Introduction

Research on literacy at school has expanded over the past few decades, when the word and the concept of literacy were inserted in the Brazilian academic field. Literacy studies have been influenced by the anthropological perspective that emerged in the mid-1980s, with the seminal research of Shirley ^{Heath (1983}), Ways With Words, in the United States, and of Brian ^{Street, Literacy in Theory and Practice (1984}), in England, among others. For this perspective, literacy is a social practice located in the use of writing, permeated by power relations and occurs in multiple social spaces, not just at school. Thus, it is most appropriate to consider it as plural - literacies.

Within the Research Group in literacy, Language and Decoloniality (GPEALE), we have been conducting researches since the early 2000s looking for an understanding of literacy focusing on reading and writing events and practices in elementary school classes. Over these nearly 20 years, we have been deepening the understanding of school literacy processes and focusing on practices and events mediated by public school teachers in four Brazilian cities: Belo Horizonte, São João del-Rei, Tiradentes and Recife.

This experience has instigated us to reflect on the theoretical and methodological issues that permeate research on the uses of reading and writing at school, indicating the need for articulation of the New Literacy Studies (NLS) with other references that enable the expansion of the epistemological lens to observe literacy at school. Thus, in addition to the NLSs, Bakhtin’s analysis of language have been present since the beginning, and Paulo Freire’s considerations on literacy and pedagogy have been incorporated more recently.

Therefore, this article aimed at contributing to the reflection on a theoretical-methodological perspective for the study of literacy in school that goes beyond the NLS, based on the assumption that the complex reality of schools and classrooms cannot be fully understood by a single research approach. Bakhtin’s studies, for example, have been incorporated over the past decades not only by NLS researchers in Brazil, such as Angela ^{Kleiman (1995}), but also abroad, including James Paul ^{Gee (1991}) and David ^{Barton (1994}).

We organized the text in three parts. Firstly, we present a discussion on the epistemological perspective constructed by GPEALE. Secondly, we describe and discuss the methodological design used in the analysis of the implementation part of a recently completed research project. Finally, we elaborate some reflections that aiming to contribute to the deepening of theoretical and methodological issues involving research on literacy in school.

An Epistemological Perspective for Literacy Studies

The NLSs inaugurated an approach to research on the uses of writing in society by breaking with the dichotomous and ethnocentric vision that separates literate subjects from non-literate, orality and writing, illiterate and literate, indicating a deeper understanding of the complexity of the social practices of reading and writing. Such a view stems from questioning the hegemonic conception of literacy as the mastery of a set of reading and writing skills that establishes a judgment of exclusive value, inferiorizing people that do not fit the standard of literacy that is dominant in the graphocentric society, and imposed by organizations such as UNESCO, which formulates literacy projects for underdeveloped and developing countries. Therefore, the main contribution of the NLS, lies in deconstructing this dichotomous vision, introducing a perspective of continuum in the relationship between orality and writing, an understanding of the plurality of practices in the use of writing in society, problematizing different aspects involving them. In contrast to this view, ^{Graff (1994}, p. 37), historian of the written culture, contributes to the NLS stating that

What needs to be learned is that the oral and the written [language], such as the manuscript and the printed [document], do not need to be put as opposed options. Human history and developments do not occur that way. Instead, they allowed the occurrence of a rich and profound process of reciprocal interaction and conditioning as literacy gained acceptance and influence [...] for many centuries, reading itself was an oral activity, often collective and not the private, silent activity that we currently know.

At the core of this perspective, lies the notion that literacy is much more an ideologically marked situated social practice, materialized by power relations (^{Street, 1984}), than a set of technical skills to be learned at school, as most of the curricular policies for literacy around the world indicate (^{Papen, 2016}). When questioning the studies of the Great Divide (^{Ong, 1982}; ^{Goody, 1977}), Street (1984) formulated the model of ideological literacy for the analysis of the plurality of reading and writing practices that occur in society in the midst of power relations. Such plurality was already identified in the first NLS studies, when Street (1984) ethnographically described three literacy practices observed in a village in Iran, just as ^{Heath (1983}) observed different practices in three communities in the United States. These ethnographies allow authors to oppose a restricted view of writing as a neutral technique, an ability of encoding and decoding that is learned and used in the same way regardless of the social and cultural context, which was named by Street as model of autonomous literacy.

