Rosch's early studies were on color. She learned of the Berlin-Kay color research midway through her own research and found that their results meshed with her own work on Dani, a New Guinea language that has only two basic color categories: mili (dark-cool, including black, green, and blue) and mola (light-warm, including white, red, yellow). Berlin and Kay had shown that focal colors had a special status within color categoraies-that of the best example of the category. Rosch found that Dani speakers, when asked for the best exampies of their two color categories, chose focal colors, for example, white, red, or yellow for mola with different speakers making different choices.
In a remarkable set of experiments, Rosch set out to show that primary color categories were psychologically real for speakers of Dani, even though they were not named. She set out to challenge one of Wharf's hypotheses,namely, that language determines one's conceptual system. If Wharf were right on this matter, the Dani's two words for colors would determine two and only two conceptual categories of colors. Rosch reasoned that if it was language alone that determined color categorization, then the Dani should have equal difficulty learning new words for colors, no matter whether the color ranges had a primary color at the center or a nonprimary color. She then went about studying how Dani speakers would learn new, made-up color terms. One group was taught arbitrary names for eight focal colors, and another group, arbitrary names for eight nonfocal colors (Rosch 1973). The names for focal colors were learned more easily. Dani speakers were also found (like English speakers) to be able to remember focal colors better than nonfocal colors (Heider 1972).
In an experiment in which speakers judged color similarity, the Dani were shown to represent colors in memory the same way English speakers do (Heider and Olivier 1972).