When South Arabian loaned a word for chicken

Today, I thought a lot about Somali.

I couldn’t tell you why, but I just did. And then, I realized something while expanding the list of pastoralism-related loanwords into Somali that had come from: almost every loanword has a cognate in the Shehri language. Shehri of course is the only South Arabian language where the speakers are avid seafarers and pastoralists, with their language being spoken not only on mainland Oman but also on the adjacent Khuriya Muriya islands in the Gulf of Aden. But before this I had never really noticed it being well, there’s also some Mehri cognates but Mehri and Shehri are a bit close in relation, which I simply settled with. But then I re-did my list:

  • “Cattle” – Somali /ħɔːlɑhɑ/ ~ Soqotri /ʔelheh/ ~ Shehri /lhoti/
  • Bull” – Somali /dibi/ ~ Harsusi /ɣɔɮəb/ ~ Shehri /ɣɔɮəb/
  • Goat” – Somali /rijɑhɑ/ ~ Mehri /tajh/ ~ Soqotri /tɛʔɛh/
  • Sheep” – Somali /idɑhɑ/ ~ Soqotri /daħ/
  • “Chicken” – Somali /degɑg/ ~ Shehri /dəgugət/

But it got even better, the root #r-n in Somali and Maay has cognates in almost every South Arabian language:

  • Meh. /ħaːra͡wn/ ~ Bat. /ʔaːʔəraːn/ ~ Soq. /ʔərəhɔn/ ~ Hob. /ħəjwəroːn/ ~ Sheh. /ɛrun/ ~ Har. /ħəwəruːn/

Of course it’s surprising that this also showed up in Maay* but it also happens to be one of the most consistent loans into Somali; with Shehri having the closest form to the Somali lexical item. This being said, I am left with this suggestive conclusion: perhaps, the language of those who went to the Horn of Africa was indeed Shehri. If this is true, this could be a possibly point of explanation as to why Somali has vowel harmony unlike any of even its closest relatives.

This is of course is a very hard thing to sum up in one post and will be left for later, but it could be evident that there is actual evidence beyond loanwords of the contact between “Proto-Somali” and a language akin to if not Shehri. If the language is Shehri or Shehri-like, then the implications are huge for the interpretation of the relationship between the Cushitic languages of the Horn of Africa and the Semitic languages of Southernmost Arabia. This would of course have to supplant the current theory of an initial migration into the Horn of Africa, and we entirely reshape how we see the language contact in such a situation. It would be indicative that it was a Semitic language that structurally changed the phonology of a Cushitic language as opposed to the current theory, which is the exact opposite. This though would be an isolated incident, being that while Somali and Somali alone shows these traits of contact with a Shehri-like language there is Cushitic languages belonging to the Lowland branch who share important isoglosses with Somali but lack these innovations, such as the rest of the Somali languages and the Saho-Afar branch that is spoke adjacent to the westernmost speakers of the Somali language.

unnamed (33)

Taking all of this into consideration, it would also reinterpret the historical view of the South Arabian languages being it would not only attest to their antiquity but it would also present them as a more dynamic and widespread branch of the Semitic family tree. It is already noted in Anonby (2013) that the Kumzari language of northernmost Oman exhibits features from South Arabian languages that are also shared in the Omani dialects of Arabic, which attests to a northern range for the languages but there is very little evidence for a southernmost expansion unless you count Soqotri and perhaps, just maybe, Himyaritic. Yet if this hypothesis does reflect historical fact it then writes a whole new chapter into how we view not only the South Arabian languages but how we view unwritten Semitic languages; which are undoubtedly lacking in proper historical analysis.

This although is an underlying theme of Semitic lingusitics: history is only understood through written attestation. This isn’t like the Indo-Iranian languages, who’s history has been reconstructed even without written attestation; hence why we have papers such as Witzel (1999) and Lubotsky (2001).  But for unwritten Semitic languages and indeed for Cushitic languages outside of the Southern Cushitic languages, we have a poor understanding of unwritten history. Cushitic languages until recent were unwritten, which leaves little room for bias towards written languages. In the case of Semitic languages though this is the actual problem. While history is easier to understand through written languages, there is always ways to find history through the languages themselves as demonstrated by the two papers on non-Indo-European substratum in Indo-Iranian languages and their implications for the unwritten history of the Indo-Iranian branch of Indo-European.

This of course is one instance where unwritten history might actually be able to be traced via the effects of this Shehri-language on “Proto-Somali“. This is one step toward, what in my opinion, is a better understanding of Semitic as a whole. Hopefully.