A few weeks ago I picked up New Approaches to Greek Particles, conference papers edited by Rijksbaron. I thought I'd give a summary of some of the more interesting papers to give people some new tools to think about Greek with.
Everyone quotes Denniston, either to confirm what he said, or to complain that his book remains the standard. He approaches the particles at the sentence level, and what grammarians everywhere are doing this days is zooming out a bit, and looking at the sort of stuff that goes into an entire conversation (a full discourse, then).
In discourse theory (at least one of them) one can recognize three levels of discourse:
- representational - semantic relationships between the state of affairs in the world represented by the statement (εἰ, ἵνα, ἐπεί)
- presentational - functional relations between discourse units (οὖν, δέ)
- interactional (also "modal" or "attitudinal") - relation of a discourse unit to its non-verbal environment (ἄρα, δή, μήν)
So, the big excitement for me is about two particles usually translated as affirmative emphatics, "truly, indeed":
- ἦ operates at the representational level; Rijksbaron speculates that ἦ is the positive counterpart to οὐ - indeed, they are never paired
- μήν operates at the interactional level - the speaker is committing to the truth of the statement personally, sometimes pre-empting a surprised response. You do find the pair οὐ μήν.
In questions, ἦ γάρ asks for confirmation out of hope, surprise, consternatin, etc.
δή is an interactional particle by which the speaker says "look, how interesting." Obviously this is too strong a translation, but helps to understand what an author is trying to say when using it. καὶ δή introduces an interesting, but not surprising additional bit of information; καὶ μήν is often used to add unexpected information that the speaker feels needs to be propped up a bit.
In another paper, the καί ... δέ pair is shown to match Denniston's original reading: "and on the other hand" where καί is connecting and δέ is adverbial.
The particles can operate at a full discourse level. In fact, one might introduce a digression lasting an entire book. Confusing particle use in Herodotus can sometimes become a bit clearer when you think paragraphs behind, not just a sentence behind.
A μέντοι B - both A and B are true, but B denies an expectation raised by A. καίτοι inverts this relationship:
He is rich μέντοι he is unhappy.
He is unhappy καίτοι he is rich.
There are a few other articles. All the ones dealing specifically with Epic are in French, and that's slow going for me, so I'm not ready to summarize them.
One very interesting article looked at a lot of non-literary Greek on stone local monuments. It appears that the florid particle use we're used to in literary Greek does not probably reflect equal lushness in the common spoken language. Also, the old assumption that particles are especially common in the dialog parts of Attic drama turns out to be demonstrably false.