Peer Collaboration to Support Chinese Immigrant Children’s Chinese Heritage Language Use and Learning in New York

Abstract

Post-monolingual circumstances are increasing reflecting the growing occurrence of multilingualism, migration, and advancing communication technologies. Cultural groups value maintaining heritage languages while also learning the ones in their new life environments. Heritage language maintenance is, for example, promoted by immigrant parents and grandparents in the Northern Netherlands, as well as in other locations. Maintaining cultural heritage language also needs innovative instructional approaches and pedagogy. This article presents a study exploring the role of peer collaboration in supporting children’s Chinese heritage language use and learning in New York City. Drawn from sociocultural theory and extended in prior research examining peer collaboration in writing development, we focused on understanding the role of peer collaboration in Chinese immigrant children’s Chinese heritage language use in an after-school (Sunday school) program. In collaboration with organizers of the school, our study involved a 30 min peer collaboration writing activity. The activity invited children to collaborate with a peer to write a letter about the fun experiences they had in the Chinese language school. Interaction processes and products were collected and analyzed to address two primary research questions including the following: What are the languages and peer collaboration processes that children in this Chinese language school used in their conversations and in the text of the letters they created together? How did children’s use of Chinese and English during peer collaboration relate to language in the texts they wrote? Results indicated that children played diverse, mutually supportive roles during their collaboration, such as suggesting narrative sequences for the letter, including confirming/agreeing, and correcting and clarifying, although different pairs balanced such strategies somewhat differently. Among numerous findings is that children used Mandarin most frequently when narrating events that they found enjoyable at the Chinese language school. The findings fill a gap in research on multilingual peer collaboration, in particular with Chinese and English in the United States. In addition to those findings, research in an after-school heritage language program provides a model for innovative research in practice.

1. Introduction and Rationale

The native Chinese-speaking population has been consistently growing in the 21st century: (Immigration to New York, 2009). New York City’s Chinese population is one of the largest Asian ethnic groups (Profile of New York City’s Chinese Americans Document, 2019). Given such a relatively large Chinese-speaking population, educators and social service providers could augment their support by learning how the young generation maintains their connection to Chinese culture and language. Post-monolingual conditions are growing along with increasingly frequent practice of multilingualism, and migration and mobility (Nam, 2024; Yildiz, 2012). One gap to be addressed in research is how young Chinese immigrant children (specifically the second-generation of Chinese immigrant families) use Chinese with their peers. This study contributes by focusing on whether and how children in a Chinese language school use their heritage language together.

Previous research has focused on parental roles in maintaining heritage languages. Research indicates that immigrant parents generally have positive attitudes towards maintaining their heritage language in the next generation and support bilingual and multilingual upbringing (Daussà & Qian, 2021; Lao, 2004; Zhang & Slaughter-Defoe, 2009). Lao’s survey of parents’ attitudes towards bilingual education highlighted parents’ motivations to enroll their children in Chinese–English bilingual preschools (Lao, 2004). Most of the English-dominant parents and Chinese-dominant parents found it important to continue improving children’s native language in school, especially for Chinese-dominant parents (Lao, 2004). An ethnographic study with 18 Chinese immigrant families in Philadelphia included formal interviews and participant observations (Zhang & Slaughter-Defoe, 2009). Findings revealed parents’ beliefs that learning heritage language contributes to children’s academic achievement and future career, connection with ethnic identity, and family cohesion (Zhang & Slaughter-Defoe, 2009).

Unsurprisingly, research has also shown that schools and educational systems play important roles in teaching and maintaining heritage languages, minority languages, second and foreign languages, and multilingual education (Ol’khovskaya, 2019; Protassova & Yelenevskaya, 2020; Yelenevskaya & Protassova, 2021). Findings of interview analyses, and a meta-analysis indicated the importance of innovative methods and flexible approaches and ethnic worldviews in language pedagogy (Protassova & Yelenevskaya, 2020). Of course, when educators and parents work together to promote heritage language values, children can integrate educators’ and parents’ values for heritage language, thereby supporting children’s enjoyment, confidence, and persistence with family culture and language in schooling (Fisherman, 1991; Lao, 2004; Otcu, 2010; Zhang & Slaughter-Defoe, 2009). Heritage language schools tend to operate as Saturday and Sunday schools, embodying shared educator and parental values for heritage language maintenance.

Research focused on a Spanish-language supporting curriculum developed and studied by Mendoza (2018), for example, aimed at creating environments where students develop communicative ability, and build self-confidence and pride in terms of the linguistic knowledge they already have in their native language. Another approach inspired by sociocultural theory involved the creation and manipulation of didactic models by students to improve their internalization of complex grammatical concepts that would otherwise require long grammatical explanations (Fernandez Parera, 2021). A total of 14 intermediate and 12 advanced learners of Spanish as a foreign language participated in a three-week peer teaching intervention about the use of the Spanish subjunctive tense. Referring to the instructor-presented model, students created their own model including Spanish subjunctive and indicative use concepts and experience to complete several assignments in class and at home. Results of pre- and post-intervention measures and a student survey indicated that the heritage language students significantly improved their interpretative abilities, and all the students found the intervention to be helpful.

Research on peer support for heritage language maintenance and learning have, however, lagged beyond these more top-down influences of parents, educators, and community schools. Peer collaboration is theoretically grounded in community of practice, social integration (Lave & Wenger, 1991), and Vygotsky’s broader sociocultural theory. According to those theories, learning occurs within a social and cultural context. Newcomers in the community develop full participation in the community through learning the specific sociocultural practices and interacting within the community in meaningful activities (Lave & Wenger, 1991). Vygotsky’s sociocultural theory emphasizes the importance of linguistic social interactions and culture in learning. Children learn through collaborative dialogues and participation with more capable others such as teachers and peers (Vygotsky, 1978). With such collaboration, learners can achieve beyond their current capability and developmental level (Vygotsky, 1978). Collaborative activities among children’s peers, such as in our study, tend to promote growth because children of similar ages are likely to operate within one another’s proximal zones of development in different ways than they learn when collaborating with a teacher (Daiute et al., 1993), and better when working together than individually (Slavin, 2016).

Peer collaboration has also been explored to offer opportunities in public education settings. A classroom-based study by Daiute et al. (1993) compared children’s collaborative conversations and writing with peers and teachers, thereby extending Vygotsky’s idea of the ZPD from cultural experts to peers. Using comparative analyses on expert–novice (teacher–child) and child–child tape-recorded conversations and writing samples, their findings offer information about how children use, balance, and change their informal conversational and writing styles with the teacher and with a peer (Daiute et al., 1993). In that study, children collaborated with the teacher and with a peer to write a narrative entry to the class newspaper about historical events they were studying and about current events in their classroom. The study contributed to a theory-based methodology for different collaborative pairing (e.g., student-with-teacher and student-with-student) and corresponding talk and text analyses to apply for insights about the processes and outcomes of each type of collaboration over time. Their study emphasized how children’s mutual and playful engagement of ideas with their peers contributed to children’s positive writing development. This active engagement with peers was important for certain writing developments, like fluency, compared to developments when working with the teacher, which led to improvements in the organization of ideas in paragraphs (Daiute et al., 1993). Peer interactions were playful and generative for text production, while interactions with the teacher focused on narrative structure (Daiute et al., 1993).

Another study by Webb et al. (2021) examined how students work with each other to solve problems in Grade 3 classrooms. The interactions of six pairs (or triads) of students during small-group work were video recorded, and written student work that showed students’ problem-solving strategies were collected for analysis. Through coding of the group interactions, including features such as the number of students who contributed to mathematical ideas, and the extent of pairs’ sustained interaction, their study found peer collaboration and interactions contributed to mathematical ideas to make suggestions for how to begin or carry out problem-solving approaches. This process also includes peers’ reacting to, or building on, others’ ideas in ways that shaped the direction or nature of the joint goal (Webb et al., 2021).

Research by Fernández-Dobao (2020) explored interactions among mixed dyads of heritage language (HL) and second language (L2) learners of Spanish in the mixed language classroom settings. Analyses of audio-recorded HL-L2 learner social interactions elicited by four collaborative writing tasks focusing on language-related episodes revealed that collaborative works benefited both HL and L2 learners of Spanish, though in different ways. For example, HL learner acted as an expert to provide scaffolding to L2 learner, resulting in L2 learner acquiring new vocabulary and HL learner consolidating existing knowledge (Fernández-Dobao, 2020). The study revealed opportunities for learners to co-construct lexical, grammatical, and orthographical knowledge of Spanish via collaborative interaction with peers.

