Reevaluating Encoding-Capacity Limitations as a Cause of the Attentional Blink Roberto Dell’Acqua University of Padova Pierre Jolicœur University of Montreal Roy Luria and Patrik Pluchino University of Padova A number of researchers have emphasized the role of distractors intervening between successive targets as the primary determinant of the attentional blink (AB) phenomenon. They argued that the AB is abolished when 3 or more targets are displayed as temporally contiguous items in rapidly presented serial sequences. In 3 experiments, the authors embedded 1-, 2-, or 3-digit targets among letter distractors in rapidly presented visual sequences. Across the experiments, both the number of targets and the lag between them were manipulated, producing different proportion of trials in which 3 temporally contiguous targets were presented in the test session. Evidence of an AB affecting the targets that followed the first target in these sequences was found in each experiment when the probability of a given target report was conditionalized on a correct response to the preceding targets, thus reinforcing the notion that some form of capacity limitation in the encoding of targets plays a central role in the elicitation and modulation of the AB effect. Keywords: Attentional blink, lag-1 sparing, multitarget RSVP streams The term attentional blink (AB; Raymond, Shapiro, & Arnell, 1992) refers to a psychological phenomenon that can be observed when two visual targets, usually embedded among distractors, are presented in close temporal succession, in a technique known as rapid serial visual presentation (RSVP). When the temporal inter- val between the two targets is shorter than about 500 ms, the identification of the first target (T1) is good, whereas the identi- fication of the second target (T2) is often poor. A related phenom- enon, termed lag-1 sparing (Potter, Chun, Banks, & Muckenhoupt, 1998), refers to the fact that T2 is often preserved from the AB when it is presented within about 100 ms of T1. Several models proposed in the past share a set of common assumptions about the potential causes of the AB and lag-1 sparing effects (see Dell’Acqua, Jolicœur, Pascali, & Pluchino, 2007, for an overview). One of these assumptions concerns the critical role played by T1 in the generation of these phenomena. Succinctly, the AB is hypothesized to result from a capacity limitation in the encoding of targets into visual short-term memory (VSTM; e.g., Chun & Potter, 1995; Jolicœur, 1998; Jolicœur & Dell’Acqua, 1998; Visser, 2007; Vogel, Luck, & Shapiro, 1998). Processing capacity allocated to T1 cannot be used for T2 during some period of time, leading to a loss of accuracy in reports of T2, if T2 is followed by a trailing mask (Dell’Acqua, Pascali, Jolicœur, & Sessa, 2003; Giesbrecht & Di Lollo, 1998; Jolicœur, 1999; Ptito, Arnell, Jolicœur, & MacLeod, in press). The selection of T1 from distractors is made possible through the activation of an attention filter that reacts promptly to T1 onset, allowing T1 to be selected efficiently for further processing, but reacts sluggishly to T1 offset, allowing the inclusion of T2 in the attention episode triggered by T1 when the two targets are temporally contiguous (e.g., Shapiro, Raymond, & Arnell, 1994; Shih, 2008). Selection Control and Distractor-Induced “Bouncing” Reactions A different perspective on the cause of AB and lag-1 sparing phenomena has recently been proposed by Di Lollo, Kawahara, Shahab Ghorashi, and Enns (2005) and by Olivers, van der Stigchel, and Hulleman (2007). In these studies, target identification accuracy for sets of contiguous targets was compared with target identification accuracy for sets in which the sequence of the targets was discontin- ued by a distractor. The case of interest emerged when as many as three targets were presented. If the targets were contiguous, T3 ben- efited from a protracted form of sparing effect: T3 report accuracy was as good as T1 report accuracy. The sparing effect was absent, however, if T2 was replaced with a distractor. In that case, T3 report accuracy was substantially worse than T1 report accuracy, suffering from an AB effect. The results of Di Lollo et al. (2005; see also Kawahara, Enns, & Di Lollo, 2006; Kawahara, Kumada, & Di Lollo, 2006) and Olivers et al. (2007) argue against the centrality of capacity limitations in the processing of T1 as the triggering factor of both the AB effect and lag-1 sparing. In contrast, as these authors have argued, the crucial element causing the AB effect appears to be the distractor trailing T1. Roberto Dell’Acqua, Department of Developmental Psychology and Center for Cognitive Science, University of Padova, Padova, Italy; Pierre Jolicœur, Department of Psychology, University of Montreal, Montreal, Quebec, Canada; Roy Luria and Patrik Pluchino, Center for Cognitive Science, University of Padova. Correspondence concerning this article should be addressed to Roberto Dell’Acqua, Center for Cognitive Science, Via Venezia 8, Padova 35131, Italy. E-mail: dar@unipd.it Journal of Experimental Psychology: © 2009 American Psychological Association Human Perception and Performance 2009, Vol. 35, No. 2, 338 –351 0096-1523/09/$12.00 DOI: 10.1037/a0013555 338