---
type: "synthesis"
articles: ["a032", "a046", "a051", "a086"]
tags: ["tacit-knowledge", "polanyi", "tension"]
id: "cross-two-theories-tacit-knowledge"
sources: ["reskilling"]
sourceVaultSlug: "hbr-seg-reskilling"
originDay: 10
articleStem: "hbr-seg-reskilling"
sourceUrl: "(unified vault: 13 sources)"
sourceTitle: "HBR — People Ⅲ-B · Reskilling / L&D / talent / restructuring"
---
The corpus contains **two competing theories of tacit knowledge** — both citing Polanyi — that a careful reader should hold in tension.

**Theory 1 (externalize it):** A032's [[concept-reverse-mastery]] argues that in the AI era, unspoken intuition is a *liability*; the valuable professional must convert tacit judgment into *explicit* criteria the machine can consume ([[concept-tacit-knowledge-d32]], [[quote-reverse-mastery]]). The [[concept-reasoning-trail]] is the artifact that forces this externalization. A086's [[concept-attribution-engine]] pushes it further: mine high performers' behaviors and codify them for everyone.

**Theory 2 (transmit it, human-to-human):** A051's [[concept-tacit-knowledge-d51]] and [[prereq-tacit-vs-explicit-knowledge-d10]] insist the deepest knowledge "lives in people and transfers through proximity and time, not documentation" — hence [[framework-distributed-apprenticeship]], not a document repository. A046's [[concept-unconscious-competence]] agrees: it is reachable only by learning "from the ground up."

**Are they contradictory?** Partly. A032 optimistically assumes judgment can be articulated; A051 warns much of it fundamentally can't. The reconciliation: *some* tacit knowledge is articulable (and should be, for AI collaboration), but the irreducible remainder still requires apprenticeship — which is exactly why cutting the base is dangerous ([[cross-broken-apprenticeship-pipeline]]).