What feels like 10 years ago I told you a helpful little tip that Levinsohn told me (ok, not in person, via 2006) about how linguistics could influence my exegesis. Specifically, we're talking about discourse analysis, one of the many sub-disciplines of general linguistics, which happens to really shine with explanatory power when dealing with understanding what a speaker is trying to communicate to some people. Levinsohn brings up 3 main points. The first was that of word order. Most commentators don't really know what to do with variations in word order but Levinsohn points outs several ways that discourse analysis can do more than just call everything emphatic.
In this post, we'll quickly cover his second point, which actually dovetails with the first. And that is, that the article in Greek can actually help you quite a bit in identifying which propositions in an utterance are considered old information (and likely topical) or new information (and likely focal) by the speaker, or author in the case of the Greek New Testament.
So without further adieu, Levinsohn (2006:16) remarks that
When a constituent precedes the verb or has been preposed, one way in Greek to decide whether it is topic-like or focus-like is to see whether the article is present or not.
- Topic-like constituents are typically associated with established information, so are more likely to have the article.
- Focus-like constituents are typically associated with non-established information, so are more likely not to have the article.
I don't think I can put it any more simply than that, so let's just get on to an example that Levinsohn gives which really showcases this general tendency at work.
Οὐκ οἴδατε ὅτι ναὸς θεοῦ ἐστε καὶ τὸ πνεῦμα τοῦ θεοῦ οἰκεῖ ἐν ὑμῖν (1 Cor 3.16)
Do you not know that you are THE TEMPLE OF GOD, and the spirit of God dwells in you?
The part we're interested in is the temple of God, which as you'll notice in the Greek is anarthrous (i.e. without an article), unlike in English. So because this constituent is preposed before the verb ἐστε and is lacking an article, we should be leaning towards thinking it is marked as focal; and if we perused the larger conversation, we'd find that this is indeed the case, as it is the new information (or proposition), which coupled with the rest of the sentence turns this utterance into an assertion (i.e. it's this part that brings something new to the table).
Now, in the very next verse we have this same Greek expression used again, but this time it does have the article.
εἴ τις τὸν ναὸν τοῦ θεοῦ φθείρει, φθερεῖ τοῦτον ὁ θεός· ὁ γὰρ ναὸς τοῦ θεοῦ ἅγιός ἐστιν, οἵτινές ἐστε ὑμεῖς.
If anyone destroys THE TEMPLE OF GOD, God will destroy this dude. For THE TEMPLE OF GOD is holy, which is what you are.
Levinsohn (2006:16–17) rightly notes that, initially, even though it might look like anyone is in the P1 position (topical) and the temple of God in the P2 (focal), this can't be the case because the latter is already presupposed information at this point, thanks to the preceding verse; and on top of this, it's got the article. That leaves the verb destroys to actually be the focal element in this verse.
Now, what about the second use of the phrase the temple of God? Honestly it's mainly the same. Simple as that. It's got the article, it's already established information and it's in the P1 slot—so it's being re-activated as the topic of the current utterance—while holy is in the P2 position and is thus a case of marked focus.
The interesting thing here, is the final phrase: which is what you are. To which we might respond: What?! What am I? The temple, or holy? How do I know, since both could make sense? And that's where discourse analysis shines again.
The predication of the sentence preceding which you are is not that we are the temple. This has already been established in vs. 16. So here, the big deal is that this temple, this place where God dwells is ἅγιός, or holy. That's the real kicker, and important enough for Paul to reiterate once again.
 The truth is, a guy like Lambrecht—a stud in the information structure world—would rightly point out some clear distinctions between 'old/new information', 'information' and 'utterances', 'topics' and 'presuppositions', 'foci' and 'assertions', but I don't want to bore you with this, and honestly, have a hard time keeping things separate as it is, so I'll opt for simplicity (and naivety at this point, for the sake of my readers, of course!). ;)