Victims and Vulnerable Groups in Southern Somalia

 

PREFACE

Due to the nature of and the difficulty in obtaining written documentation on the current situation in Somalia, Professor Lee Cassanelli of the University of Pennsylvania was commissioned to research and write this paper to address the information needs of those involved in the Canadian refugee determination process. Much of the information is contextual in nature and is designed to provide a background understanding of issues affecting the current situation of clans and other groups in southern Somalia. As indicated in the introduction, the information is based on the author's knowledge and experience, on publicly available documentation and on interviews conducted by the author with other scholars and Somalis who have left Somalia. Early drafts of the paper were shared by the author with other scholars for their input.

The paper is the work of the author; the Documentation, Information and Research Branch (DIRB) of the Immigration and Refugee Board serves as publisher. It is not, and does not purport to be, either exhaustive with regard to conditions in the country surveyed or conclusive as to the merits of any particular claim to refugee status or asylum, nor does it reflect the position of the DIRB or the Immigration and Refugee Board.

For additional information and updates on Somalia, please consult the Refinfo and Refquest databases and other sources available at the Immigration and Refugee Board's Regional Documentation Centres.

ABOUT THE AUTHOR

Lee Cassanelli has a Ph.D. in African History and has been teaching at the University of Pennsylvania since 1974, where he is currently an Associate Professor of History and Member of the African Studies Executive Committee.

Professor Cassanelli has been to Somalia several times since 1971, most recently as a Fulbright Scholar for six months in 1987. His book on the early history of Somalia, The Shaping of Somali Society, was published by the University of Pennsylvania Press in 1982, and he has written numerous articles on nineteenth and twentieth century Somali social, cultural and economic history. He was co-founder of the Somali Studies International Association, begun in 1978.

During the past three years Professor Cassanelli has conducted interviews with Somali refugees in Kenya, Italy, the UK, the US and Canada; studied the role of international emigré communities in ethnic conflicts at the Woodrow Wilson International Centre for Scholars in Washington, DC; and participated in conferences on the role of international peacekeeping in the Horn of Africa.

GLOSSARY

SAMO   Somali African Muki Organization

SDM       Somali Democratic Movement

SNA       Somali National Alliance

SNF        Somali National Front

SPM       Somali Patriotic Movement

SSDF      Somali Salvation Democratic Front

SSNM    Southern Somali National Movement

UNITAF                Unified Task Force

UNOSOM              United Nations Operations in Somalia

USC        United Somali Congress

1. INTRODUCTION: SCOPE, SOURCES AND METHODOLOGY

The purpose of this paper is threefold: first, to outline the most important causes and patterns of population displacement during the recent Somali civil war; second, to identify the groups that have appeared most vulnerable during the ongoing conflict; and third, to evaluate the real and perceived risks that certain communities continue to face. The geographical focus is on southern Somalia, roughly the area between the middle Shabeelle River valley and the Kenya border, including the coastal region from Mogadishu to Kismayu. This part of Somalia experienced the most sustained inter-clan fighting following the collapse of the Siyad Barre regime in early 1991, and it is the area where most of the 28,000 United States and United Nations peacekeepers were deployed in December 1992 (see map 1). It is also the area where a number of communities are still vulnerable, as armed militias continue to compete for access to water, land and strategic port facilities.

The information included in this paper comes from a variety of written and oral sources. Among the written sources are news reports, reports by UN agencies, NGOs and human rights groups, and accounts and letters from observers and participants. It should be noted that there is no written documentation for large areas of the country where fighting and displacement occurred. During the civil war, foreign observers and journalists were largely restricted to the capital city of Mogadishu and major relief centres like Baidoa and Kismayu, with the result that our knowledge of events in many remote areas is based only on hearsay. The present situation for obtaining information, especially on rural areas, is not much better. Only recently have efforts been made to systematically collect and photocopy documents issued by the various relief organizations that have been working on the ground in Somalia. [ The US Library of Congress Office in Nairobi has begun collecting reports and documents from organizations that were operating in Somalia at the time of Siyad Barre's fall in early 1991. An index of the 534 documents currently being microfilmed can be found in Somalia Reports, Post-Barre Period, Part I, Library of Congress Office, Nairobi, Kenya, 1994. ]

I have supplemented the available written evidence with oral testimonies, some of which I collected from Somali refugees during a two-week visit to Kenya in November 1993, and some from Somalis in Italy and the UK in April 1994. These oral testimonies must be evaluated carefully, as many of these individuals suffered extreme emotional and physical stress while in Somalia. Many of the written reports on which this paper is based are themselves based largely on interviews with refugees or displaced persons, so these need to be treated with the same care. Because of the risks of distortion or exaggeration, I have tried to substantiate all oral accounts with written sources or independent oral ones.

The crisis situation in Somalia has tended to generate many rumours and mutual accusations by members of different Somali clans. Charges of systematic intimidation, physical abuse and even genocide are regularly levelled by members of one clan against another. [ Claims of genocide and "clan cleansing" are discussed in (Mukhtar and Kusow 1993, pp. 15-22, and Prendergast Jan. 1994, pp. 7-8).] Moreover, many Somali communities have been struggling not only for physical survival but also for international recognition. They want to be regarded as "legitimate" residents of certain locales with rights to political representation on district councils, or as "vulnerable groups" entitled to priority attention from international relief agencies. Some individuals have even changed their clan affiliation to protect themselves or gain temporary advantage (for example, see Prendergast Jan. 1994, 7). In such circumstances it is often difficult to find impartial sources of information, or to confirm or disprove rumours. While it is not my intention to determine who is right or wrong, but only to decide if certain groups are indeed vulnerable in the circumstances, I have had to make judgements throughout this paper about the reliability of certain testimonies or claims. Where there seems to be more than one point of view, I try to indicate the range of possibilities.

Finally, because the situation on the ground in Somalia can change from week to week, and because military and political alliances among the contending parties are frequently only temporary, I have avoided detailed discussion of political factions and personalities, preferring to focus on patterns of relationships between communities and clans that have been affected by the war. Wherever possible I have attempted to check my information with others who have been following the situation closely, and have conferred regularly with the authors of several recent reports on contemporary conditions in Somalia. These sources are acknowledged in my footnotes and references.

2. PHASES OF THE SOMALI CONFLICT: TARGETS AND VICTIMS

Since 1988 the Somali conflict has gone through several phases, each producing distinct patterns of population displacement and distinct clusters of victims. By "victims" I mean not only those individuals and groups who directly experienced violence and human rights violations, but also those whose material survival was threatened by the overall climate of anarchy and insecurity. My conversations with Somali refugees in Africa, Europe and North America suggest that people fled Somalia for a variety of reasons. Some claim to have experienced or witnessed beatings, rapes and killings, while others claim that their personal property and means of livelihood had been forcibly seized or destroyed. Some refugees were convinced that their membership in one of the dominant clans put them at risk of abuse from military rivals, others feared that their "minority" status or lack of strong clan ties made them easy targets for the armed militias that continue to control much of the country, and still others believed they had been forced out of their settlements by outside clans who coveted their land. The basis of these fears among refugees of both major and minor clans is examined in greater detail in section 3.

