The concept of “teacher identity” first emerged in the 1980s in the field of general education as an analytic notion for studying teaching and teacher education. Teacher identity is an ongoing and dynamic social process of how teachers interpret and reinterpret who they consider themselves to be and who they would like to become (Beijaard et al., 2004) [6] in relation to different ways of “inhabiting roles, positions, and cultural imaginaries” (Holland & Lachicotte, 2007, p. 103) [17] . According to the literature, there are two common understandings of teacher identity: one is that teacher identity is not a static, invariant or unitary construct (e.g., Varghese, et al., 2005 [18] ; Beijaard, Meijer, & Verloop, 2004 [6] ; Han, 2017 [19] ) because it is based on the core beliefs one has about teaching and being a teacher, which are continuously changing and reformed through experience; The second is that teacher identity is a highly contextualized, socially defined phenomenon (Beijaard, Meijer, & Verloop, 2004) [6] that “expresses personal practical knowledge gained in experience, learned contextually, and expressed on landscapes of practice” (Xu & Connelly, 2009, p. 223) [20] . That is to say, teachers develop their identity in a context that brings social and cultural forces to bear upon that development (Holland & Lachicotte, 2007) [17] .

There has been a burgeoning interest in language teacher identity since the past decade (e.g., Beauchamp & Thomas, 2009 [21] ; Day & Kington, 2008 [22] ; Han, 2017 [19] ; Tsui, 2007 [15] ). It is shown that language teachers negotiate and construct identity under the interplay of 1) personal factors, such as teacher emotion (Xu, Y., 2013) [23] , teachers’ own English proficiency (Richards, 2008) [24] , teaching beliefs and attitudes (Farrell, 2011) [13] , or critical life events (Liu, 2009) [25] ; 2) institutional factors like work environment (Miller, 2004) [26] , teachers’ relationships with stakeholders (administrators, colleagues, parents and students) (Kanno & Stuart, 2011) [14] , and 3) social factors like national language policy and reform (Liu & Xu, 2011) [8] . These multidimensional factors shape the kaleidoscopic nature of language teacher identity. That is to say language teachers’ identity contains social and individual elements simultaneously (Han, 2017) [19] which is influenced not only by one’s life experiences, family background and values, education, peronality but also by what they do in relation to social discourses, policies, students and parents, their colleagues and their practices (ibid). However, how these factors influence teacher identity evolution and how they might result in teacher identity crisis over an extended period of time are relatively underexplored.

Huberman et al. (1993) [27] identified three career phases of teachers: 1) Year 1 - 3 career entry and early beginnings (feeling one’s way), 2) Year 4 - 6 stabilization which has profound implications for twenty-five years, and 3) (serene or bitter) disengagement. So far, teacher identity construction at the beginning career phase received much attention as discussed earlier, and studies on teacher identity evolution across the three phases are rare. Among the few relevant studies, Day and Kington (2008) [22] studied 300 secondary and primary school teachers in England over four years and found that teachers in different career phases experienced different scenarios that challenged their abilities to sustain their initial commitment. The four-year longitudinal case study of four Chinese EFL teachers by Xu, H. (2013) [28] revealed their professional identities change in the first years of teaching in K-12 schools under the mixed influences of the institutional contexts of school and the dynamic educational contexts. These studies show that teacher identity changes across career phases, however, the complicated process of teacher identity evolution still needs further exploration.

The term identity crisis was coined by German psychologist Erik Erikson (1968) [29] to describe adolescents’ failure to achieve ego identity when facing dramatic physical and mental change. Focused examination on teacher identity crisis is scarce until now (Xun & Zheng, 2014) [30] , but the handful relevant studies do suggest that teacher identity crisis can be triggered by factors such as work overload (Hargreaves, 1994) [31] , educational reform (Bolívar, et al., 2014) [3] , non-nativeness for language teachers (Shin, 2007) [32] , or work environment and low payment (Yang, 2005) [12] . For instance, Shin (2007) [32] found the “English-only policy” which legitimizes NS as ideal language teachers led to identity crises in some NNS teachers in Korea. Bolívar, et al. (2014) [3] revealed that education reform destabilized the professional practice of 10 secondary teachers in Spain and brought about their identity crisis. Similarly, Zhang and Lam (2008) [33] , using an experienced Chinese secondary school teacher of physics as example, pointed out that teachers who feel powerless to adapt to the new mode in the curriculum reform would experience identity crisis.

