A Review of Current Sociological Research on Teachers’ Emotions: The Way Forward

During the mid-1990s, teachers’ emotions emerged as an area of research in the sociology of teaching because many teachers all over the world were reported to be unhappy, dissatisfied, stressed, frustrated, and even alienated. This implies that teachers’ emotions, especially negative ones, go beyond individual factors and have become a social issue. Therefore, researchers use sociological perspectives and theories to deepen our understanding of the phenomenon. In order to advance the development of this newly-emerged field in sociology, this article reviews the sociological theories commonly applied to study teachers’ emotions: labor process theory, emotional labor theory, and theory of emotional geographies. It is suggested that the theories focus on either structural or agential influences to teachers’ emotions. It may limit their explanation about teachers’ emotions because emotions are constructed and conditioned by teacher agency and socio cultural structure. Thus, this article proposes a sociological framework which considers both structural and agential effects to teachers’ emotions. Moreover, a research agenda is also addressed in this article.


INTRODUCTION
In the mid-1990s, teachers' emotions emerged as a field of research in the sociology of teaching [1]. One likely reason for this was that more and more teachers all over the world had been identified as feeling stressed, depressive, frustrated, dissatisfied, unhappy, and even alienated [2,3]. In other words, teachers' emotions, especially negative emotions, should be considered as a social rather than a psychological issue. Thus, sociology is required to deepen our understanding of the phenomenon. Psychologists who view teachers' emotions as psychologically constructed. For example, they explain teachers' emotions by teachers' personality traits, thinking styles, emotional intelligence and competence, coping strategies with job stress [4]. On the other hand, sociologists generally argue that teachers' emotions are socially constructed [5]. That means sociologists regard how teachers feel at work is conditioned and shaped by certain social forces, such as culture, social relationship, and social system. Although this understanding of teachers' emotions is generally agreed in the literature, the studies identify different processes of social construction of teachers' emotions because of different theories or perspectives applied in the investigations, including labor process theory, emotional labor theory, and theory of emotional geographies. Thus, the aim of this article is to review the sociological theories in the research on teachers' emotions in attempts to advance the development of the sociological research that focuses on emotional phenomena of teaching. This article suggests that the explanations about teachers' emotions provided by the existing theories may be incomplete and segmented, because all the theories only consider about either structural or agential influences to teachers' emotions. Nevertheless, teachers' emotions should be the phenomenon constructed by both teacher agency and socio cultural structure. Therefore, at the end of this article some theoretical considerations for further studies will be highlighted in attempt to overcome the limitation.

LABOR PROCESS THEORY
Since the 1980s, sociologists of education have criticized education reform all over the world for its tendency to transform the labor and labor process of teachers in such as a way that they resemble those of industrial workers, resulting in poor working conditions and lives for teachers, such as heavy workload and the lack of leisure time [6][7][8]. In order to improve the situation, sociologists of education have investigated how education reform transforms labor and the labor process in teaching. This kind of research is generally called labor process theory [9]. Although labor process theory does not directly deal with teachers' emotions, its concern about the labor process in teaching offers insights into how teachers feel is conditioned by education reform.
