a. In opposition to Bertrand Russell’s view that the meaning or at least the referent of a name is determined in part by some associated definite description (the "descriptivist" account of reference).

b. Kripke claims that this flies in the face of our modal intuitions (our intuitions about what is necessarily the case or could possibly be the case).

For any co-referring definite description (the inventor of bifocals, the most accomplished student of Plato) and name (Ben Franklin, Aristotle), it is possible for the two not to refer to the same individual. Someone other than Ben Franklin could have invented bifocals, and Aristotle might not have studied under Plato (or someone else had been more accomplished).

So while it is necessary that Ben Franklin is Ben Franklin, it is not necessary that Ben Franklin is the first postmaster general or the inventor of the bifocals.

c. Kripke accounts for these intuitions be appealing to the machinery of "possible worlds." [As an aside, Kripke first rose to philosophical prominence by showing how such machinery could be employed to systematize the semantics (or "logic") of modal statements.]

i. According to Kripke, names "rigidly" refer to their referents; they refer to this referent in every possible world that that individual exists.

ii. Definite descriptions, by contrast, do not refer rigidly. They may refer to different individuals in alternate possible worlds.

iii. The "causal theory" of reference: names refer to their referents by being causally connected to them in an appropriate fashion. A particular token (use) of a name will belong to a causal chain of such tokenings (usages), which ultimately terminates in an "initial baptismal" event in which the referent is present.

- Different individuals might share acoustically similar names. Nevertheless, separate initial baptisms have created different causal chains. We can pick out the proper referent of a particular occurrence or use of a name by looking to the terminus of the causal chain to which that occurrence belongs. Occasionally, causal chains can get entangled with one another, in which case it will be hard to determine exactly who or what uses of that name refer to.

- "Empty" names can be thought of as those belonging to causal chains for which no particular referent can be determined. Perhaps no appropriate referent was present during the initial baptism.

- Our intuition that different names have different meanings (or senses) arises because different names for the same individual can be part of separate causal chains. They might trace to distinct initial baptisms.

A. Identity statements between names and definite descriptions (as well as between definite descriptions) are contingent; the definite description could have picked out a different individual in an alternate possible world.

"The first postmaster general of the US" or "The inventor of bifocals" might have referred to someone other than Ben Franklin.

B. However, and somewhat surprisingly, this treatment of names implies that identity statements between co-referring names turn out to be necessarily true.

1. Recall that names refer rigidly; they refer to that (same) individual in every possible world where that individual exists.

2. So there is no possible world in which a name will refer to one individual, while a co-referring name will refer to another (or to nothing at all).

3. Still, such identities are a posteriori; Their truth, although necessary is still a matter of empirical investigation. One might well wonder whether distinct causal chains terminate in the same individual.

4. Modal distinctions are metaphysical; as such, they need not line up with epistemological distinctions. Kripke is thus sometimes credited as "discovering" a posteriori necessities.

PAUSE: Must we accept Kripke’s account of rigid designation, and with it, the counterintuitive conclusion that many a posteriori identities are nevertheless necessarily true?

C. Extending the point to names for properties and substances:

1. Consider :

"Water is hydrogen hydroxide (H2O)."

"Heat is molecular motion (mean molecular kinetic energy)."

"Lightning is atmospheric electrical discharges."

Each also express identities between names for substances or physical phenomena. As such, these referring expressions rigidly refer. Thus they too are necessary identities.

2. But these statements seem even more contingent than identities between names of individuals. What, then, accounts for their additional layer of apparent contingency? According to Kripke, the contingency is really between heat and its sensation. What we feel as a sensation of heat might not have been association with heat. That is, on some possible world, the sensation we actually think of as a sensation of heat might have been produced by something very different. On that same world, those sensations are not produced by molecular motion either.

[Does this Convincingly capture the contingency at issue? Why have we simply dropped talk about separate reference chains?]

A. Descartes famously argued that mental states could not be identified with physical brain states because it’s possible to conceive of the one existing without the other (and vice-versa). Most philosophers who subscribe to some sort of identification between the physical and the mental (say pain and c-fiber stimulation) would accept Descartes' starting point, but merely insist that the proposed identification is merely contingent.

B. At the end of Naming and Necessity, Kripke claims that his treatment of names and rigid designation shows that this maneuver is too quick. Identities between naming expressions (say, between pain and c-fiber stimulation) really cannot be contingent.

C. Nor can we explain away their seeming contingency in the same way that we can explain the apparent contingency of other a posteriori identifications.

1. In those other cases, we made sense of the contingency by imagining possible worlds in which heat feels differently from how it feels in this the actual world. But in this case that isn’t possible, because pain cannot be imagined to feel other than how it actually does feel. Pain’s feeling (its phenomenal quality) is what pain is.

2. So the identity thesis, at least among physical and phenomenal states, cannot be made to work (although it is open whether it might work in the case of non-phenomenal mental states).