written assignment of an article

lable at ScienceDirect

International Journal of Pediatric Otorhinolaryngology 96 (2017) 77e83

Contents lists avai

International Journal of Pediatric Otorhinolaryngology

journal homepage: http://www.ijporlonline.com/

Theory of mind and language development in Japanese children with hearing loss

Hiroshi Fujino a, *, Kunihiro Fukushima b, c, Akie Fujiyoshi b

a Department of Special Needs Education, Tokyo Gakugei University, Tokyo, Japan b Shinkurashiki Ear, Nose, and Throat Clinic, Japan c Department of Otolaryngology, Fukuoka University, Japan

a r t i c l e i n f o

Article history: Received 1 September 2016 Received in revised form 27 February 2017 Accepted 1 March 2017 Available online 6 March 2017

Keywords: Theory of mind Language development Children with hearing loss

* Corresponding author. Department of Special Nee University, 4-1-1 Nukuikitamachi, Koganei-shi, Tokyo

E-mail address: [email protected] (H. Fujino

http://dx.doi.org/10.1016/j.ijporl.2017.03.005 0165-5876/© 2017 Elsevier B.V. All rights reserved.

a b s t r a c t

Objective: This study investigates the development of theory of mind (ToM) in Japanese children with hearing loss (HL) and its relationship with language abilities using the data of a large sample size. Methods: Participants were 369 children with HL, ranging from 4 to 12 years of age. The mean hearing level of the better ear was 100.7 dB. A “change in location”-type false belief task similar to the “Sally- Anne test” was given to the participants. Results: The pass rates for the false belief task were in the 20% range for 4 to 6-year-olds, 35.6% for 7- year-olds, 47.6% for 8-year-olds, and 63.6% for 9-year-olds. However, no children, even 12-year-olds, achieved a pass rate of 70%. A logistic regression analysis showed that the significant independent predictors of the false belief task performance were vocabulary age and syntactic comprehension level, and chronological age, hearing level, syntactic production level, and nonverbal intelligence were excluded. Conclusion: The results demonstrate that there is a delay in the development of ToM in Japanese children with HL. This finding is consistent with findings in English-speaking countries. Additionally, it is sug- gested that language abilities play an important role in the acquisition of ToM for children with HL.

© 2017 Elsevier B.V. All rights reserved.

1. Introduction

“Theory of mind” (ToM) is a psychological concept that relates to the cognition of mental states, including desires, beliefs, intentions, and feelings. Understanding and predicting another person's behavior by inferring their mental states is important for reciprocal communication. Although there are various aspects of ToM, false belief tasks have been used in many studies on ToM.

Wimmer and Perner developed a test called the “false belief task” to assess ToM [1]. This task consists of the following scenario. Person A puts an object into Location X. In Person A's absence, Person B transfers the object from Location X to Location Y. The subjects are asked to indicate where Person A will look for the object upon their return. Wimmer and Perner administered this “change in location”-type false belief task to typically developing children* and demonstrated that they acquire the ability to pass the

ds Education, Tokyo Gakugei , Japan. ).

task at around 4 years of age. Children with autism spectrum disorder (ASD) are known to

have difficulty acquiring ToM. Baron-Cohen et al. administered a “change in location”-type false belief task to ASD children and showed they had a significantly lower pass rate compared with typically developing children, even when their intellectual devel- opment was equivalent or higher to that of their typically devel- oping peers. Based on these findings, they proposed the ToM deficit hypothesis of ASD [2].

The development of ToM has been studied also in children with hearing loss (HL). Peterson and Siegal administered a “change in location”-type false belief task to 26 Australian children with severe and profound HL, aged between 8 and 13 years (mean age: 10 years), and they reported a pass rate of only 17% [3]. In England, Steeds et al. reported a 70% pass rate for a “change in location”-type false belief task for 22 children with profound HL, aged between 5 and 12 years (mean age: 9 years) [4], which was better than that reported by Peterson and Siegal but lower than that expected for their chronological ages.

Russell et al. also investigated 32 children with profound and severe HL in England and reported pass rates of 17% for a “change in

Table 1 Number of participants in each age group.

