A Question That Will Not Settle
Few topics in East African historiography have generated as much sustained disagreement as the question of Mijikenda origins. Did the nine Mijikenda peoples — Giriama, Duruma, Digo, Chonyi, Kambe, Rabai, Ribe, Jibana, and Kauma — migrate south from a shared homeland called Singwaya in present-day southern Somalia? Or did they evolve largely in place, in the coastal hinterland of Kenya, and adopt the Singwaya narrative later as a tool of collective identity? The debate has drawn on oral traditions, linguistic reconstruction, archaeological excavation, and colonial archival research, and after more than fifty years of scholarship, it remains unresolved. What follows is not a verdict but a map of the arguments — a guide to who has said what, on what evidence, and why it matters.
Thomas T. Spear and the Orthodox Position
The most influential statement of the orthodox view — that the Singwaya migration is substantially historical — came from Thomas T. Spear in his 1978 monograph The Kaya Complex, published by the Kenya Literature Bureau. Spear synthesised oral traditions collected across the Mijikenda, Pokomo, Swahili, Taita, and Segeju peoples, arguing that while some earlier interpretations of the Singwaya tradition could not be sustained, the tradition itself is "nevertheless valid for the Mijikenda, Pokomo, Swahili, Taita, and Segeju, where such evidence supports its basic veracity." For Spear, the convergence of multiple independent traditions pointing to the same ancestral location constituted strong evidence. If five or six unrelated peoples all say they came from the same place, the simplest explanation is that they did.
Spear's work remains the standard reference, but it has not gone unchallenged. Subsequent scholars have questioned both his methodology — particularly his handling of linguistic evidence — and his conclusions. The debate he framed, however, continues to define the field.
Rodger F. Morton and the Late-Adoption Thesis
The earliest major challenge to the orthodox view came from Rodger F. Morton, who published "The Shungwaya myth of Miji Kenda origins" in the International Journal of African Historical Studies in 1972. Morton's argument was provocative: Shungwaya, he claimed, "is actually an appended myth." His research into colonial-era records suggested that "coastal traditions recorded prior to 1897 indicate that the Shungwaya tradition entered Miji Kenda oral literature only after this date."
If Morton is correct, the implications are significant. A tradition that purports to describe events of the twelfth to seventeenth centuries may in fact be barely a century old, emerging around 1897 in connection with rising ethnic consciousness and, in the Digo case, Islamisation. Morton did not argue that nothing happened — that the Mijikenda peoples sprang into existence fully formed on the Kenya coast. He argued that the Singwaya narrative specifically, with its detailed departure order and named ancestral homeland, was a later overlay, adopted as the Mijikenda groups began to forge a collective political identity in the face of colonial pressure.
Thomas Hinnebusch and the Linguistic Critique
Thomas Hinnebusch brought the tools of historical linguistics to bear on the Singwaya question in 1976. His intervention was primarily methodological: he critiqued Spear's use of linguistic data, observing that Spear "could not confirm Shungwaya as a linguistic homeland because he misunderstood the data." Hinnebusch argued that the patterns of linguistic divergence among Sabaki Bantu languages did not neatly support a single-origin migration from one location.
Yet Hinnebusch's position was more nuanced than a simple rejection. In his later collaboration with Derek Nurse, he would actually support the idea of a northern Proto-Sabaki homeland — though not necessarily Singwaya as described in oral tradition. The distinction is important: the linguistic evidence might support a northward origin for the language family without confirming the specific narrative of a named city, an Oromo invasion, and an ordered departure.
Derek Nurse and Thomas Hinnebusch: The Linguistic History
The most sophisticated linguistic treatment of the question appeared in 1993, when Derek Nurse and Thomas Hinnebusch published Swahili and Sabaki: A Linguistic History with the University of California Press. Their reconstruction proposed that Proto-Northeast-Coastal Bantu ancestors appeared between the Wami and Rufiji rivers — far to the south of Singwaya — around the first millennium AD. Proto-Sabaki speakers then "migrated northwards across the Juba River to a location associated with the legendary Shungwaya," where they diversified into Swahili, Mijikenda, Pokomo, Elwana, and Comorian communities. They dated Proto-Northeast Coast to around 1 AD, with Proto-Sabaki emerging roughly five hundred years later.
This reconstruction partially vindicates the migration narrative while fundamentally complicating it. The direction of travel is right — Proto-Sabaki speakers did apparently move to a northern location and then disperse southward. But the timeline is far deeper than oral tradition suggests, and the process of diversification was linguistic and gradual rather than a sudden flight from a single city. The Singwaya of the oral traditions may correspond to a real phase of Proto-Sabaki concentration in the north, but it was not a single settlement in the way the tradition describes.
