Ethiopia’s Long War in Historical Perspective Part 2

Fragmentation takes a value of 0 or 1, while stalemate vs. gain/loss is also framed as dichotomous 0 or 1 variable. Crucially, a dichotomous measure of prior rebel fragmentation is included in order to investigate whether the correlation between fragmentation and gains/loss years is being driven by the endogeneity I have identified. Figure 3 presents these results. Figure 3: Baseline Model of Rebel Fragmentation Variable Full Model Gain/Loss 2.440901*** (.6318165) Prior Fragmentation -.3343384 (.4498132) N 223 Constant -3.445373 ***Significant at the .01 level, ** significant at the .05 level, and * significant at the .10 level The results suggest that despite endogeneity concerns, gains and losses remain a robust predictor of rebel fragmentation. In fact, prior fragmentation actually reduces the likelihood that fragmentation will occur again in the future. Yet while this exercise provides a useful point of departure, by itself, it does not validate the causal claims I seek to make.22 With only one control in the model, it is likely 22 One should note that gain/loss years are unbalanced, with loss years accounting for close to 2/3 of the sample of gain/loss years. When the model is run by splitting gain and loss years, loss years remain marginally significant that other confounding variables are driving our results. Furthermore, this sort of cross-sectional, time series analysis is suspect with an N of 223. And finally, our efforts ignore the sort endogeneity that results from “simultaneity” bias: since our unit of analysis is rebel year, one cannot be sure whether territorial gains and losses occurred before fragmentation, or whether fragmentation occurred before territorial gains and losses. In order to resolve these issues, I subject one of the fourteen rebel organizations in the sample to careful process tracing in a conscious effort to both elucidate the causal mechanisms that link independent variables to outcomes, and carefully weigh competing explanations.23 In this vein, the primary empirical thrust of this paper is a longitudinal study of Jebha, a rebel organization that participated in a secessionist conflict in what at the time was Ethiopia’s coastal province of Eritrea. For reasons of space, I focus on Jebha between the years 1961-1977, and trace the organization through 4 periods— its early origins, a period of gain, stalemate, and loss. I therefore exclude the last five years of the organization’s life, and a very important episode of fragmentation in 1982, when Jebha was reeling from significant military losses. While I do not formally discuss this period, I do use it to grant added leverage in investigating alternative variables. The paper uses an eclectic range of sources, but is primarily based on field work conducted in Ethiopia and Ethiopian and Eritrean diaspora communities in the US and Europe, throughout 2009. Over 70 interviews were conducted with a range of individuals, including ex-combatants, government officials, and civilians of varying partisan stripes who lived through the conflict. I’ve also utilized thousands of pages of archival material from the National Archives in Washington, DC, while gains do not, although the sign of the co-efficient is the same. Given the extremely small sample size, these results are not surprising 23 See (Eckstein 1975) and (George and Bennett 2005) for discussion of process tracing the National Archives in the United Kingdom, and the Institute of Ethiopian Studies. This data has been supplemented by an analysis of original organizational documents and available news reports. Background to Rebellion When the secessionist war in Eritrea came to a conclusion on May 24, 1991, it had been the longest running conflict on the African continent (see Figure 4 for old provincial map of Eritrea). Although the insurgency began as little more than sporadic banditry, in the end, its protagonists would be engaged in the largest mechanized battles on the African continent since World War II.24 Figure 4: Provincial Map of Eritrea 24 In a telling commentary to the BBC in the late 1980’s, the British historian Basil Davidson described the Ethiopian’s government’s defeat at Afabet, Eritrea in 1988 as “one of the biggest victories ever scored by a liberation movement since Dien Bien Phu in 1954” (Connell, Against All Odds: A Chronicle of the Eritrean Revolution 1993, 228) Any objective measure of the cost of the war illustrates this fact. Gebru Tareke estimates that Ethiopia’s civil wars, of which Eritrea was the central front, killed