CHAPTER 1:
INTRODUCTION
Ultimate language learning success is said to be influenced by many factors out of which the investigation of those grouped as individual differences in language learning have been the focus of many investigations in applied linguistics.
Nevertheless, language anxiety, as opposed to other individual variables such as age and aptitude, is still considered to be a relatively new area of study. Kleinmann’s (1977) investigation of avoidance behavior in language production is cited to date as one of the first empirical studies that pinpoint the importance of affective factors in language learning and production, which touched on the facilitating and debilitating effects of language anxiety. At the end of the 1970s, Scovel (1978) emphasized that results of investigations into the role of affect in language acquisition are still ambiguous and mixed and called for more research in the field. Since then, language anxiety research has come a long way: Scholars have defined the construct of language anxiety, they have devised theoretical models of language learning that assign anxiety an important role in language learning, they have developed self-report instruments to measure the level of language anxiety learners experience, and researchers have conducted numerous studies to gain a better understanding of how anxiety influences the language learning process.
Among the many synonymous terms of apprehension, inhibition, frustration and aversion, applied linguists have termed this affective construct as foreign language anxiety (FLA) sometimes interchangeably used with foreign language classroom anxiety (FLCA). Foreign language anxiety is defined by Horwitz, Horwitz
and Cope (1991) as “a distinct complex of self-perceptions, beliefs, feelings, and behaviors related to classroom language learning arising from the uniqueness of the language learning process” (p. 31).
Since the 1990s, with the appearance of the Foreign Language Classroom Anxiety Scale (Horwitz et al., 1991), research into foreign language anxiety has proliferated. Through studies conducted under a positivist flag with quantitative designs there is ample evidence of anxiety’s reciprocal relationship with proficiency scores (Horwitz, 1991; Horwitz & Young, 1991; MacIntyre, Noels, & Clément, 1997;
MacIntyre & Gardner, 1991b; Reid, 1999; Sparks et al., 1997; Sparks & Ganschow, 1993a; 1993b, 1995, 2000) and final course grades (Aida, 1994; Horwitz, 1991;
MacIntyre & Gardner, 1994). Researchers also conducted empirical investigations of the process of how anxiety can affect learning. The linear input-process-output learning models (Eysenck, 1979; Tobias, 1980, 1986; MacIntyre & Gardner, 1994) and the hierarchical model of willingness to communicate (WTC) (MacIntyre, 1999, 2007; Liu & Jackson, 2008) formed the basis of these investigations. Moreover, a number of correlation studies have been conducted looking at language anxiety and self-perceptions of competence (Young, 1991a; Bailey, Onwuegbuzie & Daley, 2000;
MacIntyre, Noels & Clément, 1997; Onwuegbuzie, Bailey & Daley, 1999); language anxiety, motivation and willingness to communicate (e.g. MacIntyre, Clément, Dörnyei & Noels (1998)); language anxiety and learning styles (Bailey, Daley, &
Onwuegbuzie, 1999); language anxiety and self-efficacy (Wong, 2005; Mills, Pajares,
& Herron, 2007). Many more studies on individual differences (especially on language learning strategies and language learners’ beliefs) do not explicitly aim at focusing on language anxiety, but ultimately in their concluding remarks also mention
related ideas of ‘aversion’, ‘apprehension’, ‘inhibition’ or ‘frustration’ and
‘discomfort’.
In addition to the above, studies conducted under the constructivist flag with qualitative designs have provided the field with more insights into the causes of language anxiety (Bailey, 1983; Price, 1991; Spielmann & Radnofsky, 2001) and coping strategies (Kondo & Ying-Ling, 2004) learners use to deal with anxiety.
Exploratory studies have been conducted to investigate how anxiety interacts with other individual variables (Yan & Horwitz, 2008), venturing as far as to propose a grounded theory of how individual differences influence one another and ultimately how they affect language learning.
In the continuing search to understand the mechanisms of language anxiety specifically, and the role of individual differences in language learning more generally, applied linguists by now have formed contrasting views of language anxiety. For some experts language anxiety denotes an individual variable that greatly influences foreign language learning (see Gardner’s socio-educational model of language learning in Gardner & MacIntyre, 1992, 1993). There are researchers who have explicitly declared that language anxiety is the variable that can best predict foreign language learning success (Aida, 1994; MacIntyre & Gardner, 1991a; Horwitz
& Young, 1991). MacIntyre’s (1999) cyclical model of language anxiety as a situational type of anxiety stemming from the repeated momentary experience of state anxiety clearly highlights the crucial role of this variable and the importance of dealing with it in view of successful language attainment. Last but not least, stressing the importance of foreign language anxiety, applied linguistics journals seem to dedicate special volumes to the topic (a special issue of Foreign Language Annals is planned for 2009).
