INTRODUCTION

When we talk about writing in Early Childhood Education, we do not always think about writing texts or even meaningful words such as children's names. In common sense and, unfortunately, in many educational institutions, writing is often associated either with tracing isolated letters and syllables or with copying words and even long headings (see, for example, data from research by ^{Silva, 2018} and ^{Souza, 2011}).

However, we understand that in this first stage of education, it is essential to promote experiences that lead children to reflect on why we write, what we can write, and how we can do it. This means that, alongside significant learning aimed at the appropriation of the alphabetic writing system, such as knowing and spelling some letters and words of interest to them or building the notion that we spell the sound segments of the words we pronounce, children can be stimulated to understand in which situations we write texts and for what purpose. Thus, from an early age, we consider it important to learn that we write something to be read by someone with the intention of thanking, complaining, asking, congratulating, warning, expressing feelings of joy or sadness, or even not forgetting something important, among so many other reasons.

In the beginning, when children start writing they can make repeated horizontal lines, zigzag lines, a sequence of polka dots, or other marks similar to letters. Later on, they learn to spell some conventional letters, usually starting with the letters of their names, that is, starting from a word with a high meaning, which makes its written form easy to memorize. In this process, children also come into contact with the socio-interactive dimension of language, thinking, and building hypotheses about the different situations in which writing makes sense.

In summary, despite the uninspiring picture revealed by the research mentioned above, we believe that in Early Childhood Education, it is possible to promote significant experiences with written language. By doing so, we will be contributing to the development of young children's desire to learn to read and write, something that we consider fundamental at this stage.

Since the 1980s, the Psychogenesis of Written Language has shown that very young children already reveal an active, curious, and reflective attitude towards written language. ^{Ferreiro (2001}, ²⁰¹¹) and ^{Ferreiro and Teberosky (1999)} indicated that children build and test hypotheses about writing, covering a long way of discovering this representation system. Theories and studies in the field of literacy have also indicated that very young children are already capable of expressing knowledge about different written materials, their purposes, their recipients, and the characteristics of discursive genres (see, for example, ^{Mayrink-Sabinson, 1998}; ^{Solé, 2004}). ^{Rego (1988}), when observing the path of a child in the process of building knowledge about writing from 4 to 7 years old, concluded that intense exposure to reading stories was a fundamental element for the child to appropriate the characteristic marks of written language, even before knowing how to conventionally write the texts he created on the oral level. ^{Terzi (1995}), in his research with children from “poorly literate backgrounds”, also highlighted the influence of participation in literacy events, which take place both in the family environment and in preschool, on the development of reading in the following years.

More recently, studies that focused on reading and writing practices in Early Childhood Education have deepened the relationships between children and their process of appropriation of written language and reflected on the role of the educational institution in learning related to this field. The article published by ^{Baptista et al. (2018}) shows part of the results of a research that aimed to map the state of knowledge about reading and writing in Early Childhood Education in Brazilian academic production between 1973 and 2013. The authors organized the results of their analyzes into three periods, highlighting trends of research on that topic in each of them. In the first period (from 1973 to 1989), studies were scarcer, and the works published, especially in the 1970s, were anchored in a compensatory conception of preschool education. In the second interstice analyzed by the authors (the 1990s), there was an increase in production and also an expansion of research topics, probably in an attempt to respond to the multiple demands resulting from a reconceptualization of the child as an active subject with rights. Finally, in the last analyzed period (2000 to 2013), the authors found a greater frequency and regularity in the publications. Therefore, we conclude that reading and writing practices developed with young children have become an object of interest for scientific research, and this has helped to build some references for pedagogical work. However, new questions have emerged, and, in this sense, we understand that the interactions between children in events of collective writing of texts represent a theme that deserves to be more discussed with the teachers and more focused in the scope of the research. Next, we will discuss some notes already available in the literature about this topic.

WHAT DO WE KNOW ABOUT TEXT PRODUCING WITH YOUNG CHILDREN?

