Necessity vs apriority: are necessary truths always knowable a priori, or do they come apart?

Saul Kripke, answered from the texts and cited to the page.

The position in brief

Necessity and apriority are different notions, with different subject matter, and they need not coincide. Necessity is a metaphysical notion — a proposition is necessary if it could not have been otherwise, if it is true in every possible world. Apriority is an epistemic notion — a proposition is a priori if it can be known without appeal to experience.

There are necessary a posteriori truths: "Hesperus is Phosphorus" is necessary (the two names rigidly designate the same heavenly body, Venus, and an identity between rigid designators is necessary if true) but the truth was discovered empirically by astronomers.1 "Water is H2O," "gold is the element with atomic number 79," "heat is molecular motion" — all necessary a posteriori.

There are also contingent a priori truths: if I stipulate that the length of a particular rod in Paris shall fix the reference of "one meter," then "the standard rod is one meter long" is something I know a priori in virtue of having made the stipulation, but it is contingent because the rod might have had a different length.234

Sources

  1. So it seems that we do not necessarily know a priori that an identity statement between names is true. It doesn't follow from this that the statement so expressed is a contingent one if true. This is what I've emphasized in my first lecture.
    Naming and Necessity, pp. 110–111
  2. This assertion is not trivial. The terms 'necessary' and 'a priori', then, as applied to statements, are not obvious synonyms. There may be a philosophical argument connecting them, perhaps even identifying them; but an argument is required, not simply the observation that the two terms are clearly interchangeable.
    Naming and Necessity, pp. 46–47
  3. Let us think what would be such a circumstance, using these terms here as names of a planet. For example, it could have been the case that Venus did indeed rise in the morning in exactly the position in which we saw it, but that on the other hand, in the position which is in fact occupied by Venus in the evening, Venus was not there, and Mars took its place.
    1. Identity and Necessity, p. 18
  4. More important for present purposes, though, than accepting this NAMING AND NECESSITY 57 example as an instance of the contingent a priori, is its illustration of the distinction between 'definitions' which fix a reference and those which give a synonym. In the case of names one might make this distinction too. Suppose the reference of a name is given by a description or a cluster of descriptions.
    Naming and Necessity, pp. 65–66

Where the argument is made

  • Naming and Necessity (1980), Lecture III
  • "Identity and Necessity" (1971, Philosophical Troubles Ch 1)

How the argument was received

Defended in N&N Lecture III as a central thesis of the project. The break dissolves a longstanding conflation in analytic philosophy (originating in the empiricist tradition and explicit in Quine's "Two Dogmas") between necessity, analyticity, and apriority. The standard examples — Hesperus identity, theoretical identities in chemistry and physics — are widely accepted in the post-Kripke literature, with ongoing debate about precise characterization. Soames argues in Beyond Rigidity Ch 2 that what is properly knowable only a posteriori in the Hesperus case is the metalinguistic claim "the names 'Hesperus' and 'Phosphorus' are coreferential" rather than the object-level proposition Hesperus = Phosphorus. I take Soames's worry seriously without abandoning the underlying distinction.