and only after several institutional hurdles I was able to convince the authorities that I needed to trespass into the camp in order to reach my research area west of Tongo, places like Mimi Akobo, Penshuba, Kesser and alike. I was granted permission to move through the camp, but I was told to make no contact with the refugees or do any interviews. I acceded to this request from the local authorities. the wäräda in 2008 was said to be 26,586. 64 The road system was an important issue and the government was questioned by the local population on the ability to deliver infrastructure. From total of 299 km. of roads, only 35 are all-weather roads. Kebeles remain largely inaccessible during the rainy season. 65 Major road construction carried out during the research period in 2012-14 was on the road connecting Tongo and Begi and on a road in the direction of Yangu.
Administrators often invented the existence of " Mao-Komo", or even "Maokomo" population. The wäräda has 18 primary schools (below grade eight) and six from grade 1 to 8, and thirteen only from 1-4, but there is only one secondary school (grades 9 to 10). There is a very high coverage of primary schooling, but only 5% of the approx. 2300 children at high school age can attend high schools. There are also 10 Alternative Basic Education Centres (ABE), with 740 students. Most schools are relatively basic wood/mud structures with no adequate furniture or water supply. 66 Created under the auspices of the Mao-Komo Unity Party, led by the son of abba Harun, the territory became an administrative entity, directly linked to the regional government in Asosa. Effectively this means it operates like a zone, but is a wäräda due to the small number of people it represents. This procedure is coherent with other such territorial arrangements in which minorities were incorporated in the territorial administrative setup. Territory in Ethiopia has become a spatial pattern of administrative self-determination. Hence most killils are further sub-divided in zones, wärädas, and kebeles. Usually the zones are ethnically defined and build an "autonomous tier of local government with constitutionally mandated elected councils and executive administration". 67 The official rationale behind such a territorial arrangement is the special recognition that minorities need in order to be politically meaningful. Hence the constitution of Benishangul-Gumuz stipulates that the Mao and Komo shall be given special recognition. Van  From what we have seen, Tongo poses a couple of interesting questions as to the accommodation of minorities by means of an "ethnic" territory. Historical developments challenge the tribal-model-perspective that the government has on ethnicity. They also challenge the government's approach to decentralise and advocate for self-rule of ethnic minorities.
Demographically the population of the Mao-Komo  c. Language Planning in Three Regional States Compared One way of assessing the effects of borders and boundaries is by looking at the changing administration across the border. This can be seen in the linguistic landscape, changing patterns of street signs etc. Language policies and the decisions thereupon are among the most vital rights of the regional-federal administrations.
These regional administrations can decide on the use of language and enforce it in 'mother tongue education' on the lower educational levels. For the boundaries between regional states this means that by crossing their border, the administrative language of the region changes. Language thus becomes one salient marker of the boundary between regions. Language policies are one way for the state to include or exclude. 69 The official governmental fixation on primordial markers of identity has dramatically altered language policies in Ethiopia since 1991. The history of Amharic and its connection to the nation-state project, ethnic over-rule and cultural suppression were the background for a progressive and multilingual policy choice. 70 For a long time Ethiopian identity was largely bound to the ability to speak the national language Amharic.
Since the introduction of federalism the cultural and language policies provided the background for the expansion of identity politics in Ethiopia. The regime sought to delegate educational policies and the question of the language of instruction to the regional level. Not only did the regional states decide on their choice of the working language (in Gambella and Benishangul-Gumuz Amharic was chosen in order to avoid conflicts; but in Oromia afaan Oromo became the working language). Also, the language of instruction was being decided on and worked out by regional educational bureaus. 71 Today, ethnic groups have the right to mother tongue education or the language of their choice. Mother tongue education hence depends on the capacity of the regions. Furthermore, English and Amharic are being taught from grade 1 and 3 respectively, as subjects.
Regional disparities in education were at the heart of the grief informing regional resistance that eventually toppled the Därg government. 72 In any society, and especially in multi-cultural states, language policies are a parameter to evaluate the citizen-state relationship, since language policies not only demand high political costs, but are also an important factor for 'identity planning' and citizenship expansion.
