Building professional learning communities of foreign language teachers in higher education Círculo de Lingüística Aplicada a la Comunicación

. In recent decades there has been increasing interest in identifying the critical features of effective professional learning communities (PLCs). This identification is useful for evaluating the quality of a PLC, but not for building one. This study aims at describing a conceptual model for developing a new PLC, illustrating its application with an example. The model is explained in terms of four constitutive elements (PARTICIPANTS, OBJECTIVES, MEDIATORS, and MECHANISM), and their interactive relations. The development of a PLC of foreign language teachers (PLC-FLTs) at Beijing Foreign Studies University is used to illustrate how the model can function successfully. The illustration is coupled with a discussion of the PLC’s decades of experience in collective leadership, management of the four elements, and tips for tackling various challenges in keeping the PLC going. the challenges involved in sustaining the PLC-FLTs. 5. Conclusions. Funding. Acknowledgments. References.


Introduction
Over the last three decades the notion of professional learning communities (PLCs) has drawn increasing attention from educational researchers since its introduction by DuFour and Eaker (1998). A PLC is generally defined as "a group of people sharing and critically interrogating their practice in an ongoing, reflective, collaborative, inclusive, learning-oriented, growth-promoting way; operating as a collective enterprise" (Stoll et al. 2006: 223). It is widely assumed that PLCs can make significant contributions to professional development, teaching effectiveness and ultimately student learning (Bolam et al. 2005;Vangrieken et al. 2017;Huijboom et al. 2020).
Previous research on PLCs in the field of education has explored PLCs in pre-primary, primary, and secondary education (Cheng & Pan 2019;Thornton & Cherrington 2019). However, much less is known about the presence of 1 Beijing Foreign Studies University 北京外国语大学中国外语与教育研究中心/国家语言能力发展研究中心 (China). Email: wenqiufang@bfsu.edu.cn 2 PLCs in higher education (HE), and even less about how a PLC of foreign language teachers (PLC-FLTs) may be built and sustained. It may be contended that PLCs are particularly necessary in the context of higher education, since the nature of HE environments would seem to be more likely to lend itself to isolated practice with teachers working on their own. Previous research on PLCs in HE has identified difficulties and barriers to building PLCs (Hargreaves 2007;Alles et al. 2019), reporting that it is difficult to build a positive learning atmosphere and develop mutual trust at HE level (Alles et al. 2019).
PLCs do not develop easily or quickly, and Stoll and Louis (2007: 7) suggest that "the difficulty of developing professional learning communities should not be underestimated" because of the complexity of daily educational practice (Hairon et al. 2017). It is therefore necessary for us to develop theoretical models and operational procedures for building PLCs and maintaining them when they have been formed.
Successful PLCs have five defining features, namely: 1) supportive and shared leadership; 2) shared values, vision, and goals; 3) collective learning and application; 4) shared individual practice; and 5) supportive conditions (both physical and human) (Hord 1997;Hord & Sommers 2008). These characteristics are useful for evaluating the quality of an existing PLC, but not of so much use when creating a new one. For instance, the relations among the five features are not specified in existing studies, and neither is their relative importance explained. In other words, these features could serve as general guidelines for assessing the effectiveness of an existing PLC, but they offer no operational procedures or strategies if you plan to start one.
Thus, a conceptual model is required, to elucidate the essential elements for a new PLC and their interactive relations. Moreover, this model should be grounded and explained with discussion of an illustrative example.
Against this backdrop, over the past 10 years a research team led by the first author in the National Research Center for Foreign Language Education of Beijing Foreign Studies University (BFSU) in China has been developing PLCs in HE. The first was a PLC of teachers of English within BFSU, lasting from 2009 to 2011 (Wen & Ren 2011, 2012; the second was a PLC of teachers of English from six different universities from 2011 to 2013 (Wen 2017); and the third was an ongoing PLC-FLTs at BFSU starting from 2015 (Wen & Zhang, 2019). All three of these PLCs were developed to promote collaborative research and enhance the language teachers' professional development. In this paper, we aim to present a model for developing a new PLC and demonstrate how the model was applied to the building of the PLC-FLTs in HE as an illustrative example.

