Examining temporality of language teacher identity through epistemic network analysis: the case of two EFL educators in Costa Rica

Teacher identity is understood as a dynamic process that characterizes how teachers see themselves in relation to their professional and sociocultural contexts across time and space (Varghese, Morgan, Johnston, & Johnson, 2005; Norton, 2016; Barkhuizen, 2017). Conflict and contradictions are inherent to identity formation. Research identifies internal identity tensions as those emerging from the divergence between who the teachers are, what they know and feel and who they want to be, also called performed versus imagined identities (Fairley, 2020). External identity tensions are provoked by the dissonance between teachers' desires, values or perspectives and those of their programs, institutions or other stakeholders (Varghese et al., 2005). In sum, identity tensions reflect the interaction of the social and personal sides of language teacher identity (Wenger, 1998).

One source of identity tensions for language teachers, particularly in EFL settings, is their deficit positioning as NNESTs. Extant literature shows that they have been viewed as defective communicators (Gu & Benson, 2015), perpetual language learners (e.g. Huang, 2018) or having inferior linguistic and professional competence (Braine, 2010). These inaccurate assumptions contribute to the (self) marginalization of NNEST and exacerbate feelings of self-doubt and inadequacy that lead to a sense of illegitimacy as speakers and professionals (Moussu & Llurda, 2008). Research has documented how these and related ideologies have shaped the experiences of in-service and pre-service teachers in the Global South, including the context of Costa Rica.

Solano-Campos (2014) conducted an autoethnographic study delving into the socialization processes that influenced her involvement in an international recruitment opportunity in the USA. Through her personal narratives, she explored her upbringing as a child and as an EFL pre-service teacher in Costa Rica, where she consistently received messages emphasizing the importance of English for success. Solano-Campos reflected on the pervasive discourse regarding English as a global asset, which exerted significant social pressure on her to pursue a career in English teaching and subsequently as an English as a second language (ESL) teacher abroad. By analyzing her narrative accounts, Solano-Campos gained insights into how broader ideologies within the EFL context impacted her professional journey.

Also, in Costa Rica, Barrantes Elizondo (2020) explored how the context of a rural university informed the sense of agency of nine teacher educators in an English teaching program in Costa Rica. Her narrative inquiry shows how these teachers’ agency was shaped and enacted in present, past and future temporal perspectives. The past was rooted in experiences as non-native English learners and their belonging to a rural institution. The present involved beliefs and emotions regarding their non-nativeness and their commitment to the university. The future sense of agency focused on how they were preparing the students to contribute to the social and economic development of the region.

In a previous study, colleagues and I examined identity negotiations of four pre-service EFL teachers in Costa Rica (Vega et al., 2021). Using QE, we concluded that the teacher’s identity negotiation processes predominantly involved adopting practices deemed valuable in their community of practice while demonstrating some degree of tension towards ideologies favoring a native speaker’s proficiency, primarily through unmarked speech and language correctness. Feelings of frustration and self-doubt characterized these reported identity tensions.

More broadly, professional visualizations of pre-service teachers have been studied in other regions of the Global South, including South America, Ruohotie-Lyhty, Aragão and Pitkänen-Huhta (2021) comparative study used visual methods with 120 participants engaging in a self-reflection task to explore their future intentions in Brazil and Finland. The results showed two perspectives: a desire for a professional status associated with wealth and prestige and a desire for meaning linked to a personal purpose in the profession. Sarasa's (2016) study analyzed the narratives of 24 Argentinean pre-service teachers. The results suggest that teachers' professional identities were characterized by desire, passion for teaching and imagination. Teachers’ imaginings included pursuing graduate studies and using resilience to adjust their efforts. These studies shed light on the effect of the sociocultural context on student teachers’ identity work, especially the future dimension. Ruohotie-Lyhty et al. (2021) concluded that identity is a “socio-politically situated desire revolving around an imagined future” (p. 8).