Martin Walsh and the In-Situ Thesis
Perhaps the most radical reappraisal of Mijikenda origins has come from Martin Walsh, who concluded "that this evidence is insufficient to support the tradition of a northern homeland, and proposing the alternative thesis that the Mijikenda developed in much the same area that they are to be found today." In Walsh's reading, the Mijikenda did not come from Singwaya. They evolved in the coastal hinterland of Kenya, developing their distinctive cultures, languages, and kaya-based social organisation in roughly the same territory they occupy now. They may have adopted the Singwaya narrative later, borrowing it from neighbouring peoples or constructing it collectively as a means of forging shared identity.
Walsh's work on the Segeju peoples strengthened this argument by demonstrating unexpected linguistic connections. In "The Segeju Complex?" he showed that the Mijikenda lexicon is "replete with loanwords from a Central Kenya Bantu language once spoken by the Segeju/Daiso." These borrowings are not trivial — they cover livestock production, long-distance trading, military and political organisation, and the age-set system, including the foundational word rika ("age-set") itself. If the core vocabulary of Mijikenda political and economic life was borrowed from the Segeju, then the Segeju "left a legacy of political organization and ritual practice that contributed significantly to the precolonial making of the Mijikenda." The Mijikenda identity, in other words, may be a composite — assembled from multiple sources, not descended from a single origin.
Henry Mutoro and the Archaeological Evidence
While the debate over Singwaya had been conducted through oral traditions, colonial archives, and linguistic reconstruction, Henry Mutoro introduced a new category of evidence: archaeology. In excavations conducted in 1987 and 1994 at eight makaya, Mutoro unearthed material remains that complicated every existing position. At Kaya Singwaya, his excavations "yielded pottery going back to at least the tenth century," with ceramics "characteristic of early coastal littoral settlements of the later first millennium AD."
The dating evidence is significant. If kaya sites were occupied as early as the ninth or tenth century, they potentially predate the Swahili coastal settlements — the stone towns of Mombasa, Malindi, and Lamu — that are conventionally treated as the earliest urban centres of the East African coast. This would mean the Mijikenda were not latecomers who settled in the hinterland after the coast was already developed, but early inhabitants whose forest settlements preceded or paralleled the coastal towns. It challenges both the orthodox migration narrative (which implies a relatively late arrival) and the colonial-era assumption that the Mijikenda were "bush people" marginal to the sophisticated coastal civilisation.
Dating Controversies
The question of when the kayas were established remains one of the most contested issues in the field. The range of proposed dates is striking: some scholars place kaya establishment in the sixteenth century, consistent with a late-period migration from Singwaya, while Mutoro's archaeological evidence pushes the date back to at least the ninth or tenth century. This gap of six to seven centuries is not a minor discrepancy. If the kayas are sixteenth-century foundations, the orthodox migration narrative is plausible. If they are ninth-century foundations, the narrative of a sudden southward flight from Oromo invasion becomes difficult to sustain — unless the migration happened far earlier than tradition suggests, or the kayas were established by earlier inhabitants and later absorbed into the Mijikenda system.
Oral Tradition as Adaptive Knowledge
One of the most important insights to emerge from the scholarly debate is the recognition that oral traditions are not static records. They are adaptive, living systems of knowledge that respond to changing circumstances. Thomas Spear himself acknowledged this in his work, even as he argued for the historical core of the Singwaya tradition. Morton demonstrated it by showing how the tradition may have entered Mijikenda oral literature at a specific historical moment. Walsh took the argument furthest by suggesting that the tradition may be an entirely adopted narrative — a story borrowed to serve contemporary needs.
None of this diminishes the value of oral tradition. It means that oral traditions must be read with the same critical sophistication applied to any historical source. A colonial document is not taken at face value; it is interrogated for its biases, its context, its purposes. Oral traditions deserve the same respect — and the same scrutiny. They carry genuine historical information, but that information is embedded in layers of interpretation, adaptation, and cultural purpose that must be carefully separated.
The Tension Between Fact and Charter
At the heart of the scholarly debate lies a fundamental tension between two ways of understanding the Singwaya narrative. Is it a historical account — a description of things that actually happened — or is it a charter — a narrative designed to explain and legitimise the present social order? The answer, almost certainly, is that it is both. Historical events — migrations, conflicts, the establishment of settlements — have been woven into a narrative structure that serves ongoing social and political purposes. The departure order from Singwaya establishes a hierarchy among the Mijikenda groups. The claim of Digo priority establishes their distinctive status. The incorporation of Islamic motifs reflects the reality of nineteenth-century conversion.
Scholars will continue to disagree about where the balance lies — how much is recoverable history, how much is retrospective construction. But the debate itself is productive. Each new contribution — whether from archaeology, linguistics, or careful re-reading of oral traditions — adds another layer of understanding to a question that touches on some of the deepest issues in African historiography: how peoples form, how identities are constructed, how the past is used to make sense of the present, and how traditions that seem ancient may be surprisingly young, while others that seem simple may be astonishingly old.