The ideological literacy model includes two concepts as important tools in the analysis and understanding of literacy in different social spaces: the concepts of events and literacy practices. Coined by ^{Heath (1983}), a literacy event is defined as any interaction situation structured by writing. In this sense, the interdependence between oral and written languages in literacy events, which is perceptible in Heath’s studies, constitutes one of the focuses of our analyzes, as the interactions in the teaching and learning relationships include oral and written (printed or digital) languages, both in reading practices and the production of texts.

However, the NLS warn that it is essential to go beyond the mere description of events, as it alone does not allow us to understand the complexity of situations involving reading and writing. It is necessary to infer which is the use of writing patterns present in the events in order to understand how the different literacy practices are established in society. It is also necessary to infer the inherent values, the power relations marking interactions, as well as the broader institutional aspects that take part in interactions with writing. According to ^{Barton (1994}, p. 7) literacy practices refer to “[...] general cultural ways of using the reading and writing that people produce in a literacy event”. Accordingly, ^{Street (1997}, p. 50) argues that “This term enable us to specify the particularity of cultural practices with which uses of reading or writing are associated in given contexts”.

What people do with reading and writing in everyday life is based on patterns of use historically created and recreated by subjects and institutions. Therefore, literacy events are not built; they are linked to different socially established practices, but this does not mean they are immutable. For instance, the contemporary characteristics of school literacy practices are not the same as those of the beginning of the 20th century, although we can find some similarities. School materials have changed in form and content, teachers are trained based on other pedagogical references as so have the curriculum and the formulation of educational policies. Written culture in society has radically changed, as well as subjects currently attending public schools. Thus, events are observable situations of writing use that refer to different social practices, which depend on the institution, the subjects interacting in it, the gender in use, and the historical moment when they occur, among other factors. Therefore, according to the reflections of ^{Brandt and Clinton (2002}), we consider as fundamental a methodological design that allows establishing a more consistent relationship between local and global, micro and macro dimensions involved in any literacy event.

Combined with the vision of literacy as a social practice, we add the concept of critical literacy resulting from Paulo Freire’s reflections on literacy and education (^{Bartlett; Macedo, 2015}). This thinker is responsible for introducing the educational principle of critical understanding of the reality in which we are inserted as a condition for the educational process. According to Freire, literacy must be based on critical reading of the world with the view of education for social transformation (^{Freire; Macedo, 1990}). Being situated in the third-world historical context, Freire also sought to pay attention to the political struggles for national freedom in the Americas, Africa and Asia: from Mao Tse-Tung in China to Ho Chi Minh in Vietnam, and Amílcar Cabral and Agostinho Neto in Africa’s process of Portuguese decolonization. Freire suffered a great influence from these theorists, who combined the nationalist theoretical formulation with a Marxism arising from the peripheries in strategically anti-imperialist struggle all over the world.

More specifically, Freire also assumes the perspective of the decolonization of thought, so well explained by Frantz ^{Fanon (1968}) in his book Condenados da Terra, one of the basic works of Pedagogy of the Oppressed (1974). For Freire, literacy is a political act and not the result of a process of repeating the word of the other - the colonizer - but the possibility of having his/her own voice, of humanization, of being the author of his/her own history. Decolonization is a strategy for questioning social inequality that produces exclusion and is fundamental to the consolidation of a fair, egalitarian and fraternal civilization.

Freire’s epistemological perspective points to literacy as a political action whose process is based on the pedagogy of questioning and dialogue (Freire; Faundez, 1985) towards a critical and transformative education, an education for being more. In all of his books, dialogue appears as a central category when discussing pedagogy and literacy⁵. For Freire,

If it is by saying the word ‘pronouncing’ the world that men transform it, dialogue imposes itself as a way men gain meaning as men. For this reason, dialogue is an existential requirement. And, if dialogue is the encounter in which reflection and action of subjects are addressed to transform and humanize the world, it can neither be reduced to an act of depositing ideas of one subject to another, nor becoming simple exchange of ideas to be consumed by the exchangers (^{Freire, 1974}, p. 78-79).

