Wednesday, March 4, 2015

The Oromo people: Heterogeneous

The Oromo people (Oromo: Oromoo; Ge'ez: ኦሮሞ, ’Oromo) are the single largest Cushitic speaking group in not only Northeast Africa but the entire world, outnumbering Somalis by the millions. They are Ethiopia's largest demographic to this day and inhabit its largest state of which the Capital of the country which acts as both a city and a state is centrally located within, they number into just a little over 32 million people by 2013 estimates.

The Oromia Region

Speaking a Lowland East Cushitic language (Afaan Oromo) with various dialects and derivatives; they weren't always this populace or at least this geographically prominent.

Mapping of Oromos' greater spread, breaking through national borders

The original Oromo speakers/ tribes were actually centered according to various sources speaking of the medieval to early modern Horn of Africa-> in Eastern Ethiopia where the Somali Region is now dominant, more specifically in the Southeasterly regions of Eastern Ethiopia, generally areas where the Jubba & Shebelle rivers drain.

However between the 15th to 17th Centuries CE , these agro-pastoralist tribes began to boldly expand inward into Ethiopia, conquering and assimilating in their wake with systems like their Gadaa acting as keen assets in these expansions, they even had their own pre-Islamic monotheistic belief system centered around the God Waaq, also shared by some pre-Islamic Somalis and perhaps even Afars among other groups.

They've been written about or mentioned by a plethora of writers of Ḥabesha, general Islamic-world, Portuguese and perhaps even Somali origins, many of whom encountered the fruits of their expansions one way or the other. In fact the Oromo tribes did once attempt incursions into Southern Somalia only to be pushed back by the dominant political entity of the time; the Ajuran Sultanate.

One can consult the following writers' works on the expansions if they wish to:

 Abba Paulos

Along with the Conquest of Abyssinia (Futuḥ Al-Ḥabash) by Shihab ed-Din speaking about the short-lived Islamic conquests of Abyssinia or even the chronicles of individuals such as Emperor Susenyos of Ethiopia.

The migrations were seemingly at first sporadic raids into the hinterlands that in time grew into what they were; migrations. At this time roughly in Ethiopia, the historically Muslim and mostly Somali soldiered but demographically multi-ethnic Adal Sultanate & the Abyssinian Empire ("Ethiopian Empire") who were Ethiopia's main political powers dominating different areas of what is now the modern state, had more or less bled each other into utter weakness after a terribly bloody war

Arab Slave Trade map showing the Ajuran & Adal realms

The eventual Oromo expansions are by some considered to have acted as a death blow to the Adal in particular, it and the Abyssinians along with other states in Ethiopia at the time proving too weak to prevent many of these Oromo expansions.

But this blog post is not so much about the expansions but rather the genetic impact they had on modern Oromos. You see, Oromos assimilated numerous "Horner" (Cushitic & even Ethiopian Semitic) peoples throughout their expansions. From Amharas to even Sidamic speakers, Rendilles, possibly even Agaws and in some instances Somalis.

 A prime example of this I could share for now would be Sidamics (Highland East Cushitic speakers) such as Hadiya & Sidama people who once were the progenitors of prominent Islamic states in inland Ethiopia such as the Hadiya Sultanate. These groups were once known to have been spread across the modern Ethiopian provinces of Bale & Arsi and even North Shewa only to in some of these areas be supplanted and outright assimilated by expanding Oromo speakers.

"While Bale was the first Ethiopian province, Imam Ahmad Gragn conquered after the Battle of Shimbura Kure 1529, Emperor Geladewos quickly recovered it after the Imam‟s death. However, the territory eventually became the possession of the Oromo people that had begun settling there as early as the Mudana gadaa (1530-1538), and Bale disappeared as a distinct entity by the middle of the next century. The peoples of Hadiya-Sidama who was already Muslims predominantly occupied the ancient state of Bale. One of my informants (himself an Arsi Oromo) completely disagrees with this saying. Yet, Braukamper mentioned that the region of Gadab, which is located in the western Bale, belonged to Hadiyya and was occupied by various subgroups of this people, whose Oromized descendants (for instances, the clans Doodaa, Weegee, Caatimannaa, Adamoonyee, Wosharminaa, and many others) still live there."[5, page 24]

The Hadiya Sultanate actually titled its rulers with the term "Garad" (Somali: Garaad, Arabic: الجاراد ) , an aristocratic title interchangeable with "Sultan" historically mostly known to have been used by Somalis (often along the Northern coast) as well as Muslim peoples in Medieval Ethiopia such as the Hadiya among other groups. And there indeed were seemingly diplomatic and cultural exchanges between the Hadiya entity and the Adal for example until the Hadiya became stalwart Abyssinian loyalists.

What this ultimately does to Oromos, this history of assimilation and expansion, is that the modern ethnic group is in fact made up of various historically & genetically distinct groups even with its current constituents being utterly unaware of this (many are not aware to my knowledge). Now, the populations of the Horn of Africa are quite closely related but indeed there are distinctions between them to be made.

