The Spatial Imagination of Oromia: The Ethiopian State and Oromo Transnational Politics. (Bas van Heur)

Ethiopia has always produced more imaginations than its territory could incorporate. To some, Ethiopia is the result of a thousand-year old history that has its roots in the Bible and can be directly connected to the Western civilizations of antiquity. For many Afro-American intellectuals since the 19^th century, it represents the source of their Africanicity and constitutes a powerful image of belonging and return. Seen from this perspective, many of the writings on the state of Ethiopia after the change of government in 1991 and the flight of Mengistu Haile Mariam to Zimbabwe can be seen to reflect above all the wishes, dreams and nightmares of its producers and less the factual changes inside Ethiopia. The removal of the Derg dictatorship and the arrival of the Tigrayan People’s Liberation Front (TPLF) as the main force behind a coalition called the Ethiopian People’s Revolutionary Democratic Front (EPRDF) seemed to usher in a new era of prosperity and democracy. After a thirty-year war, Eritrea became a separate nation-state under the leadership of Isaias Afewerki, previously the head of the Eritrean People’s Liberation Front (EPLF). Ethiopia, headed by Meles Zenawi, officially denounced its political tradition of centralization and started advocating and implementing a federal framework to solve the “nationalities problem” within its territory. Within a matter of months, the discourse of Marxism-Leninism was abandoned and democracy and human rights were embraced, causing the United States government to hail Isaias Afewerki and Meles Zenawi as prime examples of Africa’s New Leaders. This praise has remained remarkably constant over the last decade, although there has been a rhetorical shift away from democracy towards an emphasis on peace, stability and the fight against terrorism. During a recent visit, German chancellor Schröder reiterated this argument, emphasizing both Ethiopia’s firm commitment to fight terrorism as the fact that the new leaders had “become exemplary regarding good governance.”[1]

Meanwhile, on the other side of the global fence, a completely different set of imaginations is being produced. The implementation of the federal framework in Ethiopia accorded primacy to ethnicity as the only route to democratization and officially devolved central powers to regions below the scale of the state. Each ethnic group was accorded a separate region. In principle, this was a logical consequence of the idea of “national liberation” as used by the TPLF and other liberation movements such as the Oromo Liberation Front (OLF) during their struggle against the Derg regime from 1974 until 1991. With a background in the Marxist-Leninist study circles and organizations that emerged before the revolution, after their exclusion from the center by the Derg, they turned to the peripheries with the aim to liberate their ethnic nation from the oppressive state. In the case of the OLF, this struggle was actively accompanied by the spatial imagination of Oromia. It functioned as a sort of rallying cry around which elements such as the Oromo language (afaan oromoo) and the age-grade system called gada could be politicized and which enabled the construction of a collective identity radically opposed to the Derg regime. The change of government in 1991 suddenly created the possibility to put this imagination into practice. The OLF participated in the Transitional Government of Ethiopia (TGE) and could establish the region of Oromia within the federal framework of the Ethiopian state. After a century of oppression and rejection, afaan oromoo became the official language in this new region. Only one year later, however, the OLF had left the TGE and its role was taken over by the Oromo People’s Democratic Organization (OPDO) that was part of the EPRDF-coalition and dependent on the powerful TPLF. Since this time, one is in effect dealing with two opposing versions of Oromia, despite the largely similar contents. The OPDO, as the regional representative of the state and the Oromo people within its region, governs Oromia and is carrying out a project of cultural nationalism as originally imagined by the OLF. The OLF, however, is now involved in a discursive and military struggle against the OPDO, claiming it does not legitimately represent the Oromo people, but is merely a continuation of the earlier repressive regimes of Haile Selassie and the Derg.

How to explain this paradoxical situation? Why did the OLF leave the TGE when their spatial imagination of Oromia was being realized in socio-spatial practices? And why has every attempt at reconciliation between the parties since 1992 failed? One part of the answer has already been provided by a number of commentators. It is argued that this division is the result of a political tradition in Ethiopia that sees competition as treason and that is therefore primarily preoccupied with the monopolization of power.[2] Through a theoretically informed historical analysis of the spatial imagination of Oromia, this thesis tries to illuminate another aspect of this political impasse. The premise is that every political analysis needs to include those spatial practices and imaginations that transcend the framework of the state. My thesis is that the lack of “common ground” between the OLF and the OPDO/EPRDF is intimately related to the fact that the OLF-version of Oromia was from the very beginning the result of transnational practices: partly produced by actors not directly dependent on the resources of the Ethiopian state, the spatial imagination of Oromia could be conceptualized as radically different from this state. It is this posited difference that provides the other part of the answer to the question why the OLF left the TGE in 1992.

The academic and non-academic literature on Ethiopia that has appeared since the 1970s can only be understood by relating it to the earlier corpus it reacted against. To a large extent, this earlier literature was a state-centered discourse: Ethiopia as an ancient Christian civilization amidst a sea of barbarism. The focus of this “’Church and State’ tradition”[3] was overwhelmingly on the important role of Ethiopian monarchs, the intricacy of the Amharic culture and language and the independence and autonomy of the Ethiopian state.[4] Those periods that did not offer the possibility of a postulation of unity – such as the Era of the Princes from 1796-1855 and the short reign of Lij Iyasu from 1909 until 1916 – were interpreted as periods of fragmentation and deterioration as a result of Oromo rule or a “dalliance” (Marcus 2002, 114) with Islam. One of the results of this narrative style has surely been a somewhat conservative outlook that emphasizes and prefers continuities and vehemently rejects even the thought of historical breaks. It also informs the interpretation of contemporary history. Erlich’s exaggerated fear in 1977, for example, that “the almost inevitable result of the rising power of local Islam and the weakness of Ethiopia is the political and ideological annexation of the Horn to the Middle East” (408) might have little to do with the factual situation inside Ethiopia, but is clearly the result of this well established historiographical tradition.[5]

The deposal of Haile Selassie, the radicalization of the 1974 revolution and the central role that was accorded to the discourse of ethno-nationalism by actors inside Ethiopia changed all this. Itself the result of specific developments within the Ethiopian student communities in Addis Ababa and abroad, the discourse of ethno-nationalism provided a challenge not only to the Haile Selassie regime, but also to the mainstream interpretations of Ethiopian history. With its emphasis on the liberation of the ethnic nation from its oppressors, it radically questioned the continued existence of the Ethiopian state as it was cherished in the minds of many writers around the world. Partly as a reaction to these changes and as a way of dealing with the dramatic developments after 1974, one can distinguish – in a simplified way - three conceptualizations that have been offered to overcome the state-centeredness of the hegemonic historiographical and political tradition.

The first and most important contribution has been the conceptualization of the Ethiopian state as a center-periphery structure. It provided a way of acknowledging the diversity of the different Ethiopian societies, while at the same time embedding these societies within an overall framework that connected them to the center. Clapham (1975) provides an early example of such an approach, but the classic statement has probably been offered by Donham (1986) in his introduction to The Southern Marches of Imperial Ethiopia (Donham and James 1986). This volume contains a range of articles, which analyze the making of the Ethiopian state “from below” and these both complement and criticize the hegemonic state-centered analyses. In his article, Donham also connects the center-periphery concept to the notion of frontier as one way of connecting the two poles; a connection that has been described in more detail by Triulzi (1994). The use of the frontier paradigm, however, has its roots in older articles describing the interaction between ethnic groups.[6] Fukui and Markakis (1994) have offered an adaptation of this center-periphery approach by adding an intermediary range in-between the periphery and the center. In their view, this allows for distinguishing between conflicts at the margin that are unrelated to the state, conflicts in the center where the state is the common point of reference, and conflicts in-between that involve groups not directly interested in the state, but that seek to exploit it “for their own parochial purposes” (3). McGinnis (1999) offers a similar materialist view, but interprets it in the vocabulary of local, national and international conflicts.

Partly overlapping with this center-periphery approach is the second conceptualization that makes use of theories of space and place for a variety of purposes.[7] Donham (1993) has described some of the changes in spatial practices in the area of Maale that were the direct result of the 1974 revolution. Arnesen (1996) has used geographical theories on place to analyze the historical formation and transformation of Derra, an Oromo-inhabited area in the highlands of Shewa. The article by Cohen and Isaksson (1987) on the villagization in the Arsi-region does not explicitly refer to these theories, but its thematic is similar. Scott (1998, 247-252), in his important analysis of the imaginations and practices of the state, has made extensive use of this article and translated some of its implications into a spatial vocabulary. From a political-economic perspective, it is Clapham (2002a) who has delivered the definitive version so far, in his analysis of the different ways Haile Selassie, the Derg regime, and the EPRDF have tried to control their territory through the use of specific spatial relationships.[8]

A third conceptualization can be abstracted from those spatial practices that literally traverse the center-periphery structure of Ethiopia, namely those of the thousands of migrant workers and students that went abroad to find work or pursue their studies and the hundreds of thousands of refugees that were the result of war, famine and violence. Both these processes have caused a partial deconstruction of the center-periphery structure of Ethiopia and have led “to a more open series of interactions drawing upon partially shared and intersecting ‘ethnoscapes’ of the imagination” (Donham 2002, 2). That said, although a number of articles and monographs have been published on refugees in the Horn of Africa, the research on their imaginations is still in its infancy.[9]

This brings us to the state of literature on the Oromo and Oromo ethno-nationalism. One is tempted to say that the movement of refugees and students has also led to a fragmentation of research. If one attends both to the location of research and to the discourse that is the result of this research, one can discern a number of different approaches. The first approach takes the traditional anthropological route: it involves fieldwork among Oromo societies inside Ethiopia and results in a number of articles or monographs that try to describe these societies. Some more or less classic examples of this approach would include the work of Lewis on the Oromo in Jimma, the work of Baxter on the Arsi and Boran Oromo, Bassi’s research on the Boran Oromo and the writings of Blackhurst on the Oromo in Shewa.[10] Although I will refer to some of this literature in the following chapters, most of these writings are concerned with practices and imaginations of local societies – including their perceptions of and dealings with the Ethiopian state[11] – and only to a limited extent with analyses of Oromo nationalism.

A second approach involves research among refugees in the neighboring countries of Ethiopia that were the result of the 1974 revolution and the subsequent violence. Where Luling (1986) mainly offers a summary of a handful of conducted interviews in Sudan, more critical research on Sudan and Somalia has been offered by Braukämper (1982/83) and Zitelmann (1994a). In particular Zitelmann offers a valuable analysis of the role of the OLF in Sudan. Some of the more recent examples would include Scherrer, Namarra and Dibaba (2002) on Sudan and Kenya and Fossati, Namarra and Niggli (1996) on Djibouti. A third group of literature has also been produced in the neighboring countries of Ethiopia, but is not so much interested in a description of the situation in these areas as it is in the spatial imagination of Oromia. This is the domain of Oromo nationalists. Bulcha (1988) takes a position in between: as an Oromo nationalist, his interest is directed towards the spatial imagination of Oromia (although he tends to use the term Oromoland), while at the same time he has been conducting empirical research among the refugees in Sudan. More obvious examples are those pamphlets, articles and books produced by OLF-supporters, such as the book Oromia under the pseudonym of Gadaa Melbaa[12] and magazines such as Oromia Speaks, Bakkalcca Oromo, Sagalee Bosona (Bulletin of the OLF military activities) and Warraqqa (organ of the OLF youth league).

This brings us to the fourth and probably largest category, which consists of those discourses that are produced in the Diaspora and speak of the spatial imagination of Oromia. This version is intimately related to the Oromia-discourse among refugees in the neighboring countries of Ethiopia, in particular Sudan, since active contacts were maintained between the actors in the Diaspora and the OLF in Sudan. During the late 1970s and the 1980s, the main publications were by student organizations in Europe and North America. The Union of Oromo Students in Europe (UOSE) published Bakalćća, Sagalee Oromo and Kara Wallabuma, whereas the Union of Oromo Students in North America (UOSNA) published Wallaanso and Sagalee Wallaanso. For a while there also existed a Union of Oromo Women in Europe with a separate publication called Oromtittii.[13] Although many of these publications still exist, since the 1990s most visible activity has taken place in the United States within the academic community: the students became teachers devoting themselves to the promotion of the spatial imagination of Oromia. Asafa Jalata, Lemmu Baissa, Mohammed Hassen, Sisai Ibssa and Kuwee Kumsa are some of the more prolific writers.[14] This activity is slowly being noticed by non-Oromo academics and has led to a fifth line of investigation with a number of studies describing and analyzing the discourse and practices of Oromo communities in the Diaspora. Zitelmann (1988, 1992, 1994a) was probably the first to do detailed research in this field and his work on Oromo communities in Germany is still one of the few that seriously incorporates what Manning has recently advocated as the “homeland plus Diaspora model” (2003, 494): an analysis that incorporates both the homeland and the Diaspora and pays attention to continuities, breaks and interactions between them. More recently, Gow (2002) has presented a committed ethnographic study of the Oromo community in Melbourne, Australia. A book by Matsuoka and Sorenson (2001) deals with the construction of identity and community among refugees from Ethiopia and Eritrea in Canada. Although the main emphasis is on Eritreans and Ethiopians (that is, Ethiopian nationalists defending a unitary view of the state), one chapter is specifically devoted to the Oromo. Also, Bulcha recently published The Making of the Oromo Diaspora (2002) that interprets the Oromo from a global perspective, while at the same time incorporating centuries, from the Arab slave trade to the present.

My thesis can be seen as a contribution to this “homeland plus Diaspora model.” With an emphasis on what Braudel called the longue durée of historical time, the thesis works through a number of continuities that have, in my view, more or less stayed the same over a longer period of time and that have hardly lost any of their force. At the same time, attention is paid to decisive breaks that occurred during these spatial practices towards the Diaspora, causing a “disruptive spatiality”[15] in the process. On a conceptual level, this involves a clear rejection of the state-centrism still prevalent in most texts that deal with the relationship between Oromia (imagined or real) and Ethiopia. Although the center-periphery conceptualization did much to develop the idea of separate but interconnected societies within the state of Ethiopia, the conceptualization of Oromia is still overwhelmingly interpreted as only a reaction to oppression by the Ethiopian state and not as also the result of transnational spatial practices.[16] In contrast, this thesis tries to show the constitutive role these spatial practices 1) have played in the construction of the spatial imagination of Oromia, and 2) still play in everyday politics in Ethiopia.

This thesis is divided into two sections and eight chapters. The first section (comprising the chapters II and III) is dedicated to an exploration of a number of theoretical issues and the clarification of the central concepts that will guide the historical analysis of the spatial imagination of Oromia. Acknowledging the fact that geography – here in a physical sense: mountains, escarpments and rivers – has played an important role in the social, cultural, economic and political construction of Ethiopia,[17] I have decided to approach the diverse theoretical literature on space and place from the perspective of geographical theory. It is guided by the conviction that embedding theory in larger disciplinary traditions – thereby acknowledging its historical nature - serves a better understanding of both theory and its appropriateness to grasp social phenomena. Chapter II discusses the concepts of region, place and locale. A summary description of the different ways the concept of region has been used in a number of older traditions provides the necessary background against which the developments in geographical theory since the 1970s should be understood. Of particular importance is the concept of place used by humanistic geographers as an antithesis to the abstract and objective concept of space from spatial science. The concept is central to this thesis, since it provides a way of thinking beyond a political field inhabited by purely rational actors. At the same time, the major strength of this humanistic approach is also its main weakness. Its emphasis on the human character of place too often leads to a conflation of place with community and to an internal view of place-construction. Through a critical discussion of Giddens’ concept of locale in the final part of this chapter, first steps are made to connect this concept of place to larger structures and to deconstruct this internal-external dichotomy. Chapter III conceptualizes the role of regional, national and global processes on the construction of place, while at the same time emphasizing the importance of the agency of specific actors. The first part discusses the importance of Marxist theory on the conceptualization of space. Leaving behind a view of space as absolute or as merely a spatial structure in which social life takes place, Marxist geographers argued that capitalism produces spaces because of its expansionary nature and that this results in processes of de-territorialization as well as re-territorialization. Through a discussion of an important 1989 book by Harvey and his concept of time-space compression, I want to point out some of the limitations of this approach. The main critique is that Harvey offers a top-down approach to time-space compression. Taking my lead from feminist theory, I argue that one has to attend to the question of positionality in order to grasp these processes “from below.” Actors are not simply victims of capitalism, but actively make use of the new possibilities offered by time-space compression: through an appropriation of spaces – discursively as well as in social practice – they try to reconstruct their place. The term spatial imagination will return again and again in this thesis as shorthand for these processes of association. The third part of this chapter returns to questions of structure that delimit these spatial imaginations. It makes use of recent theorizations of scale that argue that geographic scale is actively appropriated by actors as part of their arguments and practices in order to convince others of the usefulness of a certain scale, while denying legitimacy to other scales. Combining this aspect of scale with the conceptualizations of space and place makes it possible to attend to, what I have called, the scalar positionality of spatial imaginations.

The second section comprises the case study of the spatial imagination of Oromia. The emphasis on the longue durée (as already indicated above) of the Ethiopian state and its continued relevance to the interpretation of the spatial imagination of Oromia necessitates an historical overview that incorporates much of the twentieth and even nineteenth century. That this approach is useful to get a grip on the dynamics of Ethiopia seems clear not only to me, but also to a number of other observers.[18] Chapter IV discusses the politico-economic developments of the Ethiopian state that even now form the structural coordinates for many of its subjects. The first part describes the internationalization of the state since the end of the 19^th century. Both state-controlled trade and Ethiopia’s military alliance with the United States during the Cold War (until shortly after the 1974 revolution) emphasized the connections between the scale of the state and a number of international actors, while at the same time making the state less dependent on the sub-state scales it encompassed. The second part discusses the process of centralization and territorial expansion of the state during the end of the nineteenth and the beginning of the twentieth century. Focusing on the question of control over land, it analyzes how the southern peripheries were connected to the center, developing a hierarchy with actors occupying a range of socio-spatial positions. The third part provides a critique of the center-periphery concept and argues that the conflict between the Amhara and the Oromo as represented by its elites should not be interpreted as a struggle between the center and its periphery, but instead as a struggle between two differing elements of the same center. Chapter V will explore this argument in more detail through an analysis of this center – Addis Ababa – and of the university students before the 1974 revolution in the capital and abroad, since it was in this context that the OLF discourse on Oromia emerged. Describing two transnational spatial practices – the arrival of scholarship students from other African countries and the studies abroad by Ethiopian students – it will be shown how the specific spatial imaginations of the student population, including the OLF, was the result not only of difficulties with the state, but also of these transnational spatial practices and the access they enabled to spatial imaginations not available before. The impact this had on the representation of Ethiopian history is illustrated through an analysis of an important rebellion in the 1960s and the way this rebellion was represented by the OLF. Chapter VI discusses the revolutionary period from 1974 until 1991, but is less concerned with specific developments inside the country, as it is with the exclusion of the OLF from the center. After a discussion of the early revolutionary years that were characterized by a process of homogenization – within four to five years all organizations and divergent voices were annihilated, assimilated or excluded from the center, leaving the Derg as the only power with direct access to the resources of the state – the focus will be on the OLF and their spatial position in the border area between Sudan and Ethiopia. It will be argued that this move towards the periphery should not be interpreted as a marginalization, but instead as an opportunity of globalization, since in Sudan the OLF was able to establish connections with actors all over the world. An important correlate of this argument and central to my thesis is that this embedded the spatial imagination of Oromia in another spatial network: the move towards Sudan turned Oromia into a diasporic imagination. Chapter VII discusses some of the changes concerning both the spatial imagination of Oromia and the new region of Oromia since 1991. As already mentioned, the year the OLF was part of the TGE led to the construction of Oromia as a region within the federal framework of the Ethiopian state. Building on my previous argument, it will be shown how the OLF was not only excluded from the coalition as a result of the Ethiopian tradition of a monopolization of power, but also as the result of a diasporic imagination of Oromia that could not imagine a cooperation with the Ethiopian state. The continued but separate access of both the OLF and the OPDO/EPRDF to global resources and mediators provides the key to understanding why it has been impossible for the parties to find any “common ground” since 1992. Chapter VIII summarizes the arguments made in this thesis and will briefly compare them with another case – Western Sahara – in order to offer some first impulses to a possible comparative study and to test the more general applicability of my conceptualization of transnational politics.

- Theorizing Spatial Imaginations -

II. Region, Place and Locale

The concept of region is one of the most basic and oldest in the discipline of geography. If one accepts the working hypothesis that space is abstract and universal (I will discuss different conceptualizations of space in the next chapter), then a region is always a more or less bounded space that possesses some sort of unity through which it can be distinguished from other regions. As a further specification of this general definition, one can discern between two, partly overlapping, discourses that have made extensive use of this concept: 1) the region in its loose pre-nineteenth century usage, and 2) the region in areal differentiation and spatial science.

In the first case, the region retained a loose, almost “commonsensical” usage and could refer to anything from a vast tract of land to whole continents and to areas of specialized economic activity such as port-cities and their hinterlands. Most writers working from an international relations perspective or in a tradition of political and economic history have constructed their arguments around this definition of the region. In particular the top-down imageries dividing the world into distinctive macro-regions have been harshly criticized since the 1970s. Placing “Europe” at the center of attention and installing “Africa,” “Asia” and “America” in distinctive positions within this “matrix of difference and deferral,” as Gregory called it (2000b, 687), has enabled writers to assign cultural stereotypes to each region. Excluded from these regional imaginations was the acknowledgment that these were largely structured by European exploration in the first place. In particular travel writing has been a vital source for the discursive production of these regions as bounded spaces. India became associated with hierarchy, the Mediterranean with honor-and-shame and Africa with kinship. Appadurai has usefully characterized these imageries as gate-keeping concepts “that define the quintessential and dominant questions of interest in the region” (1986, 357). It was largely the publication of Said’s Orientalism in 1978 that made many writers more aware of the imaginary nature of these geographies and their own involvement in the construction of hierarchical relationships separating the Occident from the Orient. It has since spawned a whole new field of postcolonial studies dedicated to the deconstruction of the many stereotypical images and discourses on “the Other” as produced by “the West.”[19]

The second general discourse dealing with the concept of the region brings us closer to the traditional heart of the geographical discipline. Seen from this perspective – usually referred to as areal differentiation - the region becomes one of the basic “building-blocks” of geographical inquiry. The assumption was that the world could be divided into bounded spaces that, put together, would add up to a larger totality. Geographers such as Vidal de la Blache, Hettner and Hartshorne were among the leading proponents of this approach. This “building-block” approach was shared with the successor project of spatial science. Although this new generation of geographers after the Second World War attacked the “traditionalists” for seeing locations as unique and for justifying a regional geography in which “areal differentiation dominated geography at the expense of areal integration” (Haggett qtd. in Agnew 2000, 35), both approaches were interested in constructing regions as ways of classification and organization. The main difference between them was that the practitioners of spatial science argued for a positivist view of geography. Their work was part of the quantitative revolution that dominated the larger field of the social sciences during the 1950s and 1960s. Instead of connecting a particular identity or culture to a region and its natural environment, as proponents of areal differentiation tended to do, spatial scientists were mainly interested in discerning central organizing principles within societies largely severed from its physical landscape.

It has been against this positivist tradition that most geographers since the 1970s have directed their critique. In a general way, one can discern four discourses that argued against the “spatial fetishism” of spatial science: 1) a return to traditional regional geography; 2) the development of a humanistic geography; 3) the application of structuration theory on questions of space and place; and 4) the use of a Marxist vocabulary to theorize the production of space. Although Marxism had a lasting influence on the discipline, I will discuss this approach in the next chapter. Because of the theoretical meta-nature of many of these texts, this discourse seems to have had its major influence on the definition of the concepts of space and scale.

One line of critique against spatial science involved a picking up of the earlier descriptive tradition of areal differentiation. The practitioners of spatial science had directed their harshest criticism against this approach, but never fully succeeded in convincing the traditionalists from the benefits of the quantitative revolution. In addition, the many practitioners of spatial science became disenchanted with the paradigm’s limitations and started looking for new methods. It was in this context that defenders of the descriptive approach attempted to gain more ground. Hart is often quoted as one of the central figures in this movement. In his presidential address to the Association of American Geographers, he described regional geography as an integrative discipline, which made it, according to him, “the highest form of the geographer’s art” (1982). Hart’s approach, however, is hampered by a general refusal to conceptualize his vocabulary, relying instead on the power of description. This causes at least two conceptual failures.

A central problem is Hart’s refusal to explicate the status of regions as objects of study. The study of regions should take central stage in the discipline of geography, he argues, because it is only by using the region-concept that one can tie together “all of the wildly disparate phenomena with which we deal” (18). What defines a region, however, remains unclear, since “regions are subjective artistic devices, and they must be shaped to fit the hand of the individual user” (22). This might be true, but it makes for a critical concept of limited value, since no attempt is made to incorporate questions of power that play a role in formulating the characteristics of a region. Connected to this is a second fallacy of Hart and other traditional regional geographers, namely the implicit use of a framework of naturalism that assumes correspondence between human and geographical patterns. The implication is that by describing the physical features of a region, one is able to infer the characteristics of the people living inside this region. At the same time, Hart seems reluctant to accept such a naturalistic approach and posits human beings (as a universal concept) as unconstrained by social relations. As he puts it: “I would remind you that most of the truly significant changes in our society have been initiated by individuals, or by numbers of people so small as to be totally insignificant in any statistical sense” (5). Such a view, however, ignores the embedment of individuals in social structures, unveiling, as Pudup has pointed out, traditional regional geography’s commitment to voluntarism and methodological individualism (1988, 376), thereby reducing human life to a struggle between man and nature.

A second line of critique against spatial science and one that – in contrast to traditional regional geography – privileges the subjective experience of place is the tradition of humanistic geography which emerged in the Anglo-American discipline during the 1970s. Although its roots have been traced to traditions as diverse as the French school of human geography, the Chicago school of sociology, the work of Heidegger or even the anarchism of Kropotkin (Gregory 2000a, 361), its main impetus came from a rejection of the spatial science paradigm. Dismissing the idea that human beings could only respond passively to abstract spatial structures, humanistic geography argued for an emphasis on the active role of human agency and creativity in the construction of places.