Therefore, the liberating education that Freire advocates is essentially dialogical and needs to be part of the teacher-student literacy relationship. It is an education that confronts the pedagogy of ready answers. Faundez and Freire (1985, p. 46) argue that what “[...] the teacher should first of all teach - because he himself should know it - is how to ask. For the beginning of knowledge, I repeat, is asking. And it is only through questions that one must go serching for answers, not the contrary”.

We observe a clear point of contact not only with the NLS, as already discussed, but also with Bakhtin’s dialogical/enunciative perspective on language, to whom language is a living process of verbal interaction. Working with an enunciative and/or discursive conception of language means, according to Bakhtinian thought, to conceive linguistic interaction as a result of the socio-historical context and the immediate communication situation in which the interlocutors are inserted. Thus, the meanings of concrete statements constructed in literacy events are results of these conditions of production, as discussed by ^{Kleiman (1995}).

Bakhtinian dialogism is based on a conception of language as a social practice, ideologically marked, built on the process of verbal interaction between socially organized subjects. According to the author, dialogue is the fundamental feature of enunciation and goes beyond face-to-face interaction. “Dialogue can be understood in a broad sense, i.e., not just as face-to-face communication, but any verbal communication of any kind” (1995, p. 123). Bakhtin (1995) states that “[...] every statement, even in the immobilized form of writing, is a response to something and is constructed as such (p. 98)”, thus, “[...] the dialogical relationship is a (meaning) relationship that is established between statements of verbal communication”, (1997, p. 345), being “[...] the meaning of the word totally determined by its context” (1995, p. 106). In the words of the author,

[...] the book, i.e, the act of printed speech, is also an element of verbal communication. It is object of active discussions in the form of dialogue and moreover, it is made to be actively apprehended, to be studied in depth, commented on and criticized (1995, p. 123).

This conception of dialogue and the book as an act of printed speech is vital for the reflection and understanding of literacy practices at school, since much of the material circulating in classes is printed. Based on this assumption, we can understand literacy events as marked by the dialogical relationship between students and the genres that they read and write with the teacher’s mediation. For ^{Bakhtin (1997}), understanding “[...] implies responsiveness and therefore, a value judgment” (p. 351). Thus, the responsive understanding of a genre is always dialogical. Accordingly, the language that permeates the educational processes needs to be dialogical, instigating counter words, producing critical sense and authorship, enabling students to build their own voice.

The GPEALE Methodological Operation

In this section, we discuss aspects or various facets involving the theoretical-methodological perspective that we have constructed and reconstructed over the last 15 years. Such an approach is inspired by the ethnography derived from NLS, which adopts an anthropological research perspective (^{Street, 1984}; ²⁰¹⁰; ²⁰¹⁴; ^{Barton; Hamilton; Ivanic, Roz 2000}; ^{Heath; Street, 2008}, among others), articulating it to critical conception of pedagogy and literacy of Paulo Freire and the conception of language derived from Bakhtin’s studies. We consider that this articulation allows us a more comprehensive attitude towards the teacher’s work and the complexity of literacy practices in the classroom, by identifying the principles that organize the actions of the subjects in the construction of literacy events, thus producing a written culture in/of the classroom. We start from the assumption that the answers to the question of reading and writing schooling do not come only from quantitative comparisons among schools, but fundamentally “[...] from the examination of the living processes of education that occur within classrooms” (^{Mehan, 1979}, p. 8), through the analysis of events specific to this social space, such as school literacy events. However, ^{Rockwell (1985}, p. 15) warns us that

What is important is to interpret the studied phenomenon from its relations with the broader social context and not only in terms of its internal relations. Methodologically, this implies complementing field information with information from other social orders (for example, the country’s educational structure and policy) and, on the other hand, seeking interpretations and explanations of elements that are external to the particular situations observed.

