Conceptualizing students’ learning experiences in English as second language in higher education from structure and agency

Abstract This paper narrates my research journey to conceptualize students’ learning experiences in English as Second Language program in higher education from structure and agency. It reviews literature of psychological and sociological research traditions that conceptualize students’ learning experiences. Using a critical perspective, each research tradition is reviewed in-depth for its strengths and limitations. However, the strengths and limitations are also seen from the perspective of my constructivist ontological and interpretivist epistemological stance. The paper critique the psychological research tradition of Phenomenography as well as the quantitative learning inventories descended from Phenomenography. It, then, focuses on Activity Theory, which is seen as a sophisticated analytical tool. However, the limitations of Activity Theory pave the way to move to sociological concepts of identity, community and institutional influences as theoretical lenses to conceptualize students’ learning experiences from structure and agency. These concepts are formed using Symbolic Interactionism, Community of Practices and Bourdieusian notions of Habitus, Field, and Capital, respectively.


ABOUT THE AUTHOR
Irfan Ahmed Rind is working as an associate professor and is the chair, department of Education at Sukkur IBA, Pakistan. He is also working as associate for the Center of International Education, University of Sussex, UK. He completed his PhD in Education and MA in Applied Linguistics from University of Sussex. He secured first position in BA and MA English Literature from University of Sindh and was awarded with gold medals. His teaching and research interests include methods and philosophy of research, theories in education research, development and education (MDGs and EFA), education for sustainable development, gender and leadership, monitoring and evaluation, assessment, education policy analysis and strategic planning, leadership and change management, teacher education and training, career counseling, and ESL. Irfan successfully executed different projects sponsored by British Council, USAID, Land O'Lakes, RSU, CIDA, and HEC. He has been the Lead Master Trainer in various Executive Training programs at Sukkur IBA.

PUBLIC INTEREST STATEMENT
Examining students' learning experiences in ESL programs has been of great interest for researchers, as the results have implications for the teachers, course-designers, and policy-makers. Traditionally, research in this area has been conducted from agentic (individuals) perspective, i.e. students' experiences are examined in the context of the class/program/institute. In such research, students are seen as mere learner with no other roles/identities. The other research traditions in this area focuses more on structures, i.e. how family/societal/institutional norms, values, and policies shape students' choices and actions. Students are seen as mere puppets that are played by structural forces. I have defined students' learning experiences in ESL from structure and agency, i.e. I acknowledge the structural forces that shape students' choices and actions, at the same time, I also believe that students use their agency in making their decisions. In so doing, I can get a better picture of how students experience learning in ESL programs.

Introduction
Investigating into students' learning experiences in educational programs in higher education has been of great interest to the educationalists, policy-makers, and program coordinators, since the findings can help in improving the policies, course designs, curriculums, and teaching approaches. However, most of the research in this area has been conducted using a predominated psychological perspective. Most of such research is quantitative and use learning inventories to investigate students' learning experiences by analyzing their learning approaches (Case & Gunstone †, 2003;Kember, Charlesworth, Davies, McKay, & Stott, 1997;Kurlin, 2009;Leung & Kember, 2003;MCcune & Hounsell, 2005). These studies analyze students' learning experiences from individualistic perspective ignoring contextual factors (Ashwin, 2009;Greasley & Ashworth, 2007;Haggis, 2003;Richardson, 2000). On the contrary, some studies use the social practice perspective and examine students' learning experiences from the social structure point of view alone (Berglund, 2002(Berglund, , 2004. Such studies portray students as mere puppets who have no agency, and their choices and actions are solely informed by structural forces. A very few studies adopt poststructuralist perspective to conceptualize students' learning experiences (Alhawsawi, 2014;Ashwin, 2009;Rind, 2015Rind & Alhawsawi, 2013;Rind & Kadiwal, 2016). In this paper, I offer a theoretical framework which may help exploring students' learning experiences from structure and agency. The paper reviews the relevant literature, offers an insight into different research traditions which are used to analyze teaching and learning in ESL in higher education. The paper narrates my interactions with different research perspectives and the way I have developed a theoretical framework from structure and agency. It narrates how I started my research journey using Approaches to Learning and Teaching (ALT) perspective, followed by Activity Theory and end up using the sociological concepts of identity, community, and institutional culture as theoretical lenses to understand students' learning experiences in ESL.

The approaches to learning and teaching (ALT) perspective
I started my research journey using, what Ashwin (2009) called, the ALT perspective. The ALT perspective has its foundation in the qualitative research tradition of Marton and Säljö's (1976) Phenomenography, which was later used to develop quantitative questionnaires by Entwistle and Ramsden (1983) and Biggs (1987).

