How to change one's mind again
I've finally finished rewriting my paper on self-locating belief dynamics. Here is the new version. The presentation is quite different from before. In particular, I no longer use transition probabilities in modeling Cartesian agents, which allows me to be more specific about what these probabilities are. There's also a new section where I try to show that traditional arguments in support of conditioning turn into arguments for my revised rule when self-locating beliefs are considered.