The model for building a new PLC
Emerging from our extended efforts to build PLCs is the following model ( Figure 1) (Wen 2017;Wen & Zhang 2019). It includes four constitutive elements, with the dyadic relationships among them indicated. The first element is PARTICIPANTS at the center of the model, which naturally plays the most essential role. It is a prerequisite and determinant of the other three. The second is OBJECTIVES, which are the shared goals that form the destination of the PLC. The third is MEDIATORS, which serve as devices for reaching the objectives. The fourth is the MECHA-NISM that coordinates the other three. The dyadic relations between the elements are interactive, as indicated by a double-headed arrow. In other words, no element can function in isolation, and none is passive. To develop a new PLC for collaborative research and professional development in HE, the PARTICIPANTS may include two groups. The first group contains a few experienced researchers who share collective leadership, and the other group may be classroom teachers who voluntarily participate in the PLC. The members are normally diversified in terms of their ages, teaching areas and research experience, so that they bring a variety of experiences and expertise to their discussions. The collective leadership should have a strong sense of commitment, devotion, and dedication.
Other members should show an equal willingness for full participation and readiness to offer new ideas, as well as to offer necessary assistance to the project.
OBJECTIVES clarify the orientation of the PLC and serve as the standards for evaluating its effectiveness. They should include long-term and short-term objectives. The long-term goal should be to empower the PLC members with lifelong learning competence, and to enable them to achieve whole-person development. Meanwhile, short-term goals should be tangible, specific, and measurable, such as accomplishing a research project, writing up a research paper, compiling one or two sample units of a textbook, and so on. Without these short-term objectives the PLC members might lose their sense of direction, which may cause their efforts to tail off. However, short-term objectives should also have a direct bearing on the long-term goals. Otherwise, the PLC will become purely instrumental, without humanistic values. In other words, the long-term objectives should be embedded in the short-term ones and realized through them in due time. In setting objectives, the PLC leaders should consider the PLC members' needs in advance, and the needs and objectives should be discussed thoroughly and agreed upon by all. That is to say, the objectives must be shared by both the PLC leadership and the PLC members, so that joint efforts can be made in the same direction.
Once the objectives are set, appropriate MEDIATORS need to be employed to achieve the objectives (Wertsch 1985), and also serve as a carrier for PLC members to interact with one another. They can be classified along two dimensions: either human or non-human; and either concrete or abstract. When humans serve as mediators, they may refer to the leaders, the members within the PLC, and experts outside the PLC. When non-humans function as mediators, they may refer to concrete objects such as relevant books, papers, etc., and also abstract issues such as a research topic, key problems in teaching, etc. No matter what kind of mediators they are, they should be regarded as "the right help at the right time" (Verity 1995: 3, cited in Kaur 2015, or a form of support that is "efficient, targeted and goal-oriented" (Johnson 2009: 20). The key to providing the "right" or "efficient" help lies in making appropriate choices about how a task is to be performed (Kaur 2015), and the help should be provided within the PLC members' zone of proximal development (ZPD). In Johnson and Golombek's (2011) view, the ZPD is a metaphorical space for the PLC leader to see what the PLC members might be able to do with the right assistance.
MECHANISM, as the fourth element in the model, refers to the system that sustains the development of a PLC. Figure 2 illustrates such a system that is useful for building a PLC. The system is crystalized as a pull-push mechanism that functions along both emotional and cognitive dimensions. The pulling and pushing forces are interacting with each other between PLC leaders and PLC members. As Figure 2 illustrates, pulling occurs before pushing, and the interaction along the emotional dimension occurs before the cognitive one. This means that the PLC leaders should make deliberate efforts to establish a community where all the PLC members feel confident, secure, and happy. In fact, a mutual-reliable and mutual-trustworthy atmosphere is essential for any successful PLC.
Furthermore, the figure also indicates that there must be a balance between pulling and pushing. If the pull is not strong enough the PLC members may lose confidence in being able to complete their assigned task, but if the pull is too strong, they may become over-reliant on others' help. Similarly, if the push is not strong enough, they might not meet the requirements of PLC members, but if the push is too strong, they might start to feel frustrated or even humiliated. The general policy is that PLC leaders should adjust the balance between pulling and pushing from time to time. The whole process might involve many cycles of pulling-pushing, but each cycle lasts only a short period of time. In other words, a bit of pulling and a bit of pushing work alternately, as the case demands.
To sum up, our model (shown in Figure 1) not only presents the four crucial elements (PARTICIPANTS, OB-JECTIVES, MEDIATORS, and MECHANISM) but also specifies their roles. The relations between each pair of elements are interactive and dynamic. Unlike the five defining features for a successful PLC (Hord 1997;Hord & Sommers 2008), this model is designed to inform researchers and practitioners about creating a new PLC. It has been demonstrated to be readily comprehensible and practically workable in developing the three types of PLCs developed and studied by our research team.

