Much of the criticism directed at my argument rests on a basic misunderstanding of what is being claimed, and—more importantly—what is not. This is not a denial of identity, history, or multilingual reality in Ethiopia. It is a critique of how the modern state has theorized, classified, and institutionalized language and ethnicity in ways that generate conflict rather than manage diversity.
At its core, my argument is constructivist and anti-essentialist. Following Rogers Brubaker, I treat ethnolinguistic identities as categories of practice—ways people name themselves, relate to others, and make sense of the world—not as fixed substances that naturally map onto political authority or territory. People self-identify as Oromo, Amhara, Tigrayan, Somali, or Gurage. This has never been in question. What is in question is the analytical leap from self-identification to the claim that language must therefore serve as the primary unit of political design.
That leap is neither sociologically sound nor institutionally neutral.
Identity Is Real; Institutionalization Is a Choice
Ethiopia has always been multilingual. People have always spoken multiple languages, shifted registers across contexts, and navigated social life pragmatically. A trader in Addis, a student in Dire Dawa, or a migrant worker in Jimma routinely operates with a linguistic repertoire rather than a single, bounded identity. Multilingualism is not the exception—it is the social norm.
What changed was not society, but the state.
Beginning with census practices under the Derg and culminating in the 1995 constitution, the Ethiopian state froze fluid linguistic practices into rigid administrative categories and then territorialized those categories. Languages were no longer treated as tools of communication or cultural expression; they became markers of political ownership and jurisdiction. Symbolic distinctions were converted into material stakes—jobs, borders, budgets, and authority.
This is not a natural outcome of diversity. It is the result of governing diversity with a bad theory of language.
The Error of Language-as-Essence
Many critics implicitly operate with what might be called language-as-essence thinking: the belief that language naturally defines who a people are, where they belong, and what political rights they should exercise. From this perspective, territorialized, language-exclusive federalism appears self-evident, even morally necessary.
But this rests on an analytical confusion.
To say that people identify with a language does not mean that language should determine political destiny. By that logic, a multilingual Oromo living in Addis, a trilingual Oromo trader operating across regions, and a diaspora Oromo professional would all be governed as if language, territory, and sovereignty were the same thing. Social reality simply does not work this way.
The problem is not identity consciousness; it is institutional design that mistakes identity for infrastructure.
Language as Capital, Not Property
Drawing on Pierre Bourdieu, I treat language as a form of symbolic capital—a resource whose value is produced and distributed by institutions. Some languages enable mobility, access to education, and participation in national life. Others anchor local belonging and cultural continuity. Both matter. But they do not perform the same function, and treating them as interchangeable creates inequality.
When the state assigns political ownership to languages, it turns symbolic capital into inherited privilege. Those already positioned to acquire multiple languages—urban, educated, mobile populations—benefit disproportionately. Those without the resources to adapt bear the learning burden. This is why recognition without access so often reproduces class stratification under the banner of justice.
In the Ethiopian context, treating the expansion of federal working languages as inherently just—on the assumption that the state can simply absorb the costs—ignores uneven institutional capacity and educational inequality, risks shifting adaptation burdens onto the least mobile populations, and underscores why questions of scale and distribution, rather than symbolic recognition alone, must ground a repertoire-based approach to linguistic justice.
Constructed Does Not Mean Fake
Some his critics accuse me of contradiction when I note that Ethiopia has always been multilingual while also arguing that state enumeration and classification were fictive. This reveals a basic misunderstanding of constructivism. The fact that something exists socially does not mean the state’s method of classifying it is neutral, accurate, or harmless.
Money exists socially. That does not mean central banks do not construct monetary categories that shape inequality. Languages exist historically. That does not mean censuses, constitutions, and bureaucracies merely “reflect” them. They actively reorganize social life around those categories.
Constructed does not mean fake. It means institutionally stabilized.
The English Question
The critique’s response to my discussion of English further illustrates this pattern. I never claimed that English is widely spoken, socially neutral, or a substitute for Ethiopian languages. I argued that English functions as a form of institutionally non-ethnic linguistic capital—its neutrality relative, its inequalities real, and its effects requiring management rather than denial.
Responding as if I proposed English as a magic solution is either a misreading or a deliberate mischaracterization. In either case, it avoids engaging the actual argument: that some languages derive power not from ethnic ownership but from institutional positioning.
What I Am Actually Proposing
This argument is not a call for cultural erasure, forced assimilation, or naïve post-ethnic idealism. Nor is it a defense of linguistic dominance disguised as efficiency. It is a proposal to shift how we think about language politically.
I argue for a repertoire-based model in which language is understood as practice rather than possession, and as infrastructure rather than essence. Policy should manage access to linguistic capital, not police linguistic ownership. De-politicizing language does not mean de-historicizing it. Amharic’s imperial history must be acknowledged—but ethnicizing it as the property of one group only reproduces the very hierarchies critics claim to oppose.
What Ethiopia needs is not symbolic nationalism built on flags, purity, or territorialized identity, but civic nationalism grounded in shared institutions, equal access, mobility, and participation.
The Deeper Intervention
At a deeper level, my claim is this: Ethiopia’s crisis is not primarily ethnic. It is administrative, epistemological, and institutional. Society is already multilingual, adaptive, and pragmatic. The state is behind society—governing fluid realities with frozen categories.
This is why my work consistently flips the frame:
- from identity to coordination
- from culture to access
- from recognition to mobility
- from “who we are” to “how we move”
If this unsettles established positions, it is because it makes ethnic politics look less natural and more contingent. That is not denial. It is analysis.
And it is precisely what a plural, mobile, and diverse society requires if it is to be governed without turning difference into destiny.