In Africa, researchers have examined the identity conflicts and transitions that pre-service and in-service teachers navigate in contexts that juxtapose African languages and English. In Kajee's (2016) case study of 35 pre-service multilingual teachers in South Africa, stories were used to show teachers' identities in transition as they moved from student to teacher roles, taking substantial investment in academic and English discourses. While they took in essential elements of their community of practice, their emerging professional identity primarily integrated their personal school experiences and views of the teaching profession. In a qualitative study in South African universities, Adedokun, Usadolo and Nkwatta (2024) conducted in-depth interviews with African language lecturers. They found that the instructors held an orientation to advocate for marginalized African languages based on their lived experiences, some of which related to apartheid restrictions. Results also showed how participants leveraged cultural, social and symbolic forms of capital to attain professional opportunities. Collectively, these actions were seen as “acts of strategic resistance” within their pedagogies (Adedokun et al., 2024, p. 241). Also foregrounding teachers' subjectivities as drivers of agency, Koné, Kéita and Koita's (2024) collaborative autoethnography delved into three female university English teachers' critical friendship. This study showed how they navigated professional identity tensions related to biases based on their gender. Feelings of frustration emerged when they found no community of practice. However, these emotions were productive, as they used them to access expertise and experiences valued in the profession and claim the identity of motherscholars.

Overall, these selected studies show how the complicated sociopolitical landscape of teachers in the Global South influences their emerging language teacher identities. A commonality that emerged in this literature is how teachers leveraged forms of capital and life experiences to move forward in their careers, make identity tensions productive and invest in the English teaching profession. This research highlights how non-nativeness and the economic and social capital associated with English have a weight on their decisions, desires and sense of self. Although one of the studies looks at three temporal dimensions of agency and two others look at the future aspect of teachers' visualizations, little is still known about how these three dimensions interplay in linguistic identity tensions in teachers beginning their practice. This is relevant because teachers are in a transitional phase, making critical decisions based on their trajectories and who they want to become as educators. On the other hand, while temporality and identity have been explored within the QE community (Barany & Foster, 2019; Barany, Hmelo-Silver, De Wever, & Oshima, 2021), this study focuses on language teacher identity, a field where data visualizations have not been widely used.

This study draws from two theoretical perspectives to conceptualize the process of identity formation as a site of tensions and negotiations. One is the decolonial lens offered by Santos (2016), and the second one is the sociocultural perspective of communities of practice (CoP) (Lave & Wenger, 1991; Wenger, 1998). Santos (2016) claims that over the last 200 years, the social sciences have been dominated by a Global North lazy reason incapable of capturing the immense richness of experiences and epistemologies. He posits that the lazy reason of the Global North shrinks the present in ways that leave out and invisibilize certain social experiences. This wasting of experience is caused by lazy reason's obsession with hierarchical dichotomies (e.g. man/woman, native/non-native) that create a false sense of totality. For the Global North lazy reason, nothing exists outside of this totality, even though its rationality is only one way to understand the world. In this study, this lens was used to identify the implications of the NEST-NNEST dichotomy in the teachers’ experiences. To counter the waste of experience caused by lazy reason, Santos (2016) proposes a sociology of absences, which “consists of an inquiry that aims to explain that what does not exist is in fact actively produced as non-existent, that is, as a noncredible alternative to what exists'' (pp. 171–172). According to Santos (2016), the main goal of this type of inquiry is to turn absences into presences by recovering and foregrounding the vast and rich experiences that lazy reason has intentionally failed to valorize.

On the other hand, I use CoP to conceptualize language teacher identity as a social process in which individuals build identities in relation to a community in which they engage. In this case, for teachers, it is their teacher education program and the institutions where they teach. In identity construction, teachers accept or contest the practices and discourses they find in their CoP. Particularly, I used the concepts of identification and negotiability. The former describes how teachers align their beliefs and actions to those of the community. The latter focuses on the choices, appropriations, rejections or own construction of what they observe is available in the community. Together, these two perspectives allow us to examine how teachers accept or negotiate elements from their community that have shaped their trajectories, as well as how some of their experiences and dexterity are construed as absent.

