Hong, Namkyung and Lewis, Charlie (2017) Language-specificity and young preschoolers' social-cognitive development. PhD thesis, Lancaster University.
2017NamkyungHongPhd.pdf - Published Version
Available under License Creative Commons Attribution-NonCommercial-NoDerivs.
Download (2MB)
Abstract
This thesis investigated the role of linguistic access in reference to mental states in children’s social understanding. The importance of access to, or an understanding of, mentalistic language has been stressed regarding the development of children’s social understanding (e.g., Astington & Baird, 2005). It was predicted that the exposure to the mental-state terms using specific grammatically embedded forms specifying certainty and/or the origins of information would enhance Korean children’s social understanding. There has been a vast body of research, showing the predictive role of executive function on the development of social understanding, in particular false-belief understanding (e.g., Carlson & Moses, 2001; Sabbagh, Xu, Carlson, Moses, & Lee, 2006). However, research on Korean children did not support the view on the general development between the two cognitive skills (e.g., Oh & Lewis; 2008). Thus, the current study explored the relationships between executive function and false belief understanding in response to the debate. Executive function, or higher-level self control, is necessary to fulfil goal-directed action inhibiting irrelevant alternatives (Welsh & Pennington, 1988). Children learning from adults, however, trust information selectively (Koenig & Sabbagh, 2013). As children are required to suppress distracting information for selective trust, it was expected that higher skills in executive function may predict performance on selective trust. Thus, the role of executive function on this social understanding was also examined (in Experiment 1 and 2 for false belief and 5 for selective trust). In Experiments 1 and 2 (N = 175) when a protagonist in a false-belief task expressed either his uncertainty (i.e., -keyss (-ul keya) = may) or certainty (i.e., -ci = really), the linguistic markers influenced 3- and 4-year-olds’ apparent grasp of false beliefs. The different levels of certainty (i.e., -hata = do or –ya hata = must do) were applied to the executive function measures. However, the effects of different linguistic markers on executive skills were not observed. Experiment 3 (N = 144) moved the focus from false-belief understanding to selective trust with the application of differential evidentiality in correct and incorrect speakers. Four types of tasks, presented within a 2 (certainty vs. uncertainty) x 2 (accuracy vs. inaccuracy) design, were administered (N = 36 for each task) to three age groups (3.6-4.5 years, 4.6-5.5 years and 5.6-6.5 years). In order to indicate direct access to information, -te (I saw) was used while –napo (It seems) was used for indirect information. The findings from the four tasks showed a crucial effect of accuracy over certainty in selective trust. Following on from the results of Experiment 3, Experiments 4 and 5 compared the children’s performance in epistemic trust experiments in which linguistic access to the protagonists’ mental states was specified using either two evidential markers (i.e., -te vs. – napo) identifying both certainty and the origins of the protagonist’s knowledge, or specific verb terms (i.e., know vs. think) that expressed certainty. In Experiment 4 (N = 59), the findings revealed different developmental patterns according to the use of the two types of linguistic references (evidential markers vs. explicit verb terms): sensitivity to speakers’ epistemic states using mental-verb terms was in evidence at the age four and by evidentiality around the age six. The final experiment of this work employed a battery of executive function measures along with two selective trust tests, using the same contrasting means of identifying the protagonists’ certainty and knowledge (evidential markers vs. different linguistic terms: N = 84). The findings replicated the different developmental patterns of selective trust found in Experiment 4. There were different associations between executive function and questions of two of the three levels of the standard selective trust measure. Verbal working memory predicted the children’s performance in judging who is correct when the test question used included evidential markers. Visual working memory did the same job when verbal mental-state terms were used. Finally inhibitory control predicted selective learning when verbal terms were used. Taken together, the findings suggest that (a) a grasp of certainty appears earlier than an understanding of evidentiality; (b) the grammaticalized forms of certainty and evidentiality are more likely to influence children’s linguistic access to mental states than more explicit mental-verb terms (positively in false belief and negatively in epistemic trust). These lead to the conclusions that: (c) a mastery of semantics and syntactic forms is needed in developing social-cognitive skills; (d) specific language markers identifying the sources of a protagonist’s knowledge may reduce demands of executive function in processing another’s epistemic states.