Telethon Kids Institute, The University of Western Australia, Perth, Western Australia, Australia.
Received Date: October 14, 2015; Accepted Date: November 25, 2015; Published Date: December 03, 2015
Citation: Farrant B, Specific Language Impairment and Perspective Taking: Delayed Development of Theory of Mind, Visual and Emotional Perspective Taking. J Child Dev Disord. 2016, 1:8.doi: 10.4172/2472-1786.100008
Context: The development of the ability to take the perspective of others is an important developmental milestone. There is growing evidence that language facilitates the development of children’s explicit perspective taking skills. However, there is a dearth of research investigating the development of perspective taking skills in children with specific language impairment.
Objective: The present study sought to investigate the development of explicit perspective taking skills in children with specific language impairment.
Participants: A group of 30 children with specific language impairment (M age = 63.0 months, SD = 9.0 months) participated in the study along with a group of 30 typically developing children matched for non-verbal ability, gender, and age. Main outcome measures A range of theory of mind, visual perspective taking and emotional perspective taking tasks were used.
Results: The findings of the current study are the first to indicate that the mastery of diverse desires, low verbal false belief, level 1 and ‘level 3’ visual perspective taking, and emotional perspective taking is delayed in children with specific language impairment.
Conclusions: The current results indicate delayed development of a broad range of explicit perspective taking skills in children with specific language impairment. Findings also provide further support for the importance of language acquisition for theory of mind development and add to the evidence that language facilitates the development of visual perspective taking and emotional perspective taking skills.
The ability to take the perspective of others, to understand that other people have feelings, perceptions and thoughts that are separate from and often different from one’s own, derives from an important series of developmental milestones that are associated with a range of prosocial and other behavioural outcomes [1,2]. Theory of mind (ToM) is a type of cognitive perspective taking (PT) that involves the ability to attribute mental states in order to explain and predict behaviour . Children’s ToM development and, more particularly, the understanding of false belief has been the subject of a large amount of research . Tasks that assess children’s understanding of false belief test whether children recognise that people can have mistaken beliefs about reality and whether they can predict what an individual who has a false belief will do, say, or think .
The heavy focus on the development of false belief understanding over the last three and a half decades has meant that the development of other types of PT including visual and emotional PT have received less attention . This is somewhat surprising in light of the fact that a full appreciation of other people’s inner worlds surely requires the ability to infer and understand their emotions (emotional PT) and is in spite of the fact that much of the early research into the development of children’s knowledge about the mind can be traced back to Piaget’s work  including his research into the development of visual PT . Visual PT involves the ability to appreciate the visual experience associated with another viewpoint. In the 1970s a distinction was made between the ability to infer what objects another person does and does not see (level 1 visual PT) and the ability to recognize that an object that is simultaneously visible to both self and other will nonetheless give rise to different visual impressions if their viewing circumstances differ (level 2 visual PT) . Research into children’s ability to take the emotional perspective of others also goes back several decades [10,11]. Emotional PT involves the ability to explicitly infer the feelings and emotions of other people [11,12]. Pleasingly, there is a growing recognition of the need to expand the scope of ToM research to encompass the full spectrum of mental states including emotional PT .
Language and Perspective Taking
The role of language acquisition in the development of children’s explicit false belief understanding has been underscored by the convergent findings from a number of different lines of evidence including research with profoundly deaf children  and typically developing children . The causal nature of this relationship is supported by the findings of longitudinal research [16-18] as well as the results of training studies [19,20]. Research into the relationships between language and the development of visual PT and emotional PT is much more limited. However, there is some research with typically developing children  and profoundly deaf children  that indicates the language may facilitate level 2 visual PT understanding. Similarly, there is some evidence from research with profoundly deaf children  and typically developing children  that suggests a relationship between language and the development of emotional PT skills.
A number of theories have allowed for language acquisition to facilitate the development of children’s PT skills. However, just as with the heavy research focus on ToM/false belief understanding, there are relatively few developmental theories that include a broad range of PT skills including visual and emotional PT. Two theories that do are Baron-Cohen’s modularity theory  and Harris’ simulation theory [24,25].
