- The ultimate test of a semantical theory is its ability to accommodate our pre-theoretical beliefs about language. Since our intuitive beliefs may not be sufficiently precise and consistent to be handled by a single coherent theory, we may have to give up some pre-theoretical beliefs in return for a theory compatible with other (presumably, more fundamental) pre-theoretical beliefs. Recent Millian theories of proper names, for example, are said to be more successful in their application to modal1 contexts (e.g., 'Aristotle might not have been a philosopher') than in their application to epistemic contexts (e.g., 'Sally believes that Aristotle was a philosopher'). But there has been no systematic exploration of the consequences of applying such theories to epistemic contexts; in particular, it has not been recognized that a large number of counterintuitive results of Millian theories fall into a single category – a category I call 'underprivileged access' – which arises from certain tensions between assent to a sentence and belief in the proposition expressed. And when the epistemic intuitions violated by Millian theories are seen to be of a piece, they appear to be much more fundamental than when they are taken to be relatively independent of each other.
- Kripke and Donnellan are largely responsible for the broad acceptance of Millian views of proper names. On these views, proper names in typical uses have no connotation or predicative element; the sole semantical function of proper names is to introduce their referents into the discourse. The following remark of Kripke's catches the flavor of the Millian view: Given that Cicero and Tully are the same person,
If Mill is completely right, not only should 'Cicero was lazy' have the same truth value as 'Tully was lazy', but the two sentences should express the same proposition, have the same content. Similarly, 'Cicero admired Tully', 'Tully admired Cicero', 'Cicero admired Cicero' and 'Tully admired Tully', should be four ways of saying the same thing.
- Kaplan has developed a similar view of demonstratives: demonstratives are directly referential in that they desiginate their referents without intervention of Fregean senses as meaning. Both Kaplan and Kripke distinguish between a term's being used rigidly and its being used referentially. Although all referential uses of terms are rigid, the converse does not hold: a use of a definite description, e.g., 'the square root of 9', is rigid but not referential since its referent is determined by its satisfying the associated condition. It is with referential uses of terms – ie., rigid uses which directly refer, without mediation of associated properties – that we are concerned here. (In III, I extend the argument to other views.) I use the term 'Millian' in an extended way to apply not only to views of proper names like Kripke's but also to views of demonstratives like Kaplan's. A view is Millian, in my usage, if it allows the semantical role of names and/or demonstratives to be exhausted by their designating their referents.
- On Millian views, the propositions expressed by sentences containing names or demonstratives in their typical uses are singular propositions: they have as constituents the very objects referred to by the sentences expressing them. The application of these Millian theories of reference to epistemic contexts entails that belief, in some cases, is to be construed as a relation between a believer and a singular proposition.
- When we apply a Millian theory to a belief context, it turns out that there is an important sense in which we can neither say what we believe nor be aware of what we believe, on pain of inconsistency. This is a result of a paradoxical disparity between what a person would sincerely and comprehendingly say that she believes (in direct discourse), and what beliefs others may correctly attribute to that person (in indirect discourse, construed broadly to include propositional contexts following cognitive or linguistic verbs). The effect of such tension between direct and indirect discourse is that we can truly be said to have beliefs that we sincerely deny having. Hence, if Millian theories are correct, then the intuitive connection between assent and belief is lost, and we have a peculiar kind of underprivileged access to our own beliefs.
Text Colour Conventions (see disclaimer)
- Blue: Text by me; © Theo Todman, 2019
- Mauve: Text by correspondent(s) or other author(s); © the author(s)