Neural Signatures of the Configural Superiority Effect and Fundamental Emergent Features in Human Vision

Thiago Leiros Costa,Kimberley Orsten-Hooge,Gabriel Gaudêncio Rêgo,Johan Wagemans,James R Pomerantz,Paulo Sérgio Boggio

doi:10.1038/s41598-018-32289-2

Abstract

The concepts of grouping, emergence, and superadditivity (when a whole is qualitatively different from the sum of its parts) are critical in Gestalt psychology and essential to properly understand the information processing mechanisms underlying visual perception. However, very little is known about the neural processes behind these phenomena (particularly in terms of their generality vs. specificity and their time-course). Here, we used the configural superiority effect as a way to define “emergence” and “emergent features” operationally, employing an approach that can isolate different emergent features and compare them on a common scale. By assessing well-established event related potentials in a HD-EEG system, we found that the critical processes behind configural superiority and superadditive Gestalt phenomena are present in the window between 100 and 200 ms after stimulus onset and that these effects seem to be driven by specific attentional selection mechanisms. Also, some emergent features seem to be differentially processed in different brain hemispheres. These results shed new light on the issues of the generality vs. specificity of the neural correlates of different Gestalt principles, the hemispheric asymmetries in the processing of hierarchical image structure and the role of the N1 ERP component in reflecting feature selective mechanisms.

Highlights

If discrimination of elements is aided by adding redundant uninformative context, we say that a Configural Superiority Effect has happened
That the cleanest way to assess the generality vs. specificity of the neural correlates of basic EFs is with the current approach: (i) defining EFs operationally based on their effects on discrimination performance; (ii) using stimuli and tasks that allow them to be compared on a common scale, and (iii) in a high-density EEG setup, providing the required high temporal resolution
For the P1 event related potentials (ERPs), we found a significant effect of Display Type [F2,30 = 6.68, p < 0.01, ηp2 = 0.30] and post-hoc tests showed that Superiority was significantly different from basic one (Base) (p = 0.01) but not from Inferiority (p = 0.7)

Summary

Introduction

If discrimination of elements is aided by adding redundant uninformative context, we say that a Configural Superiority Effect has happened (see Fig. 1 for a classical example). The CSE is a remarkable demonstration of the holistic nature of visual perception and the classic Gestalt motto of the whole being different from the sum of the parts In this context, Pomerantz and Portillo[8] proposed www.nature.com/scientificreports/. Gestalt psychologists have proposed an extensive number of principles based mostly on visual demonstrations and missing operational definitions, something that has proven critical in moving the field forward[1,2] Operationalizing these Gestalt principles in terms of their specific effects in discrimination performance could provide us with a common scale of measurement to compare them. In this same line of thought, a recent review by Wagemans[20] supports the idea that “not all Gestalts are equal” based both on theory and empirical findings from contemporary visual neuroscience studies

Results

Discussion

Conclusion