The Interactional Structure of Accounts During Small Group Discussions Among Autistic Children Receiving Special Education Support in Finland

Accounts are broadly understood as explanations people provide to justify actions, requests, or arguments in daily social interactions with others (Antaki & Fielding, 1981). They are commonly observed in a dialogue to “excuse, justify or otherwise exonerate the speaker from socially sanctionable behavior” that go beyond causality (Antaki, 1994, p. 43), appearing both in unexpected, controversial, or problematic actions of the speaker (Heritage, 1988) or to display an explanation of why a preferred action can or cannot be performed (Levinson, 1983). For example, accounts can appear spontaneously when one justifies their behavior, anticipating it might be controversial in the interaction context, e.g., “The garbage needs to be taken out, but I will do it in the morning; It is too cold now.” Alternatively, accounts can be solicited by others, as in the following scenario:

Speaker A: Can you grab coffee?

Speaker B: Not today.

Speaker A: Why not?

Speaker B: I already have plans.

Accounts allow us to sustain engagement and manage social relationships (Heritage, 1988) by mitigating responses to unexpected or dis-preferred conduct. Knowing when to account for behaviors, ideas, or actions, as well as identifying an account provided by others is an important conversational skill that supports maintaining reciprocity in the interaction.

In educational contexts, accounts are often encountered within student–teacher interactions following requests (Antaki, 1994), and they are important structures in social dynamics that support both learning and social interactions. For example, participating in the explanatory discourse during shared reading enables pupils to practice providing explanations for events or behavior that transcend their own experiences (Gosen et al., 2013). Therefore, investigating how children account for their ideas and behaviors provides insights into how they are making sense of the world around them, and how they are building their sense of agency in this same world.

In non-autistic children, this skill is clearly developed in early childhood and becomes more sophisticated throughout development. Children master the formal structural aspects of conversation, such as turn-taking, much earlier than the pragmatic-functional aspects, such as coherence and relevance (Blum-Kulka et al., 2010). However, for autistic children, conversational skills involving complex pragmatic-functional structures, such as accounts are potentially challenging (American Psychological Association [APA], 2013) as they could be conceptualized as relying on an ability to understand others’ mental states. According to leading theories on human cognition (e.g., Theory of Mind and Simulation Theory, see Sally & Hill, 2006), social interactions are explained by humans’ capacity to ascribe mental states (e.g., beliefs, thoughts, feelings, intentions, and desires) through inferential processes. These inferences are built either because of the similarity of human minds; by analogy and first experience, we can represent others’ minds (Theory of Mind), or by the ability to mentally simulate others’ feelings, emotions, and intentions (by putting oneself in others’ shoes) (Simulation Theory). In any case, under these theoretical frameworks, autistic people are proposed to lack intersubjective skills and perform poorly in social interactions requiring these sophisticated processes. Thus, conversations with autistic people have been characterized as involving short responses, sporadic initiations, and infrequent sharing of new, relevant information. These conversational characteristics are believed to hinder opportunities to learn from and participate in social interactions, increasing the risk of social isolation (Koegel et al., 2003), demanding auxiliary instructional intervention to improve social conversation (Hughes et al., 2012), and preventing children’s participation and achievement in the regular school context. For this reason, autistic children are often advised to learn from smaller group settings, supported by one-by-one teaching and flexible planning that can be adjusted to their developmental differences. Theories explaining social interactions through the idea of hidden intention and sophisticated intersubjective dynamics, as well as methods based on assessing the complexity of the speakers’ syntax (Price-Williams & Sabsay, 1979), still guide most research and practices that address issues related to autistic communication.

However, other theoretical views center on social interaction instead of relying solely on social cognitive explanations for behavior. For instance, an enactive view proposes that cognition is embodied and non-trivially dependent on the body and enacted (brought forth) through actual engagement with the environment (De Jaegher, 2021; Di Paolo et al., 2018). According to this framework, what drives and maintains the interactions is the principle of self-organization and the fundamental need to enact with the world (including others), not our ability to simulate or predict others’ minds. The crucial aspect of this theoretical framework is not to say that we don't mentalize; surely, we do, but it is to say that in the process of interacting with others, our bodies (and their enactions) have a predominant role in shaping reality. Indeed, investigations of autistic children's accounts using similar frameworks have shown that these children do indeed show evidence of a practical orientation to their interaction partner’s interpretations of their behavior (Henderson, 2019).

