Five-year-old children's difficulty with false belief when the sought entity is a person

https://doi.org/10.1016/j.jecp.2004.05.003Get rights and content

Abstract

A total of 153 children (excluding those who erred on control questions), mainly 5 and 7 years of age, participated in two experiments that involved tests of false belief. In the task, the sought entity was first at Location 1 and then, unknown to the searching protagonist, it moved to Location 2. In Experiment 1, performance was well below ceiling in 5-year-olds when the sought entity was a person, and this contrasted with a task in which the sought entity was a physical object. Performance was especially inaccurate when the sought person moved of his or her own volition rather than when the sought person was requested to move by a third party. Interestingly, 5-year-olds were more likely to nominate Location 1 when asked where the searching protagonist would look first than when asked what he or she would do next. In Experiment 2, however, 5-year-olds also tended to nominate Location 1 following a question that included the word “first” even in a test of true belief—a patently incorrect response. Altogether, the results suggest that 5-year-old children have considerable difficulty with a test of false belief when the sought entity is a person acting under his or her own volition. This suggests that 5-year-olds' handle on states of belief is surprisingly fragile in this kind of task.

Introduction

A considerable body of research conducted over the past two decades suggests that children begin to acknowledge false belief in an unexpected transfer task around the time of their fourth birthday (Wellman, Cross, & Watson, 2001). In the task, a sought entity is first in Location 1 and then moves to Location 2 without a searching protagonist's knowledge. Children age 4 years or above typically judge correctly that a searching protagonist will vainly search in Location 1, whereas younger children tend to judge that the searching protagonist will search in Location 2. Effectively, younger children seem to confuse their own knowledge (that the sought entity is in Location 2) with the searching protagonist's knowledge, thereby demonstrably failing to acknowledge false belief. The developmental progression toward making correct judgments has seemed so robust that some (e.g., Gopnik, 1993; Perner, 1991; Wellman et al., 2001) have proposed that children undergo a radical conceptual shift around the time of their fourth birthday. Indeed, Perner and Davies (1991) even suggested that at around 4 years of age, children acquire a theory of mind that is similar in form, if not content, to that of adults.

An alternative view proposes gradual change (e.g., Carpendale & Lewis, in press; Hala & Chandler, 1996; Mitchell, 1996). This is based on two kinds of evidence. One kind demonstrates success in acknowledging false belief before what we might have anticipated in the light of children's poor performance on a traditional test (e.g., Freeman & Lacohee, 1995; Lewis & Osborne, 1990; Mitchell & Lacohee, 1991; Robinson & Mitchell, 1995; Saltmarsh & Mitchell, 1998; Saltmarsh, Mitchell, & Robinson, 1995; Siegal & Beattie, 1991). Another kind of evidence demonstrates lingering difficulties with false belief well beyond 4 years of age. This phenomenon is the focus of the current study.

Most demonstrations of difficulty with false belief beyond 4 years of age use tasks that differ in form from the traditional test. For example, Mitchell, Robinson, Isaacs, and Nye (1996) revealed that even adults sometimes confuse their own knowledge with that of another person (see also Keysar, Lin, & Barr, 2003). In the study, participants were told about a protagonist, Kevin, who saw juice in a jug. He left the room and returned later with Rebecca, who announced that the jug contained milk. Observing participants were asked at this point whether Kevin believed the jug contained juice (as he saw) or milk (as he was told). The large majority judged that Kevin believed the jug contained juice, and perhaps this is not altogether surprising. However, in a comparison condition, a majority judged the opposite, that is, that Kevin believed the jug contained milk. In this condition, participants (but not the protagonist Kevin) were supplied with privileged information indicating that Rebecca poured away the juice and replaced it with milk. In other words, participants' judgments of what Kevin believed were profoundly affected by their own knowledge that the jug contained juice.

A study by Hulme et al.'s, 2003, Robinson's, 2003 also demonstrated difficulty in individuating between beliefs in children around 6 years of age. Children listened to a false belief story in which a thief stole George's watch while he was asleep. When George awoke, he set out to look for the thief, and observing 6-year-old participants were asked to select a picture to show what George was thinking. The children themselves had privileged knowledge that the thief was a man with curly red hair, and they correctly acknowledged that George did not know this because he had been asleep. Nevertheless, they tended to select a picture of the actual thief in preference to a person who looked like a prototypical thief or an outline of a man containing a question mark. As with the adults in Mitchell et al.'s (1996) study, these children seemed to have difficulty in separating their own knowledge from George's knowledge (see also Apperly & Robinson, 1998; Russell, 1987).

