Whether opposition parties opt for armed struggle or peaceful methods in their attempt to overthrow the existing regime in Ethiopia, they must all come up with a vision and a political solution that can heal decades of ingrained conflicts and reconstitute national unity. Since the ethnic conflict is by far the most divisive and pernicious issue of present day Ethiopia, the endeavor both to defeat the regime and establish a post-TPLF political system presupposes an approach dealing with ethnicity. It is illusory to assume that ethnicity will simply go away if the TPLF is defeated. What follows is an attempt to show how a correct theory of ethnicity and lessons from the past history of Ethiopia can help us frame a political arrangement that favors the establishment of peace and democratic governance in our country.

For one school of thought called primordialism, ethnicity is about self-determination; it is a primordial and emotional attachment to fixed social characteristics, such as blood ties, race, language, region, and custom. Such an attachment naturally longs for political sovereignty as a necessary means to protect and develop the treasured characteristics. The best way to resolve ethnic conflicts, so primordialists conclude, is to allow peoples the right to live in the state of their choice, even by seceding from existing states.

Opposed to this line of thinking is the school of instrumentalism, which argues that the solution of redrawing political borders on the basis of self-determination often advances neither democracy nor achieves the peaceful resolution of ethnic conflicts (India-Pakistan, Ethiopia-Eritrea, the former Yugoslavia, etc). It maintains that ethnic conflict is less about attachment to primary identity and more about competition for the control of state power. Ethnicity is how elites vying for state power mobilize people in the name of ethnic identity. Since ethnic conflict is primarily about politics rather than about culture, a political arrangement allowing decentralization and power-sharing can promote a peaceful resolution of conflicts.

Instrumentalism comes up against a major objection, which is that it views ethnicity as a product of elite manipulation. Such an understanding is unable to account for the emotional mood and violent methods that are often characteristic of ethnic conflicts. It is difficult to see why the masses follow with great fervor the discourse of elites unless it awakens their own deep affective longings.

In an attempt to correct instrumentalism, the school known as constructivism underlines that, rather than reviving already existing primary attachments, the ethnic discourse invents new identities. It argues that mistreatments and the need of liberation drive marginalized elites to imagine communities embellished with thrilling characteristics, thereby successfully mobilizing the people with whom they identify. The promise of deliverance activates affective components that impart an emotional dimension to what is but an invented identity.

Sustained reflections on Ethiopia’s ethnic conflicts lead me to believe that the constructivist correction of instrumentalism does not fully answer the question of knowing why the invented discourse of elites moves the masses to the point of violent confrontations. True, the element of imagination is liable to arouse emotional forces, but there is also no denying that the ethnic discourse works with past materials associated with common descent and cultural legacy to which people are naturally attached. What is achieved is thus the creation of ascriptive rights with exclusionary intent, which largely involve sentiments derived from nature rather than merely from human imaginative capacity. I also question the idea that constructivism constitutes a distinct school, all the more so as it loses much of its explanatory force if a great dose of instrumentalism does not support it.

Instead of setting apart, I propose to fuse instrumentalism with constructivism if only because such an attempt seems to recover whatever is valid in primordialism. Indeed, what is the most effective way of promoting interests if not through the mobilization of affective and cultural forces, especially when said interests are challenged or denied? Accordingly, ethnic mobilizations are better understood if cultural construction is itself an instrument whose purpose is to optimize a political claim. Such an approach retains the powerful role of culture without, however, losing sight of the material component of ethnicity. While I admit that the emotional force of ethnicity cannot be explained without appealing to primordial impulses, I argue that the impulses do not provide the inspiration; rather, they are used to maximize definite interests.

This approach insists that ethnicization is more than a mere protest against mistreatment. Indeed, had ethnicity been about the equal recognition of rights, mobilization around individual rights, as prescribed by liberal democracy, would have been the appropriate response. On the other hand, if the fight is over the control of the state, then the strategy is to mobilize group rights so as to use ascriptive characteristics (common descent, language, culture, etc.) to exclude political rivals as aliens. The use of ethnic criteria thus maps out constituencies that function as a reserved power base for vying elites.

Identity politics is all the more mobilizing when ruling elites are made responsible for economic plights of ordinary people. What is common in ethnic discourses is the framing of culprits with the consequence that it unleashes anger. The revival of traditional identities, in addition to portraying elites as saviors of their community, thereby upgrading their authority, frames social relations in terms of culprits and victims. Just as the Marxist concept of class exploitation politicized poverty, so too the ethnic discourse politicizes identities by portraying the possession of some characteristics (language, descent, religious beliefs) as reasons for mistreatment. In so doing, it stirs up anger that it directs against those who hold power.

