Unveiling the relationship between language learning beliefs , emotions , and identities 1

Several authors (Frijda, Manstead, & Bem, 2000; Van Veen & Lasky, 2006) suggest that emotions, cognitions, and identities are intrinsically related. Authors in social psychology (Fiedler & Bless, 2000; Frijda, Manstead, & Bem, 2000; Rosiek, 2003) have considered how beliefs are particularly sensitive to affective influences and how emotions, cognitions, and identities are intrinsically related. Understanding this relationship would help researchers to reveal complex key issues in beliefs research, such as the relationship between beliefs and action. Yet, although research on beliefs in applied linguistics goes back to the 70s and 80s, there has been scant connection with emotions and identities. This paper aims to reflect on the relationship between beliefs, emotions, and identities by looking at these coconstructing, overlapping concepts to advance our understanding of language learning and teaching. Through a review of studies on beliefs, emotions, and identities within applied linguistics and other areas, I illustrate how beliefs and emotions are intrinsically and interactively related, and how beliefs within a socio-historical context influence the construction of identities. Identities influence the kinds of emotions and beliefs that individuals attribute to themselves and to others. Emotions, in turn, can influence identities and how we construct them. Implications for research on beliefs, emotions, and identities are suggested.


Introduction
Research on beliefs about language learning and teaching has been recognized as a solid area of research in applied linguistics (AL) and is covered by an extensive volume of studies.Although beliefs are said to have an affective component (Nespor, 1987;Rokeach, 1968) and are believed to relate to our identities, the relationship between beliefs, emotions, and identities has not been the focus of empirical studies in the field of AL itself or in other fields for that matter.Even in psychology, with a wide literature on cognition and emotion concerned with affective influences on memory, the reference to research on beliefs is incipient (Fiedler & Bless, 2000, p. 144). 2 In this article, I wish to explore what I anticipate to be the complex relationship between beliefs, emotions and identities.Understanding this relationship is important for three reasons.
First, given the inseparability of cognition and emotion as recognized in diverse fields such as neuropsychology (Damasio, 1994;Ledoux, 1996), education (Frijda, Manstead, & Bem, 2000;Zembylas, 2004) and AL (Aragão, 2007;Dewaele, 2005;Gieve & Miller, 2006;Pavlenko, 2005), there is a need to better understand the affective dimension of teacher and learner cognition.More specifically, it is important to understand how emotions interact to shape what teachers (Borg, 2006) and learners (Kalaja & Barcelos, 2013) do, their implications for their learning (how and what they learn) (Nespor, 1987), and how teacher emotions are part of teachers' practical knowledge and pedagogical content knowledge (Rosiek, 2003) and learners' emotions are related to their beliefs (Aragão, 2011).Thus, understanding the relationship between beliefs and emotions can help us understand how together these influence teachers' and learners' actions.Within the field of language teacher research, recent work has already begun to unveil the ways in which emotions interact with teacher cognition (Golombek & Dorian, 2014;Johnson & Worden, 2014;Kubanyiova, 2012).
Second, although there are numerous studies on identity in AL (Block, 2007;Murphey, Chen, & Chen, 2005;Norton, 2001;Pavlenko & Norton, 2007;Yoshimoto, 2008) and the relationship between beliefs and identities has been 2 The collection edited by Frijda, Manstead and Bem (2000) is an exception.suggested in the literature (see following sections), none has investigated how identity, emotions, and beliefs are related.The investigation of this relationship can shed light on several issues in belief research, such as the dissonance or conflict between beliefs and actions as well as how beliefs evolve and develop.
Finally, in all theoretical models of beliefs such as Ellis (1994) and Borg (2006), there is a lack of reference to the concepts of emotions and identities.Although these models include several factors which influence beliefs, such as context, professional development and classroom practice, emotions and identities do not appear as influential factors.For example, Woods's (1996) model does not explicitly mention emotions but attributes an important role to interpretation, which can be viewed as an emotional component of appraisal.
Given these gaps in the existent literature and research, this paper attempts to bring together current literature on the concepts of beliefs, emotions, and identity with a specific focus on emotions and their interrelatedness with beliefs and identities.It is hoped that by doing so, it can outline some of the theoretical and potentially methodological issues that future work in this area will need to address.It does so, firstly, by discussing the concepts of beliefs, identity, and emotions within studies in AL, psychology, social sciences, and education; secondly, by reflecting on the interrelationships between these three concepts as suggested in the literature; and, finally, by suggesting directions for further research on the coconstructedness of these three concepts as well as by considering implications for language teaching and learning.