Although these studies are noteworthy in demonstrating lingering difficulties in reasoning about beliefs, the form of the tasks differs from that of traditional tests of false belief. Therefore, it might be necessary to deviate from the form of a traditional task to reveal lingering difficulties. However, a study by Symons, McLaughlin, Moore, and Morine (1997) demonstrated difficulty on a test of false belief that had similar form to a traditional test, albeit with different content. In these authors' “animate” task, the sought entity was not an object that transferred from Location 1 to Location 2 but rather a person. Children around 5 years of age frequently nominated Location 2 as if they confused their own knowledge with that of the searching protagonist. This was significantly more common than in a task of similar form but where an inanimate object moved from Location 1 to Location 2. Interestingly, these authors proceeded to demonstrate that it was not animacy per se that counted. In a further condition called the “animate–external” condition, the sought entity was also a person, but this person moved from Location 1 to Location 2 under instruction from a third party. As with an object transfer, it was common for children to nominate Location 1. In other words, a tendency to nominate Location 2 was largely confined to a condition in which the sought person decided to move from Location 1 to Location 2.

Symons et al. (1997) explained the tendency to nominate Location 2 in the animate condition by saying that the volition of the sought entity somehow led children to underperform. Hence, these authors viewed the tendency to nominate Location 2 as an error. If they were right, this would undermine claims made about neat developmental trends in a false belief task of this particular form (e.g., Wellman et al., 2001). Given the apparent robustness of this trend, it is worth considering an alternative possibility. Perhaps nominating Location 2 did not signal failure or that children had made an error but rather that children were engaging in rational and sophisticated reasoning that merely gave a misleading impression of incompetence. Perhaps children credited the searching protagonist with sufficient understanding and familiarity with the sought person to infer that the sought person would move from Location 1 to Location 2. Indeed, Symons and colleagues unwittingly may have included a cue in their false belief story where it mentioned that Mother's search “takes quite a while” (p. 443). Children might have taken this as a clue available to the searching protagonist that Mother was unlikely to still be at Location 1. In that context, it would seem reasonable for the searching protagonist to think that Mother had gone to Location 2. If children did make this kind of attribution of inferential knowledge, it would seem remarkable in view of research suggesting that an ability to make such attributions is difficult to identify prior to 6 or 7 years of age (e.g., Sodian & Wimmer, 1987). In summary, then, either Wellman et al. (2001) overstated the robustness of 5-year-olds' accurate performance on a test of false belief or Sodian and Wimmer (1987) were wrong to deny that 5-year-old children are able to attribute inferential knowledge. There might also be other tenable explanations for distinctive performance on the animate task, but we highlight these two as standing in contradiction to each other.

The starting point for our research was to investigate the replicability of the tendency to nominate Location 2 in the animate test of false belief. We also examined whether the effect is robust with respect to variations in question wording. Symons et al. (1997) asked participants what the searching protagonist would do next. If their performance in the animate task is fragile and unstable, as Symons and colleagues suggested, performance might improve with a question that is known to help children on this form of false belief test; namely, we asked children in another condition where the searching protagonist would look first (Siegal & Beattie, 1991; Surian & Leslie, 1999). Would they be more likely to nominate Location 1 in that condition? If, in contrast, children rationally nominate Location 2 on attributing inferential knowledge to the searching protagonist, there is no reason to suppose that a test question including the word “first” would lead them to nominate Location 1; these children would think that nominating Location 1 is an incorrect response.

Section snippets

Participants

A younger group of 52 children (the “5-year-olds”) were between 60 and 71 months of age (M=66 months), and an older group of 51 children (the “7-year-olds”) were between 84 and 95 months of age (M=89 months). The participants were recruited from five schools and one summer playgroup in the East Midlands, United Kingdom. Two of the schools had a mainly working-class enrollment, whereas the remaining three schools and the playgroup had mainly middle-class enrollments. In terms of gender, 57

Experiment 2

The purpose of Experiment 2 was to seek further evidence relating to Symons et al.'s (1997) explanation for the tendency to nominate Location 2 in animate tasks by investigating the basis of the effect associated with test question wording. We pursued this goal by adding a true belief condition in Experiment 2. In this, the searching protagonist saw Mother/Father at Location 2 through a window. Because the searching protagonist had direct visual access, it would be inappropriate to attribute a

