Modulation of implicit working memory in temporal grouping
MetadataShow full item record
A critical function of perception is the organization of temporally spaced input. This is accomplished through grouping, a process by which within-group elements are integrated with one another to form a cohesive unit. Grouping also requires boundaries to set off within-group elements from unrelated stimuli. In the temporal domain, grouping may be accomplished through use of an implicit working memory system that connects temporally spaced information. Temporal group boundaries may be created by reductions in the default integrative processes of this memory system. The present experiments probed integration strength by embedding priming tasks within temporal groups (i.e., events). Because priming also draws upon implicit working memory, priming strength should reflect the strength of integration. If modulation of temporal integration is responsible for grouping, this should be manifested as a reduction of priming across boundaries. Irrelevant feature priming tasks were used to assess integration strength. Participants responded to one of two independently varying object features. In this form of priming, change consistency of relevant and irrelevant features produces faster reaction times, resulting in a crossed interaction. This interaction served as a meter for the strength of temporal integration. The experiments included a variety of temporal grouping manipulations. Experiments involving rhythmic groups, spatial shifts, rotations, pitch, and timbre, as well as higher-level conceptual shifts, demonstrated reduced priming in across-boundary conditions. Both visual and auditory events were used, and experiments demonstrated that viewers’ interpretation of a scene contributed to the observed effects. Temporal integration does appear to be reduced at certain event boundaries, suggesting that this may be the general manner in which temporal grouping is accomplished. Motion change, a boundary from event segmentation research, did not reduce priming, indicating that the process presently under study differs from that studied using explicit segmentation procedures. The reduction of integration may correspond to a subjective, amodal experience of separation. The present technique may therefore offer an objective, implicit method to assess this sense of separation. Using this method, it is possible to reliably determine when people are experiencing temporal group boundaries even when they are not deliberately attending to them.