Abstract

In this dissertation, I explain, develop and defend a theory of belief ascriptions, which was introduced by Scott Soames and Nathan Salmon, from a type of problem concerning certain unintuitive semantic results of the theory. I call the theory 'Soames's and Salmon's Theory of Belief Ascriptions,' or 'SST' for short. SST provides both a semantics and a pragmatics for belief reports. The former is called 'Russellianism,' and the latter is called 'the Theory of Belief Report Pragmatics.' ;In the first chapter of this dissertation, I present Russellianism, and various versions of the problem of unintuitive semantics, which can be summarized in the following manner. According to Russellianism, given that 'Cicero' and 'Tully' are coreferring proper names, the following belief ascriptions Marion believes that Cicero is an orator, Marion believes that Tully is an orator, express the same proposition in all contexts. Hence, their truth values have to be identical in all contexts. However, it seems intuitively possible that in some context the truth values of the two belief ascriptions will differ. ;If so, then Russellianism is at least guilty of failing to account for our semantic intuitions about the truth values of belief ascriptions. ;In the second chapter, I present, and criticize Mark Crimmins's Theory of belief reports, which is a strong alternative to SST, for it is supposed to provide an intuitive semantics. ;In the third chapter, I introduce, and develop SST's pragmatic component. In the final chapter, I show SST's solutions to the various versions of the problem of unintuitive semantics, which are presented in the first chapter, with the help of its pragmatic component