Prior studies have also explored more specifically the role of peer collaboration in Chinese language learning and education, mostly in mixed Chinese language classroom settings. For instance, Huang (2013) studied dyadic interaction of Chinese heritage language learners (CHLs) and Chinese foreign language learners (CFLs) in an adult, mixed Chinese language classroom setting. Analyzing data sources based on language background surveys, comprehensive Chinese skills proficiency tests, and conversations of pairs through diverse tasks to study learners’ oral performance, the study found various patterns of collaboration, namely passive collaboration, active collaboration, and peer tutoring based on peer–peer proficiency gaps. Also, learners generated more language-related episodes and transferred collaborative knowledge to individual work through a high degree of interaction mutuality regardless of proficiency difference (Huang, 2013).

Another study conducted in a classroom setting by Sun (2016) focused on peer collaboration in an English/Chinese bilingual program in a city in Western Canada. With an ethnographic case study of three Grade 5 English/Chinese bilingual students based on notions of dynamic bilingualism and community of practice theories, Sun examined diverse data sources including audio- and video-taped interactions of student–student and student–teacher, participant observation and field notes, and semi-structured interviews. Four major kinds of peer collaboration forms, namely helping, practicing, sharing, and respecting, were identified in contributing to the bilingual students’ translingual practices, positive attitudes, and identities as language learners (Sun, 2016).

Such studies point toward the importance of peer influences in heritage language development, but little has been known specifically about the role of peer collaboration in Chinese immigrant children’s Chinese heritage language use processes. Gaps also remain in knowledge about the variety of native and acquired language combinations with rigorous methodologies in culturally supportive settings. It is necessary to present detailed theories and conduct rigorous analyses about the nature and value of peer collaboration such as various peer collaboration strategies used, patterns manifested, and talk–text relationships in supporting heritage language use. Based on the research rationale with supporting literature and theoretical foundations, we formulated the specific research questions and expectations as follows:

Research question 1:

What are the languages and peer collaboration processes that children in this Chinese language school used in their conversations and in the text of the letters they created together?

Expectation 1: Children will use both Mandarin and English as they collaborate with a peer to compose a letter. They will provide various supportive roles regarding both task-related and relationship-managing during the collaboration process.

Research question 2:

How did children’s use of Chinese and English during peer collaboration relate to language in the texts they wrote?

Expectation 2: Children would use Mandarin and mainly English in conversations as proficient speakers of English but also as Chinese heritage language learners. Their language use and collaboration functions during conversations will play a role in Mandarin uses and narrative development in writing.

Our study was tailored for Chinese immigrant children’s Chinese heritage language use and learning in an after-school program at a cultural Chinese language school. Children in our study grow up in a bilingual or multilingual family where their parents use English, Chinese, and other languages. In this context, the concept of translanguaging provides us with insights regarding learning children’s various language uses in conversation and in writing. “Translanguaging refers to the flexibility of bilingual learners to take control of their own learning, to self-regulate when and how to language, depending on the context in which they are performing language” (García & Wei, 2015, p. 230). Thus, this flexible naturalistic practice can contribute to a high sense of self-efficacy as students self-regulate their learning, and the belief embedded in this practice is exactly that learning is not a product, but rather a process mediated by peers and teachers (García & Wei, 2015). Children’s peer collaboration process would involve translanguaging to navigate their Mandarin use and complete a letter together.

Also, as bilingual or multilingual speakers of English and Mandarin or other language(s), language mixing is expected to occur in children’s conversations and written narratives. Language mixing is a ubiquitous phenomenon characterizing bilingual speakers. A frequent context where two languages are mixed is the word-internal level, demonstrating how tightly integrated the two grammars are in the mind of a speaker and how they adapt to each other. The word-internal level of language mixing includes, for example, speakers’ use of two different languages alternating within the same sentence, or the insertion of well-defined chunks of one language into a sentence that otherwise belongs to another language (Muysken, 2013, 2000). For instance, in the current study, children’s conversations mixed English and Mandarin words:

“and it’s 上完 SAT math 课 [/] SAT math 完了之后，我们 [/] 我们就吃了 [/] 吃了一些 [纸杯蛋糕]”

(“and it’s after finishing the SAT Math class [/] after the SAT math class, we [/] we ate [/] ate some [cupcakes”])

Also, the text children create in the peer collaboration activity conveys implicitly the important information of children’s language learning, their interactions with other children, group practices, and the physical environment. The written narratives would involve language mixing of English, Mandarin, Chinese Pinyin, or with other languages.

2. Research Design Overview

We have developed a theory-based methodology in cultural practice to examine how Chinese American children attending a Mandarin language school use Chinese and English to interact during a task about the cultural center, based on which they collaboratively wrote a related text.

A Chinese language class school setting is a practical setting where children can build on their family cultures, with the support of the shared ethnic culture of the Sunday school and, when relevant, their mainstream school experiences, as well as their peer culture. Ideally, the cultural setting with peers and adults beyond the family would extend children’s motivation to use heritage language spontaneously, as well as to satisfy parental goals. In these settings, children could possibly recall and share their fun experiences/stories with their peers.

The peer collaboration activity asked children to share their interesting experiences at the Chinese language school and eventually to produce a written letter with their peers. The prompts of the peer collaboration writing activity given to the children are depicted in Figure 1 as follows. The specific instructions to children of the peer collaboration writing activity are included as Supplementary Materials.

Studying children’s social interactions and collaborative process with their peers, and their resulting written letters, provided us with an understanding of the role of peers in supporting their Chinese heritage language use and learning. Children’s conversations and written letters would encourage and demonstrate their use of heritage language with a peer and would provide a model for future research and practice in the cultural Saturday/Sunday school setting. The activity also supports the children’s narrative writing development, an increasingly important genre in and beyond schooling.

3. Materials and Methods

3.1. Participants and Measures: Research Context and Data Collection

Participants were seven children aged from 11 to 13 years old who, after informed consent, participated in a 30 min peer collaboration (writing) activity. A brief demographic survey indicated that all the children resided with their Chinese immigrant parents in New York; 6 of the children were born in the United States and 1 child was born in Canada. The child participants are attending a Chinese language school in New York to learn Mandarin and appreciate Chinese culture (Introduction and Mission Statement of the Chinese Language School in New York, 2025).

The school is a non-profit organization focusing on teaching Chinese language and promoting Chinese culture (Introduction and Mission Statement of the Chinese Language School in New York, 2025). After identifying various Chinese language maintenance schools in New York, the first author contacted several, resulting in special interest by the director of the participating school. The Chinese language school we collaborated with is committed to teaching Chinese heritage language and promoting Chinese cultural heritage to the next generations of Chinese immigrants in New York (Introduction and Mission Statement of the Chinese Language School in New York, 2025).

Peer Collaboration Writing Activity

The data collection of the peer collaboration writing activity was a one-session 30 min activity, conducted just prior to Chinese Lunar New Year of 2024. Collaborating with the organizers of the school, the teacher worked with the first author to plan a writing class session during which the teacher’s topic and class activities were to write a letter to the extended families in China, thereby relevant to allow our writing and conversational activities to be embedded in their course activities. The context of writing the letter to the extended families in China was a teacher-designed context, so children had an idea of the audience of their letters. This was an imaginative context as the letters would not actually be sent to children’s families in China, but this context we believed could motivate children’s sharing with their peers and lead to a shared text.

The practice-based peer collaboration activity yielded two types of data sources: transcriptions of audio recordings of the peer conversations and the final written letters/narratives texts. The two sources of data were elicited in the one session. The conversational data was collected by audio-recording the entire interactive conversations among peers during the writing activity and the transcription for analysis. The original texts children created were maintained, photocopied, and then typed for analysis. In pair group 1 and trio group 2, children produced a letter together per group. In pair group 3, each child completed their own letter (so there were 2 written letters in group 3) while working collaboratively in the group.