According to labor process theory, many states in the world have attempted to increase external control over teachers' work in order to promote high quality of education, which in turn supports the development of the state [9]. The studies have found that states may increase external control through two transformation processes of labor: deskilling and intensification [6,9,10]. Deskilling is the process of devaluing and deprofessionalizing teachers' work, which in turn results in teachers being unable to define and design what they do at work. Researchers argue that managerialist education reforms may contribute to this process, because this kind of reforms tries to introduce market logics for the operation of state education [11]. For example, managerialist education reform emphasizes concepts such as quality, accountability, competition, efficiency, clients, and quality assurance [12]. A notable outcome of this kind of reform is that teachers are subject to more and more external supervision and control from the government and the community [7]. One notable example is Hong Kong. Since the late-1990s, the Hong Kong government has intensified its inspection and supervision of schools and teachers by implementing several initiatives such as: School Self Evaluation (SSE) and the External School Review (ESR) in 2003, the Quality Assurance Inspection (QAI) in 1997, and Language Benchmarks Tests for teachers in 2000 [12]. Since the teachers are subject to external control, they may become less able to control and design their teaching process in the classroom [13]. For example, they may be forced to train students' academic skills rather than to foster their intellectual, social and moral development, because the education reforms may narrowly define educational quality and effectiveness by students' results in public examinations [14]. In order to ensure students' examination result, the teachers may not be allowed to teach something outside curriculum and syllabus designed by the government and education experts [7]. As a result, they may experience negative feelings like frustration, powerlessness and meaninglessness at work because they may disagree with the narrow conception of teaching and education [9].
When teachers are deskilled, it becomes difficult for them to reject the extra duties and workload imposed upon them [10]. Thus, labor process theory claims, intensification of work is inevitable for the teaching profession during managerialist education reforms [10]. For example, a study conducted by the OECD [15] reported that teachers in the OECD countries were required to take on many responsibilities in addition to classroom teaching, such as guidance and discipline, organization of extracurricular activities, preparation of schoolbased teaching and learning materials, management of the school's public image, documentation, and writing reports for school internal and external inspection, and other administrative duties. As a result of so many duties and responsibilities, teachers may face work overload, lack leisure time, and feel stressed, and burnt out [16,17]. Moreover, studies suggest that intensification also causes teachers to experience certain negative emotions at work, such as guilt, frustration, anxiety, and meaninglessness, because they are forced to do many tasks and duties (e.g. paperwork and documentation) that are less related to teaching and education or have less educational value [2,18,19]. As a result, teachers may feel they are alienated from the work [20].
In sum, the studies on education reforms and labor process in teaching suggest that education reform may deskill and intensify teachers' work. Deskilling may result in teachers losing their control over their labor and the labor process of teaching, in which case intensification may occur. Intensification can lead to poor working conditions for teachers, such as heavy workload and lack of leisure time, so teachers may become stressed and burnt out at work. Although the studies do not directly discuss the issue of teachers' emotions, it does imply that education reforms may arouse teachers' negative emotions through deskilling and intensification. This is because the two processes together may lead to teachers being unable to define and design what they do in schools and therefore being forced to do a lot of work and duties that they think are irrelevant to teaching and that they dislike. As a result, they may feel dissatisfied, frustrated, powerless, and meaningless, and even alienated when they do their work. Nevertheless, the research on teachers' emotions from the perspective of labor process theory has a problem. It is that the research and the theory imply that teachers' emotions are purely a product of education reforms. However, many sociologists point out that emotions are not purely constructed by macro-social forces like education reform, but also by social actors themselves [21][22][23]. They claim that social actors are reflexive agents who can monitor and shape their feelings and emotions through the interpretation of the self and situation. In other words, a research that does not consider teacher agency in shaping emotions of teachers may not offer a complete account and explanation about how teachers feel at work [24]. Thus, current sociological research on teachers' emotions shifts the theoretical position from labor process theory to those that consider teacher agency.

EMOTINOAL LABOR THEORY
One of the prevalent research interests concerning teachers' emotions is emotional labor in teaching. To some extent, this kind of research aims to explore the nature of teaching and its impact on teachers' psychological and emotional well-being. Research on emotional labor in teaching is inspired by Hoch child's [25,26]work, which is influenced by dramaturgical theory and alienation theory. On the basis of the dramaturgical theory, Hoch's child notes that there is an emotional culture that consists of feeling and expression rules specifying how we should feel and display our feelings in every social setting. For example, feeling and expression rules specify that people should be sad and should not smile at funerals, but that people should be happy and should not cry at weddings [27]. If social actors cannot adjust their feelings and displays according to feeling and expression rules, they will be perceived as emotional deviants by others [28]. To avoid becoming emotional deviants, people need to manage their emotions and displays appropriately. Hoch child refers to emotion management as emotion work and identifies two strategies of emotion work: surface acting and deep acting. Surface acting involves modifying one's emotional display in accordance with what is expected in a particular situation regardless of one`s actual feelings, whereas deep acting involve strying to change one's feelings to match the appropriate emotional display [25].
