TETEGETTY.com

The Tonga-Tsonga question is this: how did one ancestral speech community — the Tonga of Nyumbane and Nkomati in the coastal zone of what is now southern Mozambique — produce five distinct language varieties that today are different enough that speakers of some varieties cannot understand speakers of others? And what role did the Zimbabwe Plateau Civilisation — specifically its Mbire language — play in shaping the divergence?

The five varieties are Gitonga of Inhambane Province (Mozambique), Xitsonga spoken primarily in South Africa, Chitonga of Zimbabwe, Chitonga of Zambia, and Chitonga of Malawi. They share a common ancestry but have diverged through different contact histories, different migration routes, and different political circumstances. Each tells a different story of how one original community’s language was shaped by the worlds it encountered.

The first critical clarification: the Chitonga varieties of Zimbabwe, Zambia, and Malawi — the Zambezi and interior plateau-fringe communities we began to examine in Volume 13 — are related to Gitonga and Xitsonga at a considerable depth of ancestral time. They are family, but distant family. The more immediately related pair — the ones whose divergence from each other is historically traceable and documentable — is Gitonga and Xitsonga. These two are the direct focus of this volume. The Chitonga varieties of Zimbabwe, Zambia, and Malawi received their own dedicated treatment in Volume 13 and will be referenced here only where their history intersects with the coastal Tonga story.

The ancestral community from which both Gitonga and Xitsonga descended was located in the Nyumbane and Nkomati coastal zone of southern Mozambique — a complex estuary and coastal plain environment where the Nkomati River meets the Indian Ocean. This was a linguistically rich environment long before European contact: a zone where coastal Bantu speech communities met Swahili trade networks from the north, where inland plateau-connected communities arrived from the west, and where the Indian Ocean brought Arabic and Indian merchant vocabulary from the east.

The original Tonga of this coastal zone spoke a Bantu language whose base grammar is preserved most clearly in Gitonga — the variety that stayed closest to the original coastal territory. The language had a noun class system, a verb extension architecture, and core vocabulary that are traceable to the same broad eastern Bantu stream as the plateau languages. But this was not a plateau language. It was a coastal Bantu language with its own distinct identity, shaped by its estuary and coastal ecology in ways that parallel how the Zambezi shaped the interior Tonga of Volume 13.

The original Tonga of Nyumbane were organised into clan groups — the Beja, Dziva, Nhlaruti, Sono, and Makunyule being among the most significant. These clans formed the social backbone of the ancestral speech community whose language would diverge into Gitonga and Xitsonga. From the same clans, in the same geographic zone, two trajectories emerged: those who remained in or near the Nyumbane coastal zone and whose language became Gitonga, absorbing Mbire, Nambya, Swahili, Arabic, and Portuguese; and those who moved southward and westward, whose language became Xitsonga, absorbing Nguni, Afrikaans, and English.

The divergence was not immediate. For generations the two streams were in contact, close enough that a single speech community covered the range. The separation deepened over centuries as the contact worlds of the two streams pulled them in increasingly different directions. By the time the Nguni impact of the early 19th century reshaped southern Mozambique and South Africa’s eastern lowveld, the two varieties were already distinct enough to be recognisably different — even if speakers at the border could still communicate.

The most significant chapter in the Gitonga story — and the one that connects it directly to this series — is the Mbire vocabulary layer that entered Gitonga through the movement of Mbire-speaking groups into the Nyumbane coastal zone. This is not a distant or theoretical connection. It is specific, documentable, and traceable in the vocabulary of Gitonga today.

The Gwamba Thobela Mbire people — who moved into the Nyumbane area through the plateau-Zambezi corridor that Nambya territory occupies (Volume 9) — brought Mbire vocabulary into sustained contact with the coastal Tonga speech community. This was not a brief encounter. It was a settlement — a sustained presence of Mbire-speaking people in the coastal Tonga zone that lasted long enough to permanently reshape Gitonga’s vocabulary in specific domains.

The Nambya of Nyai also entered this coastal story. In the movement of plateau-connected speech communities toward the Mozambique coast, the Nambya were intermediaries — carrying plateau vocabulary eastward through the Zambezi corridor. Their presence in the Nyumbane story is documented in the oral traditions of the coastal communities themselves, which name the Nambya among the groups who arrived and reshaped the local language.

The Nyumbane area was a zone of successive arrivals and overlapping communities. The Mbire people led by Gwamba of Tovera — also called Thobela — moved in from the plateau direction. The Nambya of Nyai arrived separately. Swahili and Indian-Arab traders maintained their coastal network through the zone. The Portuguese under Vasco da Gama established contact from 1498 onwards. Each of these arrivals left vocabulary in the existing Tonga speech community — and the mixing of Tonga grammar with Mbire vocabulary, Nambya vocabulary, Swahili, Arabic, and Portuguese produced the specific language we now call Gitonga.