Issues around the politicization of languages "emerge as vital and contested in the context of national and sub-national appeals for meaningful citizenship". 73 We have seen in the context of ethnized language conflicts denial to speak once language has led to the fiercest nationalistic resentments of the political centre, like in the case of the Oromo. 74 "Denial of the right to speak one's mother tongue, the language of home 70  (…) is often experienced as the most undemocratic and autocratic of all measures passed by the state". 75 Furthermore local language as a marker of identity is vital for the cultural self-esteem of a people and local language vitality even in the face of linguistic hegemony cannot be overestimated. 76 The current "National Cultural Policy" emphasizes in this regard the importance of language for culture, and language development for the conceptualization of "cultural policies": (…) Ensuring that the languages, heritage, history, fine arts, handicrafts, oral literature, customs, beliefs and other cultural elements of the nations, nationalities and peoples of Ethiopia receive equal recognition, respect and chance to development;" (…) The peoples of the country shall be provided with professional assistance in deciding the languages of instruction, mass communication and for official use at the federal, regional, zonal and when necessary, at district levels. 77 The educational policy is implemented by the Regional Bureaus of Education, and as a country strategy the policy is overseen by the Federal Ministry of Education. The administrative cost of this policy is relatively high and involved the preparation and/or translation of textbooks, language descriptions and standardization. The zealous policy has led to substantial economic cost, "and has contributed to the overall regional disparities, without necessarily lessening the political conflict over ethnicity". 78 One argument that is surprising only at first sight is that the policy will further marginalize nationalities and hinder the students in connecting to the economic development of the country as a whole. 79 80 Cohen,'Language and Ethnic Boundaries',190. are being employed. In Gambella for instance English is given additional emphasis while otherwise Anywaa and Nuer is used for primary education and Amharic is used in later stages. 81 In Benishangul-Gumuz the Ministry of Education and the Summer Institute of Linguistics are working closely to develop the regional languages for primary education. Since 2007 several pilot schools exist in which Berta, Shinasha and Gumuz were implemented for mother tongue education from grade one to four. 80 more schools were selected for pilot programs but due to financial and logistical constraints as well as a lack of trained teachers and insufficient teacher training, these developments go quite slow. 82 In comparing some observations on the development of the Mao and Komo languages, I hope to shed some light on the complex ambiguities of the policy of accommodating these minorities. I am particularly interested in two aspects of the language policy: -First, the politics of language are an important aspect of identity politics and enable the minority elites to politicise and capitalise minority status.
-The second issue is more complex and reveals quite a bit about the working of federalism as it visualizes ethnic-territorial incongruence through the linguistic landscape.
From the perspective of dispersed minorities like Mao and Komo this may give some interesting insights on the ability of minority groups to actually access equal possibilities of citizenship expansion. This is essentially so because groups are never homogenous.
In all interviews, and in all related publications both from the research area and elsewhere, the respondents clearly sympathised with the idea that their language should potentially be developed, or become a tool for education. This argument speaks about cultural recognition. If the language is being studied and eventually used for school it will help preserve group identity, or refresh it at least. People want to see their language being promoted and their children to speak it. Where there is group agency, sustained by politicians and supported by an ethnic territory, the political elite are very interested in the development of the local language.
Based on the cultural argument, they can then claim progress in terms of identity politics. The formation of a political elite is important. Elites within the minorities use language to support their argument of differentiation. In such a reading, "linguistic minorities stand a better chance of survival if they codify their language in an alphabet […] only then can they as a group present their case convincingly in national and international politics". 83 From the perspective of the Gwama and Komo two developments are worth mentioning. Apart from the work invested in the development of Gwama and Komo both by the SIL and the Norwegian Mission Society, two local politicians have been involved in their own projects of putting their respective language in the picture. In Gambella a politician and community leader has issued a school book for primary school teaching basic vocabulary of the Komo language. In Asosa, a local Mao politician and community leader has issued a book on the cultural history of the Mao and Komo (although the focus is clearly on the Gwama culture) and attached to it, in the second part of the book, are a dictionary and a list of proverbs.
The subtext of this development, the placing of the respective community in the realm of multicultural politics is quite obvious. The second example reveals this aspect quite well, as Gwama cultural reader/dictionary, opens with the statement, that "behind every nation there is unique history". Especially the latter example is an interesting reminder of the fact how identity and ethnicity became politicized in contemporary Ethiopia. It is also a reminder of the progress the political discourse has made where ethnic groups are being "assumed to be a reflection of a primordial group character, of a group as a "natural" unit in which people of a multi-ethnic state have to live." 84 More than twenty years since the introduction of "ethnic federalism", the question whether the political system actually captures the realities of identity or whether it rather creates new identities is still valid. 85  highly important in the research area, as it constitutes the largest of the Mao languages.
As we have seen in the introductory part of this thesis, Gwama is the mother tongue for people referred to as Mao (both in Benishangul and Oromia), as well as of people who refer to themselves as sit shwala (in Oromia) and for people who are referred to as Komo (the lowland Gwama) and for people who refer to themselves as Gwama. This complex and ambiguous situation is, when looked at from a political perspective, highly revealing for the understanding of the ambiguities of the federal system.
All Gwama-speakers are bi-or multilingual. Generally speaking, ttwa gwama is a matter of the house and of culture and tradition. The lingua franca for most Ethiopian Gwama is Afaan Oromo (the Oromo language), followed by Arabic rather than Amharic, and a general mutual understanding in taa komo (the Komo language); in Gambella most Gwama speak Komo and Anywaa. The patterns of multilingualism are a matter of geography, language distribution as well as power.
Both Benishangul-Gumuz and Gambella Regions employ the idea of "Nations, Nationalities and Peoples" and both regional constitutions provide in Article 6 the right of all these "Nations, Nationalities and Peoples" to use their own language. In Article 39 the regional constitutions also elaborate the right to self-administration. In Gambella ttwa gwama is spoken, but not recognized as one of the regional languages.
