Friday, May 8, 2015

Probability and normativity

The Born rule is a central part of quantum mechanics that tells us that the probability of a particle detector detecting a particle in a region U is equal to ∫U|ψ(x)|2dx, where ψ is normalized.

What exactly "probability" in the Born rule means depends on the particular interpretation of quantum mechanics. On some interpretations (e.g., on some interpretations of collapse interpretations) it will be a physical propensity and on others (e.g., Bohm and some versions of Everett) it will be something like a frequency. But any adequate interpretation of quantum mechanics needs to generate predictions, and hence needs to tell us what rational credences there are: what credences agents should assign to outcomes.

This means that a criterion of adequacy on an interpretation of quantum mechanics is that the Born rule must be understood in such a way that if a rational agent believes the Born rule, she should assign credences in accordance with it. There needs, thus, to be a bridge between the "probability" of the Born rule and the credences an agent should have.

The simplest version of the bridge is identity: take "probability" to mean an appropriate conditional rational credence. If that's done, then quantum mechanics is directly a normative theory: it tells us what we should believe.

On other interpretations, however, serious epistemology is needed to move from the probability in the Born rule to the credences an agent should have. For instance, we may need a version of the Principal Principle. This serious epistemology is normative: it is about the credences an agent should have.

Thus, either quantum physics includes normative claims or it needs further normative claims to generate predictions. (And there is nothing special here about quantum physics.)

No comments: