What's lost in inverted faces?

Repetition probability effects for inverted faces

“…Several studies suggested that face inversion disturbs the processing of configural/holistic information in a face more than that of the facial features either overall across the entire face (Rhodes et al, 1993;Cabeza and Kato, 2000;Freire et al, 2000;Leder and Bruce, 2000;Barton et al, 2001;Leder et al, 2001;Rossion and Gauthier, 2002;Leder and Carbon, 2006) or at least in the lower, mouth region (Xu and Tanaka, 2013;Tanaka et al, 2009). Other studies, however, emphasize that the processing of featural and configural information are equally disrupted by face inversion (Riesenhuber et al, 2004;Yovel and Kanwisher, 2004; for a review see Tanaka and Gordon, 2011).…”

Face Recognition in Young Children: When the Whole is Greater than the Sum of Its Parts

“…The more veridical norm representation can be used to resolve relational differences between highly similar faces, thereby leading to improved recognition of upright faces. Inversion, on the other hand, eliminates any expertise advantage by blocking norm-based encoding of relational properties (Bartlett & Searcy, 1993;Rhodes, Brake, & Atkinson, 1993;Tanaka & Sengco, 1997). Thus, the norm-based model predicts that, whereas recognition of upright faces should improve with age, recognition of inverted faces should remain fairly constant across different groups.…”

Face and object processing in autism spectrum disorders

“…The typical ability to discriminate and remember faces with high accuracy is acquired using specialist visual configural-processing strategies, rather than general strategies such as feature processing [Carey and Diamond, 1977;Diamond and Carey, 1986;Farah et al, 1998;Leder and Bruce, 1998;Maurer et al, 2002;Rhodes et al, 1993;Searcy and Bartlett, 1996;Sergent, 1984;Tanaka and Farah, 1993;Yovel and Kanwisher, 2004]. There is evidence to suggest that individuals with autism spectrum disorder (ASD) do not process faces as well as typically developing individuals [Davies et al, 1994;Hobson, 1986;Joseph and Tanaka, 2003;Klin et al, 1999;Langdell, 1978;Senju et al, 2003], but it remains unclear whether this deficit is attributable to configuralprocessing impairments.…”

“…There is still disagreement about the type of configural processing used by typically developing individuals to discriminate faces [Diamond and Carey, 1986;Farah et al, 1998;Gauthier and Tarr, 1997;Leder and Bruce, 1998;Maurer et al, 2002;Palmeri and Gauthier, 2004;Rhodes et al, 1993;Searcy and Bartlett, 1996;Yovel and Kanwisher, 2004]. Three types of configural information are considered to be important in face processing [Maurer et al, 2002]: (1) first-order information (the fixed structure of the eyes, above the nose, above the mouth); (2) holistic information (processing all componential properties without decomposition, or as a perceptual snapshot); (3) second-order information (the spatial relationship amongst the features).…”

“…Direct tests of second-order configural processing have measured participants' sensitivities to alterations made to the spatial distance between the facial features [e.g. moving the eyes further apart; Leder and Bruce, 1998;Rhodes et al, 1993]. Those researchers who support the holistic account argue that second-order configural properties and feature properties are processed interdependently to create a representation of an emergent ''whole'' [Tanaka and Sengco, 1997].…”