The Gwama speakers I met live "hidden" within the Komo population. Illustrative is the case of one civil servant at the Gambella Bureau of Education, who serves as one of the spokesmen of the Komo as a whole. He is from a Gwama clan but was drawn into the civil war with the OLF. Hence in the course of the civil war more and more Gwama came to Gambella as fighters.
In cooperation with the Summer Institute of Linguistics the Regional Bureau of Education of Benishangul-Gumuz is now developing language materials. In Benishangul-Gumuz the constitution gives special emphasis to the Mao and Komo. 86 This special consideration is reflected in the existence of the Mao Komo special wäräda. With regard to the questions of language and identity this political territory is again relatively problematic. Also, here the Gwama are a "hidden" group in the area.
Although Gwama is summoned under the label Mao together with several other (mostly Omotic) speaking groups, the regional government has issued a contract with the Summer Institute of Linguistics to develop the regional language for mother tongue education in Gwama. Berta and Gumuz are currently used in school. Works on Gwama (Mao) and Komo have been pending due to the lack of ability to send people into the field and gather the necessary data. Also, collaboration with the Gwama and Komo has been difficult since both groups hardly have enough trained youth available for teacher trainings. At the same time the other Mao languages like Bambasi Mao, which is also territorially confined to Benishangul-Gumuz, does not receive special consideration.
Oromia has a much more ambiguous stand towards minorities. The preamble makes it very clear that the regional constitution is defined for the yäOromo hizb, for the Nation of the Oromo. Similarly, Article 6 and 39 define the right to exercise the Oromo language for the Oromo people and the right to self-administration for the Oromo people. Despite the fact, that the constitution acknowledges that Oromia is inhabited by Oromo and other people (Article 2/2). Art 39 (6) of the Oromia Constitution stipulates that the people of the Oromo nation shall be construed as meaning "those people who speak the Oromo language, who believe in their common Oromo identity, who share a large measure of a common culture as Oromo and who predominantly inhabit a contiguous territory of the region." Herewith the Oromia Constitution recognizes both subjective and objective markers of the concept of 'Oromo people'. "As objective markers language, culture and identifiable predominantly contiguous territory are required, whereas believing in a common identity of Oromo is provided as subjective marker". 87 This also leaves little space for the inclusion of Oromo speaking minorities.
Hence it becomes very difficult for "indigenous" minorities to claim minority recognition, which is in practice being denied, as they are seen as an extension of the Oromo ('non-pure Oromo') and as they are speaking Oromo are assumed to become members of the wider Oromo society.
The language of instruction in schools is afaan Oromo and there is little political will to incorporate measures to support minority languages. We might think here of national minorities like the Mao (in the sense of an "indigenous population" but also immigrant groups from other regional states who live in Oromia).
The following case illustrates the problem for the maintenance of twaa gwama in twaa gwama (the family referred to themselves as Mao/sit shwala). In school, I was told, she was exceptionally good, but she would not get the promotion to college, which was explained by an unfair distribution of promotions between Oromo and Mao kids. I made sure at the school that her marks were as good as told. In the process of identifying members for a group of teachers for mother tongue education for the work of SIL, she was offered a position. But SIL was working in Benishangul and it was seen as too difficult a political manoeuvre to employ someone from Oromia through the Benishangul Bureau of Education. It was equally problematic to employ her, since in Benishangul the working language is Amharic, which she was trained in only for a relatively short time. Later on, when she had been invited to join a group of language informants, again by SIL, it dawned on the organizers that her native language skills were not sufficient and she was sent home.
The maintenance of federal borders is a matter of administrative choice as well as of the will of political actors. The study of (domestic) border cultures and people can help reveal the state of minority integration in Ethiopia as a whole. In the case of the Gwama, the border perspective reveals aspects about their place in local society, and in particular, in the different regional states. This can help to make sense of the historical experience of 'region-making' from a minority perspective. The idea of language death is not a matter of biological superiority but of social power. As much as the question of minorities is a matter of historical perception and political construction, so is the planning of educational languages and their politicization. As S. May said: Situating languages and language loss within the wider context of social and political power leads to a further recognition: that biological metaphors understate or simply ignore, the historical, social and political constructedness (emphasis in the original) of language (supra note). 88 The argument here is that administrative choice, in this case, the choice of language, is expressed by the boundary, which in return substantiates the language policy. Benishangul-Gumuz opted for a different language policy than Oromia, and so did Gambella. In Oromia, the language of instruction is Afaan Oromo and there is little inclination to support any minority languages. Whether this happens or not, is a matter of complex decisions and depends on the political will of the policy makers.
While in Benishangul-Gumuz minority languages are being promoted (Berta, Gumuz, Shinasha and Gwama are currently being made available for primary education use), Oromia has not made similar political provisions, thus flouting the Ethiopian federal constitution.
