Categorization and naming of surface texture and color

Poster Presentation 43.344: Monday, May 20, 2024, 8:30 am – 12:30 pm, Banyan Breezeway
Session: Color, Light and Materials: Appearance, categories

Angela M. Brown1 (), Delwin T. Lindsey1; 1Ohio State University

According to the Sapir-Whorf hypothesis (“Sapir-Whorf”), language controls categorical perception. Thus, Sapir-Whorf predicts that terms should be associated easily and fluently with the categories they name, and well-understood categories should have high-consensus terms. ***We studied 99 physical samples of surfaces in an incompletely crossed design. Each sample had a color (red, orange, yellow, green, turquoise, blue, purple, black, white, or polychromatic) and a surface texture (smooth, transparent, metallic, ridged, nubbly, bumpy, rough, felt, burlap, vinyl upholstery, glitter or holographic). Sixteen subjects performed unconstrained sorts of the samples, based on sample appearance, then they provided a single term for the shared quality of the samples in each group. Four people sorted by color, eleven sorted by texture, and one sorted by both texture and color. ***People who sorted by color agreed what the 9–10 color categories were (90% agreement within categories), and they provided color terms for their categories quickly and easily. The color categories and the 14 color terms were nearly 1:1, and pairs of individuals who sorted by color agreed on the color term for 87.4% of samples (SEM=±3.4%). ***People who sorted by texture also agreed what the 11–16 texture categories were (95% agreement within categories), but they provided 100 different texture terms slowly and with difficulty. Words like “shiny” and “woven” named several different textures, and pairs of individuals who sorted by texture agreed on the texture term on only 11.7% ±3.3% of samples. ***There is a long literature on the difficulties with applying Sapir-Whorf to color terms, and this study does not strongly confirm or refute it for color. However, the Sapir-Whorf hypothesis clearly cannot apply to the understanding of the surface texture of physical samples, because texture naming was difficult, the association between texture terms and categories was complex, and observers’ texture textures showed low consensus.