This article reports on a larger dissertation study that used narrative inquiry as a research methodology to examine the formation of teacher professional identity among EFL language educators in Costa Rica. Narrative inquiry (Clandinin & Connelly, 2000; Clandinin, 2006) conceptualizes stories as a means to interpret individuals and their relationships with others and the world. In my research, this methodology enabled teachers to explore their teaching and learning as lived experiences through stories that revealed aspects of their identities. Temporality is an essential concept in narrative inquiry that underscores that participants' experiences are continuous, connected and shaped by the past, present and future (Clandinin & Connelly, 2000). Thus, temporality is a dimension that requires the researcher to attend to transitions and shifts across time. In data analysis, considering temporality allows interpretations to reveal the narratives’ evolutions, turning points and the trajectories of the experiences (Clandinin, 2006).

In the present study, I include the results of the ENA (Marquart et al., 2019), a quantitative ethnographic technique for modeling the structure of connections in data. QE integrates ethnographic approaches and statistical procedures to generate a robust understanding of the meanings of human behavior. This approach generates thick description that connects emic and etic perspectives, linking the participants' discourse with systematic understandings of the phenomenon studied (Shaffer, 2017).

ENA creates weighted discourse networks showing the co-occurrences of codes identified in the qualitative data (Shaffer, 2017). These networks include (1) the nodes, which are the dots representing elements of discourse (i.e. codes, speakers – in this research, they represent the codes) stemming from the qualitative analysis. The bigger their size, the more frequent they were. (2) The lines between nodes, also called edges (Shaffer, 2017), represent the connection between the nodes, and their thickness represents how often connections across qualitative codes occur. The weight in the networks quantifies the strength of the connection between the nodes. For instance, if two codes co-occurred repeatedly in a segment of the analysis, the line connecting them will be very thick. The online ENA application is used by researchers to upload their coded data following formatting guidelines and use the platform's functionalities to create data visualizations.

In this research, ENA focused on the temporality of language tensions and negotiations. It was a visualization tool for relationships among codes in a fixed mathematical space. ENA was used during preliminary exploratory stages of data analysis to identify salient codes and refine and expand them through the understanding of their connections, and in later stages of the analysis, it served as a visualization and examination of identity temporality. The discourse networks show how teachers' tensions and negotiations are connected in the past, present and future. The analysis of these connections elucidates the role of an imagined identity in teacher identity development.

My ENA model included the following codes: feeling not good enough, feeling ashamed, feeling inferior, negotiating language tensions, imagined identity and identification (Table 1). The unit of analysis was each participant, including the temporalities of past, present and future. The past temporality refers to events occurring before they started teaching, for example, study abroad experiences and their teacher education programs; the present includes participants’ teaching experiences, and the future refers to imagined projections. I defined conversations as all lines of data included in each interview with each participant. For example, one conversation consisted of all the lines associated with the first interview with Gaby, and another conversation included her second interview. I created three networks for each participant; each network shows the connection among codes in each temporality: past, present and future.

This study includes two EFL novice female teachers who studied and, at the time of the research, worked in Costa Rica at private English teaching institutions. These participants recently completed their teaching education program, in which they received courses to develop their English and pedagogical skills. They came from rural communities where English was seen as an excellent asset for economic mobility. These participants are Spanish-English bilinguals, considered their professional formation a critical life experience and were proud to be alumni of a prestigious higher education institution. Silvia lived in a small town on the way to a country's popular tourist destination. From an early age, she showed an interest and skill for English that was helpful in the family supermarket when tourists stopped by. She grew up with curiosity for English and was delighted to be admitted to the teaching program. Gaby was also raised in a small town and moved close to the university when she was accepted. She was captivated by English music and pop icons, which she used as models for language learning.

The data sources of the larger study included three semi-structured interviews (Seidman, 2019) with each participant and artifacts. For this study, only the interviews were used. Interviews ranged from 75 minutes to 120 minutes. Examples of interview questions are: What is something that made you reflect about how you see yourself as teachers in conflicting views, struggles, changes and affirming moments? What are the lessons learned while being a teacher? How do you see yourself in the future? I used automated and manual transcription using Otter.ai, a paid transcription service. This transcript was transferred to an Excel worksheet for coding. Each participant had between 400 and 600 lines. These lines were segmented by topical boundaries.