In general, borders signify variations of cultural and socio-cultural identification among the border population. I once asked a young politician from Tongo (a Gwama) about the Gwama of Begi. He said they were "other people". A term usually applied in this regard is "Kuro" (self-designation for the Begi (Koman) Mao), but also used from the other side of the border as a designation for the allegedly Omotic Mao. I showed the same person a video of a divination process I had recorded in Begi, which he admitted was Gwama performance. The hypothesis here is that the 18 years since the border between Begi and Oromia has been created have produced a commonly accepted difference, despite the same language and culture. In both cases the border poses questions as to representation and agency. The Kuro-Gwama  I argue here that this veering away of the Gwama into Gwama and Kuro is based on the political agency in Benishangul and the lack of it in Oromia. It is based on how the administrators understand agency and which meaning they give to the question of citizenship and autochthony. Hence, I would also argue that a domestic regional border, which came into existence by a referendum in 1995, had a measurable impact on the political culture as well as on the meaning people attach to language and identity. In this sense, the border between Oromia and Benishangul would be not only a marker of administration but also a separator of a formerly "united" group. I do not want to draw too many conclusions here yet. Nonetheless, it might be safe to assume that the Gwama language will now increasingly be studied and promoted, but this will divide the Benishangul Mao and Oromo Mao further. It will probably lead to recognition of the actual status and distribution of Komo compared to Gwama, hence it might lead to political changes in Tongo in regard to the use and understanding of the terms "Mao" and "Komo" in favour of "Gwama". Local politicians are already anticipating a "  This, in turn, will affect the relation of the Gwama and Komo in Gambella. Both Gwama and Komo negotiate their political demands through Komo identity. If there will be a Gwama identity publicized in Tongo that would raise demands for a similar recognition in Gambella. Pokung, the primary Komo settlement in Gambella, is a villagised complex with a clinic and a school and only a few hundred households. The kebele belongs to the Itang wäräda and the settlement is about two hours' walk away from Itang itself.
Additionally, a relatively large number of agricultural extension workers and clinic staff lives and works in the settlement. It was recently made accessible for cars during the dry season. Even the three-wheeled Bajaj 95 can now move along a dry-weather road that connects Itang with Pokung. I visited Pokung several times in the past years. The most decisive changes were the school building at the entrance of the village, which turned from a grass-thatched wooden construction into a stone building, the sudden appearance of a small bar (a local was buying caskets of beer in Itang and serving them to the local community in his house), and the increasing appearance of agricultural workers, on the cleared countryside, where signs of ownership marked the boundaries between different leased lands, opened for commercial agriculture. There seems to be no large-scale investment by foreign companies around Pokung, but the land is leased 93 Jon Abbink, '"Land to the Foreigners": Economic, Legal, and Socio-Cultural Aspects of New Land Acquisition Schemes in Ethiopia', Journal of Contemporary African Studies 29, no. 4 (2011): . 94 Markakis,The Last Two Frontiers. 95 Bajaj are Indian-produced motorized rickshaws that have become very popular in Ethiopia. by domestic investors from northern Ethiopia. The local population finds short-term employment on the commercial farms on a day-to-day basis.
We are living together. And even I get something from them. Also, sometime we get some money for some work, to buy a coffee. We are working and get some payment, so we are happy. .96 Pokung is a villagised site. The government has brought people together in the settlement to support them with a health centre and a school. The government also provides food aid to the community and at the same time around the settlement the forests are being destroyed leaving the Komo with approximately one hectare for their household production. Destruction of the forest has severe effects on traditional economic activities of the Komo, like honey collecting.
A: The difference is, a long time ago, we used to go hunting but today it is so difficult. Since the investor came, we cannot farm honey. Because they have destroyed all the trees. The place for honey is now so far from here. They cut down all the trees. This is our problem. Q: How is the balance between payment (day labour on the farms), honey collecting and hunting?
A: This is not the same. Hunting has a lot of money. If you work one week with the investor, you get small one you can finish it with buying soap. One container of honey is five hundred birr. You can buy what you want. 97 Disregarding the cultural changes, the government is employing agriculture extension workers in order to enhance farming and foster economic self-sustainability. These people, usually very young graduated from agricultural colleges live with the local 96 Interview, Gwama elder, Pokung, 24.09.2014 97 Interview, Gwama elder, Pokung, 24.09.2014 community and are supposed to teach the community how to enhance the productivity on the land. 98 The deplorable situation is captured by the fact that many Komo in Pokung, instead of going long ways to go honey hunting now wait till they get hired for day labour (for approximately 30 birr per day), and the extension workers seem largely unable to help and enhance the cultivation of the small (0,5 hectare) fields that the community members each have available. A brief look back into the historical acquisition of land will be helpful to make this point: As the highland-Ethiopian empire expanded towards the end of the 19 th century, Gambella was among the regions where British and Ethiopian claims for territory met. In 1902, when the border between the Sudan and Ethiopia was demarcated, Gambella took its recent shape on international maps and fell on the Ethiopian side of the border. Notwithstanding the fact that Gambella had always been a commercial hub for trade between the Sudan and Ethiopia, it was the British who expressed their economic interest in the region and leased from Menilek II a plot of land to establish a trading post at Itang. 100  For the ordinary people, the arrival of the modern state meant not only economic marginalization, but also the experience of slavery, which left a lasting impression on their mode of incorporation into the wider Ethiopian society.