An illustration of how to build and sustain a PLC-FLTs
In this section, we will focus on an ongoing PLC-FLTs that has lasted for more than four years. First, brief background information about BFSU will be provided to give readers a better understanding of the wider environment where the PLC-FLTs is situated. This will be followed by a description of developing a PLC-FLTs in terms of the four elements in the model described above. The following section will illustrate the working process of the PLC-FLTs by examining one research project carried out by all the PLC-FLTs members.

Overview of BFSU
BFSU is one of the key universities in China, founded 80 years ago. It is a comprehensive university of liberal arts offering BA, MA, and PhD programs. However, it is most famous for the high quality of its foreign language BA programs, which have produced more than 400 Chinese ambassadors. The university is thus often called "the cradle of diplomats". In the past few years the number of BFSU's non-English foreign language degree programs has been increased from 40-odd to about 100, to cover all the official languages of the countries that have diplomatic relations with China. The goal here is to enhance the national foreign language capacity.
BFSU's current challenge is that a considerable number of young teachers have been recently recruited into the university who are badly in need of mentoring if they are to become competent teachers as well as researchers. Against this background, the university leaders have encouraged senior professors and experienced teacher-researchers to establish PLCs to help young faculty members. In response to this need, the PLC-FLTs described in this study was created in November 2015, and has functioned continually up to the time of writing. The following section will describe the building of the PLC-FLTs.

PARTICIPANTS: PLC leadership and members
The PARTICIPANTS in the PLC-FLTs involve two groups of people: (a) collective leadership, and (b) 32 FLTs working at BFSU (Table 1). The PLC leadership consists of one leader and two core members, who are all working in the National Research Centre for Foreign Language Education at BFSU. The PLC leader is a senior professor who has been teaching and researching for more than 40 years, with rich experiences in doing research on teachers' professional development. She has made pioneering efforts in building the three PLCs in China. One core member is working at the cutting edge of teacher education, and the other core member has expertise in linguistics. Both are associate professors.
The PLC members from nine different language departments all volunteered to participate in the PLC, rather than being forced to do so by the institutional authority. The number of FLTs has gradually increased from 14 to 32 over the four years. They teach 25 different foreign languages, and all are actively engaged in the activities in the PLC-FLTs. Their teaching experiences vary from 1 to 26 years, and the majority are female lecturers (30 out of the 32 participants). All the PLC members are highly proficient in the languages they teach and knowledgeable about the countries of their target languages. However, most of them have little research experience. Meanwhile, they face challenges and dilemmas in their professional development. For example, as teachers in newly-established BA programs, they might have neither textbooks nor dictionaries as reference materials. Nevertheless, they may be required to teach various undergraduate courses. Furthermore, they may have multiple administrative responsibilities such as designing a schedule for teaching in every semester, and arranging the itineraries for visits of foreign officials from their target language countries. In addition, they face a conflict between their commitment to teaching and their strong desire to meet university requirements for academic research. No matter what languages they teach, their promotion largely depends on their research output, since the requirements for teaching can be easily met. The reality is that the majority of the PLC-FLTs members had not published papers before they joined the PLC-FLTs. Moreover, they seldom communicated with teachers of other languages. Facing so many difficulties and challenges, they were eager to join the PLC and hoped to gain support for their professional development (Wen & Zhang 2017).
As mentioned above, the PLC-FLTs only recruited volunteers. However, they were required to sign a PLC-FLTs contract with the PLC leaders. The contract included a list of rights and obligations. For the leaders, they promised to hold at least one face-to-face conference per month to mentor PLC members collectively or individually, and to provide them with help to write research papers based on projects and grant proposals. Meanwhile, the PLC-FLTs members were required to attend meetings regularly, finish required tasks on time, and write a manuscript for publication each year and a grant proposal within three years.

OBJECTIVES: Long-term and short-term
Based on our previous experience of developing two earlier PLCs and a consideration of the needs of the PLC-FLTs, we designed long-term and short-term objectives and reached agreement with the PLC-FLTs members after a full exchange of opinions. Specifically, the long-term objective was to develop a team of applied linguists who are skillful in both teaching and researching. They will not only improve the efficiency of foreign language education, but also aim at all-round development of themselves at the same time. Ultimately, this team will enhance the competitive power of BFSU in the field of applied linguistics both at home and abroad. The short-term objectives included the following: 1) writing up one research paper that meets the standards of publication within a year; 2) compiling one or two sample units of a new textbook by applying an innovative pedagogical theory within a year; 3) drafting one or two chapters of an academic book within a year; and 4) writing a proposal to apply for a research grant within a three-year period. The PLC-FLTs members could choose at least one of these short-term objectives as their personal target.