Baron-Cohen argued that mindreading relies on specific neurocognitive mechanisms: the intentionality detector, eye direction detector, shared attention mechanism, and theory of mind mechanism . The intentionality detector enables behavior to be interpreted in terms of the volitional mental states of desire and goal. The eye direction detector detects the presence of another organism’s eyes, represents what those eyes are looking at, and interprets eye gaze in terms of the perceptual mental state of seeing. The shared attention mechanism (in combination with the intentionality detector and eye direction detector) detects the presence of shared attention by comparing the perceptual states of self and other to determine whether self and other are currently attending to the same thing. Baron- Cohen claimed that the theory of mind mechanism enables the representation of epistemic mental states (e.g., knowing, believing) as meta-representations of the form [Agent-Attitude- “Proposition”]. For example [Mr Jones-thinks-“there are Band- Aids in the Band-Aid box”].
In an update to the theory in 2005 Baron-Cohen changed the focus from mindreading to empathizing and added two more neurocognitive mechanisms to the model, the emotion detector and the empathizing system . The emotion detector detects basic emotions in others and builds dyadic representations of emotional states such as “Mother-is angry-with me”. The shared attention mechanism can receive inputs from the emotion detector and use them to build triadic representations (e.g., “Mother is unhappy that I am unhappy”). The empathizing system provides empathic reactions to other people’s emotional states and a drive to help them. It receives inputs from the shared attention mechanism and produces representations of the form Self-Affective state - [Self-Affective state -proposition]. For example, “I feel sorry that - [Mom feels sad about - the news in the letter]” .
Although Baron-Cohen allowed that the existence of language might benefit the ontological development of PT (mindreading) he did not specify the processes or mechanism by which it could do so . However, because the intentionality detector, eye direction detector, and shared attention mechanism are innate brain modules [23,27] then any effect of language on PT have would have to be via the theory of mind mechanism and/ or the empathizing system. This is because one of the most important properties, if not the essence, of modules is that they are informationally encapsulated [28-30] which means that the processing that occurs within a module (e.g., the eye direction detector) is completely independent of the information or processing in another module (e.g., a language module) or in central systems . Thus, Baron-Cohen’s theory suggests that delayed or disordered language acquisition may be associated with impaired processing of ToM and/or empathy. However, this theory predicts that delayed or disordered language acquisition should have no necessary impact on visual or emotional PT ability because, unlike ToM, visual and emotional PT is processed by an innate module (the eye direction detector and emotion detector respectively). Indeed, Langdon argued that the idea of a domain specific theory of mind mechanism was promoted as an explanation for a series of earlier findings that were interpreted as indicating that individuals with autism have impaired ToM and intact visual PT .
Baron-Cohen’s modularity theory contrasts with Harris’ simulation theory [24,25] regarding the relationships among language, ToM, visual PT, and emotional PT. According to simulation theory, PT involves simulating another’s perspective by imaginatively stepping into their shoes [25,32]. According to Harris’s theory, in the first stage of the development of this ability children can only echo another person’s intentional stance toward a target (object, person, etc.) .The child does this by feeding the other person’s visual target and/or the other person’s emotional stance toward that target into his/her own perceptual and/or emotional system. In the second stage the child uses the processes of step one, but instead of echoing the other person’s intentional stance the child attributes the simulated intentional stance to the other person such that s/he codes the other person as “seeing X” or “liking/ wanting Y” . The third stage involves using the processes of the second stage offline so that the child can imagine or anticipate an intentional stance other than their own current one without the need to monitor a visible target or the other person’s current intentional stance. This allows the child to anticipate or pretend that another person wants an object that the child does not or can see an object that the child cannot. In the fourth and final stage the child can use hypothetical counterfactual situations to drive the processes used in the third stage such that s/he can now imagine someone thinking or seeing something that runs directly counter to what the child currently thinks or sees .
Harris argued that language facilitates PT because conversation involves a constant exchange of differing points of view . Conversation encourages the individual to imaginatively adopt another person’s perspective by continually highlighting the existence of alternative points of view. Ongoing predictions as to what the other person will understand are required to maintain a well-coordinated conversation and repairs and clarifications are needed when these predictions are incorrect. Conversation improves the individual’s ability to accurately imagine what other people perceive, feel, desire, believe, etc. by providing rapid feedback concerning the success of the simulation process. While Harris (1996) initially focused on these pragmatic aspects of language use , more recently he has argued that the child’s wider verbal ability (including syntactic and semantic competence) is a factor that is likely to promote ToM development . Thus, Harris’ version of simulation theory predicts that delayed or disordered language acquisition will be associated with delayed/ impaired ToM, emotional PT and visual PT ability.
