Unraveling Somalia: Race, Class, and the Legacy of Slavery

Part I

Introduction

Chapter 1

Somalia from the Margins: An Alternative Approach

A letter arrives, telling me that every child under the age of five was now dead in the Jubba valley village in southern Somalia where I had lived several years previously. The collapse of the Somali state in 1991 ended these young lives in starvation and warfare, opening yet another violent chapter in the short history of the Jubba valley. In just the past 150 years, the people of this valley—most of whom were considered racial minorities within the Somali nation-state—had endured a series of violent encounters that shaped their relationship to the state and to regional Somali society. Such encounters—including enslavement, forced labor on colonial plantations, periodic pastoralist raids, kidnapping by the state military, and forceful land dispossession in the biggest political landgrab in Somali history—presaged their vulnerability in the violence of civil war. When the Somali state collapsed, the people of the Jubba valley disproportionately faced genocidal assault, banditry, and widespread rape. Although the valley’s population has been massively reduced through starvation, murder, and flight since 1991, the valley remains one of the most contested areas in the militia wars that continue to plague southern Somalia.

How does a place become so violent? In 1991, journalists, pundits, politicians, and academics groped for metaphors that could simply and concisely explain the warfare. The most persistent and pervasive explanation knit together popular perceptions of tribalist Africa with models derived from anthropological descriptions of northern Somali social groups to claim that ancestral clan hatreds played out in warfare both caused Somalia’s collapse and hindered future state-building efforts. Since I. M. Lewis’s (1961) classic book on northern Somali pastoralist social organization first appeared, Somali society has usually been described in academic and popular literature as an egalitarian and ethnically homogenous population of nomadic pastoralists who shared an overarching genealogical system and a common language, culture, and religion. Lewis described Somali society as consisting of six patrilineal clan-families formed by the descendants of mythical Arabic ancestors who arrived in Somalia twenty-five to thirty generations ago. His pioneering work on Somalia explained that each clan-family encompassed a set of patrilineally related clans, subclans, sub-subclans, and lineages. Historically, political activity usually occurred at the level of lineages or groups of lineages tied together by social contracts who collectively paid diya, or blood compensation, for wrongs committed by any group member. While clan-families rarely acted as a unit, diya-paying groups and lineages could join forces at higher levels against other groups of lineages for warfare and payment of diya.

Drawing upon this model (called a segmentary lineage structure by specialists), the American media and some academic accounts of Somalia’s collapse presented Somalia’s destruction as having been almost inevitable. The model of the tensions inherent to this kind of genealogically based system provided an explanation of built-in conflict, making Somali social structure appear fundamentally divisive and resistant to state-building efforts. Journalists’ reports portrayed Somalis as continuing to act out Stone Age ancestral clan rivalries, but with Star Wars military technology. Media reports were filled with evocations of ancestral violence: The clans of Somalia have regularly battled one another into a state of anarchy (Time 1992); ancient clan enmity pursued with the modern weapons that are so abundant in Somalia is at the root of the country’s conflict (Washington Post 1993); Instead of fighting with traditional spears and shields, the clans have more recently conducted their feuds with mortars and machine guns (New York Times 1992); the crisis in Somalia has been caused by intense clan rivalries, a problem common in Africa, but here carried out with such violence, there is nothing left of civil society, only anarchy and the rule of the gun (CNN 1992).¹ In refashioning academic terminology to present a portrait of Somalia’s collapse, Somalis became cartoon-like images of primordial man: unable to break out of their destructive spiral of ancient clan rivalries, loyalties, and bloodshed.

This book challenges the prevailing explanation that the violent breakdown of Somali civil and political society was the result of deep clan-based hatreds and ethnic loyalties. Most specifically, it argues that this explanation is inadequate for understanding the most violent areas of contemporary Somalia, such as the Jubba River valley. Taking the Jubba valley as my focus, I suggest that the dissolution of the Somali nation-state can be understood only by analyzing the turbulent history of race, class, and regional dynamics over the past century and a half—processes which produced a deeply stratified and fragmented society. Over the past several generations a social order emerged in Somalia that was rooted in principles other than just a simple segmentary lineage organization—a social order stratified on the basis of racialized status, regional identities, and control of valuable resources and markets. To understand the 1990s breakdown of the Somali nation-state—and the particular victimization of agricultural southerners during and after dissolution—one must acknowledge the basis and significance of these alternative forms of stratification.