Under this perspective, the most important methodological innovation must be to look at the interactions as dynamic living organisms, allowing them to reveal their own life. This means, in the case of autistic interactions, to look at it from the perspective of what participants bring to the dialogue instead of what non-autistic people would understand as relevant to the interaction. It is a new paradigm, challenging our understanding of the communication competencies of autistic people (Fantasia et al., 2014). The present study corroborates this approach; mainly, by creating a setting for autistic interactions to be analyzed.

In inclusive education, the principles of participation and achievement are guiding elements in elaborating pedagogical practices. The participation framework in the context of inclusive education is a complex concept to define, and it is pointed out as one of the main shortcomings of inclusive policies and practices (Black-Hawkins et al., 2007; European Agency for Development in Special Needs Education [EADSNE], 2011). Beyond the idea of physical presence in the classroom, participation entails the interactions established with others, the engagement in the learning process, and the meaning that schooling will have for each child (Ferreira, 2017). In a general manner, participation is understood as the quality of children's experiences whilst they are in school and, therefore, must incorporate the views of the learners themselves (Ainscow, 2016, p. 147). Consequently, the concept of achievement will entail looking at what the child is capable of experiencing in school, going beyond identifying standardized, pre-defined learning outcomes across a universal curriculum. Achievement is defined by the developmental goals in a holistic perspective, and it is not restricted to the results extracted from test or examination marks (Aiscow & Messiou, 2021; Ferreira 2017). Therefore, in the analysis of participation and achievement of autistic students, one must consider the singularity of the developmental processes and their unique learning needs. This often demands an individualized learning plan, tailored social interactions, and more hours of one-to-one teaching, enhanced in small-group teaching and learning sessions.

Small-group discussions constitute different learning settings compared to teacher-fronted activities or one-to-one lessons where the teacher’s role is prominent. Small-group discussions are effective for developing complex thinking (Applebee et al., 2003; Murphy et al., 2009), enhancing text comprehension (Maine & Hofmann, 2016), offering opportunities for students to develop collaborative (Chinn, et al., 2001) and individual reasoning skills (Mercer, 2000). All because small-group interactions facilitate the students to talk and think together, scaffolding multiple directions for the conversation (Van der Veen et al., 2015) and balancing out the benefits of a structured learning setting with a more flexible and varied composition of discursive elements (i.e., that are not reduced to teacher-controlled IRE-sequences) (Soter et al., 2008). The students are free to select in turn another student as the next speaker, creating a situation where the epistemic domain is divided among different participants. As a result, the dominant turn-taking pattern would ideally no longer have the Teacher-Student–Teacher-Student order omnipresent in teacher-fronted classroom interaction, but rather an order that reflects the multiparty character of these discussions: T–S–S–S (Myhill 2006). Our study focused on unraveling the dynamics of small group participation frameworks in the context of special education. It is important to consider that although the indication of special support is, in many cases, a common intervention for the better learning and development of the child, the ideal situation from an inclusive perspective is that all learning, including supporting the development of social and communicative skills happens in the regular classroom (Ainscow & Messiou, 2021).

The present study is part of larger project “Embodiment, (inter)subjectivity and the construction of children’s agency in learning (2021–2023)”¹ conducted by the first author, through which the construction of the sense of agency in learning is examined amongst children that are receiving special education support in Finland. What we present at this time is the qualitative analysis of small-group discussions in which autistic children participated. The aim was to understand how children made meaning of their experiences and participated in small-group discussions. We analyzed particularly the sequence organization (i.e., how turns-at-talk are ordered and combined to produce social action involving accounts provided by children during sharing circles. Different from other actions in conversation, such as requests or complaints, accounts demand awareness of others' expectations and a time-sensitive response to them within the conversation. The following research questions guided the investigation:

To answer these research questions, we build upon previous literature by applying conversation analysis (CA) as our method of research (Sidnell & Stivers, 2013). CA focuses on the “analysis of sequences of interactions and of turns with sequences” (Heritage, 1988, p. 129) as well as participants' interactional moves in their turns at talk (Hutchby & Wooffitt, 1998). This method explores how each statement is connected to previous ones and how participants understand and respond to one another, unraveling how sequences of actions and meaning-making are generated (O’Reilly et al., 2016).