[…] The people of Gambella region, therefore, first experienced 'integration' into the Ethiopian state as the loss of political autonomy, economic marginalization and the assumption of stigmatized identity. 104 Politically, Gambella was seen as a national territory, secured against British colonial aspirations, and, at the same time, a useful tool in negotiations for power and resources in the form of concessions. 105 The population of Gambella was largely seen as "in need of civilization" which the highland culture in the long run would bring. 106 We have seen in the previous chapter that the Socialist period (1974Socialist period ( -1991  The Anywaa community is very outspoken about the territorial disturbances of that period, which strongly focus on the loss of territory and national conflict. The feeling of becoming a minority in one's own land was fostered by the arrival (to refugee camps in the region) of large numbers of Nuer from the Sudan during the Sudanese civil war. Furthermore, Cuba and Russia supported large scale agricultural projects and irrigation schemes. A regional villagisation-programme alienated large parts of the local society. The imposition of a resettlement project had yet more far-reaching effects on the population. Of utmost importance for the understanding of land conflicts in the region is the perception that the indigenous population was being evicted by the rising numbers of migrant and resettles during the Därg. The land reform of 1975 had made all of Ethiopia's land national property. In so far as land was used in the political project of socialist 'hypermodernity' -socialist-oriented modelling of the social landscape -land in Gambella became a resource used to achieve social control. This land reform stood in severe contrast to the local understanding of land.
After the end of the military regime, Gambella became a killil in its own right within the federal system. The social contract between the groups living in Gambella has been weak and violent confrontations have occurred between the Anywaa and the Nuer, the Anywaa and the highlanders, and also among different Nuer groups. These conflicts centred on power sharing arrangements, land, and questions of economic and political ownership. It doesn't come as a surprise that in a rather fragile environment, conflicts over land and resources can be easily triggered by LSLI and its associated socio-economic changes-Gambella seems favourable for agricultural investment: Water from the highlands is abundant and it is especially along the water ways that the agro-industries are growing. The region is remote and in need of development and income creation. Hence based on the development strategy the Ministry of Agriculture called for investment in agricultural land and started leasing land in Gambella as well in other parts of Ethiopia  May 11, 2010) per year per hectare, which means, at the scope of the current contract, a yearly payment of 2,997,000 Birr (ca. 119, 880 Euro).
Next to the international agro-investors there are probably up to 300 Ethiopian investors active in Gambella. 109 Apart from the fact that large tracts of land are currently leased out and still others are being prepared for future investments, the government is also carrying out a villagisation programme. This programme, according to the government, aims at resettling up to 45,000 households and re-grouping them on approximately 180,000 hectares in order for them to "access socio-economic infrastructure." The target population are "those people who are settled scattered and along the riverside which are prone to flood hazards and those who practice cut & burn shifting cultivation and ultimately to enable them food secured and to bring socio-economic and cultural transformation". 110 In the view of many NGOs and parts of the local population, the current programme is related to LSLI and the land is being "cleared" for investors.
Nonetheless local concerns about the "clearance programme" are being ignored by the government in favour of its modernization project. 111 There are two interlinked developments emerging in relation to the current situation on land issues. One development is a mostly diaspora-led coalition of different opposition groups, linking international NGOs, such as Human Rights Watch, Survival International and the Solidarity Movement for a New Ethiopia (SMNE) that criticize this land appropriation. 112 The other development is the emergence and re-vitalization of armed opposition in the region, which put forward a more drastic tone like that of the Gambella Nilotes Unity Movement/Army. 113 In his concluding thoughts about, "how certain schemes to improve the human condition have failed," Scott has rightly pointed out: What is perhaps most striking about high-modernist schemes, despite their genuine egalitarian and often socialist impulses is how little confidence they repose in the skills, intelligence, and experience of ordinary people. 114 In light of the history of the failure of commercial farming in Gambella and Ethiopia in general, considering the implications of land investment on the fragile social contract between the people and the state, and in view of the devastating effects of resettlement and villagization, it is even more surprising that despite all these experiences from the past, the current government seems to largely neglect the development of new strategies and holds onto former strategies.