The teacher had one session of Chinese writing class scheduled before the researcher arrived at the classroom. When the researcher arrived at the classroom, the teacher was teaching the children a classic Chinese work of literature to learn Mandarin and appreciate Chinese traditional culture. After the teacher finished this session, she introduced the researcher to children to conduct the peer collaboration research activities in her language class, as an on-going writing session situated within the Chinese language class. A parent administrator was also present in the classroom. Children were encouraged to use Mandarin with their peers but were also free to use any languages simultaneously in the peer collaboration process, in line with Hornberger’s understanding of heritage language education, that bi/multilinguals’ learning is maximized when speakers are allowed and enabled to draw from across all their existing language skills (García & Lin, 2016; Hornberger, 2005).

Three pairs of children participated in the activity as shown in

Table 1. Child/family demographics (completed by parents) ¹.

Group—Child	Child Participant Pseudonym	Age of Child Participant	Gender	Number of Children in the Family; Their Age(s)	Years of Residence in The United States	Where Was the Child Participant born	The Number of Languages Used in the Family	Language Most Frequently Used When Communicating with Child	Language Most Frequently Used When Communicating with Partner	Code-Switching and Language Mixing Phenomenon When Communicating with Child	Personal Chinese Heritage Language(s)
Group 1—Child1	Zhuzhu	13	Girl	2; 15 and 13 years old	10	Ottawa, Canada	2	Mandarin, English	English	Have code-switching or language mixing, code switch between Mandarin and English	None
Group 1—Child2	YXG	11	Girl	2; 15 and 11 years old	18	New York, The United States	2	Mandarin, English	Mandarin	Mandarin/English	None
Group 2—Child1	ND	13	Boy	4; 31, 23, 19, and 13 years old	19	New Jersey, The United States	2	English	Mandarin	When communicating with children, there often exists language mixing and code-switching phenomenon; often when having difficulty in using Mandarin to communicate with children, parents will switch to English
Group 2—Child2	YQW	12	Boy	2; 16 and 12 years old	18	New York, The United States	3	English	Fujian dialect (Fujianese)	English/Fujian dialect	Fujian dialect
Group 2—Child3	YLW	12	Boy	N/A	N/A	N/A	N/A	N/A	N/A	N/A	N/A
Group 3—Child1	KZ	12	Girl	1; 12 years old	20	New Jersey, The United States	3	Mandarin	Mandarin	Sometimes; Mandarin and English	Shanghainese
Group 3—Child2	XXZ	12	Girl	2; 15 and 12 years old	12	New York, The United States	2	Mandarin	Mandarin	Yes, English and Mandarin	N/A

¹ In Group 2 that consists of three child participants, one child’s parent did not fill out the demographic questionnaire, the information is indicated as N/A.

. Peers’ pairings were formed by the teacher before the researcher began the research activities. As others studying peer collaboration have found, teachers use their knowledge of children’s skills and behavior to create optimum pairings (Daiute et al., 1993). Six immigrant parents of the children who participated in the peer collaboration activity filled out a demographic questionnaire with the help of the parent administrator after a consent procedure. The questionnaires were filled out outside the school and collected when parents came to the Chinese language school because of practical condition after communicating with the school. The detailed demographics of children and families that described the children’s characteristics are presented in Table 1.

The demographic information like the Chinese immigrant families’ personal Chinese heritage languages, the languages that parents use with their children, siblings, etc. provide an overview of the characteristics of the groups and participants in our study.

The researcher used both English and Mandarin when introducing the instructions to the child participants. After reading the instructions first in Mandarin and then in English, the researcher was present during the entire peer collaboration activity, walking around with each child pair, interacting with them and providing any clarification for their questions about the instructions of the peer collaboration activity. Thus, the instructions were ongoing at the beginning of the peer collaboration activity and during the entire peer collaboration process.

3.2. Data Analysis and Interpretation Methods

We adapted conversation and text analysis methods from Daiute et al. (1993) for this study. The conversational analysis categories were consistent with those by Sacks et al. (1974) and others (Hutchby & Wooffitt, 2008; Sert & Seedhouse, 2011), focusing on the construction and sequencing of conversational turns, while also adapted to peer collaboration and multilingual use. In our study, children’s conversations with peers were produced and shaped in the context of the peer collaboration activity (micro), and in the context of the classroom setting of one session of a Chinese language class at the Chinese language school (more macro), including the presence of the regular teacher, the parent, other children who were in the classroom, and the researcher who was conducting the research activity. Attention paid to such sequencing enables us to understand the whole picture of meaning the narrators wanted to convey. Daiute et al. (1993) provided specific guidelines regarding conversation and text analysis. Some of the analytic conversation categories used or adapted in the current study were derived from that research work, with new categories created and developed in the new context of the current study.

3.2.1. Conversation Analysis Process and Developing of Categories

The researcher transcribed the audio recordings of children’s conversations following CHAT Manual (MacWhinney, 2024), word for word, exactly as what children said and how they used the languages, including their laugh, pause, and playing. The researcher marked each sentence for each child’s conversational turn, and paid attention to the sequencing of each interlocutor’s words and sentences. The research team attached importance to children’s code-switching, the different languages they were speaking, and particularly the times when they were speaking/using Mandarin with their peers.

The coding process1 of each conversational turn and the developing of the conversation categories used both top-down and bottom-up approaches. On the one hand, we referred to Daiute et al. (1993) research, applying and adapting some categories from their work such as Asking Questions, Instructing, Posing Alternatives, Initiating Text Sequences, and so on. On the other hand, in the current new context of integrating peer collaboration to support Chinese heritage language, we started from the data (children’s conversations), and created and developed new conversation categories by reviewing the conversation transcriptions. New categories we created included Clarification or Correction, Confirming/Agreeing, Elaborating, Child Mandarin, Narrating the Event, and so on. We listed the most frequently occurring conversation categories with frequencies and percentages in

Table 2. Frequencies and percentages of the most frequent peer conversation categories (functions) of using English and Mandarin ¹.

	Group 1 Conversation		Group 2 Conversation		Group 3 Conversation		Total
Conversation Category	Frequency	Percentage	Frequency	Percentage	Frequency	Percentage	Frequency
Narrating the Event	96	11.05%	41	9.23%	16	9.20%	153
Narrating Process	92	10.59%	44	9.91%	6	3.45%	142
Confirming/Agreeing	81	9.32%	36	8.11%	11	6.32%	128
Talking at the Same Time	56	6.44%	15	3.38%	1	0.57%	72
Child Mandarin	42	4.83%	31	6.98%	12	6.90%	85
Talking2Neighbor	27	3.11%	14	3.15%	40	22.99%	81
Express Uncertainty	37	4.26%	17	3.83%	7	4.02%	61
Laugh: Child	33	3.80%	9	2.03%	6	3.45%	48
Prompting Memory	33	3.80%	5	1.13%	1	0.57%	39
Clarification or Correction	24	2.76%	30	6.76%	3	1.72%	57
Elaborating	22	2.53%	20	4.50%	8	4.60%	50
Instructing	9	1.04%	20	4.50%	15	8.62%	44
Asking Questions	14	1.61%	19	4.28%	9	5.17%	42
Laugh: Both/All	12	1.38%	19	4.28%	5	2.87%	36
Pause	24	2.76%	14	3.15%	7	4.02%	45
Evaluation Positive	11	1.27%	2	0.45%	3	1.72%	16
Play (Taking Break, Eating, Else)	8	0.92%	12	2.70%	3	1.72%	23
Providing Suggestions	15	1.73%	5	1.13%	2	1.15%	22
Y 2 in Mandarin	8	0.92%	4	0.90%	11	6.32%	23

¹ Percentages are out of the total conversational turns within each pair; the percentages were calculated by the frequency of each category divided by the total conversation turns of the audio-recording transcriptions of each group’s conversations (869 units in group 1, 444 units in group 2, 174 units in group 3); Table 2 presents the most frequent conversation categories. ² Y is the researcher who conducted the data collection for the current study and one of the authors of this paper.

as follows. The table presenting the definitions and illustrative examples of these categories was included as Supplementary Materials.

To ensure reliability, in the coding process and data analysis process, the authors checked and consolidated the accuracy of the conversation categories; for example, within each category, the coded conversation turns represent the same or similar features. The researchers also discussed the coding process and the codes in the research group with other research group members.