between 800,000 to 1 million people, roughly 2 percent of Africa’s second largest country (Tareke 2009, 132).This excludes the famine of 1984-85 – of which the war was the principle cause – in which another 1 million people died. If we assume that a total of 2 million people died as a consequence of the war, this would make war time deaths in Ethiopia 20 times larger than Bosnia, roughly double of the Iran-Iraq conflict, and a little more than equal to that of Afghanistan between 1979 and 2001.25 The economic costs of the conflict were not minor either. Gayle Smith, a journalist who covered Ethiopia’s internal conflicts in the 1980’s, conjectured that at the height of the secessionist war in 1982, the Ethiopian government was spending 1.5 million a day in fighting the insurgents (Smith 1983, A22). When the military government of Colonel Mengistu Haile-Mariam finally lost power in May of 1991, it had accumulated a foreign debt of USD 9 billion in trying to find a military solution to Ethiopia’s civil wars (Tareke 2009, 137). The major actors in the complex Eritrean drama – besides the central government – were two rebel organizations, the Eritrean Liberation Front (Jebha) and the Eritrean People’s Liberation Front (Shaebia).26 While both organizations were firmly committed to Eritrea’s secession from Ethiopia, they had a contentious relationship. Jebha was the first organization to emerge in 1960, and was the standard bearer for Eritrean independence until several splinter groups coalesced into Shaebia in February 1972. Shaebia’s emergence was a significant moment in the trajectory of the war, launching both Eritrean nationalist organizations into a bloody contest for supremacy from 1972-74. Despite its numerical superiority, Jebha was unable to eliminate its younger rival, and a somewhat uneasy détente between the two organizations ensued until 1980, when another dispute within the 25 26 PRIO Battle Deaths Dataset, 2005. http://www.prio.no/CSCW/Datasets/Armed-Conflict/Battle-Deaths/ “Shaebia” is the Arabic term for “people” or “popular” nationalist camp resulted in Jebha’s demise. Having eclipsed its parent organization, Shaebia defeated the Ethiopian military conclusively in 1991, and after a UN monitored referendum, declared an independent Eritrea. For the most part, this paper focuses on Jebha and leaves the relations between the two insurgent organizations untouched, except when these relations tell us something about the Jebha’s shifting military position. However, most histories of the Eritrean conflict focus a great deal on the dynamic between the two groups. The origins of the secessionist conflict in Eritrea can be found in the arrival of the Italians on the Red Sea coast in the latter half of the 19th century. Prior to the Italian colonial state of Eritrea, “Eritrea” did not exist as a political or administrative entity. While most parts of lowland and coastal Eritrea, predominantly populated by Muslim tribes, had links to Ottoman, Egyptian, and Sudanic political authorities, the central highland plateau – predominantly populated by Orthodox Christians – had historical links with an expanding Ethiopian monarchy to the south. In a sense, Ethiopia’s coastal province of Eritrea, like other parts of Sahelian Africa, was the meeting point of two very different cultural entities: one Arab and Islamic, the other African and predominantly Christian. Importantly, these different cultural heritages mapped on to two very different modes of social organization and production, with the Muslim lowlands predominantly pastoral, and the Christian highlands a sedentary, rain fed agricultural economy.27 These different political, cultural, and economic legacies largely outlasted the Italian period, and as I show, were crucial to understanding Jebha’s internal politics. Eritrea was administered by the British after World War II, and following a UN mandated inquiry into the future status of the territory, was federated with Ethiopia in 1952. While no referendum was 27 Obviously, this is a very stylized account. Eritrea’s social and ethnic demography is far more complex- the territory technically has nine ethnic groups. For more complete accounts see (Negash, Italian Colonialism In Eritrea: Policies, Praxis, and Impact 1882-1941 1987) and (Pool 2001). The above comments were only meant to contextualize what many believe to be of the major cleavage in Eritrean politics. However, many other lines of cleavage had been relevant at different points in time, particularly “awaraja” or provincial/ regional differences held on the issue, several