Polanyi established that we know more than we can tell (tacit knowledge). This paperidentifies a deeper epistemic layer: we are ignorant of more than we know we don’tknow (tacit ignorance). Where tacit knowledge concerns the barrier of articulation—knowledge possessed but unstated—tacit ignorance concerns the barrier of percep-tion: boundary conditions that practitioners do not perceive at all, because limits areinvisible until exceeded.This distinction explains a persistent puzzle in organizational knowledge transfer.When concepts migrate across boundaries—between firms, divisions, industries,or generations—their positive content (methods, assertions, conclusions) trans-fers while their negative boundaries (preconditions, validity limits, failure modes)systematically drop out. We term this mechanism Boundary Condition Dropout(BCD). Building on Scandinavian translation theory (Sahlin-Andersson, Røvik), wesynthesize eight research traditions to specify the direction of asymmetric loss thatprior frameworks describe but do not fully theorize: constraining information isstructurally more vulnerable than enabling information.Tacit ignorance renders this problem unsolvable through documentation. One cannotdocument boundary conditions one does not perceive; recipients cannot apply docu-mentation referring to conditions they cannot assess. The barrier is not informationalbut phenomenological. This is compounded by organizational imperceptibility: orga-nizations govern through information asymmetry, structurally preventing membersfrom perceiving the organization as a system. Boundary conditions—being systemproperties—are therefore doubly invisible.We develop a pragmatist response drawing on Dewey: since comprehensive priorspecification is structurally impossible, organizations require learning systems thatdiscover boundaries through practice. Architecture Decision Records (ADRs) are an-alyzed as exemplary implementations, with particular attention to why explanation2fails where participatory practice succeeds—a finding with implications for organiza-tional knowledge management beyond software engineering.
Sophia Franny Philos (Mon,) studied this question.