On the other hand, for other experts language anxiety is an individual variable that cannot be neglected, but its importance lags behind that of other individual variables such as language aptitude, language learning motivation, and language learning strategies. This is evident from the fact that overarching syntheses of individual differences normally devote relatively smaller space to discussing language anxiety (Ellis, 1994; Gardner, 2001; Dörnyei, 2005.) All in all, these reviews and other models of language learning derived from empirical studies (see Gardner &
MacIntyre, 1992, 1993; Gardner, Tremblay, & Masgoret, 1997) acknowledge that language anxiety plays a role in the underlying mechanisms of language learning, yet they provide little explanation as to how anxiety specifically influences other individual variables or how individual variables influence anxiety.
The aim of this dissertation research is to place the available pieces of a complex puzzle together by way of synthesizing what is known about language anxiety in a detailed dynamic multidisciplinary model of the evolution of foreign language classroom anxiety where (1) foreign language classroom anxiety is defined as a psychological construct within a psychological framework, and (2) individual variables that have been associated with language anxiety are accounted for and their relations are explained. It is hoped that this multidisciplinary view of language anxiety will provide those primarily interested in the dynamics of language anxiety with more insight and detail of anxiety’s developmental nature. For those interested in individual differences in language learning other than anxiety, this study intends to shed light on how language anxiety may appear in their investigations of other individual variables and language learning. All in all, the overarching emphasis is on the assumption that we can obtain more relevant information about the psychological
mechanisms that underlie language learning by considering the interplay of individual variables within a larger scheme rather than by viewing them separately.
The dissertation is divided into nine chapters. Following the introductory chapter, chapter 2 provides a theoretical overview of foreign language anxiety by explaining its psychological underpinnings, associated effects and causes of the phenomenon. In chapter 3 a theoretical model of the evolution of language anxiety is proposed with empirical evidence of previous studies presented as support. Next, the methods of the dissertation research are presented: Chapter 4 contains the research questions and hypotheses that guided the study, it elaborates on the selection and characteristics of the participants, the data collection instruments. It ends with an overview of the data collection and data analysis procedures. Chapter 8 reviews the research questions and discusses the answers to the questions based on the findings. Furthermore, the theoretical model and the structural model are compared and additional important findings concerning the instruments are presented. Finally, chapter 9 draws conclusions, emphasizes the pedagogical implications of the dissertation study and puts forth suggestions for further research.
CHAPTER 2:
BACKGROUND
This chapter sets out to present the foundations of foreign language anxiety research from a multidisciplinary perspective. First of all, the chapter will begin by outlining the typology of anxiety, placing language anxiety within the framework of trait, state and situation-specific anxiety. With emphasis on the situational character of language anxiety, psychological theory of context specific coping mechanisms will be reviewed. Turning towards the applied linguistics perspective of language anxiety, the effects and possible sources of language anxiety will be enumerated, and the most prominent models of how language anxiety influences the learning process will be presented. Finally, in an attempt to further understand how language anxiety may develop in the classroom, a summary of three qualitative pilot studies conducted by the author will be presented.
2.1 Basic Psychological Assumptions about Foreign Language Classroom Anxiety
First and foremost, similarly to all language anxiety studies, the basic assumptions underlying the construct of language anxiety were borrowed from psychology. Psychologists differentiate anxiety from fear based on the supposition that the source of anxiety is general and unknown, whereas the source of fear is usually more specific (Levitt, 1980). Another closely related concept to anxiety is
stress. Stress, in this study will be referred to as a collective term for the sum of
“circumstances calculated to arouse anxiety in the individual” (Levitt, 1980, p. 32) and the stressors will be viewed as the inducers of this generic phenomenon. Anxiety in the other hand is viewed as “the subjective feeling of tension, apprehension, nervousness, and worry associated with an arousal of the autonomic nervous system”
(Spielberger, 1983, p. 1) and is a form of “a disproportionately intense reaction” to stress (Levitt, 1980, p.30).
2.1.1 Trait, State, and Situational Anxiety
Eysenck (1979) drew a distinction between two types of anxiety: trait and state. Trait anxiety is by definition a personality characteristic, whereas state anxiety is a momentary experience of the negative feeling. Horwitz, Horwitz and Cope (1991), along with MacIntyre and Gardner (1991a, 1991b) and MacIntyre (1999), proposed that foreign language anxiety, or more precisely, foreign language classroom anxiety, belongs to a third type of anxiety, namely situational anxiety, which is state anxiety experienced to recur in the well-defined situation of the foreign language classroom.
It may be argued that language anxiety is no other than trait anxiety appearing in the foreign language classroom. However, research seems to show that trait anxiety does not predict FLCA (Horwitz, 1991; MacIntyre & Gardner, 1989, 1991b). What is more, Dewaele (2005) found that the same group of learners studying two foreign languages simultaneously manifested different levels of foreign language anxiety at an individual level, which suggests that trait anxiety should be treated separately from
language anxiety (see also Dewaele, 2002). Nevertheless, Rodriguez and Abreu’s (2003) study proposed otherwise as they found that there are no significant differences in the level of FLA of students studying different foreign languages.
However, since the researchers applied a between-participants design, only speculations can be made with reference to the anxiety provoking nature of the two foreign languages but not with reference to individuals’ personality characteristics.