That PCA plot (Principal Component Analysis)/ cluster based on autosomal DNA data above is a prime example of their heterogeneous nature. In terms of admixture levels within Pagani et al.'s original samples carried on by Hodgson et al. , Gudrasani et al. & Pickrell et al. among other papers; there are three subsets of Oromos.

One I like to dub "Oromo A" is Agaw-like in terms of fundamental admixture levels and clusters as you can see above at the fringe of the Northern Ethiopian Highland cluster with Xamir Agaws ("Afar" samples), Tigrinyas & Amharas whilst another group (Oromo B) clusters completely (overlapping) with Somalis demonstrating identical fundamental admixture levels with them to perhaps just a bit more West Asian admixture for some who straddle between us and Northern Ethiopian Highlanders.

The other is less West Asian admixed than Somalis and plots off with Wolaytas or even as one shows; proves almost less admixed than the majority of Wolaytas who can at times be comparable to Somalis in admixture levels. This group would be the Borana who are actually the least West Asian admixed Horner group tested so far.

You can even view this data quite accurately with an ADMIXTURE analysis:

Once again, one Oromo subgroup (Oromo B) is fundamentally identical to Somalis (based on more ancient and fundamental ancestral components), another comes close to Beta Israels (former Western Agaws) & the other would be Boranas, less admixed than Oromo Bs and Somalis.

The genetic data we have so far on Oromos clearly supports what the historical data tells us, that this ethnic group is an amalgamation of various distinct groups within the Horn of Africa it assimilated over the centuries. However... There are things all three of these Oromo genetic subgroups seem to share in common (something that sets them apart from other Horners).

They're all, for one; the peak of Omotic admixture among the so far tested Cushitic & Ethiopian Semitic speaking populations in the Horn of Africa:

Now, components like "Early Neolithic Farmer" or "Near Eastern" from Eurogenes K=8 or the "West Eurasian" or "Eurasian" (essentially West Asian) admixture mentioned in Pickrell et al. & Gudrasani et al. with levels much like Eurogenes' results or the East African cluster from various studies are more basal to the admixture in the Horn and show you how these groups fundamentally look in terms of more ancient (i.e. Neolithic) ancestry, corresponding more with how we plot in PCA plots however there are less ancient components in the Horn made up of these more ancient ones. 

These newer components would for example be Omotic, Ethio-Somali (representing "shared Cushitic ancestry") and so on. 

A good way to grasp this would be-> fundamentally a Somali & a Tigrinya are barely different, there's practically just a minor ~10-12% difference and that's in levels of admixture not even in that one population has ancient ancestry the other lacks. However, in terms of actual more recently shared ancestry as this older post outlines; Somalis & Tigrinyas share about ~60-70% of their more recent to perhaps post-Neolithic ancestry.

Now, that's still a lot of shared ancestry but, for example, some of the West Asian admixture in Tigrinyas while it is fundamentally the same as the admixture that was already in them (in "Cushitic") and that is in Somalis; is ultimately ancestry they have that Somalis don't.

I as an ethnic Somali do not have Omotic ancestry, I carry East African cluster based ancestry with some West Asian admixture (two dominant components of Omotic itself) but a Tigrinya has about ~12% on average Omotic ancestry and I as a Somalian Somali have virtually no "Omotic" Ari Blacksmith-like ancestry, thereupon my Tigrinya "relative" clearly has post-Neolithic ancestors I lack despite the fact that if you go back far enough (i.e. the Neolithic); some of these ancestors would share a common origin with the basal components in me and that are shared between myself and Tigrinyas nevertheless.

In that respect lies the main differences between our ethnic groups despite our fundamental similarity surpassing that between your average Tuscan and your average Englishman or your average Syrian Jew & your average Georgian.

Oromos, at least the ones who've been tested so far (not just in these studies) seem to be quite a peak for Omotic admixture in the Horn of Africa, something they all share in common over the rest of us despite their differing admixture levels.

Also engrossing would be that Y-DNA Haplogroup E-V32, a very prominent marker among ethnic Somalis (the most prominent actually) is very prominent among modern Oromos:

Somalis are also Lowland East Cushitic speakers, it is compelling to say the least that the two groups ("close linguistic relatives") share in this marker's prominence. Some such as a colleague I correspond with suggest sensibly that this maybe a mark of the "original Oromo tribes", a very probable notion as linguistic & cultural shifts of this nature almost always leave some form of a genetic impact.

A basic explanation of Haplogroups for more layman readers

This would suggest most to many modern Oromos do ultimately trace some segment of their ancestry back to the peoples who linguistically shifted them away from whatever Cushitic or at times even Ethiopian Semitic language they originally spoke to Afaan Oromo.

Nevertheless, this ethnic group with its history of expansions and assimilation proves quite heterogeneous genetically, lacking the homogeneous and inbred nature of ethnic Somalis or even the homogeneity you'd find among Beta Israels & Tigrinyas. One colleague whom I often cite in my blog posts as being someone I frequently correspond with about genetics and history (same one I mentioned earlier) went so far as to consider them "More of a Cultural Group" rather than a traditional ethnicity such as Tigrinyas or Somalis.

