Is visual perception WEIRD? The Müller-Lyer illusion and the Cultural Byproduct Hypothesis

A fundamental question in the psychological sciences is the degree to which culture shapes core cognitive processes — perhaps none more foundational than how we perceive the world around us. A dramatic and oft-cited “case study” of culture’s power in this regard is the Müller-Lyer illusion, which depicts two lines of equal length but with arrowheads pointing either inward or outward, creating the illusion that one line is longer than the other. According to a line of research stretching back over a century, depending on the society you were raised in (and how much carpentry you were exposed to), you may not see the illusion at all — an ambitious and influential research program motivating claims that seemingly basic aspects of visual processing may actually be “culturally evolved byproducts”. This Cultural Byproduct Hypothesis bears on foundational issues in the science, philosophy, and sociology of psychology, and remains popular today. Yet, here we argue that it is almost certainly false. We synthesize evidence from diverse fields which demonstrate that: (1) the illusion is not limited to humans, appearing in non-human animals from diverse ecologies; (2) the statistics of natural scenes are sufficient to capture the illusion; (3) the illusion does not require straight lines typical of carpentry (nor even any lines at all); (4) the illusion arises in sense modalities other than vision; and (5) the illusion arises even in congenitally blind subjects. Moreover, by reexamining historical data and ethnographic descriptions from the original case studies, we show that the evidence for cultural variation and its correlation with key cultural variables is in fact highly inconsistent, beset by questionable research practices, and misreported by later discussions. Together, these considerations undermine the most popular and dramatic example of cultural influence on perception. We further extend our case beyond this phenomenon, showing that many of these considerations apply to other visual illusions as well, including similarly implicated visual phenomena such as the Ebbinghaus, Ponzo, Poggendorf, and Horizontal-Vertical illusions. We conclude by outlining future approaches to cross-cultural research on perception, and we also point to other potential sources of cultural variation in visual processing.