Page 311 - Encyclopedia of Philosophy of Language

proposition), and (b) that there are no shifting con- textual features that cause the propositions specified in the premises to differ from those specified in the conclusion (i.e., that what proposition is specified by a description of the form 'the proposition that s* does not subtly vary with the evolving context of discourse). Since these assumptions are hard to defend without the help of well-developed theories about prop- ositions, semantics, and belief, it is problematic to use such a 'test' to provide data for such theories. Of course, any plausible accounts of these matters must give explanations for divergences from the natural, naive views that reports of prepositional attitudes straightforwardly report relations to propositions, and that specifying a proposition is a straightforward matter without a great deal of subtle contextuality. But there are independent reasons to think that subtle features of context can be crucial to the truth of prep- ositional attitude reports (see Kripke 1979; Crimmins and Perry 1989).
3.2 Prepositional Content
Another stumbling block in individuating prop- ositions is that intuitions about the identity and difference of 'what is expressed' in statements (or of 'what is believed' in cases of belief) can at times be uncertain, and so theories of semantics and prop- ositions that do not reflect these uncertainties have some explaining to do. There are at least two kinds of uncertainty worth mentioning.
First, between clear cases of identity and difference of proposition expressed, there may be grades and shades. Consider these pairs of statements: 'John and Bill have kissed Mary' and 'John has kissed Mary and Bill has kissed Mary'; 'John kissed Mary' and 'Mary was kissed by John'; 'John bought a car' and 'John was sold a car'; 'This triangle is equilateral' and This triangle is equiangular'; 'Line a is parallel to line b* and 'The direction of line a is the same as the direction of line b'; 'John shaved John' and 'John shaved himself.' Are these different ways of saying the same things, or are different but closely related things expressed?
Second, there are conflicts of reference and per- spective: for you to believe or say the same thing as I do when I believe or say that I am hungry, must you believe that 7 am hungry, or that you are hungry? In this particular case, there seem to be two dimensions of what we say that we can compare; perhaps if you are to say something with the same meaning you must say that you are hungry, and if you are to say some- thing with the same truth conditions, you must say that I am hungry. Things are more complicated, how- ever, with my statement of 'my fork is to the left of your spoon.' Here, there are several different can- didates for sentences you might use to say the same thing I have said, obtained by varying the different elements of the sentence to maintain either reference
or perspectival similarity. The general problem is that intuitions about saying or believing the same thing are sensitive both to reference and to perspective, and so in general these intuitions cannot be explained by citing just one 'thing said' in a statement or 'thing believed' in a belief.
4. Structured Propositions
Theories of structured propositions take propositions to be abstract, structured entities, typically containing individuals and properties as constituents (see Cress- well 1985; Katz 1977; Salmon 1986). One such theory would take a proposition to be a sequence containing a property or relation as its first constituent, and other entities as its other constituents. For instance, if FatherOf is a binary relation and John and Tom are individuals, then the sequence (FatherOf, John, T orn) is the proposition that John is the father of Tom. Logical connectives and quantifiers can be accom- modated in any of several ways, including by treating 'logical constants' as relations between propositions, as in (1):
(Or, (White, Snow), (Green, Grass))
(1)
The great usefulness of structured propositions in semantics comes from facts about the compositional way in which the proposition expressed by a statement depends on the objects, properties, relations, and so on expressed by its component expressions. The sen- tence 'John is the father of Tom' expresses the prop- osition just described because the use of 'John' refers to John, the use of 'Tom' refers to Tom and the use of 'is the father of expresses the relation FatherOf. By mirroring the syntactic structure of the sentences used to express them, structured propositions are nat- urally suited for playing the role of the semantic values of statements.
A problem somewhat special to this account of propositions is that it can seem to mistakenly dis- tinguish identical propositions. For instance, the proposition that two is less than three will count as different from the proposition that three is greater than two; and one might have thought these were the same, that the claims are identical. This might be remedied by a modification in which the argument roles of relations are not ordered (so that 'is less than' and 'is greater than' express the same relation), but other cases of this kind may persist in which prop- ositions are incorrectly distinguished.
5. Propositions as Sets of Possible Worlds
In possible-worlds theory (see Lewis 1986; Stalnaker 1984), a proposition is a set of possible worlds. The proposition expressed by 'John is the father of Tom' is the set containing exactly those possible worlds in which John is Tom's father. The actual world is a
Proposition
289