In Early Childhood Education, there are numerous opportunities for children to experience meaningful writing situations. Thus, some authors (e.g., ^{Gerde, Bingham & Wasik, 2012}; ^{Quinn, Gerde & Bingham, 2016}; ^{Rowe & Flushman, 2013}) give important recommendations for developing good writing practices with children in the final years of kindergarten. Gerde et al. (2012), for example, recommend that writing opportunities be created in children's daily routine, so that writing is not something that only happens sporadically; they reinforce that all types of writing produced by children must be accepted by the teacher - from scribbles with drawings or shapes that only look like letters. Another recommendation from the authors is to always keep pencils and paper handy when organizing make-believe play environments, such as an office, a supermarket, or other settings where writing may be necessary. In this way, it is expected that children can integrate writing into the plots of their games and discover while playing, alternatives for their use.

In line with what we have already stated, one of the fundamental indications of ^{Gerde, Bingham, and Wasik (2012}) and ^{Rowe and Flushman (2013}) is to engage children in writing proposals motivated by authentic and socially significant reasons, signaling that we can communicate ideas and information through writing. According to the authors, children may be asked to write their name on the list of who wants to use the computer in the room, or on a work of art, they have done, or to write a thank-you card to the owner of the local bakery for the visit they have made to their workspace (Gerde Bingham & Wasik, 2012).

In addition to individual writing situations, ^{Gerde et al. (2012}) and ^{Rowe and Flushman (2013}) also recommend the participation of children in writing experiences in small groups, because,“In making writing public as part of group process, children can observe what other children are thinking. It also allows children to understand the importance that print plays in documenting what they said so they can revisit their ideas and reflect on their thoughts.” (^{Gerde, Bingham & Wasik, 2012}, p. 6).

Therefore, the authors emphasize the need to create situations in which the little ones “share the pen” and write with their handwriting, maintaining physical proximity between them, at the same time that there is close monitoring of the teacher, with invitations to write and interventions addressed to the interests and needs of each one during the writing process.

Other authors (^{Girão & Brandão, 2014}; ^{Leal, Brandão & Albuquerque, 2021}; ^{Lerner, 2002}; ^{Teberosky & Ribera, 2004}) have also valued the potential of collective writing when the teacher is the scribe of the group. In these cases, she assumes a triple role: she is a scribe, mediator, and co-author in her interaction with children; and, as Girão and Brandão (2014) point out, her role is not simply to write a text dictated by the little ones. She has the challenge of mediating the production of an oral text with a written destination that meets a clear and meaningful purpose for children.

Therefore, the teacher needs to make the writing process more transparent, which means saying aloud what she is thinking while writing. She can, for example, explain what she needs to clarify in the writing of a passage of the text or that it is necessary to give more information, or even that she is in doubt about what would be the best word to use. Of course, this entire intervention process also needs to take into account the purpose of the text, who it is intended for, as well as the characteristics of the discursive genre being produced (^{Schneuwly, 2004}).

With this proposal of collective production, the expectation is that children expand their knowledge about written language, especially about why and in what situations we write texts and about what aspects we need to think about when we express in this way.

Studies that investigated the processes and practices of textual production with children who still do not read and write conventionally are scarce in Brazilian and international literature (see the research by ^{Gerde & Bingham, 2012}; ^{Gerde, Bingham et al., 2015}). In Brazil, in a survey carried out on the website of the National Bank of Dissertations and Theses, it was found that in the last ten years only three works (^{Costa, 2012}; ^{Girão, 2011}; ^{Zapelini, 2016}) specifically discussed the topic of writing texts by children under 6 years old.

The study by ^{Costa (2012}) analyzed the records produced individually by children between 4 and 5 years old in situations in which the teacher requested the written production of a certain discursive genre (for example, an invitation, a message, or a list). ^{Zapelini's research (2016}) also analyzed the individual productions of children in the same age group, from the perspective of French Discourse Analysis. The children were asked to write on their initiative and, later, in situations proposed by the teacher. The research by ^{Girão (2011}), which also involved children between 4 and 5 years old, focused on the situation of collective production of texts mediated by the teacher. The author identified 11 different types of teaching intervention during the collective writing of texts to explain and reflect on the possibilities of mediation in this type of activity. The research data also indicated the existence of a relationship between the situation of textual production, the genre being produced, and the types and frequency of the teachers' interventions.¹For example, in the production of an invitation, ^{Girão (2011)} observed a greater frequency of interventions by the teacher that promoted children's reflection on the components of the textual genre, such as conversations about the event that motivated the writing of the invitation, the date planned to be held, the location of the event and who would be invited. The author interpreted that the emergence of interventions of this nature was facilitated, probably because, in the case of an invitation, these textual components are well marked, in the form of topics.