Phenomenography
ALT emerged as a reaction to traditional, behavioristic approaches to learning second languages such as B. F. Skinner's Stimulus-Response Theory and Robert Lado's Contrastive Analysis Hypothesis (Ashwin, 2009). Marton and Säljö (1976) strongly reacted against the behavioristic notion of the fixity of learning with their phenomenographic approach. Marton and Booth (1997, p. 208) have concisely defined Phenomenography as "a change in someone's capability for experiencing something in certain ways." The phenomenographic view of learning developed as a reaction to earlier psychological research methods, which were seen as failing to provide insights into the way students "experience, conceptualize, perceive and understand various aspects of, and phenomena in, the world around them" (Marton, 1986, p. 31).
Phenomenographic investigation importantly moves away from the assumption that students' learning is a stable personality characteristic. It emphasizes the choices made by students in selecting approaches to their learning task, and indicated that the way in which students learn is a result of their interpretation of the learning context. The basic principle behind phenomenographic research is therefore that "learning should be seen as a qualitative change in a person's way of seeing, experiencing, understanding, conceptualizing something in the real world" (Marton & Ramsden, 1988, p. 271). Learning approaches (e.g. deep or surface) adopted by students can be taken as an indicator of whether or not meaningful learning has occurred (Kember et al., 1997).
However, Greasley and Ashworth (2007) criticize the contradictions inherent in Phenomenography. On the one hand, researchers recognize the need to focus on the students' experiences reflected on and discussed by the students themselves; on the other hand, the researchers understand students' accounts of their learning as falling into a number of categories (deep/surface), which few (if any) students would devise themselves. Moreover, Ashworth and Lucas (2000) and Richardson (2000) point out the lack of detail available on the actual, engaged process of phenomenographic research. They argue that, while there is considerable literature on the outcome of phenomenographic research, little attention is paid to the actual process of this form of research.

Quantitative learning inventories
Following the phenomenographic research carried out by Marton and Säljö (1976) and Pask (1976), Entwistle and Ramsden (1983) and Biggs (1987) adopted a quantitative approach to their studies, and introduced learning inventories in order to assess distinctions between different learning approaches. The main function of these inventories is to operationalize the various constructs which have emerged from qualitative insights from Phenomenography, and to generate quantitative scores on specific dimensions or scales which can show the different aspects of learning (Coffield, Moseley, Hall, & Ecclestone, 2004;Richardson, 2000). A few examples of such learning inventories are: Study Processes Questionnaire (Biggs, 1987); Revised Two-factor Study Processes Questionnaire (Biggs, Kember, & Leung, 2001), and Approaches to Studying Inventory (Entwistle & Ramsden, 1983).
The research carried out using these quantitative learning inventories focuses on students' perceptions of different variables, such as: teaching, assessment, workload, time to study, clarity and aims of syllabus/task, work space, learning resources, and interaction with peers and teachers. The findings conclude that students' perceptions of these variables significantly influence their learning experiences (Beattie, Collins, & McInnes, 1997;Matthews, 2003;Ramsden, 1979Ramsden, , 1984Ramsden, , 1987Watkins & Biggs, 2001).