By focusing on one particular community on the margins of Somali society and the Somali nation-state—the people of the Gosha area of the middle Jubba River valley—this book probes the tensions produced by these alternative forms of stratification. I will be looking beyond the fantasy of an ethnically and linguistically homogenous culture to illuminate some of the other critical fracturing points of Somali society. Such fracturing points—of race, class, and region, as well as status, occupation, and language—become clearer when we turn our focus away from the urban elite politics of the center that have dominated both media coverage and scholarly analyses. Scrutinizing the history of the margins can magnify the primary processes shaping the center; looking at state disintegration from the margins of the Jubba valley reveals the dynamism of Somali society (which is left uncaptured in the segmentary lineage model) in general, and illuminates the tense dynamics of racialized identities and stratification, in particular, which were central to state formation in Somalia and which patterned much of the violence of disintegration.

While aiming to illuminate some critical aspects of the state, this book is not an ethnography of the state, of state-building, or of state-collapse; rather it is an ethnography of the local, situated at the margins of the state. Nevertheless, one of my goals is to use the view from the margins to apprehend the institutions, discourses, and forces of the state as well as the tensions and frictions which contributed to the state’s collapse. As Michael J. Watts has argued, localities are always political and struggled over, a fact which allows us to see locality as a fundamental part of national identity and hence a repository of various rights and memberships that are regularly spoilt and fought over (1992:121).² Looking at the expression of the state on the margins and the way people on the margins experience and imagine the state can, I believe, illuminate its structures of domination, its technologies of power, the effectiveness of its discourses, the nature of its hegemonic representations, and the extent of its popular support. As James Scott has noted, referring to analyses of state collapses: And whenever . . . the ship of state runs aground on such reefs, attention is usually directed to the shipwreck itself and not to the vast aggregation of petty acts that made it possible (1985:36). Focusing on the local—where we can perhaps best see the vast aggregation of petty acts—is still, I believe, what anthropologists do best, and what we do better than any other discipline: we can use these strengths of local-level ethnography to analyze and theorize the state.

Map 1. The Jubba valley of southern Somalia.

Some of the major processes that have shaped Somali society and defined its power dynamics are inscribed in the marginal space of the Jubba valley and in the historical experiences of the people who live there. Over the past 150 years, the Jubba valley (Map 1) has been imaginatively constructed and ecologically utilized in a variety of ways by the people who live there, by the Somali national community, by the state, and by foreign authorities and donors. The valley has alternatively been viewed as a refuge, as a frontier, as a dangerous place of magic and sorcery, as a place of fantastic agricultural potential, as a desolate backwater, as a national ecological resource, and, most recently, as one of the premier shatter zones and valuable prizes in Somalia’s civil war. The Jubba valley in the mid-1990s has been the site of massacres, of famine, of years of militia wars, and of a disproportionate outpouring of the surviving local farmers into refugee camps; many of the faces shown on television screens depicting Somalia’s tragedy belong to people from the valley. The valley has now emerged as one of the most significant areas of contention, ending a century (or beginning a new century) of land occupation, expropriation, and development plans carried out by colonial governments, by the Somali state, and by international donor agencies.

The people of the Gosha have participated in, been subjected to, supported, and struggled with the significant cultural, political, and economic changes sweeping through the Horn of Africa in the past century. Their ancestors, taken from the area stretching between contemporary Kenya to Mozambique, arrived in Somalia in the nineteenth century as slaves to work on Somali-owned plantations producing cash crops for export. After escaping as fugitives or being manumitted as Muslim converts, these former slaves settled the uninhabited forests of the Jubba River valley. Twentieth-century Gosha history reflects key elements of the national experience: Gosha communities had their land taken over by colonial plantations and their bodies claimed for colonial forced-labor schemes, they lost complete rights to their land through socialist-inspired nationalization laws after independence, experienced the speculatory effects of a 1980s capitalist-inspired commodification of land, and suffered from phenomenal Cold War-linked state militarization. Although discriminated against as a racial minority on the basis of their Bantu heritage since their arrival in Somalia, Gosha people have throughout this tumultuous history continuously struggled to define their identity and a legitimate place within Somali society, using Islam, kinship, and economic relations as their tools.

