ϰάμηλος

There is a general and justified agreement that Ancient Greek ϰάμηλος ‘camel’ is a loanword from (West) Semitic, cf. Hebrew gāmāl ‘dto’, etc. (GEW I:771f., DELG:489; EDG:630; Rosół 2013a:43f., Rosół 2018a:337, all with further refs.). However, the voiceless initial consonant instead of the expected voiced one remained unexplained (the word appears with a voiceless initial consonant in Nubian and Beja, too [Rosół 2013a:43 with refs.], but a Greek borrowing from these languages is not very probable). Grimme 1925a:17 suggested a Hittite transmission, which would solve the problem, his solution, however, was ignored by the etymological dictionaries. Recently Yakubovich 2016a:87 n. 30 suggested a Luwian transmission based on his identification of Hieroglyphic Luwian kamara- as ‘camel’. Provided that this identification is correct (see the supportive discussion by Giusfredi 2020a, but cf. Bauer s.v. kamara-), Yakubovich’s solution is to be preferred, since it is based on an attested word and geographically as well as historically more plausible than a Hittite transmission (cf. Simon 2018a:393).

Giusfredi 2020a:54f., 56, however, rejects the Luwian etymology claiming that “there is no need to invoke the Luwian devoicing of the initial velar stop”, since the devoicing of the initial stop “can easily happen in contexts of adaptation and requires no formal explanation in terms of sound change (which is expected to be regular and predictable only when it occurs over time within a single language)”. This is, however, a methodological failure, since sound substitution in borrowings requires a formal explanation and sound substitution does not happen “easily”, but it is determined by the phonology of the target language (i.e. regular and predictable). Since Ancient Greek did allow /g/ in initial position, the sound substitution /g/ → /k/ in this position does require an explanation and only the Anatolian languages can provide a solution at the current level of understanding (Rosół 2013a:443 refers to the case of Greek παλλακίς ‘concubine’ allegedly borrowed from Hebrew pi/îleḡeš ‘dto’, but in fact the origin of the Greek word and the connection of these words are completely unclear [see the detailed overview by Rosół 2013a:76–79 with refs.] and a direct borrowing can definitely be excluded on phonological grounds).

Giusfredi 2020a:56 claims furthermore that the geography would also create problems, since the Luwian word is only attested in the Syro-Anatolian interface, but not in the “central and peninsular” parts of the Luwian speaking regions. However, it is exactly the Syro-Anatolian interface, more precisely the Northern Levantine, partly West Semitic speaking region, where Greek presence is attested, which provides a perfect geographical background for the borrowing.

[Zs.S.]

Last updated: 28 April 2020