Before the plateau civilisations rose, before the first stone was set at Mapungubwe or Great Zimbabwe, the Zambezi River was already flowing — and people were already living alongside it, speaking alongside it, naming every pool and current and season in a language built entirely from the river’s own logic. The Tonga of the Zambezi are among the oldest continuously settled people in Southern Africa. Their language is not a branch of the plateau. It is the valley’s own voice — ancient, river-shaped, and utterly distinct. This series cannot tell the full story of Southern African language without standing on the Zambezi’s banks and listening.
The Tonga people of the Zambezi valley — speaking Chitonga of Zimbabwe and Zambia — are a Bantu-speaking community whose settlement in the Zambezi valley predates the Zimbabwe Plateau Civilisation’s political formations by centuries. Archaeological evidence places Tonga ancestors in the Zambezi valley from approximately 500 CE, coinciding with the early Bantu arrival on the plateau above. While the plateau communities were developing the Mbire speech community that would eventually build Great Zimbabwe, the valley communities below were developing their own distinct Bantu language world — shaped entirely by the river that was their world.
The Tonga are sometimes called BaTonga — the ba- prefix being the Tonga plural class prefix for people. Their language, Chitonga, is a Bantu language of the Central Bantu sub-family — related to the plateau languages at the deep level of the shared eastern Bantu expansion ancestor, but having diverged from the plateau stream very early, before the Mbire community had fully consolidated. Tonga and Chikaranga are cousins many times removed — recognisably from the same broad Bantu family, but having been separate branches for well over a thousand years.
This distinction is critical for the Tonga-Tsonga question we will address in Volume 14. The Zambezi Tonga — Chitonga of Zimbabwe and Zambia — are entirely distinct from the coastal Tonga of Mozambique who gave rise to Gitonga and Xitsonga. They share an ancestral Bantu root at a very deep remove, and they share a name that has caused enormous confusion in colonial and post-colonial linguistics. But they are not the same people, not the same language, and not the same history. This volume is dedicated to the Zambezi Tonga specifically.
To understand Chitonga of the Zambezi, you must first understand the Zambezi River as a total environment — because no other factor shaped this language more profoundly than the river itself. The Zambezi is not simply a waterway. It is an ecological system of extraordinary complexity: seasonal flooding that transforms the valley floor twice yearly, pools and channels of varying depth and current, island communities with their own micro-ecologies, fish populations of enormous diversity, woodland and grassland ecosystems along the banks, and the constant presence of the river’s wildlife — hippopotamus, crocodile, elephant, and the birds of the great river corridor.
Chitonga developed an extraordinarily rich vocabulary for this environment. It has more words for types of water, types of river current, types of fish, types of seasonal flooding, and types of riverside ecology than any plateau language possesses — because the Tonga people needed those words for daily life in ways the plateau communities, high above the valley, did not. This ecological specificity is the most immediate evidence of Chitonga’s distinct identity: a language built for a specific world that is not the plateau world.
The Zambezi also shaped Chitonga’s social vocabulary. River people organise their lives differently from plateau farmers — the seasonal flooding creates a rhythmic calendar that differs from the agricultural calendar of the highlands, the fishing economy creates specialist roles and vocabulary absent on the plateau, and the river’s role as an east-west highway created trade relationships and the vocabulary that accompanied them that are simply absent in inland plateau speech.
Imagine you grew up on the bank of the Zambezi. Your food comes largely from the river. Your calendar is organised by the river’s flood cycles. Your movement is by canoe. Your neighbours are on the opposite bank, and to visit them you navigate channels and currents. Your language would be saturated with river vocabulary that a farmer on the plateau above would have no use for. That is Chitonga — a language built from the inside of a river world, speaking its own ecological truth. No plateau language can fully translate it, because no plateau language ever needed its words.
Chitonga of the Zambezi is not spoken uniformly across its entire range. The long east-west stretch of the Zambezi valley — from the Victoria Falls region through Lake Kariba and eastward to the Mozambique border — produces a dialect continuum: a gradual variation in vocabulary and phonology across the valley’s length. Communities at the western end and communities at the eastern end speak varieties that show noticeable differences, while each adjacent pair of communities remains mutually intelligible with its neighbours.