The historical experience of the people living in the Gosha brings into relief the critical significance of racial and growing class distinctions in Somali society during the twentieth century. Because their status as racial minorities took precedence over their lineage and clan affiliations, the Gosha peoples’ history offers a window into the ways that Somali society was stratified by constructions other than clan. Their increasing economic marginalization reveals the basis of emerging class stratification in southern Somali society during this century. By tracing historically the creation of the Gosha as a large, subordinate (and disparate) group in southern Somalia, this study highlights the problematic and contingent nature of ethnic consciousness, identity politics, and class formation. Trying to understand the brutal victimization of Gosha farmers since 1991 highlights the tragic cultural logic informing race, class, and regional dynamics in contemporary Somalia.³

Despite its often chaotic appearance, violence is never simply anarchic, but rather is informed by a particular logic and a sense of purpose. For analysts of specific cultures and places, suggests Kay Warren, the imperative is to dispel the view that violence is inherently chaotic and irrational by tracing the implications of particular forms of domination, resistance, and violence (1993a:3). Also invoking culture and history—and thereby challenging anthropologists—Fernando Coronil and Julie Skurski (1991) argue that political violence can be understood only by attending to the specific historical context of social relations, cultural forms, and memory in violent places.⁴ While we may lack the ability to explain the ruthlessness of violence, people participating in violence often do have at least a momentary sense of what they’re doing. What appears senseless is often informed by a particular sensibility that emerges in particular historically conditioned moments and is shaped by particular cultural practices and ideologies. While the act of violence may remain ultimately unapprehendable, we can seek to understand the historical and cultural tensions that surround that act and provide it with meaning. Although this book is not an ethnography of Somalia’s disintegration, taking a good look at some of the heretofore relatively masked aspects of cultural and social stratification in Somalia will, I hope, take us part of the way toward a better understanding of the historically produced social relations that have patterned much of the postcollapse violence in southern Somalia.

This book also examines the thorny problem of how to understand the widely experienced simultaneity of domination and resistance, subjugation and collusion. Analyzing Gosha peoples’ historical experiences illuminates the position of subjugated peoples living within structures of domination which define the hegemonic terrain of morality, social identity, and appropriation. This book attempts to interrogate history in order to understand how an ideology of hierarchy which assigned a denigrated, subordinate status for Gosha peoples was constructed and symbolically represented, morally legitimized, and materially enacted in Somali culture and society. Apprehending the structures of domination also means recognizing the contours of collusion: how people in the Gosha simultaneously resisted the ideological basis of denigration—which subjugated them as slaves, Bantus, and low status farmers—while accepting as legitimate the hegemonic Somali social order based on kinship.

As I will argue below, despite their denigrated position within Somali society most Gosha people did not view themselves as any kind of recognizable ethnic group set on a course of resistance against the society that denigrated them. To the contrary, rather than fostering a group sense of cultural distinctiveness and subjugation, most Gosha villagers oriented their cultural practices and identity politics toward dominant Somali patterns. A central question of this book is how to understand Gosha peoples’ choices to participate within and actively seek incorporation into a society which subjugated them, and how to understand their creative ability to manage the juxtaposition of domination and accommodation.

Gosha peoples’ historical experiences firmly situated them at the intersection of racial hierarchies, class stratification, the hegemonic Somali kinship system, and state politics—a tension-filled space within which they negotiated a personal politics of identity, culture, and social relations. Understanding this space, which will occupy us in Part III, reveals the mutually constitutive dimensions of culture, identity, and politics. Seeing in Gosha experiences the processes of nation-building and state dissolution in Somalia, we can probe the critical links between cultural struggle, identity politics, class, and the state that exists throughout Africa.

For decades, anthropologists have grappled with the concept of identity, as we have argued over how to define it, how to identify it, where to locate it (assuming it is a single entity, which we cannot). Responses to these questions have variably addressed the political and cultural dimensions of identity, usually reformulated as ethnicity. The 1970s and 1980s marked a turn in our disciplinary approach to studying ethnicity/identity, when anthropologists began rejecting earlier assumptions of cultural/ethnic primordialism/essentialism to emphasize instead the fluid and situational aspects of ethnic identity, where people made choices about their ethnic identities in different sociopolitical contexts.⁵ While transformative for the understanding of identity formation, this situational or constructionist approach often lacked sufficient attention to the role of political struggle and state-formation in setting constraints, delimiting options, or defining the possible in the shaping of identity.⁶