To some extent, emotion work is a normal act occurring in social actors' private lives. However, from the perspective of alienation theory, Hoch child [26] points out that emotion management is not only an act occurring in one's private life, but also the labor done for a wage in post-industrial societies. She found that more and more enterprises, especially service-related, tend to sell employees' emotional activities for profit making. In such a situation, employees are no longer able to exercise control over their feelings and displays. For example, Hoch child [26] illustrates how flight attendants are required by their employers to keep smiling and show warmth towards consumers because smiling and warmth are the selling points of airlines. Other studies have also had similar findings among other frontline service workers (e.g. waitresses and insurance sales), care giving workers (e.g. retail clerks and child care workers), professionals (e.g. physicians and lawyers), and public service workers (e.g. social workers and corrections officials) [29]. In other words, many people in post-industrial societies have to manage their emotions under supervision. Hochschild [26] refers to this kind of act as emotional labor: emotion management done for a wage. One possible consequence of emotional labor is emotional dissonance, which is the separation of feelings from displays [26].The higher the degree of emotional dissonance, the higher the degree of dehumanization, self-alienation, depersonalization, depression, and burnout [26,30,31].
Generally, sociological studies on emotional labor suggest that it is more likely that workers who are required to engage in face-to-face interactions with the public with love and care perform emotional labor [29]. Accordingly, educational researchers have argued that teachers are required to perform emotional labor because they need to interact face-to-face with students with love and care [32]. On the basis of the theory of emotional labor and management, the first task for educational researchers is to illustrate how teachers lose their control over their emotions in teaching. Tsang [33]has found that while there may not be explicit supervision over teachers' emotions in teaching, teachers' emotions are prescribed by different feeling and expression rules of teaching. According to Zembylas [34], the general rule requires teachers to avoid expressing too strong and too weak emotions. More specifically, Winograd [35] reveals five feeling and expression rules for teachers:  to love and to show enthusiasm for students;  to be enthusiastic and passionate about subject matter;  to avoid the display of extreme emotions like anger, joy and sadness;  to love their work; and  to have a sense of humor and laugh at their own mistakes and the peccadilloes of students.
The rules may also be related to teacher professionalism. This means that if teachers do not manage their emotions appropriately according to the rules, they will be treated as unprofessional [34,36]. Consequently, teachers have to perform emotional labor. Basically, this argument is accepted by most of education researchers. Since they agree that teachers need to perform emotional labor, they investigate the outcomes of the emotional labor in teaching with respect to teachers' psychological well-being and mentality.
According to the alienation theory, emotional labor is alienating [37]. In other words, emotional labor creates emotional dissonance resulting in job stress, emotional exhaustion, burnout, or other negative outcomes. Many empirically studies provide support for this proposition. For example, the survey studies conducted by Hu¨lsheger, et al. [38] and Iltaf and Gulzar [39]show that emotional labor may cause teacher stress and burnout, a sense of depersonalization, and turnover. In addition, the ethnographic studies of Zembylas [40,41] resulted in similar findings. These studies demonstrated that emotional labor in teaching may result in teachers' feelings of frustration, guilt, and shame, which in turn damage teachers' identity, confidence and self-esteem. For example, a teacher may dislike or even hate a student whose academic performance is bad or whose misconduct is serious. However, the teacher needs not only to suppress his or her negative emotions, but also to show love and care to the students because of the feeling rules. The suppression of negative emotions may create further negative emotions for teachers, such as guilt, regret, and shame, because they may think that it is inappropriate or even immoral for them as teachers to dislike or hate students [42]. In addition, teachers may also feel emotionally uncomfortable or selfestranged, because their displayed emotions do not correspond to their true feelings [43]. Both conditions may affect their professional identity and self-esteem, which in turn creates other intense negative emotions like frustration and depression [42].
