Nevertheless, our results argue against this model for the following reasons. First, a “switching model” in which a single spotlight travels in space predicts that it should be faster to switch attention between patterns that are close together than between patterns that are farther apart. We found the opposite (Figure 2 and Figure 3S). Second, our control experiment demonstrates an increase in RTs associated with changes

selleck in one translating RDP when its associated change probability is reduced and the change probability in the other pattern is increased. We argued that a switching spotlight should produce a RT distribution that approximates the pooled RTs distributions corresponding to both probabilities. However, we found that the pooled distribution has a higher mean than the one corresponding to 0.5-change probability targets of the main experiment.

In fact, the RTs distribution corresponding to targets with the largest change probability (0.8) was similar to the one corresponding to 0.5-change probability targets. This suggest that during the main experiment the animals devoted the same amount of attention to each target as to this website the 0.8-target of the control experiment, and that the level of attention to any of the RDPs never decreased to values similar or close to the one corresponding to the lowest (0.2) change probability target. For a switch model to account for these data animals had to switch attention between the 50-targets in ∼12 ms or less (determined by shortening the RTs corresponding to the 20-targets and repeating the pooling and comparison of RT distribution until it became nonsignificant). This is half of the estimated shift time

from our data and much shorter than the lowest value reported for stimulus driven Histamine H2 receptor (35 ms) and voluntary (∼200 ms) attention shifts in humans ( Horowitz et al., 2009). Third, and most importantly, we found that responses during tracking were decreased relative to those during attend-RF and attend-fixation when the translating stimuli circumvented the RF pattern. A switching spotlight of attention cannot account for these results. Instead, our findings suggest a relative suppression of responses to the RF pattern when it falls between the two attended RDPs. This strongly argues against models in which a single spotlight of attention travels in space, or rapidly turns on and off at the location of tracked objects ( Pylyshyn and Annan, 2006). This model proposes that when attending to multiple stimuli the spotlight of attention can split into multiple foci corresponding to each relevant stimulus and excluding distracters in between (Castiello and Umiltà, 1992, Cavanagh and Alvarez, 2005, Howe et al., 2010 and Niebergall et al., 2010). The animals’ behavioral performance in the main tracking task show that they attended to both translating RDPs.