The rise of a political economy focus in anthropology in the late 1970s and early 1980s, together with a growing postcolonial acknowledgment of colonialism’s impact on our anthropological subjects, spawned a view of contemporary ethnicity/tribalism as the result of the political economy of colonial rule⁷—a realization first suggested by an earlier generation of anthropologists.⁸ Writing from a similar perspective influenced by political economy, others suggested contemporary forms of tribalism or ethnicity resulted from violent warfare in the context of European expansion.⁹ Applying the lens of political economy has allowed some scholars to explain local forms of ethnicity as a creation of global capitalism played out in local contexts through labor relations,¹⁰ or to emphasize how the semantics of primordial sentiments are situationally used to generate ethnic cohesion in the context of warfare, violent competition, and global capitalism.¹¹ Most recently, scholars have highlighted the role of the postcolonial state in directly fostering ethnic identities and ethno-nationalisms through authoritarian rule, unequal distributions of resources, and polarized competition for political power to explain rising ethnic tensions, violence, and aggressive new nationalisms in eastern Europe,¹² Africa,¹³ and elsewhere. Again transformative in our understanding of ethnicity, tribalism, and identity because of their attention to the state and global capitalism, such approaches have been critiqued for their lack of attention to the role of (local) culture in identity formation.¹⁴

Another dimension of identity, racial constructions, has recently reemerged from hiding after a long period of prominence in anthropology. After ignoring the concept of race for decades because of our inability to define it, our discomfort with it, and perhaps because of its politicized close-to-home associations,¹⁵ anthropologists are once again confronting the significance of racial constructions in identity politics and the complex dialectic of race and class.¹⁶ The politicized and persistent recognition of racialized identities offers the strongest critique of the situational view of ethnicity (one cannot always situationally shift one’s racial identity) and the most forceful evidence of the political dimensions of identity.

For me, this academic heritage has clarified that cultural identities and practices of the subaltern result from how people construct themselves as cultural beings within fields of power shaped and directed mostly by others—whether colonial, ethnic, class, or political others. The cultural content of these identities—how people define themselves and their history—is shaped in particular political moments, under particular historical conditions, as a result of particular historical experiences, and in the memory of particular historical nightmares (to borrow Karl Marx’s [(1852) 1973:146] much borrowed image).¹⁷ In order to apprehend how cultural identities take shape within historical and political moments, we need to study, in the words of William Roseberry and Jay O’Brien, "the context of inequality and contestation; the process of struggle and the place of naturalized and oppositional historical images within it; and the production of historical images themselves" (1991:12, italics in original). We need to focus on the intersection of identity, culture, and politics in the living of daily life and to recognize how historical experiences get reworked within this intersection in meaningful ways which shape cultural practices and beliefs, political struggles, and the ways groups of people represent themselves and are represented by others.

Identities take shape from cultural content within political fields of unequal power; culture affects and is affected by political struggles over signifying, representing, and providing meaning for identities. Groups of people incessantly shape and reshape their cultures and identities in dialogue with the state, the nation, and nation-state hegemonies.¹⁸ Historical memories, historical consciousnesses, and historical knowledges saturate these mutually constituting processes. People in the Gosha lived every day in the midst of memories of particular historical nightmares; one goal of this book is to analyze the changing significance of these nightmares in shaping the confluence of identity, politics, and culture in their lives.

Before moving on to the central arguments of the book, in the remainder of this chapter I provide an overview of the past century of Somali political history.¹⁹

