Contact-induced changes in Circassian by Ranko Matasović

Ranko Matasović
Croatian linguist, Indo-Europeanist and Celticist.


The two main branches of the NW Caucasian family, Abkhaz-Abaza and Circassian, are typologically similar to a remarkable degree, although they are not genetically closely related (it is estimated that the common proto-language was spoken several millennia ago, and the time-depth of NW Caucasian is comparable to that of PIE). Indeed, both Circassian and Abkhaz-Abaza are polysynthetic languages with polypersonal verbs, very complex consonant systems and rudimentary vowel systems, with a simple syllabic structure, with ergative clause alignment and highly complex verbal morphology with categories such as causative, benefactive/malefactive, involuntative, reflexive and reciprocal, antipassive and optative, all expressed by means of prefixes (and less commonly by means of suffixes) on the verb.

The question addressed by this paper is the following: Why is it that, while Circassian is typologically so similar to Abkhaz-Abaza (and Ubykh, which is extinct), it nevertheless differs from it in a number of salient grammatical features. Indeed, those features often figure in literature about language contact, as they appear to be prone to borrowing in situations of intensive language contact.

I will review a selection of these features and show that in all cases it was the Circassian languages that innovated, while Abkhaz-Abaza retained the original situation. I will then speculate about the possible causes of the changes that affected Circassian, and claim that contact with Ossetic is the most plausible explanation.

Download the full-text document in PDF format.