2.1 Participants

Participated in this study: a) three bilingual Pomeranian-Portuguese children, two with 9 years of age and one with 10 years of age, diagnosed with Autism Spectrum Disorder of mild/moderate degree, diagnosed by a multidisciplinary team at least one year before data collection; b) the children’s mothers, one of them monolingual-Pomeranian and two bilingual Pomeranian-Portuguese; c) the head teacher of the Elementary School class of the Municipal Education Network where the child studies; d) two pedagogues from the Center of Inclusive Education; e) two pedagogues from the Pomeranian School Education Program (PROEPO).

The children participating in the research were identified through contact with the Centre of Inclusive Education, the sector that monitors and guides the team of inclusive education teachers from the municipal schools of Santa Maria de Jetibá/ES. The research participants were selected jointly with the team of pedagogues from the inclusive education sector, following the inclusion criteria of this research, which were: being aged between 6 and 12 years old, being bilingual Pomeranian-Portuguese, having a diagnosis of ASD performed by a multidisciplinary team at least one year before data collection, and be classified on the Autistic Spectrum by the Childhood Autism Rating Scale - CARS (^{Schopler, Reichler, & Renner, 1986}).

Thus, after evaluation by the CARS scale (^{Schopler, Reichler, & Renner, 1986}), the following participants were included in this research: a) Marcos, 10 years old, diagnosed with Moderate ASD, classified with a score of 36 on the CARS Scale, son of Pomeranian farmers, from the lower middle class - according to the classification proposed by ^{Paes de Barros et al. (2012)} and according to the Brazilian Association of Research Enterprises (Associação Brasileira de Empresas de Pesquisa - ABEP)⁵ as class D (10 points) -, 4th grade Elementary School student in the municipal public school system, bilingual Pomeranian-Portuguese, who had difficulty having a conversation in Portuguese, as the family usually communicated only in Pomeranian; b) Francisco, 10 years old, diagnosed with Mild ASD, classified with a score of 32 on the CARS Scale, son of Pomeranian farmers, of lower middle class - classified as class C1 (18 points) -, 4th grade Elementary School student in the municipal public school system, bilingual Pomeranian-Portuguese, who tended to use mainly the Portuguese language in his conversations; c) Cadu, 9 years old, diagnosed with Mild ASD, classified with a score of 32 on the CARS Scale, son of Pomeranian farmers, of lower middle class - classified as class C2 (16 points) -, student of the 3rd year of Elementary School in the public municipal network of teaching, bilingual Pomeranian-Portuguese and tended to use the two languages ​​alternately in his conversations.

2.2 Instruments

To collect research data, we used: 1) Childhood Autism Rating Scale - CARS (^{Schopler, Reichler, & Renner, 1986}); 2) Semi-structured interview with the mothers, along with the teachers and pedagogues (^{Hartwig, 2019}); 3) Continued Storyboards (^{Hartwig, 2019}) with the children.

2.3 Procedures

The interviews with the mothers, teachers and pedagogues were carried out individually and aimed to collect data about the child’s developmental history and to understand, mainly, the child’s communication characteristics in different contexts. To meet the objectives, a semi-structured interview was developed based on the designation of previously prepared blocks, namely: a) Sociodemographic data; b) Language development; c) Verbal and non-verbal communication. The interviews were recorded in audio and transcribed in full, the data served as a basis for further analysis of this study.

Data collection with children occurred through semi-structured interaction, using the Continued Storyboard instrument (^{Hartwig, 2019}), which aimed to serve as a visual stimulus for children to build a narrative from the image presented. The instrument, composed of four boards, was developed by ^{Hartwig (2019)} and the application of the instrument was performed by him, who is bilingual Pomeranian-Portuguese. The presentation of the boards occurred alternately so that the children could build a story in L1 and L2 with all the boards.

In order for the child to build a narrative in both languages about each board presented, it was decided to divide the application into two moments, obeying the order of application shown in Table 1.

Source: Elaborated by the authors.

Figure 1 Continued Storyboard - Affective relationship between family members. 

