Confucius was obsessed with the rectification of names — zhengming — the idea that when words no longer match reality, institutions start to rot. He wasn't being pedantic. He was pointing at something political: sloppy language doesn't just describe disorder, it produces it. We're living through a version of this right now with AI-generated communication. When a system can produce confident, well-structured prose about anything, the word 'confident' stops meaning what it used to. The prose signal — the quality that once told you 'this person knows what they're talking about' — gets decoupled from the underlying knowledge. Xunzi, Confucius's sharpest successor, pushed further: he argued that language conventions only work when communities actively maintain them, which means someone has to notice when a word has drifted. Your practical move today: pick one term you use constantly in professional communication — 'alignment', 'insight', 'quality', 'ready' — and ask whether you and your colleagues actually mean the same thing by it. In an age of AI-assisted language, terminological drift accelerates invisibly. Zhengming starts with you noticing the drift before the institution does.

What is one word your team uses constantly where, if you asked three colleagues to define it precisely, you'd get three different answers — and what decision has that ambiguity already quietly distorted?

Drawing from Confucianism — Xunzi