Not spotting that two things look alike. Spotting that they’re built the same way underneath.
Cognitive science’s structure-mapping theory distinguishes analogical pattern recognition from mere similarity: two things can share surface features (both are round, both are red) without sharing structure, and two things can share deep structure (a solar system and an atom) while looking nothing alike on the surface. Genuine pattern recognition maps relations, not attributes — it notices that the way gravity holds a planet in orbit is structurally the same move as the way a nucleus holds an electron, regardless of the surface differences.
This is what “studying history and culture” trains that studying a single technical domain doesn’t: the habit of noticing when a structure from one field — a power dynamic, a failure mode, a growth curve — recurs in a completely different one. It’s the cognitive engine behind the essay’s own central claim, that connecting disparate fields produces insight a single-domain expert can’t reach alone.
Source: Gentner, D. “Structure-mapping: A theoretical framework for analogy,” Cognitive Science, 1983.