political parties mobilized on different sides of the debate: the Muslim League and others largely rejected union with Ethiopia, while the Unionist Party, encouraged by the very powerful Orthodox Church, supported the idea. In the end, all major parties accepted the decision of the UN with varying degrees of enthusiasm. In retrospect, most recognize that the Federation was destined to fail. The lengthy Italian colonial period, and the British Administration that had followed it, had substantially changed Eritrea: it had acquired an assembly, political parties, labor unions, a well-trained bureaucracy, commercialized agriculture and a manufacturing sector that was the most substantial in the Horn of Africa. By contrast, Emperor Haile Selassie’s Ethiopia – one of only two African countries to have never been colonized – had legal slavery until the mid-20th century, was feudal, and desperately poor. Furthermore, royal absolutism and federalism were a contradiction in terms. Piece by piece, Haile Selassie destroyed the Federation, formally incorporating the territory as Ethiopia’s 14th province in 1961 amidst recriminations from several opposition parties.28 While much of Eritrea’s highland Christian population was persuaded on grounds of shared religious and cultural heritage,29 the Imperial Ethiopian Government (IEG) employed coercion as well. Opponents were jailed and killed, while others fled to neighboring countries.30 For Eritrea’s Muslim lowlanders complete union with Ethiopia was a terrible proposition, and a quick glance at the make-up of Ethiopia’s governing class illustrates why. Of the top 26 officials within the Imperial Ethiopian Government in 1973 – ministers, commanders of the armed services, 28 My assessment of this period is drawn from a large literature. See (Iyob 1995), (Selassie 1980), and (Gebremedhin 1989). For a somewhat different interpretation see (Negash, Eritrea and Ethiopia: The Federal Experience 1997) 29 Authors interview, Provincial Bureaucrat, Oct 11th, 2009. In this and prior discussions, interviewee had provided his recollection of events as early as the 1950’s 30 Pro-independence politicians like Idris Mohammed Adem, Ibrahim Sultan, and Woldeab Woldemariam all fled to Cairo during this period. Woldeab actually survived over four attempts on his life. See Issayas Tesfamariam interview with Nicole Saulsberry, Nov 7, 2002, www.shabait.com provincial governors – not one was a Muslim. Of the 101 government officials that sat directly below those with major government portfolios – deputy ministers, administrators, division commanders – only 3 were Muslims.31 This, by the way, was in a country that was 50 percent Muslim and had more Muslims than every Arab country except for Egypt, Morroco, and Algeria (Woldegiorgis 1989, 73) . The IEG’s provincial administration in Eritrea largely reproduced the pro-Christian bias at the center. In 1965, for instance, both the Governor-General and his deputy were Christians, as were seven of the eight provincial governors. So were the Commissioner of police and his deputy. So were the seven top ranking military officers in the province. So were all 5 of the GovernorGeneral’s cabinet, and both members of his secretariat. One could go on and on.32 Given the public face of Ethiopia’s administration in Eritrea and its overt discrimination against Muslims, it was no surprise that the rebellion that began in September 1961 found its home in the Muslim communities of western Eritrea. Early Rebellion, Losses, and Fragmentation: Jebha 1961-1967 Jebha was organized by a collection of Muslim exiles in Cairo in July 1960 who were disenchanted with Haile Selassie’s willful disregard for the Federation, and felt that the only reasonable response was armed resistance. The three most prominent players in the organization’s early days were Idris Mohammed Adem, a former President of the Eritrean Assembly, Idris Glawedos, his close associate and a law graduate, and Osman Saleh Sabbe, an enterprising school master from the MassawaHirgigo area. 31 See Telegram 51 from Addis Ababa to State, date unclear, RG 59, Central Files 1970-73 “Potential Leaders Biographic Reporting List- 1973-74” 32 See Imperial Ethiopian Government, Governate General of Eritrea, Officials List, 1965 The Cairo triumvirate was an interesting, if somewhat odd grouping of Eritrean dissidents. As his position as former President of the Eritrean Assembly suggests, Adem was the political heavyweight of the three, and had garnered popular appeal in the lowlands for his early opposition to Haile Selassie’s subversion of the