Right, I should have used the singular.John W. wrote:
I agree with you that, in 6.39.2, the omitted copula seems more likely to be finite - presumably ἐστιν to go with the neuter plural.
One thing I don't like about this is that the subject would have to be the ἃ phrase. But I don't think that is allowed. On my reading, we can take the ἃ phrase as a predicate where a demonstrative has dropped out--at least I think that is what I would have to say. On your reading we would have to have the demonstrative drop out of the subject which is probably not really allowed. Or perhaps you are thinking to take the entire first part beginning with ὀλιγαρχία as the subject for your ἐστιν? That's bold. Actually, I kind of like it now that I look at it. No disappearing demonstrative needed! Of course, I could probably try and smuggle in the whole first part as my object.John W. wrote: Looking at these two passages again, I think that the construction can be straighforwardly analysed as personal:
6.39.2: ἀδύνατα [ἐστιν] ... κατασχεῖν = 'are impossible to secure';
Here, the infinitve is so far out in front of the adjective, that I am uncomfortable. Although this usage is supposed to give us some sense of the infinitive being originally a dative and we do put datives out in front. So maybe not so bad.John W. wrote: 1.1.3: τὰ ... εὑρεῖν ... ἀδύνατα ἦν = 'the [events] were impossible to discover'.
I mentioned this as one of my two slippery features at the outset. "still consults it?"--Today was my first time ever consulting Goodwin.John W. wrote: For the use of an active verb after an adjective where one might logically expect a passive verb, Smith (on 6.39.2) refers us to section 763 of Goodwin's Moods and Tenses (incidentally, I never seem to see this work mentioned on here: am I the only who who still consults it?). Unfortunately Goodwin does not cite any Thucydidean examples, but he does give one from Plato's Phaedo (90c-d), λόγου ... δυνατοῦ κατανοῆσαι, 'a speech possible to understand', which seems a useful parallel.
I think to times when I am uttering long sentences. It seems that there are plenty of times when I am on a roll, especially when I start leaving out copulas and what not or when I start linking sentences in subtle ways, that there are likely more than one way to diagram the sentences.
However, I think that in these adunata examples because the plural gets used so often quasi-impersonally with the infinitive and because that is such a flexible formulation, that there is a certain semantic pull to it that almost no syntax can really control. And so even if a speaker sets out to employ a personal usage, the quasi-impersonal one comes along as a kind of freebe.