Poster #197 - Speakers' Race and Accent Influence Preschoolers’ Uptake of Irregular Nouns and Verbs

• In Event: 2-055 - Poster Session 06
  In Poster Session: PS 06 Section - Social Cognition

Integrative Statement

Preschoolers sometimes say the incorrect, overregularized forms of familiar but irregular nouns and verbs: saying childs instead of children or ringed instead of rang. Even when a previously accurate speaker provides the correct form and a previously inaccurate speaker provides the overregularized one, preschoolers consistently select the wrong word as correct (Kondrad, McKercher & Jaswal, under review). Epistemic cues, like someone’s history of accuracy, may not be enough to override children’s strong draw towards the regularization paradigm that governs the way most nouns and verbs are transformed. It is perhaps not surprising that children trust the person who provides the word that is consistent with a rule that works 90% of the time – the odds are in her favor and so children can forgive a few incorrect object labels. But gaining accurate knowledge from others is not always the goal when interacting with them. There are also social goals, such as the drive to fit in with others who are “like me”. As in the Asch paradigm, people sometimes agree with someone even if they know they are wrong to strengthen their relationship. In the two studies presented here, we asked whether children would endorse someone who shares the same race and is from the same language community as them, even if they said something unexpected. In Study 1, half of the white 4-year-olds saw a white and a black speaker and the other half heard (but did not see) a native and foreign accented speaker. They were asked to decide who was saying the right thing about four novel labels and four irregular forms. The white or native speakers provided the correct irregulars. Next, those who initially saw the speakers’ race now also heard the speakers’ accent, or the other way around, and children completed four more novel label and four more irregular trials. As Figure 1 shows, children showed no preferences for a white over black speaker’s labels or irregulars and continued to be ambivalent once their accent was revealed. Children endorsed the native speaker’s labels, t(15) = 2.39, p = .03, d = 2.05, but not her irregulars, and this preference disappeared once race was revealed. Study 1 suggests that social cues, even when stacked, are not enough to overcome the draw of the regular paradigm. The procedure was the same in Study 2, except that children first got four familiarization trials in which they learned that the white or native speaker also had a history of providing correct object labels. As Figure 2 shows, children trusted the accurate speaker who looked and sounded like them across the board, whether they had one social cue or two ts (31) > 3.92; ps < .001, ds > 1.26. Neither epistemic nor social cues alone are enough to override the strong expectations children develop about grammatical morphemes. These studies underscore the importance of considering how multiple cues together may add up to more than the sum of their parts when it comes to how and what children learn from other people.

Authors

• Marissa R Sariol-Clough, Appalachian State University
  Presenting Author

• Margaret E Hancock, University of Virginia
  Non-Presenting Author

• Robyn Kondrad, James Madison University
  Non-Presenting Author