Emotional labor theory and its research on teachers' emotions suggest that the nature of teaching as an interactive work with love and care that involves emotional culture consisting of feeling and expression rules. The emotional culture may constrain teachers' feelings and displays of feelings at work, and this constraint may create emotional dissonance resulting in negative emotional experiences and outcomes. Although the theory and research recognize teachers are able to construct their emotions through emotion management, it tends to regard emotion management as a less reflexive or normative behavior: emotion management in teaching is a behavior reacting or conforming to emotional culture of teaching. In other words, the theory and research still pay less attention to teacher agency in the analysis of teachers' emotions, although teacher agency is recognized. This weakness may affect the explanative power or validity of the theory. For example, more and more studies show that teachers may feel satisfaction, excitement, and even self-fulfillment when they successfully manage their emotions at work, because they interpret that managing emotions in teaching successfully helps them to teach and to foster students' growth [42,44,45]. The reason why preforming emotion management may arouse positive emotions is that emotion management in teaching is a voluntary practice rather than a forced labor or normative behavior [46]. That means teachers initiatively choose to regulate and manage their feelings and displays in order to facilitate students to learn and grow. For example, Yuu [47] found that Japanese teachers may choose to display anger to students, even though they are not angry, because they think such emotion management is an effective means to control students and help them concentrate on learning. Oplatka's [46] study on Israeli teachers' emotional experiences also suggested that teachers may unconditionally express and show their love and care to students because the emotional displays are meaningful and valuable for them in developing relationships with students and in fostering students' growth. His more recent research suggests that if teachers do not manage their emotions in order to facilitate students' learning and growth, they may perceive themselves as immoral and unethical, resulting in negative emotions such as guilt and shame [45]. As a result, sociologists of teaching develop another theory of teachers' emotions which emphasizes the effects of teacher agency. The theory is known as emotional geographies descried as the follows.

THEORY OF EOMTIONAL GEOGRAPHIES
In the 1990s, collaboration between teachers, parents, and students in teaching and learning was perceived as having many advantages and benefits [8]. Therefore, educational researchers, like Andy Hargreaves, have attempted to understand what makes the collaboration successful or unsuccessful at that time. They find that the emotional dimension of social interaction may be a factor. Moreover, they also identify teachers' emotions are aroused by teachers' interpretation of the social interaction and influence the social interaction. On the basis of the research findings, Hargreaves [5,[48][49][50][51][52] develops the theory of emotional geographies to understand teachers' emotions and social interaction in school settings based on the perspective of symbolic interactionism.
From symbolic interactionism, Hargreaves [53] believes that emotions are constructed by the social interactions of teachers with students, parents, and colleagues, and that in turn the emotions will produce and reproduce the social relationships and interactions. Therefore, he has attempted to investigate the relationship between teachers' emotions and social interactions in school settings in order to improve the quality of teaching and learning. More specifically, he developed his theoretical framework of teachers' emotions from Denzin's [27] symbolic interactionist and phenomenological theory of emotions. According to Denzin's theory, Hargreaves [48] conceptualizes teaching as emotional practice because teaching will activate, color and express not only teachers' feelings and actions, but also the feelings and actions of those with whom teachers interact. As an emotional practice, teaching requires extensive degrees of emotional understanding [53], which is the capacity to interpret and comprehend subjectively other people's emotions from one's own standpoint [27]. Without emotional understanding, mistrust and conflicts may permeate social relations between teachers and other school members, and in turn arouse negative emotions in them, which then affect social relationships. In order to facilitate social relationships and interaction, it is important to enhance emotional understanding between the interacting parties [27]. To enhance emotional understanding, Hargreaves proposed the concept of emotional geographies of teaching, which refer to "the spatial and experiential patterns of closeness and/or distance in human interactions and relationships" [50]. The emotional geographies of teaching include five dimensions [50]:  Sociocultural geography: the similarities and differences in cultural and social backgrounds between teachers and those with whom they interact;  Moral geography: the degree to which teachers' purposes are supported or appreciated by others;  Professional geography: the degree to which teachers hold the norms of professional interaction that prescribe coolness, reserve and emotional distance among actors;  Political geography: the differences in power and status between teachers and those with whom they interact;  Physical geography: the frequency, intensity and formalization of interactions of teachers with others . Hargreaves points out that varying the degree of each dimension will influence the overall pattern of closeness/distance in social interactions and relations. As a result, emotional understanding will also be affected. Based on this framework, Hargreaves investigated teachers' emotions in different kinds of social interactions by conducting large scale ethnographic studies in which he observed 15 varied schools of different levels and sizes and also interviewed more than 60 elementary and secondary teachers in Canada.