The application script followed an order of questions previously prepared. After rapport performing, the first Continued Storyboard was presented (^{Hartwig, 2019}) and the following command was given in L1: “I would like you to observe this image you are seeing very carefully. What do you see here?” ⁶. After the child’s response, the researcher gave the command for the child to start the story: “I would like you to tell a story from the figure you are seeing. What are these kids doing? Tell a story about them” ⁷. In addition, the participants were exposed to four other questions: 1) “Who is this person?”; 2) “What is she doing?”; 3) “How is he/she feeling?”; 4) “What else do you see?”. After the conclusion of the story, the next board was presented.

The interactions were recorded in audio, had an average duration of 30 minutes and were performed by ^{Hartwig (2019)} in bilingual Pomeranian-Portuguese in a previously prepared room. Observations related to communication indicator behaviors were recorded by the observer-researcher, who positioned himself in front of the child, in the Indicative Communication Behaviors Registration Chart (Table 2). The instrument presents a list of communication indicators, divided into four categories: (1) Non-verbal, (2) Verbal, (3) Comprehension and (4) Non-comprehension. For the definition of the categories, the study conducted by ^{Lopes (2016)} carried out at the University of Algarve/Portugal served as a base, which aimed to raise the communicative difficulties of children with ASD in the school and family context, and to verify in which of the contexts communicative intentionality was greater, also identifying which strategies facilitate this process. The instrument was adapted to meet the objectives of this research, and the indicator behaviors of communication were categorized as shown in Table 2.

In relation to the first block - Non-verbal - facial expressions, understood in this study as non-verbal communicators, were considered those behaviors that indicate primary emotions (joy, anger, sadness, fear, disgust, love), considered universal expressions. In addition to the observations made throughout the application, the child was asked how the people in the images felt, performed once for each participant, in order to understand whether the child would identify the emotional states of the images.

The behavior “drawing attention to oneself” is understood, in this research, as behaviors of addressing the other, in a non-verbal way, so that this turns the attention to the child. It was considered as a variable “to draw the researcher’s attention” without the use of verbal resources.

In this research, gesticulation is understood as a verb with intentional communication. The gestures of pointing and shaking the head in a positive or negative sign were considered in the counting. In this study, gestures are understood as a noun with the intention of expressing an idea or feeling. In this sense, the behaviors of shrugging shoulders and/or miming with intentional communication were considered. The counts were carefully made from the recordings.

In the second block - Verbal -, story is understood as the act of systematically verbalizing facts or events invented from the visual stimulus presented. To configure a story, the narrative should contain a beginning, development and conclusion. A minimum of three to five sentences, elaborated in a systematic way, without pauses, hesitations, repetition of words and/or no response. All children were exposed to the continued storyboards for an average of 30 minutes and the verbal interventions performed by the researcher occurred in the same way in all cases, taking into account only previously prepared questions.

Regarding the topic Non-comprehension: “Does not emit verbal behavior after command”, means the absence of a response from the child, after the researcher asks him/her to tell a story. Command are the phrases spoken by the researcher, as described in Table 1 previously shown.

2.4 Data analysis

The data were analyzed to identify: 1) in the interviews, the reports referring to behaviors that indicate verbal and non-verbal communication observed in family and school interactions throughout the child’s development; and 2) in interactions with children, the content of the stories reported and the type and frequency of behaviors that indicate communication.

Communication indicator behaviors, observed in the interaction with children, were recorded in the instrument Indicating Communication Behavior Registration Chart for, according to the criteria previously described. For the frequency analysis, the sums of behavior observed in the categories (1) verbal communication, (2) non-verbal communication and (3) comprehension were considered. In order to carry out the analysis of the types of behaviors, all communication behaviors observed in all categories were accounted for, in the interactions in Pomeranian Language (L1) and Portuguese Language (L2).

2.5 Ethical aspects

This research is in accordance with the guidelines and regulatory standards for studies with human beings, established in ^{Resolution no. 466/12 of the National Health Council (2012)}. The approval by the Ethics Committee was obtained under Opinion no. 1.534.339. The children only started their participation after parental consent and their own consent. In order to preserve the identity of the research participants, they will be named after fictitious names.