Reaffirming literacy as a multiple and plural process, ^{Rogers (2001}, p. 207-208) highlights the ethnographic perspective as the methodological proposition that best allows this type of approach and understanding of the phenomenon:

Ethnographic studies have thus revealed that instead of one universal literacy (a set of skills), there are different forms of literacy, different literacies [...] in place of a static definition of literacy which applies to all persons for all time (the skill of decoding words from texts), we are faced with a changing scenario in which people are both subjects of change and objects of change.

Thus, the methodological design we have built and experimented at GPEALE has as principle analyzing and understanding literacy practices at school, producing research tools that contribute to establishing relationships among classroom data, the school space and the broader context.

In this work we present as example the description and analysis of the implementation of a recently concluded research whose focus the literacy practices of a public school in the Municipal Network of Recife, state of Pernambuco, Brazil. We aimed at mapping, analyzing and understanding such practices in classrooms from the 1st to the 5th year. We have a team of teachers and students from the Post-Graduate Program in Education of the Federal University of Pernambuco (UFPE), which, under our guidance, can implement such a project.

Deciding to investigate just one school for a month was a great challenge for the research group. How to find a school that would accept such a research model, as it implied the presence of five researchers every day in classrooms? Why research just one school instead of different schools? We considered that by investigating the first years of only one school and producing data in a given time period, we could build a more expanded vision on literacy practices, thus deepening the knowledge on the literacy phenomenon in the school. Until then the GPEALE group has experienced research focusing the initial years in different schools of distinct knowledge networks. Such focus hindered, in a way, constructing an overview of the literacy practices, along with identifying each classroom patterns and the role that reading and writing occupy during the initial years of just one school.

Defining the Locus of Research

The first step was to find a school that would accept the challenge of partnering the research project. We resorted to the indications of the Recife municipal education department, as we needed schools close to UFPE to facilitate the travel of researchers. The team visited the first school, exposed the research project to the teachers and the board, enabling dialogue about any doubts or questions on the project. We attribute their refusal to the finding that some previous researches in education entailed little dialogue with the school. As the teachers said, “We open the classroom, but the researcher doesn’t come back, doesn’t even show up to thank us!”, “A girl came to attend my classes, and even my own activities I gave in to her, but I was not invited to the final presentation of her work. She’s gone! Where’s the dissertation, I await until today!”, “The people from the university came here, did their work, but we had no return!”. These statements show how much the research processes need to be reviewed with regard to the need of establishing a partnership relationship with the school subjects (^{Macedo, 1998}).

We sought a second school, as indicated by a researcher from the group who knew the pedagogical coordinator for the afternoon shift. We carried out the same process. Many questions were asked, mainly about the permanence of the researchers in classrooms and how the teachers’ practice would be presented in the academic texts to be produced. We affirmed our commitment to establish a partnership relationship, making research data available and presenting partial results throughout the process. We made it clear that our objective was not to establish a positive or negative evaluation of teaching practice, but to seek a deeper understanding of literacy processes, emphasizing the importance for the academic field of being able to implement this methodological design. After the meeting, we sent a copy of the project that would be read by all teachers to support the decision to participate in the research.

We assume that the acceptance from this school was due to the positive posture of its management and coordination in understanding our need as researchers and also in the persuasive way of the GPEALE group when presenting the research project. We acknowledged that the school understood that the presence of the university could be positive and bring benefits to it with regard to the possibilities of reflection on the teaching practice and the processes that involve reading and writing in the early years.

The Field Research

In order to prepare ourselves for an organized and cohesive work, we defined how to get to the field of research according to the availability of each researcher and in agreement with the teachers of the classes that would be investigated. We also decided to discuss ways of observing the classrooms from each other’s experience and texts on ethnography in education already read and discussed in the group, such as Ethnography as a Research Logic, by ^{Green, Dixon and Zaharlic (2001}), among others. We dealt with the importance of reflecting on the logic that guides the investigation towards an explanation of the relationship between the research questions and objectives, the data production process and analysis. Ethnographic research is not a linear process in which all decisions about the study are made a priori. It is an iterative process, through which questions are generated, revised and refined, (^{Heath; Street, 2008}) and, in this process, the data production is oriented and reoriented; thus, new questions may arise, and new research procedures can be incorporated. Decisions about fieldwork were made before, during and after the completion of the data production process, since the ethnographic research necessarily implies changes throughout the process.

