Exploring Children’s Use of Discreteness, Stability, and Labels as Cues to Essentialize Novel Categories

• In Event: 3-134 - The Essence of Biological Categories: Factors Influencing Children’s Sensitivity to Variability and Change

Integrative Statement

Background: Children treat natural and social categories as “essential,” believing that individuals have an unobservable “essence” that determines their category membership (Gelman, 2003). For categories children already represent as ‘essential’, children believe that a) these categories are discrete, with mutually exclusive, sharp boundaries (e.g., an individual cannot be “somewhat” a dog), b) category membership is stable over time (e.g., a dog will always be a dog), and c) categories map directly onto labels (e.g., the category of dogs maps onto the word “dog”). Here we explore whether children also use these properties of discreteness, stability, and labels to determine whether to ‘essentialize’ a novel category.

Experiment 1: Five- to six-year-old participants (n = 140) were introduced to a population of novel creatures that varied in shape, color, and texture. During the Familiarization phase, participants were given a cue (Discreteness, Stability, or Labels –between-subjects design) that suggested that creatures could be grouped based on texture, rather than shape or color (full procedure depicted in Figure 1). In the Discreteness condition, texture varied discretely (i.e., either fuzzy or smooth), but color and shape varied along a spectrum. In the Stability condition, texture remained stable as creatures grew older, but color and shape changed. In the Labels condition, creatures were labeled differentially according to texture, but not according to color or shape. In the Control condition, participants were given no cues. Next, during the Property phase, participants were taught about a kind-relevant property of a target creature (e.g., “this one lives in the mountains and eats berries”). Finally, during six forced-choice Test trials, we measured whether children generalized this property to a texture-matched or texture-mismatched creature, which was otherwise similar to the target creature. We found that participants in the Labels condition, but not the Stability or Discreteness conditions, used texture to generalize kind-like properties, and thus essentialize the novel category, significantly more often than participants in the Control condition (Χ2(1) = 4.76, p < .05).

Experiment 2: In a follow-up experiment, we explored whether participants are unable to use these properties to essentialize a category or simply have more difficulty detecting these cues. We provided participants (n = 81) with increased exposure to the stability of texture over time, and also included a memory check to ensure that participants encode the cue information. We found that, despite this additional exposure, relatively few participants (n = 40) remembered the cue; however, those children were significantly more successful at spontaneously inferring an essential social group based on texture compared to children who did not remember the cue (t(79) = 2.63, p = .01).

Discussion: These results suggest children most readily use labels to spontaneously infer a novel essential social category. However, children may not use cues to stability and discreteness because these properties are less salient, rather than because children do not represent this information as a conceptually relevant indicator of an essential category. Children who do encode stability and discreteness information may be able to use this information to essentialize a novel group.

Authors

• Rebecca Peretz-Lange, Tufts University
  Presenting Author

• Paul Muentener, Tufts University
  Non-Presenting Author