3.2.2. Plot Analysis on Children’s Final Written Letters/Narratives

Plot analysis applied in previous narrative inquiry studies (Conover & Daiute, 2017; Daiute, 2014b; Fante & Daiute, 2024) was applied to written narratives in the current study. Plot analysis is a helpful cultural tool to identify and analyze basic structure, sequence, and, thus, meaning of narratives (Daiute, 2014b). Plot analysis guides us to identify the meanings enacted in the narrative structure of the texts the peers co-wrote (Daiute, 2014b). Plot elements include “Setting, Character—primary and plot-crucial secondary, Initiating Action, Complicating Action(s), High Point (turning point/climax) and Resolution Strategy(ies)” (Daiute, 2014b, p. 117). Together, we applied these plot elements (categories) to the coding of children’s written letters. The plot elements applied well to our written letters data, including revealing differences in resolution strategy. As shown in previous research (Daiute, 2014b; Fante & Daiute, 2024), plot analysis provides a structural measure of oral and written narrative performance. This detailed analysis also indicates the narrative structures and logics implied in those structures in the oral and written performance by differences by participant groups (in this case pairs) and relationships between language use and narrative structure.

3.2.3. Concept of Character Mapping Analysis on the Narrating Sections of Children’s Conversations and Final Written Letters/Narratives

In addition, we used character mapping analysis on the narrating portions of the conversational data as well as on the final written letters/narratives. Character mapping is a process of analyzing meanings across various narrative perspectives. Character mapping involves identifying characters, thinking about the relation of more and less central characters by identifying frequency of character mentions and charting character enactments to examine their relative roles in narratives (Daiute, 2014a). The concept of character indicates that persons have symbolic meaning in narratives. Characters enter stories in creative manners endowed with the meaning “that authors develop as a synergy of facts that they remember, and ideas or hopes they want to share” (Daiute, 2014a, p. 3). By analyzing strands of meaning associated with diverse characters in narratives, researchers can understand diverse orientations in narrative use, such as with the relatively singular use of “I” characters compared to “we” characters, which may be relevant in these Chinese origin children. Overall, uses of characters in oral and written narratives indicate participants’ integration of social actors relevant to the stories they share (Daiute, 2014a; Fante & Daiute, 2024). In the present study, we wondered, for example, whether the children’s conversations and written letters might echo the language use, such as with characters in the Chinese language school or broader social milieu expressed in Mandarin and/or English. In brief, with character mapping analysis, researchers reflect on diverse strands of meaning operating differently and together in ways that may otherwise remain implicit.

We examined how characters are enacted with certain action verbs and psychological state (psych state) verbs suggesting character roles and their contributions to the meaning of a narrative. Characters described with actions and consciousness usually play active roles in narratives, while characters mentioned as bystanders or recipients of others’ actions and intentions play relatively passive parts in narratives (Daiute, 2014a). We identified the characters mentioned in the children’s narrating parts of their peer collaboration conversations and in their written narratives, to understand the diverse characters and their roles via analysis of characters’ action verbs and psych state verbs. We learnt about the meaning-making process of children’s narrating, and their consciousness enactment in relation with how they humanized the characters such as “we”, “friends”, their actions, and interactions among these elements. This analysis provided us with a lexical measure of narrative performance, to complement the structural plot analysis. Character mapping also provides a measure for considering language preference when describing certain categories of characters, such as whether Mandarin might be preferred for “we” versus “I”.

The following steps explain the data analysis processes of the conversations and written texts:

• Transcribe the three verbal audio recordings of children’s peer-collaborated processes of the peer collaboration activity following the CHAT Manual (MacWhinney, 2024), and translate those conversations in Mandarin to English.
• Refer to the model of peer collaboration in practical language skill—writing acquisition (Daiute et al., 1993)—and apply in our study when applicable.
• Codings and coding process were conducted in Atlas.ti (The Student Atlas.ti 2024 Version).
• Each conversation turn was coded in Atlas.ti, by applying the categories in Daiute et al. (1993) research, as well as by starting with the data and developing new relevant conversational categories.
• Develop 38 conversational categories regarding the transcripts of the conversations of three peer groups during the peer collaboration activity.
• Calculate the frequencies and percentages of some most frequently occurring conversation categories in three peer groups’ conversations, respectively, shown in Table 2. By doing so, we can capture the patterns with regard to what topics and contents children were actively talking about when they were using both English and Mandarin during the narrating process with peers.
• Apply plot analysis based on the methodology of Daiute (2014b) to analyze children’s final written letters. Use character mapping analysis on the narrating parts of children’s conversations as well as on children’s final written letters as stated in Section Character Mapping Analysis on Narrating Parts of Children’s Conversations and on Final Written Letters. Examine how children’s Chinese heritage language Mandarin was incorporated in their writing.
• Understand the connection between children’s conversations and written texts, in particular focusing on the use of Mandarin. In addition to exploring children’s uses of Mandarin and English in their conversations, this talk and text alignment examines the transfer of conversational language to written language.

4. Results

4.1. Conversation Analysis: Peers’ Functioning for English and Mandarin Uses

Children enacted a variety of conversational functions, including confirming and agreeing; qualifying (agreeing with some changes); elaborating; clarification or correction; prompting memory; posing alternatives; disagree; asking questions; and non-verbals—laughing (emotionally; to manage relationship and affiliation); and talking at the same time when conversation turns occur at the same time. Figure 2 below presents peers’ major strategies and functions during their peer collaboration process of practicing and using their multiple languages (almost all conversations were in English and Mandarin in the current context).

As illustrated in Figure 2, “Narrating the Event”, and “Narrating Process” accounted for the majority of the conversation categories in all the three peer groups’ conversations (11.05%, 9.23%, and 9.20%, respectively, in three pairs), especially in group 1 and 2, in which the two categories are the most frequently occurring categories. In addition, Confirming/Agreeing” makes up considerable proportions with 9.32%, 8.11%, and 6.32% in pair 1, pair 2, and pair 3, respectively. Other frequently discussed categories are “Talking at the Same Time” category (6.44% in pair 1, 3.38% in pair 2, and 0.57% in pair 3). “Talking2Neighbor” category is also frequently discussed, but mostly resulted from children in group 3 which accounts for 22.99% of the total conversation turns (3.11% in group 1, and 3.15% in group 2). In contrast, “Clarification or Correction” is more conveyed in pair 2 with 6.76% of the total conversation turns (2.76% in pair 1, and 1.72% in pair 3). Also, “Express Uncertainty” is another frequently discussed category (4.26% in pair 1, 3.83% in pair 2, and 4.02% in pair 3). In addition, “Instructing” and “Asking Questions” categories are frequently conveyed in pair 3’s conversation (8.62% for “Instructing”, and 5.17% for “Asking Questions”).

4.1.1. Code Groups of Peers’ Roles in Mandarin and English Uses

We then organized children’s conversation categories into four major code-group categories to represent peers’ major functionalities for children’s English and Mandarin uses through peer collaboration: Pragmatics, Narratives and Mandarin (task-related), and Non-Verbals to manage relationships such as child laughs. We found most of the children’s conversations were about pragmatics purposes (41.44%, 44.17%, and 38.89% of the total conversation turns in pair 1, pair 2, and pair 3, respectively). The “Pragmatics” code group includes conversation categories such as “Asking Questions”, “Express Uncertainty”, and “Confirming/Agreeing”. The “Mandarin” code group includes “Child Mandarin” and “Y Mandarin”, and is relatively more frequent in pair 3 children’s conversations (21.30%), for which the researcher’s use of Mandarin (Y Mandarin) also accounts for a considerable proportion for instructing purpose of the peer collaboration activity. The detailed frequencies and percentages of each code group are presented in

Table 3. Frequencies and percentages of code groups of major conversation categories ¹.

Code Groups of Conversation Categories	Peers Conversation 1		Peers Conversation 2		Peers Conversation 3		Total
	Frequency	Percentage	Frequency	Percentage	Frequency	Percentage	Frequency	Percentage
Pragmatics: Managing the collaboration	271	41.44%	144	44.17%	42	38.89%	457	42%
2. Narratives	192	29.36%	90	27.61%	24	22.22%	306	28.13%
3. Non-Verbals: Managing relationship/affiliation	125	19.11%	57	17.48%	19	17.59%	201	18.47%
4. Mandarin 2: Child using Mandarin and researcher using Mandarin	66	10.09%	35	10.74	23	21.30%	124	11.40%
Total	654		326		108		1088

¹ The percentages of each main code group are calculated by the frequency of the code group out of the total frequency within a pair (frequency is 654, 326, and 108 in pair 1, pair 2, and pair 3, respectively) as presented in this table. ² The “Mandarin” code group includes “children use Mandarin” and “researcher uses Mandarin”, and an utterance in the “Mandarin” category could also be coded as a conversation function category such as “confirming/agreeing” that served in for example in the “Pragmatics” code group. So, the code-group “Mandarin” with other code groups is not mutually exclusive.

as follows.

