A key model feature is that responses to all objects falling within the spotlight are enhanced; thus, in our experiment it predicts that during tracking responses to the irrelevant RF stimulus would be enhanced when the translating patterns circumvented but did not enter the RF. Contrary to that, we observed that when comparing the responses during tracking versus attend-fixation INCB018424 there was either no change in attentional modulation (Pr direction of translating RDPs) or a response

decrease (AP direction) in the former relative to the latter condition. Moreover, this model also predicts that when increasing the size of the attentional spotlight, the benefits of attention should GABA inhibition decrease. We found, however, that performance in the far configuration was higher than that in the near configuration (Figures 2G and 2H) and the differences in attentional modulation between tracking and attend-RF were similar in both cases or even slightly

larger in the far configuration ( Figure 4 and Figure 5). The performance differences between the far and near configurations during tracking remained when removing the RF stimulus ( Figure 3S), ruling out that stronger distracter interference in the near condition was responsible for the effect. Furthermore, during a session we interleaved trials of the three different conditions to avoid that animals could predict in advance the difficulty of the upcoming trial. Animals show a higher performance in the

easier tasks (i.e., attend-fixation and attend-RF showed higher performance than tracking), suggesting that they could not adjust their attentional effort on a trial-by-trial basis. These findings strongly argue against the zooming spotlight hypothesis. We consider at least two Linifanib (ABT-869) possible explanations for discrepancies between our results and those of studies providing neural correlates of the zooming model (Barriopedro and Botella, 1998, Heinze et al., 1994, McCormick and Jolicoeur, 1994 and Müller et al., 2003b). First, it is possible that the coarse spatial resolution of ERPs used in those studies, does not allow measuring decreases in the activity evoked by distracters. Second, it is possible that with certain stimulus configurations and task demands the spotlight of attention zooms in/out. In fact, a recent study has provided evidence that humans can adjust the size of the attentional focus depending on task instructions (Herrmann et al., 2010). This model has been very difficult to test in studies of attention (Castiello and Umiltà, 1992 and McCormick et al., 1998; Oksama and Hyönä, 2008; Jans et al., 2010 and Cave et al., 2010). It proposes that subjects attend to multiple objects by rapidly switching a single spotlight of attention from one object to another.