Notice the move towards place. Whereas the practitioners of areal differentiation and traditional regional geography mainly concerned themselves with regions, humanistic geographers explicitly appropriated the concept of place to distinguish themselves from spatial science’s concern with abstract space. Authors like Relph (1976) and Tuan (1977) interpreted place as more subjectively defined and were interested in the ways people attach elements like language, human attitudes and events to specific places. According to Tuan in a later article that focuses on the role stories and myths play in place-construction:

“[…] storytelling converts mere objects ‘out there’ into real presences. Myths have this power to an outstanding degree because they are not just any story but are foundational stories that provide support and glimmers of understanding for the basic institutions of society; at the same time, myths, by weaving in observable features in the landscape (a tree here, a rock there), strengthen a people’s bond to place.” (1991, 686)

This is an important move, since it enables a more discursive view of place that is constructed by actors through social interaction. This humanistic concept of place would later be picked up by geographers interested in structuration theory and by those working in the now burgeoning field of cultural geography, leading to a further specification and transformation of the concept of place. Of particular importance and the major strength of this humanistic tradition is its emphasis on people’s emotional attachment to places. Refusing to interpret this attachment as merely an expression of political and economic considerations by rational actors, this non-reductive approach to place acknowledges the need many actors feel to relate to something they can call their place. By stressing this human aspect, humanistic geography has in my view more than any other tradition made explicit the difference between location and place. Whereas a location can be seen as only a delimited physical area embedded in abstract space, place is always humanized i.e. it is a location to which meaning has been attached.

Despite all this praise, the humanistic version of place is by no means unproblematic. For one, it is often accompanied by a tendency to neglect power differentials, constructing a nostalgic view of place that conflates place with community. Relph, for example, used the term placelessness to describe the loss of place specificity that he saw as a result of industrialization and modernity. From such a perspective, the modern world can only seem as a threat to the existence of authentic communities and places. What this ignores is that these places and communities themselves are often highly structured and hierarchical. Therefore, while attending to the humanizing aspect of place-construction, one should at the same time address the fundamental political aspect of place. It seems to me that the main reason humanistic geography finds it difficult to discuss these questions of power has to do with the fact that it sees place-construction as an immanent process. Too often actors are seen as appropriating i.e. humanizing elements of the land on which they live in order to create an attachment to this place – to create a sense of place. It ignores, however, the external face of place; the fact that actors construct certain places to re-define their own position towards the larger spaces they also inhabit. This difficult topic will be dealt with in the chapter on space and scale.

Another line of criticism against spatial science can be summarized under the heading of structuration theory. Although other writers such as Bourdieu and Bhaskar[20] have left their mark on the discipline of geography, it is the work of Giddens that in the 1980s deeply influenced most geographers interested in theorizing the spatiality of social life.[21] Giddens identified the central problem in modern social theory as a dualism between “agency” and “structure.” Instead of analyzing structure as a framework in which different actors are able to live their lives, he argued that structure is and should be analyzed as implicated in every action. In other words, it is only through human agency and their constituting moments of interaction that structure is reproduced and transformed. This is what Giddens calls the recursive character of social life: every act of production is at the same time an act of reproduction. It was mainly Giddens’ own explicit concern to ground his social theory in a spatial context that separated him from most other social scientists and that made him such a central figure among geographers during the mid-1980s. However, despite Giddens’ initial popularity, his structuration theory in general and his conceptualization of space and place in particular know a number of deficiencies, two of which will be addressed here: first, Giddens’ conceptualization of the locale as a setting for interaction, and second, the connection of this locale to larger structures.

Giddens argues that the continuity of everyday life largely depends on routinized interactions between people who are co-present in time and space. He uses the locale-concept to describe the spatial setting within which these interactions occur. According to Giddens (1979, 206-07):

“Virtually all collectives have a locale of operation, spatially distinct from that associated with others. ‘Locale’ is in some respects a preferable term to that of ‘place’, more commonly employed in social geography: for it carries something of the connotation of space used as a setting for interaction. A setting is not just a spatial parameter, and physical environment, in which interaction ‘occurs’: it is these elements mobilised as part of the interaction. Features of the setting of interaction, including its spatial and physical aspects, […] are routinely drawn upon by social actors in the sustaining of communication – a phenomenon of no small importance for semantic theory.” (Italics in original)

Besides Giddens’ tendency to see a locale as occupied by one collective - thereby excluding the more realistic situation of a locale occupied by a number of (overlapping) collectives - there are other, more serious problems to his characterization. One is his equation of locales as settings for social actors in interaction. This, however, externalizes the formative influences of the environment and politico-economic factors such as state politics, warfare and capitalist restructuring processes on the character of these settings. Also, this approach reduces the concept of locale to one of physical location and thus misses out on the discursive representations of a locale that is constructed by specific actors. From a slightly different perspective, Kilminster (1991, 99) makes a similar criticism: “Despite the word relations, individuals are seen here only in the first person, as positions. There is no conceptual grasp of the perspective from which they themselves are regarded by others in the total social web, nor of their combined relatedness. Structuration theory is a one-dimensional view of society that does not permit the sociologist to show this combined interplay of relations and perspectives in all its richness and complex balances of power” (Italics in original). What Giddens also ignores is the fact that social actors through their (inter)action reconstruct the locales they inhabit. Although he does seem to argue something similar - he writes that actors in their social interaction “mobilize” spatial elements – he does not follow through on the consequences of this mobilization, which is that this mobilization can also lead to a transformation of these spatial elements. Here, Giddens’ theory seems limited by a view of space reminiscent of the earlier tradition of spatial science, with a locale as nothing more than a spatial location and with actors involved in the application (not transformation) of these elements in daily life. Although Giddens in general succeeds in retaining a tension between structure and agency – that was after all the main goal of his structuration theory - in his theorizing of the spatiality of social life he certainly accords too much power to a stable and indifferent spatial structure.[22]

That said, Giddens is very much aware of the connections a locale has with larger structures, but unfortunately these connections are too often theorized in a too restricted way. For example, Giddens accords central place to “dominant locales.” These should be seen as physically demarcated settings for interaction – “power containers” - that provide the major structural principles of a society. However, since Giddens puts this concept of the power container in a highly general developmental scheme – the change from city-states as centers of power to the development of the nation-state[23] – he fails to clarify, as Gregory has pointed out, “the hierarchy of locales involved in these societies (and beyond them): the ways in which […] domains nest within one another, overlap and confound another, in a changing web of interactions” (1989, 209). Although Giddens certainly does not see the state or a society as a unitary totality, through his bounded view of the power container called the state that encompasses all other locales, he tends to ignore other sources of social power that know spatial structuring processes of their own and to re-route these processes as being only a part and result of the state.

III. Space and Scale

Although Giddens in his influential work of the 1980s certainly tried to align his structuration theory with conceptualizations of space and place, his view of locales and of the power container called the state seem both too bounded and too stable to be of analytical value in a world in which locales are connected in highly complex ways. This is not to say that in a global world “everything flows,” as Castells (1996) would have it, but it is to say that every critical analysis needs to address those expressions, phenomena or practices that cannot be contained by the state.

It seems to me that one of the reasons Giddens largely fails to pursue a non-state-centered approach has to do with his difficulties in theorizing space. Despite his extensive use of a spatial vocabulary – regionalization, locale, front regions and back regions, time-space distanciation and time-space edges, among others – he remains indebted to a concept largely derived from spatial science that sees space as abstract, universal and absolute. This is reflected in his concept of time-space distanciation that refers to the processes by which societies are “stretched” over time and space (Giddens 1981, 92-97). Although this concept is clearly an acknowledgment of the complexity of modern social life – an attempt to theorize the influence of world-wide media, travel and migration on daily life and societies – it nevertheless makes social life seem simpler and more logical as it probably is, since it is couched in a covert thesis of modernization that implies the progressive dissolution of separate locales. Interpreting globalization in this way, however, reduces it to a question of physical and material integration, and ignores other conceptions of space – in particular discursive representations of space – that are not reducible to this material integration.

In this chapter I will address these problems concerning the theorizing of space and develop a conceptualization that does justice to the idea of space not as objective and abstract, but as subjective and social. Since the theoretizations of space are far too diverse to summarize under a single heading and after having discussed in the last chapter three lines of critique against spatial science – traditional regional geography, humanistic geography and structuration theory – I will here start my discussion with the fourth line of critique – Marxist geography – and from there develop a concept of space that is more discursive than the Marxist version usually allows.

Critical of the objectivism of spatial science and the non-theoretical nature of space in traditional regional geography, since the 1970s geographers started to develop concepts of space that were no longer absolute – i.e. simply there as a framework in which social life took place – but where space was seen as constituted in social practice. Similar to the move in theorizing about place and region, these new conceptualizations of space were guided by the acknowledgment that spatial analysis required socialization and that social analysis required spatialization. The traditional approach saw places as embedded in larger spaces, but the space-concept itself was usually left untheorized and the connections between space and place too often remained unclear. Many of the first attempts to reconceptualize space made extensive use of Marxist theory, leading to the development in the 1980s of a highly materialist history of space. The general argument usually was that capitalism produced spaces because of its expansionary nature, which brought along very specific processes of re-territorialization in order to make these new spaces suitable for the penetration of a capitalist space-economy.[24]

This was an enormously important contribution to the theorizing of space, but one of the major drawbacks was that many of these analyses offered a top-down approach to the explanation of socio-spatial change. Privileging capitalism as its explanatory locus, actors appeared as nothing more than spatial positions waiting to be engulfed by the next wave of capitalist expansion. The formative influence of cultural practices on the production of space was almost always insufficiently theorized. In this part I want to discuss the concept of time-space compression as developed by Harvey – one of the main representatives of the Marxist approach in geography – in order to get a grip on this conceptualization of space. After pointing out some of the weaknesses in his argument, the next part will be dedicated to the exploration of a concept of space that is more attentive to culture and agency.

In The Condition of Postmodernity (1989), Harvey offers a synthesizing history that tries to explain the state of the world we are living in. The main argument of his book is that, beginning around 1972, there has been a “sea-change” in political, economic and cultural practices, which has led to the emergence of a postmodern sensibility.[25] In causally linking postmodernism to capitalism, Harvey’s approach is similar to the famous article by Jameson (1984) that interprets postmodernism as the “cultural logic of late capitalism.” Both argue that postmodernist capitalism causes, on the one hand, fragmentation – leading to the construction of multiple and simultaneous ontologies, “a potential as well as an actual plurality of universes,” as Harvey calls it. The “disruptive spatiality” that is the result of these multiple ontologies leads to the impossibility of constructing any coherent, general narrative that could describe this fragmentation (Harvey, 301-302). On the other hand, confronted by precisely this fragmentation, people want to reach out for anything that gives them security in a highly unstable world. Harvey sees mainly dangers involved and interprets this need for attachment as an “aesthetics of space,” which tends to lead to us celebrating the “fetishisms of locality, place, or social grouping, while denying that kind of meta-theory which can grasp the political-economic processes […] that are becoming ever more universalizing in their depth, intensity, reach and power over daily life” (117).

That is in a very general way the argument of Harvey’s book. Here I would like to concentrate on two elements that are of particular importance in theorizing space and that direct us to the limits of Harvey’s approach: the concept of time-space compression and the issue of positionality. Time-space compression should be seen as the product of what Marx once called “the annihilation of space and time” by those processes acting according to the rules of commodity production and capital accumulation. According to Harvey, the term implies “processes that so revolutionize the objective qualities of space and time that we are forced to alter, sometimes in quite radical ways, how we represent the world to ourselves” (240). By explicitly referring to the changes that took place in representing space and time, Harvey incorporates an experiential dimension that is missing from the more analytical concept of time-space distanciation by Giddens. But it is exactly here – the experiential dimension that I see as subjective and social - that Harvey’s concept is too restricted. It seems unlikely that time-space compression is only the result of capital accumulation and the expansionary nature of capitalism. Here, Harvey’s argument clearly suffers from an economic reductionism that reduces all other important factors in shaping time and space – such as politics and culture – to the ultimate logic of capital. Second, a result of this reductionism is a complete failure to socially differentiate the concept of time-space compression. In Harvey’s world-view the social only exists as part of the global restructuring flows. All activity is based in capital, whereas actors are seen as on the receiving and passive end. To counter this tendency, Doreen Massey (1993, 61) has argued that one should attend to “the power-geometry of time-space compression:”

“For different social groups and different individuals are placed in very distinct ways in relation to these flows and interconnections. This point concerns not merely the issue of who moves and who doesn’t, although that is an important element of it; it is also about power in relation to the flows and the movement. Different social groups have distinct relationships to this anyway-differentiated mobility: some are more in charge of it than others; some initiate flows and movement, others don’t; some are more on the receiving end of it than others; some are effectively imprisoned by it.” (Italics in original)

There is a more general problem with Harvey’s approach, which one could describe as a refusal of positionality.[26] Despite his own analysis that interprets postmodernity as creating fragmentation, leading to an impossibility of constructing a coherent, general explanatory narrative, Harvey does precisely what his own interpretation forbids. He holds on to Marxist historical materialism as the ultimate explanatory framework and relegates other viewpoints to a subordinate place within this framework. By doing so, Harvey – as a Marxist analyst – is able to organize the world for him in an understandable way, providing, as Hall once remarked when commenting conventional Marxism and its interpretation of culture, “a way of helping you sleep at night” (1988, 72). What this top-down approach does not do, however, is to construct a view of (global) change that is socially informed. One could even argue, as Deutsche does in a marvelous fashion, that this kind of theorizing leads not to an explanation of societal and spatial changes, but that instead it leads to an explanatory image of the analyst himself. Instead of attaining a “whole view” by observing from a distance, Deutsche argues, these visual and spatial strategies themselves are constituted by a “desire for exteriority, and the image that they produce is not a reproduction of the world but a self-image” (1996, 213).

Acknowledging the critique directed at Harvey’s approach has a number of implications for a more grounded view of space. The centrality of situated subjectivities demands a view of space that is itself situated. Instead of conceptualizing space as abstract and universal – as in spatial science – or as produced by capitalism – as in Harvey’s approach – I would like to conceptualize space from the bottom up. This is, of course, a very general statement and needs to be specified, since it raises the question how places and spaces should be connected. Although I tend to agree with the view that the construction of places is increasingly dependent on – or at least influenced by - other spaces and that this can have the result that places are “drowned by other more powerful abstract spaces such as communications networks,” leading to a situation in which the “truth of an experience no longer coincides with the place in which it takes place” (Dear 1997, 58), one should at the same time attend to the other side of this argument. This side – one which is all too easily forgotten – is that actors in places are not passive, as Harvey would have it, but that these actors themselves make imaginative use of the increased availability of spaces in order to pursue certain goals and to take a stand among other social actors.

This way of conceptualizing space enables an open and political view of place, in which actors seek to reconstruct their place by reaching out to spaces – discursively as well as in social practice (through the establishment of networks). It is here that the limitations of the humanistic concept of place become clear, since in the here proposed view place-construction is not an immanent process, but, on the contrary, a construction dependent on its association with larger spaces – an externalized view of place. In this thesis I will often use the term spatial imagination as shorthand for these processes of association. Spatial imaginations are thus specific ways of imagining and / or materially constructing the embedment of your place in larger spaces. Obviously, this conceptualization avoids a conflation of community with place and chooses to see place-based political struggles not as the result of external destabilizing influences, but instead as struggles over the right to re-define the hegemonic character of place.

Taylor has usefully characterized this situation as one of place-space tension (1999, 12), but he gives a twist to it that I would like to avoid. According to Taylor, place-space tensions arise when different persons treat the same location in different ways. This is certainly correct and overlaps with my view of place-based political struggles. Unfortunately, his conceptualization suffers from a dichotomization – the result of his reliance on Tuan’s book Space and Place - that sees place as enabling and space as repressive and disenabling. The main problem with this view is that it retains a concept of space as inhuman and abstract, whereas I have argued for a view of space as social and subjective. As I see it, the tension arises not because “the same location can be both place or space depending on whose perspective is involved” (12), but it arises because of differences in spatial imaginations. There is no such thing as a place without space. The limits of his conceptualization become clear when he discusses the role of the state. In Taylor’s view the state is seen as converting places into spaces through processes of bureaucratization and territorialization. The later rise of nationalist thought made it possible to combine state and nation, merging national territory with an imagined community, which in effect reconstructed places out of the national space, creating highly effective and powerful entities. Without a doubt something like this occurred, although one can question the analytical value of such a sketchy approach. The result of this approach, however, is that the state as a spatial entity remains curiously unexamined. In Taylor’s view, bureaucrats working for the state produce spaces for no apparent reason other than as a means to create territories with populations to be taxed (14). Certainly this plays an important role, but such a view reduces the importance of cultural formations to a question of politico-economic considerations. By doing so, spatial imaginations are once again excluded.

The next part will address this problem by emphasizing the importance of scale in thinking about the state. As I see it, differences over place – my characterization of place-space tensions – often arise because different social actors occupy or have access to different scales. It is this difference that plays an important role in delimiting actors’ spatial imaginations.

Geographic scale has long been treated as a relatively fixed hierarchy of bounded spaces. Each scale represented a different level of analysis (local, regional, national, global) in which different (social) processes took place that could be analyzed. This approach has its analogy in the particular-general discussion of anthropology,[27] but has been increasingly challenged in the last decade or so by a number of (mostly political) geographers. Central to this new body of thought is that geographic scale is no longer pre-given or simply a conceptual tool of the geographer, but that it is actively appropriated by social actors as part of their arguments and their practice in order to persuade others, “to create in the minds of others a kind of mental map or image of the difference that scale makes” (Delaney and Leitner 1997, 94).

It is this spatial imagination, which always involves a strategy of re-scaling, that constructs a specific view of the spatial embeddedness of place. Since every view necessarily includes a certain perspective but excludes others, politics is essentially concerned with the promotion of specific spatial imaginations at the expense of others. According to Swyngedouw:

“These struggles change the importance and role of certain geographic scales, reassert the importance of others, and sometimes create entirely new significant scales, but – most importantly – these scale redefinitions alter and express changes in the geometry of social power by strengthening power and control of some while disempowering others.” (1997, 169)

Thinking about politics in this way clearly requires an adaptation of those discussions about politics and the state that interpret the state as the logical vehicle of political processes. In the view that is being sketched here, no longer can the scale of the state be taken for granted, but instead it has to compete with other spatial imaginations that make use of different scales. This means that one of the main functions of state representatives – in particular, the bureaucracy, the military and often the media – is to provide the people living on its territory with spatial imaginations that can successfully compete with these other non-state spatial imaginations.

In a recent article, Gupta and Ferguson (2002) – working within an anthropological tradition, but appropriating geographical theories - have discussed some of the ways in which states represent themselves – or, more precisely, how state representatives represent the state – as reified entities with particular spatial properties. From this perspective, not only are nations “imagined,” an argument made by Anderson and a host of others,[28] but states are as well. Two images are of particular importance; those of verticality and encompassment: “Verticality refers to the central and pervasive idea of the state as an institution somehow ‘above’ civil society, community, and family. Thus, state planning is inherently ‘top down’ and state actions are efforts to manipulate and plan ‘from above,’ while ‘the grassroots’ contrasts with the state precisely in that it is ‘below,’ closer to the ground, more authentic, and more ‘rooted.’” This image is combined with the image of encompassment: “Here the state (conceptually fused with the nation) is located within ever widening series of circles that begins with family and local community and ends with the system of nation-states” (982). Together these images are of central importance in the state’s attempts to naturalize its authority and position itself as “above” other centers of power.

Although Gupta and Ferguson question this state-image of vertical encompassment in order to develop a Foucault-influenced concept of transnational governmentality, let me here point out another implication of thinking about scale and the state. Taking scale seriously means that every analysis has to acknowledge the, what I would call, scalar positionality of spatial imaginations. First of all, this means that social actors always position themselves towards a particular spatial imagination. One could describe this as an anticipatory act, since it involves an association of actors with certain spatial imaginations in order to “improve” their own place. This is the imaginative part. Secondly, however, actors are never free to choose whatever spatial imagination they like, since they do not live in a void, but are obviously part of a specific socio-spatial structure. Actors do position themselves towards spatial imaginations, but at the same time others position them. Attending to the question which actors have access to what scale provides one way of understanding the politics of certain spatial imaginations and the power differentials implied by them.

To close this theoretical section, a short summary of the central concepts used in this thesis will be given. The term scalar positionality of spatial imaginations should be seen as an attempt to incorporate the concepts of location, place, space, scale and positionality within an all-encompassing shorthand and to emphasize that these concepts cannot be seen in a clearly distinct way, but instead should be seen as fundamentally interconnected. Nevertheless, for the sake of simplicity and analytical reasons they will be used as separate concepts to illuminate certain aspects of a multifaceted social practice.

Location is the most basic concept, but at the same time the most elusive. This is because the concept is supposed to describe a delimited geographical area before any meaning has been attached to this area. That is, of course, a non-possibility. The very fact of delimitation involves a human observer that does the delimiting. Nevertheless, in order to retain at least a minimum amount of ground under my feet, the concept will be pragmatically used in this thesis as a rough reference to a geographical area somewhere on this planet. Thus, if I write the location of Addis Ababa, my goal is not to provide a clear definition of how to delimit this city, but it is simply meant as a starting point for analysis.

The concept of place, then, humanizes i.e. attaches meaning to this location. A place, in other words, always evokes aspects of subjectivity, such as “belonging” and “home.” Where my concept differs from the humanistic version is that I emphasize the open and political character of place. Instead of seeing place-construction as an immanent process of attaching meaning to a location by appropriating elements already there, as it were, my version sees place-construction as a process of place-embedment within larger spaces not simply in place.

This is what I have described as a view of space “from below” and that should be seen as both a critique of and complement to the top-down version as proposed by many Marxist geographers. The concept of positionality has been offered to describe this double-sidedness of actors: being embedded in a certain space, but at the same time seeing yourself as part of other spaces. It offers a way of connecting local politics with global changes and imaginations. Thus, instead of interpreting globalization as a momentous change leading to homogenization and the destruction of local cultures, in this thesis I choose to put the agency of specific actors in the center of attention. The shorthand spatial imagination will be used to refer to the specific ways actors imagine and / or materially construct the embedment of their place in larger spaces.

Finally, the last part of the theoretical section has tried to bring both the state as well as the question of scale back into the picture. Making use of scalar theory, I have argued that geographic scale is not simply pre-given or a conceptual tool, but that it is appropriated by specific actors in order to promote a certain spatial imagination at the expense of others. From this perspective, the state is nothing more than one of these spatial imaginations. Since switching scales involves a redefinition of power – including some, excluding others – attention to the scalar positionality of spatial imaginations provides one way of analyzing the politics behind these imaginations.

- Oromia, the OLF and the Ethiopian State -

IV. The Center, its Peripheries and Spatial Imaginations

The past is of major importance in the understanding of the Ethiopian present. Of course, to a certain extent this is true for all countries, communities and biographies all over the world, but in the Ethiopian case highly divergent representations of history are used on a daily basis as strategic weapons: to politicize or to personalize certain narratives in order to take a stand – to develop a specific positionality – among and against other social actors.

This chapter will focus on the more material politico-economic developments that have led to the construction of this specific form of the Ethiopian state and in which the current actors are still, if only partly, embedded. On a general level, two interconnected changes have been of central importance: first, the internationalization of the Ethiopian state and second, its centralization and territorial expansion. Internationalization, as the name suggests, is dependent on connections between nations - which, in this context, means states - and more precisely on a connection that acknowledges the state’s right of existence by other states. In that sense, the Ethiopian state as such did not really exist before Menelik II, whose victory over the Italians at the battle of Adwa in 1896 led to a range of international agreements that endorsed the sovereignty of Ethiopia. The centralization and territorial expansion of the state was directly connected to this changed international embeddedness. It led to the incorporation of Oromo-speaking and other non-Oromo societies into the modern state in a relatively short period of time and was driven by Menelik’s success in the international trade of southern products. In this process, the southern peripheries were connected to the center in a number of ways. Northern traditions of land-control were partly transplanted onto these southern areas and its peoples and this led to the development of specific hierarchies with actors occupying a range of socio-spatial positions between, as it were, the land and the main urban center of Addis Ababa. This will be discussed in the first two parts of this chapter.

The third part should be seen as a preparation to the next chapters whose focus is on the question of spatial imaginations rather than material spatial structures. To a certain extent, it involves a critique of the center-periphery concept as described above. In my view, the main deficiency of this concept is that it involves too much structure and too little movement. Although Donham acknowledges that “local histories were never simply determined by the centre” and that “what happened was also influenced by the organization of local societies” (1986, 44) this conceptualization ignores those more powerful actors that can move from the periphery to the center and beyond. Through a conceptual critique of the frontier thesis – the frontier as a border area between the center and periphery – I want to point out some of its limitations and open up the spatial imagination of Oromia to questions of representation and power, which will then be addressed in the following chapters.

The territorial shape of the Ethiopian state until the revolution of 1974 was largely the result of processes that took place at the end of the nineteenth and the beginning of the twentieth century. Between 1769 and 1855, so the story usually goes, the Abyssinian Kingdom that would later form the core of the Ethiopian state was largely governed by a series of factional wars.[29] It was only after a governor called Kassa was able to seize power by defeating and incorporating his other regional competitors that one could again start to speak of Abyssinia as a material entity and not merely as a spatial imagination. By crowning himself “King of Kings” and taking the title of Emperor Tewodros II, he consciously positioned himself as part of the ancient Solomonic legend, tracing the legitimacy of his rule to Menelik, the son of Solomon and Sheba, who is said to have brought the Ark of the Covenant to Ethiopia. According to Levine, Kassa renamed himself Tewodros, since it was prophesied in the Fikkere Iyesus (the Interpretation of Jesus) that after a period of divine punishment a just and popular king called Tewodros would come to the throne and rule for forty years (1974, 157).[30] Rhetoric and power have always been intimately connected in Ethiopia.