Given the multifaceted nature of teacher identity, more work is still needed to understand how various internal and external elements contribute to language teachers’ identity crisis in a particular L2 context as well as how these factors shape language teachers’ identity evolution over time. Therefore, this study targets at three Chinese EFL teachers in a higher vocational college in China, and specifically addresses three questions: 1) How did their professional identities evolve over the past ten years? 2) Have they experienced identity crisis over the past ten years? If yes, what are the contributive factors? and 3) How do these factors influence their teacher identity evolution?

This study took a narrative approach to explore the three Chinese EFL teachers’ identity evolution trajectories, given that teacher identity is closely tangled with teacher experience and narrative is “the best way of representing and understanding experience” (Clandinin & Connelly, 2000, p. 18) [34] . However, as Bolívar (2002) [35] pointed out, while narrative inquiry respects the voices of the investigated subjects, this advantage can bring problems to researchers. That is, if the researcher respect the subjects’ emic discourses excessively, the researcher’s interpretation may get caught inside the horizons of the subjects. To prevent this problem, researchers need to integrate both the emic (from within) and the etic (from outside) perspectives when interpreting and reconstructing the narratives to portray the inner realities inscribed in the external context that provides meaning for the realities.

The three participants Linda, Shelly, and June (in pseudonyms) are full-time college English teachers in College A, a public vocational college in China. There are over 40 English teachers in College A, and most of them are Chinese EFL teachers who teach compulsory public English courses to Year-1 non-English major students. The three teachers were chosen because they are experienced who had been teaching in College A for over ten years by the time this study started. Second, they are representative of the English staff in education background (BAs in English, domestic MAs, and MA overseas returnees) and age group (late 40s, early 40s, and late 30s) (see Table 1 for the basic information of the participants). Also, one of the researchers (Researcher A henceforth, the other Researcher B) studied with the three teachers at a short-term training project and has maintained friendship with them since then for over a decade, so the strong rapport ensures access to their private inner stories.

The narratives of Linda, Shelly, and June were constructed and reconstructed over one year mainly based on their self-reflection journals and two rounds of in-depth interviews.

The three teachers were first invited to write self-reflection journals to recount their 1) feelings and perceptions of being an English teacher in College A over

the years; satisfaction with the profession; perceptions of their status and social recognition; relationship with the institutional environment (the school, students, colleagues); and attitudes toward socio-institutional policies and management, curriculum reform, etc.; and 2) any significant events and meaningful anecdotes of teaching, research, and life relevant to such feelings, perceptions, attitudes, and evaluations. They were given a whole winter holiday (Feb. 2017) for the writing and there were no word limits nor requirements about language or structure of the writing. They all wrote the journals in chronological order and Chinese.

After obtaining insights from their journals, Researcher A conducted two rounds of one-to-one and face-to-face semi-structured interviews with the teachers. The first round was undertaken in July 2017 during which the teachers were invited to elaborate on what they mentioned in the journals, so that the researchers could gain more insights into the teachers’ reflections and story characters, plots and contexts (Clandinin & Connelly, 2000) [34] . The second round was conducted one month later which involved more sensitive talks about the socio-institutional and personal factors (e.g., collegial relationship, family incidents or plans) or critical events that affected their identity experiences. Each interview took about two hours and was recorded with their permission. Then the interviews were transcribed verbatim in Chinese.

