Abstract
It has been suggested that processing of global and local properties are associated with neuronal oscillations at different rhythms. Recently, by employing a time-resolved behavioral measurement, several visual attentional studies have revealed neurophysiologically relevant rhythms directly in behavioral performances, suggesting that the underlying neuronal dynamics might be manifested directly at the behavioral level. Here we applied this time-resolved behavioral approach to access the temporal dynamics of global/local processing. The stimuli are a big arrow (global) composed of small arrows (local). The global and local properties are either congruent or incongruent. After cued by a text of 'Global' or 'Local', subjects reported the global or local properties of the target as fast as possible and their reaction times were recorded. The cue-to-target SOA varied from 110 to 600ms in steps of 10ms. First, we replicated typical global precedence effects (global faster than local) in slow trends of the time-resolved behavioral time courses. Second, with the slow trends removed, the global processing demonstrated a rhythmic fluctuation at alpha band (~10 Hz) and in phase between the congruent and incongruent conditions, suggesting that the global property is sampled in an alpha-band rhythm, not influenced by local properties. Relatively in contrast, the local processing showed an 'out-of-phase' relationship between the congruent and incongruent conditions, implicating that local processing is largely driven and modulated by the global property. Third, the local processing demonstrated a trend towards beta band (~30 Hz) in congruent condition. However, under incongruent condition, the local processing shifted towards the alpha-band rhythm, suggesting the strong modulations of local processing by global properties. Our results provide behavioral evidence supporting the central role of neuronal oscillations in global/local processing. Our study also speaks for a dynamic framework within which global and local properties are processed and dynamically interacted with each other.
Meeting abstract presented at VSS 2016