Federation. Glawedos, somewhat younger than his two associates, operated as the military brains behind Jebha’s early operations while Sabbe – the smooth talking, urbane scholar who would publish several Arabic texts on Eritrea – served as the intellectual and ideological force behind the movement, as well as its chief interlocutor in foreign capitals. More important than these differences in background and disposition, however, was that while Adem, Glawedos, and Sabbe were all lowland Muslims, they represented very different ethnic constituencies: Adem was a Tigre- Beni Amir from Barka, Glawedos a Bilen from the Keren area, and Sabbe a Tigre Assuarta from the coastal area of Massawa-Hirgigo. Thus, from its earliest days, Jebha represented a coalition of Eritrea’s Muslim tribes. In early 1961, the Cairo leadership appealed to Idris Hamid Awate of the Beni Amir tribe, a fiery local chief who had served with the Italians in Rome during World War II, and become well known for his bandit activities during the British period.33 Responding to the Cairo group’s call for armed resistance, Awate and a band of thirteen of his kinsman attacked Ethiopian police forces at a place called Mt. Adal west of the town of Agordat. The initial battle lasted for nearly seven hours, despite the fact that between Awate’s thirteen kinsmen there were only four guns.34 33 The Beni Amir are a Tigre tribe, all Muslim and pastoral, who reside on both sides of the Eritrean and Sudanese border. Most of the early fighters were drawn from among this tribe 34 Drawn from “Hamid Idris Awate: The Spirit of Defiance,” www.awate.com. One should note that the inclusion of Awate’s band created yet another relevant political cleavage in the organization: that between the external leadership in Cairo and locally organized fighters in the field Life was hard for the first group of fighters. Few in number, and lacking adequate armaments, the early Jebha contingents sought to avoid major confrontations with the Ethiopian police and the few army regulars in the province. As one early fighter noted, “We were short of literally everything. We were few in number; carried very few rifles; had no ammunition stock and space for movement was very limited and we had to suspect most people in the villages. For security reasons, we were covering very long distances in a single day, manly in Barka, Gash, and Setit. We were nowhere and everywhere at the same time…The first three years were terribly difficult.” 35 Where Jebha fighters engaged enemy forces, it was usually in the form of ambushes designed to capture caches of arms and ammunition. For example, during an operation at the end of 1963 Jebha commandos hijacked a bus and drove it into the town of Haicota where they took the Ethiopian police station by surprise. In the end, the rebels escaped with thirty-two guns and twelve hand grenades. Other less daring ventures were an ambush of an Ethiopian convoy in Telay in mid-1963 where seventeen guns were captured, and an ambush at Ansaba that same year, where twenty-three more guns were seized. While such early successes seem trivial in retrospect, they were vital to the early survival of the relatively isolated Jebha contingent.36 Despite operational challenges, morale amongst the Jebha’s early recruits was high.37 Not only were the early batch of fighters deeply committed to the cause, but they were buoyed by the Ethiopian 35 See Eritrean Newsletter, issue no. 44, Sept. 1981, “Interview with Mohammed Ibrahim Bahdurai” Much of the discussion of the early operations of the ELF is drawn from “From the Experiences of the Eritrean Liberation Army,” www.nharnet.com. Most of this information, in turn, is drawn from an Arabic language book published by Jebha leaders Abdullah Idris and Mohammed Hasab, “Experiences of the ELA: 1961-1981.” For other good accounts see (Markakis 1987, 104-45). Both (Markakis 1987) and (Tareke 2009) dispute the official ELF accounts, and argue that Idris Awate had no formal links with the external leadership until after the rebellion started 37 As a general point, nearly all of the Jebha fighters I interviewed revealed themselves to be deeply committed to the cause of Eritrean self-determination. Unlike Renamo or the Tamil Tigers, there was no forced recruitment. Somewhat paradoxically, a movement that suffered from significant episodes of factionalism, had a membership that was highly dedicated to the cause 36 administration’s clumsy response to the outbreak of hostilities. Dismissing the rebels as “shifta,”38 the Ethiopian government decided to deploy only provincial