The most linguistically remarkable feature of Chitonga is its ecological vocabulary — the vast, precise terminology for the Zambezi’s world that has no equivalent in any plateau language. This vocabulary is not merely a list of words. It is a complete conceptual system — a way of understanding and navigating a river environment that encodes thousands of years of accumulated ecological intelligence.
The relationship between the Zambezi Tonga communities and the plateau communities above them — Korekore and Nambya in particular — was one of sustained boundary contact: two distinct peoples, two distinct economies, two distinct language worlds, living in close enough proximity for exchange while maintaining clear identities through their different ecological territories.
The escarpment that separates the plateau from the valley was not merely a physical barrier. It was a cultural and linguistic boundary marker — the edge of one world and the beginning of another. Plateau communities came down the escarpment to fish in the Zambezi and to trade; valley communities came up to the plateau edge to acquire plateau goods — iron tools, pottery, grain — that the valley’s ecology did not produce. This trade created a contact vocabulary in both directions: plateau words for trade goods entering Chitonga, Tonga ecological vocabulary entering Korekore and Nambya.
The Tonga were never politically subordinate to the plateau in the way that tributary communities to the Mutapa state were. The valley’s ecological challenges — tsetse fly, extreme heat, Zambezi flooding — meant that plateau political authority could not easily be enforced in the valley. The Tonga maintained their political autonomy throughout the plateau civilisation’s period of dominance, and this autonomy is reflected in their language: Chitonga never became a dialect of a plateau language, never absorbed plateau governance vocabulary as a prestige layer, and never lost its river-world identity to plateau cultural influence.
The most catastrophic event in the modern history of the Zambezi Tonga was the construction of the Kariba Dam between 1956 and 1959. The dam flooded the Gwembe Tonga valley — the heartland of the Zambezi Tonga civilisation — and forcibly displaced over 57,000 people from their ancestral territories. The flooding destroyed not only homes and agricultural land but sacred sites, ancestral graves, and the specific landscape features around which the Tonga ecological vocabulary had been built over centuries.
This was a language catastrophe as well as a human one. The vocabulary for the submerged valley — the names of pools, channels, seasonal islands, fish gathering sites, sacred groves — lost its referents overnight. The landscape that the language described was placed under 40 metres of water. The Tonga communities were relocated to unfamiliar environments — higher, drier, without the river ecology their language had been built to navigate. A vocabulary shaped over a thousand years was suddenly speaking about a world that no longer existed in the same form.
The Kariba displacement is one of the most severe acts of linguistic erasure in Southern African history — not through language policy, but through the physical destruction of the landscape a language described. It stands as a warning that language and land are not separate. When you destroy a people’s landscape, you attack their language at its foundations.
The Kariba Dam did not merely flood a valley. It submerged a vocabulary. Every place name the Zambezi Tonga had given to the Gwembe valley — every term for every pool, channel, and seasonal feature of their world — was put under water. This series names that as what it was: an act of cultural and linguistic violence, carried out by a colonial administration that understood the landscape as a resource to be exploited and never thought to ask what the people who lived there called it.