2.1.2 Coping
Similarly to how the psychological construct of anxiety experienced in life- events was adapted to describe foreign language classroom anxiety, the psychological mechanisms of the coping processes characteristic of how people cope with illnesses and other life-events can be adopted and tailored to the language learning situations.
Coping strategies have been described with the use of several dichotomies such as ‘flight’ vs. ‘fight’ behaviors (Ehrman, 1996), ‘active’ vs. ‘avoidant’ (Holahan
& Moos, 1987), and ‘emotion-focused’ vs. ‘problem-focused’ (Lazarus, 1993) approaches. Of the several existing models of coping (for an overview, see Oláh, 2005) the most adequate one to adopt in the discussion of foreign language anxiety and the one the model described above builds on is the transactionalist model of coping devised by Lazarus (1993). This model views coping as a dynamic, interactive process between the individual and their environment rather than a personal trait. As such, it can add to our understanding foreign language classroom anxiety, which has been argued to be situation specific and not trait dependent.
According to Lazarus (1993), coping is an “ongoing cognitive and behavioral effort to manage specific external and, or internal demands that are appraised as
taxing or exceeding the resources of the person” (p. 237). Within this framework, we can differentiate between problem-focused and emotion-focused coping. Problem- focused coping involves the person changing one’s environment or oneself. It seems logical to assume that learners experiencing lower levels of anxiety primarily tend to use problem-focused coping strategies where learners approach the stressor as a problem that can be solved on a cognitive level. By arguing from analogy, a problem- focused coping strategy to resolve anxiety evoked by the language learning situation can be the parallel of cognitive learning strategies learners employ to solve a language learning task. If these strategies are adequate, in other words not only are the learners relieved of their anxiety but they also succeed in focusing on and performing the language learning task, it can be said that such anxiety facilitates learning.
Emotion-focused coping deals with the emotion (in the present case, foreign language anxiety) in a way that it is reduced by, for example, denial or distancing (i.e.
via use of defense mechanisms). In case of higher levels of anxiety experienced in a given situation learners can be assumed to primarily resort to emotional focused coping strategies and aim at alleviating anxiety by way of primarily targeting the emotion of state anxiety. In such cases, rather than focusing on solving the language learning task, learners will, focus on alleviating their anxiety by, for example, employing avoidance behavior. Such levels of anxiety are therefore considered debilitating as they inhibit the learner from performing the language learning task (cf.
Kleinmann, 1977; Scovel, 1978).
With respect to coping with language anxiety, a limited number of studies have been published. Kondo and Ying-Ling (2004) conducted research to explore the coping behavior of Japanese university students studying English as a foreign language. The authors identified 70 different tactics which they grouped into five
categories of affective and cognitive strategies: preparation, relaxation, positive thinking, peer seeking and resignation. Interestingly, these strategies are similar to those delineated by Oláh (2005) based on Lazarus (1993): emotion-focused strategies, resignation, tension control, self-punishment, problem-focused strategies, diverting attention and letting out emotions.
In summary, for the purposes of the present study, language anxiety and the accompanying coping strategies are viewed as situation specific. This context dependent view of the two psychological phenomena constitutes the basic assumptions of the present study on language anxiety.
2.2 Language Anxiety and Its Effects
For a dynamic view of anxiety the diversity of its psychological effects must also be taken into account. Anxiety can have contradictory effects on learning, and as such literature mentions both debilitating and facilitating types, where debilitating anxiety poses an obstacle to learning, whereas facilitating anxiety facilitates or fosters it. In psychology, the Yale theory proposes (Mandler & Sarason as cited in Levitt, 1980 p. 142) that anxiety is a learned drive or a response to a situation which can in fact involve task irrelevant reactions, this way debilitating performance. The theory also states that the effect of anxiety depends on the attitude of the instructor and “the meaning of the task as perceived by the individual” (p.132, Levitt, 1980) and not necessarily task complexity as proposed by Eysenck (1979) and Csíkszentmihályi (1997).
The Yerkes-Dodson law (as cited in Levitt, 1980), on the other hand, establishes that the emotional drive and learning are in a curvilinear (inverted u) relationship and anxiety is largely dependent on task complexity. In other words “[a]
small amount of anxiety is insufficient to improve performance. A moderate amount energizes the individual and thereby improves performance. Further increments are likely to be disruptive” (Levitt, 1980, p.148). This suggest that moderate levels of anxiety can have a facilitating effect, while higher levels of anxiety may lead to a debilitating effect on performance.
In connection to the specific experience of language anxiety, Kleinmann (1977) investigated the avoidance behavior of Arabic and Portuguese adult language learners with respect to their use or avoidance of particular syntactic structures. The author found that psychological measures were more precise in predicting the production of certain syntactic structures where contrastive analysis suggested otherwise. These results imply that facilitating anxiety encourages learners to attempt the production of syntactic structures that were predicted by contrastive analysis to be too difficult for learners to produce. Bailey (1983) further proposed that facilitating anxiety causes the learner to invest more effort into language learning, as s/he becomes more competitive.