3.1 Communication characteristics of pomeranian children

It was found from the interviews with the mothers that all children have interacted with the Pomeranian language since babies. Marcos’ mother, who spoke only in Pomeranian language, reported that the first words he said were “daddy”, “mommy” and “grandma”, and even before developing fluent speech, he perfectly understood the commands given by his mother: “He understood pomeranian before starting to speak”. Marcos learned to speak Portuguese at school and his interaction with his mother occurred only in the Pomeranian language, presenting difficulties in communicating in Portuguese. In the cases of Cadu and Francisco, the mothers reported that their exposure to the Pomeranian and Portuguese languages ​​occurred simultaneously; so, at school and at home, they interact in both Portuguese and Pomeranian. Literature comprises cases in which the child is simultaneously exposed to two languages ​​as “simultaneous bilingualism” (^{Almeida & Flores, 2017}); and, in this case, the child has two mother tongues.

Francisco’s mother also reported that, despite simultaneous exposure to languages, he was encouraged to speak in Pomeranian since birth. However, when the delay in language development was identified, the speech therapist advised that interactions in the Pomeranian language should be suspended and that the interactions should be carried out only in Portuguese for a better development of the child’s language: “When he was a baby, we talked to him in Pomeranian, but then the speech therapist at that time told us to stop talking in Pomeranian with him to see if he learned to speak faster”. Literature points out that the concept of bilingualism as a cause of pathological disorders that would affect the development of children with typical development is very common (^{Cruz-Ferreira, 2006}). ^{Macnamara (1966)} points out, in his research, that bilingual children had inferior results of linguistic skills such as the ability of verbal intelligence, which, at the time, denoted, in general, that bilingualism would be detrimental to children’s development.

More recent studies refute this view (^{Bialistok & Martin, 2004}; ^{Bialystock, 2001}; ^{Bialystok & Viswanathan, 2009}; ^{Bialystok, 2009}; ^{Carlson & Meltzoff, 2008}; ^{Genesee, Paradis, & Crago, 2004}; ^{Gonzalez-Barrero & Nadig, 2018}; ^{Green, 2003}; ^{Kay-Raining, Genesee, & Verhoeven, 2016}; ^{Pearson, 2009}; ^{Prior & Macwhinney, 2010}), showing methodological problems in the initial studies, as the researchers rarely took into account the children’s linguistic context and they often considered the performance of bilingual children in only one of the languages. In addition, they compared groups of non-homogeneous bilingual and monolingual children, with variation in many other aspects, mainly in the socioeconomic profile (^{Almeida & Flores, 2017}).

Several studies published in recent years point to the positive effect of bilingualism on children with typical development, in various cognitive tasks (^{Bialystock, 2001}); better performance in linguistic awareness tasks compared to monolingual children (^{Genesee, Paradis, & Crago, 2004}); superior reading skills to children schooled in a single language in the second grade of schooling (^{Pearson, 2009}); and, specifically, the effect of bilingualism in children with ASD, which, in a recent study, points to better performance of executive functions, especially the ability of flexibility (^{Gonzalez-Barrero & Nadig, 2018}).

Francisco’s mother also reported that, due to the speech therapist’s guidance, interactions with children occurred in Portuguese, even though communication between family members occurred in Pomeranian language most of the time. She also affirmed that, today, Francisco would prefer to speak only in Portuguese: “He speaks more in Portuguese, Pomeranian only when I argue with him sometimes in Pomeranian, then he answers, sometimes he responds in Portuguese, sometimes in Pomeranian, what he can say he says. Then he thinks a little first, then he speaks. He speaks like this, mixed up, Portuguese and Pomeranian”.

The same characteristic of mixing linguistic systems was observed in the interview with Cadu’s mother: “Cadu says everything mixed up. One minute he’s speaking Portuguese, then he’s speaking Pomeranian. He speaks the two languages ​​together, everything mixed”. The mixture or alternation of languages ​​is known in literature as code-switching or code-mixing (^{Grosjean, 1982}; ^{Mello, 1999}) defined as the “insertion or mixture of words, phrases or sentences of two different codes in the same speech act” (Grosjean, 1980 as quoted in ^{Mello, 1999}, p. 94). ^{Mello (1999)} describes that “code-switching” is a verbal behavior that must be considered as a linguistic performance skill, as it requires a high level of competence in two languages. ^{Mozzillo (2005)} understands that alternations are a means of negotiating changes in terms of distances and social approaches between bilingual interlocutors.