Crummey (1969) has argued that Tewodros II was far more serious than previous rulers in seeking to harness expatriate technology to the tasks of strengthening the state and economy. Considering later developments, one is tempted to say that the modern Ethiopian state was an international project from the day it was born, since its survival has always depended on the ability to cultivate these connections.[31] This internationalization of the Ethiopian state[32] increased when Menelik II came to power in 1889. The famous battle of Adwa in 1896, where Menelik’s army defeated the Italians confirmed on an inter-national scale the right of the Ethiopian state to exist as part of the international state system. Although Italy was allowed to maintain possession of Eritrea, it had to acknowledge the sovereignty of Ethiopia, enabling Menelik in the following years to enter into agreements with France, Britain and the Mahdists in Sudan, in which the territorial integrity of his state was ensured (Keller 1981, 528).[33] This international recognition was paralleled by measures to develop and stabilize national institutions in Ethiopia. With financial support of France and Russia the military was professionalized and as early as the 1890s young Ethiopians were sent abroad for studies to Europe, Russia and Sudan. These students would later occupy central positions in the bureaucratic and military institutions of the state (Zewde 2002, 20-34).

It is important to note that this international recognition was one on the scale of the nation-state (the terms are conflated here) and it involved the construction of a spatial imagination that was to become hugely popular among many movements around the world. The battle of Adwa, the myth of a Solomonic lineage and the later coronation of Haile Selassie were all connected to the spatial imaginary of “Ethiopia” and for a long time provided important identity-markers for many Afro-Americans, Pan-Africanists and Rastafarians.[34] Notwithstanding the liberating possibilities this imagination might have had for these actors in the face of suppression and inequality, in the case of the Ethiopian state the effect has surely been a “flattening” of history with the majority of interpretive frameworks implicitly re-scaling all social practices in order to make them fit with and part of a united image of the state.

This re-scaled spatial imagination, however, was not merely the result of fantasy. It was also consistent with material re-scaling processes that took place during Menelik’s rule, since the acknowledgment of Ethiopia as a legitimate state involved a re-routing of international connections through the mechanisms of the state. Representatives of scales below the state were now increasingly confronted with state structures that had backing from the international state system and thus privileged access to the world of commerce and arms.

Haile Selassie, first as a regent (1916-30) and then as emperor (1930-74)[35] continued and intensified many of the policies pursued by Menelik. In the attempt to internationalize the Ethiopian state, he was most successful after Mussolini’s invasion and occupation of Ethiopia from 1936 until 1941.[36] The intervention of British military troops and their training of the new national army re-established and even increased his power that had been eroded by the occupation and the strategic alliances between the Italians and those excluded or disadvantaged under the rule of Haile Selassie. His return to power was followed by a political strategy that sought to capitalize on international support in order to retain and expand his domestic control. The polarizing logic of the Cold War in geopolitical considerations certainly facilitated this process. With the USSR supporting Somalia, the United States developed a “patron-client relationship” with Ethiopia (Lyons 1990, 169). In 1953 this relationship was formalized in a series of agreements, guaranteeing US military assistance and resulting in Ethiopia receiving more than $200 million in military and economic aid over a twenty-year period, in exchange for continued access to the US communications facilities at Kagnew.[37] By the end of the 1960s, the Ethiopian state received about 60 per cent of United States military aid for all of sub-Saharan Africa (Tareke 1991, 52).

This embedment of the Ethiopian state in international relations enabled a greater control of different kinds of peripheries by the center (i.e. areas made peripheral relative to the center of the state). In the theoretical section I mentioned the importance of geographical features in the making of Ethiopian history. From a perspective of political economy, one is struck by something similar, namely the importance of land and the mechanisms of control over this land (and thus the people on it) as a central explanatory paradigm of Ethiopian history and society. Political centralization has surely been the most obvious feature of Ethiopia for the largest part of the twentieth century, but this can only be understood by relating it to the question of control over land.[38]

To understand the situation in southern Ethiopia, one has to grasp the social and material relations active in the Abyssinian heartlands, since these were later partially transplanted onto the newly conquered Oromo and non-Oromo inhabited peripheries. The question of land-ownership was central to the situation in the north and part of an earlier debate among academics about the usefulness of the feudal paradigm for explaining Ethiopia.[39] In principle – and this speaks against the feudal paradigm – it was the peasantry that owned the land. Not only did they own the land, in most cases they even had direct access to the associated means of production, which in plough-cultivating Abyssinia usually meant access to oxen. These rights were called rist and were hereditary in an ambilineal way. In other words, both the father and the mother could pass them on to their children. Although this could be seen as a sign of gender equality, in practice this ambiguity in inheritance opened up the question of land-ownership to politics, since it regularly led to one side of the family pressing claims against the other. These claims were dealt with in peasant corporations, but the communal character of these organizations should not be overestimated, since it was here that the land-question was integrated into a hierarchy that was focused towards the Emperor and the central state. Ellis has described this integration as following: “In order for one descent group – or for an individual – to press land claims against another, or for two claimants to resolve a conflict, it was necessary to appeal to the judiciary; and the judge was, of course, the gult holder. In such circumstances, it is understandable that there were a number of ‘honorary’ members of the descent group who were allocated lands” (1976, 281). The privilege of gult – which indicated a granting away of “sovereignty” (Donham 1986, 9) – was handed out by the central state and in effect constructed a military-administrative structure on top of the rist system.

This was the major difference between the Abyssinian mechanisms of control over land and the European era of feudalism. Whereas in Europe a small elite owned the land, forcing the majority of the population into a situation between slavery and wage labor (and into the growing cities), in Abyssinia the peasants controlled the land, leading to another form of domination by the elites. The gult-system provided one way of governing a large territory without actually having direct access to the land. The central state carved up all the land in gult estates and distributed these areas among loyal actors. In return for this favor, the gult holder could enjoy judicial rights over the subjects inside his territory, collect tribute from the peasants working the land and maintain a private army for his own security as well as for the center. In contrast to the system of rist, gult privileges were not hereditary. The monarchical center could and did take gult privileges away in order to give them to other, more loyal actors. This increased the power of the center and divided the actors agitating at scales below the state, creating a highly individualistic society with numerous attempts at personal advancement. Ironically, this hierarchical patron-client structure was also the main impetus behind the pervasive idea of social mobility that has been observed by many commentators.[40] In principle, everyone could become part of the nobility and receive a gult by showing deference to the Emperor. Crummey has used the metaphor of the pyramid to explain these processes: “Strongly hierarchical, Abyssinian society was not rigidly stratified in the sense that one social layer was sharply marked off from its neighbours. Quite the reverse. The bonds of patron-client relations were close; class relations were fluid. Abyssinians perceived their society as mobile” (1980, 137). At the same time, in the core areas a status quo between the peasants and the nobles seems to have developed, with the peasantry accepting a certain surplus-extraction in return for security and as protection against further exploitation by those directly working for the Emperor. This in turn enabled the nobility to claim gult lands as hereditary and a rejection of this by the Emperor would often cause rebellions by peasants that saw their relative security threatened. In other words, the granting away of sovereignty in the form of gult tended to create relatively autonomous bases of power below the scale of the state. Donham has argued that this situation provided one major internal dynamic leading to territorial expansionism. With the core areas of the state increasingly in the hands of actors acting below the scale of the state, an Emperor that wanted to stay in charge had to enlarge his empire and offer new gult grants on these new lands (1986, 13-17).

The rule of Menelik II and the territorial expansion at the end of the nineteenth century was, however, surely supported by the political-economic possibilities created by the larger, international context. With most of the northern regional rulers of the time occupied with the expansionist activities of the Europeans and Egyptians at the coastal area since the 1870s, Menelik – whose own region of Shewa was far enough from the coast to be relatively safe – could increase his own power relative to other regional competitors. Not war, but trade was to be the solution in dealing with the Europeans and Egyptians. The opening of the Suez Canal in 1869 offered major possibilities for Menelik, since he could now directly export the products he appropriated from the areas south of the Abyssinian highlands (Tareke 1991, 39-40).

The territorial expansion connected to this international trade created a situation, in which the highly diverse southern areas were connected to the center through various means. In a simplified way, one can distinguish three different kinds of linkages: semi-independent enclaves, gebbar-areas and fringe peripheries.[41]

Rulers of the semi-independent enclaves retained a certain amount of local power and were connected to the central state through the practice of tributes. In the case of the Oromo, the kingdom of Jimma that has been described by Lewis (1965) belongs to this category. These enclaves were, however, increasingly incorporated into the mechanisms of the Ethiopian state, reducing them to the status of gebbar-areas (gebbar is Amharic for peasant). With the return of Haile Selassie and the bureaucratic reforms of 1941, these areas were made part of the regularized territorial administration.

The second linkage involved a partial transplantation of the northern system of rist and gult on southern societies. The principle of control was the same as in the north, but whereas in the north it had been established as part of a centuries-long struggle between peasants and nobility, in the south it was largely imposed on people in the highland-areas suitable for sedentary production and surplus extraction during the rule of Menelik. As part of this territorial expansion gult grants were given to the northern nobility and soldiers, the so-called neftennya (Amharic for riflemen), who conquered and settled in these areas. Whereas in the northern core this system was stabilized through the ambilineal system of kinship and the idea of social mobility that tied both nobility and peasants together, in the south this stability had to be created in different ways. First, military force underlay the territorial expansion and the later system of extraction. The neftennya were located in fortified towns in strategic areas, from where the southern peripheries were further penetrated and the land was divided among the state, the Ethiopian Orthodox Church and the balabbat. Besides the military, the balabbat were the second pillar of stabilization. This term, which in Amharic literally means “he who has a father” (Haji 1994, 587), referred in the north to a local notable that occupied a position between the governor and the peasantry. In principle, this was the function they fulfilled in the south as well.[42] Local men of standing were incorporated into the mechanisms of the state through the distribution of gult grants, which gave them the privilege to take tribute from the peasants working the land. In return for this privilege, balabbat fulfilled the function of intermediaries between the northern settlers and the southern peasants, doing work that ranged from translation to the allocation of gebbar to plots owned by the nobility.

The third kind of linkage was established with the lowlands areas that were not suitable for sedentary cultivation and were inhabited by pastoralists. In these areas, balabbat were appointed, but their work was usually limited to demanding a fixed tax on cattle and, in the early twentieth century, the raiding for slaves. In these areas, the existence of an international border with Sudan, Kenya and Somalia paradoxically presented a distinct advantage for pastoralists, since it could be crossed in times of exploitation by Ethiopian officials and used as a temporary refuge. The Boran Oromo that are nowadays seen by many Oromo nationalists as “the cradle of Oromo civilization” belong to this third category of linkage with the Ethiopian state.

Besides these connections between the center and its different peripheries that largely concerned the question who controlled the land, two other intertwined processes became increasingly important during the twentieth century: the bureaucratization and commercialization of the Ethiopian state. The specificity of the Ethiopian state was that these modern changes did not radically alter the different relations of dependency at work. In the early twentieth century, a modern state was certainly being constructed, but the result was mainly an institutionalization of the pre-existent spatial structure through the introduction of a standing army, the foundation of a permanent capital and the construction of infrastructure and communication facilities. In that sense, Crummey’s metaphorical use of the pyramid to describe the interactions between peasants, gult holders and the Emperor still seems useful in the context of the modern Ethiopian state. Strongly hierarchical, bureaucratization and commercialization can be seen as attempts by the Emperor and the small and dependent urban elite to increase the power of the scale of the state at the expense of sub-state scales. The completion of the Addis Ababa – Djibouti railway in 1917 symbolized this growing importance of the center. From now on, the southern areas would contribute around 70 percent of the total external trade, whereas the northern areas lost in economic significance (Zewde 1987, 77). This trade was controlled by members of the nobility and the royalty – in itself a sign of relative homogenization of these groups – together with foreign traders (Hiwet 1975, 38-44).

After the return of Haile Selassie in 1941, the subsequent administrative reforms and the military and financial back-up from the international state system ensured a domination of Addis Ababa and the state bureaucracy - largely located in the capital - in the political economy of Ethiopia (I will come back to this in the next chapter). For those interested in gaining wealth or prestige, there were now two options available. One was the older strategy of establishing a local power-base with a certain amount of autonomy from the Emperor or central state. The other and new option was to leave the countryside in order to participate in the commercial and urban life of Addis Ababa or abroad. Local credibility was no longer needed.

This material history of internationalization and centralization of Ethiopia is now relatively well-known and has been told by others in a more detailed way. What has not so often been made explicit, however, is that these socio-spatial developments not only changed material relations of power between the center and its peripheries, but that they also led to the production of spatial imaginations that are not simply reducible to this material spatial construction.

Using a center-periphery model to describe Ethiopia surely captures many important relational dependencies and this clearly makes it a useful and popular paradigm,[43] but it does have a number of limitations. One of these is that such a model tends to exclude the fact that actors are always involved in multiple relations of dependency and thus not reducible to the structuring processes of one of these,[44] but this can in principle be addressed by incorporating a number of models in the analysis of a specific case. A more important limitation is that the development of social and material relations that are the result of a powerful center is always necessarily accompanied by its own deconstruction. For example, although it is surely true that in Ethiopia a center-periphery model was constructed with Addis Ababa functioning as the hub of exploitation, it is also true that the very fact of movement (material goods, governors, ballabat, students) between these two poles partly dissolves the border that was needed to separate them in the first place.

This brings us to the question of Oromo ethno-nationalism. Although many anthropologists have been highly sensitive to local processes of identification and change among Oromo societies, the conceptualization of Oromo ethno-nationalism is regularly hampered by the implicit or explicit postulation of a continuous link between the rural Oromo societies and those actors producing the spatial imagination of Oromia. As I see it, however, Oromia is one of those spatial imaginations that cannot be reduced to a center-periphery structure, since its actors were and are involved in a range of spatial practices that differentiate them from the rural people they claim to represent. To close this chapter and provide a number of questions with which to open the next chapter, let me here discuss some of the conceptual limitations of the center-periphery approach in analyzing Oromo ethno-nationalism by referring to an article by Triulzi (1994).

In this article, Triulzi tries to interpret Ethiopia as a frontier society – the frontier as a border area between center and periphery[45] - and uses this image as an explanatory framework for understanding the “Amhara-Oromo dispute.” On the one hand, Triulzi argues that a correct view of the frontier “involves the recognition of its being not just a fixed border line, but rather a complex multiplicity of frontier situations” (238, Italics in original), which is surely correct and emphasizes the anthropologist’s concern for detail – the ethnographic stance, as Ortner (1995) once called it. On the other hand, he still acknowledges the factual existence of this Ethiopian frontier line, which to him is mainly the result of the expansion of the Ethiopian state and the incorporation of other societies in the process. This led to the creation of a frontier signifying for the Amhara the “cultural and political divide between civilization and barbarism, the ideological marker between the law and the absence of law within the Christian realm” (237).[46] On the other side of this frontier, the subordinated groups reacted by “squeezing the dominant groups into one symbolic and political category to which they applied the ‘ethnic’ name of Amhara” (238).

Problematic is that these two aspects of the Ethiopian frontier are not connected. Whereas the frontier situations Triulzi describes are historically rich and detailed evocations of place, the narrative of Ethiopian expansionism and the reaction by incorporated societies is highly schematic and seems to hover in the air without ever touching ground. Triulzi partly recognizes this, since he writes that this narrative – the “frontier cast” – tended to “obscure ideologically the real contact situation that all frontier peoples activate” (237-38). Unfortunately, he does not follow through on the more interesting possible implication of his own argument, which is that these frontier lines are imaginary constructs produced by specific actors in specific situations and not a subsumation of these diverse contact situations. In other words, my guess would be that the use of frontier lines as an explanatory paradigm both by Amhara as well as non-Amhara actors reveals more about the connections that exist between these actors as about the differences that separate them.

This points to the limits of the center-periphery concept, since it makes explicit the cultural connections that exist between the center and the periphery. If this is accepted, then it must also lead to a re-interpretation of what Triulzi terms the “Amhara-Oromo dispute.” After rightfully noting that most Oromo scholars have simply replaced the historical and political discourse of Greater Ethiopia - which emphasized the centrality and ancient character of Amhara culture as constitutive of the character of Ethiopia – with a discourse that emphasizes the authenticity and primacy of Oromo culture,[47] he concludes as following:

“The competing views of Amhara and Oromo are not limited to the past, but en globe a critical assessment of the future. […] To sum up, ‘frontier’ history is not new in Ethiopia, and present-day politics will keep fuelling no doubt new sets of symbolic representations and opposed identities in the region. What is new is that the ‘other’ sides of the frontier are no longer silent.” (242-43)

Not asked, however, is if these Oromo scholars really do speak for the “other” side of the frontier.[48] In my view, the reversal of history that is being written by Oromo scholars - Triulzi refers to Mohammed Hassen and Mekuria Bulcha[49] - is a long-needed corrective to the Amhara-centrism of the Greater Ethiopia discourse, but it signifies at the same time some of the basic similarities between the Amhara and Oromo elites, since both are using the same elements to construct their separate spatial imaginations. That said, let us now turn to the construction of the spatial imagination of Oromia among students in Addis Ababa and abroad during the years before the 1974 revolution.

V. The Production of Oromia and the Rewriting of History

Neither locatable in the center nor the periphery, the spatial imagination of Oromia was the product of those actors that had the privilege to move between these two poles and beyond – the product of actors powerful enough to cross the frontier. This has a number of implications for thinking about the OLF and its brand of Oromo ethno-nationalism. The goal of this chapter is to analyze the period shortly before the 1974 revolution – which provided the basic elements from which the spatial imagination of Oromia was constructed – and offer an interpretation that emphasizes the transnational aspects of this imagination, while it criticizes the state-centrism prevalent throughout the field of Ethiopian studies.

The discourse of Oromia “originated” among educated Oromo in Addis Ababa.[50] This is emphasized by Baxter in an often-quoted article, where he writes: “As more and more Oromo became civil servants, Army officers and NCO’s [non-commissioned officers] and more Oromo school boys became undergraduates, and as more Oromo MPs managed to get elected, each group found that, in addition to sharing humiliating experiences, each shared a common language and similar values. The new pan-Oromo consciousness was generated in the army, the University and the Parliament itself” (1978, 290). What Baxter does not comment upon, however, is that this “new pan-Oromo consciousness” might not only be an expression of “some common types of Oromo experience” (291),[51] but that its specific shape was precisely the result of its production within these powerful institutions. The first part of this chapter describes in what ways the location of Addis Ababa played an important role. Particular attention will be given to the economic predominance of the capital city in comparison to the rest of Ethiopia and how this positioned the city as a site of privilege. The second part describes how this site of privilege delimited the spatial imaginations of the Ethiopian students in general and the Oromo students that supported the OLF in particular. This will be approached through an analysis of two different kinds of transnational spatial practices: first, the arrival of African scholarship-students in Ethiopia and second, the dispersal of Ethiopian students throughout the world with the support of governmental and non-governmental grants and loans. Although each with its own dynamics, both spatial practices were characterized by its dependency on the mechanisms of the state, but led to the establishment of spatial networks partially opposed to this state. The last part of this chapter analyzes the spatial imagination of Oromia during this period. Through a close reading of the way the OLF in their publications represented an important 1960s rebellion in Bale province, I want to achieve two things. First, emphasizing the importance of attending to positionality, I aim to show how the spatial imagination of Oromia leads to a re-scaling of history that constructs a specific geometry of social power. Second, acknowledging this political aspect should enable at least a tentative view at the more personal interests of actors that normally remain hidden behind the seemingly transparent surface of the space of Oromia.

The expansion of education after the Italian occupation coincided with a fast growth of Addis Ababa, dwarfing other cities in the process. In the first half of the twentieth century one could still speak of a certain balance between the capital city and other cities such as Harer and Dese (Addis Ababa was, respectively twice and four times as large) (Donham 1986, 32). Already in 1970, however, Addis Ababa had around 800,000 inhabitants, whereas the next ranking city, Asmara in Eritrea, only had about 200,000 (Markakis 1974, 161). Not only did the capital city literally control the majority of the economic flows of goods from southern Ethiopia to the rest of the world – mainly through the use of the Addis Ababa – Djibouti railway – most of the industrial developments during this period took place in neighboring areas. As Gilkes points out, in 1969, about half of all industrial establishments were located in Shewa, with the majority of them close to Addis Ababa (1975, 146).

This economic predominance was accompanied by a number of political and cultural characteristics that set it apart from the rest of Ethiopia. As capital city it encompassed the majority of the administrative resources of the state, which meant that about a third of all government employees – over 35,000 – worked in Addis Ababa. Most educational facilities above primary level were to be found in the capital. In addition, Addis Ababa was the only city in Ethiopia in which radio- and television-broadcasting stations were located (Markakis 1974, 162). Also, besides being the head office of the Organization of African Union (OAU), it encompassed a whole range of other international organizations, such as embassies (and the connected diplomatic community) and internationally supported private schools. Although this surely does not justify a complete isolation of Addis Ababa from the rural areas, it is nevertheless clear that these quantitative and qualitative differences provided urban actors with a whole range of elements to construct spatial imaginations that were not directly available to others.

In my view, it might be useful to embed Addis Ababa within a larger historical context and interpret the city as part of the older Abyssinian idea of social mobility with Addis Ababa representing the top of Crummey’s pyramid (see chapter IV.2 above). From this perspective, attaining the privilege to live in the capital city becomes the prize to be fought over. This interpretation is supported by Markakis who notes that “the rate of turnover among educated government officials is high because they are strongly attracted to the major cities and tend to regard their services in the small provincial towns as a temporary stage in their career” (1974, 167). It is also supported by Cohen’s remark that secondary school and university graduates usually tried to avoid government posts outside of Addis Ababa, leading to a lack of jobs in the capital city (1973, 379). Attending to the longue durée of the Abyssinian idea of social mobility also makes it possible to interpret the position of students and government officials as conditioned by an individualistic flight from the land.[52] With around 85-90 per cent of the Ethiopian population making a living either as peasant or pastoralist, attaining the privileged position of student or government official (be it in the bureaucracy or in the military) becomes the key to break with the rural past and embrace modernity.

The OLF, which originated in the academic community (Markakis 1987, 259), was part of this privileged position at the top of Crummey’s pyramid. It emerged out of earlier study groups that had devoted themselves to the discussion of Oromo culture and politics and lead to the publication of a number of pamphlets (“Kana Bektaa?” (Do You Know?) (1969-1971), “The Oromos: Voice Against Tyranny” (1971)).[53] On an organizational level, the OLF was a continuation of the short-lived Ethiopian National Liberation Front (ENLF), which was founded in 1971 and had made some attempts at guerilla warfare in the Chercher Mountains near Harer (Chanie 1998, 100; Bulcha 1997, 344-45; Zitelmann 1988, 326). Its members consisted of those Oromo-speakers propagating a highly nationalist version of Oromo history, culture and politics. At a 1976 meeting in Addis Ababa these members – largely university and secondary school students, not exceeding two hundred members, of which about thirty attended the meeting - agreed on an OLF political program, which was a revision of an earlier 1974 document (Markakis 1987, 262-63; OLF 1976).

Conceptualizing Addis Ababa as the state-version of the Abyssinian idea of social mobility makes it possible to see the students and government employees not only as part of this dynamic, but also as the main actors involved in its perpetuation. As I see it, this produces a problem of representation. On the one hand, as an analyst one is confronted with actors that claim to legitimately represent the rural people within their territory (be it the territory of the state or the ethnic nation). On the other hand, the biography of precisely these actors is informed by a desire to get away from the land - to “escape from the drudgery of farm-work and the dullness of village life”[54] - and thus to a certain extent to escape from their rural people. This creates, at the very least, a number of possible biases that need to be addressed. Focusing on education, a number of aspects seem important. For one thing, and a logical consequence of the pyramid-idea, despite the extension of education in Ethiopia after the Italian occupation, being a student automatically meant belonging to an elite. In 1970, for example, around 800,000 people were enrolled in all forms of education, from primary to university level. This reflected a little more than two per cent of the total population, whereas of this population around forty-five per cent were under fifteen years of age (Abir 1970, 49; Wagaw 1979, 169). Although primary education was to a certain extent spread out all over Ethiopia, education above this level tended to be concentrated in a small amount of provincial cities and above all in Addis Ababa. This geographical patterning of education again reinforced the logic of Crummey’s pyramid, with Addis Ababa functioning as the center of attention. Among these few students that made it to higher-level education – in 1970, less than eight thousand held a degree (Markakis 1974, 182; Ottaway 1978, 23-25) - Oromo-speaking students were clearly under-represented. It has been estimated that around fifteen per cent of all students could be considered Oromo, whereas the total Oromo population makes around forty per cent of the total Ethiopian population.[55] As most commentators have rightly mentioned, this discrepancy clearly reflected the subordinate role Oromo occupied in the Ethiopian political and cultural constellation. At the same time, describing ethnicities as relatively unambiguous and bounded entities is problematic. It ignores, for example, that close to fifteen per cent of the student population at the end of the 1960s consisted of students that were of mixed Amhara-Oromo parentage (Klineberg and Zavalloni 1969, 19). It also ignores important geographical specificities. Those Oromo-speakers that were able to pursue higher education usually came from areas like Shewa and Wellegga - where many Oromo had adopted Christianity during its incorporation into the Ethiopian empire and had participated in the further expansion southwards – and not from areas like Arsi, Bale and Sidamo – whose people were made dependent peasants in the neftennya-gebbar system (see chapter IV.2 above). Therefore, those Oromo-speakers that were enrolled as students reflected that section of the Oromo population that had cooperated with the Ethiopian state in the first place. Most Muslims were excluded and pastoralists – mainly due to the lack of state education in pastoralist areas – were ignored (Markakis 1987, 259). Secondly, the student population was dominated - both in numbers as on the level of discourse – by males. The higher the level of education, the fewer women there were to be found. In 1968, for example, primary education had around thirty per cent females enrolled. Secondary education, however, only had about twenty per cent and at the university the student population consisted of around seven per cent of females (Markakis 1974, 150).

