I've recently finished a first-pass examination of every instance of αλλα in the Apostolic Fathers. While I report numbers here, the big thing to notice are trends; the specific numbers may change as I re-evaluate things between now and November.
There are no surprises in the 352 instances evaluated.
First, αλλα usually occurs with a negated clause or phrase.
- The negator μη (or something very much like it, such as μηδε) is used 82 times.
- The negator ου and its kin are used 188 times (though note some of these are ου μη).
When I say μη or ου, I'm also including things like μη μονον and ουκ μονον and even ουδεν, μηδεν, μηκετι, ουκετι and stuff like that. Maybe not completely and technically accurate, but I have the details down in a spreadsheet I can use later to disambiguate if need be.
There are 76 'clearly' positive (so, no negator on either side of αλλα); there are six that I've found confusing enough to pass on for now. What could be confusing? Sometimes negators are involved, though it is difficult to determine if the entire context is negative, or if something else is going on. These usually involve use of μηδεν.
Recall, my submitted abstract involved examining the "positive" instances, so these instances will be followed up and re-examined.
As mentioned above, the negator occurs both before and after αλλα.
- Of the 82 instances of the negator μη, there is only one that has the negator after αλλα, though there are four instances (e.g. Ign Tral. 5.1) that have negators on both sides of αλλα.
- Of the 188 instances of ου and its kin, 21 instances occur after αλλα (αλλʼ ουκ is a relatively common formation), and seven instances that have negators on both sides of αλλα.
What have I found most interesting? Well, it has to be how the Shepherd of Hermas uses αλλα without negation. Of course, this is the largest item in the corpus of the Apostolic Fathers, but 39 of the 76 'positive' instances are found in the Shepherd. There are some pretty cool things going on in those 39 instances that have no analogue in the New Testament; I'm guessing that I'll end up working through a few of them for the paper as examples of how αλλα functions and what that means for evaluating αλλα from the perspective of discourse analysis.
What's my next step? I have similar data tables for the NT and the Apostolic Fathers. I believe my next step will be to re-evaluate the positive instances in the NT (90 clearly positive instances out of 638; but I have 35 more complex/confusing instances to re-evaluate and classify). After this, I'll be able to really start writing. I've already got a high-level outline in my head, it'll be interesting to see how it fleshes out.