Data analysis consisted of open initial coding followed by focused and axial coding (Saldaña, 2015). During this stage, I met regularly with another researcher to discuss my interpretations and guide subsequent analysis. To transition from one stage of coding to the next, I used simultaneous coding, assigning various codes to a single data point. This coding allowed me to record code co-occurrences in dense data segments that embedded more than one code in the data set. At this point, the codes of negotiating language tensions, imagined identity and identification emerged, guided by the constructs of CoP. After the second round of coding, I further revised the codebook and generated new data visualizations, which were focused on the three temporality dimensions (past, present and future). Then, I refined the codes feeling not good enough, feeling ashamed and feeling inferior, as these emotions became clearer with iterative analysis.

Regarding ethical and rigor considerations, I created opportunities to include participants' interpretations through member-checkings and revisions of transcripts. As suggested by Smythe and Murray (2000), these procedures were negotiated and carefully documented. Particular attention was paid when the researcher's interpretations were slightly different from those of the participants. Member-checkings occurred in two moments: in the third interview and when results had been written. In the interview, I co-constructed preliminary interpretations with participants by visualizing relationships among themes that stood out from the previous interview transcripts. Through this procedure, I identified emerging codes and focused on the participants' understanding.

This section presents the major findings of this study by describing the identity tensions that each of the two teachers experienced in the past, present and future. Temporality trajectories of both teachers are described individually. Each includes separate headings for (1) a narrative scenario, consisting of the qualitative data that are the basis of the analysis, and (2) the corresponding ENA analysis followed by the network visualization, which describes the codes (nodes) and how they are related to each other for each temporality. The ENA analysis suggests and scaffolds conclusions about the narrative scenario by highlighting connections between constructs that appear in the code table in Table 1. This organization aims to showcase each teacher’s lived experiences in the three temporal planes that conform to their trajectory of becoming and being English teachers. It seeks to demonstrate an in-depth application for the application of ENA in two contrasting cases. Note that the numbers referenced in parentheses in the text correspond to the description of the connections between codes included in the ENA visualizations (Figures 1–6).

Silvia described a moment when she received the comment “you write like a book,” characterizing her formality in texting, from a visiting English-speaking student at her university. Silvia added, “She's a native speaker (the person who said the comment). And she's telling me this, so meaning my English was not so as developed as it has to be.” She claimed that this observation about her made her, in her words, to “feel like so dumb” and ask herself, “Why am I writing in this way?” However, she also highlighted her enjoyment of writing classes as part of her teaching formation, saying, “My favorite part of the language is writing, and grammar […] those were my favorite classes at the university” and “My favorite part of language is writing”.

In the past temporality, Silvia’s discourse network shows the thickest lines connecting to the node of the language tension of feeling ashamed (Figure 1). The English speaker’s commentary, “You write like a book,” made Silvia feel ashamed of her writing in the context of writing text messages. This emotion triggered a sense of inferiority, perceiving something wrong with her English because a superior figure noticed a difference in her writing.

Silvia's network shows a thick line between feeling ashamed and identification (1). This relationship was shown in Silvia's acknowledgment that receiving the comment about her writing made her feel “dumb”; Silvia's judgment shows her embarrassment of her written skills. And her question – “Why am I writing in this way?” indicates her struggle with a writing competence that did not meet social expectations. On the other hand, Silvia was satisfied with her writing. Recalling experiences in her teacher education program, Silvia invested herself in developing her academic writing skills. Referring to writing as her “favorite part of language” demonstrates her enjoyment, which also aligned with the program's emphasis on this skill. Feeling ashamed shows a slightly thinner line to the negotiation of language tensions (2). This relationship explains her nervousness during tests and her decided course of action. In oral tests, she got anxious about making mistakes. She negotiated this language tension by finding temporary solutions, such as asking her mom for home remedies.

