Dissociated Forgetting Patterns in Item Versus Source Memory

Although item memory (e.g., remembering seeing your sunglasses) and source memory (e.g., remembering seeing them on the top shelf) have been dissociated both behaviorally and neurologically, little is known about dissociations in their forgetting patterns. A neuropsychological theory of forgetting (Hardt et al., 2013) predicts dissociated forgetting patterns depending on whether a memory is represented in the hippocampus or not. Extra-hippocampally represented memories, such as item memory, should be particularly susceptible to short-term, primarily interference-based forgetting. Hippocampus-dependent source memory should be protected from short-term forgetting but prone to long-term, primarily decay-based forgetting. We tested these predictions in three experiments (Ns = 164 to 210) in which participants studied object pictures in different screen locations. Item (object) and source (location) memory were tested at different retention intervals manipulated between subjects: immediately, after a short period (~10 min) of (similar or dissimilar) mental activity or intended low activity (audio-guided breathing exercise; wakeful rest), or on the next day. Across all experiments, there was no forgetting in source memory over any of the short retention intervals, whereas item memory significantly declined across most of them. There was substantial forgetting over 24 hours in both item and source memory, with more long-term forgetting in source memory after accounting for the short-term forgetting in item memory. Thus, forgetting patterns are dissociated between item versus source memory, in line with the predictions by Hardt et al. Some questions remain regarding the role of similarity-based interference in the observed short-term forgetting in visual item recognition.