Participants in this study were nine Finnish children (one with a multicultural background, the father is American) aged from seven to ten years, regularly attending school under the regime of part-time special support, in the Pirkanmaa region, Finland. Students belonged to a special classroom composed by twelve children in total. All students involved in this study were either diagnosed with ASD or under investigation to conclude the diagnosis according to criteria of the ICD-10 (World Health Organization [WHO], 1992) or DSM-V (APA, 2013). Although all children participating in the study did not present additional intellectual disabilities and could usually communicate verbally, they often could not engage in school activities involving written or spoken discussions, thus demanding special support. Special support is the third and final level of the three-tier educational support service offered to students in the Finnish Inclusive Education System (see, EDUFI, 2016; Ferreira et al., 2022). Special support can be delivered through different actions such as material adjustments, part-time small-group or individualized lessons with special education teachers, or a flexible curriculum (OSF, 2020). For the participants in this study, special support was carried out in a special classroom within a regular school. Most of the lessons are organized in small groups (from 5 to 12 children per class), and for specific subjects such as second language learning, children are placed in larger groups with children that do not demand special support.

Special classrooms within mainstream schools are uncommon in Finland. Thus, the recruitment process was purposeful; we solicited information from the municipality to find classrooms that fit the study’s requirements (small groups with autistic children) and contacted schools’ coordinators directly. Two schools were contacted, and one accepted to participate in the study. Subsequently, we invited the teacher and children to participate.

The Idea Diary as a pedagogic method was elaborated by Muniz (2015, 2020) to provide a forum through which students could bring their own ideas, interests, and knowledge about their experiences in different social contexts to the school context. It is grounded in the action tripod ‘experiencing—writing—sharing’, where children are invited to reflect on and keep a diary of their learning experiences, current interests, and important events in their lives and share these memories with the class. This activity framework supports teachers in systematically accessing and understanding the subjective productions (i.e., how children make meaning) emerging from students’ learning process, supporting creative learning (Muniz & Mitjáns-Martinez, 2019). Teachers and children establish a dialogical relationship and equally participate in producing knowledge about children’s experiences and interests.

Two fundamental activities structure the Idea Diary, first is the sharing circle. The teacher invites the children to share what they have produced in their diaries in moments called sharing circles (see Fig. 1). The sharing circles consist of small group discussions about the content in the diary. Sharing circles are organized and conducted by the teacher; during this moment, children present their productions (e.g., writings, drawings) and talk about their experiences. The teacher’s role is to support children in expressing themselves.

The sharing circles offer a unique conversational structure that combines a routinized interaction, i.e., all sharing circles are organized and conducted similarly, with the spontaneity of child-centered discussion. This structure is designed to support the development of children’s subjectivity (Muniz, 2020). Unlike classroom interactions where the pedagogical questions are known to the teachers who pose them (e.g., known-answer or test questions; see Grosse & Tomasello, 2012; Schegloff, 2007), in the dynamic proposed in the Idea Diary the epistemic authority belongs to the child (Heritage, 2012). Even though the Idea Diary was not initially designed for a particular target group, previous experience with the tool showed great potential to benefit the learning processes of autistic children (Lima et al., 2022). Sharing circles can organize the interaction, facilitating children’s participation and promoting gradual development in interacting with others. Therefore, we understand that the Idea Diary as a pedagogic method scaffolds social dynamics that build mutual trust and progressivity (Sterponi & Fasulo, 2010).

The second activity grounding the Idea Diary is incorporating children’s experiences into the classroom curriculum. The teacher then uses this knowledge, weaving children’s interests into the school curriculum and potentializing creative learning in the classroom (detailed methodology of the Idea Diary is available on the website http://​www.​lucianamuniz.​com.​br/​metodologia/​). The Idea Diary creates materials and space in the classroom routine for investigating children’s reading/writing of the world, interweaving formal learning objectives with children’s life experiences. The Idea Diary provides opportunities for children to understand reading and writing as processes of communication, production, and self-expression (Muniz, 2020). Participation, in this perspective, is identified by how children construct the Diary (producing material) and gradually incorporate references from the collective work being constructed. Achievement is observed and assessed considering transformations in participation by comparing intrapersonal changes along a timeline of events (Lima et al., 2022).