At the beginning of the interviews, Silvia noticed that she was using the word “real English” to describe the English language used by native speakers. In the last interview, she asserted, “There's no such a thing as real English.” In her current teaching, she confessed, “I don't feel bad about it [not giving students more conversational English]. […] Because what can I do?” And she resolved to “not feel bad” about not providing that in her classes. Also, she acknowledged, “I have been influenced by this academic English” in her teacher education program. She meant a form of English considered more formal and a standard for EFL teaching. Nonetheless, she confessed, “[I] wanted a more conversational English […] I think I don't have it.”

At present, Silvia’s discourse network shows a shift to negotiations of the language tensions of feeling inferior and feeling not good enough (Figure 2). Silvia associated her feeling of inferiority caused by her status as a non-native speaker with an ideal native speaker. However, Silvia's network shows thick lines between the tensions of feeling inferior and not feeling good enough (1) and negotiating tensions (2), which indicates her reconciliation with these emotions. She showed this negotiation by rejecting the word “real” English when describing her classes. She avoided using the word “real” by correcting herself because she may be becoming aware that all language varieties are valid. Silvia's second negotiation is feeling okay with offering her students what she could. She repeated the phrase “I don't feel bad” about not providing students with more conversational English opportunities in her class.

Silvia's network shows the thickest lines between negotiations and the codes of identification (3) and imagined identity (4). When making sense of her identity tensions in the third interview, Silvia acknowledged having been “influenced by this academic English.” She perceived this more formal and standard English for EFL teaching as a critical indicator of professional competence in the program. Admitting that she has been “influenced” by it means agreeing with its importance for good English teaching. At the same time, Silvia still yearned for “more conversational English.” She had this mixed desire for the academic English skills deemed valuable in her program and colloquial English. This desire reminded her of what she was missing, as evident in her phrase, “I think I don't have it.” Beginning to make sense of this complexity is part of her negotiation process; she is simultaneously trying to identify with a component of professional competence and holding a desire for an English-speaker identity.

In the third interview, Silvia discussed her desires and future plans in her teaching. She expressed her idea of “bringing a native speaker to the class.” She wanted to provide “real interactions” and “possibilities” to her students to practice their conversational skills, which she thought were lacking in their current class.

Silvia's network shows connections among the codes of identification, imagined identity and feeling not good enough (Figure 3). When visualizing her future teaching, Silvia was clear on the idea of a “native speaker” in class, thinking that what she described as “real interactions” would be created for her students. She asserted that students would have “possibilities” to practice conversational English (1). Although she did not exclusively consider bringing a native speaker, Silvia's visualization indicates that she identifies with the ideology of the native speaker as a reliable source of language. At the same time, this visualization triggered feelings about her lack of cultural knowledge, which made her an incomplete teacher (2). She wanted to offer her students opportunities to interact with native speakers.

Silvia's analysis shows that there was a desire for conversational or, as she described it, “real” English across temporalities. This manifested differently in each temporality. In the past, this desire was evident in her feelings of shame when she was a student. In the present, she desired to have conversational skills to teach her students. In the future, she aimed to provide opportunities for students to develop those skills she lacked. This identity tension was upheld throughout time, diminishing Silvia's experiences as a bilingual teacher, reducing the present and maximizing the future with a constant desire for “real English” for her and her students. On the other hand, Silvia deeply identified with discourses of academic English that were emphasized in her formation and valued professionally, which at times conflicted with her aspiration of acquiring colloquial English.

Since our first interview, Gaby used the word “real” often to denote a connection to English in an English-speaking environment. She stated, “When you learn from a real context, […] you'll learn and that stays with you because you use it, but when you learn from a book and from an activity […], you learn also, but it's not as strong.” Gaby described the “real context” as the optimal place to learn English instead of learning in books or in class. I asked Gaby if she felt a connection between the “real” English and her own English. She replied, “If I hadn't had the opportunity to go to the USA […] there would be no connection […]. The most important thing is that I'm interested in learning […] there's a connection that I, of course, want to increase.”