With respect to teacher-student interaction, Hargreaves [49] found that the elementary teachers tended to possess professional warmth and interacted frequently with students. According to Hargreaves, this may enhance the physical and professional closeness between teachers and students, and this closeness, to some extent, is the basis for the development of emotional understanding between teachers and students. On the other hand, Hargreaves also argues that the unequal social position between these two groups of people (i.e. teachers are dominant and students are dominated in the classroom) may create conflicts between them (political distance). This conflict may prevent them from developing shared emotional goals and emotional bonds, so emotional misunderstanding may occur among them [49]. However, Hargreaves [49]thinks that this political distance can be minimized if teachers and students can have more activities outside the core process of teaching and learning in classroom. This is because such activities make possible shared positive emotional experiences for both teachers and students, such as exhilaration and enjoyment.
Regarding teacher-parent relationships, Hargreaves [50] shows that teachers and parents are emotionally distant. This emotional distance is to some extent the result of (1) the difference in socio-cultural backgrounds between teachers and parents (sociocultural distance), and (2) the norm of teacher-parent interaction that prescribes coolness, reserve and emotional distance (professional distance). Sociocultural and professional distance prevents teachers and parents from developing shared understanding cognitively and emotionally. This may further foster moral distance between them, which means that parents may not understand, support, appreciate and respect how teachers carry out their classroom teaching. Consequently, teachers may feel that they have lost their professional status and power/authority (political distance). In such a situation, teachers may experience negative feelings such as anger, resignation, depression, and anxiety. In this case, it is possible that they either avoid interactions with parents or display emotions like hostility, anger and dissatisfaction towards parents. As a result, the conflicts and emotional distance between teachers and parents become deeper, and the emotional understanding between them also becomes weaker [50].
Finally, Hargreaves [51,52] has found that teachers tend to value peaceful working environments in which they may receive and enjoy rewards such as more social support and acceptance, so they should try to maintain harmonious relationships and avoid conflicts with colleagues. To achieve this, many teachers try to respect or value what their colleagues do in order to shorten moral distance [51]. In addition, they also try not to criticize their colleagues because this act may downplay colleagues' professional status and power/authority [51]. In other words, they try to maintain political closeness. As a result, emotional understanding and positive social relationship among them may be developed. Hargreaves [52] points out that if such emotional understanding and relationship cannot be fostered, it is possible for mistrust and betrayal to occur.