Beliefs, identities, and emotions
In this section, I briefly review the concepts of beliefs, identity, and emotions defining them and drawing insights from studies in a range of disciplines.Although each construct is discussed separately for purposes of organization and clarity of the article, in reality, all of them are interrelated constructs as I hope to show in the second part of this paper.
Research about teacher beliefs or cognition is a topic that has been investigated longer than research on learner beliefs in AL.The concept of teacher beliefs is situated within a research tradition in education, which started in the seventies and changed the paradigm about teachers and their work with the publication of several seminal studies such as Clark and Peterson (1986), Shavelson and Stern (1981) and Shulman (1986) on teachers' thoughts and knowledge, which portrayed teachers as knowledge constructers (Borg, 2006).Thus, teachers' beliefs and thoughts started gaining prominence in the research literature.Understanding teachers' beliefs helps us understand both teachers and learners' actions and decision-making processes in the classroom (Borg, 2003;Johnson, 1999;Pajares, 1992;Woods, 1996).Since beliefs act as a filter for human behavior, they are an important feature of reflective teaching (Richards & Lockhart, 1996), which encourages teachers to reflect upon and question their beliefs to understand how they teach.In particular, beliefs are an important feature of educational and teacher change, since a change in teachers' practices requires a change in their beliefs.
Beliefs have been investigated through a variety of methods and approaches (Kalaja & Barcelos, 2013), ranging from a more positivistic mainstream approach using closed questionnaires, with beliefs being viewed mostly as mental fixed constructs in one's minds, to a more contextual approach making use of interviews, narratives (written and visual), and observations, among others.More recent studies 4 conducted within the contextual approach have conceptualized beliefs as being dynamic and emergent, socially constructed and contextually situated, potentially paradoxical and dialectal, that is: They are social and individual, shared, diverse and uniform; and constitute a complex dynamic system that is interrelated, embedded, nonlinear, multidimensional and multilayered (Kalaja & Barcelos, 2011;Mercer, 2011).

Identities
The concept of identity has become more widely discussed in AL after Peirce's (1995) /Norton's (1997) 5 publication of her study with immigrant women in Canada.Identity can be defined as "people's understanding of their relationship to the world, the construction of that identity across time and space, and people's understanding of their possibilities for the future" (Norton, 1997, p. 410).This definition stresses how identities are related to individuals' desire for recognition, affiliation, security, and safety, which are all dependent on how material resources are distributed in society.Within a postructuralist view, identity is seen as multiple, multifaceted and dynamic.
Studies in social sciences offer other important insights about identities that may help us better understand them.Lemke (2008) presents five interesting perspectives on identities that help shed light on their relationship to beliefs and emotions.First, identities are multiple.This means that we behave differently in different situations (formal and informal) and settings (professional and intimate) and with different people (children or peers).Second, identities are multi-faceted and dynamic, they develop and change as who we are is constructed within the constraints of our interactions in different settings in our lives with different people (Lemke, 2008, p. 19).Third, identities are hybrid.Thus, we may be forced to act in more normalized ways than we would like to since there may exist "contradictions between our subjective identities, who we are to ourselves, and our projected identities, who we wish to seem to be to others."(p. 20).We compromise to reconcile the pressures between who we want to be and the pressures and forces of the multiple cultures and institutions that seek to shape our identities.Fourth, identities are contested and conflictual, not only because "there are struggles over the kinds of identities we are allowed to claim for ourselves," but also over "the kinds of identities we can conceive for ourselves" (p.31).Lastly, identities are performed.In other words, we have what Lemke (2008, pp. 24-25) has called longer and short-term identities and identity-in-practice. Longer-term identities, "inscribed in our habitus," are dispositions for action in the moment, and are themselves constituted through many actions across many moments.Our longer-term identities are performed and constituted through ways of enacting relationships to people who are significant to us in scenarios we recognize as familiar.

