I originally submitted this question to the Linguistics beta site, and those users recommended that I ask anything related to Greek here.
Although I understand that it is impossible to assign a specific time to any sound change, I am curious about the spirantization of voiced stops in Greek, particularly of beta. I'll present the evidence that I am aware of here. It seems that for beta weakening from /b/ > /v/, most scholars rely on parallel indications that the /au, eu/ diphthongs acquired fricative consonantal status, e.g. Modern Greek /av, ev/; the graphic evidence for both were confusions of αυ/ευ with αβ/εβ, e.g. ῥάυδους for ῥάβδους (2nd c. BCE.) However, I wonder if anyone has a clue as to which sound change probably occurred first, if the evidence for both is the same. Based off of analyses of the Koine papyri, Gignac (1976) argues for /b/ > /v/ (or /β/) in the 1st c. CE and /au, eu/ > /aβ, eβ/ (through intermediate /aw, ew/) generally in the Early Byzantine period (when interchanges of -υ with -β become most common, but he does list earlier examples, including from the 1st century.) However, I am slightly unsure about the evidence posed suggesting that /b/ > /v/ occurred first––the β transcription of Latin consonantal 'v' (assuming /w/ > /β/ in Latin), e.g. Φλάβιος for flavius––as these could just be a close approximation of a foreign sound. Does anyone know of additional evidence for either sound change apart from the spelling confusions I mentioned? (I also understand that dialectal variation is a further complication, e.g. Boeotian already has -υ for β in the 3rd BCE.)