Against Self-Location

TL;DR

The paper argues that rationally compelling pure self-locating credences do not exist, challenging their use in scientific reasoning about multiverses and related hypotheses.

Contribution

It distinguishes between types of self-locating credences and critiques assumptions about their rationality and derivation from non-self-locating cases.

Findings

Pure self-locating credences are not rationally compelled.

Pragmatic goals influence self-locating credences, not rational principles.

Implications for multiverse and simulation reasoning are significant.

Abstract

I distinguish between pure self-locating credences and superficially self-locating credences, and argue that there is never any rationally compelling way to assign pure self-locating credences. I first argue that from a practical point of view, pure self-locating credences simply encode our pragmatic goals, and thus pragmatic rationality does not dictate how they must be set. I then use considerations motivated by Bertrand's paradox to argue that the indifference principle and other popular constraints on self-locating credences fail to be a priori principles of epistemic rationality, and I critique some approaches to deriving self-locating credences based on analogies to non-self-locating cases. Finally, I consider the implications of this conclusion for various applications of self-locating probabilities in scientific contexts, arguing that it may undermine certain kinds of reasoning about multiverses, the simulation hypothesis, and Boltzmann brains.