Emotions
Historically, there has always been a neglect of emotions and overemphasis on the rational dimension of language, language learning and teaching.As Swain (2013) puts it, emotions are the elephant in the room: poorly studied and understood and "seen as inferior to rational thought" (p.205).However, there are exceptions and several authors have recently begun to emphasize the inseparability of emotion and reason and the central role of emotions in the process of learning a language (Aragão, 2005(Aragão, , 2011;;Dewaele, 2011;Garret & Young, 2009;Rajagopalan, 2004;Swain, 2013).
In her pioneering work, Arnold (1999) defined affect broadly as "aspects of emotion, feeling, mood or attitude which condition behavior" (p. 1).The affective domain for her includes aspects such as motivation, anxiety, extraversion, and introversion.A more dynamic definition of emotions sees them as "a complex, multifaceted psychological phenomenon," which is "constantly generated, unfolded, and changed through multiple recursive effects at any one moment" (So, 2005, p. 44).To Aragão (2011), emotions are "bodily dispositions for situated action," which "represent a variety of ways of acting in relation to the dynamics of the immediate environment" (p.302).Since the beginning of research in affective factors and emotions in AL, the field has witnessed different ways of looking into "affective variables," as shown in Table 1.(Moskowitz, 1978) Attitudes (Gardner & Lambert, 1972) Affective filter hypothesis (Krashen, 1985) First studies on anxiety (Bailey, 1983) Research on affect (affective factors as individual differences) (Arnold, 1999) Studies on language learning anxiety in language learning (Horwitz, 1990;Mac-Intyre, 1999) Schumann's work on the (1997) neurobiology of emotions Research on emotions and multilingualism (Dewaele, 2010;Pavlenko, 2005Pavlenko, , 2006) ) Emotions within the framework of the biology of cognition (Aragão, 2005(Aragão, , 2007) ) "Affective turn" (Pavlenko, 2013) Sociocultural approach to emotions (Swain, 2013;Imai, 2010;Golombek & Dorian, 2014;Johnson & Worden, 2014) Critical approach to emotions (Benesch, 2012) Research on positive emotions (Gregersen, 2013;MacIntyre & Gregersen, 2012) As illustrated in Table 1, earlier research concentrated on the works of Gardner and Lambert (1972) on attitude and motivation, and on building a more humanistic classroom (Moskowitz, 1978).Later on, the first studies on anxiety using diaries appeared (Bailey, 1983), as well as Krashen's affective filter hypothesis6 (Krashen, 1985).In the 90s, the role of anxiety in language learning began to be investigated (Horwitz & Young, 1991;MacIntyre, 1999).The premise is that too much anxiety is not conducive to language learning.Other studies during that decade focused on motivation and the neurobiology of emotions (Schumann, 1997).The first decade of 2000s saw an increase in the role of emotions as to how multilinguals and bilinguals experience emotions and express them in their respective languages (Dewaele 2010;Pavlenko, 2005Pavlenko, , 2006;;Pavlenko & Dewaele 2004a;2004b), as well as the diversity of emotions learners experience in the classroom (Mercer, 2005;Oxford, 1995), and emotions as processes (So, 2005).The last decade has witnessed a different wave of studies focusing on emotions that concentrate on emotions of speakers of other languages (Garret & Young, 2009), self-esteem (Rubio, 2014), and positive emotions (Gregersen,  2013; MacIntyre & Gregersen, 2012), 7 as well as studies based on critical theories (Benesch, 2012), and sociocultural theories (Imai, 2010;Swain, 2013).
AL has experienced the social turn (Block, 2003), and now it seems we are living an emotional turn or what Pavlenko (2013) has named the "affective turn."For Pavlenko, the affective turn replaces the single research question on how affect influences L2 learning with a range of diverse questions regarding linguistic, psychological and social dimensions of emotion that suggest how L2 learning can in turn influence emotions.This emotional turn, in my view, has raised some criticisms about how emotions have been investigated in AL.The main criticisms refer to how emotions are defined and investigated.First, emotions are marginalized and seen in a limited way only as "affective variables" or factors and "individual differences," dissociated from the social contexts, limited to motivation, anxiety and personality characteristics (Pavlenko, 2013;Swain, 2013).For instance, in books on SLA, for example, Ortega (2009), a separate section for affect and other individual differences includes items such as personality, extraversion and speaking styles, learner orientation to communication and accuracy, foreign language anxiety, willingness to communicate and L2 contact, cognitive styles, learning styles, learning strategies, and self-regulation theory, but makes no reference to other types of emotions.Second, the relationship of emotions with power relations, ideology and culture have largely been marginalized in research to date (Zembylas, 2005), with the exception of Benesch (2012).Third, most studies focus on a single negative emotion such as anxiety 8 and overlook the diversity of students' experiences and emotions as well as their dynamic and situated nature.Fourth, they heavily rely on quantitative approaches and survey instruments (Brown & White, 2010;Swain, 2013).Fifth, "affective factors" are usually discussed in terms of a causal relationship to L2 acquisition and as predictors of success (Gieve & Miller, 2006;Pavlenko, 2013).This is problematic because identity, emotions and beliefs are dynamic and social concepts, and, thus, the relationship to language learning is not one of causality but of interaction and reciprocity (Pavlenko, 2013).It is assumed that only emotions influence language learning, while the reverse is rarely considered (Swain, 2013).Finally, there is the implicit assumption that "emotion is out there (or in here) somewhere in psychological reality waiting to be isolated, pinned down and dissected" (Parkinson, 1995, p. ix).These criticisms show that the current approach to emotions still views them as decontextualized variables detached from one's mind and body ignoring their contextualized and embodied nature.I would argue that this way of looking at emotions does not do justice to the complex interplay of emotions with other constructs such as those which are the focus of this article, identities and beliefs.

