In 1931, a linguist named Clement Martyn Doke sat down in Salisbury, Rhodesia, and produced a report that would rename five distinct languages — each with its own history, its own oral tradition, its own civilisational depth — as dialects of a single administrative category he called “Shona.” The word “Shona” itself did not come from the people whose languages he was reorganising. It was, at best, a vague external label of uncertain and possibly derogatory origin. None of this deterred him. In a single administrative act, a thousand years of distinct language identities were erased in the colonial record. This volume is the story of that erasure — and of why it was wrong.
Before the Doke Commission of 1931 invented the category “Shona,” the plateau language family had been speaking under its own names for centuries. Chikaranga — the language of Great Zimbabwe and the Rozvi Empire — had been recorded by Portuguese observers in the 16th century under that name. Korekore, Zezuru, Manyika, and Ndau all had their own community identities, their own oral literary traditions, their own histories of political and cultural distinction.
These were not five varieties of one language that happened to be called different names by their speakers. They were five distinct language varieties — related as siblings within the Mbire family, mutually intelligible to varying degrees, but each carrying its own specific historical identity, its own contact history, and its own community’s sense of self. Chikaranga speakers knew they were Chikaranga. Ndau speakers knew they were Ndau. The distinctions were real to the people who lived them.
What the colonial period changed was the administrative recognition of those distinctions. Southern Rhodesia’s government needed a practical framework for administering the majority African population. Multiple distinct languages meant multiple sets of literacy materials, multiple administrative translation requirements, multiple language education tracks. A single “Shona” umbrella was administratively convenient — far more convenient than the historical reality.
The word “Shona” itself has uncertain and contested origins. The most widely accepted view is that it was an external label — possibly derogatory in origin — applied by outsiders to the plateau communities. What is clear is that it was not the plateau communities’ own name for themselves or their languages. Chikaranga speakers called their language Chikaranga. Korekore speakers called their language Chikorekore. No plateau community called their language “Shona” before the colonial administration made it official. The category was imposed from outside and above — not generated from within.
In 1929, the Southern Rhodesian government commissioned Clement Martyn Doke — a South African linguist best known for his work on the Bantu languages of South Africa — to conduct a survey of the “Shona dialects” of Southern Rhodesia and recommend a standardised written form that could be used in schools and mission publications. The commission’s terms of reference already contained the conclusion: the languages were called “Shona dialects” before Doke arrived. He was hired to systematise what had already been decided administratively.
Doke spent several months in Southern Rhodesia in 1929–1930, visiting communities and consulting with missionaries who had been working on the plateau languages for decades. His 1931 report — “A Preliminary Survey of the Bantu Languages and Dialects Spoken in Southern Rhodesia” — recommended the creation of a unified literary Shona standard, drawing primarily from Karanga (Chikaranga), Korekore, and Zezuru as the three “main dialects” to be merged, with Manyika and Ndau as further “dialects” of the unified standard.
The report was implemented. A unified Shona orthography was developed, literacy materials were produced in the standardised form, and schools across Southern Rhodesia began teaching “Shona” as a unified language. Within a generation, children who might have identified as Chikaranga, Korekore, Zezuru, Manyika, or Ndau speakers were being taught in school that they spoke “Shona dialects.”
The creation of the “Shona” category did not merely rename five languages. It severed each language’s connection to its own specific civilisational history. When Chikaranga became “Shona” in the administrative record, it lost its identity as the language of Great Zimbabwe, the language of the Mutapa state, the language of the Rozvi Empire. The 1,000-year civilisational heritage that Chikaranga carried was flattened into a generic administrative label with no history attached.
When Korekore was subsumed into “Shona,” it lost its identity as the political language of the Mutapa state, the spiritual language of the mhondoro network, the language in which Nehanda called the northern plateau to resistance. When Ndau became a “Shona dialect,” its specific history as the eastern plateau language — shaped by a thousand years of Indian Ocean contact, by the Gaza Nguni period, by its role as the plateau’s eastern gateway — was collapsed into a homogenous category that recognised none of this specificity.
Doke’s linguistic methodology was designed to find unity, not diversity. His comparisons focused on shared vocabulary and grammar — which the plateau languages do have in abundance, as siblings in the same Mbire family. His framework did not adequately weigh the historical, cultural, and identity-based distinctiveness of each variety. A linguist looking for reasons to separate the plateau languages would have found as many as Doke found for uniting them — because the evidence for both is real.
The Spanish, Portuguese, and Italian are more mutually intelligible than some of the plateau language pairs Doke unified as “dialects.” They are not called dialects of a single language. The Scandinavian languages — Danish, Norwegian, and Swedish — are in many respects mutually intelligible and share a common ancestor far more recently than the plateau languages share their Mbire root. They are named separately and treated as distinct languages. The difference between Doke’s treatment of the plateau languages and the European treatment of these European language pairs was not linguistic. It was political.
Doke was not making a pure linguistic determination. He was making an administrative recommendation dressed in linguistic clothing. The administration had already decided it wanted a single unified platform. Doke’s methodology was calibrated to produce that outcome. The result was not linguistically wrong — the plateau languages are genuinely related — but it was historically and politically wrong to use that relatedness as a reason to erase their individual identities.
Spanish and Portuguese speakers cannot always understand each other. They are different languages. Chikaranga and Korekore speakers can usually understand each other. The colonial record called them dialects. The difference in treatment was not phonological. It was political. It was racial. It was the difference between languages that colonial power respected and languages that colonial power decided to manage.