These are some of the structural elements that should be considered when writing about the OLF and the spatial imagination of Oromia. However, although the capital of Addis Ababa, the idea of social mobility and very likely even the male-biased nature of the elite is unthinkable without acknowledging the importance of the processes of centralization and internationalization as discussed in chapter IV, so far my argument has been notably state-centered. I have presented the Ethiopian students as a small elite within the center of the state destined to fulfill the role of government employees. More or less implicit in this argument has been that this structural condition also delimited the spatial imaginations of these actors. That is, however, only partly true. The privileged position at the top of the pyramid enabled a view beyond the Ethiopian state and this clearly influenced the spatial imaginations that were produced. In this respect, Markakis’ comment that the “the social and political awakening of the Ethiopian students has been a spontaneous, self-accomplished process achieved in a situation of great isolation” (1974, 358) is false and deserves a more sophisticated conceptual treatment. A first specification would be a denial of Markakis’ claim: the students were not at all isolated, but had contact with actors around the world. A second specification would be a partial acknowledgment of Markakis’ state-centered view: the students did have access with actors around the world, but this was precisely the result of their privileged position within the Ethiopian state. This is quite important to emphasize, since transnationalism was nothing radically new to Ethiopia. Pastoralists in the southern and eastern areas of Ethiopia, for example, had been traversing large areas to trade and to allow their animals to graze since centuries. The creation of an Ethiopian state with a delimited territory at the end of the nineteenth century suddenly created borders where previously there were none and turned the pastoralists into transnational nomads.[56] In contrast to the pastoralists, however, for whom the state mainly presented an obstacle to their transnational behavior, in the case of the students it was the state that made possible their transnational practices in the first place.[57] Two illustrations should suffice.

The arrival of students from other African countries at the end of the 1950s and the early 1960s, for example, was made possible by Selassies offer of two hundred scholarships at the 1958 Conference of Independent African Countries in Ghana and was informed by political considerations aimed at an increasing affiliation with Africa. The students that came - from Ghana, Kenya, Tanzania, Uganda, Sudan and Nigeria – saw themselves as the main challengers of colonial rule and as the new leaders of their countries once independence would be attained. Their own representation of Ethiopia was one that idealized the country for its independence during colonialism, its charismatic leader Haile Selassie and its long history of indigenous cultural developments. To them, Ethiopia was an important symbol of pan-Africanism. On arrival in Ethiopia, however, they were confronted with extreme poverty, educational backwardness and a student population that was politically unorganized. The effect of interaction on the Ethiopian students was a re-questioning of earlier values and the construction of new explanatory frameworks that made use of elements provided by the African students: anti-colonialism, Marx and developmental modernization.[58]

The privilege of being part of an elite also enabled many Ethiopian students to study abroad themselves – another example of a transnational spatial practice. Sponsored by the state and through the use of scholarships and other programs from foreign governments and international organizations, thousands of students pursued their education outside of Ethiopia. The largest concentration, in 1970 estimated at seven hundred, was in the United States (Zewde 1991, 222).[59] The main organizations channeling student activity were the Ethiopian Students Union in North America (ESUNA) and the Ethiopian Students Union in Europe (ESUE). It was here that these students had access to a wide range of experiences and uncensored literature, which in turn informed their representation of the situation in Ethiopia. The most well known example is probably that of Germame Neway who planned the failed coup against Haile Selassie in 1960. He had obtained a BA from Wisconsin University, an MA in political science from Columbia and in 1952 became the second president of the ESUNA. His action was informed by the conviction that Ethiopia was a backward country compared to the seemingly progressive steps taken by other African states.[60]

What were the effects of these transnational spatial practices on the spatial imaginations of the students in Ethiopia? There is no simple, unilinear answer to such a complex process, but I believe that one effect was an increase in available possibilities that could be appropriated to take a stand among other social actors. At least in this point, the Appadurian mantra that “the world we live in today is characterized by a new role for the imagination in social life” (1996, 31) retains its explanatory force. Naturally, local conditions played an important if not decisive role in the growing discrepancy between the expectations of the state and those of the students. Since secondary school and university students were educated in order to serve the expanding bureaucracy, it seems quite realistic to believe that their spatial imaginations (at least the public face of it) would largely overlap with those of the state as long as the state could redistribute jobs, money and a certain amount of prestige. The increasing difficulties to provide these privileges from the late 1960s on made many students turn away from the imagination of the state towards other imaginations.[61] But this is exactly the point. Students were confronted with a state that could no longer redistribute privileges and thus turned to the appropriation of global imaginations that were radically subversive to the further existence of this state.[62] This was a situation that had, at least in this quality, not existed before.[63]

In a general way, one can therefore say that the critique against the Ethiopian state knew two directions simultaneously. One critique was to reject the imagination of the Ethiopian state and instead opt for a spatial imagination “below” the scale of the state. In the case of the OLF, this imagination was captured in the term Oromia, which was seen to represent the largest part of southern Ethiopia. The second critique made use of elements “above” the scale of the state in order to criticize this state. At the same time, this second critique informed the contents of the first critique. It is this intertwinement of the second and first critique – of the global discourses and local struggles – that Oromo ethno-nationalism and the other nationalisms of this period differed from the older Abyssinian situation, where the lowlands provided a chance for nobles who had lost out in the political struggles at the center to increase their resources and following before returning and capturing power in the central highlands (Donham 1986, 20). The spatial imagination of Oromia was thus the result of these different elements on offer: the imagination and practices of the Ethiopian state, the structuring mechanisms of the inter-national state system and the global discourse of Marxism-Leninism. Through selectively appropriating elements of these discourses, Oromia was discursively constructed: “being Oromo” was now positioned on the crossroads of anti-colonialism, anti-imperialism and pro-modernist nationalism.

In the last part of this chapter I want to approach the spatial imagination of Oromia in the early 1970s through an analysis of the representation of the Bale rebellion in the OLF-discourse. I want to show how this rebellion is re-scaled in order to make it “fit” with the imagination of Oromia. As I see it, this involves on a discursive level a re-definition of the geometry of social power (Massey 1993, 61) that deserves to be explicated, since it is here that the question of “who speaks?” returns with a vengeance.

Bale in the 1960s was the second largest, but most sparsely populated province in Ethiopia, with about 200,000 inhabitants and a density rate of 1,6 per square kilometer.[64] This area was incorporated into the Ethiopian state in the late 1880s after a campaign that was mainly directed against the Arsi Oromo. They lived in the northern highlands of the province, whereas the Somali usually lived on the lowlands in the southern part. Both were largely decentralized pastoralist groups, although it seems that among the Oromo in the highlands sedentarization had already begun before the northerners introduced plough farming. The occupation of the Abyssinians reduced the status of the Arsi Oromo to that of a subject peasantry comparable to many other groups in the southern region. Although they were at this stage usually allowed to remain on their lands, they had to surrender a substantial part of their produce to the northern settlers. The lineage and clan heads that originally served as mediators between the different Arsi lineages and clans now became balabbat and thus took on the mediating position between the northern settlers and the peasantry. These new relationships institutionalized a context of domination and subordination, in which the northern settlers effectively could shield themselves from the peasants through cooptation of the lineage and clan heads. This created the paradoxical situation, so typical for many forms of colonial rule, in which the traditional heads were now both accomplices of the settlers as well as spokesmen for their people.

Such an unequal center-periphery relationship – a specific case of the more general “state-version” I described in chapter IV – does not inescapably lead to rebellions that endanger the perpetuation of this system, but is contingent on the keeping intact of relationships between differentially positioned actors. If the connections break, the system becomes instable. In a hierarchical system such as that in Ethiopia, the most dangerous breaks occur when the middlemen, those positioned between the central government and the peasantry, reject the system and start looking for other connections. In the case of what later came to be known as the Bale Rebellion exactly this occurred. Three of the most important men that guided this rebellion were all representatives of this position between the peasantry and the state. In the district of Wabe, Haji Isahaaq Mohammed Daddi, an Arsi Oromo and senior intelligence officer, fled to Mogadishu after having murdered an Amhara rival in 1961. Here he became affiliated with the Western Somalia Liberation Front (WSLF) – in this way gaining access to weapons – after which he returned to Wabe and started organizing revolts. In the district of El Kere, Kahin Abdi, a well-known bandit, first burned the salt mine at Afker after which he attacked a number of villages in Bale.[65] Lastly, in the district of Delo, it was Wako Gutu who deserted and at the end of 1964 made his way to Mogadishu. Wako Gutu was a village judge and his desertion was occasioned by the longstanding rivalry between him and a subdistrict administrator.

After the Ethiopian government put the province of Bale under martial rule and with external support from British and Israeli military experts, it was able to contain this rebellion. The 1969 coup of Siad Barre provided the final blow to the rebels, since the new government in Somalia ended its support for the WSLF and stopped the military support for the rebels in Bale. The center-periphery relationships were reconstructed by either a re-incorporation of the rebels into the system or by a (temporal) exclusion from this system. Wako Gutu, for example, returned to Ethiopia and was given balabbat status and a certain amount of land. Others left Ethiopia altogether and headed for Somalia where they joined other refugees from southern Ethiopia in a group called the Western Somalia Harekat. Members of this group would, in the second half of the 1970s, play an important role in the formation of the Somali Abbo Liberation Front (SALF).

Different writers construct different narratives and this becomes particularly clear in the case of the Bale rebellion. Both academic and non-academic commentators are in the business of developing arguments and these clearly inform their interpretation of historical “facts.” Whereas successive governments of Somalia interpreted this rebellion as part of a longing for the realization of Greater Somalia, the Ethiopian government rejected this reading and instead wanted to see this rebellion as the result of the imagination of a few mercenaries. Both representations, of course, say more about the positionality of the governments of Somalia and Ethiopia than it does about the Bale rebellion. Neither was the Bale rebellion a “typical peasant protest,” as Tareke (1991) has argued in his detailed description of three important rebellions.[66] As he writes, in the conclusion to the Bale chapter: “The ethnic or national question in southern Ethiopia was quintessentially a peasant question. Somalia’s stress on the more pronounced forms of social oppression – i.e. its simple presentation of the problem as a contest between Christian Amhara and Muslim Oromo/Somali – critically undermined the class aspect and the transformative potential of the revolt. […] If the rebels sometimes styled themselves sewra, perhaps a corrupted derivative of the Arabic thwar (revolution), between 1969 and 1970, they were only imitating the revolutionary rhetoric of the Somali military regime. And when they attempted to draw a parallel between the Vietnamese struggle and their own, the insurgents were only seeking wider recognition. Their movement was no more than a typical peasant protest, whereas that of the Vietnamese was a nationalist cum social revolution” (159, Italics added). Such an interpretation however, seems inspired mainly by Tareke’s attempt to embed the Bale rebellion in a larger Marxist-inspired narrative that emphasizes the oppression of the peasants. The fact of continued oppression of peasants throughout centuries of Ethiopian history is certainly of major importance, but Tareke’s ideological commitments make him obscure the translocal and transnational connections that gave this rebellion its power.

These different representations of, apparently, the same rebellion, reminds one of the argument made by Liisa Malkki, who in her work on Hutu refugees in Tanzania, has pointed out that these differing narratives involve “different regimes of truth [that] exist for different historical actors and [that] particular historical events support any number of different narrative elaborations. Such regimes of truth operate at a mythico-historical level which is concerned with the constitution of an ontological, political and moral order of the world” (quoted in Matsuoka and Sorenson 2001, 13).[67] In the case of Ethiopia, there seems to be a general problem among a whole range of authors to accept that all social praxis is “overdetermined” i.e. that it cannot be reduced to a single and linear narrative of development and decline, since social actors always occupy a whole range of positions and are dependent on a number of (often conflicting) relationships. Having said that, the analysis of these different narratives and the regimes of truth they evoke can surely tell us – if not much about history – a great deal about the actors that produced them.

In the OLF Political Program the Bale rebellion is re-scaled in order to make it part of a larger struggle against feudalism and imperialism: “Ever since their colonization by the Amharas, the heroic Oromo people have waged a relentless struggle to rid themselves of the dual burden of exploitation by feudalism and imperialism” (OLF 1976, 3). The Bale rebellion is described as following: “Eight years of guerilla struggle in Bale needed the aid of Israeli explosive experts, British Army bridge and road builders, and American Air Force advisers on precision bombing to with stand the insurgents” (4). The already mentioned pamphlet “Voice Against Tyranny,” which was published before the formation of the OLF at the time of the Oromo study circles and which was distributed in Addis Ababa, has a similar narrative logic:

“To any vigilant observer, the Zionist armed settlement in the homeland of the Palestinians and the armed settlement currently carried out in the annexed territories of Oromo and other Ethiopian nations, is exactly designed and implemented by the same agency. To implement the detestable kibutz program, the Haile Sellassie regime proceeded with the aid of Zionist experts to measure and register what it claims to be ‘government land.’ […] The intention of the regime towards the Oromo and other subjugated nations is clear. Its policy is to reduce these nations to a propertyless mass of people just like the black South Africans and Rhodesians.” (1980, 22)[68]

This quote clearly indicates the importance of attending to the socio-spatial relationships that are drawn by actors themselves in the analysis of culture, politics and ethnicity in Ethiopia. In the pamphlet, after having established the radical difference between the Amhara and the Oromo, each is connected to spatial imaginations that mutually exclude each other. The Amhara are re-scaled in order to make them part of a larger Zionist and imperialist plot with the implication that without these connections they would be powerless. On the other side of the border, the Oromo are re-scaled and related to the oppressed worldwide, in particular to those in settler colonies such as South Africa and Rhodesia where a white minority controlled a black majority. The implication in this case is, of course, that it is the oppression by the Amhara that obstructs the welfare of the Oromo and the flourishing of the Oromia nation. Textual support is also given to “all progressive forces struggling against imperialism, feudalism, racism and all reactionary forces,” by which are meant the various African liberation movements and the “Arab people’s struggle against Zionist aggression” (OLF 1976, 13). Although it is acknowledged that there have been quite a number of Oromo that allied themselves with the Amhara, they are termed neo-Gobanists and are so excluded from the spatial imagination of Oromia.[69] What is, of course, not addressed is that the very producers of this narrative – the student intellectuals in Addis Ababa – themselves occupied a position at the margins of the Oromo culture they claimed to represent.

A similar approach can be seen in the OLF’s preoccupation with the land. It must be said that for a national liberation movement, the OLF is surprisingly not interested in a delimitation of the boundaries of Oromia. Although in the OLF Program reference is made a number of times to the need to liberate Oromia, no map is included nor are there any references to specific territorial boundaries. Later publications often did include a map, but this was always of such a vague nature that its main function must have been symbolic[70]: Oromia as a partly defined container that could be related – with the relationships depending on the occasion - to a number of different social, political, cultural and religious elements. Notwithstanding the lack of territorial boundary-construction, the element of land still takes center-stage in the spatial imagination of Oromia. Besides the repeated references to the colonization of Oromia by Amhara settlers, the central connection is made between the land and the peasantry. In a move typical for most nationalist organizations, they are seen as the authentic bearers of culture, the cradle of the tradition of the Oromo. Nationalism and Marxism-Leninism are usefully combined: references abound to support the view that the peasantry is the “backbone of the revolution,” leaving the “revolutionary intelligentsia” with the “task of providing the necessary technical and ideological guidance in the forefront of the liberation of the nation” (OLF 1976, 7). Within one spatial imagination, the importance of the peasantry is celebrated, while at the same time the need for (urban) elite leadership is affirmed. This continuous reference by the OLF in the early 1970s to the importance of the peasantry as the bearers of Oromo culture, however, does raise some questions concerning the role of those Oromo that do not fit this categorization: the pastoralists. The OLF-interest in developing the Oromia nation here leads to a territorial exclusion of those actors that are at the same time seen as the bearers of Oromo-identity, namely the pastoralist Boran Oromo who – in one way or another – have still maintained their “traditional” gada-system of age-grades.[71] Although gada is celebrated by OLF-nationalists as a central aspect of Oromo culture and as an egalitarian and democratic alternative to the hierarchical Abyssinian political system, those Boran Oromo considered as most authentic are simultaneously excluded from participation in the nation. One could say that in this case the OLF treats the pastoralists in the same way as other nationalists treat women: they are “constructed as the symbolic bearers of the nation, but are denied any direct relation to national agency” (McClintock 1995, 354).

VI. Imagining Oromia Between the State and the Diaspora

The first part of this chapter continues the exploration of the center as conceptualized in the previous chapter, namely as a site of privilege. The starting point is the deposal of Haile Selassie in 1974 that ended a period of hegemonic control over the state, which had remained relatively uncontested – with the exception of the Italian occupation - since 1930. In line with the Ethiopian political tradition that sees competition as treason, the years directly after the revolution were characterized by what one could call a period of homogenization. With Haile Selassie no longer in power, the main question became who was going to take his place and thus gain the monopoly over state resources. Although the articulation of local (urban) grievances was central to the early period of the revolution, within a matter of months the majority of the population was excluded and the political struggle reduced to one between a small segment of the military and a number of “civilian” organizations that above all represented the educated. After four to five years, it was a segment of the military called the Derg that emerged as the sole survivor in the center and thus as the only organization with direct access to the state.

In principle, the combination of violence, homogenization and revolution does not distinguish Ethiopia from other countries such as France and Russia that have known their own highly violent revolutions. Donham (2002, 3) draws on the definition of Skocpol (1979) to point out that “1974 constituted a social revolution in the classic sense. That is, both class relations and the form of the state were interactively transformed.” O’Kane (2000) uses the idea of a Reign of Terror as conceptualized in the earlier “natural history” theories of Edwards (1970) and Brinton (1965) to compare the Ethiopian revolution with those that occurred in Iran and Nicaragua. At the same time, of major importance in the case of Ethiopia is that the homogenization of the center involved not only an annihilation of those not willing to submit to the Derg, but also a move of some of these oppositional voices towards the peripheries. This will be the subject of the second and third part of this chapter. In the case of the OLF the removal from the center was even more drastic than for example in the case of the TPLF or EPLF, since it was largely excluded from the territory of the state altogether. However, as I will argue, this move should not be interpreted as a marginalization of the OLF, but instead as a process of globalization, since its new position in the peripheries enabled an active re-insertion into world affairs and gave them access to global resources. It is this move towards the peripheries that also needs to be conceptualized when thinking about the spatial imagination of Oromia. Drawing on the research of Zitelmann among Oromo refugees in Sudan, I will argue that after this move one can no longer usefully interpret Oromia as only the result of an interaction with and reaction to the mechanisms of the Ethiopian state.[72] Even an incorporation of the transnational aspects of this state – which I conceptualized in chapter V – no longer suffices. Instead, the spatial imagination of Oromia should be interpreted as embedded within a spatial diasporic network that no longer needed the Ethiopian state in order to imagine Oromia.

As probably most revolutions do, the 1974 Ethiopian revolution started in a most mundane way.[73] On January 12, soldiers in a garrison in southern Ethiopia mutinied against their officers due to a lack of decent food and a shortage of drinking water. This was followed by a similar mutiny at the Debre Zeit Air Force Base near Addis Ababa and a larger revolt among the Second Division in Asmara in February. The crisis further escalated through strikes in the capital city and larger towns, with each group – students, teachers, taxi drivers and factory workers – demanding higher salaries to compensate for the increased living costs that were mainly the result of a rise in oil prices in the wake of the Yom Kippur War. The months until the formal deposition of Haile Selassie on September 12 were characterized by more strikes, concessions of the Emperor to both civilian as well as military demands, the progressive disempowerment of the old elites and a shift of power to the military. This shift was also accompanied by a move away from strikes that focused on relatively local grievances – reflecting the frustrations of the diverse urban groups – to discourses that aimed at a complete restructuring of the Ethiopian state and society – reflecting the imaginations of a number of “civilian” organizations and the military. In other words, within a few months the urban revolution was selectively appropriated and dominated by a relatively small elite. Within the urban areas, the following years were to be characterized by a struggle for hegemonic power between these elites. It involved the rise of the Derg (Amharic for committee) and the subsequent exclusion or annihilation of oppositional elite voices.

So far, the rural areas did not play a decisive role in the revolution. They were made part of the revolution, however, through the nationalization of rural land that was proclaimed by the Derg in March 1975. This was part of a more general economic policy that also involved the nationalization of banks and private-owned companies. The land reform was certainly the most radical change announced by any regime in Ethiopia. It rested on three main principles. First, the government from now on owned all land including the large areas of the Ethiopian Orthodox Church. Second, no one was to be allowed to farm more than ten hectares. And third, no one would own land individually, but was to work the land as part of a collective organized in a peasant association.[74] The results of this reform were far-reaching and led to a partial dismantling of the older patrimonial structures that connected the balabbat, governors, landlords and military commanders to Addis Ababa and Haile Selassie. Among most peasants in southern Ethiopia, this removal of state structures from their land and daily lives was embraced and often led to the violent rejection of a large number of landlords. At the same time, the discourse of the land reform reveals as much about the imaginations of its producers in the capital city as it reflects a concern for the situation of the peasants. The majority of the peasants, for example, would hardly have emphasized the importance of working the land as a collective and would have instead preferred to own the land on an individual or family basis. Thus, although many peasants were now no longer dependent on landlords, they were at the same time “captured” by the state through the institutionalization of these peasant associations (Clapham 2002, 15). Although I doubt if this capture of the peasantry was planned beforehand, it was surely one of the more useful “unintended side-effects” – to misuse the vocabulary of Beck[75] – of the actions of state actors immersed in a discourse of Marxism-Leninism. It enabled, first, the destruction of the old order and, second, the construction of new connections between the center and its peripheries through the institution of the peasant association.

The proclamation of the land reform was probably the one thing that the myriad of urban intellectual organizations could agree on. After all, the ending of feudalism had been central to the discourses of its actors since the late 1960s. Besides this general agreement however, on a change that had, at least in the beginning, no obvious and direct implications for these urban actors, the first revolutionary years were characterized by a proliferation of mutually exclusive discourses. The paradox is that many of the actors producing them did not have any difficulties with appropriating a number of these exclusive discourses simultaneously. In other words, what must strike every observer as peculiar is the highly personal nature of many of these so-called movements with their intense rhetoric of ethno-nationalism and the right to self-determination. Taking a closer look at this period shows that the exclusive discourse of ethnicity did not stop its producers from developing a range of connections that seem based mainly on attaining a position of power – or relative security - in the new Ethiopian constellation. Members of the OLF, for example, simultaneously held important positions within the Derg-regime until 1978/79 (Zitelmann 1994a, 83). Another organization, the All Ethiopian Socialist Movement (MEISON), considered by many as an important stronghold of Oromo nationalism and of maintaining connections with the OLF, also worked together with the Derg and is often credited with providing the “brains” behind the revolution. The Program of the National Democratic Revolution, issued in April 1976, was largely written by MEISON and provided the formal statement of national policy until the program of the Workers’ Party of Ethiopia (WPE) was issued in 1984. A number of MEISON members occupied ministerial and administrative posts and formed the core of civilian organizations that were organized under the People’s Office for Mass Organizational Affairs (POMOA), which, until its demise in the summer of 1977, formed a parallel apparatus to the state administration in order to conduct political education and mobilize mass support throughout the country. One of the central figures behind MEISON was Haile Fida, an Oromo from Wellegga, who provided the ideological backup to Mengistu Haile Mariam, until MEISON fell out of grace and he was arrested and executed. This same Haile Fida, however, had also studied linguistics in Paris and Hamburg and had in 1973 published the first grammar in afaan oromoo that was written by an Oromo (Zitelmann 1994, 93; Bulcha 2002, 186-88). The Oromo Baro Tumsa led another group, the Revolutionary Struggle of the Ethiopian Oppressed (ICH’AT), which had originally split from MEISON, but continued to affiliate itself with the military regime until it was also persecuted and annihilated by the Derg. Baro Tumsa also occupied a leading function within the POMOA and was at the same time related to the protestant Mekane Yesus Church through his brother Gudina Tumsa. This later led to the accusation that the Mekane Yesus Church was actually a source of Oromo nationalism, after which Gudina Tumsa was kidnapped and murdered (Eide 1996; Studer 1992).[76]

Even the Ethiopian People’s Revolutionary Party (EPRP), which is often seen as dominated by Eritreans and Tigreans and against whom much of the violence during the Red Terror of late 1976 until early 1978 was directed, cannot be radically separated from organizations as the OLF, MEISON and ICH’AT. In 1976, for example, the EPRP set up an Oromo People’s Liberation Organization (OPLO) with support from Somalia in order to win over the “Oromo masses.” In reaction, MEISON set up an Oromo National Democratic Movement (ONDM).[77] As a reminder, this was the same MEISON that incorporated a number of OLF-members. Much of the insecurity surrounding the question whether the OLF wanted self-determination for the Oromo within an overarching federal framework or if it desired separation of the nation-state Oromia from the Ethiopian state can be answered by acknowledging the pragmatism of politics during this period despite the exclusive spatial imaginations that accompanied it.[78]

The process that led to the emergence of the Derg headed by Mengistu Haile Mariam as the hegemonic force in the center was chaotic and violent. It involved, on the one hand, a power struggle within the Derg itself, and on the other hand, an annihilation and exclusion of all the important oppositional forces. It was only after three years, in early 1977, that the Derg could be considered a relatively stable entity headed by Mengistu. It involved the killing of a number of important competitors, such as Aman Andom, Teferi Bante and Atnafu Abate, who might have pursued different directions had they gained a powerful position. In the meantime, what became known as the period of Red Terror effectively killed or excluded all other oppositional forces from the center. MEISON was involved in the killing of EPRP members; the EPRP launched a campaign of urban terrorism to destroy MEISON; and in the end it was the Derg that got rid of them both and that emerged as the sole force with direct access to the resources of the state. Although the support of the United States was cancelled, in the larger context of the Cold War it was able to switch sides and receive military back-up from the USSR.

