Dynamic rationality
The standard dynamic norm of Bayesianism, conditionalization, is clearly inadequate if credences are defined over self-locating propositions. How should it be adjusted?
This question was popular at around 2005-2015. Chris Meacham and I came up with the same answer, which we published in (Meacham 2010), (Schwarz 2012), and (Schwarz 2015). I showed that the replacement norm that we proposed has all the traditional virtues of conditionalization. For example, (under the usual idealized conditions) following the norm uniquely maximizes expected accuracy, and an agent is invulnerable to diachronic Dutch books iff they follow the norm.
Nonetheless, I don't think we've convinced anyone that it is the correct norm. Why not?
In discussions, the reason soon became clear to me. It's that most philosophers don't seem to believe in dynamic norms of rationality at all. Rationality, they think, is a matter of proportioning one's beliefs to the present evidence. Past beliefs are relevant only insofar as they are part of the present evidence. There are no norms that directly relate past beliefs to present beliefs.
I don't know if this is a recent trend. How did conditionalization come to be part of the Bayesian orthodoxy, if philosophers never believed in dynamic norms?
To me, the anti-dynamic view has always looked extremely implausible. According to my preferred model, rational agents have an internal representation of the world that gets updated in response to sensory stimulation. The sensory stimulation provides the new evidence. It would obviously be insane to think that rationality is only a matter of proportioning one's beliefs to one's present sensory input. You'd have to suspend most beliefs about your physical surroundings whenever you close your eyes.
Philosophers who reject dynamic norms in favour of the evidentialist doctrine (that one should proportion one's beliefs to the present evidence) must have a different conception of evidence. Unfortunately, they rarely explain what it is. Still, it's possible to evaluate the proposal.
That's what I do in (Schwarz 2025a) (local version), which is now forthcoming in Philosophers' Imprint. I argue that any version of the evidentialist view leads to problematic consequences. I also argue that the usual arguments for the evidentialist view are unconvincing and confused.
There's a little overlap between this paper and (Schwarz 2025b) (local version). The two papers used to be a single paper, which I've unsuccessfully tried to get published since 2013. I'm glad they've finally found a good home, even though the topic has by now fallen out of fashion.