Repetition priming for multisensory stimuli: task-irrelevant and task-relevant stimuli are associated if semantically related but with no advantage over uni-sensory stimuli.

Signals presented simultaneously in two sensory modalities are detected faster and more accurately than their uni-modal presentations. We investigated the effect of repeated experience in successive test blocks (Repetition Priming, RP) for simultaneously presented multi-sensory stimuli, as compared to uni-sensory, visual, stimuli. Participants had to decide whether the order of letters in two letter-strings (the visual stimulus) was reversed or not. The visual stimuli were presented alone or accompanied by a task-irrelevant auditory or a haptic signal. The letter-strings denoted words that were either semantically related or unrelated to the auditory or haptic signals. RT measurements showed significant RP across all conditions, with accuracy at ceiling. The RP gains were not significantly different for the uni- and the bi-sensory stimulus combinations in the initial three blocks. However, in the 4th block, where instead of the paired bi-sensory stimuli the previously paired visual stimulus was presented alone, the RP gains were significantly smaller in the semantically-related stimuli (disassociation cost). Congruent bi-sensory stimuli had been shown to improve perceptual learning compared to uni-sensory stimuli when both signals were task-relevant. Our results suggest that when an additional signal, in a different sensory modality, is irrelevant for the task's performance, there may be no advantage--in terms of greater RP gains--for multisensory stimuli. Nevertheless, semantically related stimuli experienced simultaneously in different sensory modalities may be represented in an associative manner in implicit memory even when only one stimulus is task-relevant.