Abstract
Can we divide attention concurrently to two different spatial locations? To examine this question, we used the lag-1 sparing phenomenon in the attentional blink (AB) paradigm as an index of the maintenance of two concurrent attentional sets. In the AB, identification accuracy for the second of two successive targets (e.g., letters) inserted in a stream of distractors (e.g., digits) is impaired when the inter-target lag is less than about 500 ms. It is known that the AB deficit is eliminated, or is much reduced, when the second target is presented directly after the first (“Lag 1”). This phenomenon, known as lag-1 sparing, occurs only when the two targets belong to the same category. For this reason, lag-1 sparing is said to occur when the observer can maintain the same attentional set to process both targets. In Experiment 1, we presented two target letters in each of two synchronized RSVP streams of distractor digits, presented concurrently 1.7 deg left and right of fixation. They were separated by some distractors (lag 1, 2, 3, 5, or 7). Observers indicated whether the first two target-letters, which appeared simultaneously in the two streams, were the same or not, and then identified the second targets, which were also displayed simultaneously. Lag-1 sparing occurred in both streams, indicating that a given attentional set was maintained concurrently at two locations. But this result could be mediated by a single large attentional spotlight encompassing both streams. Experiment 2 tested this option. Each second-target could appear in one of three locations: Within its stream (SAME condition), displaced towards fixation (INWARD condition), or away from fixation (OUTWARD condition). Lag-1 sparing occurred in the SAME but not in the other conditions. These results suggest that attention can be deployed to two spatially non-contiguous locations at the same time.