If studies that deal with the mediation of the teacher in the activity of producing collective texts in Early Childhood Education are rare, we did not find, in the survey carried out, research aimed at analyzing the forms of children's participation during this activity. Considering that the entire process of collective writing is permeated by discursive exchanges about what to say and how to say it, we were interested in “focusing on the children's words²”, as indicated in the article's title. We call discursive exchanges the interactions between children and between them and the teacher, in the work of elaboration and re-elaboration of the discourse that is constructed during the process of production of the written text. Thus, we start from the principle that when producing the text collectively, the group talks about ideas, ways of writing, the components of the genre, creates expectations about its interlocutors, checks if they are achieving their purpose, negotiates the use of terms, makes decisions, makes choices... in short, it engages in a network of orally shared utterances, situated in a certain socio-communicative context, which will have the written text as a product.

With this intention, we qualitatively analyzed the exchanges between the group during the collective construction of texts, seeking to answer the following questions: (1) how do young children participate in the process of collective writing of texts?; (2) what knowledge about written language is revealed by them during the discursive exchanges that emerge in the process of textual production? and (3) what conditions could favor such discursive exchanges in the collective writing activity?

To address these issues, we will initially make some more general theoretical considerations on the topic of text production and then present the research data. In this direction, we will identify and exemplify, at first, the different forms of children's participation evidenced during the collective writing process. Then, to deepen the analysis of the nature of the observed discursive exchanges and the knowledge shared between them about written language, we will discuss two situations of collective production in which boys and girls are more engaged in this activity.

REFLECTING ON THE PRODUCTION SITUATION OF A WRITTEN TEXT

In the model of textual production proposed by ^{Schneuwly (2004}), when the person is faced with the need to produce a text, he/she ends up adopting a certain genre that, for him/her, best suits the specific situation of social interaction on which he/she is acting. To make this choice, we rely on certain schemes that correspond to the construction of the orientation base and to the operations of text management and linearization.

The orientation base is a set of internal representations created by the subject from the context of interaction. These representations of the situation of social interaction take into account aspects such as the social place occupied by the person, the moment and place in which the text is produced, the objectives set for writing the text, and its relationship with the interlocutors. These representations, which guide the discursive action, can change according to the needs of the communicative situation and, in this way, control the other operations involved in the textual production, that is, the text management and the linearization process.

In ^{Schneuwly's (2004}) model of textual production, the textual management process occurs through anchoring and planning operations, which can be used simultaneously. Anchoring is the central and initial form of text management, in which the individual activates the representations that define what is going to be said and how it is going to be said. Through this operation, the individual anchors the writing activity in the set of constructed representations. Planning involves activating, organizing, and sequencing the contents and also their linguistic structuring, adapted to a language model or text plan, chosen according to social interaction.

Finally, the linearization operation concerns the processes of materialization of content into linguistic units, that is, referentialization and textualization. Referencing is about creating minimal initial language structures. This operation involves the choice of terms and does not allow the production of a statement directly, but a “scheme that generates statements” (^{Culioli, 1976}, p. 91). On the other hand, textualization acts at the local level of textual chaining and includes ways of establishing hierarchical articulations of the text, such as cohesion, connection/segmentation, and modalization.

^{Bronckart (1999}) highlighted the specificity of text writing situations. For the author, the conditions of textual production are defined by physical, social, and subjective parameters, such as the place and time in which the text is produced, the position of the author and his interlocutor, as well as the purpose of the text. Thus,

it is unlikely that there is an ideal and advisable “writing process”, and writing is conceived in a flexible, dynamic, and diversified way, depending on the different situations that give origin and meaning to the task of writing. This means understanding that the writing process, carried out by an author, is only interpretable in a given context. Each scenario draws a plot of specific conditions and suggests to the author a different way of acting (^{Castelló, 2002}, p. 149, [our translation]).

Based on these considerations, we can say that the collective writing of a text involves two situations of interaction that develop in parallel: the utterances produced by the group that is writing towards the recipient(s) of the text, in a movement that goes towards the outside the writing activity, and the utterances that are built internally in the interlocution that the group establishes among themselves to structure the discourse that will be recorded in the written plan. It is these two enunciative movements (external and internal) that we seek to present in the analysis of the situations of collective writing observed.