— Tete Getty, Moyo Netombo 🇿🇼The colonial language framework, once established, proved difficult to unsettle. The “Shona” category had been embedded in education, publishing, broadcasting, and administration for fifty years by the time Southern Rhodesia became Zimbabwe. The individual language varieties continued to be spoken in homes and communities — because communities know their own names. But the official record continued to classify them as dialects rather than languages, and their specific civilisational histories continued to be absent from the formal educational narrative.
Restoration has begun, driven primarily by communities, scholars, and cultural practitioners rather than by state language policy. Chikaranga-speaking communities have asserted their distinct identity. Ndau communities have consistently resisted the “Shona” label. Manyika cultural practitioners have documented their language’s specific vocabulary and oral literary tradition. Korekore scholars have written the Mutapa state’s language back into the record.
This series is part of that restoration work. By naming each language by its own name — Chikaranga, Korekore, Zezuru, Manyika, Ndau — and by tracing each one’s distinct civilisational history, we are doing what the Doke Commission refused to do: treating the plateau languages as what they are, not as what was administratively convenient to call them.
| Language | What It Had Before 1931 | What “Shona” Erased | Status in the Record |
|---|---|---|---|
| Chikaranga | Language of Great Zimbabwe, Mutapa, and Rozvi — 1,000 years of civilisational prestige | Its specific identity as the plateau’s historical court language | Restored as Chikaranga in this series — the civilisational record reclaimed |
| Korekore | Political language of the Mutapa state; Nehanda’s language of resistance; active mhondoro tradition | Its specific Mutapa and resistance heritage; its northern plateau distinctiveness | Restored as Korekore — Volume 12 of this series |
| Zezuru | Central plateau language of the modern capital zone; its own distinct identity | Its own name — ironically, Zezuru became the de facto standard while losing its label | Named and treated as a distinct variety in this series |
| Manyika | Eastern highlands plateau language with its own trade contact history; distinct oral tradition | Its specific identity; explicitly labelled “most deviant” by colonial standards | Restored in Volume 22 of this series as a distinct language with its own history |
| Ndau | Eastern plateau gateway language; 1,000 years of Indian Ocean contact; distinct vocabulary world | Its specific history — the Nguni contact layer used to discredit it rather than understand it | Restored as Ndau in Volume 15 — its contact layers read as civilisational evidence, not corruption |
“Shona” was not the only colonial language-boxing operation in Southern Africa. The same administrative logic produced “Tsonga” — an umbrella term that collapsed Gitonga, Xitsonga, and the Chitonga varieties into a single category, which we dismantled in Volume 14. It produced “Sotho” as a cover term for the distinct Sesotho, Sepedi, and Setswana. It produced “Ndebele” as an identity frame for communities within the Ndebele state who spoke other languages. Across Southern Africa, the colonial administrative impulse to simplify, unify, and manage produced a systematic erasure of linguistic distinctiveness wherever it was administratively inconvenient.
The pattern is consistent: where multiple related languages existed, colonial administration tended to create a single administrative category and reclassify the individual languages as dialects. Where unrelated languages coexisted, colonial administration created ethnic and tribal categories that tied language to a simplified identity framework. In both cases, the result was a reduction of linguistic complexity to something manageable from an administrative distance.
What was lost in every case was the same thing: the specific civilisational history attached to each language. “Shona” has no civilisational history — it was invented in 1931. Chikaranga has a thousand years of civilisational history. The substitution of the administrative category for the historical identity was not a neutral act. It was the erasure of history in the service of administrative convenience.
This series does not use “Shona” as a language name. It uses the names the languages have always had: Chikaranga, Korekore, Zezuru, Manyika, Ndau. It recognises that these are related members of the same Mbire-descended plateau language family — but it refuses to collapse their individual identities into an administrative category invented in 1931 by a colonial government that had no interest in the civilisational histories those individual names carry. The Doke Commission created “Shona.” History created Chikaranga. This series stands with history.
Despite nearly a century of official “Shona” categorisation, the individual language identities continue to be maintained by their communities — in speech, in oral tradition, in identity assertion, and increasingly in scholarship and cultural practice.
“Shona” was built as a container. A convenient box in which to store the plateau’s language complexity so it would not complicate colonial administration. The box was administratively functional for almost a century. But the languages inside it never stopped knowing their own names. Chikaranga knew it had spoken at Great Zimbabwe. Korekore knew it had carried Nehanda’s words. Ndau knew it had been shaped by a thousand years of Indian Ocean contact. Manyika knew it had its own oral literature. Zezuru knew its own vocabulary.
The box holds administrative categories. It cannot hold civilisational memory. The civilisational memory kept speaking — in homes, in ceremonies, in the mhondoro’s sacred register, in the grandmother’s proverbs, in the dare’s formal language, in the mutupo declaration at every social introduction. The language lived outside the box while the box was being labelled and filed.
The task of this series — and of the broader work of decolonising Southern African language history — is to break the box open and restore each language to its own name, its own history, and its own civilisational depth. Not because “Shona” as a linguistic category is entirely wrong — the plateau languages are related, and the relatedness is real. But because a language name carries history. And the history that Chikaranga, Korekore, Zezuru, Manyika, and Ndau carry is a history that “Shona” cannot hold, because “Shona” has no history at all.
Chikaranga ndeyedu. Korekore ndeyedu. Zezuru ndeyedu. Manyika ndeyedu. Ndau ndeyedu. These languages are ours. They have always been ours. No commission can change that.
Leave a Reply