The code groups present the main functions of the children’s conversations. The proportions of each code group demonstrate peers’ overall major strategies and functions during the entire peer collaboration processes while practicing multiple languages, with English as the majority language and Mandarin also accounting for a considerable portion. As one of the central focuses of our study, that is, to learn about children’s Chinese heritage language use, we then specifically focused on the analysis on the “Child Mandarin” category when children were speaking Mandarin. We coded the “Child Mandarin” category (conversation turns that children spoke/used Mandarin) as co-occurring with other functional conversational categories to learn what strategies children used or how they used Mandarin in which situations more frequently, as discussed in Section 4.1.2 as follows. The co-occurrence of using English with conversational categories were also presented for reference.

4.1.2. Peers’ Mandarin Uses—General Patterns and Unique Features in Each Pair

Through analysis, we found children’s uses of Mandarin demonstrated some general patterns among all the three pairs, but each pair also used Mandarin differently such as in various occasions and mentioning different characters. Such detailed patterns and features are presented in the following sub-sections.

(a)

Mandarin Is Frequently Used When Children Are Narrating the Event, Mapping Character “We”, and During the Narrating Process

Figure 3 presents the most frequent peer conversation categories (functions) of using Mandarin and English, respectively. The presentations of co-occurrence types of “Child Mandarin” with other conversational categories vividly show the major occasions when children used Mandarin in conversations. Use of English with conversational functions were also reported for reference.

As illustrated in Figure 3, children used Mandarin most frequently when they were narrating the events with their peers. This finding is elaborated with our interpretation of the illustrative examples of conversations. For each conversational turn, the first line is the original conversation transcript with mixing of English and Mandarin, and the line below is the English translation of the entire segments:

Excerpt 1:

C22: Um ok, so it’s like 这是 SAT 老师给的

(C2: Um ok, so it’s like these were given by my SAT teacher)

C2: and it’s 上完 SAT math 课 [/] SAT math 完了之后， 我们 [/] 我们就吃了 [/] 吃了一些

(C2: and it’s after the SAT Math class, we ate [/] we ate some (cupcakes))

C3: 纸杯蛋糕

(C3: Cupcakes)

C2: （吃了）一些纸杯蛋糕

(C2: We ate some cupcakes.)

C1: stop

C2: 还吃了一些纸杯蛋糕

(C2: We also then ate some cupcakes.)

In the aforementioned conversation excerpts with several turns among the three children, we can see that when children were sharing their fun experiences with peers about one child’s birthday, they used Mandarin and discussed Mandarin words and phrases. They made efforts to speak Mandarin and incorporate these Mandarin words and phrases used in conversations into their written letter.

In addition, as Figure 3 shows, children also used Mandarin very frequently when they were elaborating for their peers about what they were talking about during conversations, as well as when conveying confirming/agreeing with what their peers said. Children also frequently used Mandarin when mentioning character “We”. Mandarin was also relatively frequently used throughout children’s narrating process to proceed their conversations and their written products with their peers, as well as when talking about instructions of the activity, and when they were repeating. Mandarin was relatively rarely used when, for example, children were qualifying, when prompting memory to recall some details of their stories, or when children were talking to neighbors. In addition, children used English more for pragmatics purposes such as when asking questions and expressing uncertainty.

Analysis of each pair’s conversations then showed that in pair 1, children mainly used Mandarin when they were mentioning character “We” (24%), when they were narrating the event (22%), during the narrating process (10%), and when repeating (8%). Similarly, in pair 2, children used Mandarin frequently when they were narrating the event (39%), mentioning “We” (11%), and expressing confirming/agreeing (8%). In pair 3, however, the analysis showed that the children did not speak Mandarin frequently when narrating the event or when mentioning the character “We”. Rather, the children in pair 3 used Mandarin more frequently when they were elaborating (42%), confirming/agreeing (25%), during the narrating process (8%), expressing uncertainty (8%), and for instructions of the activity purpose (8%).

(b)

Use of Mandarin—Conversation Turns Directed to Different Interlocutors

In analyses of categories in conversation turns directed to different interlocutors, we observed that Mandarin was most frequently used by pairs 1 and 2 to address children’s peer collaborator, as presented in Figure 4. In pair 3, children used Mandarin with the researcher mainly for instructions of the peer collaboration activity as well as with peers.

Mandarin was also directed to the teacher during the narrating process when children were asking for or confirming some details of the event(s) they were talking about. For instance, in pair 1, when children were sharing the day they met as their fun experiences, they were asking their teacher about the course she had taught them to recall the details when they met during the course:

Excerpt 2:

Zhuzhu: 老师 %com3: [calling their Chinese language teacher C], have you taught fourth grade? [/] have you taught the fourth grade?

(Zhuzhu: Hi, teacher %com: [calling their Chinese language teacher C], have you taught fourth grade? [/] have you taught the fourth grade?)

%com: [Zhuzhu was calling the teacher to inquire about some detail she was talking with her peer YXG about the teacher who taught them in fourth grade]

⌈YXG⌉ ⌊Zhuzhu and YXG⌋: fourth grade

Zhuzhu: 你教我的哥哥吗？他现在[/]在九年级

(Zhuzhu: Do you teach my brother? He is [/] now at the ninth grade.)

Mandarin was directed to the researcher regarding the instructions of the peer collaboration activity such as which languages to use, and how to write the letter format:

Excerpt 3:

Zhuzhu: 我们这个（.）我们需要在中文 （.）额 （.）我们需要中文写这个吗？

(Zhuzhu: Our this (.) we need [to write in] Chinese (.) Em (.) Do we need to write in Chinese about this?)

%com: [“this” refers to the letter produced in the peer collaboration activity]

Detailed proportions of conversation turns for directing to the three main interlocutors when children were using Mandarin and English, respectively, are presented in Figure 4 as follows.

As Figure 4 shows, in pairs 1 and 2, and the total across pairs, the amount of English and Mandarin use, respectively, with the teacher appear similar. Pairs 1 and 2 interacted with the researcher relatively less compared with pair 3, of which the interactions between the children and researcher were mainly about the instructing purpose to encourage children to share with each other verbally. For the bar showing the total across pairs, the amount of English and Mandarin use appears similar for both the teacher and the researcher. In addition, specifically within the use of Mandarin to different interlocutors in each pair and across three pairs, children used Mandarin most frequently with their peer collaborators in pair 1 and pair 2’s conversations and across the total three pairs. In pair 3, children used Mandarin most frequently to the researcher mainly for instructing purposes.

4.2. Plot Analysis on Children’s Written Letters: Mandarin Use Embedded in Plot Elements

The following discussions are the results of plot analysis on children’s written narratives, in terms of Mandarin use and narrative development across the three groups.

Specific plot elements (narrative categories) along with definitions and illustrative examples from sample letters are worth mentioning, as follows:

(1)

Setting: Place; Background; Motivation; other set up for the story;

Example (1):

“The day we met, was a day we will always remember.”

“After a couple weeks of Chinese school, 我们的老师 decided to put the tables in groups.”

(“After a couple weeks of Chinese school, our teacher decided to put the tables in groups.”)

(2)

Initiating Action: Sets the “story” in motion; where the author/speaker indicates “This is a story” (not just a description of a setting…), essay or other discourse;

Example (2):

“When we walked in we were quite Chi Jing (Chinese Pinyin-Chinese phonetic alphabet) because the tables had never been like this before.”

(“When we walked in, we were quite surprised because the tables had never been like this before.”)

(3)

Complicating Action(s): Advance the plot, that was set up in the initiating action;

Example (3):

“几分钟后，其他学生来了。”

(“After several minutes, other students were coming [in the classroom].”)

(4)

High Point/Turning Point: The point of the story, where the events/actions beginning with the initiating action integrate the author/speaker consciousness;

Example (4):

“从那一刻起，我们成了好朋友。”

(“From that moment of time, we became good friends.”)