Age (years) Range (months) Male Female Total

4 52e59 13 10 23 5 60e71 31 33 64 6 72e83 21 27 48 7 84e95 33 26 59 8 96e107 21 21 42 9 108e119 15 29 44 10 120e136 17 16 33 11 132e143 23 14 37 12 144e153 8 11 19 Total 182 187 369

H. Fujino et al. / International Journal of Pediatric Otorhinolaryngology 96 (2017) 77e8378

location”-type false belief task for those aged 4e7 years (mean age: 6 years), 10% for those aged 8e12 years (mean age: 10 years), and 60% for those aged 13e16 years (mean age: 15 years) [5]. They observed no age-related performance improvement until the age of 12 years, although there was marked improvement after the age of 13 years. They suggested that the biggest improvement in deaf children's performance on the false belief task occurred after the age of about 13 years, and there was little age-related change in false belief task performance from the age of 4e12 years. In the United States, Lundy used “change in location”- and “unexpected contents”-type false belief tasks, and two other ToM tasks, including “appearance-reality” and “misleading picture,” and re- ported that 14 children with moderate to profound HL do not ac- quire ToM ability until the age of 7 years [6].

As described above, ToM in children with HL has been mainly studied in English-speaking countries. In Japan, Ohara and Hirota (2014) investigated 12 children with moderate to profound HL, aged 4e7 years (mean age: 5 years), using a “change in location”- type false belief task and “explanation of action task” and reported a false belief task pass rate of 33%, which was significantly lower than that expected according to chronological age [7].

ToM in children with HL has been also studied from the perspective of language development. Jackson reported on the relationship between ToM and receptive language abilities in chil- dren with severe to profound HL, and pointed out that a relation- ship was not found in native signers when age was partialed out [8]. They used “change in location”- and “unexpected contents”-type false belief tasks and a false photograph test. Schick et al. investi- gated 176 children with HL, aged 4e7 years (mean age: 6 years), and showed that the significant predictors of passing the “change in location”-type false belief task were receptive vocabulary and syntactic complement processing ability [9].

To summarize the findings of these studies, we can say that ToM in children with HL is related to language abilities rather than chronological age. There are three hypotheses concerning how language contributes to the solving of false belief tasks. First, chil- dren with HL do not have sufficient conversational experience that refers to mental states [3]. Second, they have delays in acquisition of mental state vocabulary and syntactic processing of complex sentences [9]. Third, they do not have enough language abilities to understand task contents and instructions [10].

Furthermore, the problem of cultural differences has been pointed out regarding the development of ToM. It was shown that the development of false belief understanding in Japanese children was delayed compared to that of children in English-speaking countries [11,12]. Is there such a developmental delay also in chil- dren with HL?

There are very few studies on ToM in Japanese children with HL. Even from an international perspective, there is little research on the development of ToM and its relationship with language abilities in children with HL using a large sample size involving a wide range of ages from infancy to school age. Based on this background, the purpose of the present study is to investigate the development of ToM in Japanese children with HL and its relationship with lan- guage abilities as compared with the findings of studies in English- speaking countries.

2. Methods

2.1. Participants

The subjects for this study were selected from among the par- ticipants in the Ministry of Health, Labour and Welfare's Research on Sensory and Communicative Disorders Project (team leader: Kunihiro Fukushima) [13]. The study design was approved by the

ethical review board of the Association for Technical Aids. The participants in this study were 369 children, who were 4- to

12-year olds. The number and gender of the participants in each age group are shown in Table 1. The mean hearing level of the better ear was 100.7 dB (range: 70.0e135.0 dB,SD: 13.0). The scores of Raven's Colored Progressive Matrices (RCPM) as a measure of non-verbal intelligence and the use of sign language are shown in Table 2. A comparison with the Japanese children's standard [14] showed that the participants in the study did not have any intellectual disabilities.

The participants were recruited according to the following in- clusion criteria: 1) age from 48 months to 155 months, 2) congenital hearing impairment with a hearing level of greater than 70 dB (on average) appearing no later than 4 years of age. Children were recruited from schools for the deaf, schools for the hard of hearing, mainstream schools, day-care centers, and hospital training rooms.

2.2. Materials

2.2.1. Theory of mind task An animated ToM test was given to the participants. This is a

“change in location”-type false belief task such as the “Sally-Anne test” described in detail elsewhere [15]. The task consists of the following scenario: “Person A put the ball into the box and went out of the room. Then, Person B took ball from the box and put it into the bag. Then, Person A returned to the room.” Following the last scene, three questions were presented: “Where is the ball now?” (reality question), “Where was the ball at first?” (memory ques- tion), and “Where will Person A look for the ball?” (belief question). The reality and memory questions are control questions that assess the child's ability to understand the situation. The task was judged to be passed only if all three questions, including the two control questions, were answered correctly. The animation, including captions and questions, was presented visually and auditorily via a PC monitor and a speaker.