MEDIATORS: Human vs. non-human and concrete vs. abstract
As described in Section 2, the mediators serve as a means for reaching the objectives, as well as a carrier for the PLC members to work together. Among all the mediators in this PLC-FLTs, the most important were the tailored research projects designed by the PLC leadership. This tailoring depended upon the PLC leader's intimate knowledge about the PLC members' interests and abilities. The research projects designed for the PLC members (Table 2) can be grouped into three categories: discourse analysis, application of a new pedagogy, and the national language capacity of different countries.
1. Projects in the first category consist of "political discourse analysis of leaders' speeches at the UN General Assembly", "the 'Belt and Road' initiative in international media coverage", and "critical discourse analysis of reports from foreign media coverage on the Political Report at the 19 th National Congress of Communist Party of China (NCCPC)". 2. Projects in the second category aim to apply the Production-Oriented Approach (POA) to classroom teaching and teaching materials development. The POA, as an innovative pedagogy with Chinese features, has been developed by the research team led by the first author of this paper over the past decade (Sun & Wen 2018;). 3. The projects in the third category focus on the national language capacity of different countries. Each of the PLC-FLTs members chooses one country, such as the UK, Malaysia, India, Egypt, etc., as a research focus. The final products may lead to a series of books concerning the national language capacities of a particular country. The general theoretical framework of national language capacity was constructed by the research team from the National Research Centre for Foreign Language Education when doing a project on China's national language capacity.
The above projects were effective mediators because they were designed in such a way that all the PLC members could bring their own expertise into full play, and at the same time they could cooperate with one another without worrying about their innovative ideas that might be "copied". Let us take the project of "political discourse analysis of leaders' speeches at the UN General Assembly", for example. Every PLC member focused on the speech by the leader from the country of their target language at the General Assembly of the 70 th anniversary of the United Nations. All the leaders represented their own countries, and thus their speeches were aligned with the stance, principles, and social values of their governments. The PLC members could benefit from collaborative discussions, while they had to make efforts to carry out data analysis and write a paper by themselves. In addition to these research projects as the most crucial mediators, the PLC-FLTs also have academic books, lectures, and papers as subsidiary mediators for accomplishing the projects. For example, two professors from other universities were invited to deliver lectures on discourse analysis. Apart from the nun-human mediators, the PLC leadership, the PLC members, and some invited experts all serve as mediators. It is particularly important that the 32 FLTs support one another through peer-reviewing, peer-assessment, peer-checking, and so on.