Although the existence of both facilitating and debilitating anxiety has been established in psychology, besides the above two pieces of research, language anxiety literature has mostly concentrated on the latter (Scovel,1978; MacIntyre & Gardner, 1991a, 1991b). MacIntyre (1999) in his overview of language anxiety delineates the negative academic, cognitive, social and personal effects of language anxiety. With respect to language anxiety and academic achievement, researchers (Aida, 1994;
Gardner & MacIntyre, 1994; Horwitz, 1991) found negative relationship between
end-of-course grades and learners’ level of language anxiety. Furthermore, Gardner, Smythe, and Lalonde, (1984) along with MacIntyre, Noels, and Clément (1997) established that self-perception and language anxiety are inversely related. In terms of anxiety’s effect on cognition and learning, MacIntyre and Gardner (1991a, 1994) found that language anxiety exerts a negative influence at the input, process and output stages. What is more, research studies show that learners’ performance in the use of the four skills is inhibited by listening (Vogely, 1998), writing (Cheng, Horwitz, & Schallert, 1999) and reading anxiety (Saito, Garza, & Horwitz, 1999) respectively. The social effect of anxiety is mainly associated with the cultural tension arising from learning a language in a second language environment (Clément, Gardner, & Smythe, 1980; MacIntyre, 1999) and with communication apprehension resulting from avoiding communicative situations. Finally, MacIntyre (1999) suggests that the personal effects of anxiety involve low self-worth and negative self- evaluation of one’s linguistic performance.
Besides the effects mentioned above, similarly to anxiety in general, language anxiety also has physical and behavioral effects. Physical effects of anxiety are generally indistinguishable from the physiological symptoms of fear (Levitt, 1980), and they encompass sweaty palms, increased heart rate, red cheeks, trembling, fidgeting. Often anxious learners experience stage-fright and they freeze in front of others, they demonstrate avoidant behavior and in the most extreme cases oftentimes fail to attend language classes and finally drop out of language courses.
2.3 Sources of Language Anxiety
Probably the most controversial issue of anxiety entails its possible sources.
As it appears in the definition, these sources cannot be specified (cf. source of fear) and the person is normally not aware of what causes the negative feeling (Levitt, 1980). Going back to the basic assumption of language anxiety being situation specific, these unspecified sources are likely to stem from the language learning context. Csíkszentmihályi (1997) in his theory of flow suggests that in our everyday lives and in learning situations, anxiety is bound to arise due to the mismatch between task difficulty and one’s abilities to solve the task. More precisely, if the task is excessively complex to solve relative to the person’s abilities, the task poses an obstacle greater than just the optimal drive and leads to anxiety.
With regards foreign language classroom anxiety, Horwitz, Horwitz and Cope (1991) outline communication apprehension, test anxiety and fear of negative evaluation as the causes of FLCA. They hypothesize that language anxiety is a composite of communication anxiety in general, test anxiety related to the evaluative situations of the language classroom testing situation, and the fear of negative evaluation from the learners’ peers as well as the instructor.
MacIntyre and Gardner (1991a, 1991b) describe FLCA stemming from the negative expectations in foreign language learning. According to Sparks and Ganschow (1993a, 1993b), this can be tied to learners’ previous language learning experiences related to L1. The authors in their Linguistic Coding Difficulties Hypothesis (LCDH) posit that language anxiety arises from linguistic coding difficulties on the phonological, semantic and syntactic level, which can be traced
back to coding difficulties in L1 and thus it is strongly related to language learning aptitude.
Interestingly, Price (1991) in her qualitative interview study found that classroom factors play an important role in the development of language classroom anxiety. All respondents without exception named the main source of anxiety as
“having to speak the target language in front of their peers” (p.105). Making pronunciation mistakes, not being able to communicate effectively, and the difficulty of the language classes were all listed as further sources of anxiety in the language classroom. Price hypothesizes that learners feel anxious in the language class because language learning does not prove to be an easy task, and their achievement lags behind those in other areas. Price, as later Spielmann and Radnofsky (2001), also alludes to the importance of learners’ beliefs about language learning and beliefs about the necessity of a certain level of aptitude to successfully learn a foreign language. Last but not least, Price also established that the language instructor plays a significant role in learners’ feeling of anxiety.
Thus, determining the precise cause of language anxiety is not a straightforward task, especially given the diversity and complexity of the classroom situation. Nonetheless, in order to understand the role language anxiety plays in language learning, its possible sources cannot be neglected.
2.4 Models of Language Anxiety
As to the development of anxiety or how it affects language learning, very few theoretical models have been proposed. Those that have appeared in literature tend to be very broad like the socio-educational model of language learning proposed by
Gardner (as cited in Gardner & MacIntyre, 1992) or are very general like the input, process, output model of Gardner and MacIntyre (1994) and the vicious circle view of the development and effects of language anxiety presented by MacIntyre (1999). The model of willingness to communicate (MacIntyre, 2007) is one that is implicitly very much tied to language anxiety. As one of the most recent proposals for a model of language anxiety, Yan and Horwitz’s (2008) presented a grounded theory which builds on the interaction between variables, but does not cover the state and situational aspects of anxiety. These five models and their shortcomings are described below.