2.2.2. Language and other measures Receptive vocabulary was measured using the Picture Vocabu-

lary Test-Revised (PVT-R). Syntactic comprehension and produc- tion were measured using the Syntactic Processing Test of Aphasia (STA). STA is a test employed to evaluate comprehension and pro- duction of syntactic structures [16]. Each level consists of eight and three sentences to test comprehension and production, respec- tively. The children are asked to choose one of four or six pictures that are appropriate for the tester's presentation in the compre- hension test. When a child correctly comprehends seven of eight sentences of each level, the child is assumed to have passed that level. The children are asked to express a sentence corresponding to the picture that a tester indicates in the production test. When a child produces two of three sentences of each level, the child is

Table 2 Nonverbal intelligence and use of sign language.

Age group RCPM scores Use of sign

Mean (SD) Yes No

4 not available 11 12 5 not available 40 24 6 not available 28 20 7 25.69 (5.15) 27 32 8 27.88 (4.59) 32 10 9 28.86 (5.76) 21 23 10 31.42 (3.03) 17 16 11 31.59 (3.94) 22 15 12 32.26 (2.60) 10 9 Total 208 161

H. Fujino et al. / International Journal of Pediatric Otorhinolaryngology 96 (2017) 77e83 79

assumed to have passed that level. Level I is based on a semantic strategy (actioneobject or actioneobjecteverb structure), and one can understand such sentences by using only one cue; either the agent or the verb. Level II is based on a basic word order strategy in which the agent appears at the beginning of a sentence. Levels III and IV are based on a particle strategy. Level III consists of sentences without a complement clause. Level IV consists of sentences in the passive voice with a complement clause for the comprehension test, but without a complement clause for the production test. Level V consists of sentences with auxiliary verbs. This level is only for the production test. Composition and contents of STA are shown in Table 3. Fujiyoshi et al. [17] reported syntactic abilities of Japanese children with HL using STA.

3. Results

3.1. False belief task pass rates for each age group

The pass rates for the false belief task are shown in Fig. 1. The pass rates were in the 20% range for 4 to 6-year-olds, 35.6% for 7- year-olds, 47.6% for 8-year-olds, and 63.6% for 9-year-olds. How- ever, no children, even 12-year-olds, achieved a pass rate of 70%.

The correct response rates for the control questions are shown in Fig. 2. The correct response rate for the reality question was 90% or more for the 4-year-old children. The correct response rates for the memory question were 50% or more and approximately 80% for the 4-year and 6 year-olds, respectively.

3.2. Predictors of false belief task success

Logistic regression analysis was performed for the children with HL, using false belief task pass/fail as the dependent variable and chronological age, hearing level, RCPM score, PVT-R vocabulary age, and STA comprehension and production level as the independent variables. The method of analysis was forward stepwise (likelihood

Table 3 Composition and contents of STA.

Level Examples of test sentences

Comprehension I Otoko-no-ko ga aruite iru. (A boy walks.) II Ok�asan ga otoko-no-ko wo osite iru. (The mother pushes a boy) III Ot�osan ga on'na-no-ko ni hankachi wo karite iru. (The father bor IV Otoko-no-ko ga ok�asan ni boushi wo torarete iru. (A cap that bel Production I Kodomo ga hashiru. (A child runs.) II Kodomo ga ok�asan ni hana wo ageru. (A child gives a flower to h III Ok�asan ga kodomo kara hana wo morau. (The mother gets a flow IV Kodomo ga ok�asan ni k�eki wo torareru. (A child is taken cake by V Kodomo ga ok�asan ni kawa kara b�oru wo hirotte morau. (A child

ratio). The results are shown in Table 4. The significant predictors of false belief task success were vocabulary age and STA compre- hension level. Chronological age, hearing level, RCPM score, and STA production level were excluded.