Strength and limitations of ALT perspective
It is worth noting that the ALT perspective-and specifically the phenomenographic approach-has emerged in response to earlier psychological methods of research, which did not provide insights into how students "experience, conceptualize, perceive and understand various aspects of, and phenomena in the world around them" (Marton, 1986, p. 31). From the ALT perspective, importance is placed on the students (the agents), and the ways in which they process or change knowledge. ALT perspective, particularly phenomenographic research, foregrounds the concept of variability in learning a second language. Unlike the past behavioristic approaches to ESL learners, which perceive ESL learning as a mechanically fixed process, ALT and Phenomenography foreground ESL learners as individuals who have multiple ways of constructing their understanding of ESL and approaches to their learning. I was initially highly influenced by ALT perspective for two main reasons: its promotion of variability in learning, and the value placed upon ALT in higher education research, and the fact that researchers studying ESL contexts had found this approach particularly useful (e.g. Ming, 2004;Luas & Rojo-Laurilla, 2008;Xiu-juan, 2008). However, I do not adopt ALT as a theoretical framework in its entirety for two main reasons.
Firstly, the quantitative learning inventories approach conflicts with my interpretivist epistemological stance. In order to understand students' learning experiences in the ESL program in question, it is crucial to first interrogate my own assumptions about the nature of knowledge and the world, and how, as a researcher, I express or interpret reality. This process of interrogation began with questions including "what is knowledge?," and "how is it acquired?"; seeking answers to these questions led to a discovery of epistemology. Heylighen (1993) asserts that epistemology essentially attempts to answer basic questions such as, "what distinguishes true (adequate) knowledge from false (inadequate) knowledge?" In the context of research, such questions translate into issues of scientific methodology: how can one develop theories or models that are better than competing theories? My primary focus in this study is not to develop a theory or model, but to select theory/ies or model/s which support a holistic account of students' learning experiences from their perspectives, and within their context. This approach can be seen in my experience of moving from ALT to Activity theory, and then to the sociological constructs of identity, community, and institutional influences (will be discussed in detail in the following sections). The departure from ALT was influenced by the approach's detachment of the individual from his/her context; from my perspective, knowledge of students' learning experiences is inadequate when not understood within its specific context. This approach defines my epistemological position as Interpretivist, due to my underpinning assumption that "all human actions are meaningful and have to be interpreted and understood within the context of social practices. In order to make the sense of the social world, the researcher needs to understand the meanings that form and are formed by interactive social behavior" (Scott & Usher, 1996, p. 18).
As the quantitative learning inventories approach is based on questionnaires, there is little or no room for individual interpretation. As an interpretivist, I wanted to gain a holistic account of students' experiences from their perspective and within their context, which cannot be done using questionnaires alone. Moreover, survey instruments do not actually measure conceptions, but instead only measure the way in which students respond to questionnaires: "they are not sampling learners' behavior, but learners' impressions" (Mitchell, 1994, p. 8). Although, as an Interpretivist, I do realize and acknowledge the importance of learners' impressions, learning inventories are not primarily intended to offer an insight into this.
Secondly, although this approach places an emphasis on students' perceptions, it does not move beyond these insights to consider the ways in which students' perceptions, conceptualization and comprehension of different variables in educational settings are also structured and shaped by contextual factors. An important critique of ALT is therefore based on its removal of the individual learner from his/her rich, complex and multiple contexts (Ashwin, 2009;Case, 2008;Haggis, 2003;Mann, 2001). The learner is instead constructed as passively shaped by past experience, and passively amenable to reconstruction as a deep learner through a new set of molding processes which occur within the university: the learner, according to ALT, is "a human being without agency" (Haggis, 2003, p. 95). There is little acknowledgment that learners are people who may have any number of reasons to reject institutional agendas. Learners may be resistant or unable to engage with assumptions of higher education, for example, due to a sense of alienation (Mann, 2001), perceived risk or personal cost, or conflicting philosophical or cultural perspectives (Case, 2008). "In the new higher education, the learner may be a person who is experiencing tremendous difficulty in the face of unexplained norms and values; he or she may not know, for example, that facts are seen by many lecturers as the vehicle for the more abstract forms of conceptualization that are expected, but not modeled or defined. In addition, he or she may be exhausted from part-time work or parenting, distracted by family or financial problems, or lacking the fundamental confidence, self-esteem or health to engage in the ways that are assumed to be both desirable and possible" (Haggis, 2003, p. 96).
To address these issues, and to understand students' learning experiences in the ESL from a structure and agent perspective, I sought a theoretical framework that would incorporate the theory of variability from ALT into a more context-specific approach. In order to do this, I adopted Activity Theory. Activity theory accepts students' learning variability, and facilitates the analysis of their learning experiences from their perspectives, and in their own context.