Knowing the importance of observing the first day of school (^{Macedo, 2005}), all the researchers were present at the school for the opening of the school year. We also decided to daily monitor the first month to familiarize ourselves with the routine of the classes, thus capturing the regularities of the actions, with the aim of, from the second month on, reducing the intensity of observations to twice or three times in the week. Another important decision was to observe all classes and not just those that teachers normally call Portuguese classes or literacy classes, to diversify data sources and, thus, produce a broader and holistic view of literacy practices. The underlying assumption is that literacy practices at school are built on any discipline in the curriculum, and not just in “[...] Portuguese language or literacy classes”. This assumption also advocated by ^{Papen (2016}) in the ethnography he recently produced most on the first school years in the UK. Our data, for example, evidenced the presence of text reading events involving the most diverse knowledge in all classes, including specific knowledge of the writing system.

From the beginning, we knew the relevance of producing data from different sources, both in relation to classroom observation and official documents of Recife’s municipal network policy, the monitoring of public literacy and respective policies within the scope of the Brazilian Ministry of Education (MEC), the need to interview teachers, among others. Thus, we: produced field notes for all classrooms observed; interviewed all teachers once or twice, depending on the need to clarify some element of practice, which was not understood only by the field notes. According to ^{Bourdieu (1997}), with regard to scientific interrogation in research interview, however much one tries to avoid symbolic violence, all kinds of distortion crosses the research relations. Therefore, it is necessary paying attention to such distortions and trying to dominate them (being impossible to totally annul them), through reflexive reflexivity, “[...] endeavoring to make reflexive use of the knowledge acquired from social sciences to control the effects of the research itself” (p. 694). We understand that this alert extends to the entire qualitative research process that entails direct contact between researcher and subject, such as classroom observation.

It is important to highlight that many of the classes were recorded in audio, which helped the systematization of field notes and the iteration that is proper to ethnographic work. In only one of the classes, this type of data could not be produced, as it was not authorized by the teacher, showing that the data production process involves tensions and does not always occur as expected. The audio data also allowed a more in-depth approach to interactions in the classroom, approaching the pedagogical language built in this interactional space. Despite the countless benefits, we cannot fail to consider that the data produced in audio, like any research data, is selective. Behind the use of audio recordings there are the interests, questions and objectives of the researcher that condition the production of the data, defining what will be recorded. Therefore, the data production process, whether in audio or by means of other instruments, is not neutral, but is marked by the theoretical and methodological point of view of the researcher. In addition, the printed material used in classes (such as individual activities, textbooks, student notebooks) was also an object of interest.

In addition to the observations of classes, documents and interviews, we decided to apply a questionnaire to all teachers from the first to the fifth year of the Municipal Network of Recife in order to map the socio-cultural profile and learn about some trends in relation to their cultural practices such as reading, use of the internet, cinema and other cultural events present in the contemporary scene of Recife, as well as observing aspects related to professional life, such as teaching time, academic training, most used teaching materials, among others. These data are already published (^{Macedo, 2019}) and were decisive for the more global understanding of the research subjects, understanding them from the broader context in which they are inserted. The articulation between quantitative data and qualitative data was also present in previous research by GPEALE, as can be observed in Macedo (2004) and ^{Almeida (2012}).

The data production process through observations occurred throughout the 2016 school year, with the exception of the third year class, due to the teacher’s absence in the third month. In addition, observations were interrupted for a period of about 20 days due to a strike by teachers for better working conditions. These are unforeseen situations in the planning that end up interfering in the dynamics of field work, but take part of the process and results of the investigation.

