Helen E. Feibes¹, Spencer R. Loggia^1,2, Isabela Téllez¹, Bevil R. Conway^1,3,4; ¹National Eye Institute, National Institutes of Health, ²Brown University, ³National Institute of Mental Health, National Institutes of Health, ⁴University of Maryland, College Park

Where do color categories come from? Prior work suggests that consensus color categories are not innate, which raises the hypothesis that they derive from colors of consensus behavioral relevance (Garside et al., PNAS 2025). To test the hypothesis, we trained two monkeys over nine years to learn the color-shape associations of twelve 2-D objects whose colors were six hues evenly sampling CIELUV, at two luminance levels. We then tested if the monkeys’ experience with these highly relevant objects impacted color matching in a delayed-match-to-sample task (Bae et al., J Exp. Psych 2015), as predicted by the hypothesis. We used 83 equally spaced colors in CIELUV. In each trial, the animals were shown a Gaussian colored blob (“sample”); in the subsequent frame they were rewarded for touching the exact match out of four match options. We analyzed errors (75,355 trials in monkey 1, 55,104 trials in monkey 2; across 25 sessions). Mixture modeling showed evidence for 4-6 category biases in each monkey, with individual differences in their precise location in color space. Control animals showed 2 biases, attributable to underlying stimulus space nonuniformities (Garside et al., PNAS 2025). The results provide evidence in support of the hypothesis that long-term experience with a limited set of highly relevant colors contributes to the construction of cognitive color categories, which manifest in tasks unrelated to the behavioral relevance of the colors. We then used target confusability competition modeling (Schurgin et al., Nat Human Behav 2020) to try to disambiguate categorical biases from stimulus nonuniformities, but the results were inconclusive, probably because categorical biases, stimulus space nonuniformities, and individual differences interact in an unknown way. These results imply that color categories reflect an interaction of stimulus space nonuniformities and individual differences.

Acknowledgements: Supported by NIH IRP (NIH, 1ZIAEY000558 to BRC), NSF (0918064 to BRC) and NIH (R01 EY023322 to BRC). Contributions of NIH authors are considered Works of the United States Government. The findings and conclusions are those of the authors and do not necessarily reflect the views of the NIH or the U.S. Department of Health and Human Services.