Defining emotions
Before looking at the relationship between emotions and other constructs, it is necessary to define them.However, there are as many definitions of emotions as there are different emotions and their combinations as we feel.The word emotion itself implies movement and motivation (e-motion) (Hansen, 1999).Emotions color our perceptions and influence how we choose to act in the future and "trigger the appropriate biochemicals to set the internal scene in readiness for that action."Fear is the clearest example of that: It "orchestrates a massive physiological preparation for fight or flight" (Hansen, 1999, p. 214).In psychology, Izard (1991) explains that an emotion is "experienced as a feeling that motivates, organizes, and guides perception, thought, and action" (p.14).In AL, So (2005) defines emotions as, "the psychological outcome of dynamic interactions between different layers of internal and external systems -physiological, cognitive, behavioural and social" (pp.43-44).Emotions are seen as, "constantly generated, unfolded, and changed through multiple recursive effects at any one moment."So believes we should talk about emotion processes instead of emotional states since 8 According to McIntyre and Gregersen (2012, p. 195), anxiety is the most studied emotion in SLA.they are "dynamic," "self-organizing," and "nonlinear and constantly emerging through the causal interdependence among internal and external variables on different timescales" (p.5).
Instead of providing a single definition of emotion, it is best to list some of their characteristics as pointed out by Solomon (2004, p. 13).These five aspects of emotions, which are often interwoven, characterize emotions as being: (a) behavioral: When we experience an emotion it usually includes facial and verbal expressions, reports (such as I love you) and elaborate plans for action; (b) physiological, including hormonal, neurological, neuromuscular changes; (c) phenomenological, which includes physical sensations, ways of seeing and describing the objects of one's emotions, as well as "metaemotions;" (d) cognitive, which refers to appraisals, perceptions, thoughts, and reflections about one's emotions; and (e) social, referring to interpersonal interactions and cultural considerations.In addition, recent literature in education and psychology has described emotions as active, interactive and processual, hierarchically, discursive and culturally constructed.Table 2 presents a summary of these characteristics.Emotions are processes and not static entities.Interactive: Usually they involve a relationship between a subject and an object or an interaction between an individual and his/her environment, as well how we position ourselves towards these.Mesquita & Markus (2004, p. 355) Bosma & Kunnen (2001, p. 41).

Hierarchically constructed
Emotions form a complex network, an open system in which new emotions get added, but nothing gets out easily.Ekman (2004, pp. 128-129) Socially constructed Emotions are context dependent (the types of emotion one is allowed to express or not depend on national cultures, as well as on institutional cultures and emotional rules of different contexts).Emotions are not private, individual, psychological states but social and embodied.Mesquita & Marcus (2004), Zembylas (2004).Benesch (2012, p. 133) Culturally constructed Emotions are shaped by and also shape the sociocultural context (social relationships, systems of values in families, cultures and school situations).Emotion is a mode of social action, developed over time.Mesquita & Markus (2004), Zembylas ( 2004) Parkinson (1995) Discursively constructed Emotions are embedded in discursive, ideological practices, and power structures.Zembylas (2005) As illustrated in Table 2, this set of characteristics present a very different view of emotions from how they are typically portrayed in AL.Emotions are discursively constructed and seen as processes, which shape and are also shaped by the sociocultural context.As such, they are interactive, dynamic and form a complex network.This view of emotions also emphasizes the role of power, norms, rules and standards that can regulate emotions.In addition, it helps us understand how they relate to identities and beliefs, a topic I now turn to.

