Bornstein (1973) first suggested that genetic constraints could influence colour categorisation, hence populations nearer the equator having fewer colour terms. Lindsey and Brown (2002, p. 510) argue that colour terms vary cross-culturally on a “fundamentally biological basis”. They showed that the number of colour terms in a language correlated with the amount of UV-B radiation to which speakers were subjected. This ‘burnescence’ effect increases the density of the lens. Over time, populations would adapt to have denser lenses nearer the equator. Denser lenses skew hues towards green, ‘squeezing out’ the blue hues and the blue term with it (however, see Hardy et al., 2004 and Regier & Kay, 2004 for counter-arguments). Furthermore, Lindsey and Brown (2002) conducted a WCS-style Munsell-chip experiment on a culturally homogeneous population with the stimuli virtually manipulated to simulate the effects of burnescence. This was an innovative solution to controlling for individual differences. It was found that participants’ use of “green” extended further into “blue” areas with greater simulated burnescence. However, Hardy et al. (2004) repeated the burnescence experiment with the addition of a condition where older participants were given stimuli virtually manipulated to simulate less burnescence. The results did not support Lindsey and Brown’s hypothesis.
Models of genetic constraints have found that populations can converge on shared categorisations for colour if reproduction is based on discriminative success: Steels & Belpaeme (2005) constructed computational models with 10 agents with the ability to categorise colours according to perceptions (radial basis function neural networks for each colour category). The innate biases of these categories were defined by genetic encodings, but the categories did not change over a lifetime. That is, individual learning was not modelled. Agents played the discrimination game 50 times before half of the population was replaced with new agents generated by a mutation of an existing agent (asexual reproduction). The extent of the mutations were based on the inverse of the parent’s discriminatory success. That is, genes which facilitated success in the task were less likely to mutate. Category variance between agents and between populations was used to measure the extent to which agents had a common categorisation, a small value meaning that many agents share large parts of their categorisations. With these constraints, systems became totally shared within populations. However, more categories were produced than for individual learning, the synchronisation of categories took longer and the system was less robust against changes in the environment. Also, the systems were not shared across populations.
Jameson and Komerova (2009) found that modelling a heterogeneous population of dichromats and trichromats further constrained the convergence, aiding the emergence of shared categories. In conclusion, genetic constraints seem to interact with other constraints to affect colour categorisation. However,genetic change happens on a large timescale. Colour terms have evolved in a time span with very little cross-cultural genetic change, suggesting that other processes have a more immediate effect.
Bornstein, M. (1973). Color vision and color naming: A psychophysiological hypothesis of cultural difference. Psychological Bulletin, 80 (4), 257-285 DOI: 10.1037/h0034837
Lindsey, D., & Brown, A. (2002). Color Naming and the Phototoxic Effects of Sunlight on the Eye Psychological Science, 13 (6), 506-512 DOI: 10.1111/1467-9280.00489
Hardy, J., Frederick, C., Kay, P., & Werner, J. (2004). Color naming and lens brunescence Journal of Vision, 4 (8), 56-56 DOI: 10.1167/4.8.56
Regier T, & Kay P (2004). Color naming and sunlight: commentary on Lindsey and Brown (2002). Psychological science : a journal of the American Psychological Society / APS, 15 (4) PMID: 15043652
Steels, L., & Belpaeme, T. (2005). Coordinating perceptually grounded categories through language: A case study for colour Behavioral and Brain Sciences, 28 (04) DOI: 10.1017/S0140525X05000087
Jameson KA, & Komarova NL (2009). Evolutionary models of color categorization. II. Realistic observer models and population heterogeneity. Journal of the Optical Society of America. A, Optics, image science, and vision, 26 (6), 1424-36 PMID: 19488182