Page 380 - Encyclopedia of Philosophy of Language

Formal Semantics
alternatives are present. On the other hand, the names must be unique in order to preclude hidden incon- sistencies (Mister X named as 'Peter' is blond but named as 'Paul' is not). A convenient way to realize the appropriate naming is to let the objects themselves figure as their own names. Similarly, the properties and relations among the objects can be taken as 'relation signs.' Note in passing that this strategy is quite like the one used by Barwise and Perry to obtain abstract situations. The strategy undermines two objections to state-descriptions, namely: (a) that worlds are unlike sets of sentences; and (b) that maxi- mal consistent sets are not numerous enough. In the abstract, the worlds are made up of real objects and real relations. And using the objects and relations themselves removes cardinality and other restrictions which are sensible for languages (e.g., that the language is countable). Still, this liberal version of state-descriptions, far removed from Carnap's orig- inal proposal, has to face a serious problem: it takes modality as a primitive. It does so mainly via consist- ency. A set of sentences is maximal if and only if adding new sentences makes it inconsistent (i.e., not consistent). But it is consistent if and only if its sen- tences could all be true. In short, this way of explaining
modal phenomena achieves at most a reduction of all modal phenomena to a modality of a particular kind (the one taken as primitive). Whether or not this reduction is substantial depends on a further analysis of the primitive notion.
An important virtue of the proposals that identify possible worlds with concrete particulars or with maximal consistent sets is that they are specific. Often one has to make do with less detail. For instance, Kripke (1980:20) uses the image that 'possible worlds' are little more than the miniworlds of school prob- ability blown large. And in the moderate realism of Stalnaker their existence is 'inferred or abstracted from the activities of rational agents,' so that claims about possible worlds should be located 'in a theory of rational activities' (Stalnaker 1984: 50-55). His realism 'need not take possible worlds among the ulti- mate furniture of the world.' They are useful primi- tives in theorizing about activities at a certain level of abstraction. What the abstract worlds are may depend on the context, and especially on the kind of activity considered (1984:57-58). Now, these statements point to interesting ways of clarifying the use of possible worlds, but as they stand they are rather pro- grammatic (and are intended as such). Among other things, they leave in the dark how a state, an unan- alyzed 'way things might have been' is related to the
one real world, that is, what it is that makes a state not merely possible, but actual. Lewis on the 'Ersatzist Programme' (1986: §3) has an extensive discussion of the issues at hand. He also discusses other alter- natives.
5. Occam's Razor
One option is to do without possible worlds and limit ourselves to less troublesome individuals. This most radical use of Occam's razor would deprive us of a useful tool in describing semantic phenomena. After all, the metaphysics necessary to explain natural language may involve the use of entities which are suspect for the critical philosopher. For want of an analysis which is as perspicous and successful as poss- ible world semantics one has to choose among the modal realism of Lewis and the 'moderate' versions of realism. The question is how to apply the razor in this case; whether it should cut the number of entities posited, or only the number of kinds of entity. In both cases, Lewis seems to be better off. Although his worlds appear grandiose, they are all of the same kind. In contrast, the moderate realist has to distinguish between the real world and abstract world rep- resentations. And since the moderate realist will want his distinctions to be as fine grained as those of Lewis, he needs to have as many representations as Lewis needs worlds (see Lewis 1973: 87). The main virtue of the moderate realist position appears to be that the use of abstractions leaves the possibility that these may be eliminated in favor of more 'respectable' enti- ties. In the absence of such an analysis, the case is not so clear.
See also: Modal Logic.
Bibliography
Carnap R 1956 Meaning and Necessity, 2nd edn. University of Chicago Press, Chicago, IL
Gribbin J 1985 In Search of Schrddinger's Cat. Corgi Books, London
Hintikka J 1969 Models for Modalities. Synthese Library 23, Reidel, Dordrecht
Kripke S A 1980 Naming and Necessity. Harvard University Press, Cambridge, MA
Leibniz G W 1991 Monadology. Routledge, London
Lewis D 1973 Counter jactuals. Blackwell, Oxford
Lewis D 1986 On the Plurality of Worlds. Blackwell, Oxford Stalnaker R C 1984 Inquiry. Bradford Books, MIT Press,
Cambridge, MA
Voltaire 1947 (trans. Butt J) Candide or Optimism. Penguin
Books, London
358