Primitive man, aware of his helplessness against the forces of Nature but totally ignorant of their causes, would try to compensate for his ignorance by inventing hypotheses about them [...] For one who has no comprehension of physical law, but is aware of his own consciousness and volition, the natural question to ask is not: "What is causing it?", but rather: "Who is causing it?" [...] The error [Mind Projection Fallacy] occurs in two complementary forms, which we might indicate thus: (a) My own imagination => Real Property of Nature, and (b) My own ignorance => Nature is indeterminate.
The philosophical dierence between conventional probability theory and probability theory as logic is that the former allows only sampling distributions, interprets them as physically real frequencies of "random variables", and rejects the notion of probability of an hypothesis as being meaningless. We take just the opposite position: that the probability of an hypothesis is the fundamental, necessary ingredient in all inference, and the notion of "randomness" is a red herring, at best irrelevant. [...] by "probability theory as logic" we mean nothing more than applying the standard product and sum rules of probability theory to whatever propositions are of interest in our problem.
We do not seek to explain "statistical behavior" because there is no such thing; what we see in Nature is physical behavior, which does not conflict in any way with deterministic physical law.
[...] as in any other problem of inference, we never ask, "Which quantities are random?" The relevant question is: "Which quantities are known, and which are unknown?"
E.T. Jaynes, Probability Theory as Logic