3.3. Vocabulary ages and false belief task pass rates

The pass rates for the false belief task in each age group are shown in Fig. 3. The pass rates were under 40% in children with a vocabulary age of 5 years. It exceeded 60% in children with a vo- cabulary age of 6 years and 80% in those with a vocabulary age of 11 years. There were significant differences between vocabulary age of less than 3 years and 3 years (c2 (1) ¼ 5.34, p < 0.05), and 5 years and 6 years (c2 (1) ¼ 4.77, p < 0.05).

3.4. Syntactic comprehension levels and false belief task pass rates

The false belief task pass rates for each syntactic comprehension level are shown in Fig. 4. The pass rate did not reach 50% in children below Level II, but it exceeded 60% and 90% in children at Levels III and IV, respectively. There were significant differences between adjacent levels (below Level I and Level I: c2 (1) ¼ 9.32, p < 0.01; Level I and Level II: c2 (1) ¼ 7.00, p < 0.01; Level II and Level III: c2 (1) ¼ 5.70, p < 0.05; Level III and Level IV: c2 (1) ¼ 8.00, p < 0.01).

4. Discussion

In the present study, the false belief task pass rate for 4-year-old children with HL was 21.7%, which was markedly lower than that for 4-year-old typically developing (normally hearing) children (78%) reported by Wimmer and Perner [1]. In addition, the pass rate for 7-year-old children was 35.6%, which was remarkably lower than that for typically developing Japanese children, at 90% [15]. However, the correct response rate for the reality question was more than 90% for 4-year-old children, suggesting that task failure is not dependent on misunderstanding of the event presented to children.

These results demonstrate that there is a developmental delay in false belief understanding in Japanese children with HL as compared to typically developing Japanese children, which was consistent with the findings in English-speaking countries. How- ever, considered in greater detail, the pass rate for 10-year-old children, at 69.7%, was markedly higher than that for the same age group reported by Peterson and Siegel [3] and Russell et al. [5] (17% and 10%, respectively). In contrast, the pass rate for 9-year-old children, at 63.6%, was slightly lower than that for the same age group (70%) reported by Steeds et al. [4]. Therefore, there are in- consistencies in the results regarding the relationship between chronological age and the false belief task pass rate. It is suggested that a critical factor related to the passing of the false belief task

Age equivalent

3 years old 3e4 years old

rows a handkercheif from a girl.) 5e6 years old ongs to the boy is taken away by his mother) 7 years old

3e4 years is mother.) 4e5 years er from a child.) 5e6 years his mother.) 6 years gets a ball picked up from a river by his mother.) 7 years

Fig. 1. Pass rates for the false belief task in each age group.

Fig. 2. Correct response rates for the control questions.

Table 4 Summary of logistic regression analysis.

B Sig. (p) Odds ratio 95% C.I. for Odds ratio

Lower Upper

Vocabulary Age 0.022 0.000 1.022 1.012 1.033 STA Comprehension Level 0.706 0.000 2.025 1.487 2.759 Constant �3.133 0.000

Model Chi-square Test:x2 (2) ¼ 110.502,p < 0.001. Percentage correct:71.8%.

H. Fujino et al. / International Journal of Pediatric Otorhinolaryngology 96 (2017) 77e8380

might not be chronological age. The logistic regression analysis showed that the significant in-

dependent predictors of false belief task performance were vo- cabulary age in PVT-R and comprehension level in STA but not chronological age, hearing level, and nonverbal intelligence. These results suggest that the relationship between false belief task

performance and chronological age are not directly related, which supports the findings of Russell et al. [5]. However, it is suggested that there is a significant relationship between language abilities and false belief understanding. These results were consistent with previous studies that reported that false belief understanding was correlated with receptive language ability [8], and with receptive vocabulary and syntactic complement processing ability [9].

How does language ability contribute to false belief under- standing? The hypothesis of language skills necessary to under- stand the contents of and the instructions for the test is not supported. Previous studies reported that typically developing children between the ages of 4 and 5 years pass the false belief task; therefore, children with HL who have language abilities at that level are expected to be able to pass the task. The children in our study could not pass the belief question but could pass the control questions.

Figueras-Costa and Harris gave a false belief task that did not require verbal instructions to children with HL. Although they

Fig. 3. Vocabulary ages and pass rates for the false belief task.

Fig. 4. Syntactic comprehension levels and pass rates for the false belief task.