Gardner in his socio-educational model, one of the first comprehensive models of language learning, included language anxiety as an affective factor influencing language learning. Kim (2001) calls it a “passive model at least as far as the language anxiety in the language learning process is concerned” (p. 14). Indeed very little detail is provided as to how language anxiety interacts with or influences other affective and cognitive variables; in the formal context, the model suggests that anxiety has a summative effect on language learning together with motivation, strategies, aptitude and intelligence. The non-linguistics outcome of language learning feeds back into all individual variables, including anxiety. In agreement with Kim (2001), it must be acknowledged that what this model does not offer is a precise view on how language anxiety develops in the well-defined context of the foreign language classroom and how indeed the enumerated cognitive and affective variables relate to one another.
Furthermore, the model lacks elaboration on the definition of anxiety as a construct, and other individual differences variables such as self-confidence, self-efficacy, willingness to communicate, and age are not accounted for.
The input, process, and output model of the effect of anxiety (Gardner &
MacIntyre, 1994) is based on Tobias’ (1986) model of learning. This model posits that learning involves three stages: the input stage, where the learner is exposed to the information; the process stage, where the information is cognitively processed; and the output stage, which is a production stage where the material learnt is utilized.
Gardner and MacIntyre (1994) theorized that language anxiety exerts its influence at these stages respectively. At the input stage, anxiety appears as new material in form of new words, grammatical structures etc. presented in the foreign language. Here anxiety obstructs the intake by upsetting the learner’s concentration and disrupting the encoding of information. Next, as the information is processed, organized and prepared for storage, anxiety may pose an obstacle and the efficiency of these processes that involve understanding, and memorization is reduced. At the output stage, anxiety impedes the production of the linguistic material that has been learnt via hindering the retrieval of the processed material.
Although the input, process and output anxiety model aims to explain the process of language anxiety’s influence on language learning, it fails to capture the situational factors involved in the foreign language classroom context (Kim, 2001).
Interestingly, the model accounts only for the debilitating effect of anxiety on learning. What is more, the model rather concentrates on explaining the effects of language anxiety while its sources are not elaborated on. Finally, in contrast to the socio-educational model of language learning, this model does not specify the relationship of language anxiety with other individual variables that have been found to be associated with it.
The third model of language anxiety that is important to mention here is the one outlined by MacIntyre and Gardner (1989). This model describes a vicious circle
the language learner may become part of when experiencing language anxiety. First, learners encounter difficulties in language learning (e.g. difficulties in comprehension, in grammar, in learning vocabulary, etc.), then they experience state anxiety about that is directly related to this difficulty. If the experience of state anxiety is repeated, and the learners come to associate it with the language learning context (i.e. foreign language classroom), learners develop foreign language anxiety.
Like the socio-educational model, the vicious circle view of language anxiety is dynamic in nature; however, similarly to the input, process and output model, it does not explicitly mention the relationship of other factors and anxiety, nor does it allude to how other factors may contribute to the development of anxiety, save for the generalization that the learner experiences difficulties. As the input, process and output model, it concentrates on the debilitating nature of anxiety and does not explain the facilitating effect anxiety may have on language learning.
Finally, the model of willingness to communicate (WTC) must be touched upon as it tends to incorporate an element of anxiety which ultimately plays a role in learners’ willingness (MacIntyre, 2007) or unwillingness (Liu & Jackson, 2008) to communicate. It is interesting to note that the model of WTC takes the shape of a six layered pyramid and suggests a hierarchical relationship between individual variables that are associated with WTC. In the fifth layer (second from the bottom up) the author includes affective and cognitive determinants of WTC such as intergroup attitudes, social situation and communicative competence. In layer four (third from the bottom up), motivational propensities are listed, such as interpersonal motivation, inter-group motivation, and self-confidence. Anxiety is not explicitly an element of the model, but MacIntyre (2007) in his theorization does suggest that state anxiety and the volitional (moment-to-moment) aspect of WTC are very much related. Liu and
Jackson (2008) with measures of unwillingness to communicate have found statistically significant relationships between learners’ language anxiety and unwillingness to communicate from a situational point of view, but no empirical evidence of the volitional aspect of anxiety and WTC have been presented yet.
The model of WTC, although it does not explicitly incorporate an element of language anxiety, comes closest to describing the process of how language anxiety may affect language production. At the same time, it is limited to just that: accounting for anxiety’s role in momentary oral language production and not viewing the overall process of language learning. As opposed to the cyclical models of the development of anxiety and the socio-educational model, this model is linear in nature, which means that it cannot adequately describe the possible reciprocal and cyclical relationships between its elements.