Much of the current sense of collective insecurity in Somalia dates from the final years of the Siyad Barre regime, 1988-1991, when the government sought to combat a series of clan-based opposition movements by sending its troops against the civilian populations of the clans involved. [ For background information on the various Somali opposition movements and their overseas links, see (Horn of Africa Jan.-June 1990).] In addition to the well-documented war against the Isaaq populations in the north in 1988 (Cultural Survival Quarterly Winter 1989; Africa Watch 1990a), government forces launched punitive actions against Hawiyya civilians in Galkayo in November 1989 and Beled Weyn in May 1990, against Ogaden clansmen around Kismayu in May 1990, and against protesting civilians in Mogadishu in July 1989 and July 1990 (Africa Watch 1990b; Samatar 1991). In addition to killing hundreds of civilians, these attacks destroyed livestock and productive assets such as wells, storage facilities, shops and pumps upon which the communities depended for survival. This pattern of collective punishment against the civilian settlements of clan rivals continued into the post-Barre phase of the civil war. Because Somali clans have historically regarded security or retribution as a collective responsibility, individuals can feel threatened by virtue of their membership in the group, even if they have not personally engaged in overt violence against members of other clans.

* * * * * *

The period from the ouster of Siyad Barre in January 1991 to the launching of Operation Restore Hope in December 1992 witnessed the forcible and systematic displacement by Hawiyya clan militias of tens of thousands of Darods (Marehan, Dolbohante and Majerteen) from the capital city and the lower Shabeelle valley. This series of Hawiyya offensives and Darod counter-offensives involved numerous incidents of torture, mutilation, rape and execution without burial on both sides (AI Aug. 1992). Hawiyya control (shared by various clans) of Mogadishu and virtually the entire Shabeelle valley was the outcome. Darod civilians and soldiers fled south to Kismayu, Bardera and Kenya, or to their clan homelands in the north and northeast.

From their territorial bases in the Shabeelle valley and central rangelands, Hawiyya militias also advanced into the interriver regions (Bay and Bakool). The sedentary and semi-sedentary inhabitants of these regions had not played a major role in either the Barre government or the military rebellions against it. Nonetheless, in 1991-92 they found themselves in the middle of a struggle between three heavily armed factional coalitions: the Hawiyya militias of the United Somali Congress (USC), retreating government (predominantly Marehan) forces that sought to regroup in the Geedo region under the banner of the Somali National Front (SNF), and Ogaden-dominated militias based in the lower Juba and Afmadow areas that claimed to back the Somali Patriotic Movement (SPM). The Bay, Bakool and lower Shabeelle regions became major battle grounds where contending militias and their offshoots swept back and forth, confiscating livestock and food supplies, and looting water pumps, tools and even household furniture to sell for hard currency. The resulting anarchy disrupted planting and harvesting and produced widespread famine in the region, which became known as the "triangle of death." The food crisis was compounded when the militias in control of Mogadishu's port and airfield prevented imported food aid from being delivered to relief centres upcountry, both to keep it out of SNF hands and to resell it for a profit (Menkhaus 1991; ibid. July 1991; AI Aug. 1992; Mukhtar and Kusow 1993, 13-18).

Thousands of Bay and Bakool region residents died, and thousands more fled to relief camps in Baidoa and Hoddur. Although it involved tens of thousands of people, the displacement of these predominantly Rahanweyn (Reewin) clans was largely internal to the interriver region, since the area was landlocked and relatively few Rahanweyn had established substantial support networks beyond Somalia. The Rahanweyn Digil clans (see appendix I) displaced from the lower Shabeelle valley sought refuge in the bush or in Mogadishu, whose population of impoverished ex-farmers swelled despite the Hawiyya factional violence that continued in the capital. From the lower Juba, thousands of Bantu farmers fled across the border to Kenya, where they joined Darod victims of the earlier purges (Prendergast June 1994; Menkhaus 1993, No. 1).

Finally, the struggle for control of Kismayu, the major port for the entire lower Juba region, pitted militias of the SNF (Marehan/Majerteen), SPM (Ogaden), SSDF (Somali Salvation Democratic Front, representing the Harti Majerteen in the Kismayu region) and SNA (Somali National Alliance, the largely Hawiyya coalition that succeeded the USC) against each other. The prize was not just Kismayu but the routes used by livestock traders on the west bank and commercial farmers along the lower Juba to move their products to port. Kismayu became a crucial objective for Darod clansmen of all factions (SNF, SPM and SSDF) because of their expulsion and total exclusion from Mogadishu and environs (Prendergast June 1994). The multi-front struggle for Kismayu continues unabated, marked by a seemingly endless cycle of byzantine negotiations between factional leaders and prominent local elders, frequently broken by militia attacks and counterattacks. [ Detailing the shifting fortunes of the various contenders for this area over the past three years, as well as the numerous alliances and counter-alliances that have been made and broken, is beyond the scope of this paper. Ken Menkhaus has acquired a wealth of information on the lower Juba valley from his long years of research and experience as a political advisor to UNOSOM from August 1993 to April 1994. A summary of the major clans and political factions active in the lower Juba valley and Kismayu can be found in (Prendergast June 1994, pp. 10-11).] All sides have suffered heavy casualties, both among soldiers and civilians. The most consistent victims, however, have been the small farmers of the area-the Bantu minorities known as "Gosha" or "Mushunguli"-whose grain stores and personal possessions were stripped by the armed factions that criss-crossed the valley (Menkhaus July 1991; Prendergast June 1994). With few guns and no organized militias of their own, the Bantu farmers were virtually defenceless, and the large number of Bantu refugees in the Kenya camps attests to their continuing sense of insecurity (Lehman 1993).

* * * * * *

The arrival and deployment of Unified Task Force (UNITAF) and United Nations Operations in Somalia (UNOSOM) forces in December 1992 helped reduce the incidence of overt clan warfare and the outflow of refugees from southern Somalia. While media attention focused on the continuing Abgal-Habr Gedr conflict in and around Mogadishu (which resulted in a stalemate and a division of the city into two armed camps), a number of important patterns began to emerge outside the capital.

First, the major factional militias worked to consolidate control over territories where their clansmen were numerically dominant. By the start of 1994, members of the larger clans could generally find security in these "home" territories within Somalia. UNHCR data from Kenya suggests that repatriation, both spontaneous and UN-sponsored, had accelerated during the second half of 1993, especially among refugees from the Majerteen, Marehan and Ogaden clans (UNHCR Sept. 1993; ibid. June 1993; ibid. Oct. 1993a; ibid. Oct. 1993b). The US Department of State reported that as of November 1994, over 57,000 refugees had returned to various regions from camps outside Somalia's borders (Country Reports 1994 1995).

