Soph. OT 1076 Ὁποῖα χρῄζει ῥηγνύτω

{ΟΙ.} Ὁποῖα χρῄζει ῥηγνύτω· τοὐμὸν δ’ ἐγώ,
κεἰ σμικρόν ἐστι, σπέρμ’ ἰδεῖν βουλήσομαι.

It appears that χρῄζω is used here with an impersonal agent: "Break forth what will! " — Jebb. “May whatever will burst forth!” — Lloyd-Jones.

I looked in LSJ but didn’t find anything which was clearly impersonal. Perhaps “fate” (not explicitly mentioned here) is treated as a sort of personal agent.

It’s not actually impersonal. In the previous verse the chorus expresses its fear that αναρρήξει κακά, “evils will break out.” Oed’s response continues directly from that. Lit. “Whatever sort of (evils) wish/need/want/desire (to break out), let them break (out),” or “Let there break (out) what sort of (evils) want (to break out).”

PS. Or have we been tripped up by terminology again? Conventionally an “impersonal” verb is one which has no subject (whether explicit or implicit). See any grammar. “It’s raining” (e.g.) is the closest English construction. χρῄζω is not used like that. Unlike e.g. χρή, it always (I think) has a subject, an agent. Makes no difference in principle whether that is a person or a thing, though in practice of course a verb implying volition or desire or the like will usually have a person as its subject. In the English translations you quote, “what(ever) will” would not be classified as impersonal, any more than “who(ever) will” would; the syntax is identical. οποια is nominative neuter plural, sc. κακά, which you can think as quasi-personified if you wish, since the image implies something with a will of its own. This is poetry, after all. But there’s nothing at all exceptional about it grammatically.

mwh wrote:

κακά, which you can think as quasi-personified

As I suspected impersonal was the wrong word because it would be taken as grammar metalanguage rather than semantic metalanguage. What I was looking for in LSJ was an example where χρῄζω has an agent which is not a person, either a mortal or immortal. If we consider κακά in some limited sense personified then the there is no issue. I suspect there are other examples of κακά filling the agent slot with a verb which would typically take a person as an agent.

So you did understand the syntax? But evidently not, or you wouldn’t have suggested “fate.” (Note the accent and scansion of οποῖα, showing that it’s neuter plural not fem.sing.)

Anyway, verbs that most often have personal agents don’t always do so. (Isn’t it just the same in English?) When they don’t, the quasi-personification, if you want to call it that, can be semantically weak. Think e.g. βουλεσθαι or δεισθαι οr θελειν, semantically not all that far from χρηζειν. You have an instance of δειται with ἔπη as subject coming up at OT 1148. Of course poets tend to personify things more than ordinary prose does. Take a look at the Latin board on Eclogue VI.