From the interviews with the teachers, it was found that the children come from classrooms whose pairs are bilingual Pomeranian-Portuguese, and the interactions occur in both languages. It was also verified that, in all cases, there is difficulty in communication and social interaction at school, which, in addition to being a typical characteristic of children with ASD, also occurs with bilingual Pomeranian children in the early grades of Elementary School, due to the difficulty in the social use of the Portuguese language (^{Küster, 2015}; ^{Tressmann, 2005}).

It was noted that Marcos had difficulties in having a conversation with other children when he entered school, as he did not know how to speak Portuguese, which can be observed in the teacher’s statement: “When he arrived at school he did not speak Portuguese, only Pomeranian, it was a teacher at the time who ran after the diagnosis, so he could do the exams and stuff ”. Cadu had communication difficulties, and, in the absence of speech, used artifices such as pointing and searching for the toys he would like to use, which can be seen in his mother’s statement: “The teacher talks about his difficulty in communicating with children and with her too, sometimes he would point, he wanted things”. Although the motor behavior of pointing is common in children with ASD, ^{Petersen, Marinova-Todd and Mirenda (2012)} found that, compared to monolinguals, bilingual children with ASD are more likely to vocalize and use gestures, although this is not a general characteristic of this population. However, in the case of Marcos, the use of non-verbal communicators as the gestures occurred because he did not have a verbal repertoire in Portuguese that favored his interaction with others, which distanced him from the social interaction with peers, enhancing the social isolation characteristic of ASD.

The behaviors of looking and pointing are also considered indicators of shared attention. According to the literature, the look is an important level for the development of other communicative behaviors, which develop in complexity. The look underlies the interactional dyad and alternating appears as a more complex variation of this behavior; then, motor actions are added (pointing, showing), forming an interactional triad and characterizing Shared Attention. In this way, even in the absence of verbal language, communicative intentionality is perceived in the research participants from the interviews with mothers, teachers and pedagogues.

In the interview with Cadu’s teacher, it was possible to verify the comprehension difficulties, which is understood by the teacher as a difficulty related to the language: “Maybe it can be something of the language like that, if, for example, it is a shorter sentence, it is easy, but if the command of an activity is a little bigger, sometimes he gets lost a little in the middle of the interpretation”. The same is observed by pedagogues in the inclusive education sector who followed his case: “Using a more elaborate vocabulary, he does not understand. If it’s more contextual, more simplified, then he can do it”. It appears that, in addition to the common communicative difficulties in ASD, Cadu also had difficulties in the social use of the Portuguese language, as well as Marcos.

Thus, it is noted that the difficulties of interaction in Portuguese can enhance the behaviors of social isolation, which was evidenced by Marcos’ teacher’s statement “He is more isolated, the greater contact he has is with his cousin that he always talks to in Pomeranian”. The same would happen in the classroom, when Marcos was unable to make simple requests in Portuguese, such as going to the bathroom: “No, he doesn’t ask. Only in case of going to the bathroom, then he says, he even says ‘banera’ [banheiro - bathroom] , he speaks very differently”. It is observed, here, that the difficulty in social communication, a common characteristic in ASD, is enhanced by his lack of mastery of the Portuguese language, which distances him from the possibilities of interaction with his non-bilingual Pomeranian-Portuguese teacher.

^{Küster (2015)} points out, in his research, carried out with the Pomeranian population, that the difficulty of the Pomeranian child with typical development to communicate at school increases his/her shyness, a characteristic common to this population, due to the difficulty of communicating correctly in Portuguese, which, in this study, is identified as a risk factor that potentiates social isolation behaviors in children with ASD. Now, the difficulty of interacting in Portuguese, associated with the ASD framework whose central axis is the difficulty in social communication, makes social interactions less frequent at school, making monolingual education a risk for the development of the Pomeranian child with ASD.

It was found, in the teachers’ statements, a repeated mention of the characteristics of Marcos and Cadu’s shyness, which literature also points out in relation to the Pomeranian population (^{Küster, 2015}): “He has attitudes that even his peers have, right? Shyness, you know, fear remains”, reported Cadu’s teacher. The same is reported by pedagogues in the inclusive education sector: “He is shyer, he takes a while to open up” and “he is more close-mouthed, demure, he keeps his head down”. In another moment, the pedagogues reported shyness as one of Marcos’s characteristics: “He stays more isolated, he won’t open up so much, he is distrustful. He makes no eye contact, he blushes and lowers his head. He’s very shy”. Here, repeated mention is made of characteristics of shyness, frequent in the Pomeranian population; however, it must be considered that social isolation and little eye contact are common diagnostic characteristics in ASD, and their association with shyness behaviors may represent little knowledge by professionals regarding the characteristics of children with ASD, which may represent a risk factor for his/her development.

