One of the main topics of the Leibniz-Arnauld correspondence is the question how, on Leibniz’s theory, it can be true that Arnauld might have had children and been a physician rather than being a celibate theologian (see Arnauld’s letter of May 13, 1686). One of the curious things that happens in this discussion is that both Leibniz and Arnauld start talking about the many Adams and many Judases and many Arnaulds in the various possible worlds, with Leibniz insisting that none of them is identical to the actual Adam/Judas/Arnauld. In that May 13 letter, Arnauld even speaks of ‘several mes’, pluralizing the first-person pronoun.
In my view, what’s going on here is that Leibniz holds that as long as we use ‘Arnauld’ as a genuine proper name, the sentence ‘Arnauld is a celibate theologian’ is in fact a necessary truth. This is because, when ‘Arnauld’ is used as a genuine proper name, it picks out the actual Arnauld by means of his Complete Individual Concept (CIC) which includes everything about him and, indeed, everything about the world he inhabits. Of course, we don’t actually possess such a concept (at least not consciously); only God does. But somehow or other, in Leibniz’s view, we manage to use the name ‘Arnauld’ in such a way that this concept is its meaning.
However, there are other, looser contexts in which we use ‘Arnauld’ in such a way that its meaning is given not by Arnauld’s CIC but rather by some concept we actually do possess. In this kind of case, ‘Arnauld’ could be seen as abbreviating some definite description which is sufficiently detailed to identify Arnauld uniquely among actual creatures, but not including all of his characteristics, and hence not uniquely identifying him among all possible creatures. For instance, the description might be, ‘the youngest child of Antoine and Catherine Arnauld, born February 8, 1612, assigned the given name “Antoine”, who went on to become a famous Jansenist theologian.’ When we talk about alternate possibilities (‘Arnauld might have been a married physician’) or evaluate counterfactuals (‘if Arnauld had been married, he would have become a physician rather than a theologian’) we are using ‘Arnauld’, not as a proper name, but as a common name standing for such a definite description. It is for this reason that we can pluralize ‘Arnauld’ in this usage and speak of the other Arnaulds, in other possible worlds, who are married physicians. These are merely possible creatures, distinct from the actual Arnauld, who nevertheless satisfy the description which, in this context, ‘Arnauld’ abbreviates.
Today, I came across some support for this interpretation from a surprising source. It turns out there is actually a discussion of the practice of pluralizing proper nouns in the Port-Royal Grammar which Arnauld co-wrote with Claude Lancelot:
if [proper nouns] are sometimes put in the plural, as when one says the Caesars, the Alexanders, the Platos, it is done figuratively by including in the proper name all the persons who resemble them, as one would speak of kings as brave as Alexander, of philosophers as wise as Plato, etc. (part 2, ch. 4, tr. Rieux and Rollin)
This, it seems to me, is good evidence that Arnauld understood Leibniz’s view in more or less the way I have described. Of course a more careful analysis of the correspondence would be required to show that Arnauld understood Leibniz correctly.
(Cross-posted at blog.kennypearce.net.)
Does this interpretation comport with the line that Bennett gives on the Leibniz-Arnauld correspondence?
What does Bennett say? My reading is pretty close to Mondadori’s, and it does make Leibniz’s view similar but not identical to Lewisian counterpart theory. (I do think the dispute between Leibniz and Arnauld really is in part about transworld identity.)
I’ve generally been persuaded by Bob Adams’ view that one can’t make much modal semantic mileage out of Leibniz’s comments about counterfactual situations involving Arnauld.
That said, it does seem to me that we need to take seriously Leibniz’s claim that the truths about what Arnauld did are contingent if we are to get to make progress with Leibniz’s views on modality more generally. And here I don’t have anything intelligible to add to Adams either – i.e., to the thought that necessity and contingency are to be cashed out in terms of the logical structure of the propositions expressing them (something which will in turn have important epistemic consequences, but which isn’t treating all modality as epistemic modality).
This is not to say, of course, that Leibniz doesn’t think all truths are ‘necessary’ in what is, for some, a more intuitive sense of the term – given that Leibniz does accept that all the facts about the created world are fully determined to be the case by a fully determinate ground (i.e., God).
I agree that giving a semantics for counterfactuals is not what Leibniz is up to, but Leibniz’s claims about CICs are clearly at least partly semantic – concepts are the meanings of words for Leibniz. (Although I suppose there’s some room for doubt about whether CICs are actually the meanings of English proper names vs. the names that would be used in the idealized ‘general characteristic’). So I do think the analysis of ‘Arnauld might have been a married physician’ is part of what Leibniz is up to. In any event, Arnauld certainly would have been interested in that kind of question, and Leibniz would have known that Arnauld would be interested in that kind of question, since everyone had read the Port-Royal Logic. Part of what I’m talking about here is how Arnauld is understanding Leibniz, but I think we have some indications from Leibniz himself that Arnauld is getting his views mostly right.
Leibniz’s superessentialism is actually doing work in his theodicy. In both the Discourse and the Theodicy, Leibniz needs the claim that damnation is somehow essential to Judas, i.e. that Judas could not have existed without being damned. This is how Leibniz defends the claim that Judas has no just complaint against God: better to exist and be damned than not to exist at all, and those are the only possibilities for Judas. But he also needs another sense in which it is possible for Judas to do otherwise, and this on two grounds. First, Leibniz admits that the possibility of doing otherwise is necessary for the appropriateness of punitive (as opposed to merely deterrent/reformative) justice, and he believes punitive justice is in fact appropriate. Second, one of his oft-repeated considerations again Spinoza is that people have written novels which describe genuine possibilities that never have been and never will be actual. But surely he would admit that it is possible to do this with a historical novel that has real people as characters.
I don’t think I’m disagreeing very substantially with Adams here. His Leibniz book is one of my favorite secondary books on early modern philosophy, though it’s been a little while since the last time I read it. You might be right that I’m putting more emphasis on the semantics than Adams would like.