To implement the Idea Diary, teachers are trained in the method. The training entails two four-hour modules where the teacher will be introduced to the theoretical framework and presented with step-by-step video materials explaining the method, including guidance on mediating the small-group discussion to promote children’s participation. Examples of Idea Diaries are also provided, and a case study is discussed. The training aims at building a new theoretical and methodological repertoire for a student-centered approach, which the teacher can incorporate into their pedagogical approach as they judge fit.

Our dataset was derived from the project database referred to previously and consisted of 11 video-recorded sharing circles of the Idea Diary sessions, where students and their teacher discussed children’s diary entries. The duration of sharing circles averaged 25 min (a total of 240 min of video data). The sharing circles were recorded once a week throughout the Spring 2021 academic semester, according to the teacher’s schedule. Videos included only the discussions over the idea diaries. To ensure the quality of data gathering, the first author was present during the data collection and used two cameras, resulting in synchronized videos in which the teacher and all students were visible simultaneously.

Speaker A: I am making pie for lunch today; would you care to join us?

Speaker B: I don’t think I can make it today.

  I…I am running late this morning.

This analytical process focused on identifying all sequences containing accounts expressed by children across their dialogues in all sharing circles. Subsequently, the second author double-checked the structure of all selected sequences ensuring 100% agreement, which was reached by consensus. The second step of analysis consisted of interpreting the context of the accounts. The first and second authors created a codebook reflecting four types of accounts, which reflect all possible combinations of the initiator of the account (self or other) and the provider of the account (self or other): self-initiated, self-provided; self-initiated, other-provided; other-initiated, self-provided; and other-initiated, other-provided. The codebook consisted of general examples of each of the types of accounts. It was used as a reference for the analysis of individual sequences. Finally, a modified version of CA conventions (Jefferson, 1986, see appendix A for the conventions) was applied to the Finnish portion of the selected sequences and then transferred with adjustments for the English gloss. Each transcript underwent two rounds of ‘checks’ by the research team to verify transcription accuracy. It was important to apply the conventions primarily on the Finnish transcript to ensure that the analysis was properly culturally and linguistically contextualized. Additional components of each interactional turn that appeared relevant to the participant’s ability to make meaning from the interaction were also examined, including gaze, gestures, and bodily actions (figures are supplied for many extracts below to draw attention to these features as appropriate). The final analytical agreement was done through constant discussions between the two authors and their teams. The names of the students and teachers were anonymized; each child was denominated by a number, e.g., child 1, child 2, etcetera.

The accounts appeared in different sequence organizations, all, however, in a turn regularly preceded by an orientation to an object of joint attention (e.g., the idea diary of the child presenting their work to the class). By analyzing children’s engagements in parallel to the types of accounts provided it was possible to understand the different modes of participation created during the sharing circles. The analysis uncovered patterns of interaction that recurred across the collection as a whole. Therefore, the fragments shown in this paper can be considered representative of the collection of accounts. These detailed transcripts also enable readers to verify the claims made by the analysts.

Ethics approval for this study was gained from the Tampere Municipal Institutional Review Board before the commencement of the study (protocol under the name of the first author). Written informed consent for the study was obtained from the participating school, teachers, guardians, or parents of all participating children. Special attention was given to assuring children were aware of and in agreement with the study. Before data collection, the first author took part in weekly classes, meeting the students, establishing rapport, and introducing the video equipment to the children. During this time, the researcher explained the proceedings and, with the teacher’s support, inquired about the children’s thoughts concerning participating in the study.

Consent was denied for two out of eleven children. To respect parents’ and child’s decisions, we scheduled video recordings when the two students were absent from the class. Children not engaged in the study were, however, participating in the other activities of the Idea Diary, and were included in all learning opportunities that derived from the shared discussions.

