Word and Objects

The aim of this essay is to show that the subject-matter of ontology is richer than one might have thought. Our route will be indirect. We will argue that there are circumstances under which standard first-order regimentation is unacceptable, and that more appropriate varieties of regimentation lead to unexpected kinds of ontological commitment. Quine has taught us that ontological inquiry—inquiry as to what there is—can be separated into two distinct tasks.1 On the one hand, there is the problem of determining the ontological commitments of a given theory; on the other, the problem of deciding what theories to accept. The objects whose existence we have reason to believe in are then the ontological commitments of the theories we have reason to accept. Regarding the former of these two tasks, Quine maintains that a first-order theory is committed to the existence of an object satisfying a certain predicate if and only if some object satisfying that predicate must be admitted among the values of the theory’s variables in order for the theory to be true. Quine’s criterion is extremely attractive, but it applies only to theories that are couched in first-order languages. Offhand this is not a serious constraint, because most of our theories have straightforward first-order regimentations. But here we shall see that there is a special kind of tension between regimenting our discourse in a first-order language and allowing our quantifiers to range over absolutely everything.2 We will proceed on the assumption that absolutely unrestricted quantification is possible, and show that an important class of English sentences resists first-order regimentation. This will lead us to develop alternate languages of regimentation, languages containing plural See Quine (1948). For an excellent discussion on quantifying over everything see Cartwright (1994). See also Dummett (1981) chapters 14-16, Parsons (1974), Boolos (1998b) and McGee (2000).