When Does It Stop Being Blue? Children’s and Adults’ Categorization of Color Categories

• In Event: Poster Session 02
  In Poster Session: PS 02 Topic - Language, Communication

Abstract

How does language knowledge affect children’s categorization of continuous information into discrete categories? Color is one such continuous factor that humans carve into discrete categories using color words like ‘blue’ and ‘purple’. We are not born knowing these color categories: instead, we learn to categorize colors by acquiring language-specific color words, as demonstrated by differences in how speakers of different languages categorize color (Roberson, Davies & Davidoff, 2000). As such, studying children’s perception of color can give us insight into how specific aspects of language knowledge impact children’s development of categories.
In pilot work, we collected data on how children aged 4-6 years (x̅ age=4.84, SD=0.47) and adults (x̅ age=20.07, SD=1.77) categorized a range of colors that fall between adjacent color pairs (like blue-purple, green-yellow, etc.; see Fig. 1) and their ability to identify a range of colors as a test of receptive color vocabulary (i.e., by asking them to point to colors like ‘blue’ or ‘periwinkle’ from an array of possible colors; see Fig. 2). In addition, for children we collected a measure of general receptive vocabulary (i.e., PPVT-4; Dunn & Dunn, 2007) and parental estimates of a child’s specific color word knowledge. Parents’ were accurate in estimating their child’s receptive color word knowledge (b=1.12, SE=0.14, z=7.84, p<.001; N=16).
We found that children with greater receptive color vocabularies (e.g., knowledge of basic color words like ‘blue’ or ‘green’ but also more subordinate color words, like ‘indigo’ or ‘lime’) were more likely to accept a greater range of colors as belonging to a basic color category, than were children with smaller receptive color vocabularies (b=0.25, SE=0.10, t=2.41, p=.02; N=35). Interestingly, our adult data showed this same pattern (b=0.33, SE=0.13, t=2.40, p=.04; N=14). And while general receptive vocabulary was positively correlated with receptive color vocabulary in children (r(33)=0.44, p=.01; N=35), only the latter predicted children’s propensity to accept more colors as belonging to a basic color category. We propose that knowing more color words gives children and adults a better grasp of the hierarchical structure and flexible nature of categories. That is, that perceptual information can be labeled (and categorized) in multiple ways, and that labels may reflect different levels of categorization (i.e., a color that can be called ‘indigo’ can also be categorized at the more basic level of ‘blue’).
We are now collecting online data, aiming for N=40 children. Additionally, we are asking our adult participants to gauge how well specific colors (e.g., ‘periwinkle’) fit under more basic color words (like ‘blue’ vs. ‘purple’), and plan on developing a child-friendly version of this task. This data will reveal how people may hierarchically represent the colors used within our tasks. Data will be collected and analyzed by January 2021. This research will build upon work suggesting that language scaffolds category development (e.g., Waxman & Markow, 1995; Plunkett, Hu & Cohen, 2008), to examine specifically how language may support children’s learning that categories are hierarchical and flexible and, thus, that some things may be categorized in multiple ways.

Authors

• Ellise Suffill, University of Wisconsin - Madison
  Presenting Author

• Haley A Vlach, University of Wisconsin - Madison
  Non-Presenting Author