I’m trying to slog through Jaynes’s Probability Theory: The Logic of Science. It’s a refreshing change to read a book where I spend a good portion of the time clutching my head, feeling hopelessly stupid.
This is no fault of the book itself; it’s wonderfully written, and would probably make perfect sense to any reasonably clever person with a strong background in applied mathematics. Unfortunately, my background in applied mathematics is entirely nonexistent, so when Jaynes says something like “let’s assume that variable x is differentiable, so that all of these nice consequences follow during our big long chain of derivations and rewritings,“ I am forced to nod my head and grunt.
I took a completely worthless undergraduate statistics class, of which I remember little beyond its uselessness, so at least I haven’t been infected with the statistical orthodoxy against which Jaynes repeatedly inveighs.
My motivation for this was that I am currently dealing with large quantities of not entirely reliable information, and I have basically no intellectual tools for estimating the quality of my data, much less fixing it up. I hate this feeling of innumeracy, but I hadn’t previously had a compelling reason to do much about it.
Even though the Jaynes book makes me feel like a dullard, it’s very exciting to be setting myself a serious mental challenge in understanding and trying to apply it.