Illusory Chance

Despite concerted efforts by Carnap and others, including recently by Jon Williamson in an objective Bayesian setting, no satisfying philosophical justification for learning from experience has been produced thus far. I contend that finitude constraints imposed by our world on even otherwise-ideal agents suggest a new epistemic principle, Illusory Chance, which provides justification for learning in some settings. Some other benefits provided by this principle include a generalization of—and apparent improvement on—Bayesian decision theory, substantial progress toward resolving the venerable St. Petersburg paradox, and the revelation of substantive flaws in key arguments for probabilism.