Second, within each of these de facto "home" territories, weaker clans were still subject to threats from more powerful ones. In Geedo, the Marehan, whose numerical strength in the region had been deliberately built up by Siyad Barre's land resettlement programs of the 1970s and 1980s, seized farm land from the minority Gobaweyn, forcing them into dependent tenancy or flight to Kenyan refugee camps [ Some of the Gobaweyn who had sided with the USC in its initial attempts to oust the Marehan from the Juba valley were later abandoned by the USC to face retribution from the Marehan militias. This was a fate suffered by many minorities whose territories were subject to waves of invasion by successive militias.] (Cultural Survival Quarterly Spring 1994). The influx of Ogaden refugees from Ethiopia (chiefly Mohamed Zubeir supporters of Omar Jess' SPM) into the hinterlands of Kismayu after 1991 unbalanced local clan relations and threatened the Harti (Harti elders were victims of an SPM militia massacre in January 1993) (Menkhaus 1993, No. 1). Hawadle arms enabled them to dominate the riverine farmers in the district of Beled Weyn, until they themselves were ousted in July 1994 by Habr Gedr supported by local clans who resented Hawadle domination of UN-financed operations in Beled Weyn (Menkhaus Sept. 1994, 3; ibid. n.d.). The incorporation or subordination of weaker clans by stronger ones is an old pattern in Somali social relations, but the consolidation of clan power in certain areas, while bringing a greater sense of security to the majority, invariably disadvantages other local groups and minorities, who may be forced into political compliance or payment of tribute for protection.

Third, while new large-scale militia advances were deterred by the presence of international peacekeepers, the competition for rich agricultural land in the Juba and Shabeelle valleys, and for access to the port towns of Merka and Brava, intensified. This third trend has the greatest potential to produce continued long-term population displacement in southern Somalia. These strategically important resource zones were primarily inhabited by minority clans, who generally did not have the military means to defend their property rights from well-armed occupying militias. While much of the fertile and potentially irrigable farm land in the south had been appropriated by well-connected politicians during the Siyad Barre years (Cassanelli 1995; African Rights Oct. 1993), the current factions' claims to these lands are based not on questionable land titles but on force. The newcomers claim to have liberated these lands from illegal occupation by the previous regime, but they appear to have no intention of returning them to the pre-Barre owners. For example, the Habr Gedr and Hawadle militias that aggressively ousted the Darod from the Shabeelle valley in 1991-92, often with the enthusiastic support of local minorities, are now laying claim to those same districts. In places like Afgoy, Jenaale, Awdeegle, Qoriooley, Merka, Brava and Jilib, where Hawiyya settlers had previously been absent or at most had a minor presence, the SNA militias now play a dominant role. In most of these places the occupiers have tried to co-opt local leaders into accepting their presence, failing which they have employed intimidation or force to displace uncooperative residents with their own clansmen and allies. The inhabitants of all these contested zones could with some justification claim they were still at risk. Operation Restore Hope had helped suppress the level of overt warfare and stabilize the informal boundaries between rival factions, but it did not end more subtle forms of violence-such as expropriation, intimidation and extortion-practised within each region.

The "war" is now concentrated in key resource areas of the south, which are also largely, although not exclusively, inhabited by minorities. [ As one Somali put it to me in April 1994, General Aidid's SNA forces had been stalemated in Mogadishu by the Abgal, in Bardera by the Marehan, in Baidoa by the Rahanweyn alliance and in Kismayu by the Darod. Dominating the lower Shabeelle, with its militarily weak minorities, is the only thing the SNA has left.] While planting and harvesting has resumed in many districts of the south, the larger economy is one based on extortion of surpluses from the unarmed by the armed. Even if the dominant groups do not aim to forcibly displace these minorities, they do effectively dictate to the local residents, who must often turn over their surplus harvests to the young militiamen, serve as magnets to attract NGO aid, which can then be skimmed off by the powerful, and offer public support and legitimation to the leaders of the occupying forces when the latter claim representation on emerging local or district councils. Because no social contract based on clan affiliation exists between the occupying forces and the villagers, there is no assurance that benefits in the form of relief aid or services will reach the villagers themselves. For these reasons we can predict that some residents of these "occupied" districts-typically younger men with a little money-will continue to seek refuge outside the country.

* * * * * *

It is too early to say what affect the withdrawal of international peacekeepers at the end of February 1995 will have on the overall situation in southern Somalia. The UNOSOM forces on the ground had not been visibly involved in mediating local disputes for some time. While their departure precipitated a predictable series of skirmishes over the materiel left behind, the Somali factional militias did not launch any major offensives in the wake of their withdrawal. More serious were allegations that some UNOSOM military forces had provided SNA militias with fuel and supplies as a kind of guarantee for the peacekeepers' safety. Coupled with the perception that UNOSOM has tended to promote General Mohamed Farah Aidid's plans to build a coalition government in the aftermath of UNOSOM's departure, the withdrawal could be seen as leaving Aidid's SNA in a stronger position than his rivals. There seems little doubt that the SNA continues to be an "occupying" army in much of the lower Shabeelle (Menkhaus Sept. 1994, 7-9).

Other recent trends that could influence patterns of conflict and population displacement in the coming months are listed below. At this point we can only speculate on their possible significance.

1. Erosion of factional power. In a September 1994 report, Ken Menkhaus summarizes evidence that all of the factions face eroding bases of support and declining power (Menkhaus Sept. 1994, 12-16). This reinforces news reports through much of 1994 that the SNA alliance was experiencing major rifts (for example, see Somali News Update and Africa Confidential). Habr Gedr have been fighting the Hawadle clan near Beled Weyn, their former Ogaden allies near Jilib, and the Murosade militias in south Mogadishu. Several rifts have been reported within the Habr Gedr and its various subclans. Apart from the overall instability that such internal fighting will produce, civilians belonging to these subclans may be vulnerable to retaliation in the recurrent pattern that we have seen. The Murosade, Hawadle and others with good land in the Shabeelle valley may experience disruption of farming activities, although wholesale population displacements are unlikely to occur since all of the Hawiyya clans are well-armed. The rifts among the Majerteen, Marehan and Ogaden factions in the lower Juba are also likely to produce intermittent fighting, although major shifts in the territorial balance of power do not seem imminent. The riverine minorities will continue to be affected by these factional wars, and this may curtail their willingness to repatriate from refugee camps in Kenya.

2. Competition for access to export markets. In late 1994, Habr Gedr militias seized control of the port of Merka after heavy fighting with the local Biimaal clan. This followed a rift between the Dir leader of the Southern Somali National Movement (SSNM)-whose major following seems to have come from Dir fighters near Jamame on the Juba rather than from the Merka Biimaal-and his former ally, Aidid's SNA. In retrospect this struggle over Merka, a major outlet for the bananas grown in Jenaale, was only a prelude to the "banana wars" that began early in 1995 between rival Habr Gedr militias working for different international export firms (Indian Ocean Newsletter 18 Feb. 1995; Reuters European Business Report 17 Feb. 1995). These incidents appear to confirm the motives behind the consistent targeting of riverine areas by armed militias during the civil war: the aim was to occupy these productive zones in order to control the anticipated revenues from a revived export sector.

3. Incidents of summary justice by Islamic courts. While Islamic courts had passed death sentences in northern Somalia in 1992 and 1993, not until 1994 were there reports of Islamic justice being visited on thieves and adulterers in the south. Sentences involving amputations and floggings were reported from Merka and Mogadishu (Country Reports 1994 1995; AI 16 Dec. 1994). Despite periodic calls by some religious groups for the establishment of Islamic law in Somalia, it appears that Islamic courts spring up only in areas where clan courts are not functioning effectively. For example, I have heard of no summary trials based on Islamic law in the regions of Bay and Bakool, despite the religiosity of many in these areas. Where Islamic courts have emerged they seem to be filling a perceived vacuum in the administration of criminal justice, but as yet such incidents do not represent the beginning of a major trend in southern Somalia.

4. Reduction in the number of politically motivated killings. Clan retributions for abuses supposedly carried out under the Siyad Barre regime appear to be diminishing, but civilians continue to be killed in factional retaliation of more recent vintage. Extrajudicial executions, torture, home searches, evictions and looting are all cited in Country Reports 1994, but few specifics are provided. It seems that much of the continuing violence against civilians is economically rather than politically motivated, or the result of continued random banditry and extortion.

3. CATEGORIES OF VULNERABLE GROUPS

Having outlined some of the general trends that are discernible in southern Somalia over the past few years, we now turn to the situation of specific categories and communities of people who might be at risk in the current circumstances.

3.1 Major Clans

As mentioned above, by 1994 most of the major clans on both the Hawiyya and Darod sides had secured control over large blocks of territory where their kinsmen predominated. While infighting among subclan and sectional leaders continued within each territory, and while the major clan militias continued to compete for control of border towns and minority districts, the secured territories appeared sufficiently stable to offer the possibility of repatriation for clan members currently in refugee camps or elsewhere. Now that the Kenyan government has made clear its intention to close down or consolidate many of its refugee camps, many Somali refugees will likely return to their "home" territories in Somalia rather than accept relocation to other camps where their kinsmen may be in a minority. Alternatively, some of these refugees may intensify their efforts to obtain asylum in other countries. Most of the major clans successfully established extended family networks overseas in the 1970s or 1980s, and these emigré communities played important roles in both financing the factions during the civil war and rescuing many of their kinsmen from that war (Cassanelli Nov. 1994).

While it seems that most Darod, Isaaq and Hawiyya clansmen could today repatriate safely to their home regions of Somalia, there are several reasons why they might not be inclined to do so. Cost and security of travel are considerations for some, particularly for indigent individuals and single women and children who do not have close relatives to accompany them through border regions, where bandits and armed militiamen abound.

Some potential repatriates may feel there are political risks in returning to Somalia. Association with the Barre regime appears to be less stigmatizing now than it was in the months immediately after the regime collapsed. The most prominent agents of repression under the old regime have either been killed or long since fled the country, and atrocities committed by contending militias in more recent years are far fresher causes for retribution than the deeds of a now deceased dictator. At far greater risk are individuals who, during the early phases of the UNITAF and UNOSOM missions, spoke out forcefully against the "warlords," or who urged the UN to take a grassroots approach by supporting local elders, community leaders and NGOs rather than the militia leaders. In circumstances where "warlords" and their henchmen still hold sway, outspoken critics, especially those from the same clan, are regarded as a threat to the clan leaders' authority.

All of the major clans have members who, prior to the war, had lived and worked as teachers, businessmen or civil servants in urban centres like Mogadishu or Kismayu or Baidoa and had acquired a largely "detribalized" outlook. Having in all likelihood experienced "victimization by clan association" or having lost their homes or business assets to the militias, they have little desire to return to the intense tribal politics of their homeland. If they feel they cannot find a "non-tribal" role in Somalia, [ Even Somalis who work for NGOs are regarded by other Somalis first and foremost as representatives of their clans rather than as neutral employees of relief agencies. Most relief agencies become identified with the Somali clan in whose territory their headquarters is located, with the result that these agencies are expected to hire "their" clan members as local staff (Menkhaus n.d.; Politique Africaine 1992, 83-86).] they will likely remain abroad, perhaps attempting to join relatives who have already established residency overseas. Finally, there are always individuals-even among the clans currently holding power-with no immediate surviving family or whose families have broken with them, who would be extremely vulnerable should they return to Somalia.

3.2 Minority Clans

 As discussed in section 2, much of the violence that occurred after the arrival of international peacekeepers took place in areas occupied by the country's minorities. As currently used by both Somalis and foreigners, the term "minorities" refers to any clans or communities that do not belong genealogically to one of the four major "noble" clan families of Darod, Hawiyya, Isaaq or Dir. "Noble" in this sense refers to the widespread Somali belief that members of the major clans are descended from a common Somali ancestor, and that the minority clans have a different-usually mixed-parentage, with some Asian, Oromo or Bantu ancestors.

If one includes the agro-pastoral Rahanweyn clans of the interriver zone (see appendix I), the so-called minorities probably make up one-third of the Somali population. Most of them live in relatively small, distinct communities throughout southern Somalia, and often speak local dialects and marry within their own or other minority clans. Until the civil war they seldom interacted and had no sense of political solidarity, but recently some have begun to perceive themselves as sharing "second-class" status, and have begun to talk about themselves collectively as minorities who have been particularly victimized during the civil war (Cassanelli May 1994).

Historically, some of the smaller minority clans have specialized in occupations like fishing (the Bajuni and Reer Manyo), religious learning (Reer Sheekh Sufi) or metal and leatherworking (Tomal and Midgan). Their expertise in these areas sometimes earned them a certain respect, even while their small numbers made them politically powerless. On the other hand minorities like the Bantu farmers of the lower Juba and Shabeelle valleys or the Rahanweyn clans of the interriver region number in the tens or even hundreds of thousands, yet enjoyed little political influence in post-independence Somalia. As do most of the minorities, the Bantu and Rahanweyn believe the major clans have systematically discriminated against them by excluding them from important government positions, restricting their educational opportunities and neglecting social and economic services in their home districts.

As a consequence, the minorities have tended toward political and economic self-reliance and have sought to avoid central government involvement in local affairs. They speak a variety of local and regional dialects, a fact that further isolated them from the centres of power following the 1972 decision to make the north/central Somali dialect the official written national language. There is little question that members of the major Somali clans have regarded the southern minorities with condescension and even disdain, and have felt it unnecessary to incorporate them in any but a token way into the major clan coalitions that have governed the country since 1960.

These cultural attitudes almost certainly influenced the conduct of the factional militias-all of which were recruited from the major clans-during their struggle for territorial control of the south after 1991. The dozens of southern minorities were not perceived as a military threat by the Barre regime, and most did not take up arms against the regime. Ironically, this lack of involvement would later open these groups to charges from the victorious opposition militias that they had contributed nothing to the overthrow of the dictator, indeed that their apolitical and pacifist stance had helped sustain the Barre government in its later years. Thus while the minorities were not as a rule singled out as military targets by the post-Barre militias, [ A notable exception was the Abgal militia's execution of some members of the minority Yibir clan (see appendix I).] they were victimized repeatedly by armed gunmen of all persuasions. Their homes were subject to searches and looting, women were raped and elders were intimidated. The fortunes of certain minorities about which we have more detailed information are reported in appendix I. Outlined below are some of the factors that made minorities as a whole especially vulnerable to militia violence, not only during the early stages of war but also during the UN intervention.

1. Military weakness. Most minorities had few weapons and no military traditions. Some Gosha and Rahanweyn farmers had been forcibly conscripted in the Ogaden war of 1977-78, but they had few high-ranking officers and no stockpiles of weapons. They were largely defenceless.

