Hey, that was pretty nifty to start a new topic and to get the old posts included.
ὦ χαῖρε Στέφανε!
Don't forget about 1 Thes 5:1:
περὶ δὲ τῶν χρόνων καὶ τῶν κερῶν, ἀδελφοὶ...
"Now, concerning times and horns, brethren..."
It would be even more fun to do this with iotacisms, e.g. 1 Thes. 5:6
...γρηγορῶμεν καὶ νίφωμεν.
"Let us watch and let us snow."
This is a great introduction into what I call the 'carrying capacity' of a natural language. Natural languages can tolerate minor ambiguity, but when major ambiguity starts to expand the language will change and adapt, too.
for example, in the first example from Mark,
peri de ton xronon kai keron would be fairly transparent in fluent speech. χρονος and καιρος are part of the same semantic domain and synonyms. Plus, speakers of a language get to hear and feel the accents (like κέρως vs.καιρός) as well as different genre, so they would read this transliteration that I gave as χρόνων καὶ καιρῶν, not the distinct Koine prose χρόνων καὶ κεράτων for 'times and horns'.
One only needs to know that peri ton (hopefully with a high tone rather than with the low tone [grave]) is a genitive according to its context and everything flows. As mentioned, the tone would even clarify the genitive.
These AI/E interchanges, then, are of particular interest because they were part of the natural language in the first century and they were part of the language system and partially determine the carrying capacity. In a sense, AI/E would help to lead speakers to choose the prosaic κεράτων 'of horns' (frequent in LXX, Jos, NT, et al. over the form κερῶν (*κεράων). The KOINH sound change AI=E may have been a contributing factor to the preference of KERATWN over KERWN. Likewise, with οι/υ, the classical contraction for 'sheep' οἶς (*όϝις with waw/digamma, Lat ovis, Eng ewe) lost out to πρόβατον in part because people did not want to walk into a καπηλειον, order οις (lamb), but receive ὑς (pig, χοῖρος)! Certainly not εν Ιουδαιαι.
However, as more sound changes entered the language, its carrying capacity was altered and this can only affect the rest of the language: what was in common use, neologisms, neo-syntactic constructions, etc. This is most frequently viewed with HTA. In the first century HTA was still phonemic and helped to carry some of the language system by differentiating vocabulary and structures like βλέπει vs. βλέπῃ (the first century distinguished these, Machen can eat his heart out). But from p45, 46, 66, 75, and Origen on, HTA went the way of all flesh and the carrying capacity of the language was diminished. Many of the humorous misreadings above and elsewhere are based on HTA. And the notorious unaccented syllable in υμεις ημεις could not be carried at all by the time we reach the miniscule scribal revolution. (9th c)
Anyway, there are true ambiguities with AI/E, like "KENH KTISIS" out of context, and good for study. And we even have NEA available to disambiguate in live speech should the need ever arise. (It has).
I have found it helpful over the years to watch and see how ancient speakers handled such ambiguities and potential ambiguities in their speech. Stephen's list makes for some good discussion if people will chew it over.
In fact, people must chew this over if they want to understand the first century.