Absence of priming coupled with substantially preserved recognition in lorazepam-induced amnesia

Two experiments are reported on priming when subjects are in an amnesic state induced by lorazepam. The primed tasks were completion of word-stems and generation of words from specified categories. In both experiments, lorazepam subjects showed no evidence of priming; whereas control subjects showed substantial priming. Recognition by the amnesic subjects of items produced in the priming tests, although impaired, was well above the chance level. These findings contrast with those obtained with organic amnesic subjects, for whom priming is typically normal but recognition is grossly impaired. The theoretical implications of this double dissociation between priming and recognition are discussed.