Poster #196 - Children's Evaluations of People who Profess Ignorance

• In Event: 1-115 - Poster Session 03
  In Poster Session: PS 03 Section - Social Cognition

Integrative Statement

By age 3, children prefer knowledgeable sources over ignorant and inaccurate sources (Sabbagh & Baldwin, 2001; Koenig & Harris, 2005; Rakoczy, Ehrling, Harris & Schultze, 2015). This preference has been interpreted as a demonstration of vigilance against misinformation from agents who do not fit the child’s baseline expectation of accuracy. However, recent work shows that preschoolers will learn new words and facts from a single speaker who previously professed ignorance, but maintain vigilance against a single inaccurate speaker (Kushnir & Koenig, 2017). Thus, children may treat professions of ignorance as more of a situational constraint, one that does not penalize a speaker’s subsequent claims to know. To further explore this, we employed a contrasting (two-speaker) design, investigating preschoolers’ information-seeking and endorsements when a speaker who admitted ignorance was contrasted with a speaker whose knowledge state was not revealed.
Children (N= 41; 3.50-4.89 years old, M = 4.08 years, SD = .42, 16 girls) saw two speakers; one admitted ignorance of three familiar objects, and one had three neutral, non-epistemic interactions. Following this, children made predictions about each speaker’s knowledge of three additional familiar objects, chose which speaker to ask for names and functions of novel objects (two each), and then chose whether to endorse one of their conflicting labels and functions for those objects.
Children predicted that both previously ignorant and neutral speakers would correctly label and use the three familiar objects. This supports the idea that children do not use a history of admitted ignorance to infer ignorance more generally.
The contrast between children’s information seeking (i.e. “ask”) and endorsements provides further support for our hypothesis (see Table 1 for choice means and SD). On information-seeking trials, children chose to ask for novel labels from the neutral speaker with an undisclosed knowledge state more often than the ignorant speaker, t(39) = 1.69, p = .05.. However, once children were presented with novel labels from both speakers, they were equally likely to endorse the neutral or ignorant speaker’s label (M = .707, SD = .782; M = .561, SD = .808, respectively). Interestingly and conversely, children did not demonstrate a speaker preference when asking how to use novel objects, but were more likely to endorse the novel use demonstrated by the ignorant agentthan that shown by the neutral speaker, t(39) = 1.92, p = .03.
Our results show that preschoolers will learn new things from previously ignorant speakers even when a neutral informant is available. This offers support to the idea preschoolers’ view of ignorance is situational, rather than dispositional. We discuss results in light of work on epistemic vigilance, and suggest future directions examining individual differences depending on tendency of caregivers to admit ignorance and encourage children to question rather than accept all stated claims.

Authors

• Alyssa Varhol, Cornell University
  Presenting Author

• Tamar Kushnir, Cornell University
  Non-Presenting Author

• Melissa A Koenig, Institute of Child Development, University of Minnesota
  Non-Presenting Author