As the OLF was excluded from the center of the Ethiopian state, it had to reposition itself in the periphery. In doing so, it underwent a similar process as other insurgent movements such as the TPLF and the EPLF. In a 1990 article Clapham has argued that this – the fact that both the new revolutionary regime as well as the insurgent movements battling against this regime came from the center - goes a long way towards explaining the specificity of “the Ethiopian situation.” Adapting Huntington (1968, chapter five), he argues there are two ways in which revolutionary change can be brought about. One possibility is that the old regime is overthrown by an uprising in the national capital and displaced by a government of the revolutionary intelligentsia. The other possibility occurs when the old regime is capable of retaining control over the center, but is confronted with revolutionaries that try to organize the peasantry and establish control over the countryside through guerilla warfare, before finally capturing the capital and displacing the old regime. According to Clapham, “the tragedy of revolutionary Ethiopia has been that it experienced both of these forms of revolution simultaneously:”

“Accordingly, I would regard both the present Ethiopian government, which gained power through the first of these mechanisms, and insurgent movements such as the EPLF and TPLF which are seeking to gain power through the second, as different and opposed elements of the same revolution.” (224)

Although Clapham only refers to the EPLF and TPLF, his argument can be partially extended to the OLF as well. Partially, because the difference between these northern liberation movements and the OLF is that the first were actually fighting on the land they wanted to “liberate” – i.e. they occupied large areas of land inside Ethiopia – whereas the OLF was mainly active at the extreme margins of the state.

The OLF’s exclusion from the Ethiopian state was paralleled by the 1977/78 war between Ethiopia and Somalia and this war, the general escalation of violence and the villagization programs[79] led to a dispersal of Oromo (and other) refugees into neighboring countries. Besides a high level of internal displacement, most of the Oromo refugees headed for Somalia. Braukämper – using UNHCR and Oromo Relief Association (ORA) figures - estimated in a 1982/83 article that there were approximately 200-250,000 Oromo refugees in Somalia, 10,000 in Djibouti, 10,000 in Sudan, 8,000 in Kenya and 9,000 in the Middle East, Europe and North America (4). Although the majority of the Oromo refugees were thus to be found in Somalia, it was in Sudan that the OLF would have its major influence.[80] This was directly related to the local contexts in which they had to function. In Somalia, the OLF and their ORA had been allowed to operate since the end of 1979, but after disputes with the SALF were increasingly restricted in their work and in 1982 had to evacuate their office in Mogadishu. In Sudan, on the other hand, the OLF and the ORA were able to establish a structural connection to the Sudanese state, enabling the construction of a relatively competition-free environment, in which they could support those displaced Oromo that were the result of the increased violence, the resettlement programs and the general economic insecurity in Ethiopia. In contrast to Somalia, most refugees in Sudan were not acknowledged by the UNHCR and were instead controlled by the refugee bureau of the Sudanese government. Refugees came here to receive a legal status (i.e. to get papers) and to find work. The OLF/ORA and their “mass movements” had close contacts with this bureau and offered a complementary organizational structure dedicated to the support of Oromo refugees, providing shelter when necessary and functioning as a kind of social insurance institution (Braukämper 1982/83, 11). It was this privileged relation to the Sudanese state that gave the OLF its main strength and it meant that until early 1991, the main areas of activity inside Ethiopia were located at the Sudanese border in the western part of Wellegga.[81] These areas were directly connected to the Sudanese provinces of the Blue Nile and Upper Nile, from where the OLF could establish connections with the capital city of Khartoum – where its Foreign Relations Office was located – and from here with actors around the world.

To summarize all this in other words: after the exclusion of the OLF from the center it is no longer possible to interpret them as part of the Ethiopian center-periphery structure. Instead, it might be more useful to see the OLF during the period from 1974 until 1991 as embedded in a diasporic spatial structure that had its main “bridgeheads” (Hannerz 1996) in Khartoum, Sudan and a number of urban centers in Europe and the United States. It would thus be wrong to assume that the marginalization of the OLF (i.e. its exclusion from the Ethiopian center) and the thousands of refugees seeking shelter and work in Sudan implied an isolation of this group. On the contrary, the stay in Sudan gave the OLF the strategic advantage of again connecting itself with the rest of the world. From a general political perspective, this has been most clearly pointed out by Bayart who has vehemently argued against the view that Africa is (increasingly) disconnected from global processes. Counteracting this idea, he argues that African actors have mobilized resources worldwide – famine relief, armaments, ideologies, and financial support – in order to gain or consolidate power in the domestic arena. Important is that he also sees these, as he calls it, “strategies of extraversion” implicated in the spatial practice of flight:

“Movement of this sort turns the state into a political space which is both relative and highly contested, with whole regions or population groups escaping the control of the central authorities without subverting or even destabilizing the central power […]. Nevertheless, escapade is taking on a more tragic character than before, as it is motivated by very violent conflicts and sometimes takes the form of migrations provoked by campaigns of ethnic cleansing, as has happened most obviously in the Great Lakes region, but also in Kenya, Ethiopia, Angola, Liberia or Sierra Leone. People who have had resort to flight – in this case as refugees - remain in direct contact with the international system via humanitarian organizations, churches and the press […]. In other words, flight is not tantamount to disconnecting oneself from the world, as it may seem at first sight, but it is rather a mode of insertion or reinsertion into world affairs, and even of globalization.” (2000, 260-61)

Similar to the situation in Addis Ababa (as I discussed in chapter V), this “re-insertion into world affairs” is the major difference between this struggle from the margin and the older Abyssinian situation of retreating into the lowlands after losing out in the political struggles at the center. Although in both cases local credibility is needed in order to increase resources and build a following before returning and capturing power in the center, in this new situation of globality actors have an enlarged number of options available that can be appropriated and put to strategic use. In that sense, the exclusion of the OLF from the center created a range of new possibilities that had not been available before. If that is accepted, then this must surely lead to a differentiation of the idea of “forced migration” that haunts much of the literature on refugees.[82]

One differentiation can be made explicit by referring to the “power-geometry of time-space compression” (Massey 1993, 61) that I discussed in my theoretical section. In the case of Sudan, one needs to differentiate between the OLF-cadres and those Oromo refugees that were supported by and organized in the “mass organizations” of the OLF. Similar to the students in Addis Ababa and their virtual relation to the peasantry, in Sudan it were the OLF intellectuals that had privileged access to the “flows and interconnections” of this global world. Three aspects of these interconnections would be the following. First, many of the texts used for the alphabetization courses in afaan oromoo that took place in Sudan during the 1980s, were printed in and paid by actors in the Diaspora, usually Oromo student organizations in Europe and the United States. Non-Oromo actors outside of Sudan and Ethiopia were also involved in the support of these campaigns in which, according to figures of the ORA, 10,000 refugees had participated until 1989. In Germany, for example, the Berliner Missionswerke provided financial and ideological support to the liberation struggle of the OLF and its spatial imagination of Oromia.[83] In Sudan, the OLF-cadres decided which literature would be made available to its members at what stage. Second, the OLF and its Foreign Relations Office in Khartoum provided an important channel of information during the Derg-years and they, if not determined, surely influenced the worldwide public image of its own organization and the Oromo in general. Inside Ethiopia (with the possible exception of Addis Ababa), research was next to impossible due to governmental restrictions. Those interested in Oromo culture, society or politics thus had to do either research among Oromo refugees[84] or rely on the information provided by the OLF in its numerous publications - such as Oromia Speaks - and the publications and representatives of the Oromo student organizations abroad. Third, these connections were not only material or ideological, but also personal. It involved the movement of actors between Khartoum and, mainly, a number of urban centers in Europe and North America. However, this movement was highly restricted, both in numbers as well as concerning aspects of gender and class. In the Diaspora, those Oromo active in the liberation struggle were overwhelmingly educated, urban and male, since many of them were part of that group of students (as described in chapter V) that went abroad for further studies before the 1974 revolution or left the country in the early phase of the revolution. This class and gender bias was only further perpetuated by that select group of refugees of the late 1970s and 1980s that succeeded in obtaining a place in the American Refugee Resettlement Program or a visa for Canada or Europe, since most countries preferred young and educated people – which in the Ethiopian context were usually male - that were seen as more flexible and adaptable to new situations. Although US government statistics did not distinguish between “Ethiopian” and “Oromo” refugees, McSpadden (1999, 260) notes that approximately eighty per cent of all refugees originally from Ethiopia that resettled in the US were single men in their twenties.[85]

What are the consequences of these differentiations? For one, it enables a view that sees the spatial imagination of Oromia as an intellectual construct. This was, of course, true from the very beginning. Before 1974, among Oromo students in Addis Ababa and abroad, the spatial imagination of Oromia offered a sense of place to those actors that were and felt rejected by the Ethiopian state. It is important to emphasize that this sense of place was primarily an external and not an internal sense of place, as many humanistic geographers would have it. Appropriating the spatial imagination of Oromia provided one way of retaining continued access to the global scale and in this way access to the resources and attention of important others. In other words, it provided one way of not being forgotten. These more subjective elements to ethno-nationalism are also recognized in a recent article by Asafa Jalata, probably the most active Oromo nationalist in the United States, who quotes Geertz to point out that: “[t]he one aim is to be noticed; it is a search for identity, and a demand that identity be publicly acknowledged… the other aim is practical: it is a demand for progress for a rising standard of living, more effective political order, greater social justice, and beyond that of ‘playing a part in the larger arena of world politics,’ of exercising influence among the nations” (Jalata 2003, 83; the quote is from Geertz 1995, 30). After 1974 and the subsequent exclusion of the OLF from the Ethiopian state, its re-positioning in Sudan enabled a strategy of extraversion aimed at re-gaining access to resources as well as, on the individual level, a “part in the larger arena of world politics.” Although I might be stretching my emphasis on the longue durée of the Abyssinian idea of social mobility a bit too far in this case, it might be useful to interpret the spatial imagination of Oromia as harboring to its producers the possibility of escape from the land and access to a socio-spatial network that is spread out all over the world. The difference with the situation before 1974 was that during the revolutionary years the physical existence of the Ethiopian state became largely irrelevant. With important OLF actors excluded from the state, the imagining of Oromia became embedded in a spatial diasporic network of the educated.[86] At the same time, the privileged position of the OLF in Sudan as a mediator between the Sudanese state and the Oromo refugees provided a possibility of encompassment of these refugees in “mass organizations.” Through the use of these organizations and the support they offered to the refugees, the OLF could popularize the spatial imagination of Oromia that originally derived from the sense of place of a number of intellectuals. This was often accomplished in quite subtle ways; it mainly involved a re-routing of earlier separate imaginations into the overarching spatial imagination of Oromia. The connecting of elements such as gada and the language afaan oromoo to this imagination has already been mentioned. Zitelmann (1994, 120) offers another, more geographical example: the synthesizing of separate landscapes into one landscape of Oromia. As he writes: “The notion of ‘homeland,’ which previously was only connected with the immediate experience of individual refugees, gained in the discussions a new meaning through an exchange of information on what was produced in these separate ‘homelands’ and how the geographical conditions looked like.”[87]

VII. The Global Players of Oromo Ethnicity

In his analysis of the role geographical features played in the development of Ethiopia, Donham comments in a footnote that “Ethiopia is, as it were, an upside-down China” (1986, 255). Whereas in China the different rivers traditionally offered the possibility of communication between groups that usually remained separate, in Ethiopia they were seen as obstacles and barriers. Following this analogy on a political level, I propose to see contemporary Ethiopia as an upside-down state. Although the imaginations of the state still involve the “traditional” narrative strategies of verticality and encompassment (Gupta and Ferguson 2002, 982) – despite the contemporary round of restructuring processes under the TGE and the Federal Democratic Republic of Ethiopia (FDRE) - and although this clearly has an impact on the everyday life of its subjects, the actual practices of state-actors are to a large extent directed towards actors outside the Ethiopian territory they claim to represent. Partly this has to do with the fact that the state is dependent on outside resources for its very survival, while the state is at the same time virtually the sole distributor of these resources within Ethiopia. This has clearly engendered an attitude of “upward accountability” (Aalen 2002, 85) among those actors with access to the resources of the state. At the same time, however, the Abyssinian idea of social mobility I referred to a number of times in this thesis still seems to have retained much of its explanatory force. Actors are not simply made dependent on the state; many of them willingly submit to this state, since it offers possibilities for social advancement and a flight from the land. As I have pointed out, the OLF largely replicated what it rejected and can be seen as a state-structure (or at least as an attempt to construct such a structure) without the state. It was originally organized along Marxist-Leninist lines, which centralized power (in theory at least) in the hands of a small group of intellectuals. It had its own Foreign Relations Office that channeled global resources towards this elite. And its actual spatial practices were directed towards actors outside the Oromia they claimed to represent: Oromo in the Diaspora and interested or strategic non-Oromo were their main audience.

Much of this has been argued in the previous chapters. This chapter wants to tie these different strands together and synthesize the developed arguments supporting my main thesis that the lack of “common ground” between the OLF and the Ethiopian government after the OLF left the TGE in 1992 is not only the result of the monopoly mechanism of the Ethiopian state,[88] but also intimately related to the fact that the OLF-version of Oromia is embedded within a diasporic network. Some of the developments that took place since the regime change in 1991 will be discussed to lend further weight to my thesis. The first part of this chapter takes the more or less conventional state-centered route to explaining politics in Ethiopia: it discusses the political changes since 1991 as an expression of the age-old Abyssinian tradition of seeing competition as treason and thus of politics as essentially concerned with establishing a monopoly of power. There are good reasons to take this point of view, since there are numerous examples supporting it: the withdrawal of the OLF and other opposition parties before the 1992 elections due to harassment and a lack of transparent electoral institutions and procedures; the suppression of journalists and intellectuals dedicated to the freedom of speech; and the official privatization of the economy that in reality involves a transfer of these businesses to individuals linked to the ruling party, to name but a few. The second part of this chapter transcends this important but too constraining view - this “narrowly constructed world of rigor” (Migdal 2001, 3) - by incorporating the Diaspora in thinking about politics. After a conceptual critique of a number of state-centered analyses of contemporary Ethiopian politics, I will show how the OLF needs to be conceptualized as embedded within a diasporic network and how this – and not so much political changes inside Ethiopia - largely determines their political stance. The third part of the chapter adopts this diasporic line of thinking and follows some of the implications of the resulting “disruptive spatiality” (Harvey 1989, 301-02) for politics in Ethiopia. A number of other connections between the diasporic OLF and Ethiopia will be briefly discussed: armed resistance on the peripheries of the Ethiopian state; the international diplomatic community as an instrument of global legitimization enabling continued involvement in Ethiopian politics; and the role of non-Oromo mediators in promoting the OLF’s spatial imagination of Oromia. Finally, in the fourth and last part of this chapter I would like to embed both the OLF as the EPRDF/OPDO in a broader context and in this way complicate my own argumentative logic. The emphasis both the current government as the major insurgent organizations put on ethnicity and the primacy it is accorded in conducting politics – both on the scale of the ethnic nation as on the scale of the state – has led to a highly dynamic situation, in which these earlier movements might have produced politicized ethnicity, but can no longer control it. Instead, they have to compete with other imaginations of “being Oromo” that are on offer.

The implosion of the Derg regime and the flight of Mengistu Haile Mariam to Zimbabwe opened the political playing field to those liberation movements that had been fighting against this regime since the 1970s. After a period in the peripheries, they could finally occupy the center of attention as representatives of the Ethiopian state in Addis Ababa. Ethiopia, as I already mentioned, should be seen as an upside-down state, since Ethiopian politics always involves the acknowledgment of actors without the state. Legitimacy among your own people is useful, but not necessary for political survival; legitimacy among important inter- or transnational actors – such as governments, NGOs, Diaspora communities, activists and merchants – however, is. The US-sponsored London Peace Talks in May 1991 that involved the EPLF, the EPRDF, the OLF and the government exemplified this. When the government collapsed, the liberation movements remained.[89] The selection at this time seemed largely based on the US-perception of the armed strengths of these groups and on the successful attempts of these organizations to exclude others not involved in armed struggle against the Derg. Since the most successful movements that had fought the Derg saw themselves as ethno-national liberation movements fighting against the centralizing tendencies of the Ethiopian state,[90] their imaginations and those of its patrons have framed politics since 1991. Although the July conference in Addis Ababa was more inclusive, the framework as developed in London determined the proceedings. Representatives of different “ethnic groups” were selected and invited to join the Peaceful and Democratic Transitional Conference of Ethiopia. Of course, many of these ethnicities had not existed before in any politicized sense and its representatives often became those actors “accidentally present in Addis Ababa” (Pausewang, Tronvoll and Aalen 2002, 29). The EPLF did not participate in this conference, but opted for observer status, making clear that it considered the independence of Eritrea a fait accompli.[91] At the conference, the Transitional Charter was adopted, which was drafted jointly by representatives of the OLF and the EPRDF, in which the Universal Declaration of Human Rights was adopted and democracy was embraced. Most importantly, the result of this cooperation between the OLF and the EPRDF was an incorporation of the possibility of self-determination and even secession from the Ethiopian state for the different nationalities.[92] In return, the OLF dropped its implicit demand for immediate autonomy of Oromia and opted for participation within the Ethiopian state.

Participation with and subsequent rejection of – i.e. exclusion from (depending on your perspective) – the central power is a recurring phenomenon in Ethiopia and the political developments since 1991 have only corroborated this view. The removal of the Derg regime, the London Peace Talks and the Addis Ababa Conference enabled the contestation of previous dependencies between different actors, while it simultaneously directed the attention of all actors towards this state, since it was from this state that new resources could be expected. The subsequent monopolizing tendencies of the EPRDF and thus the failure of the state to fulfill the raised expectations has led many groups to reject the state and to search for a larger piece of a different cake elsewhere.[93] During these instable periods, breakaway factions from opposition groups are highly likely (McGinnis 1999, 8). The logic behind these splits, however, is based not so much on an objective analysis of the material situation, as McGinnis’ instrumentalist view would have it, but more on an expectation of what kind of advantages to expect in the near future. Naturally, the specific positionality of different actors leads them to analyze this situation in diverging ways and is strongly informed by personal animosities between actors or their embedment in multiple and only partly overlapping socio-spatial networks. In the case of the Oromo, a mixture of older and newer organizations struggled for their representation at the scale of the state during these transitional years. The Charter established the Transitional Government of Ethiopia (TGE) consisting of a legislative (Council of Representatives) and an executive (Council of Ministers). In the Council of Ministers only the major national liberation movements and a number of “technocrat individuals” (De Waal 1994, 26) were included. The leader of the TPLF, Meles Zenawi, also became the leader of the EPRDF coalition and the OLF was given the ministerial posts of Agriculture (Zegeye Asfaw), Education (Ibsa Gutama) and Information (Dima Noggo). As part of the EPRDF, the OPDO was the only other Oromo-organization represented in the executive. Founded in 1990 by the TPLF as an alternative to the OLF, which refused to pledge loyalty to the ideological program of the TPLF, in the early stage it consisted mainly of captured Oromo soldiers from the Derg army.[94] The legislative role was fulfilled by the Council of Representatives and included a larger repertoire of organizations, although out of the 87 seats 43 went to the EPRDF (within this coalition, the OPDO was accorded ten seats). The OLF, as the only larger “independent” Oromo-organization, received twelve seats. The IFLO under the leadership of Sheek Jaraa – in official existence since 1986, but the result of a 1978 split with the Sudan-wing of the OLF – was given three seats. Other organizations (each with one seat) included the earlier SALF that had in effect split into the United Oromo Peoples’ Leadership (UOPL) and the Oromo-Abbo Liberation Front (OALF). Similar to the OLF and the IFLO, the leaders of these two smaller organizations know a long history of resistance against and cooperation with the Ethiopian state. Both Wako Gutu of the UOPL and Haji Isahaaq of the OALF already played an important role in the 1960s Bale Rebellion and the later formation of the SALF. The major change has been a shift away from “being Somali” towards “being Oromo.”

The aspect of contemporary Ethiopian politics that has received most attention is the restructuring of the state along ethno-national lines. Both to foreign as well as Ethiopian observers (usually outside the country, it must be added) this seemed to constitute a break with the age-old Abyssinian tradition of centralization. To the comparatively inclined, Ethiopia became an exciting case-study that could offer interesting data on the governance of ethnicity, which might be useful in understanding and dealing with “ethnic conflicts” in other states as well. The newness of these developments, however, the usual complexity of the situation and the personal involvement of many commentators have led to a wide range of opinions on how to interpret these changes. According to Kendie (1994, 147): “the fact of the matter is, Ethiopia is disintegrating – and this disintegration is going to take others down as well, since Ethiopia’s breakup will not occur within the confines of a sealed territorial container.” Krylow (1992, 240) believes we “can expect a very unstable region from Sudan to Somalia, including Djibouti” and Engedayehu (1997, 192) writes that “the policies of the current regime, especially in respect of ethnic federalism, are likely to lead the country to ethnic disintegration rather than political pluralism.” This view is partly complemented by those that emphasize the authoritarian nature of the new government (Ottaway 1995; Harbeson 1998), although it should be added that authoritarianism in itself does not logically lead to disintegration, but can also lead to stabilization and even democratization. Young (1996, 532) also implicitly subscribes to this importance of the state when he points out that this attention to ethnic rights is a “response to the legacy of ethnic domination and marginalisation in the Ethiopian state” and “perhaps the only approach that could ensure the unity and survival of the Ethiopian state into the twenty-first century.” Henze (1998), a former American diplomat who worked at the National Security Council, has embraced the democratic potential of the state and described Ethiopia as a pluralistic and democratic society. In doing so, he largely follows Adhana (1994, 28), who has argued that the post-Derg period marked “a complete break with the 100-year-old dominant politics” and that the time had come for a “politics of ‘to be’ (revolutionary democracy) in contemporary Ethiopia.” Of course, the reference to the concept of revolutionary democracy is hardly a neutral act – it has been a central term in TPLF-discourse since the 1980s – and clearly reveals the allegiances of Adhana (and those of Henze).

As a matter of fact, no one really knows where Ethiopia is heading. Although there seems to be an increasing consensus on categorizing Ethiopia as a semi-authoritarian state (Ottaway 2003) that lacks most features typical of “democratic societies,” such an interpretation - while useful as a benchmark of the present - cannot really offer us any satisfying imaginations of possible futures. It might very well be that in the long run the devolvement of power to scales below the state is a starting point for a process of democratization (Pausewang 1994), despite the fact that in the short run the main effects are increased state control and clearly “undemocratic” relations between the state and its citizens i.e. subjects. At the same time, although a transitional period of semi-authoritarianism as a prelude to freedom and democracy (which remarkably resembles the view that sees the increase in inequality as the result of Structural Adjustment Programs as a necessary painful step to more wealth and equality for all) might be acceptable to a number of observers, it is highly unlikely that this is accepted by those actors – such as the OLF and many other “oppositional” organizations - already skeptical of the benefits of the Ethiopian state. I will come back to this aspect in the next parts of this chapter, since it is closely related to the Diaspora and the connections that exist between the Diaspora and Ethiopia. For the purposes of this part, I will adopt the view that sees post-1991 Ethiopian politics as a continuation of older strategies that mainly involves the monopolization of power. In my view, the institutionalization of “ethnic federalism” (Cohen 1995, 1997) as it is currently taking place was largely unavoidable due to the fact that the strongest opposition movements against the Derg made use of the spatial imaginations of ethno-nationalism, but it has been creatively appropriated by the central EPRDF to increase their control over the state and its territory.

The main strategy has been to use the idea of ethnic nations as separate entities - developed in a different context and embraced by most important liberation movements – but to appropriate it in such a way that paradoxically the power of the regions is weakened and the control of the center is enlarged. If anything is typically “Ethiopian,” then this is it: political actors talk about the peasants and the land, but actually mean the intellectual and the city (although it is unclear if the intellectuals are aware of this). In that sense, the transitional period after 1991 knows striking similarities to the years following the 1974 revolution. After this first revolution, students were sent into the rural areas to teach the peasants about the benefits of the revolution and to show how they should organize themselves in peasant associations. Precisely these associations, however – looking so marvelously socialist and democratic on paper in Addis Ababa – were not much later used by the central Derg regime to subject the peasants in the countryside to increased control. In similar ways, the actors that were part of the TGE talked about the law of the land and the need to liberate the ethnic nation from the oppressive state, but the result has been increased control of the subjects on Ethiopian territory through the institutionalization of the region. This time, the issue of control is inherent in the very concept of the region itself. As I pointed out in the theoretical section of this thesis, a region always involves the delimitation of an area as a bounded space i.e. as separate from other bounded spaces. There is hardly anything natural about this delimitation; instead specific actors appropriate this imagination of the region and “use” it in their daily practices as a “gate-keeping concept” (Appadurai 1986, 357), making sure that they are the privileged ones with the key to this gate. This position is privileged not only because it enables control over the subjects within this region, but more importantly because this position offers connections with other strategic actors outside this region. In that respect, the EPRDF/TPLF and the OLF concur with one another. Both organizations originate from the same 1970s intellectual environment and both have adopted a similar Marxist-Leninist view of the state that beliefs in the image of a small intellectual cadre guiding the (nation-)state and its people to better and more modern times. It is therefore also highly unlikely that the OLF would have pursued a different, more democratic course as the EPRDF in the actual institutionalization of the region of Oromia. The struggle between both parties is not so much about the actual contents of this region – debating about contents is a practice most Ethiopian intellectuals tend to despise - as it is about who controls the points of access to this region. Most of the events since 1991 have backed up this interpretation. The ancient “monopoly mechanism” (Janssen (1991, 1994) and Auf (1996)) still seems to pervade contemporary Ethiopian politics.