INTERACTIONS AS THE BASE OF APPROPRIATION OF CULTURE WRITTEN BY SMALL CHILDREN

As an artifact of culture, written language permeates social relations in a graph-centric society like ours. Therefore, the appropriation of this cultural object, in its discursive dimension, does not occur spontaneously, but through the interaction between individuals who share situations in which its use is necessary.

Considering that interactions constitute one of the structuring axes of pedagogical work in Early Childhood Education (Brasil, 2010), situations of collective writing of texts can present us with important elements about how children act and react in the face of the challenge of communicating with someone in writing. Having discursive exchanges as a keynote of the process of building a text, collective production can help us to understand aspects related to the perspective of children at the beginning of these learning processes.

Thinking about the relationship between children and written culture leads us to dialogue with Vygotsky in his description of the process of internalization of higher psychological functions. This process takes place from the interpersonal level to the intrapersonal level (^{Vygotsky, 2007}). This movement, which takes place first between people and, later, within the child, indicates that the “other” has a fundamental role in learning. In this context, by sharing the writing of a text to meet a real need with their peers and an adult, children have the opportunity to think about language with the mediation of the other. In this way, they can explain what they think about the different events mediated by writing, which representations they build about their interlocutors, and relate the context of writing with other situations experienced outside of school, among other exchanges that slide discursively and that go concretizing the authorship of the text produced by the group. From this perspective, the interactions that the group establishes are statements, and every statement is inherently dialogic since it is part of an uninterrupted discursive chain (^{Bakhtin, 1997}).

Based on the theoretical foundations presented so far, we are interested in better understanding how children react when immersed in the discursive exchanges that are established during the collective writing of texts. What aspects of written language catch your attention? What meanings do they attribute to writing when they are invited to write for someone? What aspects of mediation seem to help in the flow of interactions and the construction of a guideline for writing and the other operations involved in this process? Such issues will be discussed in the following sections, having as references both the model of textual production presented above and the specificities of pedagogical work with young children.

THE PROPOSED COLLECTIVE WRITING SITUATIONS AND THE PARTICIPATION OF CHILDREN

Children aged between 5 and 6 years old from two classes of the last year of Early Childhood Education at two institutions of the Municipal Network of Recife and their respective teachers participated in the research. Three collective productions, conducted by each group, were videotaped and transcribed. The present study sought to observe ethically informed research practices, dialoguing with the teachers and the children's families about the research procedures, as well as requesting permission for the subjects to participate through the Informed Consent Form (ICF).

Box 1 shows the three situations of textual production that were proposed by each of the educators (here called Laura and Cíntia), the duration of each textual production activity, and the number of participating children.

Source: The authors.

Box 1 - Collective writing proposals sent by the teachers³  

In the case of teacher Laura, before proposing the first activity (writing an instructional text), she made sock balls with the children and then suggested that the group write a text that teaches how to make the toy. The proposal had a clear purpose for the children; however, it was not defined who would be the interlocutors of the text produced. In teacher Cíntia's group, the first writing proposal was more associated with a communicative situation in the school sphere: rewriting of a fable known to children. The teacher also did not explain whether there would be other recipients for the text beside the children, who already knew the story well.

However, as we can see in Box 1, in the second and third activities, the teachers proposed writing situations that had a clearer socio-communicative context, which certainly made the writing activity more meaningful for the children, since there were a more defined purpose and recipients that were shared with the entire group. It is worth noting that even in the first writing proposals, the children from both groups were very interested in contributing to the production of the texts. This shows that the involvement of the little ones seems to be related not only to the immediate context in which the activities are carried out but also to the strategies used by the teachers to encourage their participation during the writing of the text and the activities that preceded this proposal (in this case, the making of the sock ball and the various moments of reading the fable, which was very much appreciated by the children).

Regarding the ways of organizing the group during the proposed writing activities, the two teachers followed different procedures. Teacher Laura wrote while standing on the whiteboard that was on the wall above the height of the children, who were seated on the floor or in small chairs, making a semicircle in front of the board. Teacher Cíntia, on two occasions, wrote on a large white paper, sitting on the floor and very close to the children, who were freely distributed around her. In the third writing activity, Cíntia wrote on a whiteboard positioned at the children's height, and everyone, including herself, sat in the small chairs next to the board. We observed that the organization proposed by Cíntia helped to create a relationship of greater proximity and interaction between the group, which could move more freely to point out letters and words recorded on the board or paper.

