Must We Mean What We Say?

I raised a question yesterday about this sentence from De Officiis I:

Ex quo intellegi potest nullum bellum esse iustum, nisi quod aut rebus repetitis geratur aut denuntiatum ante sit et indictum.

Cicero, I think, meant to write the sentence in this way, and it conveys what he meant to say. Thus, whether the sentence is 'correct Latin' is beside the point. ("Just because Cicero can do it does not mean that you can do it", as my first Latin teacher, Gail Rickert, used to say.)

To see this, consider the following:

1. A universal generalization often has the force of a claim of necessity. (The reason is that it can ask us to look beyond the mere generalization, to the reason for it.) For instance, to say "No human being ever flies" has the force of "It is impossible for a human being to fly." In the same way, nullum bellum esse iustum has the sense "It is not possible that a war be just...".

2. Frequently, when we give a list of necessary conditions, this carries with it the implicature that those conditions, although each necessary, are jointly sufficient. For instance, suppose someone asks for directions, and I say to him, "To get to the post office, you must turn right, then travel two blocks, then turn left, and then travel two more blocks". Each thing I mention is, strictly, a necessary condition of his getting to the post office. Yet in giving the list I mean to supply conditions which, taken together, suffice to get him there. (The reason is that a list is naturally taken to be complete.)

Now apply these things to what Cicero wants to say. He believes (we should presume) that each of the conditions of the fetial code is a necessary condition for the justice of the war. But he does not believe that they are jointly sufficient--because he thinks that there are other conditions, too, that must be met, if a war is to be just. Thus he wants to mention all three conditions, but without suggesting that those conditions are jointly sufficient for the justice of a war. He would not succeed in doing this if he wrote (say) bellum non iustum habetur, nisi de rebus repetitis et denuntiatum et indictum. And so he puts the apodosis in a quasi-modal form (nullum bellum iustum), and he writes the protasis in a way that invites us to consider each condition as working separately (aut...aut...).

That is why in Rep. he puts nisi in front of each condition, which suggests that he is simply giving conditions, each one of which is a necessary condition:

...nullum bellum iustum habetur nisi denuntiatum nisi indictum nisi de rebus repetitis.

(Consider here that it is natural in translation to supply 'or' rather than 'and' for each nisi: "No war may be regarded as just if there is no announcement, or if there is no formal declaration, or if there is no demand for satisfaction.")

And thus in De Officiis he uses aut...aut..., to draw us to think of the conditions as impediments singillatim: "No war even has the possibility of being just, if it fails to be the case that (nisi quod) there is a demand for satisfaction, or an announcement and declaration beforehand".