Among the numerous formal and informal theories of the at-tentional blink, the common theoretical thread is that the deficitstems from selective attention and working memory processesbeing tied up in processing the first target (T1) when the sec-ond target (T2) appears. Rusconi & Huber (2017) challengedthis view by proposing the ’perceptual wink’ model of the AB,which posits that for categorical AB tasks (e.g., number/letter)the deficit reflects a failure to perceive that T2 belonged to thetarget category. The model makes the assumption that percep-tion is ’multi-faceted’; that is, there are separate, independentperceptual representations for an item’s identity and its cate-gory, and that either representation can be used to drive per-formance (e.g., trigger attentional encoding) depending on thetask demands. To differentiate between attention versus per-ceptual accounts of the AB, we used a stripped down RSVPtask where participants were asked to either report the iden-tity or category of the third item in a sequence of characters.In support of the perceptual account, we found priming foridentity or category depending on the task. Furthermore, wefound that the category results were analogous to the AB andthe spread of sparing even though the first character was nota target and there was no need to selectively filter items intoworking memory.