As we have seen, Hargreaves illustrates in detail the emotional lives of teachers in face-toface interactions. From his studies, we understand that teachers' emotions may be determined by the inter subjectivity between teachers, students, parents and colleagues. The inter subjectivity may be influenced by the socio cultural, moral, professional, political, and physical distance between the social actors. On the other hand, teacher emotions may also affect further social interaction between teachers and those they interact with. Therefore, emotions can be viewed as a mediator in social processes at a micro-level [21]. Nevertheless, the weakness of this theory is that it overlooks the structural constraints on social interactions and feelings in school settings. Although social actors have reflexivity and agency to define the situations and in turn to behave, Kelchtermans [13] found that teachers face many structural constraints, like as school administration, school climate and curriculum, on their teaching practices and emotions. Moreover, the effects of structure and agency may interplay with each other and in turn affect teachers' emotions [54]. A more appropriate approach to investigate teachers' emotions should consider both effects of structure and teacher agency.

A PROPSED THEORETICAL FRAMEWORK
Labor process theory pays much attention to how education reforms transform and restructure teachers' labor process and in turn affect teachers' emotions. Emotional labor theory is concerned about how teachers feel is constructed and constrained by emotional culture of teaching which requires teachers to manage their emotions and displays in particular ways. Theory of emotional geographies in teaching focuses on how teachers' emotions shape and are shaped by social interactions in which each social actor has different definitions of the situation. Although each of the theories gives different explanations about the social construction of teachers' emotions, it seems that all the explanations are incomprehensive and segmented. This is because each of them only emphasizes either the structural or agential influence to emotions. However, emotions should be directly and indirectly shaped and conditioned by both structure and agency [54]. Therefore, analyses about teachers' emotions should consider both structural and agential effects. In order to overcome the shortcomings in the existing theories, I will approach teachers' emotional experiences at work by understanding the teaching purpose of teachers and social constraints on the fulfillment of the teaching purpose.

Teaching Purpose
From the perspective of symbolic interactionism, social actors will act toward objects on the base of the meanings they give to the objects [55]. In the literature, the meanings people give to the work always refer to work purpose, work goal, work value, work orientation, or work aspiration that all are interchangeable terms. Work purpose does not only imply what people understand about a work and occupation, but also what they want to achieve and fulfill in the work and occupation [56]. Thus, work purpose predisposes people to act in some ways with regard to the work [56]. Moreover, work purpose may also determine one's emotional experiences in workplace. This is because people may reflexively evaluate and interpret to what extent they achieve and fulfill the purpose in the work [57]. A positive evaluation and interpretation will arouse positive emotions and feelings such as satisfaction and self-fulfillment, but negative evaluation and interpretation will arouse negative emotions and feelings such as dissatisfaction and self-estrangement [57].
Therefore, in his classic study of schoolteachers, Lortie [58] points out that understanding the pattern of teachers' emotions in teaching needs a detail investigation of the meanings teachers attach to their work or the purpose teachers bring to teaching. The literature implies that most of teachers interpret teaching as an occupation to making a difference in children's lives, i.e., facilitating students' intellectual, social and moral development, and commit this meaning of teaching as their major purpose of teaching [18,19,58]. The teaching purpose to some extent becomes the object of teachers by which they reflexively evaluate and monitor their practices in teaching in order to fulfill the teaching purpose [24]. Thus, if they interpret what they do help them fulfill the purpose, they may feel positively; otherwise, they may feel negatively [59]. This is because the teachers may think they are good and competent when they perceive they fulfill the purpose, while they may think they are bad and incompetent if they perceive they fail to fulfill the propose [13]. To some extent, this proposition is also supported by the above reviewed theories and research. Research on education reform and the labor process of teaching implies that teachers value the work related to teaching and education, so they feel negatively if they are forced to do a lot of work unrelated to teaching and education. Research on emotional labor and management in teaching reveals that teachers will feel positively about managing their emotions if they think performing emotional labor or management is conducive to teaching and learning. Otherwise, they may feel unhappy and sad having to manage their emotions in such a way that they do not correspond to their true feelings. To some extent, all the studies from different sociological theories and studies have reached a similar and even the same conclusion that teachers' emotions may be related to the fulfillment of the purpose of teaching or the meaning they give to teaching -teaching is a profession to make a difference in students' lives.
