3.1. Experiment 2
4.4.2. Children’s selective learning
The present study used two factors to differentiate the speaker’s reliability: correctness and certainty. This section will discuss whether children take account of each factor for selective trust. There were four types of tasks. Thus, I will discuss the children’s selective learning with regard to the nature of the four tasks. First, the findings do not provide evidence that children place emphasis on certainty over uncertainty when they decide upon which is a more reliable source of information. However, the findings do not reflect young Korean children’s failure of an understanding of evidentiality. As stated in Chapter One, children distinguish different degrees of mental-state expressions by the age of four (Moore et al., 1989). However, in the CC vs. CU task in which the speakers expressed differential certainty, the children simply guessed about which speaker was correct. Even 5-6-year olds performed no differently on the task from chance distribution, t(11) = .89, p = .39. It seems unlikely that the older children did not understand the implications of un/certainty by
evidential markers. This is because the findings from two tasks in which the speakers showed conflicting accuracy (i.e., CC vs. IU and IC and CU tasks) revealed that 4-6-year olds might consider the use of uncertainty. Thus, no preference for either speaker in the CC vs. CU task might be rather the case of mentalistic inferences taking account of reasons of uncertainty as stated in the previous section. Hence, although the results did not show the children’s
preference for the confident speaker along with the use of evidentiality, there is no obvious clue to state that they did not differentiate the different levels of certainty when information was correct. However, further research, for example, using mental-verbs such as know and
think might help in clarifying young children’s ability to discriminate certainty.
Secondly, the findings on the incorrect certain vs. incorrect uncertain task, in which two speakers were both incorrect, indicate that children do not learn from either speaker when sources of information are both unreliable. The findings are consistent with those of Jaswal and Neely (2006) that showed no selectivity from unreliable speakers. It seems possible that the participants regarded both speakers as providing false statements to deceive them. Hence, the children considered that the speakers were both unreliable; therefore, they might choose randomly one of the two. Indeed, it seems that preschoolers do not consider others’ false beliefs when they judge trustworthiness. The research that was conducted on children’s understanding of lying and trustworthiness (Wimmer, Gruber, & Perner, 1984, 1985) argues that young children’s judgment on lying is dependent on a speaker’s statement rather than intention. In their studies, children were told stories about a protagonist’s false belief and asked to judge his trustworthiness. In the stories, the protagonist has a false belief about the location of an object as it has been unexpectedly moved, and he conveys
information to a third person about the location of the object which is not correct. The children then were asked to judge whether the protagonist lied or not. The results revealed that children aged four and five tended to regard the protagonist’s false statement as a lie and
this tendency was also found in children between 6- and 12-years-old, although the protagonist’s intention to deliver information correctly was explicitly stated. The findings seem to suggest that children make judgments based on statements when they involve falsity. It seems possible that if the children in the present study regarded the uncertain informant’s statement as a lie, even though the informant was misinformed and did not have any thought of lying, the children might judge that the informant was untrustworthy. Therefore, the children might not excuse the use of the linguistic expressions of un/certainty for judgments in the case of both speakers being unreliable. Taken together, the findings from the IC vs. IU task seem to suggest that children might have difficulty to discriminate differential certainty when there was no accurate information.
Thirdly, the findings of the correct certain vs. incorrect uncertain task have confirmed the previously published research that young children around the age of four are critical of what people say and selectively accept information. The children appreciated differential accuracy, preferring the previously accurate informant. Furthermore, the particular findings of the incorrect certain vs. correct uncertain task suggest that children place emphasis on one’s knowledge status over mental states when they have to consider two different aspects of sources of information such as between correctness and certainty. The results are consistent with prior research (e.g., Corriveau & Harris, 2009a; Jaswal & Neely, 2006), indicating that when two factors are placed together one aspect may override the other; that is, a speaker’s accuracy is a stronger element in deciding trustworthiness over other sources such as certainty, familiarity or authority figure.
It seems necessary to consider why accuracy is a predominant factor for successful judgment or communication. Children may use their knowledge about the world when judging statements (Sobel & Kushnir, 2013). It would be obvious that preschool children have already established certain knowledge about objects based on their experience. There
are two phases in the selective trust paradigm, and children encounter conventional objects in the first phase which are considered to be known by them as young as infancy. Therefore, it is possible that children intuitively favour a person who possesses information in accordance with their basic knowledge, rather than one who holds a different point of view. This
established judgment on the person who shares knowledge with children may continue onto the next step when they endorse labels of novel objects. Indeed, there is evidence suggesting that young children develop sensitivity in favour of the ingroup (Corriveau, Fusaro, & Harris, 2009; Kinzler et al., 2011). They may show an initial preference for a speaker within a group among others or for a person within their own group. In contrast, however, children’s
development in understanding of others’ or their own mental states or beliefs may not firmly develop at an early age (Wellman et al., 2001). Therefore, it is conceivable that preschoolers’ selective trust strongly relies on accuracy over certainty, given children’s early consensus about people according to their own knowledge.
So far, this section discussed implications of the four types of tasks. The findings indicate that the young children might have shown epistemic trust considering the speakers’ certainty, but had difficulty in discriminating when two sources of information were
unreliable. Finally, accuracy trumps inaccuracy and certainty in selective trust. In the next section, I will consider whether the use of evidential markers influenced the young children’s selectivity to information.