Nevertheless, the short period of cooperation between the OLF and the EPRDF from 1991 until a few days before the elections of 1992 might prove to be fatal to the rule of the EPRDF in the long run, since it was during this year that the first important steps were made to realize the spatial imagination of Oromia in practice. In doing so, the region of Oromia was not only connected to the center of Addis Ababa as part of the federal framework of Ethiopia, but also to a more radical diasporic network of OLF-supporters. The problems this entails for conducting politics in Ethiopia will be discussed in the next parts of this chapter. Let me here briefly discuss some of the “OLF-elements” introduced during this period. The conversion of the spatial imagination of Oromia into the region of Oromia involved negotiations between the OLF and the EPRDF and the use of a 1984 language survey (itself partly based on linguistic classifications of European scholars) (Schlee 2003b, 346-47).[95] Of course, hardly any area within this region satisfied the ideal of a pure ethnic composition and the final result was based as much on “ethnic” criteria as on questions of power and the access to material resources.[96] Within this contested region, however, all Oromo organizations decided to use afaan oromoo - written not in the Ethiopian but in the Latin script – as the language of communication in the administration and primary education of Oromia. Before 1991, only the OLF in the western areas of Wellegga and Sudan and the Oromo Diaspora in Europe and the United States had used the language in this script and accorded it centrality in their national liberation struggle. The result of this 1991 consensus was that, according to Mekuria Bulcha, “textbooks were prepared in a matter of four months and school teachers were given crash courses in the Oromo language” (1993b, 16). For the first time in more than fifty years, Oromo had the opportunity to use afaan oromoo in public. This is one of the major achievements of the OLF that can hardly be underestimated. At the same time, the language issue in combination with the restructuring of the state along ethno-national lines has caused its own dynamic. Although the OLF left the TGE in 1992, command of afaan oromoo now offers to many Oromo easier access to education and employment, since – at least within the region of Oromia – they no longer face direct competition from other non-Oromo.[97]

The cooperation between the OLF and EPRDF was, however, troubled from its very beginning. At the London Peace Talks it had been agreed that the different liberation fronts should continue to administer the territory they had already controlled before the flight of Mengistu. Other areas were to be controlled by the EPRDF coalition. This seemed the most practical solution, but almost immediately after this agreement their ensued a debate on the actual size of the areas controlled. Just one month after the Addis Ababa Conference, the EPRDF and OLF were fighting each other in the southeastern part of the country. A series of agreements in order to end this violence followed, but this was largely unsuccessful, since the parties did not always control or could not always communicate with their own troops (De Waal 1994, 27-28). During this period, banditry, local grievances and national ideologies intermixed, creating a highly volatile as well as confusing situation. The question who was to control the land and thus have the monopoly on representing the people on this land became central during the months before the local elections of May 1992. The largest part of Oromia was controlled by the OPDO and in these areas OLF offices were closed down, its representatives harassed and in some cases even killed. At the same time, in the only OLF-administered area around Dembi Dollo in western Wellegga, the OPDO faced similar conditions: their office was closed in the week before the elections and its representatives were harassed (Pausewang, Tronvoll and Aalen 2002, 30-32). The strategies of the two parties were not very different, but the EPRDF had a quantitative advantage, since it had a larger military force available and could thus control the major part of Ethiopia. As a result of the obvious harassment and faced with this political hegemony of the EPRDF, the OLF decided to withdraw from the elections and start an armed resurrection: the usual route for Ethiopian actors after losing the struggle for political power in the center. Although EPRDF forces almost immediately put down this first military attempt and around 20,000 OLF-members (suspected or real) were imprisoned (Young 1998, 194), since 1992 the OLF has been involved in scattered military opposition on the peripheries of the state and occasional meetings with the government in Addis Ababa or urban centers in Europe and North America.

The example of the OLF was followed by several other opposition parties, including the IFLO, enabling the EPRDF to gain an easy victory in the 1992 elections and establish hegemonic control of the center over the different regions. Aalen (2002) has most comprehensively shown how the federalist structure of the Ethiopian state has been used by the central government to enforce static control over its territory. Her report adds a practical component to earlier articles that mainly focused on the constitution and official proclamations from the TGE and FDRE.[98] Others have investigated some of the fiscal and financial developments since 1991.[99] Despite a range of opinions on how to interpret the policies of the current government, it is clear to all that the different “ethnic regions” will remain dependent on the center for quite some time, if only because of the federal government’s domination of revenue collection. According to Cohen: “[t]he center’s domination of revenue is due to the fact that the most significant sources of revenue come either from import and export taxes, which are constitutionally outside the jurisdiction of the regions, or from excise taxes levied on manufactured goods, which are largely collected in Addis Ababa. If external assistance is considered, the federal government controls approximately 90 percent of total revenue” (1997, 145-46).

Instead of reiterating these important analyses in a few lines, however, let me here briefly discuss another aspect of Ethiopian politics that has remained relatively constant over at least the last forty years or so. The tendency in Ethiopia to monopolize politics has been observed by many authors, but this seems to be conducted in such a radical fashion that politics in Ethiopia essentially involves a progressive exclusion of elites from the center of power. A case could be made for interpreting the regimes of Haile Selassie (at least the last decade of his rule), Mengistu Haile Mariam and Meles Zenawi not so much as a politics of monopolization that creates stabilization and maybe even democracy, but instead as a politics of exclusion that constantly creates differences between elites that did not necessarily exist before. To many outside observers, this often makes from Ethiopian politics a kind of farce with actors engaging in discursive battles that seem far removed from the day-to-day problems of the overwhelming majority of Ethiopia’s inhabitants.[100] One of the latest episodes took place in 2001, when a crisis split the TPLF Central Committee into two opposing camps and also affected the parties in the EPRDF-coalition, in particular the OPDO. Although the seeds of the crisis were sown during the war with Eritrea, it reached its culmination when, during a TPLF self-criticism meeting, Meles Zenawi presented a paper entitled “Bonapartism is the main danger,” in which he argued that the TPLF leadership was decaying and becoming distant from its constituency. Quite a realistic analysis, it must be said, but once again the practical result was the precise opposite of the presented discourse. Similar to the “capture” of the peasant associations by the Derg and the federal structure based on ethnicity by the EPRDF, the result was not a consensus to close this gap between party and constituency, but paradoxically the assuming of a position of unchallenged supremacy by Meles Zenawi (Tadesse and Young 2003, 397).[101]

The 2001 crisis within the TPLF also had consequences for the OPDO. Negasso Gidada of the OPDO, President of the FDRE since 1995, and long-time Oromo nationalist in Ethiopia as well as the Diaspora, had to resign from his post. The OPDO split into those who continued supporting the line of Meles Zenawi and those who opposed him for differing reasons. About half of the central committee of the OPDO defected to the OLF (Aalen 2002, 104). This was almost ten years after the OLF had left the TGE in the days before the 1992 elections. At that time, many of the OLF-fighters were arrested by the EPRDF and put in camps, but the OLF-ministers left the country to join the Diaspora in Europe and the United States (Pausewang, Tronvoll and Aalen 2002, 33). As I see it, it is at these junctures that politics and globalization meet. They also point to the limits of a state-centered approach in understanding contemporary politics.[102]

One disadvantage of the state-centered approach is that it tends to ignore that political actors nowadays have a range of options available: the state is only one of the routes to political action and personal advancement. If actors are excluded from the center of power, as continuously happens in Ethiopia, they can almost always – providing they are not killed or imprisoned - pursue their political goals elsewhere. This also implies, however, that it is highly likely that actors within the political system of the state will weigh one route against others. If the state no longer seems the best avenue to fulfillment of their imaginations, they will either turn to scales below the state - as has taken place throughout Ethiopian history - or leave the territory of the state and go into exile. In other words, it is necessary to accord activity and powers not only to the state, but also to the Ethiopian actors supposedly encompassed by this state, since these actors establish connections with and are embedded in socio-spatial networks that cannot be controlled by the mechanisms of the state. This seems particularly true in the case of Oromo nationalism. As I have pointed out before, the period between the 1991 Addis Ababa Conference and the 1992 elections embedded the region of Oromia – and the actors perpetuating the existence of this region – not only in the spatial structure of the state, but also in a diasporic network that had become used to imagining an Oromia without Ethiopia. In a simplified way, one could say that this has created an identity crisis for Oromo intellectuals in Ethiopia: every act of suppression by the EPRDF/OPDO of “Oromo culture” will be interpreted both through the framework of the OPDO as through the framework of the OLF. Since the OLF is not involved in the day-to-day politics of running a state, it is quite likely it will be considered the moral winner of the game.

On a conceptual level, this also complicates a materialist explanation presented, among others, by Markakis (1994, 2003). As he writes: “There may be a host of contributing elements in the generation of a conflict […] however, the prime cause and catalyst, more often than not, is a struggle for access to scarce resources, i.e. a struggle for survival. […] In the Horn, state power is the most direct and effective means of gaining access to scarce resources. Consequently, a share of state power is most often the bone of contention at first instance, and most conflicts involve the state in one form or another” (2003, 362). Personally, I doubt if this materialist interpretation is (still) applicable to many of the conflicts in the Horn and in particular to the conflict between the OLF and the EPRDF. Markakis’ view works with a notion that sees society and culture as directly connected to a material base, whereas I argue that to understand the conflict between the OLF and the EPRDF one has to acknowledge the disjuncture between these material and socio-cultural aspects. If one accepts that the region of Oromia is connected not only to the Ethiopian state, but also to a global diasporic network, can one still argue that the state is the “bone of contention” or is it more likely that the state is the unintended victim of clashing networks? There seem to be good reasons to adopt the latter view.

This is because ever since its original exclusion from the center of the Ethiopian state in the late 1970s and even more so after its renewed exclusion just before the 1992 elections, the OLF has relied on a Diaspora audience for ideological (and financial) support. As I pointed out in the previous chapter, the stay of the OLF in Sudan gave them the advantage of direct connections with Oromo student organizations and other supporters in Europe and North America. These connections did not break during the short period of cooperation between the OLF and EPRDF in the TGE, but played an important role in the “self-inflicted exclusion” (Abbink 1995, 154) of the OLF from the Ethiopian state in 1992. The harassment of the OLF by EPRDF forces is beyond doubt, but fact of the matter is that the OLF had to sell its further cooperation with the Ethiopian state not so much to the Oromo people inside Ethiopia, but to Oromo in the Diaspora. Partly, this is acknowledged by Leenco Lata, who participated as a representative of the OLF in the drafting of the Transitional Charter, when he writes that protests against this harassment “failed to stop rising TPLF/EPRDF aggression, and the OLF leadership was facing mounting pressure from an already skeptical community and rank and file. The history of European and American support to Haile Selassie and of Soviet support to the Derg to quell previous Oromo resistances was being recalled to underline the risk of putting any trust in the US and its allies. Endless cases of bitter Oromo experiences with Abyssinian treachery were also being recollected to emphasize the danger of vesting any kind of confidence in either the TPLF/EPRDF or the EPLF” (1998, 63; Italics in original). Of course, as an OLF-representative he does not differentiate between the Diaspora and the “homeland,” but this interpretation of events is clearly informed by over a decade of Diaspora experience. The OLF has changed a number of times its official position on whether to fight for a sovereign nation-state of Oromia or whether to accept a federal solution within the Ethiopian state, but it has always had to walk a tight-rope between the imaginations of its audience in the Diaspora and what it thinks are the wishes of the Oromo inside Ethiopia. In addition, most of the OLF-representatives reflect the position of their audience, since they are themselves a product of the 1970s student movement that was forced into the peripheries and fought against the Derg. In a 1997 interview, the secretary general of the OLF, Galassa Dilbo, still interpreted the conflict in Ethiopia as following: “One of our problems in Ethiopia is that we are dealing with two opposed cultures: on one side there is the democratic culture of the Oromo and other peoples in southern Ethiopia and on the other side there is the autocratic culture of the rulers of north Ethiopia” (Melchers 1997, 2).[103] In 1999, Galassa Dilbo was replaced by Daoud Ibsa Gudina who has undergone a similar rite of initiation as most others in the OLF. Born in the western part of Wellegga and educated at Addis Ababa University, he joined the first batch of OLF fighters in the regions of Bale and Harer. Although he was captured by the Derg and imprisoned, he managed to escape. When the Derg regime imploded in 1991, he returned to Addis Ababa and worked for the OLF until the movement was once again excluded from the center.[104] Needless to say, it seems unlikely that this selective admission procedure to the higher ranks of the OLF will lead to a more conciliatory policy towards the Ethiopian state: the parties involved speak separate languages, making use of different and exclusive “spatial ontologies” (Harvey 1989, 301-02).

Therefore, instead of interpreting the conflict between the OLF and the TPLF/EPRDF as a “struggle for the state,” it might be more useful to conceptualize both parties as socio-spatial networks that tend to perpetuate what Bourdieu would call a specific habitus: a set of generative schemes that produces practices and representations without reference to overt rules and without necessarily the rational and conscious selection of goals and ways of achieving them.[105] This is what makes the disagreement between the OLF and EPRDF so intransigent: not only have both parties been able to develop separate networks that support their existence, but the habitus in which the actors are immersed has naturalized this separateness. This can be illustrated by taking a closer look at the situation among Oromo in the Diaspora. In a recent book by Matsuoka and Sorenson on the construction of communities among Horn of Africa refugees in Canada, one chapter is devoted to an analysis of Oromo Studies Association (OSA) meetings between 1989 and 1994.[106] The following quote clearly reveals the emotional and personal aspects at stake in constructing a spatial imagination of Oromia that is radically different from Ethiopia:

“During the sessions, the Ethiopian government is condemned and audience members are urged to identify fully with Oromo identity and with the OLF’s political program, which demands an independent state. Speakers pose explicit ‘challenges’ to the audience: to commit themselves, to contribute financially, to help the OLF and the Oromo Relief Association (ORA), to learn more about Oromo language and culture, and to adopt authentic Oromo identity to replace the artificial Ethiopian identity that was imposed through violence and indoctrination. The audience members respond to these challenges, not only by endorsing (and thus encouraging) pro-Oromo statements with applause and comments, but also by making emotional public pledges of commitment and support. Imbedded in this pattern of assertions, exhortations, and responses are persistent vilifications of the Abyssinian or Ethiopian Other.” (2001, 182-83)

Similar situations have been reported by Zitelmann in his work on Oromo refugees in Sudan (as discussed in the previous chapter) and Germany (1992a) and by Gow (2002) in his research on the Oromo community in Melbourne, Australia. This is the audience to a certain extent created by the OLF, but it is at the same time the audience the OLF is dependent on for support. This became clear during a 1999 meeting of Leenco Lata and Dima Noggo – at that time both ex-leaders of the OLF - in Atlanta City with Oromo from North America. Leenco Lata and Dima Noggo argued that the OLF should abandon its traditional position that Oromia was colonized by the Ethiopian state and therefore should fight for independence, since this was unacceptable to western governments and a majority of the people in Ethiopia, leading to an isolation of the OLF. A number of Oromo present at this meeting, however, “notably more radical ones from Minnesota State,” rejected this conciliatory stance and further defended the traditional position of the OLF.[107] Although one should certainly not reify its importance, at least in this case the impact of spatial constructions in determining actors’ behavior seems clear.

This importance of the Diaspora in influencing the policies of the OLF and thus in contributing to the political impasse between the OLF and EPRDF begs the larger and more general question how the Diaspora and Ethiopia are connected: how do these diasporic imaginations influence the Oromo “back home” in Ethiopia? Although research on this aspect of Oromo nationalism is virtually non-existent, a number of aspects seem to play an important role. In an idealtypical way, one can distinguish three types of connections.

The first connection takes the old-fashioned route typical of Horn of Africa politics and involves the search for support among neighboring regimes. In the case of the OLF, Eritrea nowadays plays a central role. During and after the 1998-2000 war with Ethiopia both sides have once again started supporting each other’s opposition movements. This kind of strategy has stayed constant throughout the decades and usually involves the more or less effective combination of culture and arms. In the closing months of 1998 Eritrea added a program in afaan oromoo to its radio station “Voice of the Broad Masses of Eritrea,” which is located in the capital of Asmara. Not much later it sent five Antonov planeloads of arms to Somali warlord Hussein Aydeed, who forwarded a certain amount to the OLF and possibly also to Wako Gutu of the UOPLF.[108] It seems very likely that this Eritrean support has also led to some sort of union between different Oromo oppositional movements. During a September 2000 conference held in Asmara, the United Liberation Forces of Oromia (ULFO) was formed and Wako Gutu was made chairman. Other parties included the familiar OLF, IFLO and OPLO and smaller or newer (or renamed) groups such as the Oromo Liberation Council (OLC) and the Oromo Peoples’ Liberation Front (OPLF).[109] Of course, these splits and shifting alliances are nothing new in the Horn and the specific details are less important than the observance of a general shift away from Somalia towards support by Eritrea. Among writers in the Oromo Diaspora, the formation of the UFLO has also been digested, but it is effortlessly included in the older binary imagination of Oromo resistance against the Ethiopian state.[110]

The second and third type of connection relate to the idea of Ethiopia as an upside-down state. Sorenson (1993) has rightly identified the trope of the “foreign agent” used by most opposition movements as well as the EPRDF government to deligitimize diverging opinions of others and to establish the authentic and autonomous character of the own organization. In the analysis of the OLF’s representation of the Bale rebellion (see chapter V.3), I have tried to show how this applies to the OLF as well. Once again, however, the imagination stands in direct contrast to the practices of the OLF that, to a large extent, take place within the realm of international diplomacy – the second type of connection. Henze is partly right when he argues that the attempts to bring the EPRDF and opposition parties together and to achieve a degree of power-sharing “had the effect of strengthening the pervasive rejectionism of opposition politicians, who voiced ever more provocative demands on the government for cooperation, while concentrating their attention on foreign diplomats and NGOs rather than on the Ethiopian electorate” (1998, 47). However, in his overwhelming enthusiasm for the sincerity of the Ethiopian government he ignores the fact that his critique of the opposition movements is equally applicable to the government as well. What is essentially involved is an externalization of legitimacy: both the EPRDF and the OLF need these “external” connections (and not necessarily those with their subjects inside Ethiopia), since a break would endanger their very survival. At the moment, however, it is the EPRDF that possesses the best cards, since the current emphasis on stability and the fight against terrorism in the discourse of the United States and European governments has enabled the EPRDF to present itself as the protector of a relatively stable state in the middle of a geopolitical region of instability. In the eyes of its donors, anything is better than an Islamic regime in Sudan and an imploded Somalia harboring Islamic terrorists. That this current Western fascination for order, structure and stability perfectly overlaps with the older image of an Ethiopia defending its Christian civilization amidst a sea of barbarism probably remains unnoticed.

The power of this older image can hardly be underestimated, since it has to a certain extent become part of the collective imagination of most actors worldwide.[111] This also meant that it has been easier for those claiming to represent “the Ethiopians” and “Ethiopia” to establish connections than it has been for those claiming to represent “the Oromo” and “Oromia.” The supposed naturalness of the international state system as a global ordering mechanism only tends to reinforce this logic. The third type of connection, therefore, bypasses the official representatives of states and supra-national organizations such as the United Nations and the African Union (that both defend the concept of state-sovereignty) and involves other transnational actors that take on the role of mediator between the OLF and an interested public around the world. Within the academic context, authors such as Greenfield and Baxter, who were among the first to address the “problem of the Oromo” (Baxter 1983) in the Ethiopian state, have taken on this role and have in this way provided a possible alternative reading of the Ethiopian state that was too often ignored by its defendants.[112] At the same time, the “problem of the mediators” is precisely that they more often than not adopt the view – either knowingly or unknowingly - of the “native intellectuals” that themselves stand at the margins of the group they claim to represent.[113] Hasselblatt from the Berliner Missionswerke in Germany – as discussed in the previous chapter – clearly occupied this role during the 1980s. In their book, Matsuoka and Sorenson describe the talk at the 1993 OSA conference of a Rev. Ward who had visited Oromia in that same year and who reiterated the well-known OLF argument: “Our enemies are not sleeping. The EPRDF is setting up a puppet OPDO government in Oromia while doing anything to discredit Oromia” (2001, 192). Trueman who leads the Oromia Support Group (OSG) in the United Kingdom offers another example.[114] Although OSG claims to be a non-political organization that attempts to raise awareness of human rights issues in Ethiopia and that urges governments to withdraw support from the EPRDF (how such a goal could be non-political is not quite clear to me), Trueman’s allegiances are clearly on the side of the OLF.[115] A final example would be Pollock, a Scottish aid worker and human rights activist, who visited the region of Oromia between December 1995 and January 1996 and produced a report on the human rights situation. What makes her not a commentator on the situation of the Oromo, but above all a mediator for the OLF, is not the falsification of facts, but the embedment of these facts within a narrative structure originally constructed by the OLF. And it is this narrative that is being presented to a larger public through the publication in different media. Extracts of Pollock’s report were published in The Herald (Glasgow), New African (London), Share (Toronto) and Urjii and Nabalbal (Addis Ababa) (Pollock 1996, 5). In this way, networks between actors are constructed that encompass the world and that transport the spatial imagination of Oromia. The strength of these mediator-based networks is that its actors often personally identify with the cause being promoted. Above all, these networks are independent from the mechanisms of the Ethiopian state, but can at the same time bring about changes in the behavior of actors inside Ethiopia. An indication of this might be given by quoting from a letter Pollock (2003) send to Urjii, a newspaper in Addis Ababa that is close to the position of the OLF and for this reason regularly suppressed by the EPRDF. As she writes:

“How wonderful and inspiring to see such new developments towards the Oromo cause. […] I wish to see more dedicated Oromos rise up and take their place. Be proud to be an Oromo, for there is no shame in it. If you do not know who you are, how do you know where you are going? Knowing your identity is central to this question. Every Oromo needs to examine themselves and ask themselves this question honestly.”

Although it is clearly impossible to prove a causal connection between this letter and the behavior of actors inside Ethiopia, one need only imagine how such a call to reflect on the centrality of your own ethnicity is likely to reinforce the already existing polarizations in Ethiopia.

When a distinguished academic commentator writes about Oromo nationalism in Ethiopia and uncritically quotes the articles and books of Oromo-writers in the Diaspora, then one can surely say that the mediators have been successful.[116] They have succeeded in at least partially displacing the Ethiopianist discourse that occupied a near hegemonic position for the largest part of the twentieth century. This is quite an accomplishment, since the words Oromo and Oromia were simply unknown to most observers (and to many of those that would now describe themselves as Oromo) before the 1970s, whereas thirty years later the first steps have been made to make these words part of the collective imagination of actors around the world. In that sense, Asafa Jalata is actually right when he writes that “Oromo nationalism developed into a mass movement in the early 1990s” (2003, 80). At the same time, he is mistaken in believing that this Oromo nationalism can still be contained by one or more political parties, since one of the “unintended side-effects” (Beck, Bonss and Lau 2003) of the struggle of the OLF and the still incomplete institutionalization of the OLF’s imagination of Oromia under the EPRDF/OPDO rule has paradoxically been a curtailment of the powers of these parties. Politicized Oromo ethno-nationalism is everywhere nowadays, but it has also become public.

The consequences of this popularization of Oromo ethno-nationalism for the future of Ethiopia are quite unclear and defy any coherent narrativization. Of course, this is true for each depiction of “history” and one of the goals of this thesis has been to emphasize how much Ethiopian history is composed of texts and not of facts. As a distanced observer with no stake in the political and personal games of Ethiopia, it is quite astounding to see how exclusive spatial imaginations are being written on the basis of a handful of academic articles and books. There seems to be a widespread reluctance to except at least some of the arguments of historians like White and Ankersmit that history-writing is not so much a reflection of historical truth simply out there, but that this writing is always informed (at the very least) by the specific positionality of the author.[117] The “going public” of Oromo ethno-nationalism has only complicated matters, but once again this has not deterred writers from offering transparent analyses of contemporary Ethiopia. For example, it is surely more complicated than the image presented by Joireman when she writes that “the religious institutions are divided along ethnic lines with the Tigrayans and the Amhara predominantly following the Orthodox faith, while the majority of the Oromo and the Somalis are Muslims” (1997, 405). Such a conflation of ethnicity and religion can, of course, nicely be embedded within a gloomy Huntington-inspired vision of a “clash of civilizations” (1993), but it also involves adopting an image of Ethiopian society as separated along clearly defined lines. In doing so, it not only ignores important regional differences, but also the specific histories of the various political movements that have contributed to the formation of the Ethiopian state along ethno-national lines. Although the IFLO as an “islamic” party might be categorized as a “Muslim-party,” the OLF surely cannot, since it explicitly defines itself as a secular and modernist organization, while it also emphasizes the “traditional” elements of Oromo culture, such as the gada age-grade system and afaan oromoo. It seems highly unlikely that the OLF would suddenly turn “islamic,” although nothing is unthinkable in the Ethiopian situation. In the meantime, the largest private investor in Ethiopia is Mohammed al-Amudi, a Muslim from Ethiopian-Saudi background. Despite this religious affiliation, however, he is hardly interested in an “islamization” of Ethiopia or in a separation of the Somali- and Oromo-areas from the state. Instead, he has embraced a unitary image of Ethiopian development that brings him very close to the position of many Orthodox Christian Tigray and Amhara (Abbink 1998, 121). In other words, nothing is as simple or oppositional as it may sometimes seem in Ethiopia and it is unlikely that a static top-down approach such as the one offered by Huntington is of much use in grasping the dynamics of Oromo ethnic politics in- and outside of Ethiopia. Rejecting this top-down approach, in the remaining part of this chapter I will continue my attempts to offer a bottom-up approach to the construction of spaces and offer some provisional thoughts on how to think through Oromo politics between Ethiopia and the Diaspora and in- and outside of the official political parties.

Although the struggle against the Derg regime and the subsequent experiment of “ethnic federalism” (Cohen 1995, 1997) has thoroughly shaken the older center-periphery structure of the Ethiopian state, it has not reduced the importance of Addis Ababa as the main hub of economic, political and cultural activity. If anything, this might even have increased, since the end of the Derg regime also ushered in a period of migration towards the capital city that now has around five million inhabitants. In chapter V.1 I conceptualized the city as a site of privilege and as part of the older Abyssinian idea of social mobility with Addis Ababa representing the top of Crummey’s social pyramid. Such a characterization still seems useful, although this time for slightly different reasons. Since 1991 the city is a site of privilege, because it is the only area within Ethiopia, in which the EPRDF government allows a minimal form of democracy as a kind of concession towards its international audience. As Pausewang, Tronvoll and Aalen (2002b) write: “The formal structures have allowed some free space in Addis Ababa, and the party has tolerated the capital to develop into a kind of democratic showcase, within strict limits, until it disturbed the central hierarchy” (242-43). Such a performance is possible as long as the embassies, the African Union and the majority of the NGOs remain based in the capital city and as long as other governments pretend this limited form of democracy in the center applies to the peripheries as well.

