Interactions between Face Identity and Speech Processing in 5-year-olds and Adults

• In Event: Recent findings on the development of face processing and preferences in early childhood

Abstract

In the first year of life, infants’ face and speech processing abilities improve for own-race faces and native language but decrease for other-race faces and non-native languages. This process is described as perceptual narrowing (e.g. Anzures et al. 2012; Pons et al. 2009). Previous research indicates similar developmental trajectories both in face and speech processing suggesting domain-general processes driving perceptual narrowing (e.g. Maurer und Werker 2014; Krasotkina et al. 2020, 2018). Hereby, perceptual narrowing is discussed as the beginning of a process in which information processing in the face and speech domain gets more and more separated. However, until now, it is unclear whether interactions between face and speech processing in older children and adults still exist or whether both systems are completely separated after the narrowing process. For this reason, we investigated face and speech processing with own-race faces and native speech in 5-year-olds and adults with a computer-based speeded sorting task (Spangler et al. 2010; Schweinberger & Soukup 1998).
Our final sample consisted of 34 adults and 51 five-year-olds. Participants were asked to sort own-race faces, while ignoring native speech contrasts (face task) and vice versa (speech task). Reaction times served as dependent variable. The task compromised a control, a correlational and an orthogonal condition. In the control condition, the irrelevant dimension was constant, in the correlational condition, the irrelevant dimension was consistently paired with the relevant dimension and in the orthogonal condition, both dimensions were paired unsystematically (see Table 1).
For adults, reaction times were similar across both tasks and all three conditions (all Fs < 3, Control: M = 241.67 s, SD= 75.97 s; Correlational: M = 243.65 s, SD = 66.93 s; Orthogonal: M = 251.96 s, SD = 76.17 s). For the 5-year-olds, we found a significant interaction between condition × task, F(2, 48) = 6.74, p = .007, ηp2 = .14. Follow-up analyses for the face task showed no significant main effect for condition, F(2,24) = 0.23, p = .797. However, for the speech task, the post hoc ANOVA showed a significant main effect for condition, F(2, 23) = 9.59, p ≤ .001, ηp2 = .47. Post hoc paired t-tests showed a significant difference in reaction times between the control and correlational condition, t(23) = 2.86, p = .009, Cohen’s d = 0.39, as well as the control and orthogonal condition, t(23) = 2.52, p = .019, Cohen’s d = -0.29 (see Figure 1).
Our results show that adults are able to sort faces independent of speech and vice versa. Hence, our results speak for a separation of face and speech processing in adults. Similarly, the 5-year-olds can process face identities independent of speech. However, sorting speech was dependent on face processing. In the correlated condition, 5-year-olds were faster in sorting native speech contrasts compared to the control condition. Additionally, they were slower in the orthogonal condition compared to the control condition. Thus, speech processing seems to be dependent on face processing indicating that interactions between face and speech processing still exist in 5-year-olds.

Authors

• Özlem Sensoy, Justus-Liebig-Universität Gießen
  Presenting Author

• Anna Krasotkina, Justus-Liebig-Universität Gießen

• Antonia Götz

• Barbara Höhle, University of Potsdam

• Gudrun Schwarzer, University of Giessen