The relationship between emotions, beliefs, and identities
How has the relationship between beliefs, identities, and emotions been addressed in the literature?In this section, I first address some studies in social psychology (and a few in AL) that have suggested a relationship between each set of constructs.First, I discuss how beliefs and identities are related.Second, I present studies that suggest the relationship between identities and emotions.Finally, I show how emotions and beliefs are related based on studies from social psychology.

Beliefs and identities
Beliefs and identities are related in intimate and intricate ways.We are what we believe and our beliefs make up our identities.This complex relationship between beliefs and identities has been suggested in psychology and philosophy.As early as 1968, Rokeach used the metaphor of an atom to compare to the structure of beliefs.According to Rokeach, beliefs systems are grouped into core and peripheral beliefs.Core beliefs, besides being more resistant to change, have the following characteristics: (a) They are more interconnected with others, and, because of this, they communicate with each other more easily and may bring more consequences to other beliefs; (b) they are more related to a person's identity and "self."This resonates with Dewey's (1933) concept of "pet beliefs," that is, beliefs to which we get attached to and which we do not let go of easily.These pet beliefs are more related to our identity and emotions, smililarly to the core beliefs.Thus, the more central a belief is, the more related it is to our identities and emotions and other beliefs.
Although there are no specific studies that have investigated the relationship between beliefs and identities in AL,9 some researchers have hinted at or suggested this relationship in their work (Barcelos, 2000;Murphey, 1995;Oxford, 2008;Sakui & Gaies, 2003;Singh & Richards, 2006;Woods, 2003).As early as 2003, Woods had already stated that beliefs "seem to be intrinsically related to people's selves and identity" (p.225).In respect specifically to teachers, Sakui and Gaies's study (2003) showed how the beliefs of a teacher are related to her professional identity.Studies in ecological linguistics have also suggested the interrelationship between beliefs and identities since to perceive the world is to coperceive oneself (van Lier, 2004citing Gibson, 1979).This means that when "we perceive something, we perceive it as it relates to us" (van Lier, 2004, p. 91).Thus, to believe is to ascribe meaning to the world and to ourselves, and when we do this, we are constructing our identities in the world.In short, as illustrated in Table 3, so far the relationship between identities and beliefs has only been suggested in AL.This relationship is best characterized up to now as interactive.Thus, we can conclude that beliefs are related to one's identities and the more central a belief, the more important it is to one´s identities.

Table 3 The relationship between beliefs and identities Education/philosophy
Applied linguistics Ecological linguistics Core beliefs are more related to a person´s identity and self (Rokeach, 1968) Pet beliefs (Dewey, 1933): more related to identities and emotions Beliefs are related to self and identity (Woods, 2003) A teacher's identity is made up of her beliefs (Sakui & Gaies, 2003) To believe is to ascribe meaning to the world and as we do this we construct our identities (van Lier, 2004)

