Preferences form the basis for how we see the world. At the most fundamental level, organisms can be seen as preference computers. They interpret sensory stimuli as attractive or aversive, gravitating toward the former. Opinions are more or less complex preferences. They add nuance to our movements toward and away from things and situations. Without them, we might as well be the simplest automata like a neutrino careening into the abyss.
From a utilitarian perspective, some basic (e.g., “biological”) preferences drive a complex of preferences and opinions, which in turn engages logical reasoning to maximize the return on caloric investment. This, I believe, is the relationship of reason to opinion. Reason itself is only a means to an end and cannot be an end in itself. Kurt Gödel’s incompleteness theorem cast doubt onto the ability of formal systems like arithmetic and Aristotelean logic to perfectly undergird an epistemology (and, I would venture to say, an ethic).
However, I do see value in cleaning the lens, again from a utilitarian standpoint. To use a computer metaphor, a computer program or algorithm is opinionated in that it drives toward a certain goal, which the programmer decides. However, the steps in its procedures should be strictly deterministic and free from arbitrary preferences. Hence, we should work to enhance the reliability of the mechanical aspects of information processing and to remove the aberrations of the lens through which we see to work toward our goals.