(5)

Resolution Strategies: The plot resolves with these statements, not necessarily the real-life situation but the story, which must, after all, end (Daiute, 2014b);

Example (5):

“Thus, a loved of friendship emerged, which exists to this day,”

The above plot elements (narrative categories) based on Daiute (2014b) plot analysis methodology were applied to the diverse narratives that children curated, including indicating differences in resolution strategy in children’s letters based on their peer collaboration process. We coded each grammatical sentence with a specific plot element, resulting in the total plot elements as the total grammatical sentences (the maximal number of sentences are 11 in the current study). The specific plot elements in each pair are presented in Figure 5 below. Our analysis showed that the occurrence of the plot elements sequencing was following the sequence of the above-stated elements, that is, the children’s stories usually started with settings, then initiating action, followed with some complicating actions, then with high point/turning point, and finally with a possible resolution strategy when the story after all would complete.

As can be seen in Figure 5, children developed various written letters in the three pairs with different proportions of plot elements based on their collaboration and language use processes. Such differences reveal the structural differences of children’s oral and written performance. Their letters also consisted of various kinds of languages including English, Mandarin, and code-mixing of English, Mandarin, and Chinese Pinyin, as can be seen in

Table 4. Number of sentences in English, in Mandarin, and with code-mixing in each group’s letter.

Language	Group 1 Written Letter (Frequency; %)		Group 2 Written Letter (Frequency; %)		Group 3 Written Letter (Frequency; %)
English	8	73%	6	55%	4	20%
Mandarin	0	0%	5	45%	10	50%
Code-mixing	3	27%	0	0%	6	30%
Total number of grammatical sentences	11	100%	11	100%	20 1	100%

¹ Children in group 3 wrote two letters about their fun experiences as they worked together, so the total number of grammatical sentences is 20. We counted the percentages.

as follows. In addition, children’s uses of English and Mandarin regarding the specific plot elements also showed distinct patterns across the three pairs.

Then, we observed Mandarin use was embedded in children’s narrating and various plot elements. For instance, Mandarin use was connected with the high point/turning point when the story or actions beginning with the initiating action integrated the author or speaker’s consciousness such as in Example (4) above. In addition, as can be seen in the plot elements and examples above, children’s letters were developed somewhat with code-mixed uses of Mandarin, Chinese Pinyin, and English, and some sentences were written all in English or all in Mandarin. Specifically, we also found that in group 1, children collaborated most frequently through conversations, children had more conversation turns, and creatively used Mandarin in their high point as shown in the following excerpt:

Excerpt 4:

我们四个人 seemed to hit it off right away, “using”4 papers from YXG’s notebook, writing notes to each other.

(The four of us seemed to hit it off right away, “using” papers from YXG’s notebook, writing notes to each other.)

The above narrative excerpt is the children’s high point/turning point in their written letter, when the four friends, including the two children who were participating in the activity as the primary characters in the children’s stories, started to do something together that they found happy, and thus developed friendship together. Children also mentioned “the four of us” using Mandarin in the high point.

In group 2, the children’s narrative was solely in English for the high point/turning point, but, remarkably, the children used Mandarin frequently in the “Settings”, and “Complicating Action(s)” such as in the following excerpts from the children’s written letter:

Excerpt 5:

我一进中文课时，我拿了一盒纸杯蛋糕。—Setting

(At the time when I entered the Chinese language class, I brought a box of cupcakes.—Setting)

这是从我的 SAT 老师给的。—Setting

(This was provided by my SAT teacher.—Setting)

Before Chinese class, my SAT classmates ate some cupcakes and ended up not eating 5 cupcakes.—Initiating Action

Those 5 cupcakes were saved for my Chinese classmates.—Complicating Action

几分钟后，其他学生来了。—Complicating Action

(After several minutes, other students were coming [in the classroom].—Complicating Action)

In the above narratives, children used Mandarin in settings and in the complicating action in their written letter. The contents generated in the written letter were closely related to the narrating parts in children’s conversations. Children also mapped character “other students” when mentioning their friends using Mandarin.

In group 3, one child’s narrative started with Mandarin. Her narrative settings, initiating action, and complicating actions had code-mixed Mandarin and Chinese Pinyin, and code-mixed English, Mandarin, and Chinese Pinyin. For example:

Excerpt 6:

我的爸爸 yi qian gao xu wo [Chinese Pinyin] 很多次

(My dad told me for many times)

The high point and resolution strategy of her written narrative were in English. The other child in this group almost used Mandarin in the entire written narrative.

Character Mapping Analysis on Narrating Parts of Children’s Conversations and on Final Written Letters

Embedded in the plot analysis, we used character mapping analysis on the narrating parts of children’s conversations and on final written letters. Through analysis, we found various character mappings combined with action verbs and psych state verbs in each group. Overall, the mapping of “we”, “they”, “our”, etc. symbolizing children’s consciousness of peers, along with psych state verbs such as “Cognition”, and “Feelings” are noteworthy findings.

For each character, we listed the mention of the character; counted the number of conversational turns that mentioned the character, along with the frequencies of their action verbs; and studied the same with psych state verbs in each pair. “Action Verb” contains all kinds of action verbs. Regarding psych state verbs, six major psych state verbs categories were developed in Atlas.ti: “Cognition”; “Attributing”; “Thoughts”; “Like/Love”; “Capability”; and “Feelings”. Our analyses indicated that children provided various important roles regarding psychological states when they were collaborating with their peers to support Chinese heritage language use and learning and to complete the written letters.

For example, “Cognition” includes memories traces, decision, and choice making when children were sharing and discussing the details of their events, talking about the context of their event in which other character’s decision making like their teacher in their story also came across.

The “Thoughts” category shows children’s personal viewpoints. “Attributing” conveys who said which contents, who did what action, who swears to confirm the factual feature of what they are talking, etc., all contributing to children’s details in the conversations and in their written letters. “Capability” indicates children’s reported or expressed ability to do something such as “I can text her”; “I cannot say cupcakes in Chinese”.

“Like/Love” is children’s emotional feedback and comments toward their peers’ conversations. The “Feelings” category includes children’s personal feelings and interpretations of experiences and emotions based on their social and cultural understanding and their individual memories during the peer collaboration activity.

Singular use of “I” characters compared to “We” characters and third-person peer’s names and teacher all offered us an understanding of diverse characters and their actions and consciousness enactment in narrating. Children’s psych states like thoughts were enacted in practicing Chinese heritage language and English. Their frequent mentioning of “we” suggests their consciousness of cultural identity and sense of self in the context of heritage language practicing with peers during activity.

For example, character “we” was frequently mapped with psych state verb “Cognition” (e.g., regarding memory):

Excerpt 7:

我们都忘记了

(We all forget)

Character “we” was also frequently mentioned with psychological verb “Feelings” such as “我们成了好朋友” (“we have become good friends”).

Character “I” was also mentioned frequently with both psych state verb “Cognition”, and “Feelings” such as “I [/] I [/] I [/] I will always remember” with personal feelings to always remember the day they met. In another example “I all could be … couple of years tortured in Chinese school ☺”, the child humorously expressed her feelings about the experiences of learning Mandarin together at the Chinese language school: though challenging, it was unforgettable and they had happy learning experiences.

Also, we captured that children used and repeated using Mandarin for those they found interesting, as follows:

Excerpt 8:

YXG: When I walked in, I was very surprised. When…

Zhuzhu: When (.) [/] when I (.) walked (.) in, I (.) was (.) what?

[%com: writing while repeating the sentence]

Zhuzhu: 吃惊，吃惊，吃惊， 吃☺ [/] 吃惊

(Zhuzhu: Surprised, surprised, surprised, sur ☺ [/] surprised)

YXG: ☺

YXG: 吃惊

(YXG: surprised)

Zhuzhu: 吃惊

(Zhuzhu: surprised)

[%com: children were happy to verbally repeat the Chinese word together 吃惊 meaning surprised in English]

Zhuzhu: 吃惊！

(Zhuzhu: Surprised!)

YXG: Yeah. %com: [confirming]

Zhuzhu: I have to [/] I have to put in Chinese words, ok? So 吃惊

(Zhuzhu: I have to [/] I have to put in Chinese words, ok? So “surprised”)

The phrase in Mandarin 吃惊 (surprised) means “eat the surprise” in Mandarin literally, which in English means “we are surprised”. We found frequent character mappings of “I” with the psych state verb “Feelings” “are surprised” when children were sharing their fun experiences with their peers, and when they were trying to use some Mandarin words and phrases in their written letter.

