Abstract
Using the DRM paradigm, our experiments examined the activation and monitoring of memories in semantic and phonological networks. Participants viewed lists of words and/or pseudohomophones (e.g.,
dreem). In Experiment 1, participants verbally recalled lists of semantic associates or attempted to write them as they appeared during study. False recall was reduced in the written modality only for pseudohomophone lists. In Experiment 2, hybrid (i.e., semantic and phonological associate) lists were read, and participants attempted to spell the items as they appeared during study. False recall was negatively related to the number of pseudohomophone phonological associates contained in the list. We posit that this result is due to a reduction in orthographic overlap between pseudohomophone associates and the critical lure. In both Experiment 1 and 2, the critical lure was usually written as a pseudohomophone when participants viewed a pure pseudohomophone list. Experiment 3 demonstrated that false recall produced by lists of phonological associates is related to the degree of orthographic overlap between the list items and the critical lure. We conclude that list context can provide discriminating information that can be used to reduce false recall via source monitoring. Also, when reading lists, the critical lure can be activated in a semantic network via phonology, and its activation in a phonological network is enhanced via orthography.