2. Vulnerable assets. Most minorities depended on fixed assets-fishing boats and nets, shops in the towns and villages, cultivated land and grain stores in the countryside-for their livelihood. These were easy targets for the armed militias and the roving bandits that followed in their wake.

3. Social isolation. Almost by definition, minorities had few marriage or kin ties to members of the major clans [ The reasons for this varied. Most of the established urban families tended to marry among themselves, both as a cultural preference and perhaps as a way of keeping wealth "in the neighbourhood." The Bantu and Rahanweyn minorities also enjoyed few cross-clan alliances, but largely because they were shunned as marriage partners by the "noble" clans.], ties that might have mitigated the treatment they received.

4. Political neutrality. The fact that the minorities were not part of any major clan coalition ironically made them suspect by all sides. When one faction replaced another as temporary master of a region, local residents were accused of harbouring or backing its rival. Incidentally, this served as a convenient pretext for the successive rounds of looting and intimidation experienced by most minorities.

5. Limited support networks beyond their home communities. Their past exclusion from government positions and overseas scholarships, coupled with their attachment to sedentary pursuits, meant that few minorities had established substantial family networks in other parts of Somalia or beyond its borders. If they lost their local productive assets, they had nowhere to turn for material assistance, and if they were displaced from their places of residence, they had no place to flee but to refugee camps. [ An exception needs to be made for certain prominent religious and commercial families of Mogadishu, Marka and Brava who were historically linked by trading interests, and sometimes marriage ties, to similar families along the Arabian and Kenyan/Tanzanian coasts.]

In general, the minorities have had little recourse against the successive waves of militiamen and bandits that scoured the countryside. In time of war, when force is all that matters, communities without weapons, allies or connections enjoy little respect or status, and there is evidence that minorities' property and persons were treated with disdain. There are indications that as the various militias continue to infiltrate the districts where minorities live, their political leaders are attempting to legitimate the takeovers by claiming that coastal gibil'aad ("white skin") and rural Bantu peoples are not really Somalis. Thus the threat to the minorities is not only physical, it also touches on their very identities as Somalis.

3.3 The Situation for Women and Children

 Women have been particularly vulnerable since the outbreak of civil war, both as victims of rape and as disproportionate sufferers from food shortages. Rape was extremely widespread in both the initial Hawiyya sweep against the Darod and Darod reprisals against the Hawiyya. These rapes were not just a product of young soldiers being freed from the constraints of their elders, they also seem to have been deliberately used to humiliate the men of the targeted clan, who were often forced to watch helplessly as their sisters and daughters were violated before they themselves were executed. As children in Somali society generally belong to the clan of the father, [ According to Lewis, "a woman's husband has the right to claim any children the wife produces during their marriage whoever begets them. Children borne out of wedlock by divorced women, or prostitutes, take the affiliation of the man who begets them. Where the father is unknown the mother invents a genealogy for her children; they never take her affiliation" (Lewis 1994, 75, n. 23).] it was widely rumoured that some of the assailants intended to impregnate the women of rival clans in order that they would bear sons for the victors. At the same time there were reports of pregnant women being forcibly aborted, or of fetuses being cut from the wombs of their mothers (AI Aug. 1992).

Women of minority tribes also seem to have been vulnerable to rape, not so much because of vendettas against these clans as a whole, but because these largely unarmed communities could do little to prevent it. Most Bravanese I interviewed in Mombasa were convinced that the repeated violations of Bravanese women by occupying militias were intended to create such intolerable conditions that family heads would abandon the town and leave it to the new occupiers.

Not only was violence against women used as a weapon in clan war, it was also a symptom of the wider breakdown of the traditional system that protected women through the collective support both of her own kin and her husband's. The extremely mobile character of militia warfare in post-1991 Somalia meant that men of fighting age were frequently away from their home settlements, placing their women at greater risk from rival militias and roving bandits. Women left in charge of farming and small stock herding found themselves in the precarious position of being both the producers and defenders of local food supplies. The majority of famine victims in the Bay region were women and children unable to plant or collect food from the nearby fields, while roving soldiers ate off the land or extorted food from relief convoys. One rarely saw pictures of armed youths going hungry.

In Somali society, married women traditionally have served to link the clans of their fathers and brothers, to whom they always belong, with those of their husbands, to whom the children always belong. Most of the nomadic clans practised some form of exogamy-marriage outside the clan-to help strengthen alliances with "outsiders." Wives were exchanged even between clans and clan sections that were prone to fight over water and pasture (like the Habr Gedir Saad and the Majerteen Omar Mahmud, or the Habr Gedir Ayr and the Marehan in central Somalia). These ties helped mediate disputes between clans, since there were always families with in-laws on the other side that would have an interest in the peaceful resolution of conflicts.

Somalis told me that even in the vendettas following the fall of Siyad Barre, occasionally certain individuals among targeted groups would be spared because an attacker discovered that he was in some way related by marriage. Notwithstanding these cases, the recent war seems to have accelerated a trend among most clans toward excluding outsiders and closing ranks: ties of blood (through the male line) have become the primary markers of identity and loyalty. The foremost anthropologist on the Somali has noted that in the current conflict, "in areas formerly characterized by clan heterogeneity, with people of different clans living together harmoniously and intermarrying, marriage outside one's own clan became the exception rather than, as formerly, the rule" (Lewis 1994, 51). Although I cannot confirm this information, I have been told by several Somalis that in the name of clan solidarity, some of their friends have been pressured by kinsmen to divorce their wives from other clans and to take new wives from within their own clan.

The impact on mothers and their children has been equally devastating. In Mogadishu, "women who had married outside their own clan found themselves at a serious disadvantage, they and their children being disowned and left unprotected by both sets of kin" (Lewis 1994, 51). Around Kismayu there were reports of heated domestic quarrels over efforts to recruit the children of mixed clan marriages into the militias of one-usually the father's-side of the family, and even unconfirmed rumours of mothers being pressured by their kinsmen to kill their children because they belonged to the father's clan (Menkhaus n.d.).

Finally, for many women, even flight to Kenya and relocation to refugee camps offered little real security. Incidents of rape have been widely reported, with the assaults coming not only from male Somali refugees but also from Kenyan policemen and security personnel. Women who have to go out to collect firewood are also at risk from Kenyan or Somali bandits (African Rights Sept. 1993; UNHCR Oct. 1993b; Prendergast Jan. 1994, 11).

APPENDIX I: NOTES ON SELECTED MINORITY COMMUNITIES

A. Bajuni

The Bajuni are a small community of perhaps 3,000 or 4,000 who are predominantly sailors and fishermen. They live in small communities along the Indian Ocean coastline and on some of the larger offshore islands between Kismayu and Mombasa, Kenya. There are also substantial numbers of families living in Kismayu, and smaller numbers in Mogadishu and Brava. Older Bajunis frequently identify themselves by their island of origin (Koyama, Ngumi, Chovayi, Chula, etc.), although most have now resettled on the mainland. Most anthropologists believe the Bajuni represent an admixture of Arab, Bantu, Somali and perhaps Malaysian backgrounds. Many have lighter skin and hair than other Somalis, although local Somalis usually distinguish them by their primary language, kibajuni, which is a dialect of Swahili. Those who live or work in the mainland towns also speak Somali. Only a handful have received a western education.

