Cookies: sometimes or anytime?

Over at the only official blog of Clayton Littlejohn, the eponymous author has an interesting post about biscotti (you know, the Italian cookies). "Biscotto," as I might have noticed had I been paying attention, translates as "twice-baked"; but "on Aristotelian grounds," he says, Clayton objects to what might seem the natural way of understanding this:

Contrary to what some cooks might say, Biscotti are never baked. [...] The Biscotti are numerically distinct from the hunks of dough which are the twice-baked objects. Biscotti needn't be baked even once. Similarly, one needn't toast toast to make toast, one should instead toast bread.

Although metaphysics, Aristotelian or not, is not my long suit, to me this just looks like a mistake. Biscotti aren't "numerically distinct from the hunks of dough which are the twice-baked objects" -- they're numerically identical with them. They're qualitatively distinct, having undergone a qualitative change in the oven (which is after all what it's for: to affect things qualitatively, e.g., making them crunchy, by heating them). You put the hunk in the oven, and it -- the same entity -- changes from uncooked to cooked (or cooked to recooked). The successive cookie-stages have to be of one numerically identical object or we can't see that thing as having undergone a qualitative change.

What seems to be troubling Clayton is this: Are the hunks "biscotti" on the way in? After all, they haven't been cooked yet, and how is it a "bis-cotto" if it hasn't been cooked even once? Doesn't its identity as a biscotto depend on a future contingency which might very well never happen? My answer: Well, it depends. What am I asking when I ask "is that a biscotto?"? Maybe I mean "I heard biscotti are really good with coffee -- can I try one [reaching for it]?" (Answer: no, not yet; wait til they're done.) But maybe I mean "I get my Italian goodies mixed up -- are those sfogliatelle or biscotti?" Here I don't see any reason not to say: they're biscotti (not: they'll be biscotti after they come out of the oven). In fact, they'd be biscotti even if they never get baked. Similarly, although this case is a bit different, when on CSI they take all the goop out of the victim's stomach, and someone asks "what's that?", a perfectly good answer is: "That's a biscotto" (not: it was part of a biscotto until she ate it (the biscotto); now it's a half-digested, non-crunchy mass).

We may be tempted to say that "it's a biscotto" in these cases is elliptical (i.e. for "that which will be a biscotto once it's cooked," staying with the first case) and not literal -- that ontologically speaking, it isn't really a biscotto yet. But I agree with Austin (on my reading) that to say this misconstrues what we are doing when we refer to things. (One reason I mention this is because in a subsequent post, Clayton tells us that he just got Austin's Philosophical Papers -- a righteous purchase). Everyone should also check out Cavell's chapter on "What a thing is (called)" in The Claim of Reason.

More on that in a bit. Let me explain Clayton's worry a bit more. A commenter, ian, seems to start to say what I just did, but ends up someplace else. Referring, by "D1", to "the packets of space time that at some point undergo the baking process," ian says:

When you say "Biscotti are twice-baked", you mean "Biscotti are D1 that have been baked twice." D1 is the subject of the predicate "twice-baked", not the biscotti. Clearly, to make biscotti you don't take something that is both D1 and biscotti and bake it again. You take something that is both D1 and [hunks of dough] and, by baking it, make it D1 and biscotti.

ian finishes his comment by remarking:

Now I get why some people think philosophers are jerk offs.

Yes. Well. (What, only some people think this?) I certainly agree that it sounds weird to say that to make biscotti you take biscotti and bake them (twice, no less). But that's because to do so would be to use the term in two different ways at the same time, deliberately inviting confusion, for no reason. That would be the act of a jerkoff. But we don't have to say this in order to preserve the idea, as ian's version at least succeeds in doing, that there was some thing that went into the oven and was numerically identical with what came out. On the other hand ian says (I didn't quote this part) that this is thing is D1, the "packets of space time," but not the hunks of dough therein, which confuses me. Clayton, on yet another hand, is

a bit worried about the pair of claims that (a) D1 is the biscotti but (b) D1 but not the biscotti is the subject of predication. This suggests that when it comes to predication, the truth of a predicative judgment depends on what sort of description the speaker is thinking of the subject under. This will create all sorts of untold havoc when we think about identity.

Here I agree that if D1 is the biscotto then the biscotto is the subject of predication, which means that biscotti can be uncooked, which is fine with me (in theory, that is, as above). With respect to havoc untold, however, it is important that we lay the blame at the proper door. Said havoc results, in my view, not from the perfectly straightforward idea that the truth of a judgment depends on what (i.e. not simply what thing, under no particular description) the speaker is talking about. It is only when this is combined with tendentious (if also traditional) metaphysical assumptions that havoc results. The identity in question here is the identity of an object, and what objects are -- the only things they could be -- are the referents of our concepts, as used in judgments. An object is what it is in being that thing of which the judgment that it falls under such-and-such a concept is true. And, correspondingly, concepts get their content from their use in judgments about the things they (are used to) refer to in (appropriate generalizations across sets of) particular cases.

Here I may sound more like Davidson (or Davidson as modified by McDowell) than Austin (or Austin as modified by Cavell). But Cavell is working the same side of the street. In the chapter I mentioned above, he suggests that one of his targets is "[a] radical distinction between what is a question of language and what a question of fact" (p. 67); or, in Davidsonian terms, the scheme/content dualism. And of course the context of this quotation is a discussion of Wittgenstein's use of the concept of "criteria" -- which recalls my earlier post on the centrality of the concept of (acts of) judgment for understanding the relation (i.e. overcoming the traditional metaphysical dualism, of which Cartesian substance dualism is only the surface manifestation) of mind and world.

My guess is that this suggestion -- that it makes no sense to consider the properties of an "object" under no particular description -- must sound like idealism, especially in this post-Kripkean era in which "essentialism" (let alone "realism") is no longer a bad word. And in the post-Kantian context which McDowell and his (better) critics (Brandom, Andrew Bowie) operate, this may not even be a bad thing, even if only for historical reasons. But if the mind/world dualism is our target, surely so must be the realism/idealism one as well. If we could ever agree even on this much, then I would be happy to hand cookies all round. But I'm not holding my breath.