However, research suggests that there are different purposes of teaching among teachers, such as interest in the relationship and interaction with children, love for children, interest in subject-matter field, making contribution to society, salary, stable working condition, no better choices, social reputation and social mobility [58,60]. The first five types of purpose together with the purpose of making a difference in students' lives are commonly known as intrinsic aspiration and the remaining as extrinsic aspiration [61]. Therefore, people who decide to teach may not aspire or only aspire to make a difference in students' lives. For example, research indicates that Hong Kong people chose to teach because of purely intrinsic aspiration, purely extrinsic aspiration, or mix of them. For example, Lai, Chan, Ko, and So's survey [62]shows that many secondary school students wanted to become teachers because they thought teaching as an occupation with good working hours, good job security, and longer holidays. On the other hand, Lam [63] discovers that many Hong Kong people who chose to teach had mixed types of purposes and aspirations. She also finds that some of Hong Kong student-teachers tended to commit to the intrinsic types of purpose or aspiration and some of them tended to commit to the extrinsic types of purpose or aspiration. Therefore, a first step to understand the pattern of teachers' emotional experiences at work is to explore the teaching purpose of the teachers and then explore how the teaching purpose shape the teachers' interpretation of the work as well as their emotions at work.

Social Constraints
It is assumed that teachers attempt to achieve and fulfill their teaching purpose in teaching [59]. However, Kelchtermans [13] illustrates that there are many social constraints on teachers' fulfillment of their teaching purpose. As mentioned above, labor process theory points out that education reform is a macro-social constraint affecting this fulfillment and another macro-social constraint is emotional culture in teaching in terms of the feeling and expression rules indicated by emotional labor theory. On the other hand, theory of emotional geographies suggests that social relationships and interactions among school members may be micro-social constraints.
However, a significant social constraint at meso-or instructional level is missed in the reviewed theories and research. The meso-social constraints should be school administration. Sociologists of school organizations have debated whether school administrations are bureaucratic ornon-bureaucratic systems [64]. A bureaucratic school system exercises many controls over teachers' work, but a non-bureaucratic school system provides teachers with a lot of autonomy. This is not to say that the sociological research on teachers' emotions should resolve this debate. Rather, it is suggested that the research should try to understand how teachers feel in different types of school administrative systems. It is commonly believed that bureaucratic system is associated with teachers' negative emotions, while non-bureaucratic systems is associated with teachers' positive emotions [64]. This is because teachers enjoy higher autonomy in loosely coupled system than tightly coupled system. However, it is the lack of investigation about how the two types of school administration affect the fulfillment of the teaching purpose of the teachers and in turn their emotions.
In addition, career stage may also be a social constraint that is less recognized in the research on teachers' emotions. The life history research on teachers' life shows that teachers in different career stages may have different teaching purposes and perspectives about their work and school life, because of their different experiences, roles, positions, and authorities across career stages [65,66].Since teachers' emotional experiences at work may relate to teaches' interpretation of their work and teaching purposes, career stages may also shape teachers' emotional experiences at work. However, there is limited investigation on the relationship. Moreover, the social constraints at all levels may not function alone. Theoretically, they may interact with one another and then affect teachers' emotions. Therefore, sociological research on teachers' emotions should also pay attention to how the social constraints interact and interplay and in turn shape and condition how teachers work and feel in teaching.
Accordingly, the present research will consider both agential and structural influences to teachers' emotional experiences at work. First, this study should focus on what is/are the purpose(s) teachers hold in teaching and how the purpose(s) affect teachers' teaching practices and interpretation of their work. Then, the study will also explore what social constraints the teachers encounter on the fulfillment of their purpose(s) and what is/are the emotional outcomes of the success and failure on the fulfillment caused by the social constraints.

RESEARCH AGENDA
Accordingly, researchers may be concerned with the following research questions in their research projects: (1) What are teaching purposes schoolteachers have? (2) How do the teaching purposes shape how schoolteachers work and feel at work? (3) What are social constraints on the fulfillment of teaching purposes in education system? (4) How do the social constraints interact with one another and affect teachers' emotions?
