Abstract
Navon hierarchical letter stimuli are often used to measure interference between global and local visual information, with global-to-local interference effects being usually interpreted as evidence for perceptual grouping (e.g., Behrmann & Kimchi, 2003). The relationship between this measure and holistic/configural face perception, namely the encoding of the multiple features of a face as a whole representation, remains elusive. In a recent study, Gao and colleagues (2011) showed that global processing of Navon stimuli increased the sensitivity to incongruence between the upper and the lower halves of a composite face and concluded that global processing of Navon stimuli augments holistic face processing. However, rather than using a standard composite face matching paradigm, the authors used a congruency paradigm with composite faces (e.g., Richler et al., 2008), namely a paradigm that has a built-in decisional confound and includes responses on ‘different’ trials likely reflecting part-based judgments (Rossion, in press). Here we reanalyzed Gao et al. (2011)’s data by examining only the ‘same’ trials, as in a standard composite face measure, and found that global processing of Navon stimuli did not increase the composite face effect, neither in accuracy rates nor in correct RTs. The effect initially described is due only to a priming of the ‘different’ composite face trials, reflecting part-based judgments. These findings reinforce the view that the congruency paradigm with composite faces measures general processes that should not be mistaken with face-specific holistic/configural perceptual processes measured in the standard composite face paradigm. More generally and importantly, they support the view that processing global aspects of hierarchical Navon stimuli is functionally distinct than holistic/configural perception of individual faces (Busigny & Rossion, 2011) and suggest that correlations between the two measures across individuals or impairments at the Navon task in cases of congenital prosopagnosia (Behrmann et al., 2005) essentially reflects general visual processes.
Meeting abstract presented at VSS 2013