police to deal with the attacks, instead of regular army units who were better equipped to deal with the demands of counterinsurgency (Erlich 1983, 35). As Jebha began to make a name for itself amongst the relatively friendly Muslim communities of the western lowlands, recruitment began to increase. Two groups were particularly important during this phase. The first was a contingent of Eritreans who had served in the Sudanese army, who brought with them the formal military expertise that Jebha desperately needed. Of this group, were Mohamed Idris Haj, Omar Izaz, Abu Tiyarra, and Abu Rijeila, all of whom became important military commanders during the 1960s and 1970s. A second contingent was a group of Muslim Eritrean policeman who defected to the rebels and brought with them not only technical expertise, but an intimate understanding of the enemy and its strategy.39 By the end of 1963 Jebha had 250 men under arms, and by the end of 1964, they had around 800. Additionally, the leadership in Cairo had achieved some success in attracting external aid, particularly from Baathist regimes in Syria and Iraq, which was then funneled to fighters in the field by land and sea.40 By 1965, Jebha’s began to successfully recruit beyond the pastoral communities of the western lowlands. Among the new sectors to join the nationalist organization were urban, predominantly Muslim recruits from the coastal area of Massawa-Hargigo, an increasing number of Christian students from the highlands, and younger, foreign trained militants of various ethnic and religious backgrounds. In order to accommodate this new diversity, Jebha sought to reorganize along ethnic 38 39 40 Shifta is the Amharic word for “bandit” “From the Experiences of the Eritrean Liberation Army,” www.nharnet.com and (Markakis 1987) “From the Experiences of the Eritrean Liberation Army,” www.nharnet.com and (Markakis 1987) and tribal lines in 1965 and 1966. Using Algeria’s AFLN as a model, the Cairo triumvirate – which had now relocated to Khartoum after running afoul Egyptian President Gamel Abdel- Nasser – established five zonal commands, in which commanders and recruits from the ethnic and tribal constituencies that resided within these zones were given responsibility for operations (Markakis 1987) (Iyob 1995) (Pool 2001). Although designed to promote the effective use of local resources while enhancing the operational flexibility of local commanders, Jebha’s restructuring effectively tribalized the organization, discouraging coordination across zonal units and undermining the development of solidarity across Jebha’s ethnic and tribal groupings. As growth and internal differentiation progressed, the Ethiopian government became increasingly alarmed. The very limited expansion of Jebha activities onto the economically vital highland Christian plateau, which was linked to Jebha’s recruiting successes amongst historically proEthiopian Christian communities and its successful extortion of local merchants and wealthy farmers, was a particular cause for concern, and spurred the Ethiopian government to action. In 1965 Gov. Gen Asrate Kassa assembled a paramilitary force exclusively comprised of Christian Eritreans called Commandis 101. Trained in the latest counter-insurgency techniques at an Israeli run military academy in the highland town of Decamhare, the Commandis became a formidable force, consistently bearing the brunt of counter-insurgency responsibilities in the rural parts of the province, while regular military units – mostly comprised of soldiers from Ethiopia proper – remained garrisoned in the major towns. The Commandis close connection with Christian communities in the highlands made them very effective, and their presence made it virtually impossible for Jebha to establish an effective presence on the highland plateau until 1974 when the monarchy collapsed.41 While the emergence of the Commandis signaled that tougher times were on the horizon for Jebha, it was a major Ethiopian Army offensive in March of 1967 that nearly crippled the young insurgency and unleashed an orgy of factional conflict within Jebha’s ranks. What the Ethiopians had long treated as a problem of policing, know became an issue of full- fledged counter insurgency. Consistent with this strategic shift, the government redeployed two extra brigades of army regulars to the province, and launched a series of crisp offensives that were as brutal as they were effective. Jebha units acquitted themselves poorly, as the five zonal commands were so decentralized that they were unable to coordinate any meaningful