A qualitative thematic analysis of the data was carried out, which was interwoven with data collection. First, a “preliminary exploratory analysis” (Creswell, 2005, p. 237) [36] was undertaken. Both researchers first read through their narratives several times to obtain a general understanding. Then major recurring themes and categories were identified and manually coded. As discussed earlier, bothemic and etic views are necessary for interpreting subjects’ narratives. Researcher A’spersonal close friendship with the teachers over the years kind of gives her a double identity of “insider” and outsider, which, though in some way opened an opportunity to a unusual access to the meaning reconstruction (Liu & Xu, 2011) [8] of the teachers’ narratives, might bring subjectivity into the interpretation that can soften her judgment or even “mute the emic voice” if untamed (Peshkin, 1988, p. 21) [37] . To avoid this potential problem, the preliminary analysis and coding were carried out by both researchers independently. Then both researchers discussed and recoded those disagreed places.

Three thematic phases of the teachers’ identity evolution emerged from the preliminary analysis: the honeymoon phase (characterized by meaning constructs like happiness, satisfaction, luckiness, etc.), the confusion phase (characterized by loss, disappointment, hopelessness etc.), and the dilemma phase (characterized by struggle, indecisiveness, predicament, etc.). To group the categories that contributed to thesethematic phases, Day’s (2011) [38] categorization of the macro-, meso-, and micro-level influential factors of teacher identity were applied as a guiding compass: 1) socio-cultural/policy, reflecting cultural, social and political expectations of teachers and teaching, as well as teachers’ educational ideals, and ethical and moral purposes; 2) workplace or socially located influences, including the micro-politics and social relationships of school, department or classroom contexts and local conditions (pupil behavior, the quality of leadership, and support and feedback in teachers’ immediate work context); and 3) personal influences, located in life outside school like personal history, present life, family, and a personal sense of efficacy and vulnerability. This preliminary analysis revealed contributive factors to the three teachers’ identity experiences from the above three levels: curriculum reform, college and school policy on teaching, collegial cooperation, classroom satisfaction, research, students, promotion of professional title, personal life experience, etc. The thematic phases and the contributive factors formed the guiding coding scheme for the later main coding, as shown by Figure 1.

The actual main coding was implemented with the assistance of Nvivo 10. The reflective journals and interview transcripts were imported into the software and coded according to the coding scheme as shown above. This round of coding also, to some extent, “triangulated” (Liu & Xu, 2011, p. 184) [8] the themes and categories derived from the manual coding, through which several new categories (such as family situation, teacher appraising system, social status) were

identified and added. Finally, the teachers’ narratives were rewritten with the three thematic phases of identity evolution functioning as the storyline. Mini-stories concerning different influencing factors were recounted and weaved under the three phases. The reconstructed narratives of Linda, Shelly, and June were sent back to them to ensure the narratives reflected their experiences (Creswell, 2012) [39] .

A brief sketch of each of the three teachers’ “personal and contextual contours” (Xu, Y., 2013, p. 380) [23] will be presented here to help readers better understand their identity evolution over the years.

Linda was born in mid 1970s in a countryside of an underdeveloped province in China. Being the middle child in her family, she often played the role of “mother” who did most of the housework due to her mother’s poor health condition. She first taught in a local primary school, during when she got married and had a son. Several years later, she went to an economically advanced city in China where she self-studied English and obtained a college diploma. To reunite with her husband who then moved to City A, Linda took the postgraduate entrance exam twice and was finally admitted by a university in City A, majoring in English Education. After she gained her master degree, Linda found the job in College A in 2006.

Born in early 1980s, Shelly’s journey of academic pursuit was smoother than Linda’s, who studied in City A from kindergarten to college. Though Shelly is a city girl, the financial condition of her family had never been sound as her father passed away long ago and all life burden fell onto her mother who ran a small grocery store. After gaining her BA of English, Shelly studied at a university in UK majoring in Business Management for a master degree. After graduation, she came back to City A and first worked in a private college and then a year later transferred to College A in 2007.

June has been teaching in College A for the longest time among the three. She was born in early 1970s in a countryside family in an inland province and similar to Linda, she shouldered much family responsibilities since childhood. After obtaining her BA of English from a four-year normal university in City A, she became a teacher in College A which was then a secondary vocational school. When Researcher A first met June in 2006, June was already an established figure in the school as one of the few associate professors in the English faculty of College A.