The mechanics of this mixing are precisely those described in Volume 2 of this series when we examined language divergence generally. The Tonga grammar — the noun class system, the verb extension architecture, the basic sentence structure — remained the skeleton. The vocabulary from each contact source filled in the flesh. Gitonga is therefore Tonga grammar speaking Mbire, Nambya, Swahili, Arabic, and Portuguese vocabulary simultaneously — a multilingual contact product that is none of those languages individually but carries all of them inside it.

The most powerful evidence for the Mbire-Gitonga connection is the traceable word trail — specific vocabulary items that can be followed from their Mbire origin on the plateau through the Nambya corridor to their arrival in Gitonga on the coast. This is not linguistic reconstruction at a distance. It is a step-by-step vocabulary journey, readable in the evidence of living languages today.

By the time the divergence between Gitonga and Xitsonga was complete, the two had absorbed such different contact vocabularies that speakers of one variety cannot always hold a sustained conversation with speakers of the other. This mutual unintelligibility is not the failure of a language. It is the evidence of different histories — two communities that began speaking the same language and then lived through different centuries in different directions.

A Gitonga speaker from Inhambane and a Xitsonga speaker from Maputo share: the same ancestral Tonga grammar structure, the same clan origins in Nyumbane, and some core vocabulary preserved from before the divergence. What they do not share: the Mbire vocabulary layer that entered Gitonga but not Xitsonga (or not with the same depth), the Nguni vocabulary that entered Xitsonga heavily through the Gaza Nguni state but barely touched Gitonga, and the different Swahili and Portuguese forms that each variety absorbed through their distinct coastal positions.

The comparison in everyday words is telling. Gitonga says kuhuma duva for “the sun comes out.” Xitsonga says kuhuma dyambu. The verb root is shared. The sun word is different. And that difference — duva versus dyambu — encodes the entire history of the divergence: duva is the Mbire root that entered Gitonga through the plateau corridor; dyambu is the form that emerged in Xitsonga after its separate path of contact and change.

The claim that “Tsonga is not a language” requires precise unpacking — because it can be misunderstood as a denial of these communities’ linguistic identity. It is the opposite. It is an insistence that each variety deserves its own proper name rather than being collapsed into a generic label.

“Tsonga” as a linguistic category is a colonial administrative convenience — an umbrella term applied by outsiders to a group of related but distinct speech varieties whose internal differences the administrators either did not notice or did not care to preserve. The same logic produced “Shona” for the plateau languages in 1931. The same logic produced “Sotho” as a cover term for Sesotho, Sepedi, and Setswana. In each case, administrative consolidation erased the specific identities of distinct communities by grouping them under a single name.

What should we say instead? Each variety has its own name: Gitonga for the Inhambane variety, Xitsonga for the South African variety. The Chitonga varieties of Zimbabwe, Zambia, and Malawi each have their own names and their own distinct identities. When we discuss the whole group, we say “the Tonga-Tsonga language family” — acknowledging the family relationship while refusing to collapse the members into a single false identity.

This naming is not pedantry. It is the foundation of accurate history. A Gitonga speaker from Inhambane and a Xitsonga speaker from Limpopo Province have different ancestral contact worlds, different oral traditions, different vocabulary for their different coastal and lowveld environments. To call them both “Tsonga” is to erase those differences and flatten two distinct histories into a single convenient label. This series refuses that erasure.

The Tonga-Tsonga story is this series in miniature. One ancestral community, one original language, multiple centuries of movement and contact, five distinct voices by the time history catches up. Each voice is real. Each voice deserves its own name. Each voice carries a specific history that the generic label “Tsonga” cannot contain.

For this series specifically, the Tonga-Tsonga story is the fullest demonstration of the plateau’s language reaching the Indian Ocean. We traced that reach from the Mbire root in Volume 3, through the Mutapa trade network in Volume 10, through the Nambya corridor in Volume 9, and now we have arrived at the coast — in the vocabulary of Gitonga, where duva for the sun and guthuma for work preserve the plateau’s speech in a coastal language that developed thousands of kilometres from Great Zimbabwe.

The plateau’s language did not stop at the escarpment. It did not stop at the Zambezi. It did not stop at the Mozambique border. It reached the Indian Ocean — in the mouths of Gitonga speakers whose language carries the Mbire root in its vocabulary, mixed with Nambya, Swahili, Arabic, and Portuguese, all sitting on top of a Tonga grammar base that is itself ancient Bantu. That is the full story. That is the civilisational reach of the Zimbabwe Plateau Civilisation’s language.

The confusion is resolved. The five varieties are named. The plateau connection is traced. The colonial category is dismantled. The languages are restored to their own names and their own histories.

Gitonga ndiyo Gitonga. Xitsonga ndiyo Xitsonga. Vanhu ava vane nhoroondo dzavo. Each is itself. Each people has its own history.