H. Fujino et al. / International Journal of Pediatric Otorhinolaryngology 96 (2017) 77e83 81

found improved performance with this nonverbal approach, the pass rate was still lower than that of the children's normally hearing peers, suggesting that the task failure was not attributable to the factor of understanding verbal instructions but to problems with understanding mental representations [10]. These results suggest that children with HL fail the false belief task due to diffi- culty in understanding ToM rather than in comprehending test instructions.

The pass rate of the false belief task exceeded 60% for those with a receptive vocabulary age of 6 years. There was a significant dif- ference between the pass rate in children with a vocabulary age of 5 years and 6 years. The results suggest that a receptive vocabulary age of 6 years is a critical point in the development of false belief understanding in children with HL. These results were approxi- mately consistent with Ohara and Hirota's study in Japanese

children with HL, in which the median language developmental age of children with HL who passed the false belief task was reported to be 5 years and 10 months [7].

Many studies reported that the vocabulary age of 9 years is a critical point for false belief task success for children with ASD. In Japan, Fujino et al. showed that the false belief task pass rate for children with ASD was 0% in those with a vocabulary age of 6 years or below and that it exceeded 50% in those of 8 years using the same task as in the present study [18]. This result is almost consistent with studies in English-speaking countries [19e21]. These findings suggest that, to understand false belief, children with HL need language abilities that are higher than in typically developing children but lower than in children with ASD.

Why does vocabulary facilitate false belief understanding? More than 60% and nearly 80% of the children with HL with a vocabulary

H. Fujino et al. / International Journal of Pediatric Otorhinolaryngology 96 (2017) 77e8382

age of 6 and 9 years, respectively, passed the false belief task. These ages overlapped with the age at which children gain the ability to understand abstract words. While concrete words are learned by associating a word with an object, abstract words are learned through metalanguage manipulation, in which the meaning of a word is explained using other words. Such metalanguage manip- ulation may require the same mental process used for meta- representation, and it is possible that metalanguage manipulation can be the basis for such metarepresentation ability. It is considered that ToM is based on the metarepresentation ability [1].

Syntactic ability is also related to false belief understanding. The logistic regression analysis showed that syntactic comprehension was the predictor of false belief task success. The pass rate in STA comprehension did not reach 50% until Level II, and it exceeded 60% and 90% at Levels III and IV, respectively. There were significant differences between each adjacent level. There is a tendency for the pass rates to increase with the improvement in the STA level, and almost all children with HL at Level IV passed the false belief task. The sentences in the Level IV task consisted of passive sentences with complement clauses, e.g., “A cap that belongs to the boy is taken away by his mother.” This result suggests that syntactic ability, especially comprehension of sentences with complement clauses, is an important factor for false belief understanding in children with HL.

The relationship between false belief understanding and syn- tactic ability has also been reported in studies of children with ASD. Fisher et al. showed that false belief task performance was strongly correlated with receptive vocabulary and syntactic comprehension in children with ASD [22]. Lind and Bowler examined the effects of memory for the complement clause on false belief task perfor- mance, and they indicated a strong correlation between the two and a moderate correlation between task performance and recep- tive vocabulary [23].

How does syntactic ability relate to false belief understanding? The hypothesis of De Villiers et al. is as follows [24]. The false belief task has a structure such as “Person A believes that object X is in location Y,” which corresponds to the sentence with a complement clause. The complement clause can be either true or false. Based on this hypothesis, they considered that the ability to understand a sentence with a complement clause is related to the ability to pass the false belief task. This relationship can be explained through the following example. The complement clause “it will rain” in the sentence “Person A said that it will rain,” can be either true or false. If it does not rain, one realizes the proposition “it will rain” was false, and one also realizes Person A had a false belief.

In line with this hypothesis, the inferential process in the false belief task can be understood, in that Person A does not know the object had been transferred to Place Y (the true proposition); therefore, Person A considers that the object was in Place X (the false proposition). Typically developing children can be expected to intuitively understand a mental representation without language mediation [25], and even if they cannot do this, they can complete the task using the complement clause as a symbolic representation of mental states. In addition, the predictors of false belief task success were not the expressive aspects of language abilities as measured by STA production but the receptive aspect of language abilities as measured by STA comprehension. This result can be explained from the viewpoint of the STA's test composition. The level IV task of the comprehension test requires understanding of sentences with a complement clause. The processing of a comple- ment structure is a critical factor for a comprehension test, but not for a production test. It is considered that such a difference between both tasks was reflected in the results. Furthermore, the result supports the findings on the importance of complement processing ability for false belief understanding [9,24].