On top of deriving the emotional component from the construction of imagined communities, my approach thus adds the important factor of the vilification of ruling elites, which often results in the them/us dichotomy with high normative overtones. The use of moral qualifications turns the confrontation between ethnic groups into a struggle between the good and the bad, the virtuous and the vicious. This moral classification is then used to justify the resort to violent means.

To understand the wide impact of ethnic discourse, one must go beyond the negative role of inciting anger by adding its restorative function. Discriminatory treatment as a result of the hegemony of one ethnic group has a deep impact on the self-representation of dominated or marginalized groups, since it activates feelings associated with humiliation. This explains why ethnicity is so violent when compared to class conflict, which is mostly about justice and fair distribution, and not about human pride. Not only does the ethnic construction highlight humiliation, but it also proposes a curative solution in the form of self-determination or self-rule. While the solution supports the political ambition of elite groups, it is also largely accepted as a necessary step toward the removal of humiliation. According to the logic of ethnicization, pride is restored only when governments by non-kindred people, however democratic they may claim to be, are replaced by governments of kindred-people.

The significance of my hypothesis transpires as soon as one asks what specific ideas it contributes to the paramount issue of the peaceful resolution of conflicts. The importance of having the correct approach is that it enables us to find relevant solutions: if we know what causes ethnic conflicts, then we can devise institutional mechanisms that remove the causes and, therefore, ease ethnic tensions.

The primordialist approach has no other option than the secessionist solution, since it reduces ethnic conflicts to cultural incompatibilities. The instrumentalist approach has the merit of deriving ethnic conflicts from elite rivalries for the ownership of the state. In agreement with instrumentalism, my approach suggests that the main solution to ethnic conflicts is to open up the power game by devising institutions that decentralize power, as in the case of federal arrangement with large autonomy. Nevertheless, my analysis of the cultural dimension as a maximizing factor argues that autonomy should go to the extent of allowing the implementation of group rights and self-rule. I thus take into consideration the powerful emotional forces unleashed by the ethnic discourse. Unless these forces are appeased, a mere decentralization will not be enough.

In addition, my view, which can be termed “maximism,” suggests the possibility of displacement (in the Freudian sense of the word). One way of reducing tensions would thus be to shift the emotional forces to trans-ethnic or multiethnic institutions and symbols. My assumption is that multiethnic institutions can supersede ethnic exclusiveness if access to higher levels of national government represents, not the surrender of ethnic identity, but its graduation from local to national statures. Such institutions together with the celebration of diversity will help cultural conversion to multiethnicism as an imagined community.

My thesis, namely, ethnicity as a maximizing factor in elites’ struggle for the control of power, finds a perfect confirmation in both the origin of ethnic conflicts in Ethiopia and Ethiopia’s experiment with ethnic federalism. A strong argument for this would be the fact that the Ethiopian system, besides being imposed, is deliberately established to encourage ethnicization. Whereas other countries, such as Nigeria, India, etc., used federalism as a devise to dilute ethnicity so as to safeguard national unity, all the practices and constitutional provisions in Ethiopia tend to strengthen ethnic identity to the detriment of national integration.

The explanation springs to mind: both to mobilize the Tigrean people so as to overthrow the dominance of the Amhara elite and to establish a federal system that favors it, the TPLF had to fracture Ethiopia along ethnic lines, thereby speaking of the country as an ensemble of nations and nationalities. So fractured, the political struggle becomes focused on self-rule and the control of regional states, leaving the federal government to the TPLF. Such a system develops local elite groups that have common interests with the ruling power without, however, making them competitors.

Scholars who study the Ethiopian case marvel about the radical nature of ethnic federalism, but they also observe shortcomings. They thus underline a disparity between theory and practice, especially when it comes to the autonomy of ethnic regions. This disparity proves that the wrong understanding of ethnicity actually inspires those who speak of shortcomings. A consistent and comprehensive view of the discrepancy is achieved only when it is admitted that ethnicity is less about democracy than it is about the control of state by elite groups.

The primordialist position is completely unable to explain the disparity between practice and theory. If primordial sentiments exclusively motivate ethnicity, then the victory of the TPLF should have led to the secession of Tigray or the implementation of a real system of decentralization and self-rule. What is more, the TPLF wholeheartedly supported the Eritrean independence on the basis of primordialist criteria, but refuses to recognize the claim of secessionist movements in the regions of Oromia and Somalia. These apparent contradictions vanish if it is shown that calculations of interests condition the TPLF’s decisions.

The involvement of interests becomes manifest when we remark that, though the Ethiopian system encourages ethnicization, it remains very centralized. The centralization is realized through a party system, the EPRDF (Ethiopian People's Revolutionary Democratic Front), which is a coalition of ethnic parties in which the TPLF is the dominant partner. Thanks to the democratic centralism governing the coalition, the TPLF thus controls the whole federal system and intervenes extensively in the administration of regional and sub-regional governments. What comes to mind is the Soviet model of federalism based on the tight control of the communist party.