Many of the problems relating to land can be traced back to the way Gambella was incorporated into Ethiopian state territory since the end of the 19 th century. This process had, and still has, an imprint on the local population. The rhetoric of 'exploitation' and 'alienation' has its roots in the experience of the national incorporation of Gambella itself. Today, resistance against LSLIs and villagisationlocal and globally -is framed by the fear of loss of autonomy and the infringement on minority rights. On the side of the government the desire for a strong state in the periphery remains a policy guide: a new villagisation project is being run, despite the fact that the last villagisation project was abandoned by the previous government after 112 See: http://www.solidaritymovement.org/index.php. The spokesperson of the SMNE is Mr. Obang Metho, who has ever since 2003 been actively involved in human rights issues in Gambella and managed to keep the incidents of 2003 and its aftermath in both the media and the public consciousness. 113 http://www.gambellanum.org/ (last accessed 20.04.2013); the movement was created in August 2011 as a self-proclaimed reaction to the LSLI. The extent and public acceptance of the movement is hard to measure. The contestation on land policy and land appropriation is also fought out in the virtual spaces of the Internet; here the issue is by various opposition groups turned into one on the general political future of Ethiopia. 114  The experiences with the Socialist project's "fear and anger", has strongly inspired the return of the rhetoric of powerlessness among the different groups. 116 At the same time it is exactly this fear of exploitation that has inspired the rhetoric of "liberation" and "suppression" in which the land-grabbing debate is now placed.
Land has been one of the most important features in Ethiopian history. Securing and administrating it has been the aim of all consecutive Ethiopian governments since the empire expanded under King Menelik II. The federal structure has emphasized decentralization as a means of integrating Ethiopia's different ethnic groups into the national territory. This decentralization, the devolution of political and economic powers to the regional states, has naturally raised high hopes in rural populations concerning the ownership of land. But the land remains in the hands of the government. To make sure that centralization and the distribution of land through the land bank does not mean a break with the federal experiment, open, fair and sound redistribution processes of land, including public consultation have to be observed.
Citizenship in diverse societies is an issue of great and enduring salience. James Holsten pointed out that, "the worldwide insurgence of democratic citizenships in recent decades has disrupted established formulas of rule and privilege in the most diverse societies." 117 As we saw, struggles over citizenship have gained importance in 115 The Provisional Military Government of Socialist Ethiopia: Relief and Rehabilitation Commission: The Gambella Settlement Project: A WFP Assisted Settlement Site, Addis Ababa, July 1981; I am not engaging here in the discussion of forced or voluntary resettlement. Nonetheless it is important to note that the issue of LSLI has brought back the rural population's old fears concerning the results of the villagisation project, despite a similar plan to increase the well-being of the local population. These observations today call attention to an historical political space that, as a frontier region between Sudan and Ethiopia, was dominated by highly dynamic processes of inter-ethnic relations, migration and cultural change. Migrants from Sudan, merchants and fakis integrated into the indigenous societies from which grew Arabized indigenous communities. To the south, the Oromo kingdoms, soon to be important regional competitors, had developed hierarchical political structures, based on territorial control and subsuming various indigenous communities. This process led to an 'Oromization' of the indigenous communities.
This frontier region formed the space into which the Ethiopian empire expanded after the 1880s. Since the conquest of the Ethiopian army in 1898, the region was annexed and seen as an economic reservoir to the emerging Ethiopian state. Local rulers became middlemen between the periphery and the centre. The peripheral peoples forcibly supplied the centre with all kinds of goods, from gold to slaves. In the course of these economic demands, local authorities, such as Sheikh Khojali al-Hassan, rose to power and, as autonomous rulers, administered their lands, delivering the tributes in exchange for relative political independence. In the western-most Oromo provinces Jote Tullu had risen to become such a local leader. Thus, driven by the economic and political pressure of the centre, the gradual integration of the periphery unfolded. This in part was the beginning of the penetration and construction of central control into previously remote and autonomous regions. Increasingly, the indigenous communities were subjected to economic demands. Also, lesser principalities and -Cultural citizenship in Ethiopia is a 'place-making' project.
-As a project of place, citizenship -despite being initially a legal construction and status, as constitutionally outlined -is 'negotiated' on the regional level and depends largely on the cultural neighbours and the historically built majorityminority relations.
-Through its primarily territorial approach to ethnicity, the political system fails to integrate fragile minorities.
To begin with, citizenship is more than the "legal statues of members of a national political community". 2 The focus on the margins of the society enabled an ethnographic approach to federalism and multi-culturalism. The ethnography of citizenship practice brought out the processes behind membership of a given political community. Being observant to this process is as important "as the end result itself." 3 At the same time it illuminated the obstacles of citizenship expansion in the given case.
a. Citizenship in Ethiopia is a place-making project Minorities in Ethiopia have to be territorial to be politically meaningful. This is expressed in the ethnic designation of territories, on all administrative levels from killil to the special wäräda. It is also expressed in the many interlocking competitions and conflicts over boundaries and territories in Ethiopia today.
The Mao and Komo have been localized in various overlapping space-making projects, processes I have referred to as the competition for territorial control. 4 The history of the Mao and Komo is deeply interwoven with the western Ethiopian frontier, as well as consecutive territorial claims by various groups and entities. Appadurai emphasized the fact that space making is "hard and regular work". 5 Work that includes spatial practice, the building of houses, maintenance of borders, or symbolic practices of all sorts. 6 The Komo and Gwama made place where they moved, until the next force drove them somewhere, displaced, as we have seen, by the intertwined civil wars, settlement schemes, etc.
Today the Mao and Komo are being localized in the ethnic landscape of Ethiopia.
