The attentional blink is not a unitary phenomenon.

Identification of the second of two targets is impaired if it is presented less than about 500 ms after the first. Three models of this second-target deficit, known as attentional blink (AB), were compared: resource-depletion, bottleneck, and temporary loss of control (TLC). Five experiments, in which three sequential targets were inserted in a stream of distractors, showed that identification accuracy for the leading target depended on an attentional switch whose magnitude varied with distractor-target similarity. In contrast, accuracy for the trailing target depended on similarity between the target and the trailing mask. These results strongly suggest that the AB is not a unitary phenomenon. Resource-depletion was ruled out as a viable account. The effect of attentional switching was handled naturally by the TLC model, while bottleneck models offered the best account of the effect of backward masking.