Activity theory
Activity theory was developed from the work of Vygotsky (1978). Vygotsky's triangular model of "a complex, mediated act" (Vygotsky, 1978, p. 40) was fundamental in illustrating the relationship between the human agent and objects of their environment, as mediated by cultural means, tools and signs. As the model in Figure 1 depicts, the mediation process is two-way (represented by two-sided arrows); human beings actively shape the forces which are active in shaping them, and this process "lies at the heart of the many attempts to develop our understanding of the possibilities for interventions in processes of human learning and development" (Daniels, 2004, p. 121). Leont'ev (1981) developed a second generation of Vygotsky's model, which distinguished between individual action and collective activity. Leont'ev viewed the activity system as incorporating three hierarchical processes, driven by different individual or communal motives. These ideas were further developed by Engeström (1987), illustrated in Figure 2.
Within the Activity theory approach, learning and other human activities are understood as integrated into a larger system, or an activity. An activity is a theoretical entity, discerned, described, and used by the researchers to consider the socially based nature of human activity. The interacting constituents of an activity continuously develop into new forms which together create an activity. Used in this way, Activity theory serves as a clarifying and descriptive tool rather than a strongly predictive theory (Nardi, 1996), and does not prescribe any particular way of understanding learning or a fixed ontology (Berglund, 2002(Berglund, , 2004Havnes, 2002).  Kaptelinin, Kuutti, and Bannon (1995) point out that to understand Activity theory, it is important to understand a set of key concepts which together describe what constitutes an activity. Berglund (2002) has discussed in detail some of these basic principles of activity theory, based on the works of Engeström (1987), Kuutti (1996), Bannon (1997), and Havnes (2002). The principles Berglund (2002) discusses are: (1) An activity as a context: An activity is a theoretical construct which serves to describe and analyze a complex system as a unified whole. The activity gives "meaning to seemingly random individual events" (Engeström, 1993, p. 65); an activity is therefore a context, in which participating individuals and different events are integrated. Following this approach, a part of the activity cannot be studied as a separate unit without seeing it as belonging to and interacting with the whole system. The activity is the context, and analyzed as a unified whole, an activity is the unit of analysis. However, Berglund (2002) argues that this does not imply that the activity is a homogenous and harmonious entity, but has a multi-voiced character and an embedded history, containing different, often contradictory, expressions and events.
(2) The dynamic structure and historicity of an activity: An activity is a dynamic entity that is always being developed. Its current state is related to its history, and a historical perspective is therefore required.
(3) The object as a reason for existence of an activity: Engeström (1987) claims that an activity is defined by its object, or the motive for the existence of the activity. An individual's actions may seem meaningless if the object of the activity is not considered.
(4) The role of mediation: Tools (abstract or physical) play a fundamental role in human activity. Tools or signs have mediating functions between the subject and the object, and are "at the same time both enabling and limiting" (Kuutti, 1996, p. 27). In other words, an individual interacts indirectly with an object by using mediating tools. In this way, "the use of signs leads humans to a specific structure of behavior that breaks away from biological development and creates new forms of a culturally-based psychological process" (Vygotsky, 1978;in Engeström, 1987, p. 59).
(5) The role of inner contradictions: The multi-voiced, continuously changing character of an activity means that inner contradictions between and within its constituents are inevitable. Kuutti (1996, p. 34) describes these tensions as "Contradictions manifest themselves as problems, ruptures, breakdowns and clashes. Activity theory sees contradictions as sources of development; activities are virtually always in the process of working though contradictions." Engeström (1987, p. 91) similarly argues that contradictions serve as driving forces in an activity system. He suggests that "Contradictions are not just inevitable features of activity. They are principal of its self-movement and […] the form in which the development is cast. This means that new qualitative stages and forms of an activity emerge as solutions to the contradictions of the preceding stage [or] form." (6) Individual actions as parts of an activity: Long and complex activities consist of smaller entities, and can be described in three levels that express the "hierarchical nature of human activity" (Engeström, 1990, p. 172). Engeström (1990) describes the three levels, distinguished by Leont'ev (1978) as "Activity is the molar unit, collective in nature and driven by a complex motive of which the individual actors are seldom aware. Activity manifests itself in the form of goal-oriented individual actions in which the subject is consciously aware of what he or she is trying to accomplish. Actions, in turn rely on automatic operations, dependent on the conditions at hand" (Engeström, 1990, p. 172). An activity should therefore be understood as a collective, changing context that both consists of and gives meaning to individual actions. Engeström (1987, p. 66) stresses that "we may very well speak of the activity of the individual, but never of individual activity; only actions are individual." Ashwin (2009) proposes the use of Activity theory as an analytical tool to analyze relations and identify contradictions within an educational program, in order to see how these relations and contradictions support or undermine students' learning and teachers' teaching experiences. His analysis focuses on understanding the interaction of students and teachers within an educational program from a structure-agent perspective. Ashwin (2009) maintains that generating the structure of an activity system (whether for learning or teaching) is an empirical matter, which means that the researcher must discover different variables through research and use Activity theory for further analysis. My generic model of students' learning activity in the ESL program, based on preliminary findings, can be seen in Figure 3.
This model includes elements of Activity theory reinterpreted for my own study. These elements include the reinterpretation of the "subject" as student identities; "artifacts" as teaching, assessment, textbooks, physical space and other learning resources; the "object" as gaining knowledge, improving practices, passing exam, passing time, and maintaining good image in the classroom; "rules" as institutional rules, assessment criteria, and tacit rules; "community" as peers and teachers; and "division of labor" as students as learners vs. other roles, and students as learners vs. the teachers' role.
After placing these variables in different nodes, the initial analysis highlights the importance of the sociological constructs of identity and community. Moreover, each node of Activity theory is the product of another Activity theory (Engeström, 1987), which suggests that each node in Figure 3 is shaped by structural influences (or the institutional influences in this case). I decided to use these three concepts (e.g. identity, community, and institutional influences) for further analysis. This focus was chosen largely due to the fact that it would not be possible to consider all the elements represented in Figure 3 by analyzing them individually. Moreover, this individual approach to analysis would represent a return to traditional ways of analyzing assessment (Genesse & Upshur, 1996;O'Malley & Pierce, 1996), teaching (Hughes, 2003), and syllabus (Jahanbakhsh, 1996). While I acknowledge the contribution of these traditional analytical approaches to assessment, teaching, and syllabus, I adopted a structure-agency perspective as theoretical approach in order to examine students' learning experiences. Therefore, I focused on the notions of identity, community, and institutional influences (see Figure 4).