They are granted territories but only as long as the land is not needed by the government for other territorial or national endeavours, like land investment, etc. The Mao and Komo are being localized in village complexes, as we have seen in the last chapter. On the other hand, the relation between territory and identiy has provided certain ethno-elite groups within the complex Mao and Komo population to push for administrative positions and political recognition within the confines of an ethnic territory. In the quest to define the uniqueness of the people, such ethno-elites emphasize cultural traditions.
Through their territory the Gwama have entered the realm of modern-day identity politics in Ethiopia. A singer by the name Kush Kush very recently produced a typical cultural pop song similar to much regional-specific music. It is mainly in modern pop music-style and parts of the chorus' lyrics say, Turana tongo, usiti tongo -"Tongo our homeland, the people of Tongo". The video shows the singer in Tongo town. It opens with a scene of a Gwama divination process, and is accompanied by Gwama dancing and traditional areophone instruments (which are not part of the music, though). I see this as an example of the importance of expressing group identity and reference so as to place to be part of the mainstream cultural landscape of Ethiopia today. In the ethnic territory, the Mao and Komo (mainly Gwama) are able to perform this form of cultural citizenship. 7 In such a way, the construction of identity and ethnicity become strategies to be politically meaningful. This also reflects in the attempt to push for their own cultural and learning materials in the major settlements. The paradox that this situation entails lies in the opposing place-making processes.
While the Mao and Komo have been split apart over various territories, and both international and regional frontiers, the ethno-territory is confined to serve only those groups that inhabit it and that are able to claim autochthony in it. This situation weakens their political relevance in an already politically and economically marginal environment.
b. Citizenship is negotiated on the regional level The issues of why some groups stand in the shadow of others are complex and cannot be easily described. Minority groups, numerically inferior and politically insubstantial in the multi-cultural competition for political power and recognition, naturally find it more difficult to present alternative political solutions opposed to the political mainstream, simply because they are less vocal in the political bargaining.
For the timeframe of this thesis, the first frontier movement that the Mao (Gwama) were exposed to was the migration of the Berta in the 17 th century. Another frontier, the Oromo frontier, pushed northward from the Anfillo and other Omotic Mao regions. The local history continued to be one of constant displacement, as the recollections and interviews have shown. All Mao, Gwama and Komo felt the pressure of invading groups. They fled or were drawn into new emerging polities, and eventually into the Ethiopian state. All groups, the Mao -both Omotic and Nilo-Saharan -as well as the Komo remember slavery and the tax in humans (lij gibbir) as a major humiliation. In this thesis I have shown that patterns of historical marginalization are being transmitted from generation to generation. These feelings are memorized and internalized by the Mao and Komo, but similarly they live on in the memories of the former masters who are still the dominant group. Memories about the "other" and the "self" define the way people perceive each other today; and this extends to the practices of citizenship, inter-regional and political cooperation and conflict. Sometimes, perceptions about 'the other' are strongly expressed: What do you want from the Mao, they were our slaves: they don't have a history. If you want to know about them talk to us. 8 Often such recollections are concealed or hidden, while at other times people try to redefine the history of their forefathers. During a group interview with Fadashi, a subgroup of the Berta and descendants of the watawit, we discussed the relation of the Historical frontier processes and migration, as we have seen throughout the thesis, have fostered such complexities. The political framework of a territorial identity is diametrically opposed to the historical frontier processes, and as such dramatically alters the historically grown inter-ethnic relations. It illuminates the small extent to which the focus on ethnicity and territory in Ethiopia is able to capture the finer points of cultural complexities. The spectre of oblivion I was surprised to learn from one young, educated Komo (probably aged twentythree) in Gambella that he had never before heard stories about slavery. He was not the only one. Hence, my description of slavery based on the memories of elders is insufficient. Not only am I unable to capture the finer points of its legacies, but I am also unable to capture the silence concerning the inter-ethnic relations today. Similarly the memories of elders do not capture the nascence of the younger generation. People born after the civil war in the past twenty years may have more demanding questions regarding socio-political developments, culture, and identity, which they can hardly explain with reference to the history of marginalization alone. They nonetheless may experience racism and marginalization in the schools and universities both on the regional as well on the national level in the interaction with youngsters from other ethnic backgrounds.
Regardless of the historical injustices and possibilities, the practice of cultural citizenship in Ethiopia today does not offer any special rights for historically marginalized groups, and neither does Ethiopia see any need for special consideration of indigenous or autochthonous groups. Ethnic politics in Ethiopia today are based on a strict post-modern paradigm: given the provision made by the constitution under the umbrella of 'nations, nationalities and peoples' is deemed sufficient for cultural development. This paradigm states that the current government, after overthrowing the Amhara-centric, socialist, military regime (as a successor of the earlier imperialfeudal regime) and reorganizing the state into decentralized federal sub-entities which all have the right of "self-determination up to succession," 13 has liberated the ethnic groups of Ethiopia, giving them their dignity and right to cultural expression. But history has always been a state-centered project in Ethiopia. The political and administrative reorganization of Ethiopia along ethno-linguistic lines and the current politics of diversity, being promoted on all political levels, have led to an official neglect of marginalization as part of the historical interethnic relations in the name of peace and stability through a proclaimed unity in diversity.
