Abstract
Studies on visual object memory have historically used unisensory stimuli during encoding and retrieval. However, research has shown that later visual object recognition is improved when the initial encounter is multisensory, such as when a to-be-remembered object is presented as an image along with an identity-congruent sound during encoding. However, these studies have used non-realistic objects, such as line drawings, and explicit instructions to memorize items. The present study uses realistic objects, and introduces a novel task to investigate how auditory information enhances visual encoding by probing incidental memory for visual objects. Participants were presented with visual images of 3D objects (e.g., dog) alongside an object-congruent (e.g., “bark”), object-incongruent (e.g., “ring”), or white noise control sound, and were asked to categorize each as animate or inanimate. In a subsequent, surprise visual recognition task, old images were mixed with new images, and participants were asked to judge whether each object was new, or whether it was old and had initially been presented with a congruent, incongruent, or control sound. Incidental visual memory for previously seen objects was better for objects initially presented with object-congruent auditory information than with incongruent sounds or with a control sound, though there were no significant differences between the incongruent or control sound conditions. These results extend previous findings of the benefits of multisensory experiences on visual object recognition to longer term incidental visual memories with realistic objects. The results suggest that visual representations are more robust at encoding when accompanied by identity-congruent sounds and that these benefits extend into memory, regardless of intention to encode visual objects into memory. Future investigations will evaluate this effect in increasingly naturalistic situations, with variable delay periods, and with the presence of visual distractor objects during encoding.