“Initial-a-final”, Luwian low vowels, and language contact in the Syro-Anatolian world

Abstract The Anatolian hieroglyphic inventory features two phonetic signs conventionally translated as “a”: 〈a〉 = L. 450 and 〈a〉 = L. 19. Their distribution has eluded explanation due to the complicated conventions governing the use of the 〈a〉 sign, which can be placed at the end of a word (“initial-a-final”), replaced with 〈a〉 (“a-substitution”), or deleted (“a-deletion”). Some words also regularly change between 〈a〉 and 〈a〉 within their paradigm. Based on a more refined understanding of these conventions, which vary by period and location, and comparison with data from the closely related Lycian language and Luwian names in Greek transmission, it will be suggested that 〈a〉 and 〈a〉 corresponded to the low vowels /a/ and /ɑ/ respectively. It will be argued that distinction between the two vowels in Luwian ultimately arose from the Proto-Anatolian contrast between the short and long non-high vowels, but also in some instances from the Luwic umlaut (mostly anticipatory backing of /a/ to /ɑ/). New evidence is also put forth that the vowel system of some dialects of Late Luwian underwent a change due to contact with Semitic languages. This resulted in an orthographic reform beginning in the late 9th century BC, which has practical implications for the dating of certain inscriptions.