Uncanny valley as a window into predictive processing in the social brain

NEUROPSYCHOLOGIA

卷:114

Article

Urgen, Burcu A.^1,2  Kutas, Marta^1,3  Saygin, Ayse P.^1,3 

[1] Univ Calif San Diego, Dept Cognit Sci, 9500 Gilman Dr, La Jolla, CA 92093 USA

[2] Univ Parma, Dept Med & Surg, Via Volturno 39, I-43125 Parma, Italy

[3] Univ Calif San Diego, Neurosci Program, 9500 Gilman Dr, La Jolla, CA 92093 USA

关键词: Uncanny valley;    Predictive processing;    N400;    Action perception;    Social neuroscience;   

DOI  :  10.1016/j.neuropsychologia.2018.04.027

来源: Elsevier

【 摘 要 】

Uncanny valley refers to humans' negative reaction to almost-but-not-quite-human agents. Theoretical work proposes prediction violation as an explanation for uncanny valley but no empirical work has directly tested it. Here, we provide evidence that supports this theory using event-related brain potential recordings from the human scalp. Human subjects were presented images and videos of three agents as EEG was recorded: a real human, a mechanical robot, and a realistic robot in between. The real human and the mechanical robot had congruent appearance and motion whereas the realistic robot had incongruent appearance and motion. We hypothesize that the appearance of the agent would provide a context to predict her movement, and accordingly the perception of the realistic robot would elicit an N400 effect indicating the violation of predictions, whereas the human and the mechanical robot would not. Our data confirmed this hypothesis suggesting that uncanny valley could be explained by violation of one's predictions about human norms when encountered with realistic but artificial human forms. Importantly, our results implicate that the mechanisms underlying perception of other individuals in our environment are predictive in nature.

【 授权许可】

Free