MECHANISM: Pull-push
As mentioned before, the pull-push mechanism involves both emotional and cognitive dimensions. In terms of emotional pulling, on the one hand the collective leadership is honest, friendly, and respectful, aiming to create a safe and helpful community environment with mutual trust. On the other hand, the collective leadership is extremely sympathetic towards the members about their heavy workload, and appreciative of their courage to face challenges and their devotion to their duties and research. Moreover, emotional pulling comes before cognitive pulling. For example, at the first two-weekly meeting of the PLC-FLTs, we discussed how to work together to achieve mutual benefits and build mutual trust. One of the leaders said: As a teacher of English, I have considerable knowledge about English-speaking countries. However, my knowledge concerning the other countries is limited. For instance, I know little about Iceland, Norway, etc. One of the advantages of our PLC is that the members teach diverse languages and are familiar with their target countries' political systems, social customs, and economic development. I am sure I can learn a great deal from all of you here, and at the same time you can learn from one another.
This example strongly emphasizes that each member of the PLC is an asset to the group and can make a unique contribution to its growth and maturity. In addition to boosting their confidence and helping them realize their own value, the leadership also shared experiences with them of how PLC-FLTs members could put forward critical comments on others' work in a professional way, without hurting others' self-esteem. They were advised to use a sentence pattern such as "If I were you, I would…" when raising suggestions. Furthermore, a set of rules were set up for the behaviors of the PLC-FLTs members at the very beginning. For instance, whatever shortcomings of members might be noticed during discussions, other members are not to gossip about them, either within or outside the PLC-FLTs. In general, the emotional pulling is contingent on several factors including, but not limited to, understanding teachers' difficulties and needs, reducing their anxiety, inspiring, encouraging, motivating, and enthusing each and every member for the sake of the group's common goals.
The cognitive pulling may be demonstrated in several aspects. First, it means providing easy access to academic resources, timely guidance, patient mentoring, and requested assistance of any kind. Let us also take the research project of "political discourse analysis of national leaders' speeches at the UN General Assembly", as an example. First, several books and papers on discourse analysis were recommended to the PLC members, such as Discourse, Grammar and Ideology: Functional and Cognitive Perspectives (Hart 2014), as well as journal articles on sociocultural linguistic approaches (e.g., Bucholtz & Hall 2005). Second, the collective leadership dissected the project into small and concrete tasks, and provided scaffolding. In practice, the project was carried out bit by bit, lasting a whole year in total. Third, the members also pulled the collective leadership in the sense that they were, so to speak, force-fed with information about multilingual cultures introduced by non-English-speaking members. Fourth, the community pulled one another by learning from each other's varied expertise through collaborative discussions, and by reading and revising each other's papers. In addition, two lectures were organized on discourse analysis; some members were asked to sit in classes on systemic functional linguistics; and 16 PLC-FLTs meetings were held in 2016. All these activities served to raise awareness about doing research in one way or another.
"Pushing" includes activities such as discussions and negotiations that urge the members to make conscientious efforts to comply with accepted regulations and requirements. Emotional pushing may take place in various forms. First, a contract was signed with each PLC-FLTs member, which prescribed their duties and rights. Deadlines were also set for all the assignments. Meeting attendance was monitored, and absence without justifiable reason was not allowed. Furthermore, members were given friendly reminders of approaching deadlines for specified tasks. Consequently, the members were prompted to complete the tasks in good time.
In addition, the members could push one another by peer pressure. When one member submitted their work in a WeChat group or by email, those who had not submitted their work would feel pressurized. Meanwhile, the PLC leaders were also sometimes emotionally touched by the members. They were deeply moved when they found an email with a draft of a paper, sent at 2:00 a.m. with a message saying: "I am very sorry to submit the draft two hours after the deadline, because I could work only after my kid went to sleep." The message affected the leaders and reinforced their feelings of responsibility to help the members of the PLC-FLTs.
Cognitive pushing refers to high demands related to the quality of the finished task. For example, a single paper would normally be revised at least ten times by a PLC-FLTs member and the leadership together. The revisions generally moved from the macro-level to the micro-level, and from the content to the language. Finally, all the references provided and listed in each paper would be checked for validity by the collective leadership. The leadership also sought feedback from the PLC-FLTs members, for example, if there were points that were not quite clear in their papers. In these situations the leadership needed to read more related literature and gather more information before they could provide proper feedback.
In summary, the two "pull" and "push" forces form a mechanism that is an indivisible unit in our model. They work alternately, and simultaneously influence the running of the community. At the very beginning of the PLC-FLTs the members were first pulled, and then pushed. However, once the PLC-FLTs started functioning, pulling and pushing did not happen in a linear or exclusive manner. They started to form a cyclical process in which pulling and pushing were intertwined. In fact, when pulling and pushing were no longer distinguishable, this was the moment when magic results were achieved.

Working process of the PLC-FLTs
The following section illustrates the working process of the PLC-FLTs (see Figure 3), also using the research project on political discourse analyses of national leaders' speeches at the UN General Assembly as an example. As Figure 3 illustrates, there are three phases in the whole working process of the PLC-FLTs. The first phase focuses on selecting a topic and drafting the manuscript; the second phase is primarily concerned with revising the manuscript; and the third phase places strong emphasis on submitting the manuscript and responding to critical comments offered by reviewers. Every phase involves three major tasks which were presented in a linear sequence. In fact, each pair of tasks in each phase involved several cycles and these relationships were interactive in nature (indicated by a double-headed arrow in the figure). Similarly, the three phases were not clear-cut, but often interacted with each other. However, a verbal description of each phase and each task within each phase is easiest to express in a linear fashion. We hope the following illustration does not lead to misinterpretation of the working process.