Finally, a grounded theory was presented by Yan and Horwitz (2008) of the interdependent relationships of individual variables in language learning that arose from a qualitative study based on learners’ own perceptions of how anxiety is affected by other individual variables. This theory incorporates, among others, the components of aptitude, language learning strategies and motivation. Yan and Horwitz, based on learners’ accounts, suggest that peer influence, learning strategies and motivation directly influence language anxiety; whereas, parental influence, teacher characteristics and beliefs about aptitude exert indirect influences on language anxiety. The theory also takes a new approach in that it views motivation and anxiety’s relationship as being reciprocal. Most probably this theory comes closest to describing how anxiety and other variables interact. Yet, the basic psychological assumption and the state-to-situational evolvement of language anxiety are not considered.
Thus far, it has been theorized that language anxiety is a construct distinct from trait anxiety, and its facilitating and debilitating effects on language learning are not to be neglected. It is also clear that not much has been written about the coping mechanisms involved in learners’ attempts to overcome language anxiety. Last but not least, existing theories do not seem to adequately describe the mechanisms underlying the development of language anxiety, as they fail to incorporate in necessary detail the situational or contextual factors at play. In order to explore the sources of foreign language classroom anxiety in the Hungarian secondary school context, studies with qualitative design preceded the development of a theoretical model of language anxiety.
2.5 Exploratory Pilot Studies
In the following section, three qualitative studies will be summarized that preceded the dissertation research. They were conducted with the aim of exploring what may be at the source of Hungarian high school learners’ language anxiety and what kind of coping strategies they resort to. The results informed the construction and testing of a theoretical model of the development of language anxiety (see Chapter 3).
2.5.1 First Study
The first study conducted by the author sought to investigate whether foreign language classroom anxiety is context dependent and explore how learners feel in different foreign language learning classrooms (Piniel, 2004). The participants were 34 secondary school students (23 girls and 11 boys) studying at two different high schools in Budapest. All students were enrolled in two parallel foreign language classes (English and German). Data were gathered through the use of questionnaires (Foreign Language Classroom Anxiety Scale (FLCAS) (Horwitz, Horwitz & Cope, 1991), Attitudes and motivation scale (Kósáné, Porkolábné, & Ritoók, 1984), semi- structured interviews (based on Price, 1991) and observations. Based on the results of the anxiety scales with regards the two foreign language classrooms, eight students whose scores depicted the greatest discrepancy were selected for interviews and were the focus of subsequent classroom observations.
Results showed that the same individual can experience different levels of foreign language anxiety in different foreign language contexts (two different foreign language classrooms). As to the possible causes of these differences, students perceived their previous experiences, self-ratings (self-perceptions), the personality and role of the instructor, how error correction was carried out, evaluation of the teacher, quality of instructions, types of tasks employed during a lesson, motivating content of the lessons to be all important in determining their feelings of inhibition at a language lesson. These results suggest that language anxiety is very much influenced by problems and conflicts arising from the contextual factors of the language learning classroom.
2.5.2 Second Study
The second study was a replication of the first one with a larger sample and with slight modification in the instruments utilized (Piniel, 2006). The two main foci were identical to the first study, namely, investigating language anxiety’s stability across different foreign languages among secondary school students and the possible causes of foreign language anxiety that stem from the foreign language classroom.
Similarly to the previous study, a two-phased research was conducted with the participation of 62 ninth grade Hungarian students who were studying two foreign languages simultaneously at high school. The first phase of the study followed a quantitative, while the second followed a qualitative design.
First, the validated Hungarian translation of the FLCAS (Tóth, 2003, 2008) and the Trait part of Spielberger’s (as cited in Sipos, Sipos & Spielberger, 1994) State-Trait Anxiety Inventory (STAI-Trait) were administered. Through purposive sampling, based on the results of the first phase, six respondents were selected to participate in the qualitative part of the study. The students selected did not possess high levels of trait anxiety, yet demonstrated high levels of anxiety in one foreign language classroom but not in the other. Semi–structured interviews were conducted to explore the possible classroom sources of foreign language anxiety. Findings suggested that foreign language classroom anxiety may develop irrespective of low levels of trait anxiety and independently of the level of anxiety experienced in the other language classroom. From the interview data, the role of the instructor emerged as a central theme in students’ perceptions of the source of their feelings.
The categories that could be set up revolved around the following four main issues: course requirements, language input, the instructors’ personal relationship with
the learners, and the teacher’s personality. The first group of perceived causes of language anxiety seems to originate from the academic expectations of the language course. Participants reported that the methods of assessment, the inconsistency of evaluation, and the lack of clarity with reference to teachers’ expectations were all experienced sources of their anxiety. The second, and the most elaborated category comprised the characteristic of foreign language input. Learners mentioned that the lack of comprehensibility of linguistics input in terms of the quality and quantity of explanations that accompany the presentation of the material was often the source of their frustration. Besides not being able to understand the foreign language, learners’
beliefs about its utility, their dislike of certain topics based on which linguistic input was presented and the materials used to present the new elements of the foreign language also prompted feelings of uneasiness. The final two sources learners’
perceived as anxiety provoking were the teacher’s relationship with the students and the teachers’ personality.
