The Attentional Bottleneck: Constraints, Neural Correlates and Reward-Dependent Plasticity

Through this series of experiments, we unveiled major determinants in multiple target selection and defined which critical features allow a target to win competition for attentional resources. We then revealed the electrophysiological correlates (ERPs) of the attentional bottleneck. Moreover, we induced durable reward-based changes in the attentional priority of specific spatial locations, resulting in an imbalanced probability to identify a target in one or the other of two critical spatial locations in a pair, and revealed an interesting influence of individual traits affecting performance during the training session. In the first experiment, a series of critical elements were found to influence competition for attentional resources and allow a target to enter the attentional bottleneck. Firstly, an influence of target position within the stimulus array was revealed, with the best performance for the most lateralized positions, resulting in a visual anisotropy. We then observed an effect of targets’ relative position, namely a significant advantage when the targets were displayed across hemifields. Eventually, we demonstrated that the relative competitive strength of two target locations in a pair can be predicted on the basis of performance measured in the single target condition for the given locations. In the second experiment, we measured the Event-Related Potentials (ERPs) in order to directly unveil neural correlates of the attentional bottleneck. Interestingly, a number of critical factors appeared to affect the occurrence and amplitude of the N2pc component. Specifically, the N2pc was not modulated by the number of presented targets, but only by the number of consciously identified targets, as well as by their relative spatial position. Furthermore, we demonstrated that the better performance at the behavioral level (corresponding to correctly reporting both targets vs. one in the double target condition) could be explained by an earlier attentional engagement in the given trial. Finally, we revealed a suggestive modulation of the amplitude of the N2pc waveform depending on the number of identified targets. We then investigated the influence of a reward-based learning protocol on the attentional priority of locations in space. Specifically, through the use of a reward-based training protocol, we were able to produce enduring changes in priority maps that are responsible for directing spatial attention and for arbitrating target selection. In particular, spatial locations associated with an overall more positive outcome during training were prioritized with respect to spatial locations associated with an overall less positive outcome. Importantly, the observed changes in performance were measured after a four-day delay from the training phase, likely reflecting durable plastic changes in spatial priority maps. Finally, we investigated whether the entity of the acquired reward-based attentional bias could depend on specific individual attributes and personality traits. An interesting influence of individual traits appeared to affect performance during training. Specifically, differences in performance (in terms of accuracy and of reaction times) between reporting a target displayed at an 80Hh position vs. a target displayed at a 20Lh position appeared to be predicted by a series of factors concerning participant profile. In particular, we identified two main individual traits having an influence on performance, namely gender and individual scores at the Drive subscale of the Behavioral Activation System scale. Altogether, this research contributes interesting results with respect to the behavioral and neural basis of the attentional bottleneck that characterizes spatial directed attention. Moreover, it adds critical new evidence in the field of reward-mediated attentional learning, demonstrating that spatial priority maps can encounter durable plastic changes.