        Bajunis do not seem to have been targeted for personal or political attack during the fighting. Because of their maritime mobility, they have not been in as much danger as other Somali minorities (Menkhaus July 1991). Some even earned money-as much as US$400 per passenger-transporting refugees from places like Brava and Kismayu to Kenya in their fishing boats. Still, in November 1993 I met many Bajunis who had taken refuge with relatives in the coastal villages of northern Kenya, and hundreds of others who ended up in the Jomvo refugee camp near the Mombasa airport. Most observers regard the Jomvo camp as the least desirable refugee site in Kenya.

Many Bajunis relocate to Kenya because their capacity to market fish and repair their boats in Somalia continues to be threatened by the presence of armed militias, who typically steal their equipment and resell it at exorbitant prices. A former UN official recently told me that a quick impact project that donated nets to help restart a fishing cooperative in Kismayu was handed over to the leaders of the dominant factions in the area. When the fraudulent recipients realized they did not know how to use the nets, they simply cut up the nets of local Bajuni fishermen, thus compelling the latter to buy the donated equipment (Menkhaus n.d.).

In these circumstances, many impoverished Bajuni will not be able to repatriate to Somalia with any promise of security. Their ability to earn a living in Somalia may be at an end. At the same time, with only a handful of educated professionals and overseas residents, the Bajuni are unlikely to have the means or the contacts to emigrate in large numbers.

B. Bravans (Barawans)

While sharing the vulnerability factors common to other Somali minorities (see above), residents of the coastal town of Brava seem to have been singled out for harassment from the early stages of the civil war. [ According to Menkhaus, "Brava was singled out for especially harsh treatment by the USC, with murder and rape widely reported" (Menkhaus 1991).] In 1993 a UNOSOM official wrote that "no other ethnic or social group has experienced such sustained and brutal treatment at the hands of roving bandits and shifting armed factions." He added that "other coastal peoples in southern Somalia, all associated with long-standing commercial activities, have experienced similar atrocities. Most of the 'Hamar-Weyn' community in the old quarter of Mogadishu has been driven out of the country, and the small Swahili and Indian population in Kismayu has been forced to flee as well." Menkhaus held out little hope that Bravans who fled to Kenya could safely return to Somalia given that "Somalis have made it abundantly clear that, historical evidence notwithstanding, they consider the Barawans and other coastal peoples to be foreigners" (Menkhaus 3 May 1993). The Bravans I interviewed in Mombasa in November 1993 reported a consistent pattern of looting, rapes and intimidation by the succession of militias that moved in and out of Brava. A former mayor of Brava recorded no fewer than 10 occupations of the town by various militias between January 1991 and May 1992. Most of these occupations involved the expropriation of vehicles and local food supplies, and random acts of terror and rape (Zen Jelani 1992). UN peacekeeping forces never appeared in Bravan, even after UNITAF was launched. According to Bravanese sources, on two occasions food being distributed by a Korean NGO was expropriated by the armed militias in control (Menkhaus July 1991, 43; Zen Jelani 1992).

There are several possible explanations for why Brava's inhabitants were targeted. Even in colonial days the town and its people had a reputation for being peaceful, enterprising and somewhat insulated from the rest of Somali society. Most residents spoke Cimini, a dialect of Swahili, rather than Somali as their primary language. [ Many Bravans have family connections in Swahili communities along the Kenyan coast. These connections often helped them relocate after fleeing Somalia, and also explain the confusion over the ethnic identity of recent Bravan refugees to the UK; because they arrived from Kenya and speak Swahili as a first language, some have suspected them to be Kenyans rather than Somalis. If they speak Cimini (or Chimbalazi), however, they are almost certainly Bravans from Somalia (Cassanelli May 1994).] They also tended to marry among themselves and had a reputation for coexisting with, if never fully trusting, whatever national government was in power. Many lived in gracious stone houses and were reputedly quite wealthy. These characteristics, coupled with their military weakness, made them objects of resentment and contempt on the part of the invading militias. [ Brava was carefully monitored by Siyad Barre's security forces, probably because most of its inhabitants seemed to lack "revolutionary" fervour, and because Brava was a holiday spot visited by many expatriates. I witnessed the arrest of a young Bravan cook who had accompanied my son and me for a weekend holiday to Brava; he was apparently taken into custody for escorting two gaals (foreigners) around town, where his family and many relatives lived. I had to bribe the local security men to have him released.]

The displaced Bravans that I interviewed in Kenya and the UK repeated stories about their gold and household goods being looted, Bravan women being raped and Bravan men being subjected to random acts of violence (see Asha n.d.). They believed these acts of violence were intended to shame the men and drive them and their families out of the town, and to intimidate the elders who remained into recognizing the authority of the USC, and later the SNA (both the Hawadle and Habr Gedr factions). A score of Bravan families could be found living abroad prior to the war, and many more will likely seek to join them in the near future.

C. Somali "Bantu"

 In the Somali context the term "Bantu" was historically applied to individuals presumed to have come from east Africa, and it usually connoted slave origins and low status. More recently the term has acquired a more positive political content, as it has been taken up by Bantu Somalis who have formed their own political organization-Somali African Muki Organization, or SAMO-to press for international recognition. Some Bantu leaders also use the term "Jareer," which formerly was applied to all dark-skinned "African"-looking Somalis, to mark their identity as part of a conscious political movement (Luling 1994).

The Bantu or Jareer who live along the lower Juba, collectively known as Gosha (people of the bush), include distinct groups like the Zigua, Zaramo, Magindo, Makua, Manyasa, Mushunguli and Yao. These tribal names can be identified with peoples living in modern-day Tanzania, Malawi and Mozambique. In all likelihood these are the tribes of origin of their ancestors, who came to Somalia as slaves in the nineteenth century. Over the next few generations many converted to Islam, learned to speak Somali (usually the southern af-maay variant) and even assimilated into local Somali clans, although they retained their low status as ex-slaves in the eyes of the Somali majority.

Other Jareer are long-time inhabitants of farming villages along the Shabeelle river from Qorioley to Beled Weyn, and of interriver settlements like Bur Heybe and Baidoa. Historically, most of these groups have been "clients" of dominant pastoral clans and have depended on these pastoralists for protection and trade. While some of the Shabeelle valley Jareer have suffered from the warfare of the past several years, they have undergone much less displacement than the Bantu minorities (Gosha) of the Juba valley.

The Gosha were hit harder by looting during the early phases of the civil war than was any other social group in the Juba valley, and by June 1992 they had been displaced into some 50 camps (Menkhaus July 1991; Prendergast June 1994, 17). The collapse of the fruit plantations and state farms that had served as supplementary sources of food and employment further eroded their subsistence base, and those who remained in their villages were often terrorized by successive waves of armed militias.

Although SAMO claims to speak for the Somali Bantu as a whole, its constituents reveal different degrees of attachment to the notion of Somali citizenship. Some Bantu recently stated they want to return to their ancestral homelands in eastern Africa, fearing they will never again have security of tenure on their land and will be perpetually considered second-class citizens under any Somali government. A UNHCR field officer reported that roughly 75 per cent of some 10,000 Bantu residents in Kenya's refugee camps in late 1993 said they did not want to repatriate to Somalia (Lehman 1993).

