Less Wrong is a community blog devoted to refining the art of human rationality. Please visit our About page for more information.

poke comments on Reductive Reference - Less Wrong

20 Post author: Eliezer_Yudkowsky 03 April 2008 01:37AM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (39)

Sort By: Old

You are viewing a single comment's thread.

Comment author: poke 03 April 2008 04:35:33PM 1 point [-]

I'm curious, Eliezer, do you introduce Bayes into the argument as a means of warrant/justification or do you see it playing a causal (descriptive) role? Historically speaking, for example, the causal reason a telescope was used in your example might be, say, Kuhn's notion of exemplary research ("Galileo did it so it's good enough for me") or some other (perhaps more likely) psychological/cultural explanation and the scientist might be completely disinterested in justifying his belief based on priors (and he might not do such a thing unconsciously either).

On the topic: I think one of the reasons this sort of thing is unnerving to contemplate is the specter of skepticism that tends to still haunt people. Most people believe that if their beliefs do not have some exacting connection with reality (some metaphysical relation) then they can never "reach the world" so to speak. I think this is incorrect. The view that we can never reach the world is predicated on a strong sort of subjectivism; that "inside here" is the privileged starting point of inquiry when, in reality, anything can be our starting point (no point is privileged).

If you reject the infallibility of introspection, it loses its privileged place as the starting point of inquiry, and there's no reason to continue thinking you're bound to it; taking the quantum mechanical view of the world as the starting point for inquiry is no different than taking your subject-bound observations as the starting point for inquiry (and, for that matter, moving from one to the other no longer presents a problem; 3rd person accounts do not need to be justified in terms of 1st person accounts).

It's the looseness of reference that allows us to penetrate beyond our psychology; it's the fact that our words are related to the world by convention that has allowed us to create this elaborate system (science) that establishes a series of conventional relationships between terms that eventually, through a path we can trace and re-trace, takes us to the world.