Identities and emotions
Regarding the relationship between emotions and identities, there have been very few studies in AL that have addressed this issue.In identity research in AL, Norton (2001) has examined identities together with the construct of investment.Investment is defined as "the socially and historically constructed relationships of learners to the target language, and their often ambivalent desire [emphasis added] to learn and practice it" (Norton, 2001, pp. 165-166).Although desire is mentioned in this definition, the investigation of desire itself as an emotion has not been the focus of studies in AL.10 Indeed, research on identities has not yet adequately dealt with many aspects of emotions, despite the fact that aspects of our identities are shaped by pain, fear and desire: "We are what we fear, we are what we desire" (Lemke, 2008, p. 27).These persistent fears and recurrent desires across time and situations help define our long-term identities (Lemke, 2008, p. 27).For example, learners' fears of speaking English in class or outside and their desires to go abroad to become fluent are related to their identities.
In social sciences, identities are related to emotions by the "emotional factors and pressures that lead us to 'identify' only with one or a few of the available social types in our communities" (Lemke, 2008, p. 37).The emotional identification and emotional reasons influence us to assume identities that may seem conflicting and incompatible, play at them, or try them on in private (Lemke, 2008).In addition, we use our emotions to make claims about our identities (such as when I say "I love English," I may be making a claim about my identity as someone who loves the language and may feel good about studying it).
In the field of education, some researchers (Day, 2004;Zembylas, 2004Zembylas, , 2005) have emphasized how emotions are inextricably tied to teachers' identities and have an essential role in understanding teacher thinking, reasoning, learning, and change.Teachers' identities are influenced by how they feel about themselves and their students.Their professional identity helps them position themselves in relation to students and make adjustments in their practice and their beliefs in order to engage with students (Day, 2004, p. 46).Emotions and teaching are epistemologically and constitutively interrelated (Zembylas, 2004) and, thus, teaching is defined as "a way of being and feeling, historically, in relation to others" (Zembylas, 2005, p. 469).In Zembylas's (2004) study of a secondary school teacher, the findings reveal how cultural, political, social and institutional factors influenced the experiences and expressions of the teacher's emotions and played a role in the construction of her "sense of teacher-identity " (p. 195).In this way, the study showed how emotions are significant aspects of identity formation and vice versa, and essential in the structures of power to "constitute some teacher-selves while denying others" (p.936).Thus, like identities, emotions are also performative, that is, the ways in which teachers understand, experience, perform, and talk about emotions are highly related to their sense of identity (Zembylas, 2005).In short, as shown in Table 4, identities and emotions are closely related either as part of the selves or as aspects of identities.We are shaped by the emotions we feel (fear, desire, joy, love), and these in turn shape the kinds of identities we construct of ourselves.
Table 4 The relationship between emotions and identities

Applied linguistics
Social sciences Education Self: includes desires, hopes and dreams (Ryan & Irie, 2014) Self: dynamic system of beliefs and emotions (Mercer, 2014) Identities: shaped by fears and desires Emotions help us identify with social types (Lemke, 2008) Emotions are aspects of teacher identities Teacher identities: composed of how they feel about students (and their jobs, and themselves) Teachers' identities are revealed through their way of talking about emotions (Zembylas, 2004)