In the next section, we will deeply discuss the children's participation in collective writing activities, analyze their speech in the interactions established between the group (teacher and children), and explore their relationship with the knowledge about written language involved in the different situations of textual production.

MODES OF CHILDREN'S PARTICIPATION DURING THE JOINT PRODUCTION OF TEXTS

As we have already announced, in this article we will focus on the children's speeches, but they will be analyzed in the flow of the group's talks, having as a guideline the verbal interaction and its dialogical dimension (^{Bakhtin, 1997}). The comments of children and teachers were categorized from the perspective of Content Analysis (^{Bardin, 2002}); and the analysis performed here did not intend to quantify speech turns, but to approach them qualitatively. Therefore, in order to define the children's degree of participation during the activities, we considered the extension and the alternation of the shifts (longer or shorter shifts, predominance of the teacher's speech or of a specific child and greater alternation between the participants).

The analysis of children's participation in the six productions observed also pointed to some more general questions:

Considering the aspects mentioned above, we can say that the collective productions conductedby teacher Cíntia, in general, provided greater participation of the children, both in relation to the alternation of shifts and in terms of children's autonomy in speech and in the diversity of children who contributed to the writing. The data in Box 2 reinforce our assessment:

Source: the authors.

Box 2 - Children's participation during written productions 

Analyzing the interactions qualitatively, we can infer that some factors contributed to the more intense participation of the children in teacher Cíntia's group. They are: the spatial organization of children, as we mentioned earlier; the smaller number of children participating and the teacher’s mediation style (which tended to produce shorter and more reflective speech turns than Laura. She generally returned questions to the children and directed her questions to specific children, requesting individual participation). We believe that these factors helped to build a more open and favorable climate for children to express themselves. However, we noticed that these ways of children's participation also fluctuated in the three activities proposed by the same teacher, something that may be related to the situations of text production, as we will discuss below.

Considering the dimensions of the textual production process - the generation of textual content, the textualization, the compositional form of the discursive genre, the textual review, and the alphabetic writing system -, the observation of the six collective productions indicated that the children of the two groups interacted with writing exploring these different dimensions.

Next, we will present examples extracted from the videos in which it is possible to verify this participation of children. Finally, as we have already announced, we will deepen the discussion on discursive exchanges approaching two of the six collective productions analyzed.

CHILDREN IN THE COLLECTIVE WRITING PROCESS: ENGAGEMENT IN ACTIVITY AND SHARED KNOWLEDGE IN DISCURSIVE EXCHANGES

The data presented in the previous section provide evidence of children's participation in the different operations that involve the production of a written text. However, we observed that, in the last two activities conducted with Cíntia's group, the children, in general, were more engaged in the writing activity, producing a greater alternation in the discursive exchanges, signaling a movement of more autonomy in the flow of interactions. In other words, in the two writing activities that will be discussed here, the children were not limited to just answering the questions and responding to the teacher's commands. They suggested and discussed the ideas that were raised, made their points of view explicit, and made decisions more independently.

In addition to a meaningful writing context for the children (purpose, genre, and addressees clearly presented), a second explanatory hypothesis for this broader participation of the little ones is the mediation style of teacher Cíntia. As we mentioned before, she conducted the activities of written production in a more dialogic way, questioning the children and stimulating their reflection.

These two conditions provided a more favorable space for discursive exchanges to be established and the co-authorship process to take on greater prominence among the children. We observed that teacher Cíntia's mediation was characterized by the use of proposing and problematizing language, by the teacher's availability to listen to the children, and by the openness to collective decision-making.

Next, we will discuss the data from the two selected activities, showing how teacher mediation favored discursive exchanges and the construction of knowledge by children.

Discursive exchanges and collective decision-making

Producing a text requires making decisions throughout the writing process. A characteristic observed in the last two collective productions experienced by teacher Cíntia's group was the effective participation of children in the choices that gave materiality to the text. This aspect has already been evidenced in previous topics (when we talk about the discussion about textual content and the textualization process). However, we would like, once again, to highlight this “more horizontal” mode of interaction of Professor Cíntia, constantly inviting the group to think and decide together about what and how to write.