Hearing loss is a sensory disorder and not a neuro- developmental disorder like ASD. Then why does a developmental delay in ToM occur even in children with HL? Moreover, what is the difference between HL and ASD? Insufficient conversational expe- rience is considered to be a common factor between them. In addition, there is a central social cognitive dysfunction in children with ASD. Therefore, children with ASD need language mediation to compensate for weak social cognition. That is to say, children with ASD have problems related to social brain dysfunction and insuf- ficient communication experience, whereas children with HL only have insufficient communication experience. Therefore, children with ASD might show more severe problems related to ToM than would children with HL.

According to Peterson and Siegal, ToM is acquired through conversational experience with familiar people in the early stages of development [3]. The important point is whether topics con- cerning mental representations, such as the thoughts and feelings of another person, are involved in conversation. When children do not have such experience, they are likely to have difficulty in acquiring ToM. This explanation seems especially applicable to cases of children with HL with normally hearing mothers. Some studies have pointed out that normally hearing parents tend to talk to their children with HL only about things that are visible during the language acquisition period [26,27]. Thus, the lack of conver- sation experience concerning topics including mental representa- tions leads to difficulty in acquiring ToM for children with HL.

However, native signers are reported not to have any problems with ToM [3,9]. In such cases, most of these parents also have hearing impairment and use sign language for daily communica- tion, through which the children acquire sign language as their primary language. When both children and parents are sign lan- guage users, their topics of conversation involve invisible mental representations, as is the case with normally hearing children and mothers. This may be why the children acquire ToM. We therefore consider that ToM development is determined by the quantity and quality of conversation, irrespective of whether the communication mode is spoken or signed language.

In the present study, we were not able to identify native sign language users among the participants, which is a limitation. Further studies regarding the relationship between quality of communication from the early stages of development, including communication with sign language, and development of ToM are required, especially in Japan.

Therefore, the hypothesis that conversational experience is the most important factor for the development of theory of mind in children with HL cannot be fully verified from the results of this study. However, it is suggested that enhancing language ability including vocabulary and syntax promotes the development of theory of mind. That is the clinical implication of this study.

4.1. Authors' note

Normally developing children are described as “typically developing children,” and children without hearing loss are described as “normally hearing children.” In the present manu- script, these two terms were used as contextually appropriate, but both refer to children without disabilities.

Conflicts of interest

None.

Acknowledgment

This study was supported by a Health and Labour Science

H. Fujino et al. / International Journal of Pediatric Otorhinolaryngology 96 (2017) 77e83 83

Research Grant (as part of the Research on Sensory and Commu- nicative Disorders Project) from the Japanese Ministry of Health, Labour, and Welfare. Grant Number H19-018.

References

[1] H. Wimmer, J. Perner, Beliefs about beliefs: representation and constraining function of wrong beliefs in young children's understanding of deception, Cognition 13 (1983) 103e128.

[2] S. Baron-Cohen, A.M. Leslie, U. Frith, Does the autistic child have a 'theory of mind'? Cognition 21 (1985) 37e46.

[3] C.C. Peterson, M. Siegal, Deafness, conversation and theory of mind, J. Child Psychol. Psychiatry 36 (1995) 459e474.

[4] L. Steeds, K. Rowe, A. Dowker, Deaf children's understanding of beliefs and desires, J. Deaf Stud. Deaf Educ. 2 (1997) 185e195.

[5] P.A. Russell, C.D. Hosie, G.C. Scott, N. Hunter, J.S. Banks, M.C. Macaulay, The development of theory of mind in deaf children, J. Child Psychol. Psychiatry 39 (1998) 903e910.

[6] J.E.B. Lundy, Age and language skills of deaf children in relation to theory of mind, J. Deaf Stud. Deaf Educ. 7 (2002) 41e56.

[7] S. Oohara, E. Hirota, Analysis of factors related to meta-representational development in hearing-impaired children, Jpn. J. Logopedics Phoniatrics 55 (2014) 17e25.

[8] A.L. Jackson, Language facility and theory of mind development in deaf chil- dren, J. Deaf Stud. Deaf Educ. 6 (2001) 161e176.

[9] B. Schick, P. de Villiers, J. de Villiers, R. Hoffmeister, Language and theory of mind: a study of deaf children, Child. Dev. 78 (2007) 376e396.

