Picking-up and exploiting spatial and temporal regularities in the occurrence of sensory events is important for goal-directed behaviour. According to the “Predictive Coding Hypothesis” (Friston Philosophical Trans R Soc B 360(1456):815–836, 2005), these regularities are used to generate top-down predictions that are constantly compared with actual sensory events. In a previous study with the Attentional Blink (AB) paradigm, we showed that the temporal and probabilistic uncertainty of T2s that are presented outside the Attentional Blink period, i.e. at least 400 ms after T1, improves the conscious report of T2 that are presented inside the AB. The study of ERP correlated showed that this improvement was associated with a prolonged storage of pre-conscious T2 traces in extra-striate areas (Lasaponara et al. Cortex 71:15–33, 2015). Here, we tested whether variations in the probabilistic cueing of the position of a primary T1 visual target in a 4 × 4 letter array, modulate the retention of memory traces evoked by secondary letter targets (T2) that were presented in other positions of the array. Most important, in each trial, the identity of T2 was specified to participants upon disappearance of the array. We show that high probabilistic cueing facilitates T1 detection and improves the corresponding sensitivity index (d′). In contrast, retention and conscious report of secondary targets (T2) improves when the probabilistic cueing of T1 position is poor. These results suggest that uncertainty in the upcoming position of primary targets boosts the strength of memory traces evoked by secondary targets and improves the possibility that traces of secondary targets gain full access to conscious processing.
ASJC Scopus subject areas