In the imperial era of the late nineteenth century, Britain, Italy, France, and Ethiopia subdivided the territory inhabited by Somalis—people self-identified as members of the Somali genealogy traced to common ancestors—among their governments. In the division of Somali territory, Britain claimed two areas of Somali-inhabited territory; the northern coast and inland area along the Gulf of Aden inhabited by pastoralist Somalis became British Somaliland and the area west of the Jubba River, also inhabited by pastoralist Somalis as well as by farmers on the west bank of the river, became part of the Northern Frontier District of British Kenya. Agricultural, agropastoral, and pastoralist southern Somalis from the east bank of the Jubba River to the boundary with British Somaliland found themselves living in Italian Somalia. In 1925 Britain ceded to Italy much of Somali-inhabited territory west of the Jubba, drawing a boundary between Italian Somalia and British Kenya’s Somali-inhabited Northern Frontier District. A British military administration took over southern Somalia following Italy’s defeat in World War II until 1950, when southern Somalia was returned to Italy (against much local opposition) as a United Nations trusteeship. Independence in 1960 united into one nation the southern Italian colony and northern British Somaliland, although the maintenance of colonial boundaries with French Djibouti, Kenya, and Ethiopia left much Somali-inhabited territory within these neighboring states. The colonial legacy of illegitimate boundaries took on enormous significance in a pastoral economy where family members were separated from each other and from critical grazing areas. Furthermore, different colonial histories, languages, and political systems contributed to continuing regional distinctions in the postcolonial era.

The decade following independence was marked by increasingly fragmented politics, as Somalis splintered into dozens of clan-based political parties that drew support from a patronage system well maintained by massive injections of foreign aid. In 1969, army officer Siyad Barre led a successful coup that overthrew the postcolonial parliamentary democracy, and established rule by the Barre-led Supreme Revolutionary Council under the banner of scientific socialism, which Barre defined as a commitment to equality, economic independence, and economic growth.²⁰ The political intricacies of Barre’s rule have been explored elsewhere;²¹ a brief overview must suffice here. With substantial military aid from the Soviet Union through the mid-1970s, Barre embarked on a set of ambitious programs to foster nationalist sentiment and consolidate state power. One of Barre’s earliest acts was to abolish tribalism through outlawing the recognition of clan distinctions, ethnic distinctions, patron-client relationships, and any other formal relationships based on social inequality. Siyad explained this drastic decree as a step toward nationalist unity that would nurture economic, political, and social progress through erasing social inequalities, clan conflicts, and loyalties built along clan or ethnic lines. Many Somalis initially welcomed this political ideology, probably in part because of their desire to move beyond the kind of clan politics that had characterized the post-independence parliamentary democracy of the 1960s.²²

To ensure that the populace followed the law and gave their loyalty only to the state, Barre developed several branches of the government police and judiciary structure to oversee, adjudicate, and punish disloyalty or treason under the National Security Laws of 1970. The National Security Service (NSS) investigated all suspected security offenses, maintained its own detention centers, and became the main agency supplying defendants for trial in the National Security Courts (I. M. Lewis 1988:213). Lewis describes how the NSS and the National Security Courts jointly dealt with a wide range of ‘political’ offences including nepotism and tribalism . . . as well as with such charges as ‘lack of revolutionary zeal’ and treason. Members of the National Security Service, under a Sandhurst and K.G.B.-trained commander, enjoyed arbitrary powers of arrest, sometimes following the denunciation of a suspect by his personal enemies. Members of the public services were kept under surveillance and N.S.S. reports played an important part in promotion and demotion (1988:212).

The NSS and National Security Courts reported directly to the president, as did the police, the prisons, the military police (who also served as the presidential guard), the military intelligence service, and the militia-styled Victory Pioneers (civilians who were to maintain revolutionary fervor and who could arrest suspects for the NSS; Rakiya Omaar [1992:231] likened them to Haiti’s Tontons Macoute).²³ Amnesty International summarized the enormous powers thus held by the president: The President is in direct control of all facets of state power, including the security institutions and all branches of the judiciary, and the Constitution is so drafted that no provision obstructs his exercise of all decision-making should he wish it (1990:9). In addition to these policies and institutions designed to oversee and maintain ideological control, Barre introduced a torrent of laws and policies to abolish local forms of authority and leadership—in control over land, water, health, marriage, inheritance, mediation, commerce, production, speech, political participation, and the media—and replace them with state control. Public executions of prominent secular and religious people who spoke out against the regime and its new policies sent unequivocal messages of the fate awaiting those who did not support the government, its ideological orientation, and its laws.