While some portions of the Bantu population have over the past century partially assimilated into Somali society and still consider Somalia their home (for example, those long-resident in Kismayu, the Bay region and along the Shabeelle), others, particularly along the lower and middle Juba, have tended to preserve their distinct cultural and linguistic traditions, and thus are more likely to continue to identify with their east African tribe of origin. Moreover, although all minority landholders seem to be increasingly at risk of land alienation, over the past two decades the Bantu of the Juba valley have been victims of land alienation and state confiscation to a greater extent than have their counterparts in other districts (Besteman and Cassanelli 1995). There were some attempts to repatriate "Gosha" refugees to the Kismayu region in early 1994, but continued violence and insecurity along the lower Juba threatens to create a new exodus of Bantu minorities when UN forces abandon Kismayu.

D. Rahanweyn (Reewin)

The term Rahanweyn or Reewin refers to about three dozen clans that inhabit the interriver zone of southern Somalia. They are concentrated in the Bay and southern Bakool regions, but they also form substantial minorities in Geedo and the lower and middle Juba and lower Shabeelle regions. They are distinguished from other Somalis in their use of af-maay, a regional dialect that serves as a lingua franca among the various Rahanweyn groups, several of whom also speak distinct local dialects, and some of their Bantu riverine neighbours. Af-maay differs substantially from the af-mahaad version of Somali that was made the official written language of Somalia in 1972, and which most Rahanweyn feel disadvantages them in their efforts to achieve literacy and full participation in public life. The Rahanweyn also differ from the other large Somali clan-families in having a mixed agro-pastoral economy. The majority of Rahanweyn families rely on both farming and livestock for subsistence, with the result that particular Rahanweyn groups more closely identify with specific local settlements. Through intermarriage and clientship Rahanweyn clans historically have absorbed individuals and families from other parts of the country into their local communities. With the intense politicization of clan identities during the recent civil war, this practice has led many non-Rahanweyn clans to claim rights to land and representation in Bay and Bakool (Mukhtar and Kusow 1993).

The numerous and genealogically heterogeneous Rahanweyn are usually divided (according to their own categories) into the Digil and Merifle. The Digil consist of seven clans (including Geledi, Begeda, Tunni, Jiddo, Garre and Dabarre) that inhabit the inland districts adjacent to the lower Shabeelle and Juba valleys. The Merifle divide into two large segments: the Sagaal ("nine"), which includes the Hadama, Luway and Gasargudda of Bakool and Geedo regions, and the Siyeed ("eight"), which includes the Harien, Heraw, Emid and Elay in Bay region. Each of these clans is divided into major and minor lineages that occupy relatively distinct territories and often have attached to them dependent groups of artisans, hunters and cultivators whose ancestors may have been slaves. The hierarchical social structure of many Rahanweyn clans, coupled with their heterogeneity, probably made it more difficult for them to mobilize cohesive alliances to defend their region during the early stages of the civil war, and it certainly facilitated the manipulation of clan and status divisions by outside factions competing for territorial footholds in the Bay, Bakool and valley districts.

The susceptibility of the Rahanweyn to famine and displacement in 1991-92 was partly due to their location in the path of the retreating Darod militias and their Hawiyya pursuers, all of whom looted livestock and grain stores indiscriminately. The weaker clan and extended family loyalties of the Rahanweyn and their absence of strong support networks outside the country also contributed to their vulnerability at the height of the crisis (Prendergast June 1994, 36). The arrival of UNITAF forces in late 1992 helped restore order and security in the main population centres around Baidoa, Hoddur and Biyooley, and many Rahanweyn clans have subsequently set up local police forces and courts, and have resumed farming, herding and marketing. There has also been considerable progress in establishing an overarching council of Rahanweyn clan elders that includes elements of the Somali Democratic Movement (SDM) (Mukhtar and Kusow 1993).

However, there are potential problems. Local residents continued to express concern in 1994 that Marehan, Ogaden and Habr Gedr militias were poised to reoccupy the area as soon as international troops left (Prendergast Jan. 1994, 9). Because these regions are landlocked, Rahnaweyn leaders have had to make informal agreements with SNF, SPM and SNA forces to assure access to the lower Shabeelle, where they get fruit and vegetables, and to middle Juba, where some take their herds for seasonal grazing (Mukhtar Mar. 1995). The most vulnerable Rahanweyn groups are those living adjacent to territory controlled by the heavily armed militias of the SNF (near Bardera and Sakow Weyn), SNA (around Bur Hakaba and Wanle Weyn) and SPM (around Dinsor and in the districts north of Hoddur). While these "buffer zones" protect the Rahanweyn of Baidoa and Hoddur from outside clan militias, they also put local residents at risk of intimidation, random banditry and political pressure to join rival factional coalitions (the SDM is divided into at least three subfactions). Rahanweyn clan leaders in these outlying areas have therefore begun to negotiate with Marehan and Ogaden elders in middle Juba and Bakool frontier areas (Prendergast Jan. 1994, 8). Thus while the Rahanweyn as a whole are much less at risk than they were in the early phases of the conflict, some groups, especially in the border areas with other clans, will remain vulnerable to predation and displacement in the event of renewed warfare for the valuable interriver pasture and farm districts.

E. Caste Groups

There is one category of minority clans about whose fortunes in the civil war it is difficult to generalize. These clans, collectively known in the ethnographic literature as "sab," include the Tomal, Midgan and Yibir. Outside observers have sometimes considered them "outcastes" because traditionally they could only marry among themselves and other Somali clans considered them ritually polluted. Living primarily among the nomadic populations of Somalia but in their own distinct settlements, they performed specialized occupational services such as metalworking, tanning and midwifery for the dominant clans in the area (Cassanelli 1969).

In more recent times many "sab" families have migrated to the cities, where they have been employed by politicians in more powerful clans as drivers, bodyguards and spies. For example, Siyad Barre elevated several Midgan to important positions in the ministries of defence and education. With no independent clan base or status of their own, such appointees could be trusted to carry out orders. Other clans also employed Tomal, Midgan and Yibir families.

While as a group these "sab" minorities did not pose a significant threat to any other Somali group, particular individuals and families who had visibly supported the old regime were vulnerable to retaliation. The most striking example occurred when Abgal militiamen of the USC executed several Yibir who had formerly been their clients. In its waning days the Barre regime reportedly armed many Yibir and sent them into the local market to kill Abgal clansmen in retaliation for Abgal rebellion against the government. After Abgal forces succeeded in helping oust Siyad Barre, they turned on their former Yibir clients and massacred them. Having broken with their former patrons, the Yibir had no one to turn to for protection (Menkhaus n.d.).

While I have been unable to find evidence of other cases of systematic retaliation against these minorities, other Midgan, Tomal or Yibir may have been targeted during the civil strife. Because they have no natural clan allies in the wider society, and no collective voice in political circles, they can be attacked with impunity.

APPENDIX II: MAPS

Map I: Operation Restore Hope

See original

Map II: Population Density

See original

Map III: Distribution of Dialectal Groups

See original

REFERENCES

Africa Watch. 1993. Somalia: Beyond the Warlords. New York: Human Rights Watch.