resistance. Attempting to exploit Jebha’s organizational weaknesses, the Ethiopians employed infantry, armor, and aerial bombardment to strike and isolate each zone, moving west from the highlands through Gash-Barka, and then north-eastwards towards Keren in an attempt to strangle retreating Jebha units. Confused and bewildered, Jebha cadres fled in a haphazard and disorganized fashion, leaving the predominantly Muslim communities that had given them food and shelter to face government troops that heretofore they had never really encountered in great numbers. The result was predictable. The Ethiopian’s created strategic hamlets of 3000 inhabitants each to cut the rebels off from the population. They burned villages and massacred civilians, causing over 20,000 civilians to flee to Sudan, the first refugee wave of many that would follow over the course of the next 24 years. The discontent among the population was palpable, especially among women. As former Jebha cadre Haile Woldense recalls 41 Airgram 18 from Asmara to State, March 2, 1970, RG 59, Central files “Eritrean Commando Police Now at Full Strength.” Also, authors interview, Jebha Intelligence Officer, Sept 11, 2009 “So the first protests came from the mothers. This is very important. The mothers were saying: ‘We are feeding the fighters while leaving our children hungry. But when the enemy came, rather than at least not be inside to give a pretext for the enemy to burn the village, the worst thing the ELF (Jebha) fighters were doing was they stayed in the villages, fired some shots, and then escaped. And they wouldn’t even try to do some fighting so the elderly, the children, and the mothers could get out.’ And of course a lot of people went to the Sudan and they were saying that the fighters had no performed well. There was no support.” (Connell, Conversations with Eritrean Political Prisoners 2005, 48-49) For an organization that had traded on its popular appeal in the western lowlands, this was a devastating development. Yet the poor showing in combat was not all Jebha’s fault. The Ethiopian offensive, either by intention or happenstance, coincided with the Six Day Arab-Israeli War of June 1967, in which Jebha’s Iraqi and Syrian allies were thoroughly trounced. The result of developments in the Middle East was two-fold: not only were Jebha’s Arab benefactors no longer in a position to offer the support that they had provided previously, but the Israeli’s were able to close shipping lanes in the Red Sea. Ominously, Jebha was incurring acute shortages of supplies at precisely the moment that they could least afford them.42 Even Jebha sources signal that 1967 was a tough year. In otherwise effusive historical publication, Jebha writers noted that in 1967 “The very continuation of the revolution was in danger…Fighters in the field, who had little food and arms supplies, were being overwhelmed with desperation.”43 42 INR Memo, George C. Denney to Acting Secretary, April 1, 1968, RG 59, Central Files “Ethiopia: Kagnew and the ELF.” Also see Airgram 174, Asmara to State, Oct 9, 1968, RG 59, Central Files, “The ELF and Kagnew: An Assessment.” See also Jeffrey Lefebvre, “Middle East Conflicts and Middle Level Power Intervention in the Horn of Africa, Middle East Journal, Summer 1996 43 See Eritrean Newsletter, issue No.44, Sept 1981, “Diary of the Revolution” Detailed American declassified intelligence reports from the period largely echo these observations, noting that Ethiopian military efforts during 1967 had been “consistent success story” and that Jebha had seen “a loss of momentum.” These reports go on to say, given the success of the 1967 offensive, the Ethiopian military would phase out of the province the majority of army regulars— some 6000 in all.44 To Ethiopian authorities, the insurgency in Eritrea had been dealt a severe blow. In territorial terms, Jebha’s losses were significant. On the eve of the offensives of March 1967, Jebha’s guerilla army effectively operated in about half of the country—in Sahel, Barka, Gash-Setit, Seraye, and parts of Hamasein province.45 After the Ethiopian offensives, Jebha units were forced to operate in a small, largely uninhabited corner of land in northern Sahel bounded by the Sudanese border to the west and the Red Sea to the east. While Jebha had lost popular support and credibility, its territorial losses were truly astounding. What had been an organization with national scope was now confined to a dusty outpost in a remote, strategically irrelevant district that had more hyenas than people. Figure 5 and Figure 6 provide district level maps that give an indication of the size of Jebha’s territorial losses in 1967. 