When your word for “language” is a loan

Oh out-dated “Sabaean theory“, you’ve been able to birth some of my favorite posts.

From what it seems after years of reading, new information coming out almost every year, and just writing these posts putting it all together, I can gladly say that I have every reason to doubt the Sabaean theory and have every motivation to end up writing a book on why it’s false. But, alas, that won’t be for years, and anyone interested is stuck with whatever is on my WordPress. Regardless, every single part of my examination of the theory has been directed towards the idea that Central Cushitic influenced Ethiopian Semitic, with only a light review of the reality of the case which turns out to be the other way around.

In light of this, I spent about two days reading and formulating this post. Somehow, someway I was going to write this post. This involved reading a 70+ page paper on an obscure-until-recent dialect of Awngi, reviewing papers by Paul Fallon for perhaps the 30th time in the last 2 years, and spending the past 4 hours this morning trying to figure out how to word this. But in all regards, the argument can be made and I’m going to do it primarily through examination of 2 fairly obscure Central Cushitic dialects: Kayla and Kusili; as described by Appleyard (1996) and Leyew (2011). These two dialects belong to two respective languages belonging to each of the two main branches of Central Cushitic, Qemant (North Central Cushitic) and Awngi (South Central Cushitic), and represent fairly well the picture of language contact between Ethiopian Semitic and Central Cushitic.

So in the second “The Sabaeans, who never colonized Ethiopia” post I had addressed fact that no known Central Cushitic language has ejective consonants that do not originate from Ethiopian Semitic influence, and thusforth would make it nearly impossible to have had been the origin of ejectives in Ethiopian Semitic. This was significant at the time because of the fact that it means that ejectives in Ethiopian Semitic languages are better said to be a Semitic-originated trait than a Cushitic-influenced one, and thus meaning that the Central Cushitic languages also lack a diagnostic Afroasiatic trait: the 3-way voiced, voiceless, emphatic contrast on most stops and affricates. This trait is rarely lacking but there is cases, most of which are Cushitic. The Beja language of Sudan, Eritrea, and Egypt lacks the traditional 3-way distinction (in most cases emphatics are ejective or pharyngealized) and has a phonemic voiced, voiceless, retroflex contrast on alveolar polsives; and in another interesting case numerous Lowland East Cushitic languages such as the Oromoid languages, Somali, Maay, Afar-Saho, and Dahalo all lack phonemic /z/. So while Central Cushitic is not alone in its phonological oddity on Afroasaitic-terms, it nonetheless is important to isolate the lack of the an emphatic series when examining Ethiopian Semitic loans into even basic vocabulary in Central Cushitic languages. As cited in Leyew (2011), the following examples can be found for Kusili:

  • green pepper” – Amh. [k’et’o] > Kus. [keto]
  • shoe” – Amh. [t͡ʃ’ami] > Kus. [t͡ʃamma]
  • mud” – Amh. [t͡ʃ’ik’a] > Kus. [t͡ʃika]
  • salt” – Amh. [t͡ʃ’ɐw] > Kus. [t͡ʃiwi]
  • cotton” – Amh. [t’ɨt’] > Kus. [tɨti]
  • language” – Amh. [kʷ’ankʷ’a] > Kus. [kʷankʷi]

Of course there’s many more words which can be found in the wordlist found at the end of Leyew’s paper, but the interesting thing about this that many of the loans cited by Leyew (2011) are agriculture related, which will have to be discussed later on in a post about “Proto-Central Cushitic”, but is significant nonetheless.


When one looks at Kayla, for which there is no extensive lexical list but some lexical items available throughout the paper such as the lexical item for “ask“, [t͡ʃ’ɐw-], need more explanation outside of simple comparison. In all likelihood, [t͡ʃ’ɐw-] is a loanword. At the time of Appleyard’s source of the collection of data on the Kayla dialect the language was succumbing heavily to Amharic and likely was in many ways reduced in native lexicon by those who could speak it due to such linguistic pressures. What makes this word interesting is the fact that this word is formal in nature via it’s origin in Amharic. the root [t͡ʃ’-w] in Amharic has to do to some degree with civility or politeness, or as seen above, salt. Of course, this term does not have an origin with salt. The term [t͡ʃ’ɐwa] in Amharic can mean:

  • Adj. – Civil, urbane, “cultured”
  • Noun – A polite person

And then there’s words like ከጨዋው [kɐt͡ʃ’ɐwamu] which also have to do with civility and being in a position of civil (i.e “formal”) office, but in this case the root [t͡ʃ’-w] in Kayla could very well better be translated as “ask formally” or “ask politely”. This of course is likely, but is probably impossible to confirm due to the reality that it’s likely no one speaks Kayla in the present day. It’s also hard to confirm that any of the words in question from Appleyard (1996) were actually pronounced this way, being that Appleyard was not the person who collected the data first hand; Appleyard was working with an unpublished set of data collected and written down by Jacques Faitlovitch in the 1920’s during his stay with the last of the native speakers, who happened to be members of the ever-mysticized Beta Israel. Not to be ignored, it’s likely that Faitlovitch’s transcriptions were off because they were in the Ge’ez script, a script at this time that not many Orientalists specializing in Ethiopia understood very well. Regardless I do maintain a degree of confidence being that Appleyard does list Xamt’anga as also exhibiting the lexical item [t͡ʃ’ɐw-] and the non-Kayla dialects of Qemant exhibiting the lexical item [[ʃiw-]; the latter representing a sound change from /t͡ʃ’/ > /ʃ/ in other Qemant dialects and thus a preserved form in Kayla.

This shows another interesting feature of loans into Central Cushitic that contain ejective consonants. There is no entirely streamline measure of what plain stops or fricatives Ethiopian Semitic ejectives will assimilate to in Central Cushitic languages, or if such a change will even happen. While above a sound change of /t͡ʃ’/ > /ʃ/ is mentioned in regards to non-Kayla Qemant dialects, one can also find examples of the preservation of /t͡ʃ’/ in non-Kayla Qemant as well:

  • finish” – Amh. [t͡ʃ’ɐrrɐsɐ] > Qem. [t͡ʃ’ɐrɐs-]
  • yellow” – Amh. [bɨt͡ʃ’a] > Qem. [bɨt͡ʃ’a]

And despite in most cases the Ethiopian Semitic loans and influence being from Amharic, this is not universal. Much of the contact related influence in Xamt’anga and all of the influence in Bilin can be traced back to Tigrinya, resulting in loans for the same words having different results:

  • yellow” – Tig. [bit͡ʃ’a] > Bil. [bit’a], Xam. [bɨs’a]

And even more interestingly, the loan in Xamt’anga very well could be from Ge’ez! Because of the likelihood that the Xamt’anga language was or is the descendant of the language spoken by the Zagwe dynasty in the 10th-11th centuries it’s equally as likely that the loan could originate in the Ge’ez word [bes’a]. But the assimilation of /t͡ʃ’/ isn’t the only ejective that has an uneven result. Let’s talk about /k’/ and its labialized variant /kʷ’/.

If one takes the loans into Kusili that would have /k’/ in Amharic, one will see a pattern:

  • skin” – Amh. [k’orbɐt] > Kus. [korbɐt]
  • dirt” – Amh. [k’oʃaʃa] > Kus. [koʃaʃa]
  • bean” – Amh. [bak’ela] > Kus. [bakela]

And then, you get the word for crow:

  • crow” – Amh. [k’ura] > Kus. [qura]

And of course, there’s more than just that example but they’re not very common. The only other one listed in Leyew (2011) is the lexical item for “mule” but to some degree there is a variation between a pronunciation with /k/ and another with /q/. This is mentioned in Fallon (2009) in regards to Bilin, which has phonologically been greatly influenced by Tigrinya and exhibits a variation between /k’/~/q/ and /kʷ’/~/qʷ/:

  • whistle” – [fikʷ’ɨr-]~[fiqʷɨr-]
  • sickness” – [ʃɨkʷ’ɨda]~[ʃɨqʷɨda]

Both of these lexical items are also not loans, but this is not something distinctly problematic in the way that Fallon (2009) attempts to put forth. Bilin, Xamt’anga, etc. have all been heavily influenced by Ethiopian Semitic languages after centuries of being the basiclects or (as of recent) mesolects of the societies in which they exist and it’s only natural that phonology would be one of the first things touched by this linguistic pressure. The presence of native lexemes showing these variations, if anything, shows that pressure from Ethiopian Semitic languages has begun to create a class of emphatic consonants within Central Cushitic, which has lacked them before the extensive contact.

And all of this, from top to bottom, is when when talking about Kusili you can confidently say first and foremost that it’s a language with a loaned word for language itself.


5 thoughts on “When your word for “language” is a loan

    1. Oh my God I’m from Agame and my mother is of partial Saho descent! I identify as ethnic Tigray being my father is a Jeberti and my mother spoke Tigrinya but it’s so cool! I had written some stuff before on how Saho phonology and syntax is inherently influenced by Ethiopian Semitic languages but I never finished it.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s