The question “How do I know?” has rescued many intelligent people from themselves.

Four words interrupt a great deal of confident error. Asked seriously, they force you back to the actual source: who told you, what you observed, what you inferred and then forgot was an inference. Intelligent people are especially vulnerable without them, because they are better at constructing persuasive support for whatever they already think.

Build the question into anything consequential. Before you act on a belief, say how you know and see whether the answer holds up. Often the trail runs out much sooner than expected, and finding that out beforehand is considerably cheaper than finding it out afterward.