To sum up, regarding pair 1, Zhuzhu and YXG, the singular first-person character “I/me/my/mine Zhuzhu” was the preferred character (16.03%), followed by character “We” (11.16%). Third-person peers’ names accounted for a considerable proportion, for example, “Char Peer S” (7.26%); “Char she child” (4.49%); and Char Peer KZ (3.63%) when children used third-person pronouns to mention their peers. Other frequent character mappings were first-person character “Our” (3.21% of the total conversation units); and “Char they/their/them/people/person” when children mentioned their peers and friends (1.36%).

In group 2, YLW, YQW and ND, we found different patterns. First-person character “I” was mostly mentioned with action verbs when using Mandarin, such as “我一进中文[/]我一进中文课时” (“when I went in, when I went in the Chinese class”); and “我拿 [/] 我拿[/] 我拿了一盒 [/] 一盒” (“I brought, I brought a box, a box of …”). “I” was also frequently mentioned with psych state verb “Attributing” such as “I mean college high school”; “No, I am [/] I am kidding; and I am saying I wanted to distribute the cupcakes”. Also, second-person character “You” was highly mapped with action verbs such as when language mixing occurred between English and Mandarin: “like that you went into the classroom, and SAT [/] SAT 数学 那个class” (“like that you went into the classroom, and SAT [/] SAT math class”). In addition, children used Mandarin when mentioning character “I” with psych state “Cognition” regarding memory, such as in the following excerpt:

Excerpt 9:

C2: 我还记得，对对对

(C2: I still remember, yeah yeah yeah)

%com: [talking to C1 and C3]

In the above conversation turn, the first-person character “I” was one of the children, and the conversations were directed to their child peers in the group, when sharing with his peers that he remembered (“Cognition”). Children’s memory traces were context-based regarding classes, teachers, peers, and so on in the context of their storytelling and sharing with peers.

In group 3, KZ and XXZ, the most frequently mentioned character was first-person character “I”, with action verbs such as “Once I went into the forest”. Also, a frequently mentioned was character “I” with psych state verb “Feelings” such as “So [/] so my story is about um like how I made friends, like in Chinese school.” In addition, character mapping of “we” with action verbs was also frequent, such as “um we went into [/] like a woods behind the school.”; “Um and we were like went deep into the woods,”; and “Um. And [.] um we took some pictures of the deer”. Also, character “we” was frequently mapped with psych state verb “Cognition” and character “I” was mentioned with psych state verb “Thoughts” and “Cognition” such as in the following excerpts:

Excerpt 10:

XXZ: I don’t think that happen because, I kinda of that some lost

KZ: Ohh

XXZ: but I found some way back, as there was a tall building

KZ: That’s cool!

In the conversation in excerpt 10, first-person character “I” was mentioned with psych state verb “Thoughts”: “I don’t think that happen…”, expressing the child’s viewpoints. Character “I” was also mapped with psych state verb “Cognition” such as “I found some way back…” when the child was sharing her experience in the Chinese language school with her peer KZ while working collaboratively in the group. In general, children talked much less in pair 3; thus, they also used fewer Mandarin words compared with the aforementioned two pairs, but their written letters were achieved with Mandarin uses abound based on their collaboration.

The following specific results of character mapping analysis with illustrative examples of the original excerpts of the children’s written letters also highlight how Chinese heritage language uses are emphasized in this analysis. Mandarin uses were embedded in children’s character mappings with action verbs and psych state verbs. For example, in pair 1, character “we” was highly mentioned (11.16%), with action verbs such as met and sat at:

Excerpt 11:

我们五年级的第一天，我遇见了 Zhuzhu, YXG, 和 S。

(The first day at our fifth grade, I met Zhuzhu, YXG, and S.)

我们一起坐在一个桌子。

(We sat at the same table together.)

We can see the character mapping of “我们” in Mandarin, which means “we” in English, and action verb “坐在” in Mandarin, which means “sat at” in English.

In addition, children used Mandarin for mapping character “we” with psych state verbs such as “Feelings”:

Excerpt 12:

从那一刻起，我们成了好朋友。

(From that moment of time, we became good friends.)

We can see the character mapping of “we” with psychological verb “Feelings” “我们 成了好朋友” (“we became good friends”). This finding again consolidated our discussion that the children were developing an identity and landscape of consciousness of “we” and “friends”. This landscape of self-concept (Bruner, 1986) was being formed in children’s Chinese heritage language use and collaborative processes in the context of peer collaboration activity.

In summary, the analysis revealed a rich variety of structural (plot development) and conceptual (character development) narrative expertise among this group of multilingual children.

4.3. Analyzing and Connecting Conversations with Written Letters/Narratives

In this section, we connected children’s peer collaboration conversations with their written letters, in understanding their inter-relationships. For instance, did peer interactions and collaboration improve their writing, and if so, how? Did the children’s collaboration and sharing with pairs facilitate their Chinese heritage language uses in writing?

4.3.1. Peer Interactions Advance the Development of Written Letters/Narratives

We found, generally, that peer interactions and collaboration played a supportive role in children’s written letter length, the development of letters regarding plot elements, and Mandarin uses in written letters.

Regarding the connection between conversations and written letter length and Mandarin uses in written letters, the results of our study indicated that some conversational categories made a difference to children’s written letter length and the use of Mandarin words and sentences in letters. Children who were narrating the event with peers more (indicated by the higher frequency of the “Narrating the Event” category), and who had more abundant narrating process with their peers (higher frequency of “Narrating Process” category), wrote a letter with more grammatical units/sentences. Likewise, “Confirming/Agreeing” and “Elaborating” categories that helped children to proceed their conversations and write-ups of their letters also increased the children’s length of written narratives. In addition, children who provided more “Clarification or Correction” during the collaborative works managed to use more Mandarin words in written letters.

4.3.2. Correlation Between Conversation Categories and Plot Elements of Written Letters/Narratives

This part of the analysis connects the analyses of the conversations and texts for the three pairs of young collaborators overall and for each pair. We analyzed the relation between some representative conversation categories and the plot elements of children’s written letters. Based on the conversation analysis stated earlier, there was a large number of proportions of narrating with Mandarin including “Narrating the Event” and “Narrating Process”. Based on qualitative interpretations5 along with the contingency

Table 5. Conversation categories and plot elements of children’s written narratives.

	Conversation Category (Frequency)										Plot Elements (Frequency)
	NE 1	NP	CA	CM	T2N	TAST	EU	CL	EL	LC	Setting	Initiating Action	Complicating Action(s)	High Point/Turning Point	Resolution Strategy(ies)	Total Sentences
Pair 1	96	92	81	42	27	56	37	24	22	33	2	1	5	1	2	11
Pair 2	41	44	36	31	14	15	17	30	20	9	3	2	5	1	0	11
Pair 3	16	6	11	12	40	1	7	3	8	6	3.5	1	2.5	1.5	1.5	10

¹ The full names of the abbreviations of the conversational categories that are presented in Table 5: NE: Narrating the Event; NP: Narrating Process; CA: Confirming/Agreeing; CM: Child Mandarin; T2N: Talking2Neighbor; TAST: Talking at the Same Time; EU: Express Uncertainty; CL: Clarification or Correction; EL: Elaborating; LC: Laugh: Child.

presented as follows, we found conversational utterances categorized as “Narrating the Event” and “Narrating Process” related to text analysis category (plot element) “Complicating Action(s)”. For example, in pair 1, “Narrating the Event” and “Narrating Process” categories accounted for the highest frequencies, and the “Complicating Action(s)” plot element also accounted for the highest frequency. The same pattern applied to pair 2, with the highest frequencies of categories “Narrating the Event” and “Narrating Process”, leading to the highest frequency of the “Complicating Action(s)” plot element in children’s written narratives.

In addition, we found, generally, that children who had more conversation categories like “Narrating the Event”, “Narrating Process”, “Confirming/Agreeing”, “Express Uncertainty”, and “Elaborating”, produced a written narrative with more clear and well-developed plot elements with well-identifiable setting, initiating action, complicating action(s), high point/turning point, and resolution strategies such as in group 1’s written narratives. Pair 1 who had the most pragmatics functional conversation categories including “Confirming/Agreeing”, “Express Uncertainty”, most “Narrating Process” and most non-verbals such as “Laugh: Child” also had most resolution strategies in their written narratives.

