Högre seminarium i teoretisk filosofi: Peter Pagin
Belief and communication de se
There is a gap between objective knowledge and "perspectival" knowledge. Thought experiments have indicated that a subject X can know all objectively true propositions about the world without knowing who X is, where X is, or when X is (what time it is). Belief de se is belief that is essentially about oneself (I-beliefs). Two main types of solution have been proposed to bridge the gap; either to supplement belief content with something else, such as belief states, or else change the very model of belief content. David Lewis was the first to propose the second, with his idea of centred-worlds propositions. It was later pointed out by Robert Stalnaker that Lewis's model leads to problems for communication. Several people tried to save Lewis's discussion by modifying it. This was intensely discussed in the 2010s. In an earlier paper (Pagin 2016) I criticised these attempts and concluded that they all (at the time) fail.
Here I shall instead propose an alternative solution to the gap-bridging problem. This solution agrees with Lewis's in characterising de se as a feature of content. It uses classical propositions, unlike Lewis's. It makes use of two additional tools. The first is that of doxastic worlds, which can characterise a subject's belief content but need not be metaphysically possible. The second is that of a model of communication that allows communicative success when contents are similar but not identical.
Senast uppdaterad: 2026-04-09
Sidansvarig: Department of Philosophy