The accounts in this category were related to the identity of objects, situations, and contents produced by the child. The sharing circles in the Idea Diary usually started with children showing their entries in the diary—drawings, collages, and writings that represented their experiences, learned contents, or interests in the past days. These accounts appeared immediately after two types of requests which are similar to those identified in previous works focusing on small-group discussions (Willemsen et al., 2018); (1) requests formatted as open invitations projecting descriptive answers, such as “what do you have here” and “what is this”; and (2) requests as open invitations projecting specific types of response, for example, “why have you done x”, which inherently targeted the child’s perceptions of their own experience.

When producing other-initiated, self-provided accounts, children positioned themselves as part of the group by choosing whether and what to tell about the content represented in their diaries. Although the participation in these cases was guided (i.e., the teacher is requesting explanations, inviting children to join the sharing process) and constrained by the structure of a wh-questions, children made use of their position to explicate their thinking process, or to exert control over how they will conduct the sharing process according to their intentions. To illustrate, we present examples from the very first sharing circle (see Excerpt 1, Fig. 3) and sharing circle 6 (Excerpt 2, Fig. 4).

In excerpt 1—The letter “p” like the pirate, we see an example in which open invitations projecting descriptive explanations (Willemsen et al., 2018) support the child to share not only the content but the reasoning (the reflective process of why they have noted something in a particular way) when registering ideas in the diary. The teacher identifies that child 2 assigns a random word that starts with the letter “p” to each of his drawings, which creates specific patterns of registering contents in the idea diary (see Fig. 3).

The teacher was particularly interested in the context behind the use of the word ‘evil’ (paha) and requests an account in line 12, which is then immediately provided by child 2; ‘I write a word below each picture. The child does not provide a direct explanation of how the word evil was thought about, but rather suggests in his turn that the word is random. This account is structured without the use of common explanatory conjunction (e.g., because) or reflective verbs (e.g. I think) featured in this type of adjacency pair. The account offers fragments of the information that was requested by the teacher (i.e., why something happened), demanding the interlocutor to interpret that the explanation is contained in the sentence “I write a word below each picture”. The dialogue continues focusing on the pattern established using the letter ‘p’ in all the words (lines 16 and 20), prompting a reflective process. In the following turn (line 22), child 2 provides the reasoning on how he engages with the construction of the diary and interestingly reveals the connection he made between the storytelling (The pirate of words, which tells the story of a pirate that collects words with the letter P) used to introduce the idea of using the diary and the actual construction of his diary. Once again, we see that the explanations provided by child 2 are elaborated as single words or short sentences bringing pieces of information (e.g., turns in lines 20–22), findings similar to other studies of young autistic children’s conversational moves (Bottema-Beutel, 2021; Ochs et al., 2004; Paul et al., 2009). From the perspective of the framework proposed in the Idea Diary (Muniz, 2015), participation and achievement are signaled in the prompt response and the adequacy of turn in responding to the teacher’s request for an account (i.e., the turn is “type fitted” to the sequence). The entire conversational structure and the consistent adjacency pairs formed by the child and the teacher reveal that child 2 has awareness of the processes involved in the construction of his idea diary and takes the opportunity of the small-group discussion to share his ideas. Participation, although guided, is assured both through engagement in the activity and the maintenance of communication in the small-group discussion.

Additionally, the constant use of the letter ‘p’ could be a sign of repetitive interest commonly associated with autism (APA, 2013; Bottema-Beutel et al., 2015), and our data manifests in a manner that supports the construction of the diary. The phenomenon of linguistic repetitiveness in autism is not limited to echolalic utterances (Dobbinson et al., 1998); it can also manifest in the repetition of a specific structure or a particular topic in the dialogue. For our analysis, it is relevant to consider that accounting for such patterns of behaviors creates an understanding of the child’s reasoning and execution of the diary, revealing the child’s reflective process about their work. The repetition is also a source of engagement in the Diary activity; it creates the regularity, consistency, and familiarity that autistic children seek in routines. Furthermore, considering the diary as a tool to express one’s ideas, a parallel can be made to the role of repetitions in autistic communications. Similar to the palilalia (repeats of one’s own prior talk) identified in autistic dialogues (Stribling et al., 2007), this child brings the reference of the letter ‘p’as a repetition of his own ‘turn’, creating a continuation for its engagement.

Another example of other initiated, self-provided accounts starts with a request from peers. In excerpt 2 (Excerpt 2)—My knowledge about the blackholes displayed below, child 8 inquired about specific content in child 3’s diary (line 1, see Fig. 4).