By contrast, children who shared less frequently with their peers in conversations or those who had more off-topic conversations produced written letters with less identifiable plot elements and writing quality. For example, in group 2, the narrative did not have a resolution strategy with the narrative ending with the high point/turning point. Also, in group 3, in child XXZ’s narrative, the settings were long, while the complicating actions were relatively shorter. Detailed connections between some frequent conversation categories and plot elements across each pair are presented in Table 5.

5. Discussion and Conclusions

5.1. Peer Collaboration Activity Promoting Chinese Heritage Language Learning and Education in Cultural Context

In conclusion, this study indicated the spontaneous rich and varied use of Chinese and English among multilingual children. The study thus suggests recruiting more peers in this setting to explore whether and how the observed patterns might be consistent with one or another of those identified here. For example, Zhuzhu and YXG were the most active narrators who talked with their peers and shared their experiences as shown from their vast number of total conversational turns. Sparing no effort to use sufficient Mandarin in their conversations, children also creatively integrated Mandarin words and phrases in their written letter and used Mandarin in the high point/turning point in their written narrative. They also mentioned “us” using Mandarin as “我们四个人” (“the four of us”) as indicated in character mapping analysis, suggesting their landscape of consciousness and cultural identity of “peerness” when using heritage language. Children YLW, YQW, and ND also played an active role in sharing their fun experiences verbally. Though there were some off-topic conversations, they managed to use Mandarin in both their conversations and written narratives. Similar to peer group 1, in group 2, the frequently discussed Mandarin words in conversation were used in their written letter, such as “纸杯蛋糕” (“cupcakes”) and “中文课” (“Chinese language class”). Children KZ and XXZ, relatively, did not talk much, but they played an active role in “Narrating the Event” sharing with each other about their fun experiences at the Chinese language school verbally. This collaborative narrating of the event with a peer process led to Mandarin uses in their written letters: KZ’s letter was almost all written in Mandarin, and XXZ also used a considerable amount of Mandarin in the letter, with some Chinese Pinyin (Mandarin phonetic system) such as “gao su wo 告诉我” (“told me”), and “dong wu 动物” (“animals”) and sound-like Cantonese Chinese Pinyin expressions. These promising results confirmed the supportive roles and contributions that children’s peers at the Chinese language school have played in their Chinese use and learning in the meaningful peer collaboration activity context.

In our study, we examined how peer collaboration and related writing in Chinese heritage language and English proceeded spontaneously among seven children. Children used both Mandarin and English as they collaborated with peers to compose a letter. Although they mainly spoke English, their Mandarin uses accounted for a considerable portion of the total conversation turns. Children’s peers played diverse supportive roles in their Chinese heritage language use and learning. The analysis also revealed a range of patterns among the peers, which we will examine further as we add more peer pairs following this same process.

Our research design of the children-centered collaborative learning of Chinese heritage language in the U.S. of multilingual settings contributes to the new ideas of teaching Chinese immigrant children’s minority language and immigrant language with cultural heritage preservation. We embraced and encouraged children’s various and free peer collaboration and interaction patterns in all the three groups, consistent with Webb et al. (2021) research supporting a variety of interaction patterns in collaborative groups that lead to productive participation by all members of the groups (Webb et al., 2021). The results of our study were consistent with Webb et al. (2021)’s suggestions for teachers and researchers to work on creating opportunities for students to engage with each other during collaborative work that make sense to students at the moment (Webb et al., 2021), like in the practical context of our peer collaboration writing activity in the Chinese language classroom of the Chinese language school.

The various functionalities and roles of peer-collaborated learning for Chinese heritage language enhanced the heritage language instruction and education for multilingual learners along the migration journey of Chinese immigrants in the United States. Functional bilingualism (Protassova, 2008) requires support of multiple languages including heritage language at multiple aspects, and the supportive roles of immigrant children’s peers who are also born in the immigrant families and learn Chinese heritage language together in the after-school Sunday Chinese language school will be a cogent addition to the heritage language education and its cultural heritage preservation.

In writing, children used English, Mandarin, and code-mixed different combinations of English, Mandarin, and Chinese Pinyin. Plot analysis of children’s written letters with specific plot elements also offered insights that peer collaboration indeed facilitated children’s use of Mandarin in their writing through collaborations and social interactions among peers. Their use of Mandarin in specific plot elements varied across the three groups. Character mapping analysis of the narrating portions of children’s conversations and of their final written letters indicated the children’s diverse patterns of Mandarin uses in characters enactment while narrating with their peers. The findings also suggest that children’s identity and landscapes of consciousness (Bruner, 1986) were illustrated via their Chinese heritage language use with their peers in this context. Children were frequently using character “we”, representing their “peerness” consciousness and identity construction when they were using Chinese heritage language with their peers. The children’s state of being conscious of themselves with peers and friends were developing during their process of using Chinese heritage language with their peers in the practical context of the peer collaboration writing activity. More specifically, the findings of this study echo Daiute et al. (1993) which highlighted the spontaneous conversation among peers connected with structural and semantic generative written text composing.

5.2. Contributions, Implications, Limitations, and Further Directions

This research provided important data and findings regarding the supportive roles and functionalities that peer collaboration have played in Chinese immigrant children’s Chinese heritage language Mandarin use, learning, and education. It is meaningful for our study to contribute to sociocultural theory, including Vygotsky’s ZPD (Vygotsky, 1978), and situated cognition (Roth & Jornet, 2013; Brown et al., 1989) with new Chinese heritage language learning practices and sociocultural activities context. Our study drives innovation towards diversity by adding Chinese heritage language use and learning processes, and the diaspora of Chinese immigrants in New York. Our research design used children’s narrating as a protocol for data collection in informal context embedded in usually structured Chinese language class to study children’s Chinese heritage language use and peer collaboration strategies. Our study provides educational implications for Chinese heritage language education and instruction of immigrant children in the U.S. The diverse and novel patterns of peer interactions we described contributes to Chinese heritage language learning in classroom settings, and to Chinese immigrant families’ everyday language use contexts to maintain and teach their Chinese heritage language. The supportive functionalities of peers during collaborative writing activities are also consistent with Fernández-Dobao (2020)’s research that confirmed the conducive opportunities for both heritage language and L2 learners of Spanish via collaborative writing activities.

With all the contributions, the current study also has limitations. For example, we did not add children’s Mandarin proficiency for describing the pairing and descriptive findings of their performance. Also, a larger sample size would benefit the correlation analysis between some frequent conversation categories and written letter length, and Mandarin use in written letters, and the choice of more appropriate statistical techniques. In addition, the process of instructions to children for the peer collaboration activity could be more informative, such as which language to use in conversation and in writing. The transcriptions of the audio-recordings of the children’s conversations may also be improved in further studies to ensure more reliable qualitative analysis and interpretations.

To improve the current study and point toward future research directions, we may study children’s use and development of Chinese heritage language and peer collaboration strategies for the mutual goal of composing the written letter across years. It will be meaningful to follow-up, to study the same group of children to learn about their development in using languages and utilizing strategies in group works, as well as to include and study some new pairs of children participants for more suggestive descriptive findings connected with the data analysis and findings we have gained in the current study. We will also include the component of children’s Mandarin proficiency, to describe their performance of language uses along with their collaborative styles and patterns. In addition, we will study the relation among the children’s language use in conversations and in writing, with the demographics elements such as the language that parents use with their children (from the demographics questionnaire), and children’s Mandarin proficiency.

Our study used innovative narrating as a research protocol for data collection to study Chinese immigrant children’s use of multiple languages with a focus on Chinese language, and their varieties of peer collaboration strategies and functionalities to accomplish joint work of writing a letter with their peers. We presented a practice-based study with conversational and written data sources. We employed conversation analysis, plot analysis, character mapping analysis, along with correlation analysis to understand the extensive data of the multiple discourses that children generated in peer collaboration conversations and writing. The contributions of the current study align with Daiute et al. (1993) research that advocated for more innovative research on peer collaboration that delves into diverse formats and nature to support practical language instructional design and pedagogy. In detail, we described and discussed the children’s peer collaboration patterns, styles, and language uses in each pair and across the three groups of children. We fully appreciate children’s productive and unique participations in our peer collaboration research activity, as a way to help our understanding of how peer collaboration enhances their Chinese heritage language use and learning in the practical context that children enjoy and the school community supports. Elaborating conversation analysis informs us about the children’s dynamic language use processes. Through creation of the conversation analysis model for mix-language settings, our research contributes to multilingual education focusing on Chinese heritage language and English in New York in the cultural school program context.