On the fate of visual stimuli masked by object substitution

Human visual awareness is inherently limited. When we look at a visual scene, we are conscious of only a small fraction of the available information at a particular point in time. Given this, experimental psychologists have long been interested in the factors that determine what and when stimuli are registered consciously, and the extent to which stimuli are processed in the absence of awareness. A number of methods have been devised to study these limitations and associated mechanisms in the laboratory. One such method is object-substitution masking (OSM). In OSM, a common-onsetting, temporally-trailing four-dot mask can interfere with the perception of a briefly presented target, even though the target and mask do not spatially overlap and they have different contours. Study 1 documents a previously unreported phenomenon: The recovery from OSM with prolonged mask exposure (Goodhew, Visser, Lipp, & Dux, in press). While all previous reports suggested that the magnitude of OSM increased monotonically with trailing mask duration (Di Lollo, Enns, & Rensink, 2000), we found that target identification can improve with prolonged (e.g., 640 ms) relative to intermediate (e.g., 240 ms) mask durations. That is, there is a recovery from OSM with prolonged mask exposure, producing a U-shaped masking function across time. This demonstrates that stimuli which initially lose the ‘competition for consciousness’ are not irrevocably lost, but in fact, can be recovered and consciously perceived. Study 2 further investigated the mechanisms underlying this recovery. We eliminated explanations based on temporal separation of target and mask offsets (e.g., Macknik & Livingstone, 1998) or object-individuation cues triggered by prolonged mask exposure (e.g., Lleras & Moore, 2003). Instead, we isolated the recovery to an inherently target-processing time-based phenomenon, which reveals the prolonged temporal dynamics of conscious object perception (Goodhew, Dux, Lipp, & Visser, submitted). Study 3 examined the level of processing of the target in OSM. Here, we found, contrary to several previous reports (Chen & Treisman, 2009; Reiss & Hoffman, 2006), that there is implicit semantic processing of a masked target; when observers are unable to report the identity of the target, or even detect it presence. This tells us that the brain is capable of extracting semantic meaning information from stimuli that are not consciously perceived, and that this information can implicitly guide behavior (Goodhew, Visser, Lipp, & Dux, 2011). Together, these results highlight the prolonged iterative dynamics of recovering a masked target in OSM, and show that this process is invariant to key physical properties of the mask. Furthermore, they demonstrate that there is extensive high-level processing of successfully masked stimuli, prior to them reaching conscious awareness