Every model of reasoning is also a policy — it applies not just once, but to every situation that looks like this one. Charles Sanders Peirce noticed something underappreciated about how habits of thought work: they don't just describe past decisions, they pre-commit you to future ones. The secondary models you run — the heuristics, the mental checklists, the 'second-order' lenses — aren't just tools you pick up and put down. They reshape the questions you'll even think to ask next time. Peirce called this the 'self-correcting' nature of inquiry, but the knife cuts both ways: a flawed secondary model doesn't just give you a bad answer today, it trains you to ask the wrong question tomorrow. The practical move is to occasionally audit not what your secondary models conclude, but what they cause you to ignore — the questions they make invisible by virtue of having already 'answered' them.

Pick one secondary model you rely on regularly. What category of evidence does it structurally exclude — and when did you last encounter something that should have triggered it but didn't?

Drawing from American Pragmatism — Charles Sanders Peirce (How to Make Our Ideas Clear, 1878)