Strengths and limitations of activity theory
Activity theory is a useful theoretical framework which can help in understanding students' interactions with different variables from their perspectives and within their own context. The theory also works as an analytic tool to understand students' experiences in terms of their interactions with different aspects of the ESL program. However, for further analysis, I focus upon three factors which emerge from the conceptualization of students' learning in the ESL program using Activity theory (Figure 4). These factors are students' identities, the community of ESL classroom, and the institutional influences on the policies and practice of the ESL program. I, therefore, use Activity theory to a limited extent for two main reasons. Firstly, although Activity theory allows a focus on students' interactions with different variables of the ESL program, and shows the resultant outcome of such interactions, it does not explain how these variables are shaped by structural (institutional) influences. For example, while it helps in the analysis of students' interactions with rules, and the effects of these interactions on the objectives of the ESL program, Activity theory does not explain why these rules are specifically formed with the context of a particular institution. Additionally, while Activity theory allows students' interactions with teaching, syllabus, physical space, and other tools, it does not explain how institutional influences shape these tools.
Secondly, adopting a constructivist ontological and interpretivist epistemological stance, I consider students as social agents, who therefore have multiple identities. From this perspective, the subjects (i.e. students) in Activity theory are not merely embedded learners, but social agents with multiple identities. However, Activity theory does not offer any conceptual understanding of identity. Similarly, the idea of community is central to Activity theory, but the theory does not offer any conceptual definition of community. I therefore limit the use of Activity theory, viewing it as a helpful framework for the identification of variables which shape students' ESL learning experiences, but not as a main theoretical framework. Students' identities, the community of the ESL classroom and institutional influences are instead used to analyze students' learning experiences.

Conceptualizing identity from structure and agency
Several researchers have asserted that in almost all ESL teaching situations, teachers and researchers do not simply deal with language, or with learners and their cognitive and affective characteristics, but also take into account the relational aspects of ESL learning (Arnold, 1999;Jackson, 2008;Parkinson & Crouch, 2011). Arnold (1999, p. 18) argues that ESL learning and use is a transactional process, which he defines as an act of reaching out beyond the self to others. As such, it is intimately connected with learner's emotional self. The way in which individuals consider who they are is significantly formed by their social identity. A social perspective on ESL highlights the fact that learners are not anchored to a fixed state, but are conditioned by social forces which affect their sense of self (Arnold, 1999). This conceptualization of the self begins with an assumption that there is a reciprocal relationship between the self and society (Stryker, 1980). Based on this understanding that the self emerges in and is reflective of society, Brewer (2001) argues that a social understanding of the self means that the society in which the self is acting must also be understood. Moreover, the idea that the self is always acting in a social context in which other selves exist must be taken into consideration.

Figure 4. Conceptualising students' learning experiences using identity, community and institutional influences.
A review of the literature suggests that the concept of self and identity has been adopted in numerous ways in the social sciences. For example, according to the traditional symbolic interactionist perspective, known as the situational approach to self and society, society is viewed as constantly in the process of being created through the interpretations and definitions of actors in particular situations (Blumer, 1969). Individuals identify the factors which need to be considered for themselves, act on the basis of these identifications, and attempt to fit their lines of action with others in particular situations to accomplish their goals. From this perspective, it is assumed that individuals can freely define situations, which means that society is always thought to be in a state of flux with no real organization or structure (Stets & Burke, 2005, p. 129). By contrast, the structural approach does not see society as tentatively shaped (Stryker, 1980), but assumes that society is stable and durable, as reflected in the "patterned regularities that characterize most human action" (Stryker, 1980, p. 65). These contradictory approaches nevertheless share the notion that identity is embedded in social interactions; while the former perspective over-emphasizes individual agency, the latter questions individuals' capacity to make choices. There are researchers who suggest a third way, acknowledging the deterministic features of structure but also leaving room for individual choice. For example, Giddens (1991) has suggested that even in the most extreme life conditions, space for individual choice remains. Cultural anthropologist Mathews (in Block, 2007, p. 865) has similarly argued that identities are not innate personality traits, but assumed and worked out by individuals. Identities are constructed in a "cultural supermarket," which is not free from social constraints, and individuals face these constrains in different ways. Giddens (1991) and Mathews' contribution to this discussion is that although identity is conditioned by social interaction and social structures, it simultaneously conditions social interaction and social structures. In short, it is constitutive of and constituted by the social environment. Using this third approach, I conceptualized the notion of identity in terms of the extent to which it is shaped by structure, but also the extent to which agents can exercise choice and agency in constructing identity. Identity is also used to refer to multiple, contradictory, and contested understandings of self in relation to agencies.