In many ways, these results echo the results of the first study: teacher evaluation (error correction) and methods of assessment seem to be corresponding areas, quality of instructions can be matched with the category of foreign language input in the second investigation, types of tasks employed during a lesson and motivating content of the lessons appeared in the latter study as materials used and content (topics) discussed in the lesson; moreover, the personality of the instructor also appeared as an anxiety related issue in both studies.
2.5.3 Third Study
In the third study, qualitative research was carried out to explore the different ways anxious and non-anxious students deal and cope with foreign language anxiety (Piniel, 2005). With the participation of three secondary students (two anxious and one non-anxious) learning two foreign languages simultaneously at a secondary school in Budapest, data were collected using the validated Hungarian versions of the two anxiety scales (the FLCAS and the STAI-Trait), and semi-structured interviews were conducted according to the techniques suggested by Patton (2002) and Szokolszky (2004). Official documents such as the grade books in the school also provided invaluable data. For the convenience of retrieval, an audit trail was kept of the research process.
The case studies showed an interesting picture of how students with foreign language anxiety are in need of larger inventories of coping strategies that involve more problem-solving strategies than defense mechanisms. Interestingly, the most successful language learner of the three participants (based on her academic achievement) had the largest number of suggestions for coping strategies which she could and apparently did use to cope with the anxiety provoking atmosphere of the language lessons.
Although the other two participants employed coping strategies, they did not seem to resolve the conflicts in the long run. Their pleading emotional support or merely denying the responsibility they had in their language learning and thus fleeing from the problem seem to have long term effects on these students’ language learning achievement. The results of this study suggested that learners do not have an adequate inventory of coping strategies they could resort to in order to resolve conflicts of
foreign language classroom anxiety in a more rational way. Without assistance, students are not willing to take risks which lead them to withdraw from coming into contact with the foreign language as much as possible. Relative to their peers, their linguistic competence does not develop to such an extent and this ultimately decreases their self-perception of competence, giving them the sense that they lack efficacy to learn a foreign language. This eventually gives rise to language anxiety.
2.6 Conclusion
The above chapter described language anxiety as a psychologically based construct that affects language learning to a great extent. Language anxiety has been established as a situation specific phenomenon, which entails that its sources are dominantly context dependent. For this reason, it is important to shed light on the possible sources of language anxiety that stem from the language learning situation.
The diversity of the possible sources of foreign language classroom anxiety could be seen from the literature and the pilot studies that were carried out involving foreign language anxiety in the Hungarian context. Anxiety does not only seem to interweave the process of foreign language learning but also several factors appear to be related to it (self-perceptions, age, personality, learning styles, perceptions of self-efficacy, motivation, etc.).
Although theoretical models incorporating language anxiety exist, they fail to capture the details of its mechanisms and how language anxiety interacts with other individual variables. As an attempt to organize the basic psychological assumptions of language anxiety and the individual variables that seem to be associated with it, a
developmental model based on the above theoretical considerations of foreign language anxiety was drawn up. This theoretical model of the development of language anxiety is presented in the following section.
CHAPTER 3:
A PROPOSED THEORETICAL MODEL
The previous chapter established that language anxiety is a type of situation specific anxiety whose sources are, consequently, context dependent. As previous models fail to describe the development of anxiety by providing detail about its sources and its dynamics, the following chapter will propose a theoretical model of the development of language anxiety. In the first part of the chapter, the model and its elements will be presented in an overview and the second half of the chapter aims to provide theoretical justification of the hypothesized model.
3.1 Overview of the Model
The model of the evolution of foreign language classroom anxiety is founded on previous published empirical evidence, on theoretical considerations in applied linguistics, and on the dynamic mechanisms that characterize anxiety as a psychological construct. The model proposed here consists of six phases (see Figure 1). In phase 1, comprises what Dörnyei (2005) calls the learners’ initial motivation to learn the language. This initial motivation rests on previous experiences, social effects of the learners’ environment (e.g. parents), and ultimate personal goals (ideal L2 self) among others.
Phase 2 represents the language learning situation where the learners’
interaction with themselves, their instructor, and their peers (also see Young, 1991b,
1992) are considered. These play an important role in the self-regulatory processes, as we will later see in phase 3. The points of interaction are viewed as possible sources of stress for the learners and appear as stressors. More precisely, such individual variables are considered here as self-perceptions; learners’ beliefs about language learning in general, more specifically about age and aptitude; learners’ personality traits; learning styles; and learners’ willingness to communicate in view of teaching styles and teachers’ beliefs.
In phase 3 the dynamics of the psychological process of coping come into play. As the stressful situations are encountered, the individual evaluates them in respect of his ‘self’. Lazarus (1991) calls this the ‘cognitive appraisal’. Primary cognitive appraisal evaluates how the event (the stressor) relates to the individual’s personal goals and motivation, whereas secondary appraisal considers the coping resources available to deal with the stressor and the emotion it invokes (cf. Bandura’s concept of self-efficacy).