Minority citizenship can, to a certain degree, be realized on the regional level but cannot be sufficiently claimed on the national level. Furthermore, only groups that fit 13 "The Constitution of Ethiopia", Art. 39.4., 14 the criteria of cultural citizenship of nations, nationalities and people may also be able to exercise political citizenship. Thus, repeating a familiar historical process, fragile groups may wither away into any of their more stable neighbouring groups or recede into oblivion.
-Field notes of Alessandro Triulzi (Asosa-Begi, a-bg 4). Diesen Prozess beschreibt die Arbeit anhand der politischen Entwicklungen in der Peripherie und vor dem Hintergrund der Geschichte des Mao-Königs   These observations open up a political space that, as a frontier region between Sudan and Ethiopia, was dominated by highly dynamic processes of inter-ethnic relations, migration and cultural changes. Migrants from Sudan, merchants and faqis (teachers of Islam) integrated into the indigenous societies from which grew Arabized indigenous communities, as well as a local elite, the watawit.
To the south the Oromo kingdoms, soon to be important regional competitors, had developed hierarchical political structures, based on territorial control. This frontier zone, which I describe by means of the frontier theory by Igor Kopytoff, forms the space, into which the Ethiopian empire expended. Since the conquest of the Ethiopian army in 1898, the region was annexed as an economic reservoir to the developing Ethiopian state. Local rulers became middlemen between the periphery and the centre.
The peripherals forcibly supplied the centre with all kinds of goods. Annually, the periphery delivered gold, and in addition slaves, honey and other items, to Addis Ababa. In the course of these economic demands, local authorities, such as Sheikh Khojali al-Hassan, rose to power and, as an autonomous ruler, delivered the tributes in exchange for relative political independence. In the western-most Oromo province, Jote Tullu had risen to become such a local leader. Thus unfolded, driven by the economic and political pressure of the centre, the gradual integration of the periphery. This in part was the beginning of the penetration of central control into previously remote regions. The small groups were subjected successively to the same economic demands. Also, more fragile groups and smaller chieftaincies were integrated. The fought together against the Ethiopian government. The war led to further fragmentation of the groups. At the end of the war, an alliance of regional liberation movements had prevailed against the Ethiopian military. This was followed by the consolidation of post-war Ethiopia based on an ethno-federal arrangement after 1991. In this federal system ethno-linguistic territories were created which served the political deconcentration of national politics. The more than eighty ethnic groups of Ethiopia were, based largely on primordial aspects, such as language, or kinship, etc., divided in respective ethno-territories. Where small groups should be integrated, further sub-territories were established in order to follow this ethno-federal principle.
The Mao and Komo were incorporated into both the federal states of Benishangul-Gumuz and Gambella. Mentioned in the regional constitution as nations, nationalities and peoples, people of these regions they were thereby (constitutionally) made citizens of Ethiopia. Due to the long period of marginalization and expulsion, members of these groups also live in other regions, notably Oromia, without the corresponding political manifestations. In addition, other Omotic groups, also known as Mao, live both in Benishangul-Gumuz and in Oromia. They bear the name of ethnic Mao; however, except to the common history of marginalization they have almost no contact with the Mao and Komo in Benishangul-Gumuz. Against this ethno-political background, the present work discusses the scope of cultural citizenship and political integration in a multi-ethnic state. Minority citizenship is the vehicle to illuminate both the historical manifestation of state encroachment as well as the inter-ethnic history of the Mao and Komo. The multiethnic state of Ethiopia and its multicultural policy is thereby questioned from the perspective of the minorities. Their sense of history, their experience of marginalization, and their experiences with the state on the margins, are at the centre of the historical perspective of this thesis. Conceived as a chronology of the encounters between state and minority, this text is dedicated to a synchronic and diachronic perspective of the experiences of the Mao and Komo in the context of state intervention, which directly or indirectly influenced the regional inter-ethnic relations.
The final chapter of the text describes, based on observations and interviews collected during repeated field research, how the Mao and Komo remember the history of marginalization and especially how the history of slavery affects their sense for today's society. Moreover, the analysis of the political system looks at the possibilities and difficulties that arise for such scattered minorities who are still marginalized on the regional level. Examples discussed hereafter concern the effects of the land policy, and how the federal system limits the political rights of scattered minorities.
This thesis concludes that the concept of citizenship, framed by Western democratic thought, as the membership of groups in a national whole, is very suitable as an analytical tool to describe the interaction between the state and minorities. It especially is able to highlight the limits of such interaction. However, as this thesis shows, political participation is not met by the ethnicity-identity-territory approach to citizenship proposed by the Ethiopian political system. Hence approaching ethnopolitics through the perspective of minority citizenship shows a significant divide between regional and national belonging.