Conceptualizing community from structure and agency
Learners construct different identities, which are socially, culturally, politically, and historically situated, within ESL classrooms. In fact, ESL classrooms are viewed as important sites in which learners acquire English language skills in a formal or instructional setting, and in which teachers support students to gain particular linguistic and academic competencies. In order to acquire these competencies, students' participation in classroom activities is seen as essential (Barnawi, 2009;Hirst, 2007;Morita, 2004).
Classroom participation implies that students are engaged actors on socio-cognitive planes, where cognitive and social properties are reciprocally connected and essential to solve given problems. Socio-cognitive interactions are inevitably more complicated in a classroom made up of linguistically and culturally diverse students (Hirst, 2007). Morita (2004, p. 573) points out that "understanding how these students participate in their new academic communities and acquire academic discourses in their second language has become critical." The crucial importance of acquiring academic discourses on ESL programs requires in-depth analysis; however, for this analysis to take place, the notion of community within an ESL classroom must first be conceptualized.
The notion of community, as with the concept of identity, has been defined in many different ways; however, early conceptualizations of community were predominantly structuralist. For example, Ferdinand Tönnies, a German sociologist, defined community as Gemeinschaft. Gemeinschaft is a social configuration in which individuals are as oriented to the wider association as they are to their own self-interest. Individuals are regulated by common beliefs about appropriate behavior, the responsibility of members of the association to each other, and to the association at large. Associations are therefore marked by a unity of will (Harris, 2001). Moreover, Gemeinschaft community involves ascribed status, i.e. individuals are given a status by birth. For example, the child of a farmer will assume their parent's status until death.
Reacting against this structuralist conceptualization of community, a number of researchers (including Cohen, 2001) developed an approach which views individuals as symbolically constructing communities. Lave and Wenger (1991) are also among the cognitive anthropologists who developed Community of Practice (CoP), an approach which considers both structure and agency when conceptualizing community. Like Tönnies' Gemeinschaft, individuals in Lave and Wenger's CoP are oriented to the larger association as much as to their own self-interest, but unlike Gemeinschaft, individuals develop themselves personally and professionally through the processes of sharing information and experiences with the group.
Wenger explains CoP as "groups of people who share a concern or passion for something they do and learn how to do it better as they interact regularly" (cited in Toohey, 2000, p. 1). Smith (2009) maintains that a CoP involves much more than the technical knowledge or skill associated with undertaking some task. Members are involved in sets of relationships over time, and communities develop around things which matter to people. Describing Lave and Wenger's conceptualization of learning, Hanks (1991, p. 41) notes that "rather than looking to learning as the acquisition of certain forms of knowledge, Lave and Wenger have tried to place it in social relationships-situations of coparticipation.
[…] Rather than asking what kind of cognitive processes and conceptual structures are involved, they ask what kinds of social engagements provide the proper context for learning to take place." Learners are therefore not seen to acquire structures or models through which they understand the world, but rather they participate in frameworks which are structured. Learning involves participation in a CoP; in this sense, participation "refers not just to local events of engagement in certain activities with certain people, but to a more encompassing process of being active participants in the practices of social communities and constructing identities in relation to these communities" (Wenger, 1999, p. 4).
A number of researchers have used CoP in the context of ESL. Barnawi (2009) used CoP to explore Saudi students' English language learning experiences at a North American university. According to Barnawi (2009), membership in a CoP situated in an English-medium academic classroom always alters as newcomers (students who have no or minimal experience of attending English medium schools and have no access to the power associated with English language skills) interact with old-timers (students who have been to English-medium schools, and already have the access to the power associated with English language skills). This interaction involves language learning and socialization, whereby competency and membership are required in order to participate in the discourse community (Lave & Wenger, 1991;Morita, 2004). Here the process of participation-legitimate peripheral participation-has significant implications for the membership in the discourse community. Wenger (1998a) argues that the notions of "peripherality" and "legitimacy" are vital as both notions facilitate newcomers' actual participation in the classroom in order to help them move to the center of the community, which in turns enables them to access greater power. Peripherality provides "an opening, a way of gaining access to sources for understanding through growing involvement" (Lave & Wenger, 1991, p. 37). This idea suggests that individuals can be members of the CoP in various ways, and their roles and positions in the community are subject to change over place and time.
Meanwhile, legitimacy affects the way in which individuals gain access to particular CoPs (Barnawi, 2009). Newcomers must be given a sufficient sense of legitimacy in order to be seen as viable potential members. Wenger (1998b, p. 101) argues that "only with legitimacy can all their inevitable tumbling and violations become opportunities for learning rather than cause for dismissal, neglect, or exclusion." Legitimacy therefore also plays a vital role in the teaching and learning process, as it facilitates access to power. The higher the level of legitimacy granted to learners in a certain classroom setting, the more they will be able to negotiate and construct their identities in the CoP (Toohey, 2000). A significant number of research studies used CoP to examine different factors affecting the legitimacy students get in the form of class participation in the ESL classroom. These studies found that students' learning experiences are shaped by teachers' roles and strategies (Christiansen, 2010); teachers' previous experiences (Schweisfurth, 2011); student teacher ratio in the classrooms (Cuseo, 2007) and the opportunities for students to interaction with teachers (Carbone & Greenberg, 1998). However, most of these studies used teachers' perspective to understand students-teachers interactions in the ESL classrooms. I use the CoP to conceptualize the ESL classroom from students' perspective. This means that the focus is on how students understand the community of ESL classroom and different power dimensions within this community; how they define the role of teachers in the community; how teachers' pedagogic strategies enhance or limit students' opportunities to participate, and therefore increase or decrease their legitimacy in the community; how large classes affect student-student and student-teacher interactions, and how these interactions affect students' participation and legitimacy in the community.