Africa Watch. 1990a. Somalia: A Government at War with Its Own People. New York: Human Rights Watch.

Africa Watch. 1990b. Somalia: Evading Reality. New York: Human Rights Watch.

African Rights. October 1993. "Land Tenure, the Creation of Famine, and Prospects for Peace in Somalia." Discussion Paper No. 1.

African Rights. September 1993. The Nightmare Continues...Abuses Against Somali Refugees in Kenya. London: African Rights.

African Rights. May 1993. Somalia: Operation Restore Hope: A Preliminary Assessment. London: African Rights.

Amnesty International (AI). 16 December 1994. Somalia: Amnesty International Protests Cruel Punishments. (AI Index: AFR 52/WU 02/94). London: Amnesty International.

Amnesty International (AI). 30 April 1993. Somalia: Update on a Disaster-Proposals for Human Rights. (AI Index: AFR 52/01/93). London: Amnesty International.

Amnesty International (AI). 15 February 1993. "Amnesty International's Concerns Regarding Human Rights Violations Against Minority Tribes in Somalia." Toronto: Amnesty International.

Amnesty International (AI). August 1992. Somalia: A Human Rights Disaster. (AI Index: AFR 52/01/92). London: Amnesty International.

Asha (Refugee Somali woman MD), London. n.d. "The Sufferings of the Somalis." Personal interview. (transcript)

Besteman, Catherine and Lee Cassanelli, eds. 1995. The War for Land in Southern Somalia. Boulder, Col.: Westview Press.

Cahiers d'Etudes Africaines [Paris]. 1993. No. 2. Roland Marchal. "Formes de la violence et de son controle dans un espace urbain en guerre: les mooryaan de Mogadishu."

Cassanelli, Lee. 1995. "Explaining the Somali Crisis," The War for Land in Southern Somalia. Edited by Catherine Besteman and Lee Cassanelli. Boulder, Col.: Westview Press.

Cassanelli, Lee. November 1994. "The Role of Somali Diaspora Communities in Homeland Politics." Revised version of a paper presented to Workshop on Ethnic Group Conflict, Woodrow Wilson International Center for Scholars, Washington, DC, August 1994.

Cassanelli, Lee. May 1994. "History and Identity Among Somali Refugees: A Recent Example from Coastal Kenya." Revised version of a paper presented to the Folklore Seminar, University of Pennsylvania, January 1994.

Cassanelli, Lee. 1969. "The Sab Low-Status Groups in Northern Somaliland." Seminar paper, University of Wisconsin, Madison.

Country Reports on Human Rights Practices for 1994. 1995. United States Department of State. Washington, DC: United States Government Printing Office.

Cultural Survival Quarterly [Cambridge, Mass.]. Spring 1994. Vol. 18, No. 1. Alex de Waal. "The UN and Somalia's Invisible Minorities."

Cultural Survival Quarterly [Cambridge, Mass. Winter 1989. Vol. 13, No. 4. Robert Gersony. "Why Somalis Flee: Conflict in Northern Somalia."

Horn of Africa [Summit, NJ]. Jan.-June 1990. Vol. 13, No. 1-2. Daniel Compagnon. "The Somali Opposition Fronts: Some Comments and Questions."

Indian Ocean Newsletter [Paris]. 18 February 1995. No. 659. "Somalia: The Bloody Banana Bust-Up."

Indian Ocean Newsletter [Paris]. 11 February 1995. No. 658. "Somalia: Now They're Shooting for Bananas."

Lehman, Daniel J. 1993. "Resettlement of the Mushunguli, Somali Refugees of Southeast African Origins." Paper sent to the Fifth International Congress of Somali Studies, Holy Cross College, Worcester, Mass., 1-3 December 1993. (With permission of the UNHCR country representative in Nairobi)

Lewis, I.M. 1994. Blood and Bone: The Call of Kinship in Somali Society. Lawrenceville, NJ: Red Sea Press.

Luling, Virginia. 1994. "The Origins of the 'Jareer,' People of the Shabeelle: The Implications of Some Rituals." Unpublished paper.

Menkhaus, Ken. n.d. Personal communication.

Menkhaus, Ken. September 1994. "Trip Report-Somalia."

Menkhaus, Ken. 1993. UNOSOM "Lower Jubba Strategy" Briefing Papers: No. 1. (July) "The Historical Roots of the Current Crisis. Patterns of Settlement and Displacement"; No. 2. (June) "The Economic Dimension of the Kismaayo Conflict"; No. 3. (June) "Foreign Assistance in the Lower Jubba Valley: The Past Record."

Menkhaus, Ken. 3 May 1993. "Statement Regarding the Status of the Barawan (Bravan) People of Coastal Southern Somalia."

Menkhaus, Ken. 1991. "Notes on the Current Situation in Somalia-April 1991." Compiled on the basis of talks and interviews with foreign aid expats, reporters, Somalis in Kenya and several written reports.

Menkhaus, Ken. July 1991. "Report on an Emergency Needs Assessment of the Lower Jubba Region (Kismaayo, Jamaame and Jilib Districts), Somalia." Submitted to World Concern.

Mukhtar, Mohamed Haji. March 1995. Personal communication.

Mukhtar, Mohamed Haji. 1992. "Ergada Mission to the Horn of Africa (The Inter-river Regions of Somalia)." 18 August-14 September 1992.

Mukhtar, Mohamed Haji and Abdi Mohamed Kusow. 1993. "The Bottom-Up Approach in Reconciliation in the Inter-River Regions of Somalia." A Visiting Mission Report. 18 August-23 September 1993.

Politique Africaine [Paris]. 1992. No. 46. Roland Marchal. "La guerre a Mogadishu."

Prendergast, John. June 1994. "The Gun Talks Louder Than the Voice: Somalia's Continuing Cycles of Violence." Washington, DC: Center of Concern.

Prendergast, John. January 1994. "The Bones of Our Children Are Not Yet Buried: The Looming Spectre of Famine and Massive Human Rights Abuse in Somalia." Washington, DC: Center of Concern.

Reuter European Business Report. 17 February 1995. (Internet)

Samatar, Said S. 1991. Somalia: A Nation in Turmoil. London: Minority Rights Group.

Somali News Update [Uppsala, Swe.]. 1992. (An independent newsletter devoted to critical analysis of political and humanitarian developments in Somalia. Edited and published by Dr. Bernhard Helander, Uppsala University, with support from the Scandinavian Institute of African Studies)

United Nations High Commissioner for Refugees (UNHCR), Geneva. October 1993a. Information Bulletin. "Kenya-Cross Border-Somalia."

United Nations High Commissioner for Refugees (UNHCR), Geneva. October 1993b. Refugee Women Victims of Violence: A Special Project by UNHCR.

United Nations High Commissioner for Refugees (UNHCR), Geneva. September 1993. "Appeal for Repatriation to Somalia."

United Nations High Commissioner for Refugees (UNHCR), Geneva. June 1993. Information Bulletin. "Kenya-Cross Border-Somalia."

Zen Jelani. 1992. "Un breve cenno storico sulla Communita Bravana." Includes overviews and brief notes on the successive occupations of Brava from December 1990-May 1992.