Effects of hearing words, imaging hearing words, and reading on auditory implicit and explicit memory tests

In four experiments, we examined the degree to which imaging written words as spoken by a familiar talker differs from direct perception (hearing word s spoken by that talker) and reading words (without imagery) on implicit an d explicit tests. Subjects first performed a surface encoding task on spoke n, imagined as spoken, or visually presented wards, and then were given eit her an implicit test (perceptual identification or stem completion) or an e xplicit test (recognition or cued recall) involving auditorily presented wo rds. Auditory presentation at study produced larger priming effects than di d imaging or reading. Imaging and reading yielded priming effects of simila r magnitude, whereas imaging produced lower performance than reading on the explicit test of cued recall. Voice changes between study and test weakene d priming on the implicit tests, but did not affect performance on the expl icit tests. Imagined voice changes affected priming only in the implicit ta sk of stem completion. These findings show that the sensitivity of a memory test to perceptual information, either directly perceived or imagined, is an important dimension for dissociating incidental (implicit) and intention al (explicit) retrieval processes.