Phase I: Choosing a topic and drafting the manuscript
Phase I included three main tasks: 1) reading the related literature and discussing each individual's possible topics for the research project; 2) analyzing data individually and determining the focus of the analysis; and 3) drafting manuscripts. Among the three tasks, the first was the most demanding, since the PLC-FLTs members came from different departments, taught diverse languages, and had varied academic backgrounds. Clearly, it was difficult to find one project which could fit with each member's interests, abilities, and research backgrounds. Therefore, the initial two or three months were allocated to brainstorming sessions on feasible projects, engaging all the PLC-FLTs members.
To enhance the efficiency of the discussions, the collective leadership usually held a short preparatory meeting in advance, at which resources were shared and a proposed agenda was set for the upcoming PLC-FLTs meeting. Take a PLC-FLTs meeting held in December 2015, for example. The PLC-FLTs leaders spent a lot of time in selecting a project out of several options, making sure it could be carried out by all the PLC-FLTs members. The provisional proposal put forward by the PLC leaders was to analyze the speeches of national leaders from different countries.
At the PLC-FLTs meeting in January 2016 when the project was presented, there was a heated discussion during which some critical questions were put forward: "Who can be regarded as the leader of a country, the president or the premier? What kind of speech is worth analyzing? What occasions produce the speeches that are the most suitable for the project? What is the purpose of such an analysis?" These questions indicated that the PLC-FLTs members were interested in the project, but were confused about its significance and the specific choice of a speech.
After this discussion, initial agreement was reached on a number of issues. First, the speeches by different leaders should either center on the same topic (e.g., celebrating New Year), or be given at the same occasion (e.g., the UN General Assembly). Second, the project will be of significance in the sense that it may help us to arrive at a better understanding of different beliefs and values, and to investigate the discourse characteristics of the leaders concerned. However, the PLC-FLTs members were not ready to make a final decision about the speeches for analysis. The PLC leaders did not push them, recognizing that it was hard to agree on a topic or an occasion. Instead, it was suggested that they could each choose one leader's speech in their own mother tongue and translate it into Chinese for a discussion at the next meeting, which was scheduled to take place after the winter vacation.
Once the group leaders had received a number of translated speeches in Chinese, they had another preparatory meeting to make an initial analysis. They immediately realized that speeches on the same topic were not comparable. Take New Year speeches, for example-in these speeches the leaders of different countries were addressing diversified national issues of their own, which hardly shared anything in common. Instead, speeches on the same occasion, such as multilateral conferences, seemed to be more promising for comparison. The second preparatory meeting proposed an analysis of leaders' speeches at the UN General Assembly in 2015, which took place on the 70 th anniversary of the UN.
This proposal was fully supported by all the PLC-FLTs members at a meeting in March 2016. It had a couple of advantages: first, it removed the difficulties of choosing one speech for analysis for all members; second, this project enabled the PLC-FLTs to help and learn from one another, since their tasks had a lot in common in terms of data analysis. After the meeting, the PLC members started to analyze the discourse and draft their manuscripts.

Phase II: Discussing and revising the manuscript
Phase II was critical and essential for the development of the PLC-FLTs. Once the participants entered this phase they were less likely to withdraw from the PLC. Therefore, Phase II required more efforts from both the leaders and the voluntary participants. As Figure 3 illustrates, Phase II involved three steps: 1) collaborative revisions of a few samples; 2) individual revisions and peer revisions; and 3) revisions guided by the PLC-FLTs leaders through a process of face to-face mentoring.
Collaborative revision of drafts was a good learning opportunity for the PLC-FLTs members. In this process, a few good-quality manuscripts that had already been completed were selected by the leaders, and then the authors shared their experiences and difficulties during the writing process. During the discussion, all participants were able to raise questions or offer suggestions for revision. For example, the teacher of Japanese with a PhD degree in History first demonstrated how he analyzed the speech by Shinzou Abe at the UN from a historical perspective. Then he reported that this project had been "really difficult" for him because of his lack of linguistic expertise. In his presentation his profound knowledge of Japanese history impressed all other PLC members and leaders. However, the challenge he experienced involved shifting his perspective from historical to linguistic, so that the paper might be accepted by a linguistic journal. The discussion resulted in two suggestions: first, he could employ the strategy of "erasing" (Stibbe 2014) in discourse analysis; second, he could use his historical knowledge to discuss the reasons why such a strategy was employed by Shinzou Abe. The collaborative revision session gave him more confidence in revising the paper. After he finished his own revision the manuscript was checked once more by other members, and finally the leaders read his manuscript and gave suggestions for final revisions.
There were often several rounds of the first and second task before the members moved on to the third task in this stage. Take the manuscript written by the teacher of Sinhalese (Jiang 2017) as an example. This manuscript was revised 11 times concerning the title, the theoretical framework, content, wording, abstract, and references. During this process the leaders alternatively pulled and pushed the teacher of Sinhalese, increasing her confidence and making her determined to get her paper published. The following interview data demonstrates this process: It was a challenging job and I wanted to give up at the very beginning of the revision. I remember it was almost the deadline of 12 am when I finished preparing my lesson for the next day. Later I thought that I had already spent much time writing and I shouldn't give up. What about trying it once more? Then I tried my best to revise my manuscript based on the feedback provided by the PLC leader, and sent the revised version to the PLC at 3 am via email. Then the PLC leadership gave me detailed suggestions for further revisions. The feedback was encouraging and became a driving force for me to keep working at the manuscript. (interview) (Wen, Ren, & Zhang 2021) Like the teacher of Sinhalese, all the PLC-FLTs members went through a long process of drafting and revising. During Phase II, pulling and pushing worked alternately both at the cognitive and emotional levels. In response to members' difficulties, the PLC leaders provided one-to-one step by step guidance in addition to collaborative discussions and peer revisions. The PLC-FLTs leaders also encouraged members to make their own efforts first, and then to seek support from the people around them. By doing so, the PLC-FLTs members gradually became more and more independent.