During the observations, another unforeseen situation had to be considered. The research team begun to realize the importance of the reading room in the development of literacy practices in the school, since, weekly, the classes had activities in this space, in a partnership between the class teacher and the reading room teacher, a very common function in other schools at the municipal network of Recife. Thus, a redefinition of the research was necessary to incorporate the reading room as a locus of investigation.

In addition to the observation and recording of classes, the team met weekly to discuss data, make decisions and share experiences. It was a collaborative and dialogical process in facing the demands of research, a challenge of participating in the process of each class without assuming the teaching function, as some teachers asked us for.

The Treatment of the Data Set

The analysis of the questionnaire⁶ applied to all teachers in the early years of the Municipal Network of Recife sought to identify trends in the profile of the teachers, including the level of schooling of the family and themselves, family income, preferred cultural activities, the relationship with reading of varying genres and with the internet, time in the profession, number of shifts and schools in which they work, participation in continuing education processes, among others. This process of analysis was carried out concurrently with work with classroom data and interviews.

As for the data of an ethnographic nature, each researcher made a general mapping that was discussed at the research group meetings in order to select the most significant ones in the representation of patterns and routines constructed in each classroom. Once we chose the classrooms, the next step was mapping the literacy events of each, focusing on situations involving reading, writing, analysis of the writing system, in any of the observed classes (Portuguese, mathematics, science, history, geography, etc.).

In possession of the event maps, each researcher selected the most significant ones to be analyzed in depth. The description of the events included not only a narrative of what happened, but the researchers’ inferences based on the theories and concepts that support the research with a view to producing an understanding of literacy at school and its various aspects, and the power relations involved in the teacher-students relationship. In some cases, it was possible for us to move towards mapping discursive sequences (^{Macedo, 2005}; ^{Macedo; Mortimer; Green, 2004}), which were important tools for a better understanding of the interactions between teacher and students. The event map combined with the use of the discursive sequence, allowed us to show how the interaction between students and teacher was built during literacy events, the recurring interactional patterns in the classroom and how the time was spent in carrying out such events. Some of these results are already published in the thesis of Magda ^{Dezotti (2019}) and in the dissertation of Érica F. ^{Lima (2017}).

We must emphasize that the work of transcribing the audios in the classroom was thorough and challenging due to the simultaneity of actions taking place there. As is known, every transcription requires choices that reflect the researcher’s point of view; therefore, it is not an impartial process. It reflects the conditions of production of the research, i.e, the aims, the theoretical-methodological references that support it and the researcher’s beliefs about the participating subjects. Therefore, the transcription of a text always involves the transcription of a context as indicated by ^{Bucholtz (1999}).

The appreciation of context in literacy practices and events is also clear in ^{Freire’s contributions (2011}, p. 43) when referring to reading and writing. According to this author, the context is decisive to understand the writing practices, as he thinks of literacy as an act of knowledge, creation and a political act that demands effort to read the world and the word. Freire also states that “[...] currently, text without context is no longer possible”. In the same perspective, ^{Bakhtin (1995}) draws attention to the social character of language and discourse, which is not limited to an abstract system of norms, but constitutes the participation of subjects in the verbal interaction chain.

Thus, the transcription and organization of classroom data was performed at three levels. The first refers to the mapping of all classes observed, with brief descriptions of the literacy events that occurred in each class, as indicated above. The second refers to the selection of classrooms to be analyzed and the elaboration of event maps. The third refers to the identification of discursive sequences.

In view of the huge volume of data, we decided that from the most general reading of each class, we would choose a focus of analysis according to the recurrence patterns identified, in order to highlight the characteristics of the literacy practices prevalent in each classroom. Thus, in the first and second year classrooms, our analysis focused on events related to the appropriation of the alphabetical writing system, a recurring aspect in both classes, since the children are in the process of literacy. In the third year classroom, we chose literacy events focused on the production of texts (^{Lima, 2017}) because we considered that the teacher’s practice showed innovative aspects in the treatment of this writing process. In all the events analyzed, it was possible to observe the focus on the conditions of production of the written text, especially the definition of objectives, interlocutors and the specificity of the genre, indicating an approximation with the concept of language as a social and discursive practice. In the fourth year (^{Araújo, 2017}), the predominant focus was on reading events, understood as a practice that goes through most school events in all areas of knowledge. In the fifth year (^{Dezotti, 2019}), the focus was on literacy events mediated by literary texts, and we observed that this practice was recurrent in the class and in the library space.

