Amanda K Robinson, Greta Stuart, Sophia M Shatek, Adrian Herbert, Jessica Taubert
{"title":"Neural correlates reveal separate stages of spontaneous face perception.","authors":"Amanda K Robinson, Greta Stuart, Sophia M Shatek, Adrian Herbert, Jessica Taubert","doi":"10.1038/s44271-025-00308-4","DOIUrl":null,"url":null,"abstract":"<p><p>The human brain rapidly detects faces, even in inanimate objects-a phenomenon known as face pareidolia. While this illusion reveals the automaticity of face detection, it also presents a paradox: how does the brain process stimuli that are simultaneously faces and objects? Here, we combined behavioral experiments with electroencephalography to track the temporal dynamics of face pareidolia processing. Using a large stimulus set of human faces, objects containing illusory faces, and matched control objects, we show that perception of face pareidolia is remarkably flexible and task dependent. When making spontaneous similarity judgments, participants (N = 338) perceived illusory faces as intermediate between faces and objects. However, in explicit categorization tasks (N = 22), the same stimuli were predominantly classified as objects, while rating face-likeness (N = 20) bolstered the representation of face-like features. Neural responses (N = 20) tracked this perceptual flexibility: early visual processing (90-130 ms) correlated with face-like judgments, while later activity (150-210 ms) aligned with object categorization. This temporal progression demonstrates how the brain maintains multiple levels of representation, integrating early face detection with subsequent object recognition to support flexible behavior. Our findings demonstrate that face pareidolia exemplifies the brain's capacity to resolve perceptual ambiguity through dynamic processing, with task demands determining how competing representations contribute to perception.</p>","PeriodicalId":501698,"journal":{"name":"Communications Psychology","volume":"3 1","pages":"126"},"PeriodicalIF":0.0000,"publicationDate":"2025-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12361498/pdf/","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Communications Psychology","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1038/s44271-025-00308-4","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
The human brain rapidly detects faces, even in inanimate objects-a phenomenon known as face pareidolia. While this illusion reveals the automaticity of face detection, it also presents a paradox: how does the brain process stimuli that are simultaneously faces and objects? Here, we combined behavioral experiments with electroencephalography to track the temporal dynamics of face pareidolia processing. Using a large stimulus set of human faces, objects containing illusory faces, and matched control objects, we show that perception of face pareidolia is remarkably flexible and task dependent. When making spontaneous similarity judgments, participants (N = 338) perceived illusory faces as intermediate between faces and objects. However, in explicit categorization tasks (N = 22), the same stimuli were predominantly classified as objects, while rating face-likeness (N = 20) bolstered the representation of face-like features. Neural responses (N = 20) tracked this perceptual flexibility: early visual processing (90-130 ms) correlated with face-like judgments, while later activity (150-210 ms) aligned with object categorization. This temporal progression demonstrates how the brain maintains multiple levels of representation, integrating early face detection with subsequent object recognition to support flexible behavior. Our findings demonstrate that face pareidolia exemplifies the brain's capacity to resolve perceptual ambiguity through dynamic processing, with task demands determining how competing representations contribute to perception.