[10] B. Figueras-Costa, P. Harris, Theory of mind development in deaf children: a nonverbal test of false-belief understanding, J. Deaf Stud. Deaf Educ. 6 (2001) 92e102.

[11] H.M. Wellman, D. Cross, J. Watson, Meta-analysis of theory-of-mind devel- opment: the truth about false belief, Child. Dev. 72 (2001) 655e684.

[12] M. Naito, K. Koyama, The development of false belief understanding in Japa- nese children: delay and difference? Int. J. Behav. Dev. 30 (2006) 290e304.

[13] K. Fukushima, N. Kasai, K. Omori, A. Sugaya, A. Fujiyoshi, T. Taguchi, T. Konishi, S. Sugishita, W. Takei, H. Fujino, T. Ojima, K. Nishizaki, Assessment package for language development in Japanese hearing-impaired children (ALADJIN) as a test battery for the development of practical communication, Ann. Otol., Rhinol., Laryngol. 121 (Supplement 202) (2012) 3e15.

[14] A. Uno, N. Shinya, N. Haruhara, M. Kaneko, Raven's Coloured Progressive Matrices in Japanese children: as a screening intelligence test for children

with learning disorder and acquired childhood aphasia, Jpn. J. Logopedics Phoniatrics 46 (2005) 185e189.

[15] H. Fujino, A. Moriwaki, K. Kamii, M. Watanabe, T. Shiiki, Developmental trend in theory of mind in school-aged children of typical development and with high-functioning autism spectrum disorder: application of the animated version of the ToM test ver.2, Bull. Tokyo Gakugei Univ. Educ. Sci. 64 (2) (2013) 151e164.

[16] I. Fujita, T. Miyake, Syntactic Processing Test of Aphasia (STA), Committee for Researches on Communication and Aphasia, Tochigi, Japan, 2009.

[17] A. Fujiyoshi, K. Fukushima, T. Taguchi, K. Omori, N. Kasai, S. Nishio, A. Sugaya, R. Nagayasu, T. Konishi, S. Sugishita, J. Fujita, K. Nishizaki, M. Shiroma, Syn- tactic development in Japanese hearing-impaired children, Ann. Otol., Rhinol., Laryngol. 121 (Supplement 202) (2012) 28e34.

[18] H. Fujino, T. Matsui, Y. Tojo, H. Osanai, Theory of mind and receptive vocab- ulary in school-aged children with high- functioning autism spectrum dis- order, Bull. Tokyo Gakugei Univ. Educ. Sci. 66 (2) (2015) 311e318.

[19] U. Frith, F. Happ�e, F. Siddons, Autism and theory of mind in everyday life, Soc. Dev. 3 (1994) 108e124.

[20] F.G.E. Happ�e, The role of age and verbal ability in the theory of mind task performance of subjects with autism, Child. Dev. 66 (1995) 843e855.

[21] R. Eisenmajer, M. Prior, Cognitive linguistic correlates of “theory of mind” ability in autistic children, Br. J. Dev. Psychol. 9 (1991) 351e364.

[22] N. Fisher, F. Happ�e, J. Dunn, The relationship between vocabulary, grammar, and false belief task performance in children with autistic spectrum disorders and children with moderate learning difficulties, J. Child Psychol. Psychiatry 46 (2005) 409e419.

[23] S.E. Lind, D.M. Bowler, Language and theory of mind in autism spectrum disorders: the relationship between complex syntax and false belief task performance, J. Autism Dev. Disord. 39 (2009) 929e937.

[24] J. de Villiers, Language and theory of mind: what are the developmental relationship?, in: S. Baron-Cohen, H. Tager-Flusberg, D.J. Cohen (Eds.), Un- derstanding Other Minds. Perspective from Developmental Cognitive Neuro- science, second ed., Oxford University Press, Oxford, 2000, pp. 83e123.

[25] H. Tager-Flusberg, R.M. Joseph, How language facilitates the acquisition of false-belief understanding in children with autism, in: J.W. Astington, J.A. Baird (Eds.), Why Language Matters for Theory of Mind, Oxford University Press, Oxford, 2005, pp. 298e318.

[26] D.B. Harris, Language Experience and Early Language Development, Erlbaum Associates, Hove, U.K, 1992.

[27] M. Marschark, Psychological Development of Deaf Children, Oxford University Press, New York, 1993.