In Marcos’s case, the teacher reported that he interacts more when the conversation takes place in the Pomeranian language: “I feel that he feels more comfortable when it is in Pomeranian, you know, you feel that he understands. I, for example, explain it to him in Portuguese and Pomeranian”. ^{Schweers (1999)} believes that the student’s mother tongue is the main means of communication and cultural expression and that its use in the classroom has shown positive indicators regarding the learning process. Thus, in this study, it is considered that the bilingual educational context in the schools of the Pomeranian community is an important means for favoring social interactions in children with ASD, since most of the schools in the city are attended by the bilingual Pomeranian-Portuguese population. In addition, a bilingual school environment enables cultural expression and the strengthening of the language of this population.

At the end of the interview, the teacher reported the affective relationship that Marcos would have with the Pomeranian language: “And in Pomeranian you see his excitement, he likes to talk about fish, you will observe, he talks about fishing, fish, his ox, the reality that lives around him”. ^{Mozzillo (2005)} refers to the emotional factor as one of the motivators to use the mother tongue, understanding that there are people who can only demonstrate their true feelings using the mother tongue. This was also observed in the interactions carried out with children, described below, when emotional expressions occurred more frequently in interactions carried out in the Pomeranian language, when compared to those carried out in Portuguese.

3.2 Interactions from the continued storyboards

In the interactions carried out with the Continued Storyboard, it was observed that the communication behaviors were more frequent, in all cases, in the interactions performed in Pomeranian Language (Figures 2, 3 and 4).

Source: Elaborated by the authors.

Figure 2 Cadu’s communication indicator behaviors. 

Source: Elaborated by the authors.

Figure 3 Francisco’s communication indicator behaviors. 

Source: Elaborated by the authors.

Figure 4 Marco’s communication indicator behaviors. 

Cadu alternated the linguistic system according to mediation, although he tended to tell his stories in Portuguese, mixing words, phrases and expressions from the Pomeranian language. He also tended to contextualize the boards with his reality, for example, when he mentions his animals, in Portuguese, on the Affective relationship board among family members: “The child has a dog, I have too, they call themselves Feroz [ferocious] and Girafa [giraffe]”. Cadu showed the ability to start, develop and build his stories without the help of the mediator; with just one command given, whether in Pomeranian or Portuguese, he started the narrative.

The easy switching between the two language systems denotes the ability to be flexible, in addition to being an indicator of communicative intentionality with the interlocutor (^{Mello, 1999}). It was observed that Cadu’s verbal repertoire was more varied in the Portuguese language, which is possibly related to his literacy process and the social relationships experienced at school (^{Tressmann, 2005}).

Francisco and Marcos were not able to develop a story fluently; however, they communicated in a non-systematic way. Francisco took the initiative to communicate only in Portuguese, emitting words and phrases that did not constitute a story. The words were related to what he visualized on the boards, and his tendency was to correlate them with his personal experiences. The fact that he only interacts in Portuguese may have occurred because he identified the researcher as a teacher. The same was observed in Marcos; however, he communicated only in the Pomeranian language, even when the linguistic system was changed by the researcher. In addition to his narratives being based on the characteristics of the visualized image, his communication occurred only in the Pomeranian language.

Although Francisco and Marcos did not build a story in a fluent way, both were able to report what they observed in the visualized images and showed more behaviors indicating non-verbal communication in the mediations performed in Pomeranian language, mainly on the Affective relationship board among family members, in which it was identified the greatest indicator of non-verbal behaviors of both. They smiled, made eye contact and showed that they were motivated to say what they saw on the board. In addition, in all participants, expressions of joy, eye contact and attention sharing were more frequent in interactions conducted in the Pomeranian language, compared to those performed in Portuguese, which demonstrates greater satisfaction in the use of the mother tongue.

It was also found that the highest frequency of non-comprehension indicator behaviors occurred when the interaction was performed in Portuguese, indicating that, for these participants, understanding is more effective when communication is performed in the Pomeranian language.