My understanding of the history is that PIE had active and middle voices, while the passive was a later innovation. Therefore it seems that in the early language as preserved in Homer, we see the middle voice being used with semantics that we as English speakers would expect for the passive. Example:
For example, in Iliad 8.513, Hector says:
ἀλλ ̓ ὥς τις τούτων γε βέλος καὶ οἴκοθι πέσσῃ βλήμενος ἢ ἰῷ ἢ ἔγχεϊ ὀξυόεντι νηὸς ἐπιθρῴσκων,
Buckley's translation:
"but [let us] take care that each of them [the Greeks] may have to heal a wound at home, being stricken either with an arrow, or with a sharp spear, bounding into his ship"
Here, the ending of βλήμενος is middle, but the semantics are what I as an English speaker would expect for the passive voice. Cunliffe, who is ordinarily a fanatical semantic splitter, doesn't note anything about the semantics of the middle of this verb that would suggest that this is noteworthy. Smyth (p. 689) says: "Epic forms ... of the 2 aor. mid. as pass. ἐβλήμην (... part. βλήμενος)."
The web site hellenisticgreek.com, however, has an additional twist on this. They say that even in koine, it's idiomatic to use the middle with the semantics of the passive voice. The example they give is from Acts 4:3: ἔθεντο εἰς τήρησιν. The problem here is that it's not really obvious to me that this is a convincing example. Both the actors and the people being acted upon are plural, so we can't tell from the plural verb form who is the subject. There are two verbs with a και in between, and the object is given only before the και. In English, for the active verb, we would need to repeat the object for the second verb ("and placed them in custory"), but I imagine that in Greek it might be grammatical to omit the object for the second verb.
Are there any less ambiguous examples that show that the use of the middle with passive semantics persisted into koine?