Researcher may regard the suggested research questions as starting points. The essence of these generic questions is opening up of possibilities for investigation and keeping the important issues about teachers' emotions in researchers' mind. When a research project unfolds, the researchers can formulate and ask more specific questions in order to deepen the analysis. As Van Manen [67] notes, truly research questions always come from researchers' experience in the field when the research is unfolding. For example, if researchers identify emotional culture as the most important social constraint on teachers' work and emotions, they may further consider the questions like: how is the emotional culture constructed? How do teachers learn and internalize the emotional culture? To what extent, do teachers resist or reconstruct the emotional culture? As a result, the research findings and analyses may deepen our understanding about the phenomena of teachers' emotions being studied.
In-depth qualitative studies may be particularly helpful in investigating the phenomenon of teachers' emotions, because the qualitative methods can provide us a deeper understanding about teachers' interpretations and purpose of teachers' work, the possible social constraints at different levels that affect the fulfillment of the teaching purpose, and about how the agency and social constraints influence their action and emotions in schools. Indeed, various qualitative methods can be used depending on the research purposes. For example, life history method may be suitable for the researchers who want to study the pattern of teachers' emotional experiences throughout their careers [68]; if researchers attempt to understand in detail how teachers feel in particular school settings, they may use participant observation or the case study method [69,70]; if researchers want to investigate teachers' emotional experiences in a broad range of settings (e.g. different types of schools), they may interview teachers who come from different school settings [71]. Nevertheless, it does not mean that quantitative data is irrelevant for investigation. This type of data may be useful if it is used to determine the distribution of teachers' definition or meaning of teachers' work, interpretation of working conditions and social constraints, and their feelings among the teaching population in order to assist in the exploration of the qualitative findings [72].
In our investigation, we should also consider the following factors: teachers' age, gender, and teaching experience. The reason is that these factors may mediate teachers' emotions [73]. For instance, teachers of different ages, genders, and teaching experience may have different duties and positions in schools and different interpretations about their work resulting in different emotional experiences [65,74]. As a result, research should select informants or cases for investigation with purposive sampling technique. This sampling technique may help in maximizing variation of cases and in identifying the information-rich cases for study in depth [75]. In this way, it will be possible to develop a better sense of the phenomenon.
To summarize, we should qualitatively study teachers' emotional experiences. This kind of investigation will provide us with thick descriptions about teachers' school lives and experiences. Through these thick descriptions, we will not only obtain a relatively clear and comprehensive understanding about what teachers' work, interpretation and emotions should be, but also develop propositions and identify the patterns of teachers' emotional experiences. Such findings will contribute to our understanding and further investigation into the phenomenon of teachers' emotional experiences.

CONCLUSION
In this article, three sociological theories about teachers' emotions are reviewed, including labor process theory, emotional labor theory, and theory of emotional geographies. Each of the theories considers either structural influences (e.g. education reform and emotional culture) or agential influences (e.g. definitions of situation) to teachers' emotions. However, teachers' emotions are constructed by both structure and agency so that this article argues the explanations provided by the reviewed sociological theories may be incomplete. In order to overcome the limitations, this article purposes a sociological framework to study teachers' emotions. This framework suggests that researchers should consider what teaching purpose teachers hold and how the teaching purpose influence their interpretation, teaching practices and emotions at work. On the other hand, the framework notes that there are certain social constraints which affect teachers' fulfillment of their teaching purpose and in turn their emotions at work. Thus, researchers should explore what and how social constraints condition teachers' work and emotions. Accordingly, qualitative methodology is recommended to study teachers' emotions, because the methodology enables us to have an in-depth investigation about what teaching purpose hold by teachers, how the teaching purpose influences teachers' interpretation, work and emotions, and what and how social constraints affect teachers' work and emotions.