First: Some Key Terms Concepts.

Contingent: Something is contingent if either (i) it does in fact exist but need not have or (ii) it does not in fact exist, but it could have.

Most of the things in this world are contingent--rocks, trees, computers, desks, you, me, etc.--because each one of these things does in fact exist, but could easily have not existed (the computer you are looking at might have not been made in the factory where it was in fact made, your parents might not have met, and then there would be no you, etc.). There are also, intuitively, a host of things that do not exist but might have. Your parents might have had one more child than they actually did, and so you could have had another sibling. This sibling that does not exist but could have is also considered a contingent thing.

Contingent things are often contrasted with necessary things. Something is necessary if it does in fact exist, and could not have failed to exist. There aren't many things that people think that are necessary. Some think that God is necessary--he does exist, and could not have failed to exist. Other think that certain truths are necessary--e.g., 2 + 2 = 4 is in fact true, and it could not have failed to be true.

Possible Worlds: In class, and on this handout here, we discussed how talk of what could have been and had to have been and could never be--i.e., possibility, necessity, and impossibility--is often discussed in terms of possible worlds. If it's possible that p, then there is at least one possible world where p. If it is necessary that p, then in all possible worlds, p. If it is impossible that p, then in no possible world, p. 

So put in terms of possible world talk: A contingent being is a being that exists in at least one possible world (which may not be the actual world). A necessary being is one that exists in all possible worlds.

Gibbard symbolizes possibility by a diamond; necessity is symbolized with a box. So, for example, this statement:
 

(Again, apologies for writing out the symbol "[diamond]"; I couldn't get the symbols to post correctly.)

This statement means that a statue, s, and a piece of clay, c, are identical (e.g, s = c), but that it is also possible that--i.e., there is a possible world where--s exists and c exists but s and c are not identical. In other words, Gibbard wants to endorse the claim that sometimes an object, x, and an object, y, are in fact identical (identical in the actual world), but that they need not be (i.e., x and y are not identical in some possible world that is not the actual world).

Rea's Assumptions vs. Our Commonsense Truisms

In the introduction, Rea posited what he called The Necessity Assumption:

This is the assumption that Gibbard is dropping in his "Contingent Identity."

However, given how we formulated the material constitution puzzles by our 7 commonsense truisms, we don't have a Necessity Assumption. But we do have (7), Leibniz's Law. We have seen that Leibniz's Law can apply to spatial properties, temporal properties, and persistence conditions. It can also apply to modal properties--i.e., what is and is not possible or necessary for an object. In this way, we can subsume Rea's Necessity Assumption under our principle (7), Leibniz's Law.


Goliath and Lumpl

Gibbard makes heavy use of the Goliath and Lumpl puzzle, which was discussed on this handout here.


A Posteriori Necessity

Consider some a posteriori identity claim such as Hesperus = Phosphorus. It used to be, long ago, that people thought that Hesperus and Phosphorus were different stars: they thought that Hesperus was a star that rose in the evening, while Phosphorus was a different star that rose in the morning. In fact, Hesperus and Phosphorus were the same star, Venus. So we discovered that Hesperus = Phosphorus; it was an a posteriori identity claim.

Now, prior to Saul Kripke (1971, 1972), many thought that a posteriori identity claims were contingent. After all, it seems true that it could have been that Hesperus was not identical to Phosphorus; it seems true that 'they' could have been two distinct stars. And if so, then it is contingent that Hesperus is identical to Phosphorus.  However, what Kripke pointed out was that there is a difference between epistemic and metaphysical possibility, and that a posteriori identities are metaphysically necessary.

For consider: it could be that the astronomers were wrong and that Hesperus is indeed distinct from Phosphorus. If so, then it seems that it is possible that Hesperus (H) is not identical to Phosphorus (P). And in this way, for all we know, it is possible that H is distinct from P. Yet notice that this can only happen if we consider a situation where the actual world is one in which--unbeknownst to the astronomers--H is in fact distinct from P. But given that in the actual world, H = P, then it must be that necessarily H = P.  Kripke further showed that "[w]hether something is a necessary truth...is not a matter of how we can know it, but whether it might have been false if the world had been different..."  (p. 93-4 MC)

(We will discuss this in more detail in class.)

Kripke's Theory of Proper Names and Rigid Designators

Kripke claims that when we use proper names such as "Goliath" and "Lumpl" that these names are rigid designators--they rigidly pick out certain individuals in every possible world (in which that individual exists). And we can then track these individuals across possible worlds when we are considering the particular individuals modal properties. So, for example, when I say "George W. Bush could have been a circus clown" I mean to say of the individual, picked out by the name "George W. Bush", that he could have been a circus clown. In terms of possible worlds, this means that there is a possible world with George W. Bush being a circus clown.

For more information than you want to know about Kripke, go here.


Gibbard does not want to dispute Kripke, but he does want to maintain that sometimes--e.g., in puzzles cases such as Goliath and Lumpl--there are indeed cases of contingent identity.

Gibbard argues for the identity of Goliath and Lumpl by appealing to three reasons:


His then turns to a theory of proper names (an alternative to Kripke's theory) to account for the contingency of an identity statement such as Goliath = Lumpl. Gibbard admits that if Kripke's theory of proper names is correct, then Gibbard could not maintain his view about contingent identity. However, Gibbard proposes a different view, insisting that it (a) is a more plausible theory of proper names and (b) that it helps us to solve material constitution puzzles such as Goliath and Lumpl.

One of the key moves here is to concentrate on just what is going on when we say that some thing in one possible world is "the same thing as" some thing in the actual world. On p. 99 Gibbard says:

Moreover, he thinks that such a question only makes sense when we consider what something is designated in a certain way. He claims:

So sortals, for Gibbard, are important because they take a name like "Goliath" and make the name rigid. Once we know what kind of thing something is, then, and only then, can we track it across possible worlds and figure out its persistence conditions.

As we will see in class, David Lewis will elaborate on this Contingent Identity view, although with a slightly different (and radical) view of possible worlds.

---