As we saw in Box 1, the second writing proposal launched by the educator was to write the recipe for “Joy Cookie” for the children to take the text home along with a bag of cookies. So, one day the teacher made the cookie with her class, and, the next day, the text was produced.

In the video in which the activity was recorded, the interest of the children was evident throughout the process, which involved a collaborative experience, mobilizing joint decision-making at various times, such as the name given to the cookie recipe, as we see in the dialogue below:

[...] Teacher - (...) yesterday we chose the name of the cookie.

Carla - It's happy cookie.

Teacher - Happy cookie? No...

Alisson - It's a cookie...

Joanna - Happy.

Teacher - It was.

Alisson - Happy Cookie.

Teacher - Cíntia asked for a suggestion to all the children in this room, see, Rafael? Didn't she, Alisson?

Alisson - Yes!

Teacher - I asked for a suggestion for the name of our cookie. Two suggestions came up: “happy cookie” and “joy cookie”. And then we voted, each child gave their opinion between “happy cookie” and “joy cookie”. Which onewon?

Juliana - Happy!

Calina - Joy!

Teacher - Joy! Joy cookie won. That's the name of our cookie! [...]

Besides the choice of the name for the cookies, the children also took part in other decisions: do you think it's important for us to put the recipe together with the cookie? We also highlight the moment when, at the end of the text, two girls suggest inserting a reminder for people to wash their hands and tie their hair up before preparing the cookies. Given this, the teacher launched the suggestion of the two for the group's approval and then continued in her role as scribe. Here's a snippet of that dialogue:

[...]

Carla - Because if you hold it with dirty hands...

Juliana - Hair will fall out...

Carla - Hair will fall into the food, and it will get dirty. And with dirty hands, the dough will get dirty.

Teacher - That's it! So, when we deliver this recipe, we'll remember to explain that we need to wash our hands well, right? And tie...?

Carla - The hair.

Teacher - Tie the hair up when making this recipe, because it can contaminate, which is dirty, contaminating is the same thing as dirtying...

Carla - Get the hair...

Luana - When the hair falls out it will get dirty. The food will get dirty. You have to wash your hands.

Teacher - That's it! Exactly. You have to wash your hands and tie your hair up, right? Do you want to put this here? Do you think it's important that we put here in our recipe that we should remember to wash our hands? Do you think it's important, Calina?

[Calina nods.]

Luana - Yes, if not, we will forget. [...]

Source: The authors.

Photo 1 - Final version of the recipe for joy cookie 

It is also worth mentioning a reflection made on a question addressed by a child, during the written production of this recipe. Faced with a question asked by teacher Cíntia, the child looked at the teacher thoughtfully and asked: how can we say to you? The child's question draws attention to at least two aspects. The first is the recognition that the teacher is an important interlocutor in this collective writing process. She is a reference, someone more experienced with written language, who is available to listen and build the text together with the group. The second aspect is related to the reflexive and interactive character of the collective production of texts, that is, a situation in which there is someone for whom we need to explain the thought and re-elaborate the discourse according to a diversity of ideas that are presented. We understand that providing such an experience to children is, without a doubt, something fundamental in their process of discovery in the functioning of written language. In this regard, ^{Bakhtin (2006}) reinforces that

this orientation of the word in terms of the interlocutor is very important. In reality, every word has two faces. It is determined both by the fact that it comes from someone and by the fact that it is directed towards someone. It is precisely the product of the interaction between the speaker and the listener. Every word is an expression for one in the other. Through the word, I define myself as the other, that is, ultimately, about collectivity. The word is a kind of bridge thrown between myself and others (p. 115).

Therefore, by getting involved in discursive exchanges with their peers and with the educator, the children in teacher Cíntia's class began to internalize the idea that the process of writing a text requires decision-making by its authors and that, in the specific case of the collective text, these decisions need to be explained and agreed among the participants of the group.

Finally, the analysis presented here converges with the study carried out by ^{Pascucci and Rossi (2005}) about the role of the teacher, who goes beyond being a scribe, acting as a facilitator of children's discursive production and assuming the function of a reader to the extent that she asks the children to explain the information needed to understand the text. We can say, then, that the teacher acts as a link between what we call the internal and external movement of utterances during the collective production of texts. In the same way, we observed in the present study what Pascucci and Rossi highlight as children's agency. This means that each child is an “responsible agent of his/her own words, who participates in a human technology - language as discourse - within a small community of meanings, in which the teacher also participates” (^{Pascucci & Rossi, 2005}, p. 185).