As part of his scientific socialist package, Barre inaugurated a variety of social programs to support his nationalist vision of an egalitarian country united behind his leadership. His 1973–75 literacy campaign established the first written script for Somali and sent students and civil servants throughout the countryside to teach Somali citizens to read. He targeted women’s participation in nation-building by establishing a national organization for women that provided social programs and revolutionary teaching to women, by introducing new laws governing marriage, divorce, and inheritance which gave women greater individual rights backed by the state rather than their kin, and by outlawing the universal practice of female infibulation. He established orientation centers throughout the country where people could learn about the revolution, where marriages were to take place, and where rituals supporting his revolutionary aims could be staged. He mandated the participation of the citizenry in public displays of revolutionary fervor, including marches, dances, songs, and poems which proclaimed him the Father of the nation, associated him with great Somali leaders of the past who Barre claimed had fought against external enemies on behalf of the Somali nation, identified him as part of the new holy trinity of Marx, Lenin, and Barre (I. M. Lewis 1988:211), and praised his leadership by elaborating a vigorous personality cult (Omaar 1992:231) around his image as head of nation and state.

One of his greatest nationalist aspirations was to unite Somalis throughout the Horn into one expanded nation-state. Following the fall of Ethiopia’s Emperor Haile Selassie, the Somali-inhabited Ogaadeen region claimed by Ethiopia became the focus of his most intensive irredentist effort. Somalia invaded the Ogaadeen in 1977 to reclaim Somali territory from Ethiopian control, but was forced to retreat when the Soviets (and Cubans) backed the new Marxist Ethiopian government with massive military support against the Somali incursion. Following Somalia’s defeat and retreat in early 1978, Barre all but severed links with Moscow and distributed arms among the civilian population of the north for self-protection against Ethiopian incursion across the disputed border.²⁴

Many observers see the loss of the Ogaadeen war as the beginning of Barre’s declining popularity, loss of ideological vision, increasing manipulation of clan politics, and rising tactics of state repression. Nationalist spirit, at its height during the early months of the Ogaadeen war,²⁵ began to disintegrate in the face of economic crisis, political insecurity, state repression, and lack of ideological leadership. The Somali government had lost the military and economic support of its superpower patron and had drained state coffers for the war effort. Refugees poured into northern Somalia, making further demands on state resources exhausted by the war and creating tensions among northerners over access to land and water. Domestic politics dissolved into blame and recriminations as to who lost the Ogaadeen (Cassanelli 1993a:13). An unsuccessful coup against Barre led by military officers of the Majeerteen clan resulted in a purge of the military, the formation in exile of a Majeerteen clan-based anti-Barre guerrilla group, the tightening of domestic security, a reshuffling of government officials, and a consolidation of control over government institutions by Barre and his closest relatives.

Over the next decade, two significant trends came to dominate the political economy of Somalia: (1) a massive influx of foreign aid, and (2) a virtually complete disintegration of popular trust in the government as a result of flagrant human rights abuses, state-backed terror, unpredictable political appointments and demotions, and the increasing concentration of state power by a small circle of Barre’s relatives and supporters.

In the aftermath of the breakdown in relations between Somalia and the Soviet Union, Barre sought support from the other global superpower, the United States. Cold War geopolitics provoked U.S. interest in Somalia, which was seen as strategic because of its proximity to the Middle East and the Persian Gulf.²⁶ The late 1970s saw the fall of the Shah of Iran, the humiliating takeover of the U.S. embassy in Iran by revolutionaries, and the U.S. perception of a growing Soviet influence in the Gulf area.²⁷ In order to maintain military bases in Somalia that could monitor affairs in the Gulf, the United States government provided $163.5 million in military technology and four times that much in economic aid during 1980–88.²⁸ By the late 1980s, Somalia was receiving 20 percent of U.S. aid to Africa.²⁹ The growing U.S. interest in Somalia accompanied large influxes of foreign funds from other donors to alleviate the refugee crisis and to welcome Somalia into the anticommunist fold. Throughout the 1980s donor funding to Somalia soared into the billions with contributions from Italy, Britain, Germany, and Saudi Arabia (as well as from China, who had provided economic support to Somalia since Barre’s rise to power). The value of arms alone imported by Somalia during the two decades of Barre’s rule totaled nearly two billion dollars.³⁰ The United Nations High Commission for Refugees (UNHCR) spent hundreds of millions of dollars to maintain refugee camps for those displaced by the Ogaadeen war.