Emotions and beliefs
Next, I wish to explore more deeply the relationship between emotions and beliefs.If we look at the root of the word belief, we see that it already reveals the interrelatedness with emotions.Syal (2012) explains that the word belief comes from two words: Be -and -life.Be comes from being, To be is to live.lief comes from the Indo-European word leubh which means love.Thus, belief means to be in love with.A true belief is something that resonates both in heart and mind.
Despite this etymology, and despite the fact that beliefs and emotions have been shown to be interrelated, with emotions being influential for thought and action, scant attention has been paid to this relationship between the two in empirical studies.
In social psychology, the relationship between beliefs and emotions has been characterized by researchers (Frijda et al., 2000;Hannula, Evans, Philippou, & Zan, 2004) as interderdependent and reciprocal (Hannula et al., 2004).According to Frijda et al. (2000, p. 3), emotions "are at the heart of what beliefs are about," and this intimate relationship can be seen in the different ways that emotions affect beliefs.
First, emotions provoke changes in the mental operations such as mental image productions in the brain as well as bodily changes (Damasio, 2004).Thus, when one is sad, there are fewer new images being produced and a constant rerun of the same old sad images; when one is happy, new images of different scenarios and possibilities are constantly created.Second, emotions validate and provide evidence for beliefs and guide our attention towards information that is relevant to our goals (Winograd, 2003).In this case, we search for evidence that suits our agenda and disregard the existence of evidence that contradicts our story.Third, emotions awake, intrude into, and shape beliefs, creating, altering, amplifying, and making them more resistant to change (Frijda et al., 2000).
Four features may explain the influence of emotions upon beliefs (Frijda & Mesquita, 2000).First, there is instrumentality, which refers to "entertaining thoughts to help us achieve our emotional goals" (p.64).Thus, our emotions make us accept existing beliefs, or strengthen beliefs based on their functions.Thus, when we are in love, "small signs of attention are taken as signs of affection because they open the gate to further approach.An ambiguous statement ('I really like you') is understood as an understatement ('she really likes me!') because it encourages action" (p.65).Second, emotions have motivational force, which refers to our motivation to get rid of discomfort, or to achieve pleasure and harmony.Third, there is a concept called control of the scope of thought, which has to do with how we seek information that is relevant to our goals and avoid wasting time and energy attending to information or details that are irrelevant.Emotions therefore direct our attention to goals and information which is deemed most relevant.Fourth, motivated bias refers to the processes whereby humans entertain beliefs and look for, retrieve and generate information that supports their emotional goals.This motivational bias makes beliefs more resistant to change as the person becomes closed off to and not interested in alternative views.
Other studies can be found in the field of language teacher education, with emotions seen as a functional component of a teacher's cognitive development (Golombek & Dorian, 2014;Johnson & Worden, 2014) and as essential to teachers' conceptual changes (Kubanyiova, 2012).In language learning, the increasing recognition of the inseparability of emotion and cognition stems largely from a sociocultural approach to these constructs (Imai, 2010;Swain, 2013), from studies based on the biology of cognition originating from the works of biologist Humberto Maturana 11 (Aragão, 2011), and from a more critical/poststructuralist approach to emotions (Benesch, 2012;Motha & Lin, 2014).Aragão (2011), for instance, has argued that the relationship between embarrassment and self-esteem influenced by beliefs associated with a student's self-concept "play[s] a fundamental role in the way students see themselves in class and how they behave in their learning environment" (p.303).These studies show the importance of the interrelationship between emotions and beliefs in language learning and teaching.
To sum up, the review so far has suggested that the relationship between beliefs and emotions is dynamic, interactive and reciprocal, with cognitions influencing emotions and vice versa (Clore & Gasper, 2000;Frijda & Mesquita, 2000;Parkinson, 1995).Feeling is believing (Clore & Gasper, 2000, p. 30) and "emotions are believable" (Winograd, 2003, p. 39).In this "emotion-belief spiral," emotions generate and sustain certain beliefs, and these in turn further support emotions (Frijda & Mesquita, 2000, p. 49).Accordingly, beliefs are part of emotions as they give meanings to events.In other words, beliefs are inherently emotional.In addition, emotions not only involve the beliefs formed long ago or at a given moment but also stimulate the elaboration of beliefs.In other words, emotions are processes in which information generates emotional responses, which, in turn, generate new information.This rumination helps prolong the life of temporary beliefs, making them more persistent.The more you think about some beliefs, the stronger they become in the neural pathways, making them more resistant to being challenged.

