Where the Somalis come from: genealogies, language, migration theories and genetics

In short

Where the Somalis come from is answered differently by genealogists, linguists, historians and geneticists. This page sets out what each says, with its source. The older names used for the people of the coast (Barbaria, Berber, Bilad al-Barbar) are covered on Greek and Roman writers on the Somali coast, and the people who lived in the Somali lands in the Stone Age on The Stone Age in the Somali lands.

Genealogies: what Somalis say about their ancestors

Somali clans trace their descent through long genealogies in the male line. Four of the five main clan families (Dir, Isaaq, Darod and Hawiye) are often grouped as Samaale, after a shared ancestor; the Rahanweyn are grouped as Saab, after their ancestor Saabe (EUAA Country Focus Somalia, 2025, glossary, as cited on Clans and lineage).

Several genealogies run back to holy men said to have come from Arabia. In 1854 Richard Burton recorded these traditions in the north (Burton, First Footsteps in East Africa, 1856, Project Gutenberg ebook 6886):

• Sheikh Ishaaq. "The Sherif Ishak bin Ahmed left his native country Hazramaut, and, with forty-four saints… landed on Makhar", the coast between Karam and Cape Guardafui, about "fifteen generations or 450 years ago". His tomb is at Maydh.
• Darod. "Darud Jabarti bin Ismail bin Akil (or Ukayl) is supposed by his descendants to have been a noble Arab from El Hejaz, who, obliged to flee his country, was wrecked on the north-east coast of Africa, where he married a daughter of the Hawiyah tribe."
• Dir and Hawiye. Burton was told that Dir, Aydur, Darod and, "according to some", Hawiye were the "old" genealogies, and that Dir and Aydur were the ancestors of several northern clans.

Who says what. Burton himself wrote that the Somalis were related to the Oromo (whom he called by the older name "Galla"), "Islamised and Semiticised by repeated immigrations from Arabia" (same source). The archaeologist Sada Mire notes that the Walashma rulers of Ifat and Adal "justified their rule and expansion in the Horn of Africa through Islam and their direct ancestral links to Sharif Yusuf Al-Kawnayn", the saint of Aw-Barkhadle (Mire 2015, African Archaeological Review 32). The genealogies themselves are on the Somali clan tree.

Language: a Cushitic people

Somali belongs to the Cushitic branch of the Afroasiatic language family. The Glottolog catalogue of the world's languages classifies it as Afro-Asiatic, Cushitic, East Cushitic, Lowland East Cushitic, Southern Lowland East Cushitic, Omo-Tana, Eastern Omo-Tana (Glottolog, "Somali"). Its nearest relatives are spoken in southern Ethiopia and northern Kenya; Oromo, Afar and Saho are more distant Cushitic relatives. See The Somali language.

Two migration theories

I. M. Lewis, 1960: from the Gulf of Aden southward. The anthropologist Ioan Lewis described "the expansion of the Hamitic Somali from the shores of the Gulf of Aden to the plains of Northern Kenya over the last ten centuries". He based it on oral tradition, which he found "so abundant, and from a variety of sources both inside and outside Somaliland, so much in agreement" that the migration could be reconstructed "with what I believe is a fairly high degree of probability" (I. M. Lewis, "The Somali Conquest of the Horn of Africa", Journal of African History 1.2, 1960, pp. 213–230, abstract).

Herbert S. Lewis, 1966: from southern Ethiopia. The anthropologist Herbert Lewis (no relation) challenged this. He described the "traditional view" as being that the Oromo "occupied most of the Horn of Africa until the Somali, beginning about the tenth century, swept south and south-west from the shores of the Gulf of Aden driving the Galla before them". His own thesis was "that both the Galla and the Somali originated in southern Ethiopia, that the Somali expanded to the east and north much earlier than the Galla, and that the Galla lived only in southern Ethiopia and northern Kenya until their migrations began about 1530" (H. S. Lewis, "The Origins of the Galla and Somali", Journal of African History 7.1, 1966, pp. 27–46, abstract; "Galla" is the older outside name for the Oromo).

E. R. Turton, 1975: the Juba–Tana country. Turton argued that "Garre or proto-Garre Somali gained control of the area between the Juba and the Tana rivers before the Galla arrived", pushing Bantu-speaking peoples back to the Tana, and that the Orma Oromo reached the coast in about the 16th century from southern Ethiopia by way of the Lorian Swamp and the Tana (Turton, "Bantu, Galla and Somali migrations in the Horn of Africa", Journal of African History 16.4, 1975, pp. 519–537, abstract).

Archaeology. Sada Mire argues from the rock art, burials and ruined towns of the north that the region had long local continuity, and that its medieval towns were built by "local kingdoms" of "settled people", not by Arab outposts (Mire 2015). See Laas Geel and Somali rock art and Ruined towns of the Somali interior.

Genetics

Genetic studies sample living people and estimate how their ancestry was formed. They answer a different question from the genealogies: not who a clan's founder was, but where the bulk of a population's ancestors lived.

Y chromosome (the male line), 2005. A study of 201 Somali men found haplogroup E3b1 (now called E-M78 and related branches) in 77.6% and K2 in 10.4%. Most carried one closely related set of markers. The authors concluded that "the male Somali population is a branch of the East African population, closely related to the Oromos in Ethiopia and North Kenya", with lineages "introduced into the Somali population 4000–5000 years ago", and estimated that about 15% of Somali Y chromosomes come from Eurasia and about 5% from sub-Saharan Africa (Sanchez et al., "High frequencies of Y chromosome lineages characterized by E3b1, DYS19-11, DYS392-12 in Somali males", European Journal of Human Genetics 13, 2005, abstract).

Genome-wide, 2012. A study of 235 people from ten Ethiopian populations, South Sudanese and Somalis found "a match between genetic data and linguistic affiliation". It identified a non-African component in Horn of Africa genomes that "may represent gene flow into Africa, which we estimate to have occurred ~3 thousand years ago", closer to populations of the Levant than of the Arabian Peninsula (Pagani et al., "Ethiopian genetic diversity reveals linguistic stratification and complex influences on the Ethiopian gene pool", American Journal of Human Genetics 91, 2012, abstract).

Genome-wide, 2014. A study that added Yemeni samples found a distinct non-African ancestry it named "Ethio-Somali", "found at its highest frequencies in Cushitic speaking Somali populations" and also common among the Afar, Amhara, Oromo and Tigray. It found this ancestry "significantly differentiated from all neighboring non-African ancestries in North Africa, the Levant, and Arabia", "estimated to have diverged from all other non-African ancestries by at least 23 ka" (thousand years ago). It notes that the Arabian lactase persistence gene, which arose about 4,000 years ago, is "almost absent in the Somali (1.6%)". The authors conclude that the migration that brought this ancestry into the Horn "is most likely pre-agricultural", before about 7,000 years ago (Hodgson, Mulligan, Al-Meeri and Raaum, "Early Back-to-Africa Migration into the Horn of Africa", PLoS Genetics 10.6, 2014, e1004393).

Who says what on dates. The 2012 study dates the non-African gene flow to about 3,000 years ago using one method (linkage disequilibrium). The 2014 study accepts that this method gives dates "within the last few thousand years", but argues from other evidence that the ancestry itself is much older.

Related pages

• Before 1300: the ancient Somali lands
• Clans and lineage
• The Somali language