"Threat-unrelated" properties: An ill-defined concept. A reply to "The danger of interpreting detection differences between image categories" (Gayet, Stein, & Peelen, 2019)

Abstract

In our previous work (Gomes, Soares, Silva, & Silva, 2018), we evidenced that snakes (vs. birds), an archetypal threat-related stimulus, have an advantage in accessing visual awareness during continuous flash suppression. This advantage was observed when the images were filtered in low spatial frequency (LSF) but not in high spatial frequency and interpreted as supporting the role of a subcortical pathway (superior colliculus-pulvinar) to the amygdala in threat detection, thought to be sensitive to LSF but not to high spatial frequency information. Recently Gayet, Stein, and Peelen (2019), using stimuli without differences in the threat they represented (bicycles and cars) but with an analogous perimeter-area ratio as snakes and birds (respectively), evidenced that bicycles had a faster access to visual awareness than cars. As in our study, this advantage relied on LSF information. The authors argued that the images' perimeter-area ratio (a characteristic intrinsically related with stimuli shape referred to by the authors as threat-unrelated properties), but not the threat itself, may have driven the detection advantage observed in our study and the association with the action of the subcortical pathway to the amygdala. We propose that the results obtained by Gayet et al. do not undermine ours and discuss the relevance of higher perimeter-area ratios (such as images with elongated shapes; e.g., snakes) as a threat-related property required for a fast-initial snake processing. We argue that stimuli shape is a prime feature for the fast-initial threat processing, relevant to grab early visual attention, an effect that might explain the obtained results. (PsycINFO Database Record (c) 2019 APA, all rights reserved).