44 45 Airgram 174, Asmara to State, Oct 9, 1968, RG 59, Central Files, “The ELF and Kagnew: An Assessment” These are the names of Eritrean provinces, of which there are nine Figure 5: Jebha Areas of Operation, by District, March 1967 In the wake of the Ethiopian offensive, the breakdown in cooperation amongst Jebha’s membership was almost immediate, and largely played out along sectarian lines.46 Osman Hishal, Deputy Commander of the 5th zone, summarily executed twenty-seven Christians for failing to perform their duties during the Ethiopian offensive. For Hishal, and many Muslim lowlanders like him, the influx of Christians into Jebha raised the prospect of penetration by the Ethiopian authorities. The successful Ethiopian offensive, and the territorial losses it wrought, represented Christian collusion with the Ethiopian government. Significantly, the executions were carried out without the approval of the zone’s commander, Woldai Kahsai, who himself was a Christian. (Markakis 1987) 46 INR Memo, Thomas Hughes to Secretary, April 8, 1969, RG 59, Central Files, “Ethiopia: ELF takes a Radical Turn”- memo draws direct connection between military failures of 1967 and out- break of “internecine fighting” Figure 6: Jebha Areas of Operation, by District, November 1967 The summary executions served as a trigger: Woldai Kahsai, who was in the Sudanese- Eritrean border town of Kassala at the time, took a contingent of Christians and defected to the Ethiopian embassy in Khartoum. For the remaining Christian cadres within the 5th zone, the defection of their commander was too much to bear. In an effort to build a coalition Christian cadres appealed to many new arrivals from the 3rd and 4th zones, and together these constituencies began to demand reform. Criticisms were leveled at the external leadership in Khartoum, who were accused of absenteeism, at the zonal command structure, that lacked coordination, and at a nationalist program that was Arab centric and tended to exclude Christians. Calls for reform evolved rapidly into open insubordination as units in the field began to make efforts to re-organize independently of the external leadership. What followed in the months ahead was an incredibly complex set of maneuvers between the external leadership, their supporters in the field, and the reformers. Politically attuned and ever resourceful, the triumvirate of Adem, Glawedos, and Sabbe attempted to block reforms that would inevitably reduce their power. In a series of meetings in 1968 and 1969 that were ostensibly designed to reform the organization, they were able to mobilize support and successfully infiltrate the ranks of the reformers. Many were purged and killed. Interestingly, as this process played out, open splits emerged amongst the external leadership as well. Indeed, as Jebha descended into internal chaos, the military picture did not get much better. At Hal Hal in 1968, Jebha units from the 2nd zone suffered some of the organizations worst losses of the war. For Sabbe, who had been Jebha’s most prolific fundraiser abroad, the writing was on the wall. In 1969 he broke with leadership and created what he called the “General Secretariat” in Amman, Jordan, and actively began to encourage others to defect from Jebha ranks. In the field, the losers of the factional disputes at the meetings of 1968 and 1969 exited the organization in a bid for self-preservation. A group of Christians under the direction of Abraham Tewolde set out for the Ala plains in the eastern part of the highlands, while a collection of younger cadres from the Massawa-Hirgigo area, many foreign trained, were reorganized by Sabbe in Yemen. Yet another faction, led by Adam Saleh, set off for the Obel River to re-organize.47 It was these diverse and disenchanted elements who would form the nucleus of Shaebia in 1972.48 47 The departure of Adam Saleh and what became known as the “Obel Group” is quite interesting, because this group could not properly be considered reformers given their ideological orientation. Their departure, more than anyone else, signals what I am trying to argue, which is that this episode of factionalism was largely a result of Jebha’s military failures 48 The preceding four paragraphs are drawn from a range of historical works (Markakis 1987), (Iyob 1995), (Pool 2001). For a primary account of these events see “Nhanan Elaman,” which can be translated as “We and Our Objectives.” The manifesto was written by Christian dissidents who split from Jebha in the early 1970’s