Specific Language Impairment and Perspective Taking
Research into the development of PT in children with specific language impairment (SLI) is important to inform clinical and intervention programs and it also offers a useful way to gain insight into the relationship between language and PT . SLI involves delayed or disordered language development in the context of intact nonverbal ability . Accounts of SLI traditionally focused on phonological and syntactic difficulties . A range of subtypes have been proposed including grammatical  and pragmatic  language impairment. However, there is some evidence that individual patterns of impairment change over the course of development  and the debate about SLI subtypes is ongoing.
Although a few studies have found that children with SLI have impaired/delayed ability to recognise emotional facial expressions in terms of identifying what emotion is being expressed [39-41] and one study has found impaired/delayed ability to infer the emotion experienced by a story character  I am unaware of any published research that has assessed the emotional PT abilities of children with SLI using tasks that lead the child and the story character experiencing different emotions. This was the first aim of the current study.
The findings of research using explicit false belief tasks in the 1980s and 1990s were initially interpreted as demonstrating intact ToM in children with SLI [43-45]. However, it has since been pointed out that because this research involved children with SLI whose average chronological age was much older than the age at which explicit false belief understanding typically develops (4-5 years) it left open the possibility of delayed ToM development in children with SLI [46,47]. Indeed, while some more recent ToM research using false belief tasks  has continued to involve children with SLI (often as a control group in studies of autism) who are well over the age at which explicit false belief understanding typically develops a growing number of studies with younger children has found that SLI is associated with delayed development of ToM/false belief understanding [34,47,49-51]. Furthermore, the findings of studies with older children with SLI and age appropriate ToM tasks indicate that this delayed ToM development may extend to 12 years of age and beyond depending on their educational/intervention experiences [52-55]. The second aim of the current study was to further investigate the nature of the delayed ToM development in children with SLI using a battery of ToM tasks.
To my knowledge only one published study has investigated the development of visual PT in children with SLI. This study found that a group of 20 children with SLI aged between 58 and 68 months scored significantly lower on level 2 visual PT tasks than a group of typically developing children who were matched in terms of age, gender and non-verbal IQ . This was the first study to demonstrate delayed visual PT development in children with SLI. However, this study was unable to assess whether children with SLI are delayed on other visual PT tasks that typically developing children master in earlier and later stages of development. The third aim of this study was to fill this gap in our understanding. Answering these questions about the development of explicit PT abilities in children with SLI would provide important information to theorists and practitioners alike.
The overarching aim of the current research was to further investigate the nature of the delayed development of explicit PT skills experienced by children with SLI. This was achieved by comparing a group of children with SLI with a group of typically developing children matched for non-verbal ability, gender, and age on a range of ToM, visual PT and emotional PT tasks. Harris’ simulation theory predicts that the children with SLI will have delayed/impaired emotional PT, ToM, and visual PT ability. In contrast, Baron-Cohen’s theory predicts that although children with SLI may have delayed/impaired ToM their visual and emotional PT abilities should be intact.
The data reported here represent part of a larger research project. All of the Australian children who participated in the present study spoke English as their first language. Thirty children with SLI (26 boys) aged 48-74 months were recruited from a metropolitan Language Development Centre (LDC). In order to receive a place in the LDC, children must have a significant primary language disability in the presence of normal non-verbal intelligence and sound adaptive behaviour skills. A diagnosis of autism was an exclusion criteria for admittance to the LDC. The placement committee (comprised of the LDC Principal, School Psychologist and Speech Pathologist) assess eligibility on the basis of standardized cognitive assessments, speech pathology referral, and information from teachers and carers. In most cases, linguistic ability was assessed with the Clinical Evaluation of Language Fundamentals-Preschool  and non-verbal intelligence via the performance scale of the Wechsler Preschool and Primary Scale of Intelligence – Third Edition . The group of children with SLI were matched in terms of non-verbal ability, gender, and age to a group of 30 typically developing children drawn from a sample of 105 (52 boys) who were recruited from three primary schools spread across working class to upper middle class areas.