What this means is that regional autonomy is not how a region is allowed to decide and control its affairs; rather, the system creates client parties that allow the center to maintain its controls through dependent local elites. That is why, as I wrote in a previously published article titled “The Underside of the Eritrean Issue,” it is perfectly sound to state that the TPLF politely but firmly expelled Eritrea from Ethiopia because it understood that the EPLF will never agree to become a dependent partner. The system and the way it works make sense only if we assume that it is purposely designed to maintain the hegemony of an elite group claiming to defend the interests of a minority ethnic group.

The presence of interests in ethnic claims is also attested by the fact that there is no shortage of elite-groups seeking to become clients. To the imposition of ethnicity as a primary criterion of federal arrangement, local elites responded by creating political movements that endorsed the criterion. So that, ethnic identities that used to be weak are restructured as primary for the simple reason that the TPLF-dominated federal government rewards ethnicization.

Be it noted that instrumentalism cannot explain the ethnicization of Tigray without interpreting ethnicity as an imaginative reinvention of identity. Though Tigray has been part of Ethiopia (Abyssinia) for at least 3000 years and Tigreans and Amhara–– the dominant ethnic group during Haile Selassie’s long reign–– share the same culture and political system, the TPLF constructed Tigray as a nation by emphasizing language difference. While this reinvention supports constructivism, a complete view is achieved only if it is inserted into my interpretation of identity politics as a maximizing factor.

The use of ethnic criteria to reinforce a political goal is what explains the deep contradiction of ethnic movements in Ethiopia. Whether we take the Eritrean, Oromo, Tigrean, or other ethnic movements, all trace their emergence back to the imperial regime, which they defined as the imposition of Amhara culture and interests in a tightly centralized political system. The democratic solution would have been decentralization together with the recognition of Ethiopia as a multiethnic country. Ethnic movements did not opt for such a solution; instead, they brandished self-rule and group rights. The definition of ethnic groups as nations and nationalities means that they revert back to the nation-state model that they had previously rejected in the name of multiethnicism. Only the goal of capturing state power by amplifying cultural incompatibilities can explain the reversal.

The factor of maximization becomes fully manifest when we notice the rise of dissident ethnic parties accusing the TPLF of not being consistent. Such movements are often secessionist and they become so by stretching cultural disparity, that is, by adopting an even more primordialist language. Dissident ethnic parties cannot hope to compete successfully against client elites working with the TPLF unless they change identity into a primordial commitment overriding everything. In particular, the works of intellectuals of Oromo origin clearly show how they combine vilification and utopia to create the “Oromo” nation. The vilification inherent in the thesis of Abyssinia’s colonization of Oromia and the myth of democratic Oromia before the colonization both testify to the invention of Oromia as an imagined community.

Since democratic rules guaranteeing the proper application of federalism are not followed in Ethiopia, ethnic federalism, as it is now implemented, only succeeds in radicalizing and multiplying dissident ethnic groups. As a result, there is a growing danger of disintegration that will lead to violent confrontations, not only inside Ethiopia, but also in the entire Horn of Africa, unless a reverse process toward reintegration is put in place. In other words, what puts the country in danger is less ethnicity than the lack of democratic governance, itself originating from an eccentric group’s shortsighted and vain goal of preserving indefinitely the control of power.

The theory of maximization and its attendant, namely, the possibility of displacement, suggest a way out through the creation of national symbols and institutions encouraging ethnic cooperation. In other words, the crystallization of ethnic identity could be diluted if national offices are made dependent on moderation. The lure of higher political rewards through moderation could thus produce a displacement mitigating the exclusionary practice of identity politics.

This means, of course, that the main solution to ethnic conflicts is the democratization of the state through decentralization and large local autonomy. However, I emphasize that the autonomy must go to the extent of allowing the implementation of group rights and self-rule, the only way by which the affective element can be dealt with. Since in denouncing alien rule, the ethnic discourse has awakened the feeling of humiliation, only the provision of a local or regional administration controlled by culturally kindred elites can satisfy both the masses and the competing elites.

My thesis also predicts that, as soon as grudge is removed through the granting of self-rule, ethnic groups will lose their original compactness and give way to diversity and the appearance of sub-group elites vying for the control of local power. In due course, this will reintroduce issues of individual rights that will be useful both to democratize the local community and to rebuild the national unity.

My solution is then to open up the power game in conjunction with the creation of institutional mechanisms that work toward unity. The tendency to unity should grow from the political dispersion, that is, from the implementation of group rights, itself leading to intra-ethnic rivalries. From this diffused power game must rise national ambition forcing elite groups to moderate their views if they want to extend their power and influence beyond their ethnic groups. Moderation as a prerequisite to national leadership can also be used to prevail over local rivals.

