In “On What There Is” Quine begins with a problem. Suppose that he and another philosopher, whom he calls McX , disagree over ontology. McX holds that a certain thing, say Pegasus, exists while Quine maintains that it does not. The difficulty arises, for Quine, when he attempts to formulate the way in which his opinion differs with that of McX . If Quine tries to describe it by saying that there is a thing (Pegasus) that McX believes exists and he does not, then it seems that Quine has already affirmed what he is trying to deny: that Pegasus is. This is what McX will accuse him of. Of course, as Quine points out, McX cannot honestly believe that an actual flesh and blood winged horse exists in any region of space-time. Therefore, McX will have to say that the Pegasus that exists is “an idea in men’s minds” (133). It is clear, however, that there is a difference between Pegasus and a Pegasus idea. Surely Quine is not denying the existence of the idea of Pegasus when he says that Pegasus is not. McX is stuck.

Having dispensed with McX , Quine introduces another hypothetical colleague, Wyman. Wyman has another way of tackling the Pegasus problem. For him, Pegasus exists as an “unactualized possible” (133). When we say that Pegasus does not exist, what we are saying is that Pegasus does not have the characteristic of actuality (or existence.) Statements such as these commit Wyman to a messy ontology. Quine gives us at least three reasons why we should prefer a “desert landscape” to Wyman’s ontological slum. The first of these reasons is the issue of Pegasus. Wyman will try to save his argument by positing a distinction between existence and subsistence (134). But, then, why say that Pegasus exists at all? It seems to be an artificial move. What does it mean for something to exist as an “unactualized possible”?

The second reason that Quine offers for the rejection of Wyman’s ontology concerns the issue of identity. According to Quine, the concept of identity cannot apply to the denizens of Wyman’s slum. When we speak of a possible fat man in the doorway what do we mean? Say we posit a possible fat man and a possible bald man in the doorway. Is there just one possible man in the doorway with the attributes fat and bald or are there two possible men, a fat one and a bald one? Say we speak of two possible fat men in the doorway. The only attributes of these men are what their definitions entail; each one is a) possible, b) fat, c) a man. If their sets of attributes are exactly the same and they do are not extended in space or time, how are we to say that they are two distinct men?

The third reason that Quine offers for rejecting Wyman’s slum is that even if we allow for the existence of unactualized possibles we are still left with some non-existence statements that cannot be explained, namely, those that describe logically impossible entities such as “the round square cupola on Berkeley College” (134). If we follow the line of reasoning begun with McX , it seems that when we say that the round square cupola on Berkeley College does not exist we are saying that something is such that it is the round square cupola on Berkeley College and it does not exist. But what can this something be? Surely Wyman does not want to commit to there being a realm of unactualizable impossibles (which is what the round square cupola would have to be given that it is a logical impossibility)? If he does make this move, he can eventually be trapped in to saying strange things about these entities; like that there are some that are round and square (134). Misguided as he is, Wyman is too smart to make this move. Instead he will say that the phrase “round square cupola” is meaningless. The doctrine of the meaninglessness of contradictions, however, has its own problems (135).

Later on, Quine comments, “Whatever we say with the help of names can be said in a language that shuns names altogether” (138). Quine believes that any name that we use in common speech, Pegasus for example, can be converted to a description. When we say “Pegasus” we can express the same thought by using a bound variable and a description: “there is something which pegasizes” or “there is something which is a winged horse.” Quine says that the only way in which we involve ourselves in ontological commitments is through our use of bound variables (138). I can say “x pegasizes” without committing myself to saying that there exists anything which can stand in for the variable x. If I use a bound variable “there is something” or “there exists an x” then I am making an ontological commitment.

Quine goes on to discuss the problem of universals. He says that the medieval controversy over universals has arisen again in the modern philosophy of mathematics but we now have a more explicit standard by which we can decide whether or not a view entails commitment to universals (138). According to Quine, “a theory is committed to those and only those entities to which the bound variables of the theory must be capable of referring in order that the affirmations made in the theory be true” (138). It is not exactly clear, to me at least, what this statement means. Quine next presents the three main medieval points of view concerning universals and explains how they map up with three main doctrines in the modern philosophy of mathematics.

The first view is known as realism in the medieval terminology. Realism holds that universals (abstract entities) exist independently of the mind. While the mind can access universals, the mind does not merely create them. Quine says that medieval realism is similar to modern logicism as represented by Frege, Russell, Whitehead, Church, and Carnap (139).

The second view, conceptualism, says that although there are universals, they are created by the mind. This doctrine is similar to modern intuitionism, which has been defended by Poincaré, Brouwer, and Weyl (139).

The third view, the modern version of which is known as formalism, finds intuitionism unsatisfactory. Like the medieval nomialist, the defender of formalism may object to saying that abstract entities exist at all. Concerning mathematics, the formalist will have to say that it is a notation system that is useful but meaningless. We can use mathematical concepts to model the world of extant entities but to do this we needn’t assume that the mathematical concepts themselves are meaningful in the sense of referring to anything that exists (139).

Quine thinks that we should choose an ontology (and thus decide whether or not to affirm the existence of universals) based on the simplicity principle (140). He says that our acceptance of an ontology should be, in principle, similar to our acceptance of a scientific theory: “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” (140). Put differently, we should posit only as many entities as are needed to explain experience. If we can explain every phenomenon (including the way in which we speak about things like Pegasus) without appealing to the existence of universals, then we should not posit them.

Quine, however, openly admits that simplicity can be an ambiguous concept. For example, in one sense the phenomenalistic conceptual scheme is simpler than the physicalistic conceptual scheme. We already accept that there are phenomena (all that is required to accept a phenomenalistic conceptual scheme) but to offer a physicalistic account we need to posit not only phenomena but also physical objects. Thus, the former seems, at first glance, simpler. However, the physicalistic conceptual scheme simplifies our account of experience in that it explains why various different sense impressions (phenomena) come to be associated with what seem to be single objects (140). Quine says, then, that physicalism is “a convenient myth” from the point of view of phenomenalism. In much the same way, he says that an ontology that includes classes or attributes of physical objects can be seen as a convenient myth from the point of view of a strictly physicalistic conceptual scheme (140).

Although he is careful to state that the question of which ontology to accept remains an open one, Quine concludes that of the conceptual systems that he presents, the phenomenalistic is better from an epistemological point of view (141). If one accepts the phenomenalistic account then the ontologies of physical objects and mathematical objects are myths (141). I presume that Quine thinks that the ontologies of classes or attributes are myths on both the phenomenalistic and physicalistic accounts.

-- AnnieS - 10 Oct 2005