— Tete Getty, Moyo Netombo 🇿🇼| Concept | Chitonga (Zambezi) | Chikaranga (Plateau) | Language Layer / Relationship |
|---|---|---|---|
| Person / people | muntu / bantu | munhu / vanhu | Proto-Bantu root *-ntu shared — diverged in prefix and phonology but identically descended |
| Water | mabwe / maanza | mvura | Different roots — Tonga water vocabulary shaped by river-world specificity; plateau mvura is general rain/water |
| River / the Zambezi | Mujila / Muyobo | rwizi | The river named specifically in Tonga with a personality and identity; plateau has generic rwizi for any river |
| Flood / rising water | chifukuto / musindi | — | Distinct Tonga flood vocabulary — no equivalent in plateau languages that do not experience Zambezi-scale seasonal flooding |
| Canoe / boat | bwato | bwato | Shared trade word — plateau Korekore acquired bwato through valley contact; one of the clearest plateau-Tonga borrowings |
| Fish (general) | nsomba | hove | Different roots — Tonga fish vocabulary is primary and elaborated; plateau hove is secondary and limited |
| Ancestor / spirit | mizimu | midzimu | Near-identical — deep Bantu root for ancestor spirits, preserved across the plateau-valley boundary with minimal change |
| Land / territory | nyika | nyika | Identical — the ancient proto-Bantu word for one’s land preserved in both valley and plateau speech |
| Chief / headman | sikatongo / mwami | mambo / sabhuku | Distinct political title vocabulary — the Tonga political system was not organised under the mambo structure; their own titles reflect their independent political tradition |
| Hippo | mvuu | mvuu | Shared — one of the most widely distributed Bantu words for hippopotamus; confirms ancient shared ancestry despite the languages’ long divergence |
Before we close this volume we must address the naming confusion that has generated significant scholarly and popular misunderstanding — the use of “Tonga” as a label for both the Zambezi valley communities examined here and the coastal Mozambique communities whose language gave rise to Gitonga and Xitsonga. This confusion is not innocent. It has consequences for how both communities understand and are understood in their own histories.
The Zambezi Tonga — Chitonga of Zimbabwe and Zambia — and the coastal Tonga of Mozambique (from whom Gitonga and Xitsonga descended) are not the same people. They do not speak the same language. They have different histories, different ancestries, and different contact worlds. The only things they share are a similar name and a very distant common Bantu ancestry that they share with hundreds of other language communities across Africa.
How did they get the same name? The honest answer is that “Tonga” appears to be an ancient Bantu root used in various regions to describe people who live alongside large bodies of water — whether rivers, lakes, or coasts. It is a descriptive term that was applied independently by different people in different places, not a marker of shared ethnic or linguistic identity. Colonial linguistics compounded the confusion by treating the shared name as evidence of shared origin — a methodological error that this series corrects.
Colonial ethnography and linguistics consistently collapsed the Zambezi Tonga and the coastal Mozambique Tonga into a single confused category — treating the shared name as evidence of shared identity. This produced cascading errors in how both communities were classified, administered, and understood. The Zambezi Tonga — the BaTonga of the Gwembe valley and Zambia’s Southern Province — are a distinct people with a distinct language rooted in the Zambezi valley, whose history is their own and whose identity does not require the coastal Tonga for context. Volume 14 will map the coastal Tonga story in full. They are a different story. Both stories deserve to be told on their own terms.
Despite the Kariba displacement and over a century of colonial and post-colonial marginalisation, the Zambezi Tonga maintain their language — carrying the river’s vocabulary in the mouths of communities now living partly above the waters that submerged their world.
The Zambezi Tonga remind us of something this series has insisted upon since Volume 1: the Zimbabwe Plateau Civilisation did not exist in isolation. It was surrounded by, connected to, and enriched by other language communities whose histories are not subordinate chapters in the plateau’s story but parallel narratives with their own depth and their own civilisational logic.
The Tonga of the Zambezi were here before the plateau states rose. They will be speaking when the colonial categories that tried to erase them are forgotten. Their language carries in it a river world of extraordinary ecological precision — a vocabulary for the Zambezi’s life that no plateau language ever needed to develop, and that the world would be poorer without.
The Kariba Dam submerged a valley. It did not submerge the language. The Tonga carried it up the escarpment and into the resettlement communities and into their children’s mouths. The river was dammed. The words were not.
In Volume 14 we turn to the coastal Tonga — the communities of Mozambique’s Nyumbane and Nkomati zones from whom Gitonga and Xitsonga descended. That is a different river, a different coast, a different contact world. But it is a connected story — and the thread that connects it to the plateau runs through the Mbire vocabulary that reached the coast through the corridors we have been tracing since Volume 3. The full Tonga-Tsonga question is finally ready to be answered.
Mujila inotsvedza asi inotaura. The Zambezi flows on, and it is still speaking.
Leave a Reply