Sentences, Propositions, and Beliefs (Draft)

We offer an account of beliefs as patterns of behaviour, including linguistic behaviour. Such an account of beliefs is implicit in Ramsey’s treatment of probability, but is extended here to a much wider domain. Various puzzles then appear to dissolve. What is it that we know or believe? Is it sentences? Or is it propositions? In other words, in “Jack believes that X”, what kind of object takes the place of the second argument? And is believes actually a two place relation between something and something else? The answers to these questions lie at the heart of issues like logical omniscience and Frege’s problem. Let us temporarily assume that believes is indeed a two place relation and we are not too concerned about the nature of the first argument (although in my view we should be at least in some contexts). Let us consider the nature of the second argument. If it is indeed sentences which we believe then it seems that two people who speak two different languages could not believe the same thing. And that seems unreasonable. Surely, if they are both watching a sky full of dark clouds then very probably they both believe it is going to rain and the difference in language should not matter. Moreover, someone who believes that Jack is married to Jane also believes that Jack is married. But if these are just two sentences which are believed, then nothing seems to prevent someone believing that Jack is married to Jane but is not married. (We will ignore strange situations where we might wish to say this). That seems to make a case that it is propositions which are the objects of our beliefs. Then it is possible for people speaking different languages to believe the same proposition We put aside for the moment the question of what propositions are in the first place.