Abstract
The striking similarity in basic color naming systems observed worldwide is often attributed to universal biases in the perceptual representation of color (Berlin & Kay, 1969) that guides color lexicon creation and evolution. Using a modified version of Boster's (1986) paradigm for exploring this idea, we asked native English speakers to sort, in separate trials, representative color palettes into n=2...6 piles based on perceptual similarities. In Study I, 45 university undergraduates sorted 145 Munsell chips (5 achromatic, 140 chromatic) that uniformly sampled the variations in hue, chroma and value of the World Color Survey (WCS) chart. Cluster analysis across all n-sorts revealed 15 canonical classification templates, some of which were similar to universal WCS lexical categories, red, green, grue, warm, cool (Lindsey & Brown, 2006). Others, like dark and light, were unlike any universal WCS categories, or those found in English (Lindsey & Brown, 2014). When the classified sorting data were united into color category systems (one for each subject and each n-sort), there were a modest number of motifs that recurred across the n-sorts. Some, but not all, resembled WCS motifs (Lindsey & Brown, 2009). Similar results were obtained in Study II, where a diverse group of 90 adult subjects at our local science museum sorted colors from a palette of 30 representative WCS colors displayed and manipulated on custom-programmed iPads. In Study III, 55 additional subjects sorted 25 test colors varying in hue but similar in Munsell chroma and value. Cluster analysis of Study III revealed 10 canonical classification templates. Some of these resembled universal WCS lexical categories, whereas others resembled minority WCS categories that deviated from predictions of Berlin & Kay's theory. This is interesting because the minority categories come from preindustrial cultures yet are manifest in the n-sorts of technologically modern English speakers.
Meeting abstract presented at VSS 2018