Theory of Mind
Children’s ToM ability was measured using two batteries of tasks, each including one of the versions of Wellman and Liu’s  ToM scale used by Farrant et al.  and one of Woolfe, Want and Siegal’s  low verbal false belief tasks (see Appendix A). These tasks range in difficulty from the diverse desires task that assesses whether the individual understands that other people can have desires that differ from his/her own through to the contents false belief task. Previous research has found that typically developing children, with few exceptions, pass the Wellman and Liu tasks in a fixed succession (Peterson, Wellman, & Liu, 2005; Wellman & Liu, 2004). For the present study, the two versions of Wellman and Liu’s (2004) ToM scale were modified to include control questions on the diverse desires and diverse beliefs tasks and an own false belief question in the contents false belief task.
Visual Perspective Taking
Visual PT ability was assessed using two batteries of tasks, each including the level 1, level 2, and array visual PT tasks used by Farrant et al.  and a new ‘level 3’ visual PT task developed specifically for the current research project (see Appendix B). For the present study, all visual PT tasks were modified to include a control question. Earlier research has demonstrated that most typically developing children pass the level 1 visual PT task before they pass the level 2 visual PT task [9,59].
Emotional Perspective Taking
Children’s emotional PT ability was assessed using Harwood and Farrar’s  tasks (see Appendix C). The child’s ability to receptively and expressively identify the relevant emotions (happy and sad) is initially assessed via an emotion labelling task. Then, after identifying his/her best friend, the child is told short stories involving him/herself and the identified friend in situations that would lead to the child and the friend experiencing different emotions (one happy, the other sad) and the child is asked to identify how s/he would feel and how the friend would feel.
The Fluid Reasoning composite score from the Leiter International Performance Scale-Revised  was used as a measure of children’s nonverbal ability since it was not part of the psychometric testing performed by the LDC. The Fluid Reasoning composite score is calculated from age-adjusted scores on the Sequential Order and Repeated Patterns subtests and provides a measure of the child’s nonverbal IQ.
Consent forms and information sheets were sent to parents/ guardians of all kindergarten and pre-primary children attending the three mainstream schools. Children were individually tested in a quiet room at the child’s regular centre. For both groups of children, each of the three testing sessions lasted approximately 15 minutes, and the maximum period between the first and final session for each participant was three weeks. The ToM tasks, emotional PT tasks, and the nonverbal ability test were completed in separate sessions. The visual PT tasks were completed in the same session as the ToM tasks in a counterbalanced order. The sessions were completed in a counterbalanced order and the within-session order of presentation of the individual tasks within each task set was also counterbalanced. Approximately half of the children completed one version of each of the emotional PT, visual PT, and ToM task sets and the other half completed the other versions.
Approval to conduct the research was obtained from the University of Western Australia’s Human Research Ethics Committee prior to the commencement of the study. The study protocol conforms to the World Medical Association Declaration of Helsinki - Ethical Principles for Medical Research Involving Human Subjects. Written prior informed consent was obtained from parents of all children involved in the research.
The group of children with SLI and the matched subgroup of typically developing children had the same gender distribution (26 boys, 4 girls). The age of the children with SLI (M = 63.0 months, SD = 9.0 months) was not significantly different, t(58) = 0.07, p = .94, from that of the matched subgroup of typically developing children (M = 62.8 months, SD = 8.4 months). Similarly, the nonverbal IQ of the children with SLI (M = 105.67, SD = 13.74) was not significantly different, t(58) = 0.12, p = .91, from that of the matched subgroup of typically developing children (M = 105.27, SD = 12.84).
Each ToM, visual PT and emotional PT target question was scored as 1 (passed) or 0 (failed). To be counted as passing each ToM and visual PT target question the child also had to answer the relevant memory and/or control questions correctly. Children had to be scored as correct on both target questions (self and other) to pass the contents false belief task and had to answer all four target questions correctly to pass the emotional PT task. As explained above, there were two task sets for each of the ToM, visual PT and emotional PT tasks. In each case, performance on the two task sets did not differ significantly for the typically developing children (all ts (28) < 0.66, all ps > .52) or for the children with SLI (all ts (28) < 0.29, all ps > .77). Therefore for each of these tasks sets, the results were collapsed across the two versions.