This centered democracy, however, raises some questions regarding the stability of the current regime. As modern Ethiopian history shows us, it is from the center that a regime can expect its most vocal critics. The peripheries are violent, but not powerful enough to replace those in power in the center. When a regime loses control over the center, however, it will almost inevitably lose control over the state as well: the 1974 revolution certainly provides the most vivid memory of this. The relative privilege actors in Addis Ababa nowadays enjoy, however, has made from the city a kind of locus in which all kinds of different imaginations and practices – ranging from “democratic” to “fundamentalist” – come together and provide a highly dynamic, but also instable situation. The newness of the post-1991 situation is that the (imaginations of the) peripheral movements fighting against the state can be found in Addis Ababa in the form of newspapers, NGOs, commercial companies and politicians also active in legal political parties (Zitelmann 1997, 66). Of course, networks between the center and the peripheries did exist before 1991. The OLF, for example, although its spatial imagination was clearly the result of interaction between intellectuals in Addis Ababa (and urban centers abroad), developed from its very beginning connections with actors in Bale and Hararghe. But in these early years the spatial imaginations largely involved the appropriation of the global discourse of Marxism-Leninism with a touch of “traditional” elements. Since 1991, actors in the center accord these “traditional” elements from the peripheries central place in their spatial imaginations. In both cases, however, it remains unclear if the spatial practices of these actors in any way overlap with their spatial imaginations. Judging on the basis of the intellectual history of Oromo ethno-nationalism I described in the previous chapters, they very likely do not or only to a limited extent. Nevertheless, their ability to produce a radical spatial imagination – and to attract other actors and resources - could threaten the survival of the EPRDF government.

This brings us (back) to the question of transnational connections and the Oromo Diaspora. Not only has Addis Ababa grown in size, the Diaspora has as well. If one combines this with the fact that virtually all Oromo intellectuals either live in the capital city or have migrated – forced or voluntary – to urban areas in Europe or North America, one is tempted to say that the current discourse surrounding the “Amhara-Oromo dispute” (Triulzi 1994) is produced between Addis Ababa and the West. Considering also the importance in Ethiopian history of small groups of intellectuals to bring about change in Ethiopia (Zewde 2002) and the awareness that transnational practices played an important role in the radicalization of the students in the 1960s and 1970s, one needs to investigate more thoroughly the implications these connections have for contemporary Ethiopian politics. What does it mean when much of the “alternative public sphere” (Gilroy 1987) of Oromia is produced in the Diaspora, but “used” in day-to-day life in Ethiopia?[118] In an article on the use of Internet during the Ethiopian-Eritrean war, Schmidt (2001) has offered us some clues. Treating Internet as a political medium, he has described how students and intellectuals in Eritrea and Ethiopia used the different websites of their liberation movements and anti- or pro-government websites to find out about details on the war. A larger audience was reached through the practice of printing and photocopying these online-publications and distributing them among friends and acquaintances. His conclusion is sobering: “Those who expect enlightenment from the Internet can see in the case of Ethiopia and Eritrea that this – at least in the short run – does not necessarily mean enlightenment in the direction of freedom and democracy, but above all enlightenment against the enemy” (5).[119] Similar strategies of “blaming the Other” can be observed in the case of most sites catering to an Oromo-public.[120] All these websites, however, are based on servers in the United States and the contents of these sites are to a large extent produced by actors in the Diaspora.[121] Similar to the letter from Pollock to Urjii, it is unclear if these actors are aware of the implications their own diasporic writings might have in an Ethiopian context.

That said, it is possible – with some effort - to detect a move away from this polarized OLF-inspired diasporic imagination towards views that appropriate elements of “being Oromo” in a more productive and positive fashion. The growth of the Diaspora – particularly in the United States – has not only amplified the voice of “the Oromo,” but it has also led to a process of differentiation that can no longer be subsumed under the heading of the OLF. The OSA, for example, which has always been close to the philosophy of the OLF, is no longer the only academic organization, but has to share this place with the recently formed Oromian National Academy (ONA).[122] At the same time, one can observe an “islamization” of the secular discourse of the OLF. Oromo Muslim Communities have been founded and one can even download audio-files with the Koran in afaan oromoo.[123] Feminist Oromo-writers now address the male-dominated character of the OLF and traditional Oromo culture.[124] Oromia Online, the first website of the Oromo Diaspora and closely related to the OLF, is no longer able to represent all Oromo in the Diaspora, but is now complemented by a whole range of community organizations that involve more local representations of Oromo ethnicity.[125] Although all this should not necessarily imply a rejection of the OLF, it seems reasonable to assume that the growth of these organizations leads Oromo in the Diaspora to reduce their interest in the OLF and increase their activities in the community organizations. Even in Minnesota State, for example, which I described earlier (see VII.2) as a haven for radical Oromo, the unconditional support for the OLF needs to be questioned. Since access to the higher ranks of the OLF seems largely based on experience in the 1970s student movement and the struggle against the Derg, it is quite likely that those radical Oromo in Minnesota are part of this same generation. If that is the case, the largest problem for the perpetuation of OLF-support in the Diaspora is a generational one, since it could be difficult to attract younger actors with no direct experience of the struggle against the Derg.[126] This speculation seems to receive support from the fact that Minnesota has also undergone a process of differentiation. One now finds there a range of Oromo organizations, such as the Oromo Muslim Organization of the Twin Cities, the American Oromo Community of Minnesota, the Oromo Women Organization of Minnesota, the Oromo Evangelical Lutheran Church, the Oromo Seventh-Day Adventist Church and the Ormeda-Oromo Relief Medical Educational Development Association.[127]

The situation in Addis Ababa, in the meantime, is different once again. Although, in principle, actors in the capital city do have access to these diversified diasporic imaginations of Oromo ethnicity, the more radical one offered by the OLF still plays an important role. The 1992 exclusion of the OLF from state politics seems to have reified the image of this organization as a radical alternative to the politics of the EPRDF/OPDO. The obvious monopolization of Ethiopian politics by the EPRDF and the progressive exclusion of elites from this monopoly have only made this antithesis more attractive. The 2001 purge and subsequent defection of OPDO-members to the OLF confirm this interpretation. Nevertheless, the recurring discrepancies between imaginations and practices deserve a somewhat skeptical analysis of this embrace of the OLF. In a previous chapter (see VI.3) I interpreted the spatial imagination of Oromia as a “strategy of extraversion” (Bayart 2000) and discussed three connections between OLF-intellectuals in Sudan and other actors worldwide: material, ideological and personal. As I have shown in this chapter, these connections have lost none of its relevance in the post-1991 era. Incorporating this aspect of globality in the analysis of Ethiopian politics also makes it possible to expose some of the more personal motivations behind the popularity of the OLF and their spatial imagination of Oromia. Of course, the defection to the OLF reflects a rejection of the Ethiopian government. At the same time, however, this move could be seen as one way of getting access to a global network; not so much to fight the Ethiopian state – as the OLF-supporters in the Diaspora might hope – but to get away from the state and join the relative comfort of other Oromo in Europe or the United States. In that sense, the spatial imagination of Oromia offers a way out.

VIII. Conclusion

Transcending the framework of the state in the analysis of political processes that clearly refer to this state is no simple matter. A similar problem arises in the analysis of an ethno-national movement such as the OLF that explicitly uses the spatial imagination of Oromia as a rallying cry. In both cases it is by no means clear if the continuous reference to the state or the ethnic nation actually involves the wish to physically control these territories or if they mainly function as a sort of container in which to transport other dreams and demands. To a large extent, I tend to adopt the latter view and follow Albert (2001), who writes that “it is the principle of territoriality itself and its deep embeddedness as a cognitive and epistemological framework which mandates that all new articulations of ethnic or national identities do base themselves on a strong territorial component and do evoke the belonging to a specific territory over which control is claimed as a central component of identity construction” (2). In other words and simplifying grossly, identity and not the struggle for resources and/or the state - as Markakis (1994, 1998, 2003) would have it - is the key to an understanding of the Ethiopian present.

Such a state-centered analysis is further complicated if one acknowledges the importance of those processes and connections usually summarized under the headings of transnationalism and globalization. In this thesis I have conceptualized these connections through a historical and theoretically informed focus on the interactions between the OLF and the Ethiopian state. Although one could certainly interpret Ethiopia as a dependent state within a global world-economy, I have instead opted for a bottom-up interpretation that accords power to specific actors in appropriating certain spatial imaginations in order to take a stand among other actors. In doing so, I have tried to make clear how in a situation of globality the state is only one of the many possible spatial imaginations and how this complicates contemporary Ethiopian politics. Until 1974, the Ethiopian state was relatively successful in limiting the privilege of transnational connections to those actors in the center. In doing so, it was able to construct a stratified system of control supervised by Addis Ababa. These transnational connections were not destroyed, however, during the revolution and the flourishing of “ethnic” liberation movements should thus not be seen as a relapse into primordial habits by actors that the state could no longer support. Instead, as I have shown in my discussion of the OLF in Sudan, this invoking of ethno-nationalism needs to be interpreted as one way of re-gaining access to the global scale that many had temporarily lost after their exclusion from the center. The flight from Ethiopia towards the Diaspora that accompanied this exclusion provided another way of access to this scale and enabled a struggle against the state without the immediate danger of repercussions.

From an analytical perspective, one of the consequences of this move towards the Diaspora has been a complication of the politics of resistance. Of course, the time when resistance tended to be seen as an unambiguous category, as “the other side“ of domination, is long gone. Thirty years of Cultural Studies and Foucault have made most commentators aware of the intricate and ambivalent nature of power. Resistance, nowadays, is always partial, contradictory and embedded in a play of representations.[128] But even an author like Scott (1990) who wrote in a magisterial way about the everyday arts of resistance largely limited himself to situations that occurred within a more or less bounded geographical area – the city, a forced labor camp or a shipyard, for example. “Being Oromo,” however, is promoted all over the world by a diverse range of actors and these various imaginations interact with one another. Some of these actors – such as those in the higher ranks of the OLF or academics in the United States – have access to all scales from the local to the global through their participation in powerful networks. Others do not have this privilege and are often limited to the production of “being Oromo” on a local scale. The question should not be which version is more true or authentic, but who has the power to impose a certain version on others and through what kinds of practices is this accomplished.

In my view, one of the more striking developments behind this spatial imagination of Oromia and its promotion by actors all over the world has been a growing disjuncture between cultural expressions and its material base – a classic Marxist impossibility, in other words. Considering the fact that Ethiopia is one of the poorest countries in the world, among Ethiopian intellectuals there is surprisingly little talk of economics. Instead, culture is everywhere.[129] It seems quite likely that this fetishization of culture and the perpetuation and even reification of conflict this entails is related to the global popularity of imageries of traditional culture as a radical alternative to the insecurity and dynamism of modern times. Much of the interaction between “native intellectuals” and “outsiders” (missionaries and NGOs, in particular) seems to be based on a common interest in the promotion of such an alternative, although the intentions of each might very well differ. This kind of romantic inclination is, of course, hardly limited to those actors involved in the promotion of the spatial imagination of Oromia, but can be observed in quite a number of other situations. Although each with its own specificities, a more general logic emerges. Here, as a fist impulse to a possible comparative study, I will briefly discuss the case of Western Sahara.

“Western Sahara: The last African colony,” it read on a pamphlet I happened to come across a few months ago.[130] It is always interesting to see how the various liberation movements can be utterly unaware of each other’s struggles, but nevertheless use exactly the same vocabulary – same imagination, different region. The main difference between both is a legal one. Whereas the international community never acknowledged Oromia as a colony, the Western Sahara did receive this status through its inclusion in the 1963 United Nations (UN) list of non-self-governing territories. In 1964 the UN Decolonization Committee adopted its first resolution of Western Sahara, in which the colonial power Spain was urged to start preparing for decolonization. When Spain conducted a census in 1974 in preparation for a referendum that should allow the inhabitants to decide on the future of its territory, Morocco consulted with the International Court of Justice in The Hague for an opinion on the legal status of the Western Sahara at the time of Spain’s colonization in 1884. The Court held in 1975 that Sahrawi leaders had had ties of allegiance to the Moroccan monarchy but that these were insufficient to grant it sovereignty over the territory. Following this rejection, King Hassan II responded by sending 350,000 civilians and 80,000 soldiers into the area – the famous Green March. As a result of the violence, nearly half of the total Sahrawi-population fled from the area into exile in western Algeria.

Besides these historical and geographical specificities, however, the subsequent events have been informed by a similar logic as in the case of the OLF and its spatial imagination of Oromia. Similar to the Oromo refugees in Sudan and their relation to the OLF, the Sahrawi in western Algeria were “captured” by the Frente Popular Para La Liberacion De Saguiet El Hamra Y Rio De Oro (POLISARIO), an organization founded in 1973 by a number of students in order to conduct armed struggle against the Spanish and the Moroccans. Where the OLF received back-up from the Sudanese state, POLISARIO was given logistical and military support by Algeria. Fighting stopped in 1991 after both parties accepted a UN settlement plan, but intractable disagreements on who actually is a Western Sahrawi and thus eligible to vote in a referendum that would determine the status of the territory have led to a political impasse that seems irresolvable. Although most analyses of this conflict see Morocco as the obvious aggressor, a complementary reading seems useful here as well. As Maghraoui (2003) has nicely pointed out, one factor behind this image is a differend concerning the nature of the conflict. The protagonists simply speak two different languages: “it is the language of the right to self-determination versus the language of national integration and national sovereignty” (116). Similar to the conflict between the OLF and the Ethiopian state, this differend has over the years been cultivated on the Sahrawi-side through a range of interactions between the “native intellectuals” (both in the refugee camps and abroad) and interested “outsiders.”

In the attempt to establish connections with strategic actors abroad, the POLISARIO has had it easier than the OLF. The interpretation of the Western Sahara as a colonial question by the UN has enabled the self-proclaimed representatives of this territory and its people to capitalize on the situation: the Sahrawi Arab Democratic Republic (SADR) that was founded in exile in 1976 has been recognized by 75 countries and has a seat as a full member state in the African Union (Williams and Zunes 2003, 3). Of course, this privileged position was nothing natural, but the result of intense lobbying by the POLISARIO/SADR (Pazzanita 1994). Because of this position, however, the organization could function as the sole voice of “the Sahrawi people.” It is also this voice that has been heard and further amplified by many NGOs and solidarity committees around the world: besides the already mentioned Society of Friends of the Sahrawi People in Germany, one finds a range of organizations in Austria, Norway, Slovenia, Sweden, Italy, Switzerland, the United States, Australia, the United Kingdom and other countries, but above all in France and Spain.[131] This transnational network of actors involved in the promotion of “being Sahrawi” is further strengthened through the normalization of exile in the refugee camps in Algeria and the role of the SADR in governing these camps. Governance, of course, is always both repressive and enabling: it controls people through the imposition of certain rules and regulations, but within this framework it creates possibilities of social advancement for those willing to accept these structuring rules. The SADR has, for example, enabled hundreds of students to study abroad, but they are expected to contribute to their community in the camps and the promotion of the image of a Sahrawi nation (Farah 2003, 22).

Similar to the OLF and their spatial imagination of Oromia, the struggle against Morocco has involved a rewriting and rescaling of history in order to make the different available narratives “fit” within the encompassing narrative of one Sahrawi people and their nation. In both cases this involves a narrative strategy aimed at creating differences between actors or collectives that were not necessarily there before. Kamal Fadel, the POLISARIO representative in Australia, exemplifies this approach: “The Sahrawis see themselves as a nation separate and distinct from neighbouring peoples in what is now North Western Africa. […] The arid terrain, mainly desert, molded their culture and shaped their distinctiveness. They speak a dialect of Arabic called Hassania, which is unlike the Tashelhit dialect spoken by the Berbers of Morocco. Other features such as social customs, diet and clothes emphasize their distinctiveness” (1999, 7). With all these distinctive features, one does wonder sometimes how these intellectuals are able to survive in western and urban surroundings. Such a polarized image does, however, create problems in finding any common ground between Morocco and the POLISARIO.

What this brief comparison between the spatial imagination of Oromia and that of Western Sahara shows is that although each conflict might have local roots, the interpretation of this conflict is immediately and necessarily embedded in global imaginations, which in turn can inform the actions of the major protagonists and either contribute to or hinder the search for a solution to this conflict. The development of spatial imaginations is not so much a natural response to oppression, but above all the result of grievances being articulated in a pre-formulated cognitive and epistemological framework. The task, therefore, should be to go beyond imaginaries based on an intimate connection between the people and the land and to find solutions to these (intellectual) conflicts that are able to translate grievances in non-territorial ways.

[1] See Haile-Gebriel Emdeshaw, “Germany to Step up Dev’t Cooperation with Ethiopia. Gerhard Schröder Concludes Two-Day Visit,” January 20 (2004) Mailing List Pol.Ethiopia. Available at: http://lists.sn.apc.org/mailman/listinfo/pol.ethiopia (February 4, 2004).

[2] Some examples would include Auf (1996), Lata (1998) and Pausewang, Tronvoll and Aalen (2002a).

[3] The term “Church and State” is derived from the book title of Tamrat (1972) and is also used by Triulzi (2002) in his article on Ethiopian historiography.

[4] For examples of this approach, see Ullendorff (1960), Marcus (2002), Rubenson (1976), Abir (1968), Levine (1974) and Crummey (2000a and 2000b).

[5] For important critiques of this tradition (and of many of the counter-histories that were a reaction against this tradition), see Clapham (2002b), Triulzi (2002) and Sorenson (1992, 1993).

[6] See, in particular, Knutsson (1969) and Blackhurst (1980).

[7] Spatial theories have a long history, but they are increasingly popular within a wide range of disciplines. One of the earlier anthropological examples is Thornton (1980). Important newer theoretical texts from an anthropological perspective are Gupta and Ferguson (1997 and 2002). The theoretical chapters of this thesis provide a more detailed analysis of geographical theories on space and place.

[8] Because of its time span not of direct relevance to this thesis, the article by Lagopoulos and Stylianoudi (2001) nevertheless discusses an interesting aspect of Amhara culture and space, namely the spatial organization of royal and military camps since the fourteenth century. According to the authors, this spatial model reinforced the material and symbolic power of the king and his representation as the single regulator of this model.

[9] The Horn of Africa in general is a relatively neglected area of research. Research on transnational spatial imaginations has, until recently, focused on Eritrea that adopted an extended view of citizenship, including those persons born to a father or mother of Eritrean origin, but living abroad. See Iyob (2000).

[10] A selection would include Lewis (1965), Baxter (1978b and 1979), Bassi (1996) and Blackhurst (1978).

[11] For an interesting recent example, see the dissertation by Frejacques (2003).

[12] This book in a way reflects the movement of the OLF towards the Diaspora. The first version appeared in Addis Ababa in 1980, the second version in 1988 in Khartoum, Sudan and the most recent version (from 1999) can be bought on http://www.amazon.com and has been printed by Kirk House Publishers, Minneapolis, MN (last date of access: 2 February 2004).

[13] See Biyyaa (1984) and Oromo-Frauen-Vereinigung in Europa (1985).

[14] See in particular Jalata (1993a, 1993b, 1998a, 1998b, 1998c, 1998d, 2003), Baissa (1998), Hassen (1990), Holcomb and Ibssa (1990) and Kumsa (1997, 1998). Although most writers publish in a wide array of journals and magazines, two journals are of particular importance: The Journal of Oromo Studies and The Oromo Commentary. The 1996 volume Being and Becoming Oromo presents an interesting mix of contributions from Oromo nationalists and those anthropologists that did research among Oromo societies in Ethiopia.

[15] A term employed by Harvey (1989) to describe the development of multiple ontologies that are the result of postmodernist capitalism. Harvey’s text will be discussed in the theoretical chapter on space and scale.

[16] Examples of this state-centered approach in interpretations of Oromo ethno-nationalism include Chanie (1998), Keller (1995, 1998) and to a certain extent Joireman (1997).

[17] For an exploration of the role of geographical features in the development of Ethiopia, see in particular Donham (1986).

[18] See, for example, the works of Janssen (1991, 1994) and Auf (1996) that use Norbert Elias’ theory on The Civilizing Process to understand the current difficulties the Ethiopian state is facing.

[19] The amount of research being undertaken under the heading of postcolonial studies is huge and extremely diverse. However, the following works deserve to be mentioned: from the perspective of travel literature, see Pratt (1992); the flourishing sub-field of subaltern studies is provocatively represented by Spivak (1988); from a somewhat more psycho-analytic literary perspective, see Bhabha (1994); the question of Africa and (post)colonialism is addressed by Werbner and Ranger (1996) and in a recent article by Abrahamsen (2003).

[20] See Bhaskar (1979). Bourdieu’s work – in particular his concept of habitus - is still important in the discipline of geography, but the original influence can probably be dated to the publication of Bourdieu (1977).

[21] Two important texts that appropriated structuration theory are Pred (1984) and Agnew (1987).

[22] The major problem behind this approach is Giddens’ repeated characterization of structure as a “system of generative rules and resources.” By doing so, he makes an analogy between the application of the rules of language in the act of speaking and social practice as an implementation of systemic rules. This approach, however – based on an illegitimate analogy between language and social life – tends to posit a structure outside of space and time. For a critique, see Thompson (1989). De Certeau (1984) also uses the language analogy, but in a much more productive fashion.

[23] See in particular Giddens (1985).

[24] Lefebvre was probably the first to theorize the connections between capital and space. See his magisterial Production de l’espace (1974), which has been translated as The Production of Space (1991). Other important contributions to this Marxist tradition include Smith (1984) and Harvey (1982). In this book, Harvey develops his concept of spatial fix, which is meant to describe the expansionary nature of capital and the geographical restructuring processes this involves all around the world: capital destroys the old, but to facilitate investment it needs, for a certain period of time, a stable environment. In other words, after de-territorialization, capital always needs re-territorialization.

[25] The possible criticism that Harvey’s Marxist approach can hardly be considered relevant to the Ethiopian situation can be answered in the following ways. First of all, most commentators on Ethiopia agree that one of the major defining features of twentieth-century Ethiopia has been its “insertion” into the world-economy. Second, my thesis is concerned with the spatial imagination of Oromia and, as will be argued in the second section, this imagination is largely an intellectual construct. Although it might be true that many actors inside Ethiopia cannot be considered postmodern, I do not doubt that the Oromo intellectuals discussed in this thesis are global postmodern citizens.

[26] The concept of positionality takes its lead from feminist theory and should be seen as a challenge to the ideology of universal and objective knowledge. It is often used to emphasize that knowledge is a relational construct and that these relations involve questions of power that should be made explicit. Here it is used to question Harvey’s top-down approach to space and provide the basis for a development of a concept of space that is intimately related to the positionality of specific actors. For a discussion of this concept within geographical theory, see Rose (1997).

[27] In the case of southern Ethiopia, this is discussed by Blackhurst (1980).

[28] See Anderson (1983), Hobsbawm (1992) and Gellner (1983) to name but a few.

[29] For an overview of this period, see Abir (1968) and Rubenson (1976)

[30] For a concise analysis of the ways these myths supported the legitimacy of the imperial Abyssinian crown, see Crummey (1988).

[31] Holcomb and Ibssa (1990) have radicalized this logic and argued that the Ethiopian state is the direct result of a collaboration between Ethiopian rulers and the agents of capitalist imperialism. Although an interesting theory, this is, of course, an extreme simplification, which is mainly guided by the attempt to interpret Oromia and the Oromo as powerless victims of an Amhara-dominated Ethiopia. In doing so, however, they offer an interpretation reminiscent of earlier interpretations of colonialism that deny any agency to its subordinated subjects.

[32] Although there is no clear historical break between the entity of Abyssinia and that of Ethiopia, in this thesis the first is used to roughly denote the period before Menelik II, whereas the second term refers to the period from Menelik until the present, which should imply that “Ethiopia” is that entity firmly embedded in international relations.

[33] Marcus (1995) 77-173 provides more detailed information on this period of Menelik’s rule.

[34] For an analysis of the different ways the terms Haile Selassie and Ethiopia have been used in the discourse of Rastafarianism, see Van Heur (2003). An extensive description of Afro-American reactions to the Italian occupation of Ethiopia has been given by Scott (1993)

[35] Actually, it was only at his coronation in 1930 that he took his baptismal name of Haile Selassie. Before this period he was known as Ras Tafari.

[36] Sbacchi (1985) provides an extensive description of this period of occupation.

[37] For a decent overview of the relations between Ethiopia, the United States and Great Britain, see Marcus (1983). Concerning the use of numbers and statistics in this thesis, please consider them as roughly indicative of certain trends and not as directly reflecting the reality on the ground. Statistics in Ethiopia are notoriously unreliable, influenced by political considerations and dependent on the whims of specific writers.

[38] For a general overview of the role land and the control over land played in the development of Ethiopia, see Crummey (2000)

[39] See, in particular, Crummey (1980), Ellis (1976), Cohen (1974) and Hoben (1973), in particular chapters 7-9. Although this debate had its main momentum during the 1970s, it is still unsettled. Tareke (1991, 57-58) acknowledges this, but nevertheless places Ethiopia firmly in the feudalist category. However, he does not offer any arguments for this and his analysis here seems to be based more on a Marxist attempt to conceptualize the peasantry as a separate and oppressed class than on empirical data.

[40] See, for example, Crummey (1980, 137), Donham (1986, 7) and Ellis (1976, 287). Perham (1969, 164) describes another important route to social advancement: military life. As she writes: “It meant the escape from the drudgery of farm-work and the dullness of village life; it held out hopes of loot and advancement.”

[41] This distinction is based on Donham (1986, 37-46). Clapham (2002) has criticized this center-periphery approach and has argued that other spatial structures – such as the networks of important Islamic families throughout southern Ethiopia - have also defined Ethiopian societies. I will come back to these non-state spatial structures in the following chapters. Here, however, I am mainly interested in the ways different areas have been incorporated into the Ethiopian state mechanisms and will thus follow Donham’s approach.

[42] For a more detailed description, see Markakis (1974, 104-140) that describes the incorporation of the southern areas into the Empire and gives a discussion of the use of balabbat in this process. Haji (1994) gives a discussion of balabbat in Arsi Oromo area.