The request was followed by two interesting actions. First, child 3 in line 3 doesn’t offer a direct verbal explanation, but rather indicates where the peer can find the answer to his question. The lexical item ‘katso[look]’ expressed simultaneously with a deictic gesture (e.g., pointing) within a situation where the peer who uttered the question is already visually engaged is interpreted not as a directive to joint attention (which is already secured), but as an embodied supportive turn that mediates turn taking and projects an upcoming action (Bottema-Beutel et al., 2022). Beyond supporting the embodied engagement to the activity in place (i.e., encouraging peers to lean forward and continue looking at the diary), this turn in the dialogue shows that child 3 is expecting that child 8, like the rest of the group, can identify the picture since he has written the information right next to it (see Fig. 4). This is a clear example that child 3 has an idea of what other children should be able to do and organizes his materials around the expectation that other children can read. The following turn, expressed by child 5, answers child 8’s request for an account and confirms the embodied directive given by child 3. From a pedagogical point of view, child 3 is incorporating written language into his strategies to communicate thoughts and ideas, which is new behavior for any of the children in this group, and an important learning event for this child.

Second, this excerpt also illustrates the importance of recognizing each child’s epistemic authority (and their perception of others’ epistemic access) regarding their diary entries as part of supporting their participation and agency. When child 5 provides a possible explanation for 'black hole', the teacher reformulates the question (requests confirmation), pushing for a verbal confirmation (line 7), and by doing so, places the epistemic authority back on the child who made the drawing. This process prompts an account of why the drawing was made, informing the group about the topic, content, and experience in a broader way (see Excerpt 2, first arrow). The situation created by this sequence positions child 3 in the center of the learning dynamic; his experience, interests, and knowledge are recognized as relevant to the group and the learning process in place. Moreover, this context supports autonomous and intentional participation, which are further connected to developing a sense of agency as the understanding or awareness of the consequences of one’s actions in a certain event (Gallagher, 2007). Child 3 prepares the material by considering the sharing circle and the interaction with his peers; he not only writes the name of the figure (which could be a self-oriented behavior), but he uses the sign of the arrow to further evidence (transmit the information) of what he has drawn. The process of sharing specific knowledge is intentional; even further, he expects that others will recognize it and assures that peers and the teacher understand why he has drawn a “black hole” when accounting for the drawing on line 12.

In this category, children accounted for others’ behaviors or works and often portrayed children’s recalling events and reflecting on what seemed relevant to them about others’ productions. These types of accounts were usually offered as ‘candidate’ explanations that were up for negotiation. This framing is critical, because interlocutors do not usually have epistemic access to the reasons guiding their interaction partners’ behavior, and therefore cannot authoritatively provide accounts for their conduct. When children provided candidate explanations for their classmates’ conduct, this offered opportunities for interaction about characters and events on a meta-level, as well as explanations about the process of registering and sharing their ideas in the diary. Participation in these situations was still guided by the teacher’s or peers’ requests for accounts, but with fewer constraints of direct wh-requests, children’s participation could be expanded allowing the expression of their personal preferences, feelings, and opinions important to them. Resembling what happens when children are reading books together (Gosen et al., 2013; Rogoff, 1990), the conversations about the contents in the idea diary provided participation in cognitively challenging talk, and the extension of children’s knowledge related to the topics, concepts, and situations presented in the diaries.

In excerpt 3—Such a long word! we see an example of an other-initiated, other-provided account that provides a child’s perception of the competencies of the peer. The account doesn’t change the configuration of children’s participation, but it discloses the understanding that children may have of each other within the group. In this small-group discussion, children are starting to explore child 2’s diary. Child 2 has drawn a figure that he invented (see Fig. 5) and the teacher is interested in the meaning behind the rather distinct drawing.

The conversation starts with child 2’s initiative in requesting the floor by raising his hands during a pause in the previous sequences (line 1). The teacher acknowledges the request, and invites him to speak by uttering his name with high intonation (line 2) and focusing her gaze on the child. Child 2 takes the turn and directs the attention of the group toward his diary “I have done here two new drawings”, pre-setting the subsequent invitation for the group to explore child 2’s diary, which is completed with the bodily action in line 4. This sequence shows that child 2 understands the conversational structure of the sharing circle and is acting accordingly to what is expected in a small-group discussion—identify a moment to change turns, request a turn, and propose action. Although the turns in this sequence are not clearly articulated in verbal utterances, the bodily actions perform the invitation for others’ to participate.

