| The structure of human memory is heatedly debated. While some researchers claim that there are two or more memory systems (Graf & Schacter, 1985; Mitchell & Brown, 1988; Tulving, 1985), others believe that memory is unitary (Bentin et al., 1992; Roediger, 1990). Behavioral and neuroimaging research presented in this dissertation examines a unitary model of human memory using implicit and explicit memory tests for object locations. Specifically, the behavioral experiments tested (a) if implicit memory could occur without explicit memory, (b) if implicit memory correlated with explicit memory, and (c) if both tests were equally affected by the strength of memory traces. It appeared that implicit memory for object locations was greater for explicitly recalled, compared to not recalled, items. Implicit memory was correlated with explicit recall. Stronger memory traces were associated with higher recall and higher implicit memory for object locations. The fMRI experiment tested (a) if increases in neural activity were related to enhancement of implicit memory, and (b) re-engagement of encoding-related brain regions during implicit memory test for locations. Consistent with the findings reported in the earlier studies of explicit memory, improvement of implicit memory for locations was associated with increased neural activity in task-specific and attentional networks. In addition, implicit spatial memory depended on the magnitude of reactivation in the encoding-related brain regions. These findings indicated that implicit and explicit memory may rely on similar facilitatory mechanisms. Taken together, behavioral and neuroimaging evidence supported a unitary model of human memory and suggested that both, implicit and explicit, spatial memory tests access common memory representations. It was proposed that reactivation of memory traces is a key factor for performance on implicit and explicit memory tests. Implicit and explicit tests may differ in their ability to reactivate memory representations. This difference may account for dissociation between performance on implicit and explicit memory tests observed in the previous studies. |