Appropriate institutional mechanisms can further fortify the appeal of moderation. So that, the peaceful and lasting solution to ethnic conflicts seems to be the device of a political system in which centripetal forces (national institutions and symbols) counter centrifugal forces (ethnicity). While federalism with large autonomy and self-rule should satisfy ethnicity, political institutions making national positions dependent on moderation should encourage unity. As much as I support the political recognition of ethnicity, unlike primordialism, I think that the reconstruction of unity is also necessary for a lasting peace.

One way of balancing centripetal and centrifugal forces is the creation of a presidential figure with large political and symbolic meanings. If the election of the president depends on majority vote of the people, in addition to encouraging the expression of individual rights in conjunction with group rights, candidates for the presidential office will have to become attractive to voters outside their ethnic groups. This arrangement encourages moderation, but also creates national figures.

My theory of ethnic management finds a validating argument in the proposal that it is but a modernized version of the political system of traditional Ethiopia. Seeing the long duration of the political system, which even resisted repeated colonial assaults, it is sound to contend that the provision of an open power game based on the interplay of centrifugal and centripetal forces was the secret of the long survival of Ethiopia (for detailed explanation of the traditional system, see my book, Survival and Modernization).

Indeed, while the nobility with often ethnic definitions controlled regional power, the imperial throne and the Orthodox Church represented centripetal forces. Another crucial centripetal force was the active role of the national intelligentsia (debtera), which was the product of a common system of education whose pillars were use of the Geez language, the centering of Ethiopia, and the propagation of its divine mission (the Kibre Negast).The system defined the emperor as king of kings: the recognition of regional leaders as kings meant nothing less than the acceptance of large autonomy and self-rule. That Tigray preserved its language and ruling elites for centuries even though the Amhara were numerically superior and often in control of the imperial throne proves how extensive was the autonomy that regions enjoyed.

What is more, regional lords could freely compete for the imperial throne, since the system did not institute any exclusive definition of the heir to the throne, except for the vague and inclusive concept of Solomonic descent. Decentralization and competition for the imperial throne encouraged intra-ethnic competitions resulting in the emergence of rival sub-regions in Amhara and Tigray. These conditions never allowed the crystallization of ethnic identity; instead, they enabled the emperor to emerge as a divine-elected protector of Orthodox Christianity and unifier of a multiethnic community. In other words, political dispersion or regional autonomy was coined as a source of rivalry setting the stage for the intervention of God’s express choice of the emperor. Often based on military prowess, God’s choice became formal the moment the Church anointed the elect.

The working principle required not only the respect of large local autonomy with self-rule, but also that the various regions of Gondar, Gojjam, Wollo, Shoa, and Tigray had comparable powers. Witness: when the central system collapsed during the Era of the Princes, no one was really able to prevail until the rise of Tewodros, who also failed partially. Menelik was able to triumph because the southern expansion of Showa created an imbalance that favored the Shown nobility. The loss of balance changed the political game: the political dispersion necessary to set God’s choice in motion was replaced by entitlement derived from the Shoan hegemony.

The unrivalled power of Show cleared the way for the establishment of Haile Selassie’s autocratic rule and his hereditary monarchy. In the name of modern nation-building, Haile Selassie put an end to the decentered power game through a tight political centralization and Amharization that naturally favored the Amhara nobility. Its outcome was the slow but steady exasperation of ethnic conflicts through the instigation of elites from marginalized ethnic groups.

The traditional system teaches us that wisdom lies in creating regional units that are balanced, but also open to intra-group competitions. The latter together with centripetal institutions and symbols prevent the crystallization of ethnic identity to the benefit of multiethnicism. The shift results from the open power game that defines national positions as graduations of ethnic identities to trans-ethnic representations.

The present policy of the TPLF prevents the emergence of national ambitions and intra-ethnic group competitions by the method of democratic centralism, which protects client parties from competition. Moreover, the principle of balanced power does not command the establishment of federal units. In particular, the two big regions of Amhara and Oromia create a serious imbalance endangering national unity. Wisdom advises the fracturing of these two regions into smaller units as a necessary condition of promoting ethnic cooperation.

What we learn from the traditional system is thus the recapture of the culture tolerating diversity, which culture was sidelined by the uprooting imitation of Europe’s model of the nation-state. The expression “Amhara or Tigrean hegemony” would be incomprehensible to the people of traditional Ethiopia who understood ethnicity in terms of rivalry, and not as a system of hegemonic government. The other important lesson is the need to couple ethnicity with centripetal institutions and visions, whose outcome is the promotion of multiethnicism. A strong presidential figure who would be elected on the basis of majority vote among all ethnic groups would be to the modern system what the emperor was to the traditional polity.

June 9, 2009