Impressive. I've felt this way for a long time, and the author articulates it well. My attempt at a solution to this problem is to study logic and rationality.

Impressive. I've felt this way for a long time, and the author articulates it well. My attempt at a solution to this problem is to study logic and rationality. The LessWrong wiki and blog is a great place to start doing that. I think that studying reasoning in general can help to spot where and how you're using propositions in your inference. I regularly do sanity-checks where I go through my propositions to check how close they are to my core domain knowledge. It then becomes a balancing act: how suspicious do you want to be of relatedness to your own domain? The more suspicious you are, the more work there is to check. This process has saved me from making serious errors on several occasions. I think that the "suspiciousness tolerance" needs to be set according to context, but that "good-ish" (ie, acceptable risk of error, acceptable amount of work to check domain status of propositions) values are probably possible in a good number of situations.