In the context of huge foreign aid support, growing popular dissension against his government, domestic tensions in the wake of the war, and the influx of millions of refugees, Barre seemed to abandon his vision of socialist development, equality, and nationalist unity by increasingly utilizing terror and repression to retain personal power and personal control over state resources, including the distribution of foreign aid funds. By the early 1980s, the Somali state was one of the most militarized in Africa, having grown from 3,000 troops at independence to a suffocating 120,000 by 1982 (Adam 1995:71). In Hussein Adam’s words, The army of liberation had been converted to a huge army of repression (1995:71). In tandem with state militarization, Barre began to concentrate power in the hands of a small circle of supporters drawn largely from the clans of his father (Marehan), his mother (Ogaadeen), and his son-in-law, the head of the NSS (Dulbahante) (all part of the Darood clan-family). As head of state, he controlled most of the lucrative resources of his country (such as foreign aid, refugee relief, and land, as well as political appointments), and his redistribution of these resources to political supporters came to be popularly perceived as overtly clan-based rather than ideologically socialist. The prominence of Marehan, Ogaadeen, and Dulbahante clan members in his government was reflected in the clandestine code name ‘M.O.D.’ given to the regime. . . . Although no-one could utter the secret symbol of General Siyad [Barre]’s power openly, the M.O.D. basis of his rule was public knowledge and discussed and criticized in private (I. M. Lewis 1988:221–22). Ever the political strategist, however, Barre ensured that any perceived M.O.D. favoritism was offset by other political appointments and demotions which kept people uncertain about the requisites for success in the Somali government. As Lee Cassanelli explains,

The President did not want to turn the struggle into one of his own Mareehan (with its Dulbahante and Ogaadeen allies) against a series of united clan opponents. From the beginning, and continuing even into the mid-1980s, the President sought to distribute favors to erstwhile representatives of every clan. Acknowledged (even by his avowed enemies) as a master manipulator of the country’s byzantine internal politics, Siyad consistently managed to isolate influential individuals while wooing less prominent rivals from the same clan. He was able to coopt many with funds and contracts his regime was accumulating from the substantial foreign relief assistance and military operations in Somalia. Each individual who joined the President’s fold was expected to bring the support of his lineage or sub-clan with him. (1993a:16)

Such political practices produced a pattern of uncertain promotion and demotion as members of a rotating elite [became resigned to a] trajectory [which] extended from prison through ambassadorships to ministerial office (I. M. Lewis 1988:250).

The perception of a highly unequal distribution of state resources based on clan favoritism, combined with the overall sense of personal insecurity generated by the pervasive security system and the unpredictable pattern of political appointments and imprisonments, began to have effects most clearly in the north, where state-backed Ogaadeen refugees were beginning to challenge Isaaq dominance.³¹ By the late 1980s, Barre policies in the north following the loss of the Ogaadeen war had begun breeding revolt among the Isaaq, the largest northern clan-family, who felt excluded from politics and state resources.³² Barre’s efforts to wrest away control over livestock exports and qat (a popular narcotic) sales from Isaaq merchants and entrepreneurs,³³ and his resettlement of Ogaadeeni refugees in Isaaq-dominated territory, were seen in general as part of a continuing subversion of northern interests by a southern-dominated government,³⁴ and in particular as an attempt to undermine the economic and political strength of the Isaaq clanfamily.³⁵ This growing sense of Isaaqi alienation (I. M. Lewis 1994:179) prompted Isaaqs and others in exile to form a resistance group, the Somali National Movement (SNM), against Barre’s government.

Following his inability to fully control the lucrative northern commerce in qat and livestock and aware of growing Isaaq-led challenges to his authority, Barre embarked on a variety of harsher measures against Isaaqs in the north that contributed to further tensions between refugee Ogaadeenis and resident Isaaqs. The Somali government armed Ogaadeeni refugees, conscripted refugees into the military to be used as soldiers against Isaaqs, and encouraged Ogaadeeni refugees to claim the property of displaced Isaaqs.³⁶ The SNM’s low-grade guerrilla war against the government culminated in several attacks on UNHCR refugee camps for Ogaadeenis and on government installations in the north in 1988. Barre responded in a series of savage counterinsurgency tactics (Human Rights Watch 1989:22), ordering the bombing and strafing of northern towns, villages, rural encampments, and even people fleeing on foot.³⁷ In its retaliatory attacks, the government killed tens of thousands of its own people; almost half a million northern Somalis fled from Barre’s repression into Ethiopia, and over half a million were displaced within the north.³⁸

By the final years of the 1980s when the Somali state began to wage open war against its own citizens, its international patrons could no