General discussion

In replication of Symons et al. (1997), it seems that children were more likely to nominate Location 2 in an animate test of false belief than in a traditional object-based test of false belief, as demonstrated in Experiment 1. They were also more likely to nominate Location 2 when the sought person in an animate task decided to move from Location 1 to Location 2 than when another person asked the sought person to move. This trend was apparent in both experiments, although it was significant

Cited by (20)

  • The knowledge (“true belief”) error in 4- to 6-year-old children: When are agents aware of what they have in view?

    2023, Cognition
    Citation Excerpt :

    The word “first” in the test question may encourage use of PAR “by implying that the first search would be wrong” (Fabricius et al., 2010, p. 1414). This argument is supported by findings of Rai and Mitchell (2004), who, in a knowledge task in which Lisa stays after the teddy's displacement, found that 5-year-olds were 85% correct when asked a “will do next” test question, but were only 25% correct when given a “look first” test question. However, this objection does not apply to Rakoczy and Oktay-Gür (2020, Experiment 2 and Experiment 4, TB first condition), where a knowledge error was found when children were asked where Lisa believed/thought her teddy was; contrary to what the PAR account predicts.

  • Accurate inferences of others' thoughts depend on where they stand on the empathic trait continuum

    2019, Personality and Individual Differences
    Citation Excerpt :

    Mindreading (known otherwise as mentalizing, empathic accuracy) refers to people's (perceivers') ability to infer what another person (the target) might think, feel, and know for the purpose of interpreting and predicting their behavior (Flavell, Miller, & Miller, 1993; Premack & Woodruff, 1978). Past research on mindreading has explored people's ability to infer others' mental states (e.g., Cassidy, Ropar, Mitchell, & Chapman, 2013, 2015; Ickes, Stinson, Bissonnette, & Garcia, 1990; Pillai et al., 2014; Pillai, Sheppard, & Mitchell, 2012; Sheppard, Pillai, Wong, Ropar, & Mitchell, 2016; Wimmer & Perner, 1983); but much of this research largely ignores the characteristics of the target – the person we are making inferences about (Andrews, 2008; Rai & Mitchell, 2004; Wu, Sheppard, & Mitchell, 2016a, 2016b) – as if we only need to focus on the features of the situation in order to explain mindreading. The empirical work reported here is novel in seeking to explore the possibility that some aspects of target traits might affect how accurately we make mental state inferences.

  • Brain Mechanisms Underlying Reading the Mind from Eyes, Voice, and Actions

    2018, Neuroscience
    Citation Excerpt :

    Traditionally, research on ToM processing has focused exclusively on young children and utilized tasks which appear to be mastered by children between the ages of 4 and 7 years (Apperly, 2012) such as false belief (Wimmer and Perner, 1983; see Wellman et al., 2001 for review), strange stories (Happé, 1994), and social faux pas (Baron-Cohen et al., 1999). Several researchers, however, argue that although children are able to pass classic ToM tasks such as false belief and strange stories at a young age, their performance and nuanced understanding of the mental states and thoughts of others improve throughout childhood (Wellman et al., 2001; Rai and Mitchell, 2004; Calero et al., 2013; Devine and Hughes, 2013) as well as adolescence through young adulthood due, in part, to corresponding increases in cognitive ability (Frischen et al., 2007; Dumontheil et al., 2010; Valle et al., 2015). In addition to the debate on the developmental progression of ToM, ToM has also been interpreted to reflect a wide range of behaviors and has been studied extensively using behavioral and neuroimaging methods.

  • Children's reasoning about other's intentions: False-belief and counterfactual conditional inferences

    2016, Cognitive Development
    Citation Excerpt :

    Perner & Wimmer, 1985). Similarly, 5 year old children make mistakes about false beliefs when the entity that changes location is a person, especially when the change of location is brought about by the person’s own decision rather than by another’s instruction (e.g., Rai & Mitchell, 2004; Symons, McLoughlin, Moore, & Morine, 1997). They also make mistakes about interpretative beliefs, such as one person believing the soup tastes good and another believing that it tastes bad (e.g., Carpendale & Chandler, 1996).

View all citing articles on Scopus
View full text