RESEARCH ARTICLE Atypical Perception in Autism A Failure of Perceptual Specialization? Bat Sheva Hadad , Eugenia K. Goldstein, and Natalie N. Russo We examined whether reduced perceptual specialization underlies atypical perception in autism spectrum disorder (ASD) testing classifications of stimuli that differ either along integral dimensions (prototypical integral dimensions of value and chroma), or along separable dimensions (prototypical separable dimensions of value and size). Current models of the per- ception of individuals with an ASD would suggest that on these tasks, individuals with ASD would be as, or more, likely to process dimensions as separable, regardless of whether they represented separable or integrated dimensions. In contrast, reduced specialization would propose that individuals with ASD would respond in a more integral manner to stimuli that differ along separable dimensions, and at the same time, respond in a more separable manner to stimuli that differ along integral dimensions. A group of nineteen adults diagnosed with high functioning ASD and seventeen typically developing participants of similar age and IQ, were tested on speeded and restricted classifications tasks. Consistent with the reduced specialization account, results show that individuals with ASD do not always respond more analytically than typically developed (TD) observers: Dimensions identified as integral for TD individuals evoke less integral responding in individu- als with ASD, while those identified as separable evoke less analytic responding. These results suggest that perceptual rep- resentations are more broadly tuned and more flexibly represented in ASD. Autism Res 2017, 0: 000–000. V C 2017 International Society for Autism Research, Wiley Periodicals, Inc. Keywords: ASD; autism; perception; perceptual specialization; color perception; separable and integral perception; predictive coding Introduction Perceptual processing of ordinary scenes often involves identifying the structure of the sensory input and decom- posing multidimensional sensations into separate psy- chophysical dimensions [e.g., Jones & Goldstone, 2013; Schyns, Goldstone, & Thibaut, 1998]. This allows, for example, the extraction of shape representations, spatial positions, colors, and the orientation of environmental objects independently from one another. Object repre- sentations that are invariant across changes in identity- preserving variables, such as orientation, are critical for invariant object recognition [Soto & Ashby, 2015]. The perceptual system becomes specialized with age or expertize, learning to carve a scene into useful compo- nents that comprise the perceptual building blocks used for representing objects [e.g., Goldstone, 2003]. The spe- cialized perceptual system thus tends to integrate stimu- lus components together if they co-occur, and to separate these bundles from other statistically independent bundles. Accordingly, perceptual decomposition occurs more readily for separable dimensions—properties that are conjunctions of values on component dimensions, but less for integral dimensions—properties that are phys- ically multidimensional but psychologically difficult to analyze (e.g., color dimensions). Considerable evidence suggests that higher-level cognitive mechanisms, such as selective attention [Garner, 1970, 1974; Goldstone, 1994] and working memory functions [Bae & Flombaum, 2013] are applied differently when stimuli differ along separa- ble dimensions rather than along integral dimensions. Studying perceptual separability and integrality is thus important for understanding visual perception and high-level cognition. Individuals diagnosed with an Autism Spectrum Dis- order (ASD) demonstrate superior processing of detailed local information on a variety of perceptual tasks and across different sensory modalities. For example, as compared to typically developing (TD) individuals, they are better able to detect local targets in visual search tasks [Plaisted, O’Riordan, & Baron-Cohen, 1998b], ignore the influence of increasing numbers of dis- tracters during visual search [O’Riordan, Plaisted, Driv- er, & Baron-Cohen, 2001], and identify fine stimulus From the Edmond J. Safra Brain Research Center, University of Haifa, Haifa, 31905, Israel (B.S.H., E.K.G.); Cognition, Brain and Behavior, Syracuse University, 403 Huntington Hall, Syracuse, NY (N.N.R.) Received December 22, 2016; accepted for publication March 20, 2017 Address for correspondence and reprints: Bat Sheva Hadad, Edmond J. Safra Brain Research Center, University of Haifa, Haifa 31905, Israel. E-mail: bhadad@univ.haifa.ac.il Published online 00 Month 2017 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/aur.1800 V C 2017 International Society for Autism Research, Wiley Periodicals, Inc. INSAR Autism Research 00: 00–00, 2017 1