Mencius argued that moral perception — the capacity to recognize what a situation actually calls for — is more like a sense organ than a skill. You don't train it by memorizing rules; you train it by repeatedly noticing when your first reading of a situation was wrong. What makes this unsettling for anyone serious about mental models is that the same repetition which sharpens a model also slowly anesthetizes the sense that was supposed to catch the model's errors. The Mencian term for this perceptual faculty is *liangzhi* — roughly, 'innate moral knowing' — and Mencius spent considerable energy arguing that it degrades not from lack of use but from habituation to noise. For leaders and thinkers who accumulate models deliberately, the implication is precise: your models are becoming the noise. The corrective isn't adding more frameworks — it's periodic exposure to situations your current models find genuinely uninterpretable, where you have to sit in confusion long enough for the underlying faculty to reassert itself. Call it structured disorientation. Build in encounters — with unfamiliar industries, unfamiliar people, unfamiliar problems — not to gather new data, but to temporarily destabilize the perceptual apparatus that's been capturing all your data through the same filter.

What's the last situation that genuinely confused you — and did you sit with the confusion, or immediately reach for a familiar frame to resolve it?

Drawing from Confucianism (Mencius) — Mencius (Mengzi)