Skip to main content
. 2021 Sep 27;11:19095. doi: 10.1038/s41598-021-98550-3

Figure 1.

Figure 1

(Panel A): Map of languages and populations in our data. Each circle represents one language and corresponding population, with the presence (blue) or absence (yellow) for a specific term for ’blue’. The circle radius is proportional to the population frequency of red/green abnormal color perception. The intensity of the background magenta is proportional to the incidence of UV-B radiation as given by the NASA Total Ozone Mapping Spectrometer (TOMS) for the year 1998. This map was generated with the package maps in R version 4.0.559, which uses public domain data from the Natural Earth project https://www.naturalearthdata.com. (Panel B): Posterior density plots of the best predictors of the presence of a word for blue across Bayesian regression models. (Panel C): Posterior density plots of the best predictors of the frequency of red-green abnormal color perception across Bayesian regression models. The plots in panels (B) and (C) are based on the analyses detailed in the “Results” section and were generated using R59. (Panel D): The lens brunescence hypothesis, as detailed in this paper. UV rays cause the lens to opacify and become yellower (top), filtering blue light, so that the perceived spectrum (on the right) contains less blue and more green, compared to the non-affected eye (bottom). (Panel E): A simulation of different types of color perception abnormalities. From left to right: normal vision; one type of red-green abnormal color perception (deuteranopia); strong lens brunescence (yellow filter and darkening of the image); both lens brunescence and deuteranopia, hypothesized to result in decreased biological fitness relative to other color vision types. Background image cropped from a photo by Joydeep Pal on Unspalsh (https://unsplash.com/photos/i6kccimZz_8) under a free-to-use license (https://unsplash.com/license). For deuteranopia we used Coblis—The Color BLIndness Simulator (https://www.color-blindness.com/coblis-color-blindness-simulator/) and checked the results with the Colorblind Web Page Filter (https://www.toptal.com/designers/colorfilter). The lens brunescence filter is inspired by30: we used Adobe Lightroom classic version 10.0 with a yellow filter and darkening to simulate a Kodak yellow filter. For more details, see the “Results”, “Materials and Methods” and Supplementary Materials.