Recall of general-knowledge targets remains remarkably high 9 days after a
5-s exposure, although these targets could not be recalled (marginal knowle
dge) or recognized (submarginal knowledge) just prior to exposure. Large am
ounts of educational content (e.g., foreign language vocabulary, mathematic
al formulas) become marginal or submarginal after formal study ends. The au
thors discuss the potential benefits of brief maintenance interventions des
igned to stabilize long-term access to such knowledge.