I'm inclined to agree. But keep in mind that the peoples who were assimilated to form the modern ethnic group are essentially already closely related peoples (as all Horners seemingly are) so it's not like some Azeris went and assimilated Bengalis & Moroccans but rather that Saudis would have assimilated Lebanese individuals and Yemenite Jews. Still a clear distinction to be made but relatively close peoples anyway; genetically, linguistically & culturally but in the end many Oromos have ancestors from just a few centuries ago who likely wouldn't have spoken a single word of Afaan Oromo or any of its dialects and derivatives.

Reference List:


1. This is also why I tend to sadly keep Oromos out of things like the averages of shared ancestry in my blog post about the shared ancestry between the so far tested groups in the Horn. Studies like Pagani et al. & Hodgson et al. don't divulge the results of single individuals within these populations, they just share general averages. Therefore it's really impossible to give someone an accurate look at the results for each Oromo subgroup within their sample sets. I didn't at all keep them out to "disregard" what is a very important demographic within the Horn of Africa.

2. Notably, Pagani et al. noticed this Oromo heterogeneity and even had two subsets of Oromos in this PCA plot (Oromo 1 & Oromo 2).


  1. Borana's are not the least Eurasian tested Horner group. That title goes out to the Gumuz and Ari. Yes, they are Horner as they live inside Ethiopia. Also, I believe the Hamer have once been tested and came out similar to the Ari, but their data is not publicly available.

    1. Ah, well to be honest my use of the term "Horner" is more descriptive of groups like Agaws, Habeshas, Somalis, Afars etc. I know one would count Aris as "Horners" since they live in the Horn but I meant it more in that Boranas are the least admixed among the so far tested Cushitic (discounting former Omotic, Bantu & Khoisan speaking peoples who've been "Cushitized") or Ethiopian Semitic speaking group. I define Horners as basically those peoples.

      But I'll make an edit soon and say 'Cushitic & Ethiopian Semitic peoples' as "Horner" can be a pretty broad and ambiguous term to some.

    2. Horn of Africa is a geographic term not an ethnic term. Since the Omotics and Gumuz are actually native to this region, they are just as much ''Horner'' as others.

    3. I'm just too used to some circles online where people refer to populations like Cushites & Ethio-Semites as "Horners".

      I agree that it's a disingenuous label given that the Horn is not solely populated by our populations. But it just works better than constantly outlining what I'm saying like "Cushitic speakers of not originally Arabian ("Benadiri"/ Reer Ḥamar in Somalia), Bantu or Khoisan origins" + Ethio-Semitic speakers.

  2. Not all of Behar's Oromos are Borana. Most are, but one isn't.

    GSM537014 was sampled in Haraghe (per e-mail correspondence with Behar).

    Notice how he has less of the Omotic cluster than the Ethiopian Somali outliers. Arguably the only Oromo sample that is similar to Somalis out of the many.

    1. Hmm, I had noticed that one sample standing out but didn't realize he was sampled in Hararghe. It's interesting that other than the Omotic admixture-> he or she looks pretty Somali-like.

      Do you know anything about GSM537021? The sample looks pretty Habesha or Agaw like and has what in Bandar's run seems representative of a least some of the newer West Asian admixture in Habesha & Agaw populations ("Northwest Asian").

    2. That one was sampled around Moyale with the other Borana. But since the deviant nature we can assume it's an urban immigrant from elsewhere in Oromiya.

    3. PS. The Hararghe GSM537014 one has the highest % of the ''Cushitic'' cluster in Bandar's runs out of all the non-Somalis.

      We can assume that Oromos from areas near Bale and Hararghe are the closest to Somalis.

    4. "That one was sampled around Moyale with the other Borana. But since the deviant nature we can assume it's an urban immigrant from elsewhere in Oromiya."

      Lol_Race posited as much on ABF. Seems like a reasonable position.

      "PS. The Hararghe GSM537014 one has the highest % of the ''Cushitic'' cluster in Bandar's runs out of all the non-Somalis."

      To be honest, Bandar's run becomes somewhat useless after K=7, I wouldn't make much of its data (actual percentages) beyond the "clime" (i.e. this group is more "Cushitic" than this group and so on). The run is mostly exemplary in its usefulness when we're dealing with levels of Omotic admixture, it's not very reliable otherwise.

      However your point does stand about GSM537014 (Hararghe residing Oromo) in that it looks pretty Somali-esque. In fact; compare that Oromo at K=7 to this Ethiopian Somali sample:


      Omotic: 7.3%
      Nilo-Saharan: 47.6%
      Arabian: 44.2%

      GSM537014 (Hararghe Oromo):

      Omotic: 7.0%
      Nilo-Saharan: 49.7%
      Arabian: 43.2%

      Not really a notable difference between them at all. Albeit ESOMALI4 is not entirely representative of Ethiopian Somalis and is at the highest bound (once you remove the two "outliers") for their Omotic levels, as you know some Ethiopian Somalis even lack Omotic entirely.

      "We can assume that Oromos from areas near Bale and Hararghe are the closest to Somalis."

      I'd assume some Bale Oromos would be more Sidamic-like (those more west in the area)? And while I find your assumption likely, one sample can't entirely be taken as representative of all Hararghe Oromo.