This paper summarizes relevant literature on visual attention, beginning with a discussion of how attention is shifted to different locations of focus - both consciously and reflexively. Two experiments investigate the mechanisms controlling attention shifts. Tones perceived to originate from a specific location inside the head were created by manipulating the amplitude of tones presented dichotically through earphones. Evidence of selective attention to these intracranial locations suggests that the premotor theory of attention must be revised. Attention is shown to give rise to a facilitation of processing power for items located within the area of focus. This effect decreases as the area of focus increases. Before any facilitation effects occur attention must select an area or object that will receive the boost in processing power. How this mechanism functions is discussed. The ability to distribute attention flexibly - to multiple non-contiguous locations, areas of various size and shape - appears, in part, due to the influence of distracting information that is present within a visual scene. Studies that have addressed the allocation of attention have, for the most part, utilized impoverished stimuli - letters, numbers or simple shapes displayed on a blank background - ignoring the possible influence a richly detailed scene might have on attentional processes. Two more experiments are run which investigate how attention is allocated to a detailed, photograph quality scene. Participants focused on an object located amongst distractors with all objects set against a photograph quality background. While focused on the object a probe occured briefly at either the target, distractor or background locations and participants responded to the direction of the probe arrow as quickly as possible. Response times were faster for object locations, slower for distractor locations and slowest for background locations. Two possible patterns of inhibition were not found - inhibiting information similar to the target the most or inhibiting information immediately surrounding the target the most. Near background locations were faster than far background locations and all distractor objects, regardless of similarity to and distance from, the target were inhibited equally. The pattern of response times supports a spread of facilitation from the target and inhibition from distractors to surrounding background locations. When a scene is created that better represents the natural world gradients of facilitation and inhibition are still found - the mechanism is not restricted to spreading into blank locations. When a more visually complex scene is viewed participants are able to allocate attentional resources flexibly - different locations within a scene receive different levels of inhibition - based on the demands of the task.

The perception and understanding of human behavior is fundamental for social interaction. Mostly grounded in the theory of embodied cognition, the study of intentional action has focused on different aspects such as action planning and execution, action identity, and action prediction. Based on principles of dynamical systems theory (DST), the goal of this study is to test the method of using eye tracking to expand our knowledge of the temporal dynamics of human intention perception by investigating the time-locked sequence of eye movements during social interaction to investigate the online decision making process during an action observation task. Stimuli consist of 5 sec videos portraying reach and grasp actions, which are either intentional (pour coffee) or unintentional (coffee spills), cooperative action (serve other) or non-cooperative (serve self). In condition 1 participants are asked to determine whether an action presented is intentional, condition 2 whether the action is cooperative while collecting eye tracking data. This study hypothesized that participants' eye movements will be sensitive to the task demands, predicting that (H1) participants in the intentional group will be more likely to attend to the object interactions as compared to the cooperative condition and (H2) participants in the cooperative group will be more likely to attend to social cues between the agents as compared to the intentional condition. The results show that the intentional group was more likely to focus on the object interactions in support of hypothesis 1. Furthermore, results also showed that the cooperative group focused more on social cues in support of hypothesis 2. Therefore, the results of this study strongly support the theory that cognitive processes such as decision-making during an intentional or cooperative action are emergent and the temporal dynamics can be made visible through eye tracking. In support of the dynamical systems theory, external influences such as task demand were shown to have an effect on the viewing pattern, duration of fixations, as well as attention to target details while observing a dynamic, natural and social interaction.