24.211 Metaphysics October 13, 2005 Time was when something called "contingent identity" was a staple of analytic metaphysics. The idea was that you could have an entity y and an entity z that were identical in fact, but not necessarily. This was the relation obtaining between (a) Nixon and the US President, nine and the number of planets, etc(b) statue and clay, ship and planks, etc.(c) water and H20, electrical current and the flow of electrons, etc.(d) pain and c-fiber firings, itchiness and d-nerve palpitations, etc.Start with examples of type (a), which grew out of Quine's critique of de re modality. Quine as you may know said that whether a condition φ(x) holds necessarily of x depends on how x is denoted or picked out. A good example is the condition that x is identical to so and so: x = the number of planets. This holds necessarily of the number of planets, but contingently of the number 9. The number of planets is necessarily the number of planets, but the number 9 is only contingently the number of planets. Something has gone wrong here, Kripke thinks. One can prove after all that the objects are necessarily identical if identical at all. The proof goes like this: Suppose y = z. Then by Leibniz's Law, y and z have the same properties. z has the property of being necessarily identical to z. So y has the property of being necessarily identical to z. So it holds necessarily that y = z. Here then is a philosophical puzzle: on the one hand, you've got the appearance of contingent identity, and on the other, you've got a proof that contingent identity is impossible. Solution proposed by Kripke is simple. It's the proof that's correct. Objects are nec. identical if identical at all; contingent identity is impossible. But what are we imagining when we seem to imagine y = 9 and z = the number of planets coming apart? The answer in this case seems pretty clear. We do not imagine de re of y and z that the one comes apart from the other; we imagine de dicto that the 9 should fail to be the number of planets. A certain identity-stating dictum strikes us as possibly false. But this isn’t enough for contingent identity! What is contingent is that a certain statement, which relative to this world expresses an identity between 9 and 9, is true. That’s not to say that the identity it actually expresses is contingent. So, distinguish two ways an identity statement can be contingent: first, the identity it actually expressesfails in other possible worlds; second, the identity it expresses w.r.t. some other world fails in that world. The first would be contingent identity. The second, the contingent truth of an identity statement, is so far allwe've got. Kripke’s Hypothesis: all so-called contingent identity is really just contingently true identitystatements. Now, if we have a contingently true identity statement, then it would seem that in the worlds where it isfalse at least one of the terms involved must refer to a different object than it refers to in this world. Rigidvs. nonrigid. Unless you use rigid designators, that the statement is only contingently true signifiesnothing about the nature of the identity. For contingent identity you’d need a contingent identity statementinvolving rigid designators. Are there any? First we have to ask, which designators are rigid? Definite descriptions typically aren’t.But according to Russell and others, names are disguised def. descriptions. So they ought to be nonrigidtoo! Kripke has two main arguments that they’re not. Spose “Einstein” is a disguised form of “thediscoverer of relativity.” (i) Necessity objection. Possible that Einstein didn’t do it. (ii) A priorityobjection. Should be a priori that Einstein discovered it. But what if it turns out that the discoverer was Smith. Then Smith is Einstein! For it’s apriori that Einstein was the guy who did the job. So, “Einstein” is not a definite description but a tag: its whole meaning is that it stands for that guy.Question now becomes, are there true contingent identity statements between names? Again, it wouldseem like we have an argument that there can’t be. Names are rigid, so if they refer to the same thing inany world, they do in every.So far so good. But now Kripke notices that it is not just identity statements between descriptions that canlook contingent. Identity statements between names can also seem this way. Hesperus = Phosphorus.Certainly took experience to discover. And we want to say, it might have come out otherwise. Other examples: Everest = Gaurisanker. Water = H20 Time to regroup. Contingent identity statements between descriptions don’t give us contingently identicalobjects. But contingent identity statements between names do, or rather would. Explaining these away willnot be so easy. But let's try. The argument that these statements, e.g., "water = H20", are contingent,remember, is that (i) it took experience to confirm them, and (ii) things could have come out the other way. Take the first formulation first: they are not a priori. But, must all necessary truths be knowable a priori?Terms often used equivalently but the concepts different, one from epistemology, other from metaphysics.So it’s a substantive claim, not a trivial one. Mathematical examples (Goldbach) already give grounds fordoubt. Essentialist examples clinch it. This lectern was first made of wood. This is a cat. This is gold. Second formulation trickier: it could have turned out otherwise. One is tempted to say that it could not have turned out otherwise: given that water is H20, it could not have been anything else. This suggests that"it might have turned out that P" -- in the sense in which it might indeed have turned out that water wasn'tH20 -- is not to be understood as claiming that the identity actually expressed by "P" could have failed toobtain. Need to look for another interpretation. How do names get their meanings, if as Kripke maintains their meanings are just the objects? One can separate out two questions here: where did the meaning originally come from? and how is it passed along?Initial baptism plus historical chain (Homer, Cicero, Tully); or reference-fixing description plus historicalchain (Neptune, Hesperus, Phosphorus). What could have been otherwise is not Hesperus’s relations withPhosphorus; rather the way we pick out these objects could have been a way of picking out distinct objects.. Same