To help ensure that any group differences on the experimental tasks were not the result of the linguistic or working memory demands of the tasks, comparisons were made of the groups’ performance on the control questions. All children in both groups were able to successfully identify the pictures used in the visual PT tasks and they understood the meanings of the terms ‘rightway- up’ and ‘upside-down’ in the level 2 and ‘level 3’ visual PT tasks. Furthermore, the performance of the children with SLI on each of the control questions from the visual PT and ToM tasks were not significantly different from the performance of the typically developing group (Mann-Whitney U tests (58), all ps > .13).
The percentages of children who passed each of the PT tasks are displayed in Table 1. The children with SLI scored significantly worse on the diverse desires, knowledge access, contents false belief and low verbal false belief tasks compared to the matched group of typically developing children (U (58), all ps < .04). However, the difference between the groups on the diverse beliefs task was only marginally significant (U (58), p = .07). The typically developing group performed significantly better than the children with SLI on the level 1, level 2 and level 3 visual PT tasks (U (58), all ps < .03). However, the difference between the typically developing children and the children with SLI on the array visual PT task was not significant (U (58), p = .15). This likely reflects the fact that the difficulty of the array visual PT task resulted in a floor effect where all but a few of the more advanced typically developing children failed the task (Table 1). As predicted, the typically developing children also performed significantly better than the children with SLI on the emotional PT tasks (U (58), p = .04).
|SLI (n = 30)||Matched Typically Developing (n = 30)|
|Task||% Pass||% Pass|
|Contents False Belief||13||37|
|Low Verbal False Belief||20||47|
|Level 1 visual PT||70||93|
|Level 2 visual PT||17||47|
|Level 3 visual PT||13||40|
|Array visual PT||0||7|
Table 1: Percentage of Children Passing the Tasks.
The current study sought to further investigate the delayed development of explicit PT skills experienced by children with SLI. The performance of a group of children with SLI on a range of ToM, visual PT and emotional PT tasks was compared with the performance of a group of typically developing children matched for non-verbal ability, gender, and age. Harris’ simulation theory predicted that the children with SLI will have delayed/ impaired ToM, emotional PT and visual PT ability. In contrast, Baron-Cohen’s theory predicted that although children with SLI may have delayed/impaired ToM their visual and emotional PT abilities should be intact.
Delayed Development of PT Skills in Children with SLI
Consistent with the predictions from Harris’ simulation theory [24,25], the matched group of typically developing children performed significantly better than the children with SLI on the diverse desires, knowledge access, contents false belief, low verbal false belief, level 1, level 2 and level 3 visual PT, and emotional PT tasks. Indeed, the findings of the current study are the first to indicate that the mastery of emotional PT, level 1 and ‘level 3’ visual PT, diverse desires, and low verbal false belief is delayed in children with SLI. The present results are consistent with previous research demonstrating delayed development of knowledge access and false belief understanding in children with SLI [34,46,47,49-51,61] as well as research supporting a relationship between language and explicit false belief understanding in profoundly deaf children  and typically developing children . Furthermore, the current findings indicate that the previously reported finding of atypical development of false belief understanding in the current sample of children with SLI  is also evident on a low verbal false belief task and that the delayed ToM development experienced by children with SLI extends to the understanding of diverse desires as well as knowledge access. The difference between the performances of the groups on the diverse beliefs task was also in the expected direction but was only marginally significant.
Consistent with previous research demonstrating slower development of level 2 visual PT understanding in children with SLI  and other research suggesting that language facilitates typically developing and profoundly deaf children’s level 2 visual PT performance [14,21], the present results indicate that the mastery of level 1, level 2, and ‘level 3’ visual PT are delayed in children with SLI. Thus, the current findings suggest that children with SLI experience delays in the development of a range of visual PT skills. Similarly, although the findings of the current study are the first to indicate that children with SLI experience delayed development of emotional PT this finding is consistent with the results of research with profoundly deaf children  and typically developing children  that suggest a relationship between language and the development of emotional PT skills. Thus, the current findings provide further support for a relationship between language and the development of explicit PT skills and indicate that children with SLI experience delays in development across a range of different kinds of PT skills.