In the following sequence (lines 6–9), the teacher initiated a new adjacency pair (type offer-acceptance; Stivers, 2012) offering to assist child 2 in showing the diary (line 6 “Can I show and you tell”) and scaffolded the following turns by elaborating on a more direct question to guide child 2’s presentation (continuation of line 6 “what do you have there, child 2?”). With this combination of utterances, the teacher sets the ground for exploring child 2’s diary; not only does child 2 have the floor to present his drawings, but he also is supported on how to begin the explanation of his doings (line 9). Breaking down complex interactional settings into more direct and simple ones supports the continuation of the dialogue (Bottema-Beutel, 2017; Rendle-Short, 2014; Sterponi & Shankey, 2014), thus increasing engagement and participation. This type of interactional scaffolding is commonly identified in conversations with autistic people (Rendle-Short, 2014) and it occurred several times in our data set.

This type of account appeared formulated as spontaneous explanations about children’s own (within the dyad child-teacher) or others’ work (interjecting an ongoing dialogue between the teacher and another child). The accounts either added new information to the ongoing discussion or added remarks of personal opinions about peers’ work. Expansions precede, intervene in, or follow the base of the sequence, and can indicate stance, managing affiliation or alignment, or dealing with issues of intersubjectivity (Stivers, 2012). Structuring spontaneous initiations can be a challenge for autistic people due to the demand of considering various factors involved in social dynamics (Colombi et al., 2009; Fujiki & Brinton, 2009; Koegel et al., 2003). However, conversation analysis has revealed the communication competencies of autistic people on different occasions (Conn et al., 2018; Rendle-Short, 2014; Stiegler, 2007) showing that people on the spectrum may deploy unconventional means to perform the same pragmatic function in the dialogue as conventional conversational structures.

In our data, the self-initiated, self-provided accounts often followed confirmation statements or were structured within a single turn at talk. The latter occurs when a speaker perceives that something in their utterance requires explanation, before such an explanation is requested (Heritage, 1988). The first example of a self-initiated, self-provided account can be found in excerpt 2, presented previously. On lines 29 and 30 (emphasized with the second arrow), the teacher reaffirms the relevance of the turn presented by child 3 and immediately proposes that he chooses the topic of next week’s classwork (which is part of the methodology of the Idea Diary). Prior research has already shown that formulating proposals so they reflect what the child is already doing increases the likelihood that the child will continue engaging with the task at hand (Bottema-Beutel et al., 2018). Child 3 reflects on the proposal (identified by the time between the first response ‘well’ and the account that follows) and accounts for his decision not to choose the topic of the black hole (the topic that child 3 writes on the board, just after the dialogue presented in excerpt 2). Although the formulation of this account is not clearly structured as a reflection (e.g., ‘I will choose this because I am afraid of choosing that’), child 3 anticipates that a justification is necessary (provides the account before writing the word on the board) for what he is about to do next. The account is configured around the fundamental motivation behind the child’s decision, and gives evidence that the child orients to how his interaction partners make sense of his conduct.

In a slightly different manner, In excerpt 4—Establishing the dynamic of the presentation, child 2 explains (line 13) his desired way of participating after the teacher proposes a different way of participating (line 1). The child anticipates the rules of the group dynamic and positions himself as the ‘one ahead of the process of sharing’, conducting the presentation. In this interaction, it is important to highlight how the teacher accepts the child’s explanation and flexibly follows the child’s proposal for the presentation (shown in line 27). To illustrate the dialogue and the embodied actions that support this interpretation, we present excerpt 4 and Fig. 6.

Overall, self-initiated accounts offer evidence that autistic children anticipate their interlocutors’ perception of their conduct and,, different from what previous studies have found (Angus et al., 2015), are capable of providing the answer to an imagined request for explanation. From the conversational point of view, self-initiated, self-provided accounts were used to inform the group of something relevant about the topic, personalizing the experience and revealing motivational stances surrounding the construction of the Diary.