FINAL CONSIDERATIONS

The examples presented and discussed here show that significant situations of collective production of texts in Early Childhood Education institutions can be good opportunities for children to build knowledge about different aspects of written culture.

The study data also show us that children can bring their previous experiences with the discursive genres produced, expose ideas, and agree and disagree on their initiative, which signals an active posture ⁵during the activities of collective production of texts analyzed here. As in life, we understand that knowledge about written production is mobilized by children in the educational institution in the context of its uses. In this way, the different dimensions of the text construction process, discussed above, were the children's learning objects as they interacted and shared the challenge of communicating with someone through writing. Therefore, it is not about “teaching a class on text production in Early Childhood Education”, presenting, for example, the characteristics of different discursive genres. On the contrary, the idea is to offer the little ones interesting and meaningful opportunities for them to express themselves in writing and have contact with the reading and production of different texts.

As we have already stated, some research has revealed the absence of proposals for the production of texts in Early Childhood Education (see, for example, ^{Cabral, 2013}; ^{Gerde & Bingham, 2012}; ^{Silva, 2018}; ^{Souza, 2011}). Perhaps this represents certain insecurity by the teachers to propose writing activities for children who have not yet been literate. This study highlightsthe mistake of this idea and presents elements so that we can better understand how the little ones participate in this type of proposal, helping to reflect on the possibilities of mediation between them and the written language.

Teacher Cíntia, especially in two of the three text production activities she proposed, showed that a more dialogic posture can favor more effective participation of children in the writing process. In addition, once reaching this higher level of participation, which we call here “engagement”, we saw that children can get involved in the co-authorship of texts, even contributing to the development of more complex operations related to the writing activity, such as the discussion about the content and the textualization.

The teacher's availability to listen to the children, welcome their ideas, problematize them and incorporate them into the text, encouraging the expression and listening of different opinions, and developing the skills to negotiate and make decisions, were also actions that proved to be essential for them to recognize each other in the text construction process.

During the collective production of texts, although the written notation may draw the children's attention, the recording of the text is carried out by the teacher. With this, it is expected that they will focus more on other dimensions of written culture, such as sociodiscursive aspects. Thus, both in situations in which children write by hand in the way they know and in those of collective writing, we found that the little ones build representations about the activity and seek to mobilize their knowledge to express themselves in writing, as we saw in our research.

We also saw that mediating the collective writing of a text with young children presents great challenges, such as guaranteeing everyone the space for speech and observing the reactions and engagement of children in the activity. In this sense, we can question the pertinence of forwarding collective writing with all the children in the class at the same time. Analyzing the six proposed activities and specifically observing the mediation of teacher Cíntia, who had fewer children in her group, we saw that she made use of interventions closer to each one. Thus, we ask ourselves: would the collective textual production conducted with a small group, which could be more interested in that writing situation, favor a more qualified mediation, also allowing a greater engagement of children? The data in Box 2 point in this direction, indicating a tendency towards an increase in participation when there are fewer children involved in the collective writing activity.

The studies mentioned here also converge on the idea that the production conditions interfere in the writing of texts and that “the other”, be it the mediator adult⁶ or the other interlocutors, “has a fundamental place in the children’s writing process, or that is, a huge influence on his saying” (^{Costa, 2013}, p. 2012). We agree with this author when she states that skills related to the conventional notation of writing “are not requirements for the production of texts, but it is through the production of texts that children understand written language in its entirety: as a form and as a discourse”. (^{Costa, 2013}, p. 31).

In summary, the research data show that discursive exchanges in a context in which the group effectively engages in the production of the text can create favorable conditions for the construction of knowledge related to writing, which constitutes an important basis in the process of developing readers and writers, which begins in Early Childhood Education. In this way, the collective production of texts can be a moment of encounter with the other in a significant experience of learning written language.

In conclusion, the analysis of the interactions between children and teachers when they collectively produce texts also shows that building a text from the relationship with the other is to assume the principle of collectivity, which presupposes the exercise of listening, respect, and acceptance, dimensions fundamentals of human formation.