Gaby's past network shows the strongest connections among the codes of feeling inferior, negotiating language tensions and imagined identity (Figure 4). Regarding feeling inferior (1), like Silvia, Gaby referred to the English spoken by native speakers as “real” English. She made a contextual difference that positioned non-native English-speaking teachers in a subordinate status because they did not have authentic English use opportunities for conversational skills. Gaby expressed this hierarchy in her description of the difference between learning English in “a real context” and learning “from a book and from an activity.” Although she claimed that both settings lead to learning, she emphasized the former, providing a hierarchical relationship between the two contexts.

For Gaby, learning “from a book and from an activity” was a less desirable way to learn English because a “real” interaction was more meaningful. Gaby developed this imagined notion that a “real” context was the perfect setting to develop a desired target proficiency in English (2). She could achieve this level of English by interacting with proficient or native speakers, preferably in an English-dominant place. Gaby's perspective of finding a connection to English by being in the USA highlighted her appreciation for the opportunity to narrow the gap between “real” English and her English. Gaby found little access to more conversational opportunities in Costa Rica, so she reconciled that wish by traveling to an English-speaking context (3). Thus, she gained the experience she longed for with the hope of reaching her imagined target proficiency.

When reflecting on her English proficiency, Gaby stated, “I still have the feeling I have to improve. […] I would never say that I'm 100% a good English teacher, like I managed the language 100% or 90%” […] I feel I still have a lot to learn. Further, she stated, “In a way, I'm happy with where I am right now […] I am aware that I'm not perfect. And probably no one is perfect, like to reach 100% of our potential. So, I know I'm good.”

Gaby's discourse network for the present temporality shows the strongest connections between the codes of negotiating language tensions, imagined identity and feeling not good enough (Figure 5). Gaby's words, “I would never say that I'm 100% a good English teacher,” reflect her continuous desire for improvement (1). This desire revealed her imagined identity as an English teacher with a highly proficient level or reaching 100% competence in the language (2). In turn, this wish points to her never-enough level, never reaching an ideal 100% and considering she has “still a lot to learn.” This connection is represented in the discourse network by the lines between imagined identity and not feeling good enough.

Gaby has started reconciling these ideals and feelings by acknowledging that she was not “perfect” and no one is. She recognized that perhaps the 100% ideal proficiency is unrealistic (2). This statement represents the connection between negotiating language tensions and feeling not good enough. Gaby's negotiation is dynamic and emerging as she feels “happy” with her level, but she still desires improvement. Negotiations were prominent in the present temporality since the node is the biggest in the network.

Because I felt Gaby's dissatisfaction with her linguistic competence, I further inquired what she thought and how she felt about it. She asserted, “Sometimes I pronounce things, and I know that's not correct. […] And I feel bad in that sense, because I know that I know how to pronounce that. But I just didn't do it.” She added, “And I want to feel good with myself my level of English and also to be able to have like, this standard in the sense of other people understanding me.”

Gaby's discourse network for the future temporality shows connections between the codes of imagined identity and feeling ashamed (Figure 6, number 1). In the comment above, Gaby refers to this relationship by describing how bad she feels when she makes a linguistic mistake and what she would like instead. Gaby's words about feeling bad after committing an error suggest she feels ashamed. She wanted her performance to follow her competence because she wanted to avoid linguistic mistakes, saying that she knew, but she “just didn't do it.” At the same time, her wish for a “standard” level emerged. In a way, that may be achieved by minimizing errors and approximating to standard English, which she probably still desires.

Gaby's analysis reveals the evolution an identity tension contrasting an imagined English proficiency with her current linguistic level. From past to future, this tension manifested in a desire for a “real context” to learn, a negotiation to reconcile linguistic ideals and a future yearning for “better English.” This identity tension was present in different ways across temporalities, highlighting what was absent – not standard, perfect or 100% – in Gaby's experience rather than uplifting or making what she did have more visible.

