On one use of the term, "propositions" are objects of assertion, what successful uses of declarative sentences say. As such, they determine truth-values and truth conditions. On a second, they are the objects of certain psychological states (such as belief and wonder) ascribed with verbs that take sentential complements (such as believe and wonder ). On a third use, they are what are (or could be) named by the complements of such verbs. Many assume that propositions in one sense are propositions in the others.
After some decades of skepticism about the worth of positing propositions, the last quarter of the twentieth century saw renewed interest in and vigorous debate over their nature. This can be traced in good part to three factors: the development in intensional logic of formal models of propositions; (not altogether unrelated) attacks on broadly Fregean accounts of propositions; and a spate of work on the nature of belief and its ascription.
"Possible-worlds semantics" is a collection of methods for describing the semantical and logical properties of expressions such as necessarily ; these methods developed out of work done by Saul Kripke, Richard Montague, and others in the 1960s. It illuminated the logic and semantics of modal terms such as necessarily, of conditionals and tenses, and other constructions as well. In such semantics one assigns a sentence a rule that determines a truth-value relative to various "circumstances of evaluation" (possible worlds, times, whatever); a sentence such as "it is necessary that S" has its truth-value determined by the rule so associated with S. The success of such accounts made it natural to hypothesize that propositions, qua what is named by expressions of the form "that S," could be identified with such rules—equivalently, with sets of circumstances such rules pick out.
Such a conception of proposition provides too crude an account of objects of belief or assertion: It implausibly makes all logically equivalent sentences express the same belief and say the same thing. A partial solution to this problem supposes that propositional identity is partially reflected in sentential structure, taking propositions themselves to be structured. Given the working hypothesis that a proposition's structure is that of sentences expressing it, critical to determining the proposition a sentence (use) expresses are the contributions made by sentence parts (on that use).
Gottlob Frege (1952) suggested that associated with names and other meaningful expressions are "ways of thinking" or senses of what the expressions pick out; one might suppose that sense and sentence structure jointly determine proposition expressed. Sense, in the case of names and other singular terms, has standardly been taken to be given by describing how one thinks of the referent. For example, the sense of "Aristotle" for me might be given by "the author of the Metaphysics "; if so, my uses of "Aristotle taught Alexander" and "the author of the Metaphysics taught Alexander" would, on a Fregean view, express the same proposition.
During the 1970s Kripke, David Kaplan, and others argued convincingly that this view is untenable: It is obvious, on reflection, that the truth conditions of the assertion or belief that Aristotle was F depend on Aristotle in a way in which the truth conditions of the assertion or belief that the author of the Metaphysics was F do not. So either ways of thinking are somehow tied to the objects they present (so that the way I think of Aristotle could not present anything but Aristotle), or the contributions of expressions to propositions must be something other than senses.
The success of accounts of intensional language that ignored sense in favor of constructions from references, along with the apparent failure of Fregean accounts, led in the 1980s to debate over the merits of what is variously called direct-reference theory, Millianism, and (neo-) Russellianism, espoused at various times by a wide variety of theorists including Kaplan (1989), Mark Richard (1990), Nathan Salmon, and Scott Soames (1988). On such views sense is irrelevant to individuating a proposition; indeed, it is irrelevant to semantics. In particular, what a name contributes to a proposition is its referent: The proposition that Twain is dead is the same singular proposition as the proposition that Clemens is.
Neo-Russellians identify the object of assertion and the referent of a "that" clause with a Russellian proposition. They allow that there is such a thing as a "way of grasping" a proposition and that belief in a singular proposition is mediated by such. Against the intuition that, for example, A: Mo believes that Twain is dead, and B: Mo believes that Clemens is dead, might differ in truth-value, direct-reference accounts typically suggest that a pragmatic explanation is appropriate. Just as an ironic use of a sentence can convey a claim without literally expressing it, so a sentence about Mo's beliefs might convey information about Mo's way of grasping a singular proposition, without that information being part of what the sentence literally says. If this is so, intuitions about A and B are explained pragmatically.
Those unhappy with this account of propositions have looked elsewhere. Many accounts of propositions identify the proposition determined by S with some construction from linguistic items associated with S and the semantic values of S's parts. James Higginbotham has identified the referents of "that" clauses with phrase markers that may be annotated with referents; Richard has suggested that the referent of a "that" clause be identified with something like the singular proposition it determines paired off with the sentence itself. In making linguistic items constitutive of propositions, these views run counter to ones, like Frege's and Bertrand Russell's, that closely tie meaning and synonymity to propositional determination. On linguistic views of propositions the synonymity of groundhog and woodchuck does not assure the identity of the proposition that groundhogs are pests with the proposition that woodchucks are. Other theorists (Gareth Evans, for example) have attempted to revive a version of Frege's views of propositions.
Many philosophers continue to doubt the utility of positing propositions. Quineans argue that meaning and reference must be determined by behaviorally manifest facts but that such facts woefully underdetermine assignments of meaning and reference; they conclude that there is nothing about language that need or could be explained by positing propositions. Stephen Schiffer has argued that propositions are a sort of "linguistic posit": that we accept nominalizations of the form "that S" as referring to singular terms and have coherent criteria for using sentences in which those terms occur is itself sufficient for its being true that there are propositions. Such a deflationist view implies neither the possibility of a substantive account of propositions (on which, for example, the proposition expressed by a sentence is compositionally determined), nor that propositions play a substantive role in explaining semantic phenomena.
Evans, G. The Varieties of Reference. Oxford: Clarendon Press, 1982.
Frege, G. "Uber Sinn und Bedeutung." Zeitschrift fur Philosophie and Philosophische Kritik 100 (1892): 25–50. Translated by P. Geach and M. Black as "On Sense and Reference." In Translations from the Philosophical Writings of Gottlob Frege, edited by P. Geach and M. Black. New York: Philosophical Library, 1952.
Horwich, P. Truth. Cambridge, MA: Blackwell, 1990.
Kripke, S. Naming and Necessity. Cambridge, MA: Harvard University Press, 1980.
Lewis, D. "Attitudes de dicto and de se." Philosophical Review 88 (1979): 513–543.
Quine, W. V. O. Word and Object. Cambridge, MA: MIT Press, 1960.
Richard, M. Propositional Attitudes. Cambridge, U.K.: Cambridge University Press, 1990.
Salmon, N., and S. Soames, eds. Propositions and Attitudes. Oxford: Oxford University Press, 1988.
Schiffer S. "A Paradox of Meaning." Noûs 28 (1994): 279–324.
Stalnaker, R. Inquiry. Cambridge, MA: MIT Press, 1984.
Mark E. Richard (1996)