Phase III: Submitting the manuscript
Phase III almost marked the ending of the research project that would decide whether the PLC project would bring visible achievements to the PLC-FLTs members. Figure 3 lists three main tasks: 1) the PLC leaders invited external experts to review the manuscripts; 2) the PLC leaders advised and helped the members to submit their manuscripts to journals; and 3) collaborative revising in light of the reviewers' feedback.
To ensure the quality of the finished papers, experts outside the PLC were invited to read the manuscripts and provide feedback, and the PLC leaders also helped to check the format of the papers before submission. The second task, organizing the papers for submission to journals, was very challenging and demanding. The major reason was that key language journals in China seldom publish papers related to foreign languages other than English, because these papers are less likely cited by other authors. To overcome this difficulty, the PLC leaders tried to organize papers to relate to new and relevant topics. For example, our first project generated more than 10 papers on the national images projected by UN General Assembly speeches. The topic was appealing to journal editors because it opened a new field of research for language teachers. They believed that if more people carried out research on national image projection, these papers would certainly be cited.
After receiving major or minor revision feedback from the journals, the PLC leaders and the members met to discuss how to polish the papers, and then the members went back to revise on their own. Such discussions and revisions sometimes proceeded over several rounds until the paper's acceptance.
Many of the PLC-FLTs members had little experience in writing an academic paper, and therefore one of the major jobs for the leaders was to boost their confidence and stimulate their passion for research. For example, the teacher of German admitted that she "had no experience of writing academic papers, had not published anything academic on her own, did not know how to organize an academic paper, and did not know what words to be selected" (interview). However, at the end of the first project she found to her surprise that she "could write something and got it published" (researchers' research journal). This demonstrated that the PLC-FLTs had the potential to ameliorate its members' initial fears and anxieties about academic research and writing.
In sum, the PLC-FLTs members started from scratch and learned to do academic research with advice and assistance from the leadership. Along with growing numbers of research projects completed, the PLC-FLTs members' confidence in and independence of conducting research progressively increased, while the outside support and guidance they needed gradually reduced.

Reflections
The paper will conclude with our reflections on why the PLC-FLTs has been so successful over its four years. This study records how a PLC-FLTs was successfully created and sustained by following the model presented in Figure 1. Reflecting on our practice, we have further realized that it is crucial to have a sound understanding of the essence of the model, so as to ensure that a new PLC runs on the right track. The essence of the model, in our opinion, lies in the high quality of the collective leadership, combined with strategic management of the four elements in the model. Moreover, sustaining a PLC in HE still faces a number of challenges which need to be dealt with.

Collective leadership as a key to the success of the PLC-FLTs
Among the four elements of the PLC, the participants, consisting of the leadership and the members, were at the center of the framework. A closer look at these participants shows that the collective leadership was the backbone for the success of our PLC-FLTs. We, as leaders, held the firm belief that we should be accountable for the progress of the community. The leadership consisted of three people who were all enthusiastic about the work, and who all had a firm conviction and great passion. By engaging in collective leadership, the three leaders' wisdom was fully tapped so that the plans made at preparatory meetings were comprehensive, workable and could easily gain support from the whole PLC. Furthermore, responsibilities were shared and fulfilled satisfactorily. Last but not the least, as the leaders, we also had our own learning and research agendas. In other words, we grew together with all the other PLC members while producing research output concerning the research project and the building of PLCs (Wen & Zhang 2019;Zhang & Wen 2020). In a word, the relationship between the leaders and the members was mutually beneficial.
As a Chinese saying goes, "it is the engine that decides the speed of the train." Similarly, the success of a PLC largely depends on its leaders. In fact, the leaders not only played a leading role, but also acted as role models. Both the words and the deeds of the leaders had an enormous impact on the PLC-FLTs members. Our conscientious and concerted efforts to discuss the research plan, the drafts and revisions of research papers, and so on, strongly motivated the members to conduct research while also making them determined to do similar things to help younger teachers in the future (Zhang & Wen 2020).

