Quine is writing in response to Carnap, but Quine's article does more stage setting, and so we'll read him first. The positions of Carnap and Quine on ontology are essentially the only ones anyone subscribes to today.

The two views are rivals, regarded as extremely different.

• Ground clearing

• • Plato's beard: It seems impossible to deny the existence of something without also affirming it:

Pegasus exists.

Pegasus has wings.

Pegasus does not have a trunk.

Pegasus does not exist.

Anything that has a property, exists. To say of Pegasus that he "doesn't exist" is to attribute a property to Pegasus, and hence to say that he exists (at least, according to the view Quine is presenting in order to refute it).

There is a long-standing debate about whether existence is a property essentially for this reason.

• • Ideas

• • Unactualized possibles

• Semantic ascent: Instead of talking about a problem, talk about how we talk about the problem.

• • Name: Russell already solved this problem for "definite descriptions," like "the present king of France. We know what to do with

The winged horse Bellerophon rode does not exist

The internal version (The ... has the property of not existing) is false, but the external version (It is not the case that the ...) is true.

If names were definite descriptions, we would have dissolved the problem called Plato's beard. We can arrange that:

Say that x Pegasizes if and only if x is Pegasus. Now we can replace Pegasus with the definite description The x that Pegasizes. End of problem.

Never use names but only descriptions when discussing things whose existence is in doubt.

• • "Attributes" (properties, classes, …)

There are red houses, sunsets, fires, therefore, there is something they have in common, namely, the attribute red.

This argument is parallel to the Plato's beard trap: we treat the word "red" as if it is a name, and then infer its existence from its meaningful use. How can we avoid the trap? There are red things, houses, sunsets, and fires, and some of the houses, sunsets and fires are among the red things.

What might the attribute of redness be? Following the paths discussed above, there are two possibilities:

• • the red idea (meaning)

• • the class of all possible red things in all possible worlds
    

• Ontological criterion (p. 9):

"Existence is what the existential quantifier expresses.

Example: To the extent we are committed to the truth of the sentence "There is a winged horse," we are committed to the existence of winged horses. In a formalized language Quine likes, that sentence would be, "%$\exists x(Hx\land Wx)$%.

• How to answer ontological questions (back to first part):

Our acceptance of an ontology is, I think, similar in principle to our acceptance of a scientific theory, say a system of physics: we adopt, at least insofar as we are reasonable the simplest conceptual scheme into which the disordered fragments of raw experience can be fitted and arranged.

For Quine, metaphysics (ontology) is contiguous with science. Our language (conceptual scheme) is not freely chosen by us but constrained by experience: we adopt a way of talking about things that makes them as easy as possible to understand.

-- ShaughanLavine - 28 Aug 2008