A Friendly Amendment to the Standard Analysis of Supposition

On what I will call the standard analysis of supposition, when an agent supposes that φ, she temporarily adds the proposition that φ to the set of propositions she accepts, and goes on to reason in the way that she would if she believed that φ. This kind of analysis is given in a number of places in the literature on the norms governing belief revision. I think that this picture is mostly right, but I also think it needs what I take to be a friendly amendment. In this paper I will present the standard analysis, introduce a problem for it, and propose a revision to the standard analysis that solves the problem. I will then argue that my analysis captures some important facts about the way supposition works. Finally, I will argue that my analysis gives us a straightforward defense of the Ramsey test against an interesting challenge presented by certain indicative conditionals. Here is a short introduction to the problem I will discuss. We are all familiar with MooreParadoxical sentences like “It’s raining but I don’t believe it.” One thing that’s remarkable about such sentences is that although they are often true, we cannot appropriately assert them or believe them to be true (except in some very special circumstances). But we can suppose that they are true. For example, suppose I have some inconclusive evidence that it’s raining, and I wonder whether I should accept that it’s raining. I might say to myself