Looking from the perspective of children’s participation, self-initiated accounts in this study supported children to experience the construction of continuous dialogues more autonomously and spontaneously, acting, expressing ideas and perceptions, and contributing to the small-group discussion. Participation promoted by self-initiated accounts served the purpose of entering and influencing others’ sense-making processes, focalizing on a specific topic of the conversation, and effectively regulating other peers’ and the teacher’s behaviors. It is within the self-initiated, self-provided accounts that children are revealing the meaning-making processes, providing the opportunity for the teacher to understand their world of significance and from that standpoint further work with the child.

The educational value of exploring conversation among autistic children is, we believe, self-evident. The first point of contact between educators and peers in the school is through interaction, then a detailed analysis of that interaction is bound to be a useful undertaking, especially considering that conversation with autistic children can often be challenging for non-autistic people who are not experienced in appropriately adapting their conversational repertoires. The findings presented in this article indicate how autistic children can build on their understanding of social dynamics when the epistemic authority necessary for knowledge construction belongs to them, showing issues of broader significance within educational research in autism. In particular, the shift in emphasis away from the autistic person as the source of trouble in interaction, and re-focus on the interaction itself can allow the possibility of modifying the interaction in such a way as to enhance the communicative exchange for all participants. Therefore, the present study contributes to the long-standing discussion on the importance of supporting social interactions, group learning, and peer collaboration in educational contexts for autistic children (Fantasia et al., 2014), arguing alongside others that these children are socially competent (Caldwell, 2006) and that deficit-based practices only reinforce the idea that non-autistic interactions are more valuable. The present study indicates that pedagogical approaches that use learning strategies in which autistic children lead the activity and hold the epistemic authority of knowledge construction support autistic children’s interactions, evidencing interactive dynamics that feature accountability, mutual engagement, and agency.

Moreover, the study also offers interesting inputs on using CA in dialogues among Finnish autistic children for two reasons. First, it addresses a phenomenon (small group interactions among autistic children) that has rarely been explored with the use of CA. For example, according to the Finnish Conversation Analysis Archive (https://​metashare.​csc.​fi/​)—a collection of recordings of Finnish speech in interaction under the domain of Helsinki University and widely used by CA researchers in Finland, there aren’t any samples of small-group interactions among Finnish autistic children. Second, the study explores the possibilities and challenges of applying CA conventions created for English speakers in a completely different language, such as Finnish, pointing out the need to further develop CA methods and conventions in languages other than English. In this study, we identified the need to adjust Jefferson's (1986) conventions. For example, different from English, Finnish relies on specific grammatical structures (kö/ko) instead of intonations to create certain types of questions, or it demands the pronunciation of every letter in the word, which can sound like a prolongation of the word. Thus, the translation work must be culturally sensitive.

Children participating in this study were not diagnosed or evaluated systematically by any of the researchers, evaluations were instead carried out by the school personnel. Specificities of the spectrum such as parameters of language development and social skills were not used to determine the inclusion criteria or baseline for this study. Therefore, it is important to consider individual variations and recognize that case study analysis is designed to explore a given phenomenon in-depth, and not to generalize findings beyond the data. While keeping in mind the qualitative and small-scale character of the present study, we reflect upon two important considerations for future studies – the role of teacher’s communication strategies for supporting group learning among autistic children, and the further exploration of the dynamic system created among autistic children (peer-to-peer). Teachers’ communication strategies can vary significantly even within the same activities (Willemsen et al., 2018). However, in this study, we noticed that the teacher kept a consistent way of interacting with children that involved flexibility and acceptance of the ways by which they decided to engage, which will be presented in a different publication. How many of the advances children made concerning their participation are due to the flexible approach of the teacher? This perspective will be explored in the future by using the Enactive theories of intersubjectivity (Di Paolo et al., 2018) to analyze the dynamics created by this communication strategy.

In conclusion, the present article reports an interesting case of an alternative student-centered pedagogical tool for working with children receiving special support. The study evidence important communicative competencies of autistic children that support their learning and participation in small-group discussions in school activities. Namely, our findings showed that autistic children engaged in constructing explanations for their and others’ conduct, which illustrates their sensitivity in attending to the practical needs of their interaction partners in order to maintain intersubjectivity while propelling the interaction forward.