Teaching English in the complicated sociopolitical landscape of the Global South does not come without identity struggles for teachers. Attending to the need to examine harmful dichotomies that do not valorize the experiences of English teachers in this region, this study answers the research question: What identity tensions regarding language do novice EFL teachers experience in different time frames of their trajectories? The discourse network analysis shows the dynamic development of identity tensions and negotiations in different temporalities. Predominantly, these were internal tensions arising from dissonances between who teachers were – performed identities – versus who they wanted to be -imagined identities- (Fairley, 2020). In the past, teachers experienced identity tensions due to linguistic shame (Liyanage & Canagarajah, 2019), an embarrassment of the identity associated with NNEST. For example, for Silvia, this shame manifested in having a deficit perspective about her English skills, and for Gaby, in considering English-dominant contexts superior. In other words, teachers desired and identified with an imagined target proficiency or an ideal language learning context (Motha & Lin, 2014), which contrasted with their self-perceptions. In the present, teachers negotiated this language tension by developing less unrealistic expectations about their proficiency. For instance, they started to challenge ideals and reconcile them. After investing years in learning English to the best of their ability, they realized that their imagined target proficiency was idealistic and even unnecessary in their teaching. In the future, although teachers brought a more realistic desire for improvement, it did not entirely disappear. This suggests they still perceived that they had to change something about themselves to reach legitimate competence.

The imagined identity at the center of these identity tensions shows how the teachers' descriptions, such as “real English,” “real context and” “real interactions,” denote hierarchical dichotomies that erase or waste their experience (Santos, 2016). As Ruohotie-Lyhty et al. (2021) assert, identity is a “socio-politically situated desire revolving around an imagined future” (p. 8). While previous studies have viewed this desire as productive and essential for finding meaning in teaching (e.g. Sarasa, 2016; Koné, Kéita, & Koita, 2024), this study shows that imagined identities also created challenges as teachers learned English and transitioned into their teaching roles. It is worth attending to imagined identities as spaces for possibility and for invisibilizing language teachers’ social experiences.

The temporal view in the discourse network analysis visualizes the role of teachers’ imagined identity in mediating language tensions and negotiations across temporalities. The analysis showed that teachers’ idealistic imaginings incited challenging emotions, but they transformed them when they realized how constraining these notions were. As such, imagined identities enabled teachers’ identities when unrealistic language expectations were interrogated; however, they caused tension when teachers felt that they must adhere to language ideals.

This study has implications for teacher education, particularly for preparing teachers navigating the linguistic hegemony of English. As identity formation is a place for contradictions and conflict (Barkhuizen, 2017; Fairley, 2020), it is essential to understand the nuances of such tensions and their implications for teachers as they transition into new stages of their trajectories. Imagined identities may have a productive side. As such, they can lead to small or more significant “acts of strategic resistance” (Adedokun et al., 2024, p. 241) that challenge oppression. For language teachers, these could be meaningful ruptures of dichotomic views. However, imagined identities can also involve desires that reinscribe hierarchical categories, reducing teachers' experiences and producing them as non-existent. Teachers may come to see their knowledge and strengths as illegitimate because they don't match idealized standards. Teacher education should look at both sides of imagined identities, enabling teachers to explore the extent to which their desires foreground their experiences as speakers and teachers of English.

This study contested the epistemic waste (Santos, 2016) in EFL. Valuable knowledge has been invisibilized, resulting in the waste of experience that could contribute to building epistemological diversity. The power systems sustaining the hegemony of English in the Global South have relegated teachers as subordinates of the English language (Makoni & Pennycook, 2005). This privilege of language over people has marginalized the knowledge stemming from teachers’ experiences. To this end, this research provides an understanding of teachers' localized identity tensions and negotiations that otherwise would be unnoticed and left out in teacher education. The lived experiences are profound and revelatory of oppression and emerging liberation. The discourse and personal imaginations had a significant place in the analysis. These elements were in constant conversation with an evolving theorization of language teacher identity situated in participants’ lives. ENA amplified the conclusions of this research by systematically modeling and visualizing the connections in different temporalities of identity development. ENA allowed a nuanced examination of teachers' discourse, specifically looking at their imagined identities from multiple perspectives.

This paper forms part of a special section “Quantitative ethnography in education research and evaluation in low- and middle-income nations”, guest edited by Drs Eric Hamilton, Danielle Pascual Espino, Seung Lee and Kristina Lux.