Conceptualizing institutional influences from structure and agency
As observed in Figure 3, students' learning experiences in the ESL program are influenced by numerous variables (i.e. teaching, examination, syllabus, physical space and so on). However, these variables are not free from structural influences. For me, structure refers to the institute where I conducted my research, and structural influences suggest institutional influences on the ESL program. Thus, it is important for me to conceptualize institutional influences in a way that complements my structure and agency stance to analyze students' learning experiences in the ESL.
A number of studies have examined the institutional (i.e. universities and other institutes of higher education) influences on the teaching and learning in higher education in general (Ashwin, 2009;Barratt-Pugh, 2007) and on ESL programs in particular (Gao, 2005;Flowerdew & Miller, 2008;Kelly, 2010). Some studies have focused on analyzing the relationships between specific variables within the institute, and the impact of these relationships on learning and teaching. These studies refer to institutional influences as institutional culture. For example, Marginson (2008) discovered a link between institutional culture and the position of institute in the field of higher education. Exploring the global field of higher education and drawing on Bourdieu (1993), Marginson divides the field of institutes of higher education along two axes. The first axis is a continuum from elite research universities to commercial vocational education, and the second axis is continuum based on the institute's focus on global or local markets.
Within the field of higher education, different institutes develop their own cultures which are strongly influenced by their positions in the field, and which highly influence the teaching and learning within these institutes. Kezae (2006) finds that the size of an institute can be an important institutional influence that greatly impacts the way in which students engage with a program. Similarly, Gibbs and Dunbas-Goddett (2007) establish a link between assessment and institutional culture, arguing that different approaches to assessment on different programs are related to institutional culture. Jones, Turner, and Street (1999) discuss the relationship between the use of English and institutional culture, noting that there are few differences in the quality of English produced in two different institutional cultures; however, the ways in which it is produced differ. Another relationship is established by Crozier, Reay, Clayton, Colliander, and Grinstead (2008), between entry requirement for students and institutional culture. Using Bourdieu's (1990, p. 56) notion of "habitus," they argue that the implicit understanding of an institute influences who can be "legitimate" students within particular program within particular university. Henkel (2000) discusses the relationship between teachers' identities and institutional culture, describing differences in her examination of academic identities in different universities. Likewise, other studies analyze the impact of students on institutional cultures. These studies suggest that, for a number of reasons, students of different social classes tend to attend different types of higher education institutes, which also affect the institutional cultures (Ashworth, 2004;Brennan & Naidoo, 2008;Crozier et al., 2008). Ashwin (2009) has conducted research integrating all of the above-mentioned variables, and linked it within institutional culture. Ashwin (2009) used a Bourdieusian structural approach to conceptualize the influence of macro-structure (i.e. the influence of higher education on the policies of a university; the influence of university policies on its departments; the influence of departments on the structure of the programs offered) on the micro-structure (i.e. the interaction between students and teachers) in higher education. He suggests that using the Bourdieusian notions of habitus, field, and capital, it is possible to consider the position of different institutes in the field of higher education, and how these positions impact on their institutional habitus.
This notion of institutional habitus was borrowed from the works of McDonough (1997) and Reay (1998). McDonough (1997) develops the concept of institutional habitus as a link between institutions and the wider socio-economic context, and it is this link which, she argues, differentiates institutional habitus from institutional culture. She argues that the formation of institutional habitus constitutes a complex mixture of agency and structure, and could be understood as the impact of a cultural group or social class on an individual's behavior as it is mediated through an organization (McDonough, 1997). McDonough (1997) further claims that institutional habitus is established over time and develops its own history. It is, therefore, capable of change, but by dint of its collective nature, institutional habitus is less fluid than individual habitus (Bourdieu, 1993, p. 78 describes it as "a power of adaptation"). McDonough (1997) argues that institutional habitus is developed in relation to each institute's position in the field of higher education. Each institute has a form of capital (economic, cultural, and social capital that takes on different forms of symbolic capital) which are developed or maintained in the field of higher education. The ways in which institutes attempt to develop or maintain their capital inform the notions of what is reasonable. In her study of how schools in the United States influence their students' choice of college, McDonough (1997) argues that institutional habitus informs: (1) an institute's sense of its students' expected identities, (2) the courses institute will offer; and (3) which progression routes constitute reasonable uses of the capital developed by students. Ashwin (2009) argues that institutional habitus can be articulated in different forms of expression, which constitute institutional settings for the development of students as learners. These forms of expression include: the selection criteria for entry to an institute and onto a program within institute; the particular form of the program offered by institutes; institutional teaching and learning quality regimes, which include the general standards of institute; teaching standards; examinations criteria; curriculum; and the quality of learning space (i.e. the size of classrooms, the quality of libraries or laboratories, etc.). All these forms of expression determine the ways in which institutes perceive the role of students. Some institutes perceive students as "consumers" (Naidoo & Jamieson, 2006), while others adopt approaches to develop "independent learners" (Smith, 2007). Ashwin (2009) argues that institutional settings do not act in a deterministic way, but instead, different students can respond to the same institutional setting in different ways, depending on the relationship between students' identities as learners and their other personal identities.
This review foregrounds the importance of institutional influences on teaching and learning in higher education institutes. At the same time, it highlights the importance of an institute's position in the field of higher education, its institutional habitus, and different forms of capital that an institute tries to maintain. Thus, I use Bourdieusian notion of field, capital, and habitus to conceptualize institutional influences from structure and agency. Using this theoretical framework, I can conceptualize the field of higher education; the position of the institute within this field; the status of the ESL program within the institute; different policies and practices of the institute; and the influences of such policies and practices on the teaching and learning of ESL.

