When two masked targets (T1 and T2), both requiring attention, are presented within half a second of each other, report of the second target is poor (Broadbent & Broadbent, 1987;Raymond, Shapiro, & Arnell, 1992). However, T2 report is good when T2 is presented more than half a second after T1 or when T1 does not require attention (Raymond et al., 1992). Raymond et al. (1992) named this pattern of results the attentional blink (AB). AB experiments require participants to make unspeeded responses after the presentation of all stimuli has ended. The fact that there is no need for on-line response selection has helped to convince many researchers that the AB results from relatively early processing limitations that occur before central (amodal) processing bottlenecks. In early AB experiments, only visual stimuli were used as targets and distractors, and these were often presented using the rapid serial visual presentation (RSVP) technique in which letters or words appeared rapidly one at a time in the same spatial location. Owing to the use of visual stimuli and the assumption that the AB resulted from relatively early processing limitations, the visual nature of the stimuli and their processing became incorporated into various theories of the AB (e.g., Duncan, Ward, & Shapiro, 1994;Raymond et al., 1992;Raymond, Shapiro, & Arnell, 1995;Shapiro, Raymond, & Arnell, 1994;Ward, Duncan, & Shapiro, 1996). For example, in the attentional dwell time theory of Duncan et al. (1994) and Ward et al., the AB is taken as a measure of the time course of visual attention. Shapiro et al. (1994) and Raymond et al. (1995) explained the AB as the result of confusion in an overcrowded visual short-term memory.Arnell and Jolicoeur (1999) tested visual theories of the AB by having participants attend to concurrent but independent visual RSVP letter streams and to compressed speech rapid auditory presentation (RAP) streams. The participants were asked to report the identity of a T1 number and the presence or absence of a subsequent X. T1 modality (visual or auditory) was fully crossed with T2 modality (visual or auditory), producing two withinmodality conditions (both targets visual, both targets auditory) and two cross-modality conditions (T1 visual and T2 auditory, T1 auditory and T2 visual). One quarter of the participants received each modality combination. Arnell and Jolicoeur found a reliable AB in each of the four modality combinations. The ABs were larger for the visual T2 conditions than for the auditory T2 conditions,but the AB sizes in the within-modality conditions were roughly equal to the AB sizes in the cross-modality conditions. Arnell and Jolicoeur concluded that the AB is not a uniquely visual phenomenon. They argued that the robust ABs observed in the cross-modality conditions provided good evidence that central (amodal) processing limitations were responsible for the AB. Specifically, limitations on stimulus consolidationin working memory were proposed.Using a variety of presentation conditions, other researchers have a...