Strategic management of the four elements
Strategic management of the four elements refers to a careful top-down design of the development of the PLC-FLTs, which requires the PLC leaders to capture the crucial features and details of each element so that it may be strategically dealt with. A successful PLC is not an accidental gathering of a group of people who have complete freedom to come and go. Therefore, a question arises of how to establish requirements for disciplined behaviors that will be accepted by all the PLC-FLTs members without making them feel uncomfortable.
To do this, we adopted two strategies. First, voluntary participation was guaranteed as a prerequisite, and a detailed explanation of the objectives, mediators, and mechanism was presented by the PLC leaders before they formally joined the PLC-FLTs. Second, a contract had to be signed between the leaders and the PLC-FLTs members, after they were informed of the specific content in the contract. The contract stated clearly the leaders' obligations, as well as requirements for the members. Clearly, everyone's decision about participation was free and without pressure. However, once they had signed a contract, they were required to keep their promises.
We felt that the contract was a formal bond, but it was not sufficient to enforce the expected disciplined behaviors to follow. Strategically, when we discussed the objectives with them, we started with the short-term ones since all the participants were eager to learn how to do research and how to write a research paper. For the long-term objectives, we carefully embedded them in the short-term ones. For example, the members were assigned to do research in collaboration with other PLC members so as to cultivate their cooperative competence. As illustrated in Figure 3, in the second phase all the PLC-FLTs members formed groups based on their similar research topics. They were expected to provide feedback on one another's manuscripts, and engage in professional discussions on their own and others' research. Through these activities, they could facilitate their own learning while a helpful and trusting PLC culture gradually evolved.
The group's objectives cannot be realized without mediators. From the four years of practice, we have recognized that identifying a single suitable research project at the beginning is enormously important. For example, the first project, asking the PLC members to analyze speeches from the UN General Assembly in 2015, could engage every member of the group. This meant that they were all able to share their experiences as well as their difficulties. The project thus acted as an appetizer for later projects, and an effective mediator that reduced their anxiety. After the first project was finished, the atmosphere of mutual trust and benefit was created.
The pull-push mechanism acts as a monitoring system. Without such a system, even if everything else was in place, the PLC might not have worked as successfully as we hoped. Pulling coupled with pushing is easy to set out, but difficult to implement. Two strategies were very useful in our practice. The pull-push mechanism caters to individual differences, but this also means that the amount of force that should be exerted in pulling or pushing is extremely complex and delicate. The leaders observed the PLC-FLTs members carefully so as to balance their pulling and pushing. Second, the leaders would pay special attention to the timing of pulling and pushing. Pulling would normally occur at the beginning of a new task when the PLC-FLTs members did not know what to do, or in the process of carrying out the task when they felt frustrated by seemingly insurmountable difficulties. They would be pushed when they were not strict with themselves, or if they became complacent about their work.

Suggestions for tackling the challenges involved in sustaining the PLC-FLTs
In spite of successfully sustaining the PLC-FLTs for more than four years, we still face some challenges. First, the majority of our PLC-FLTs members were young mothers with small children to look after. They were thus often torn between their roles: PLC-FLTs assignments, teaching and administration in their own departments, and their household responsibilities. Because of this overload, they sometimes could not meet deadlines for the PLC assignments. Second, it was difficult to find a suitable time for all the PLC-FLTs members to attend a collective meeting, since they all had different schedules.
To solve these problems, we offer the following suggestions. First, the PLC leaders should flexibly deal with the degree of pulling and pushing. For example, members with small children or babies were allowed to hand in their assigned work slightly behind schedule. Second, the PLC leaders should model dedication; if no suitable time was available for all the members, the PLC leaders chose weekends or holidays for collective meetings. Sometimes the leaders would also offer time for one-to-one mentoring. Third, the PLC members should play an agentive role in the PLC, since without their agency and commitment, no matter how much force was exerted in pulling and pushing, the PLC would not survive.

Conclusions
The productive and collaborative research carried out by the PLC-FLTs members generally indicates that our model is applicable and effective. We suggest that the four elements (PARTICIPANTS, OBJECTIVES, MEDIATORS, and MECHANISM) should be strategically managed for a number of reasons. First, they are indispensable, but they do not work in isolation. Second, they need to be considered sequentially when building PLCs; normally their order should not be changed. That is to say, PARTICIPANTS serve as the starting point, OBJECTIVES as shared goals, MEDIATORS as the means for achieving the objectives, and MECHANISM as a driving force. It is suggested that the above recommendations be more suitable for voluntary PLCs in higher education contexts.
Thanks to the advancement of technology, the first author of the paper has founded a virtual PLC which has recruited about 140 volunteers from 90 universities across China, with the second author of the paper acting as a core member of the collective leadership group. We hope that the development of this virtual PLC will further test and enrich our current model, and that more virtual PLCs will be formed in the future.