Conclusion
The ESL teaching and learning in higher education have mostly been examined using psychological tradition of research, known as the ALT perspective. The ALT perspective is based in the phenomenographic research traditions of Marton and Säljö (1976), which was further developed by Entwistle and Ramsden (1983) and Biggs (1987) using quantitative learning inventories. Despite its popularity, the ALT perspective has been importantly critiqued for its individualistic approach to understanding students' learning in ESL. Nevertheless, unlike previous behavioristic and neurological approaches to understanding ESL teaching and learning, the phenomenographic research tradition has tried to consider issues relating to context. These contextual issues are only covered to a very limited extent in phenomenographic research, but this research tradition can be credited with moving from a notion of fixity to the notion of variability in ESL learning. I, therefore, aimed to locate myself within this theory of variability in learning, but at the same time I sought to use this notion of variability within a more context-specific understanding of ESL teaching and learning. In particular, I was interested in understanding how certain variables impact upon ESL teaching and learning from students', or agents' perspective. I, therefore, needed to find a theory which could explain the different variables affecting teaching and learning in the context of ESL.
To this end, I found Engeström's (1987) Activity theory to be a sophisticated tool for the analysis of individuals' experiences of learning from their perspective and within their own context. I, therefore, adopted an Activity theory framework to generate a model of students' learning activity in ESL program by identifying different variables. This model highlighted important factors such as students' identities, the community within the ESL classroom, and institutional influences on different variables that shape students' learning experiences. However, Activity theory does not offer conceptual insights into these notions. My use of Activity theory in the current study has therefore been limited to explaining variables in students' learning experiences; it has not been adopted as an overarching theoretical framework for my research. Nevertheless, my brief interactions with Activity theory were useful in enabling me to finally decide to focus upon the use of identity, community, and institutional influences for further analysis. Thus, I decided to use the concept of identity in relation to students and teachers in ESL context; the concept of community to examine interactions in the field space of ESL, and the concept of institutional influences in order to understand how the policies and practice of ESL are shaped by an institute's position in the field of higher education. As I intended to understand students' learning experiences in the ESL program from a structure and agent perspective, I conceptualized identity, community, and institutional influences using theories, such as Symbolic Interactionism, CoP, and Bourdieusian notions of habitus, field, and capital.