Purchase this article with an account.
Leslie Blaha, Thomas Busey; Electrophysiological substrates of configural learning. Journal of Vision 2007;7(9):796. doi: 10.1167/7.9.796.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Perceptual unitization, or “perceptual chunking,” of novel object features results in a fundamental shift in perceptual processing strategy, as evidenced by the qualitative shift in processing capacity found by Blaha and Townsend (VSS, 2006). Unitization thus provides a potential learning mechanism for the development of configural processing strategies or representations. Evidence from studies of both real-life and laboratory-trained experts demonstrate N170 differences for the visual response to objects of expertise. Often these are similar to the responses to faces. Researchers proposed this neurological response of visual expertise results from configural processing strategies, to which the N170 is sensitive (Busey & Vanderkolk, 2005). We propose that the observed changes in information processing over the course of configural learning should be accompanied by changes in neurological measures of perception; in particular, we expect the N170 to change as configural object representations are developed in a unitization learning task. Participants unitized novel object features in a categorization task requiring conjunctive processing for correct category identification. Capacity analyses of response times replicate our previous findings of shifts from extreme limited to extreme super capacity processing over the course of learning. On the initial training day, participants exhibited no N170 differences between novel objects. Seven days of training resulted in an N170 amplitude change for the unitized object. Unitization thus alters this neural response when objects can be processed in a configural manner, as indexed by the behavioral measures. These results provide converging evidence that configural learning via perceptual unitization fundamentally alters the perceptual processing strategies or representations employed in object identification. Further, this study may serve as a platform for developing models to bridge response time behavioral measures and the neurological signals in ERPs.
This PDF is available to Subscribers Only