Contextual consistency facilitates long-term memory of perceptual detail in barely seen images

It is long known that contextual information affects memory for an object's identity (e.g., its basic level category), yet it is unclear whether schematic knowledge additionally enhances memory for the precise visual appearance of an item. Here we investigated memory for visual detail of merely glimpsed objects. Participants viewed pairs of contextually related and unrelated stimuli, presented for an extremely brief duration (24 ms, masked). They then performed a forced-choice memory-recognition test for the precise perceptual appearance of 1 of 2 objects within each pair (i.e., the "memory-target" item). In 3 experiments, we show that memory-target stimuli originally appearing within contextually related pairs are remembered better than targets appearing within unrelated pairs. These effects are obtained whether the target is presented at test with its counterpart pair object (i.e., when reiterating the original context at encoding) or whether the target is presented alone, implying that the contextual consistency effects are mediated predominantly by processes occurring during stimulus encoding, rather than during stimulus retrieval. Furthermore, visual detail encoding is improved whether object relations involve implied action or not, suggesting that, contrary to some prior suggestions, action is not a necessary component for object-to-object associative "grouping" processes. Our findings suggest that during a brief glimpse, but not under long viewing conditions, contextual associations may play a critical role in reducing stimulus competition for attention selection and in facilitating rapid encoding of sensory details. Theoretical implications with respect to classic frame theories are discussed.