Under the interplay of various macro-, meso-, and micro-level factors (as will be explicated in the next section), Linda, Shelly, and June’s teacher identity evolution in College A displayed a similar pattern over the past ten years: first the honeymoon phase during which they accepted being a higher vocational college English teacher, then the confusion phase when they doubted the meaning and value of that identity, and finally the dilemma phase during which they struggled over staying in or out of that identity.

Recently, there has been a curriculum reform in higher vocational college English teaching since China’s Vocational College English Curriculum Requirement (2009) promoted “General English (GE) + Professional English (PE)” as the new mode. GE aims at cultivating students’ general English abilities of speaking, listening, reading, and writing and PE targets at English language competence relating to students’ future profession. This new mode asks for Double-qualified teachers who not only have knowledge in English language and teaching but also other subject knowledge like accounting, logistics, tourism, etc., which poses challenges for mostvocational college English teachers, since the majority of them specialize in English language or English education and have no experience nor qualification in other disciplines.

This curriculum reform appeared repetitively in the teachers’ narratives. The glim chance of becoming a double-qualified teacher made June to doubt her value to the school:

All they (the school administrative leaders) talked about now is “Double qualified teachers”, making us feel that we are worthless to the school! They urged us to learn other professional knowledge, but our teaching workload was still as heavy as before. They offered us no help nor chance to access other fields. How could we find the extra time and energy? How could we transform into double-qualified teachers all by ourselves?

Shelly said the national propaganda for double-qualified teachers has put public English teachers in College A in a more unfavorable situation:

We public college English teachers are already peripheral members in the college compared with those professional course teachers. Now, with this “double-qualified teachers” thing going on, we became even more marginalized. The college already started to hire those who once worked in a professional field with good English proficiency. How could we survive?

Another reason mentioned by the teachers is the ongoing credit reduction of college English teaching in China. Due to the emphasis on the practical learning in vocational college education, the credits of public courses like college English are commonly reduced in recent years. In College A, the credits of college English decreased from 12 to 8, this made English teachers feel insecure about their job, as Linda remarked:

College English course is one of the “endangered” courses and always the victim in the reform. Whenever they need to cut the credits, they would think of public English first! We feel in peril here!

The soundness of promoting double-qualified teachers and reducing public course credits is beyond the scope of the present discussion. But these broad social factors clearly exerted negative impact on the three teachers’ identities, who felt powerless to adapt to the new requirement. This finding is consistent with the literature (e.g., Bolívar, et al., 2014 [3] ; Zhang & Lam, 2008 [33] ; Van Lankveld et al., 2017 [41] ), showing that the higher-up educational and curriculum reform could constrain teacher identity development if teachers are not well-prepared and assisted.

The institutional environment plays an unnegligible role in shaping teachers’ identity (Tsui, 2007 [15] ; Richards, 2008 [24] ). Among the many aspects, those relating to the direct workplace, to teaching and students, and to staff development are found to be major contributors to the three teachers’ identity crises.

There is a developing awareness of connections between teachers’ private lives and their professional thoughts and actions (Goodson, 2003) [44] . An important self of teachers are family situated, therefore, family situation, family plan or family change can influence teacher identity development. In Linda’s case, family responsibility caused her to give up her own career pursuit, and the second child distracted her attention even changed her understanding of teaching: “I now see myself mainly as a mother and less as a teacher! Teaching may not be as important as I used to think.”

Similarly, Shelly’s family financial situation puts her in a predicament, which also drove her into a prolonged preparation for leaving the college. June had a divorce one year ago which seemed changed her way of looking at teaching drastically: “The divorce made me realize that I don’t need to be fully devoted toteaching, there are more important things!” Clearly, June’s case resonates with what Liu (2009) [25] found that dramatic occurrences in one’s private life exert great influence on teacher identity shifts.