Unveiling the relationship between beliefs, emotions and identities
In this paper, I have attempted to bring together the concepts of beliefs, emotions, and identities, trying to unveil their interrelationships.In order to do that, 11 Humberto Maturana is a Chilean neurobiologist who conceptualized the biology of cognition, "a theoretical framework which characterizes living beings as dynamic systems operationally closed to information from the environment, treated as beings in constant transformation in their living and in permanent structural coupling with the environment" (Aragão, 2005, p. 104).
I have briefly reviewed studies in AL and other areas considering how AL has dealt with these three aspects separately and in combination.Although the relationship between these three concepts has been a concern of psychologists for about two decades now, in AL the discussion of this relationship between all three remains scarce so far.For the most part, AL still tends to view these three concepts in isolation.Studies examining beliefs rarely recognize them as having an emotional component and, similarly, studies on identities tend not to explicitly include an emotional or affective component.
However, the review of the studies in psychology has suggested that emotions and beliefs are connected in complex ways, where beliefs and emotions influence each other interactively.Emotions shape beliefs intensifying them, making them stronger or weaker, creating beliefs, altering them, and these beliefs in turn shape our emotions.In changing our beliefs and our emotions, we are constructing different identities within what is possible or allowed to construct given the power structures in society in general.In addition, emotions are part of our identities.We are what we believe and what we feel.Beliefs form and are part of our identities, and emotions lie at the core of what beliefs are about.
Perhaps we can talk about emotions-beliefs-identities-in-practice to indicate how these have been formed over the years from interactions in different settings and are continually being reconstructed through our past and present situated experiences.Interestingly, all three concepts are usually described in similar ways in recent studies: Identities are multiple, dynamic, hybrid, contested and conflictual; beliefs are complex, dynamic and contradictory; and emotions are active, interactive, processual, and culturally, socially and discursively constructed.It may be that they are all part of one network and cannot be looked at separately; or rather, they are perhaps part of the same continuum, or parts of the same nucleus that forms our "selves."As was suggested by Woods (1996) with his use of the term beliefs-assumptions-knowledge (BAK), we can speculate and say that we may be talking about beliefs-emotions-identities (BEI).They may all be part of a complex system (Larsen-Freeman & Cameron, 2008), with nonlinear behaviors that "change and adapt in response to feedback" (p.2).We could also interpret this relationship from the perspective of distributed cognition, which sees "languaging as inherently affective" (Jensen, 2014, p. 1).According to Jensen, "emotion and affect are studied, not as inner mental states, but as processes of organism-environment interactions."Thus, emotions, beliefs, and identities change and adapt in response to any changes within themselves and within each other.They interact at different times but also simultaneously.
I suggest that looking at these three aspects together may bring interesting insights especially to the issue of teacher/learner cognition change and the relationship to action.Change is not only cognitive; it is also emotional and it affects our identities (Kubanyiova, 2012).Studies in education have already started looking at how emotions are related to the process of change and teacher identities (Zembylas, 2004(Zembylas, , 2005)).In this respect, there is clearly a gap in AL.The gap refers to not seeing emotions as part of identities and beliefs and not looking at this triadic relationship.
In terms of research, issues of terminology need to be discussed.If these concepts are part of one network, or are part of a continuum, how should we talk about them?How should we investigate them?Should we use, then, the inclusive concept of self, as defined by Mercer (2014)?What implications does this bring for research?How can we observe this in the classroom and other language learning and teaching contexts?Rubio (2014) goes further and asks: "Which comes first -the feeling or the thought?Can one go without the other?Can you do research focusing on just one of them independently of the other?"(p.42).He believes we do not have answers to these questions yet.I contend that identities, beliefs, and emotions work together to modulate our actions in language learning and teaching.Therefore, we need empirical studies that investigate these combined interrelationships in the actual life of language classrooms and other learning contexts.
In terms of teaching, because beliefs, emotions, and identities are embodied in contexts, if we want to see their interrelationship we need to see that what happens in the classroom influences how learners construct their identities, emotions, and beliefs in that group.Thus, our concern should be not only about what learners or teachers bring to the classroom but also about the sorts of emotions that are constructed in the interaction within the classroom and inside schools, and in the discourses and practices in the classroom, as Zembylas (2005) suggests.It is a social constructivist perspective.How do learners and teachers see themselves?What kind of identities, emotions, and beliefs are they constructing in our classrooms?Which emotions are they allowed to express and construct?How do these emotional rules or norms shape their practices and discourses about learning English in and outside a classroom?In which ways do learners and teachers help (or not) in the construction of their (imagined) identities, emotions and beliefs?Furthermore, what beliefs are available in our community for our learners and teachers?
In practical terms, it is possible that talking about these issues in class may help to bring them onto the surface and help teachers and their learners to construct more learning opportunities in class.Discussions could be prompted by reading texts which help learners and teachers metacognitively reflect about their own emotions and beliefs (see Andrés &Arnold, 2009 andMurphey, 1998 for excellent discussions of learners and teachers' emotions).Another suggestion is to ask learners to write their language learning histories or draw how they feel, what they believe, and how they see themselves in relation to learning a language (Mercer, 2006;Oxford, 1995).We listen to our students' emotions and beliefs not in order to change them, but to "recognize them and use them to help us understand when -and how -it is appropriate to intervene" (Swain, 2013, p. 11).Having learners analyze and reflect about their own data and discuss them with others in class is also good practice as recent research has suggested (Murphey & Falout, 2010).Finally, I would add that we can ask learners to reflect on the relationship between their own beliefs, emotions, and identities, such as by asking them to work with collages, mini-autobiographies and reflective questions to help them understand their desires and emotions in EFL learning and teaching, as has been suggested by Motha and Lin (2014).
For AL, this emotional or affective turn (Pavlenko, 2013) may mean a new way of defining and researching emotions either based on a sociocultural (Imai, 2010;Swain, 2013) or critical approach (Benesch, 2012), or simply understanding emotions as part of cognition and, as such, as embedded in our identities and in how we relate to the world holistically.This understanding may help us learn, teach and research emotions-beliefs-identities in interaction.That is what I hope.

Table 1
Timeline of research on emotions in AL

Table 2
Characteristics of emotions