We defined as a data analysis principle the contrast among classrooms, classes, literacy events and discursive sequences. When we contrasted the five classrooms, we observed the specifics as well as the similarities among literate practices built there. Through contrast, we were able to identify the patterns of interaction mediated by writing that recurred among classes and in each one of them. Contrast also allowed us to observe and understand the focus of the teaching action and realize how reading and text production only appears in the classroom more intensely from the third year on, probably due to the belief that children from the first and second years who have not yet mastered the writing system cannot write texts or read individually. By contrast, we were able to highlight a recurring pattern in all classes: all reading and text production events were held collectively. The reading and writing of individual texts did not occur in the observed classes. The immersion in the field, with daily observations in the first month and on average three times a week from the second month on made it possible to verify that the teaching of the writing system, in the first two years, occurred in all classrooms of all disciplines and not only when working with literacy recognized by teachers.

To what extent did the school participate in the data selection and analysis process? From the beginning, we proposed teachers to attend our meetings at the university and have access to the data being produced. However, the day-to-day teaching work does not allowed it to happen. There were several difficulties for teachers to leave the school, even for a day, to reflect on their practices, either because of a lack of teachers or they have to work in the school. In addition, teachers participate monthly in training meetings offered by the Paulo Freire Training School, of the Municipal Secretariat, which makes it difficult for them to be released on a different day. On the other hand, although in disagreement with what was planned, the dialogue about the data occurred at the time of the interviews when the researcher dealt with observed situations, in informal conversations during the observations, in which the teacher used to ask the researcher’s opinion on some subject or question dealt with at that moment, and after the conclusion of the first version of the thesis or dissertation, when the researchers returned the analysis to the group of school teachers before the end of their work. Finally, a return of the questionnaire data was carried out in a lecture for teachers from the municipal network who participated at least in one of the meetings of that training center. The published article was later forwarded to all teachers of the network.

Conclusions

In this text we present a theoretical-methodological perspective for the study of literacy in school that was based on NLS, Paulo Freire’s Pedagogy and Bakhtin’s enunciation theory. The articulation of the three epistemologies in the construction of a lens to observe the phenomenon of literacy starts from the assumption that only the concept of ideological literacy, coined by ^{Street (1984}), events (^{Heath, 1983}) and literacy practices (Street, 1997; ^{Barton, 1994}), are not sufficient for the analysis of the complexity of the literacy phenomenon at school. For this reason, we based our research on the concept of language as a phenomenon that occurs in verbal interaction, coined by Bakhtin, and the concept of literacy as a political act, presented by Freire from the beginning.

Taking literacy at school as a socially situated practice and built on the interaction between student and teacher subjects, we consider fundamental the use of a research methodology that allows immersion in the context of the school and the classroom, with an extended presence in classes of all disciplines since reading and writing are practices that cross the entire curriculum in action. Some of the research findings were only possible because of this format, such as the perception of the importance of the reading room in the training of students, the strong presence of children’s literature in the students ‘daily lives, and the finding that most of the students’ time in the first two years in the school are dedicated to teaching the writing system. We also include the perception that reading is performed in predominantly collective situations, with a significant absence of individual and silent reading activities, among others. Thus, it was possible to observe the multiplicity of literacy events in elementary school, reaffirming the plurality of uses and functions that reading and writing assume in society and at school.

Such format also allowed us to understand more deeply the trends in the profile of teachers of the researched school in relation to the more general profile of teachers in the early years of Elementary Municipal Network of Recife and to understand why literary reading is part of the school’s daily life. It is not only because there are public reading policies that have affected the school in recent decades, but because most of the network’s teachers identify themselves as readers and subjects who participate in various cultural activities in the city, in addition to attending continuing education actions. Finally, we hope that this text will contribute to a more in-depth theoretical and methodological reflection on literacy studies at school.