[43] Besides Donham (1986) and the contributors to his edited book (in particular, Triulzi (1986), Turton (1986) and McClellan (1986)), an earlier example would be Clapham (1975).

[44] Wallerstein’s world-system theory (itself making use of a center-periphery conceptualization on a global scale) is also an example of such a monocausal approach.

[45] Triulzi’s article can be seen as a follow-up to Donham (1986), since Donham explicitly mentions the concept of frontier as a connection between center and periphery (p. 4), but in the remainder of his text the concept is never further explored. In his article, Triulzi refers to the classic text by Kopytoff (1987) on the African frontier and a number of texts that deal with the American frontier. It was, of course, Turner (1893) who has delivered the original formulation of the frontier as a contribution to identity. Unfortunately, Triulzi does not incorporate any new work on the American frontier, since this would have made possible a much more sophisticated conceptualization of the frontier that incorporates questions of power and space. For examples of this approach, see Cronon, Miles and Gitlin (1992) and Limerick (1987).

[46] One of the main boundary markings would become the distinction between “Galla” and “Oromo,” with Oromo nationalists employing the term Oromo in order to distinguish themselves from other actors that used the older and derogatory term Galla. See Hultin (1996). At the same time, as Zitelmann (1996b) argues, this Gallo/Oromo distinction was also used on lower levels within Oromo societies to distinguish between those who were considered an organic part of the group and those still partially considered strangers.

[47] Also see Triulzi (2002). In a lecture at the French Embassy in Addis Ababa, Clapham (2002) offered a somewhat similar argument and critique of Oromo nationalism.

[48] A question also, but somewhat more cautiously, asked by Sorenson (1998, 247-248): “In my view, many of these Oromo intellectuals are motivated by the best impulses of democracy and justice. Nevertheless, while we need not impugn the motives of those intellectuals in the diaspora who support Oromo nationalism, we do need to recall that the relationship between who speaks and the Oromo subject that is spoken of or for (i.e. ‘the people’) is neither completely transparent nor free of contradictions and difference and that the concerns, perceptions and goals of impoverished Oromo peasants in rural areas may not be exactly the same as those of university-educated intellectuals in North America.”

[49] It was the publication of Hassen (1990) that caused a controversy between Oromo nationalist scholars and defenders of the Greater Ethiopia tradition who accused Hassen of distorting Ethiopian history. See Marcus (1992) and the reply by Hassen (1992) and Baxter (1992). Also see: Ibssa (1993) and Birru (1993).

[50] I use the term (spatial imagination of) Oromia here as a shorthand. It must be said that the word Oromia does not occur in any texts before 1976. In 1976 it appeared in OLF (1976) and, according to Zitelmann (1996a), in the patriotic poetry of one of the leaders of the OLF. In my view, however, most of the elements that were later used to construct this spatial imagination – in particular the language afaan oromoo and the nationalist nostalgia – were already in place since the late 1960s.

[51] In a sense, my argument is the precise opposite of the one made by Lewis (1996, 39) in his article on Oromo political consciousness, who writes that “ethnicity begins at home – but can be extended to others who are seen to share at least some of the same characteristics and symbolic elements.” In contrast and with a certain irony, I argue that Oromo politicized ethnicity begins at a distance – far away from home – but that it can be extended to others who are seen to belong to this imagined home.

[52] The term longue durée derives from the historian Fernand Braudel and has been applied to Africa by Bayart (1993) in order to conceptualize African states as the product of their own societies and not as the result of European colonialism. For an appreciation and critique of Bayart’s approach that also refers to Ethiopia, see Clapham (1994).

[53] This 1971 pamphlet has been re-printed under the same title in The Horn of Africa (1980).

[54] A citation from Perham (1969, 164) I also used in my discussion of land and social mobility (see chapter IV.2)

[55] As I already mentioned, figures and facts are notoriously unreliable in Ethiopia. This is especially true in the case of statistics on ethnicities, since higher numbers imply a larger share of state resources. In the case of the Oromo, the argument is further complicated due to the politicized nature of every discussion. Therefore, the numbers tend to depend on the personal position of the commentator. Baxter, Hultin and Triulzi (1996, 9) mention that there are probably around twenty million Oromo, which would mean around 40 per cent of the total population. This is the number I have used in this thesis. It should hopefully become clear, however, that the question of ethnicity is – despite the rhetoric – quite flexible in the Horn of Africa: in one lifetime, one can be and become Oromo, Somali, Christian and Muslim. In that respect, Ethiopians are highly postmodern, switching identities and engaging in a play of subjectivities whenever deemed appropriate.

[56] For an important overview of the ways African states “manage” their territories and in particular for an argument that pays attention to the differences between the idea of the sovereign state and African ideas of overlapping sovereignties and extraterritoriality, see Herbst (2000). Mbembe (2001) offers a much more negative and even cynical view of the contemporary situation in Africa.

[57] This centrality of the state in allowing globalization to happen is emphasized by one strand within economic globalization theory. See, for example, Gilpin (1987) who emphasizes the important role of the state as an actor on the world market that allows its subjects to engage in transnational practices. At the same time, it is also the state that can end this permissive situation.

[58] Several authors comment on the importance of these African students in politicizing the Ethiopian students: Balsvik (1985, 74-91), Greenfield (1965, 366-67), Zewde (1991, 221) and Koehn and Hayes (1978). By using this example I do not want to imply that Ethiopian students were apolitical or non-reflexive before they met the African students. All I argue is that contact with these African students led to a re-questioning of earlier values, which means that a questioning of values was in itself nothing new.

[59] Markakis (1974, 154-55) mentions the following figures, applicable for 1968: United States (523), France (193), U.S.S.R. (131), United Kingdom (119), Germany (118), Italy (114) and the United Arab Republic (108).

[60] See Zewde (1991, 211-15), Balsvik (1985, 93-101), Zewde (1994) and Greenfield (1965, 337-374).

[61] A point also stressed by Schlee (2003a, 338).

[62] Bayart (2000, 227-228), whose concept of extraversion I will more extensively discuss in the next chapter, also acknowledges this dialectic between the local and the global: “Extraversion can also take a cultural form, since combatants adopt simultaneously the cosmologies and forms of symbolic representation of their local territories or home-areas […] and the imaginary figures of globalization.”

[63] Of course, the so-called Japanizers, a term coined by Addis Hiwet, in the early decades of the twentieth century appropriated the spatial imagination of Japan as a model for Ethiopian development. Besides their smaller number, however, the main difference with the later Ethiopian students was that the first group could be incorporated into the mechanisms of the Ethiopian state under Haile Selassie, whereas the second group could not and was excluded. See Zewde (2001)

[64] This section on the Bale Rebellion is based on Markakis (1987, 191-201), Ottaway (1978, 90-94), Tareke (1991, 125-159) and Gilkes (1975, 214-219).

[65] As Tareke (1991, 16-18) has pointed out, banditry was a common phenomenon that was, however, viewed in an ambivalent way by the peasantry. As he writes: “Banditry was the quintessential expression of a compartmentalized and highly individualistic society. […] Bandits operated within the social system and more often than not were easily coopted. Their arrogant disregard for the established order was a short-term phenomenon, for bandits held to the value of the dominant classes to which they normally aspired” (17).

[66] Besides the Bale rebellion, he analyzes the Tigrean uprising of 1943 and the 1969 revolt in the northwestern province of Gojjam.

[67] In an earlier article, Malkki (1992) gives a compact overview of the impact of being a refugee and the situation in the receiving area on the conceptualizations of nationality and territoriality.

[68] In this 1980 edition of the journal Horn of Africa one also finds an interview with an OLF member and an important article by Greenfield and Hassan that provides the basic building blocks with which the imaginations of Oromo nationalists were built: “Oromia Speaks” (1980) and Greenfield and Hassan (1980).

[69] The term “neo-Gobanist” derives from Ras Gobana, who was an important Oromo military commander in the army of Menelik and who conquered many parts of southern Ethiopia in the late nineteenth century. Among OLF nationalists, his name is invoked as a symbol for treason. Schlee (2003b), however, points out – being somewhat annoyed with the OLF rhetoric - that Gobana was not an individual traitor, but an Oromo lord and commander of an Oromo force. In addition, he provides a whole range of other examples of powerful Oromo that cooperated with Amhara or that oppressed other Oromo on their own.

[70] The OLF publication Oromia Speaks, for example, usually included a sketchy map that roughly delimited the boundaries of Oromia. This map was complemented by statistics that nicely point out the vagueness of the spatial imagination of Oromia: “Area – 600.000 sq. kms. (approximately),” “Population – 18.000.000 (Estimate).” See OLF (1979). Within a year, however, in Oromia Speaks the population increased by a million. See OLF (1980).

[71] Gada refers to a system of age-grades that controls and divides political power based on the succession of “generations.” Male individuals in different periods of their life belonged to specific generation-groups and each group was accorded a number of collective rights and duties. See Legesse (1973) and Baxter (1987b and 1979). Helland (1996) and Bassi (1996) offer a discussion of the political role of Gada. Among OLF-nationalists Gada is seen as a central symbol of Oromo culture and as a positive, egalitarian and democratic alternative to the hierarchical Abyssinian political system. See, for example, UOSNA (1981), Biyyaa (1984, 8-18) and Shunkuri (1995).

[72] As already mentioned in the introduction to this thesis, examples of this state-centered approach include Chanie (1998), Keller (1995, 1998) and to a certain extent Joireman (1997). Keller (1995, 632), for example, writes: “[…] contemporary Oromo national identity is the product of the Ethiopian colonial experience and the persistent fight for citizenship rights within the context of a multi-ethnic nation-state. The reason the notion of an independent Oromia seems to have become so salient for some Oromo is because they fear continued repression and exploitation at the hands of yet another Ethiopian régime.” Although I do not question this fear, such a view ignores the possibility that the imagination of Oromia might have been constructed outside of the Ethiopian state and is thus not only the result of repression.

[73] This section can do no more than present a highly compressed version of this early period of the revolution. A large amount of literature has appeared on this topic in the meantime and these paragraphs are largely based on the following texts: Ottaway (1978), Ottaway (1987), Markakis (1987, 237-44 and 258-64), Tiruneh (1993) and Clapham (1988, 38-57 and 196-218). Andargachew Tiruneh provides the most comprehensive account available so far, makes extensive use of both Ethiopian and non-Ethiopian sources, and also critically reflects on earlier accounts of the revolution.

[74] See “Ethiopia: Stresses and Strains,” in: Africa Confidential, vol. 16, no. 10 (1975) pp. 4-5.

[75] A useful term, since it captures some of the unintentional problems and possibilities that arise from day-to-day politics. A misuse, however, because it is does not really refer to Beck’s encompassing idea of reflexive modernization. See Beck, Bonss and Lau (2003).

[76] In his important study on the Mekane Yesus church during the revolutionary years, Eide (1996, 160-61) also points out some of the ideological differences between Baro and Gudina Tumsa. In a lecture called “The new political ideologies and the church,” Baro had tried to explain how a materialist and atheistic ideology could co-exist with the church. After this lecture, Gudina replied as following: “It must be understood that there can be no reconciliation and no compromise between what the church believes and materialism. Marxism-Leninism and the church can never be friends. Materialism thinks and lives from below, from matter, but the church lives from the Spirit of God, who comes from above.”

[77] See Zitelmann (1994, 86-87); “’Ethiopian Parties’ and ‘Militia Politics’,” in: Africa Confidential, vol. 18, no. 16 (1977) pp. 8; “Ethiopia: Dergue Comes out on Top,” in: Africa Confidential, vol. 18, no. 18 (1977) pp. 4-6; “Ethiopia II: Abo and Oromo Movements,” in: Africa Confidential, vol. 19, no. 11 (1978) pp. 5-6.

[78] See, for example, Cappelli (1997, 87-88).

[79] An important article that discusses this villagization in the Arsi (Oromo) Region is Cohen and Isaksson (1987). The monumental book of Scott (1998, 247-252) discusses, among many other things, the villagization programs in Tanzania and compares these to those in Ethiopia, making extensive use of Cohen and Isaksson’s article.

[80] The analysis of the work of the OLF in Sudan is largely based on Zitelmann (1994), Chapter 7.

[81] Originally, the OLF (and its partial predecessor the ENLF) started operations in the Chercher Mountains, near the city of Harer. Resistance in this area under the banner of Oromo nationalism continued during the entire period from 1974-1991 and was led by Abdulkarim Ibrahim Hamid (Sheek Jaraa). Although Sheek Jaraa and the activists in Addis Ababa (and a number of fighters in Bale that during the Somali invasion of 1977 became part of the SALF) were united under the banner of the OLF, in 1978 a clash between the chairman Magarsa Barii and Sheek Jaraa split the OLF into two wings. Both kept the name OLF until in 1986 Sheek Jaraa started using the name Islamic Front for the Liberation of Oromia (IFLO). The other wing moved its main activities to the border area between Sudan and Ethiopia and has my main interest here. See Zitelmann (1994, 83-84), Markakis (1987, 263), Clapham (1988, 219) and “Ethiopia: The Oromo factor,” in: Africa Confidential, vol. 25, no. 15 (1984) pp. 1-4.

[82] This is also my main critique of Bulcha (2002) that carries the slogan of forced migration in its title. Throughout the book this term is used to describe “the Oromo” as the victims of Abyssinian conquest and the behavior of the Ethiopian state. It must be acknowledged that Bulcha, in contrast to most other Oromo nationalists, accepts that exile does have its advantages: “it means exposure to and communication with the outside world” (30). He does not, however, acknowledge that this privilege of exile was originally mainly available to highly educated, young and male individuals and that this might have influenced the interpretation of “being Oromo.”

[83] This was in particular the result of the efforts of Hasselblatt, who worked at the Berliner Missionswerke. His own writings are interesting specimens of an embrace of the OLF-Oromia-imagination with a Christian touch. See Hasselblatt (1981a, 1981b, 1982, 1990 and 1992).

[84] As was done by, among others, Zitelmann (1980, 1988, 1992, 1994a), Braukämper (1982/83), Luling (1986), Clay and Holcomb (1985), Clay, Steingraber and Niggli (1988) and Bulcha (1988).

[85] Matsuoka and Sorenson (2001, 60-61) also provide an example of this selective approach by the receiving countries: “Immigration officials oversees weed out those they consider unsuited to life in Canada, and illiterate African peasants with no English-language abilities, no experience of waged employment, and little familiarity with urban life were clearly seen as a low priority.” Also see Karadawi (1983, 545) and Koehn (1991).

[86] For more detailed information on the production of literature and the development of afaan oromoo in exile during the late 1970s and 1980s, see Zitelmann (1988).

[87] Personal translation. I have translated the word “Heimat” as homeland, but it also evokes the imagery of country and home. The original German version is as following: “Der Begriff der ‚Heimat’, der zunächst nur mit der unmittelbaren Erfahrungswelt einzelner Flüchtlinge verbunden war, erhielt in den Diskussionen eine neue Bedeutung darüber, daß ausgetauscht wurde, was in den einzelnen ‚Heimaten’ produziert wurde und wie die geographischen Bedingungen aussahen.“

[88] A term that derives from Norbert Elias and which has been used by Janssen (1991 and 1994) and Auf (1996) to explain the homogenizing tendency of Ethiopian politics.

[89] A description of the proceedings of these London Peace Talks is given by Haile-Selaisse (1997, 312-330).

[90] One could argue that this is not the case for the EPLF that was involved in a struggle to separate Eritrea from Ethiopia and make it a sovereign state. However, in a similar way to other movements, the EPLF also wanted to get away from the centralizing tendencies of the Ethiopian state.

[91] In a 1993 referendum this was made official.

[92] The term “nationalities” still reveals the Stalinist conflation of ethnicities and nations so popular among the intellectuals of the late 1960s and 1970s.

[93] I will complicate this materialist interpretation of events in the next part of this chapter.

[94] See “Ethiopia: An Elusive Victory,” in: Africa Confidential, vol. 31, no. 14 (1990) pp. 3-4 and “Ethiopia: From Rebels to Rulers” in: Africa Confidential, vol. 32, no. 11 (1991) pp. 1-3.

[95] Also see “Ethiopia: New Government, New Map,” in: Africa Confidential, vol. 32, no. 22 (1991) pp. 7.

[96] Nor is the shape of the region really final, since adjustments have been made in certain border areas to accommodate specific and local conflicts (themselves highly influenced if not the result of the realization of this region in practice).

[97] For more detailed information on the language issue and the development of afaan oromoo, also see Bulcha (1993a), Bulcha (1997) and Hassen (1994). More data on the reasons behind the choice of the Latin script for the language and the polemics surrounding this choice in the Ethiopian context is provided by Gamta (1993) and Zitelmann (1994b).

[98] See Brietzke (1995), Cohen (1995, 1997), Abbink (1997), Young (1996, 1998) and Paul (2000).

[99] See, in particular, Chole (1994), Harvey (1996) and Dercon (2000).

[100] For a critique of this mode of conducting politics, see the recent article by Moussa Iye (2003).

[101] Also see „Introduction: Meles Zenawi on the Razor’s Edge,“ in: The Indian Ocean Newsletter, Ethiopia: The Top 100 People No. 1 (2001).

[102] The exclusion of global i.e. transnational processes and the adoption of a closed state-centered view causes some analysts to believe that the state is something one simply cannot do without. Zartman (1995) for example, in the closing chapter to the book Collapsed States, writes: “In the search for answers, it is first necessary to reaffirm that reconstruction of the sovereign state is necessary. [...] State functions cannot be left to even a well-functioning society, any more than society can abdicate its activities to the state” (267). A similar view concerning the Ethiopian state is adopted by Keller (1995) in his contribution to the book edited by Zartman. Clapham, on the other hand, has recently written about “thinking the unthinkable” and described some of the elements from which a new African political order could be constructed. See Clapham (2001).

[103] Personal translation. The original reads as following: “Eines unserer Probleme in Äthiopien ist, daß sich zwei Kulturen gegenüberstehen: auf der einen Seite gibt es die demokratische Kultur der Oromo und anderer Völker in Süd-Äthiopien und auf der anderen Seite die autokratische Kultur der Herrschenden von Nord-Äthiopien.”

[104] See “New OLF Leader,” in: The Indian Ocean Newsletter, no. 884 (1999) and “Daoud Ibsa Gudina,” in: The Indian Ocean Newsletter, no. 885 (1999).

[105] The notion of habitus takes center stage in the work of Bourdieu. For a concise description of this and other important terms, see Bourdieu (1992). Carter Bentley (1987) has used this notion of habitus to contribute to a practice theory of ethnicity in his case study on the Philippines.

[106] To be precise, in 1989 an Oromo Studies Committee was formed under the auspices of the Union of Oromo in North America (UONA), a “mass organization” of the OLF. This committee later became the OSA. Since 1993 the OSA publishes The Journal of Oromo Studies. For more information on this earlier period and some of the struggles that accompanied it, see Tuso (1993).

[107] See “Ex-OLF Members in a Row,” in: The Indian Ocean Newsletter, no. 873 (1999).

[108] See “My Enemy’s Enemy,” in: Africa Confidential, vol. 40, no. 4 (1999) pp. 5.

[109] See “Common Front for Oromos,” in: The Indian Ocean Newsletter, no. 920 (2000).

[110] See, for example, Jalata (2003), in which he embraces the UFLO. In doing so, he ignores his depiction in earlier articles of the OLF as the only true representative of the Oromo and his rejection of other parties such as the IFLO that are now part of the UFLO-coalition.

[111] Thus, Henze (2002), in his history of Ethiopia, begins with the geological foundations of the Abyssinian Empire and ends with the contemporary period. Such a teleological and unitary view is repeated over and over again in newspaper articles, Afro-American discourse and travel literature.

[112] This is not to say that Baxter and Greenfield simply repeated OLF-propaganda, but their allegiances clearly lay with “the Oromo” and this brought them very close to the position of the OLF. Baxter, in his articles on Oromo nationalism (1978a, 1983, 1994 and 1998) has always embraced this construction of an Oromo national identity. Over the years, however, he has also grown more critical of the possible negative aspects of this and has started emphasizing the need to attend to cultural diversity within this Oromo national identity. In the work of Hassen, friend and colleague of Baxter, one can see a similar development. See, for example, Hassen (1999).

[113] This is what the anthropologist Clifford means when he writes about “the informant”: “[…] I’m always running up against a problematic figure, the ‘informant.’ A great many of these interlocutors, complex individuals routinely made to speak for ‘cultural’ knowledge, turn out to have their own “ethnographic” proclivities and interesting histories of travel. Insider-outsiders, good translators and explicators, they’ve been around.” See Clifford (1997, 19).

[114] On the OSG website http://www.oromo.org, the numbers of killings and disappearances in Ethiopia is counted. Gow, who published the ethnographic study on the Oromo in Melbourne, Australia, also established an OSG in Australia, but quit after indifference from the Australian media and negativity from the Oromo community partly due to the fact that Gow’s main collaborator was a university educated Oromo woman. See Gow (2002, 25-26).

[115] As usual, it is not so much historical facts, but the embedment of these facts in a certain narrative that is important. See, for example Trueman (1995) and Trueman (1998), where he writes: “The OLF has made repeated efforts to negotiate with the TPLF-led regime in Addis Ababa but, as talks in October recently showed, the TPLF is intransigent” (4).

[116] I am referring here to Edmond Keller, who in two articles (1995b, 1998) specifically devoted to the question of Oromo ethno-nationalism and the Ethiopian state, largely constructs his argument on the basis of texts by Oromo-writers in the Diaspora, such as Jalata, Hassen and Holcomb and Ibssa. Although he combines these with other articles and books on the Ethiopian state not written by these Diaspora intellectuals, he in no way clarifies the status of each text and instead simply constructs a homogenous and linear narrative.

[117] This is even more astounding, since this critical literature on the historical enterprise is hardly part of a recent fashion. Many important works were already published in the 1980s (and earlier). See, for example, Ankersmit (1983) and White (1987).

[118] Clifford’s discussion of Diaspora culture also refers to Gilroy and makes explicit the position this implies towards the state: “Diaspora discourse articulates, or bends together, both roots and routes to construct what Gilroy (1987) describes as alternate public spheres, forms of community consciousness and solidarity that maintain identifications outside the national time/space in order to live inside, with a difference” (1997, 251).

[119] Personal translation. The original reads as following: “Wer vom Internet Aufklärung erwartet, kann am Beispiel Äthiopien und Eritrea erkennen, dass dies – zumindest kurzfristig – nicht immer Aufklärung hin zu Freiheit und Demokratie bedeutet, sondern primär Aufklärung ’gegen’ den Feind.

[120] One need only take a look at the websites of the different liberation movements, such as the OLF (www.oromoliberationfront.org) and the OLC (www.gumii.org), and the “original” website of the Oromo diasporic community (www.oromia.org).

[121] Based on a search with VisualRoute, which is software that traces the connections your computer makes in order to reach the server of a specific website.

[122] See OSA (http://www.oromostudies.org) and ONA (http://www.oromiannationalacademy.org).

[123] Oromo Muslim Community Seattle: http://freewebhosting.hostdepartment.com/g/gandaoromo; Manamarii: http://www.geocities.com/manamarii; Bilisummaa Oromiyaa: http://www.bilisummaa.com.

[124] See, for example Deressa (1990) and Kumsa (1997 and 1998) and the interview with her in a Canadian Newspaper: Paul Irish, “A Survivor’s Long Road to Freedom: Endured Torture for her Journalism,” in: Toronto Star, March 20, 2003.

[125] This importance of community organizations is also emphasized by Bulcha (2002). See for example the websites of the Oromo Youth Association (http://www.oromiayouth.org), Oromo Community Organization, Washington DC (http://www.oromoco.org), All Oromos (http://www.allaromos.com) and Oromia TV (http://producers.mtn.org/oromiatv).

[126] In Ethiopia, this could also apply to the OALF and the UOPLF whose roots go back to the 1960s. Wako Gutu, for example, is over 70 years old.

[127] Recently, a site of the Oromo Community of Minnesota went online that tries to cater to the general Oromo audience. See: http://www.oromocommunitymn.com.

[128] The question how much this interpretation is itself part of a current academic fashion for fragmentation and diversity is surely interesting, but will not be addressed here.

[129] Another interpretation I cannot pursue here, but that would surely be interesting is to interpret the developments in Ethiopia as an example of the functional differentiation of world society (as has been theorized by Luhmann). Such a view argues that what is occurring is a development of global functional systems with each system producing a range of global semantics and images.

[130] The pamphlet was an invitation to an informational meeting in March 2004 in Berlin, Germany, organized by the Society of Friends of the Sahrawi People (Gesellschaft der Freunde des Sahrauischen Volkes) and dedicated to the support of their self-determination struggle. The event was moderated by Axel Goldau – a journalist for the magazine Critical Ecology (Kritische Ökologie) – and involved two talks: one by the Sahrawi sociologist Salama Emhemed-Zeyou and the other by Daniela Hinz, a political scientist.

[131] See, for example, the mailing list Sahara Update, which is moderated by the Norwegian Support Committee for Western Sahara (http://groups.yahoo.com/group/Sahara-update). Other examples include the Australian – Western Sahara Association (http://www.awsa-westernsahara.org), Sahara Occidental en Direct (http://www.sahara-occidental.org) and Sahara Info (http://www.sahara-info.org) in France, Bureau International pour le Respect des Droits de l’Homme au Sahara Occidental (http://www.birdhso.org) in Switzerland and Friends of the Western Sahara (http://www.friendsofthewesternsahara.org), Western Sahara Online (http://www.wsahara.net) and Homeland International (http://www.homelandonline.org) in the United States. Most NGOs dedicated to the support of the Sahrawi are located in Spain, such as the Asociación Saharagalicia (http://www.saharagalicia.org), Associació Catalana d’Amics del Poble Sahraui (http://www.sahara-acaps.org), Um Draiga (http://www.umdraiga.com) and the Asociación de Familiares de Presos y Desaparecidos Saharuis (http://www.afapredesa.org).

I. Introduction