Thus, the findings of the current study significantly expand the growing number of findings that are contrary to the prediction drawn from Baron-Cohen’s modularity theory [23,26] that although children with SLI may have delayed/impaired ToM their visual and emotional PT abilities should be intact. Baron- Cohen could argue that his theory is neutral with regard to the effect of delayed/disordered language acquisition on visual PT ability because he does not consider the intentionality detector, eye direction detector, emotion detector, and shared attention mechanism to be classical Fodorian modules  in the sense of being informationally encapsulated. However, as pointed out by Gopnik , softened versions of modularity without encapsulation rob the notion of modularity of any epistemological interest. This is because, “if a ‘module’ just means a functional unit in a cognitive system then all knowledge is modular, and modularity claims are uninteresting” . Regardless of whether or not Baron-Cohen’s theory [23,26] is truly modular in the Fodorian sense, the theory would require revision in order to explain how delayed or disordered language acquisition could have any effect on visual and emotional PT.
Methodological Strengths, Limitations, and Directions for Future Research
The major strength of the current research was the inclusion of an array of measures of emotional PT, visual PT, and ToM in a design including both typically developing children and children with SLI. The sample sizes used in the current study are also fairly large for research of this type. A further strength was the use of closely matched groups of typically developing children and children with SLI. Consistent with Mervis and Robinson’s  argument that p levels should be at least .50 to avoid incorrectly accepting the null hypothesis and thereby ensure that groups are matched on control variables, the groups used in the current research project had the same gender distribution and were well matched in terms of chronological age and nonverbal ability (both p levels greater than .90).
Nevertheless, the inclusion of a broad range of emotional PT, ToM, and visual PT tasks and time limitations meant that each of the sub-concepts (e.g., knowledge access, level 2 visual PT) was assessed via a small number of observations. It could be argued that this increases the likelihood that children’s responses are influenced by chance. However, the fact that the patterns of task mastery observed in the current research (proportion of children passing each task) are largely consistent with previous research [4,9,59,64] makes this explanation less plausible. Similarly, it would be difficult to argue that the impaired performance of the children with SLI on the PT tasks is due to the linguistic demands of the tasks. For a start, all of the children with SLI were able to successfully identify the pictures used in the visual PT tasks and they all understood the meanings of the terms ‘right-way-up’ and ‘upside-down’ in the level 2 and ‘level 3’ visual PT tasks. Secondly, there were no significant differences between the performance of the children with SLI and the performance of the group of typically developing children (matched for non-verbal ability, gender, and age) on each of the control questions from the visual PT and ToM tasks. The fact that the children with SLI had impaired performance on the low verbal false belief task in the present study also indicates that PT development is genuinely delayed in children with SLI.
However, our understanding of the relationships among language and the development of PT skills observed in the present study would benefit from future research that included a larger number of observations for each task type. Research of this type that included larger samples of children would also help clarify whether children with SLI experience delayed understanding of diverse beliefs. Future research should also investigate whether children with SLI experience delays in the mastery of other PT tasks. One promising example is the colour filter visual PT task which recent research has found is passed by the majority of 36 month old typically developing children . Research with older children could also investigate whether children with SLI have impaired performance on the array visual PT task. Another promising avenue for future research would be to explore the relations among implicit false belief understanding, language and explicit implicit false belief understanding children with SLI.
The current study found that, compared with the performance of typically developing children matched for non-verbal ability, gender, and age, children with SLI performed significantly less well on a range of ToM, visual PT and emotional PT tasks. Thus, the present results indicate that children with SLI experience delays in the development of a range of different kinds of explicit PT skills. These findings add further support to the importance of language for children’s ToM development [see 15, for a metaanalysis] and add to the evidence that language facilitates the development of visual PT [14,21,47] and emotional PT [14,22].
The author declares that no conflicts of interest exist.
All Published work is licensed under a Creative Commons Attribution 4.0 International License
Copyright © 2019 All rights reserved. iMedPub LTD Last revised : March 21, 2019