Whether the stressor is relevant to the goals of the individual is determined in phase 4, and emotions emerge accordingly. Consequently, if the stressor is not congruent with the goals of the learner, negative emotions appear. In terms of ego- involvement, if the ‘self’ is threatened by the stressor, we can also speak of the negative emotion of anxiety (see Lazarus, 1991). The first time this anxiety appears, as in MacIntyre’s (1999) model, it is characterized as state anxiety, in other word, anxiety experienced at a given moment.
In the following phase, phase 5, depending on the learners’ motivation to learn the language, the threat the conflicting situations pose to the individual’s self, and the resources available to resolve these conflicts, the learner will try to cope with the
Figure 1. A model of the development of foreign language classroom anxiety. Phase 5.
PRE- LANGUAGE LEARNING LANGUAGE CLASSROOM EXPERIENCE Phase 4.
Phase 1. Phase 2. Phase 3.
Phase 6.
EFFECTIVE COPING strategies(in terms of the stressor and the negative feeling of anxiety)
LEVEL OF (State)ANXIETY COGNITIVE
APPRAISAL INITIAL
MOTIVATION to learn a foreign language
STRESSOR(S) Points of conflict:
learner-self learner-instructor learner-peers
COPING STRATEGIES directed at the feeling evoked and/or the stressor How threatening
to the self?
What resources are available?
(self-efficacy)
INEFFECTIVE COPING strategies were ineffective, anxiety persists and/or source of stress has not ceased
LEVEL OF (State- Situational) ANXIETY
EFFECTIVE COPING strategies (in terms of the stressor and the negative feeling of anxiety) COGNITIVE
APPRAISAL STRESSOR(S)
COPING STRATEGIES directed at the feeling evoked and/or the stressor
INEFFECTIVE COPING strategies were ineffective, anxiety persists and/or source of stress has not ceased FOREIGN LANGUAGE ANXIETY
due to repeated experiences of ineffectively resolved conflicts (repeated experiences of state anxiety)
situation by targeting the evoked feeling of anxiety and the stressor itself. In the attempt, the learner can resort to several ways of coping.
Phase 5 shows that in either case, there is a wide spectrum as to the outcome of the coping behavior, ranging from effective to ineffective strategies (from the point of anxiety and the stressor). It must be noted, however, that an effective coping strategy from an affective perspective does not necessarily mean a constructive one from the language learning point of view: for example, quitting the course may be an effective way of alleviating anxiety and eliminating the stressor, but it will most probably not lead to successful language learning.
Finally, in phase 6, if the anxiety evoked is not dealt with effectively (i.e. the conflict does not cease to be a stressor) and/or other conflicts further appear as stressors, learners experience repeated instances of state anxiety associated with the language classroom. Thus, foreign language anxiety a situation-specific anxiety particular to the foreign language learning situation evolves (MacIntyre, 1999).
It is important to note that the proposed model is not an alternative one to existing models but rather a more detailed one which builds on the results of previous research in applied linguistics and psychology and attempts to position all that is already available on language anxiety into a more comprehensive dynamic and situational perspective.
This section of the dissertation presented a detailed dynamic multidisciplinary model of the evolution of language classroom anxiety. The dynamic nature of the proposed developmental model cannot be stressed enough. As opposed to previous models of language anxiety, the present model is more detailed in terms of accounting for underlying mechanisms, and as part of its dynamism it suggests that there can be a way out of the vicious circle. More precisely, it also accounts for instances when
through the appropriate choice of language learning strategies the stressor is eliminated, the feeling of anxiety is alleviated and the learner experiences success which can in turn be motivating to continue learning the language and it also enhances self-efficacy which in turn (in the cognitive appraisal) will defend the learner against experiencing higher levels of anxiety.
It must be noted that between the two extreme experiences (the vicious circle of foreign language anxiety and the ultimate conquering of the feeling of anxiety), in reality there is great variability in terms of how threatening the stressor is to the self (in form of a task, of a peers’ evaluation etc.), how and to what extent the learner can combat the stressor, and how the learner approaches new arising stressors or recurring ones.
The proposed model of the evolution of foreign language anxiety also aims at presenting the phenomenon as part of a larger multidisciplinary scheme. Many of the basic elements of the model stem from psychology but the language learning context justifies their relevance in applied linguistics and more specifically in psycholinguistics.
Finally, the model aims to build on previous research conducted on language anxiety to take a more focused path and propose an explanation of language anxiety’s relationship with other individual learner variables while also capturing the mechanisms that underlie this network. The intention was to synthesize and account for the role of individual variables that have been associated with language learning, noting the importance of learning strategies as coping strategies, learners’ personality, aptitude, age, motivation, beliefs, and learning styles.
Like all models, the one presented here is not without flaws. It leaves many questions unanswered, the most important one being what the critical levels of the
quality and quantity of the conflicts (stressors) are that appear in the foreign language classroom respective of the level of FLA. Nevertheless, it can be stated that the model fills a gap in the field of FLA research, as its development has not been dealt with in such detail with regard to individual differences and the foreign language context.
Hence the model presented here on the evolution of foreign language classroom anxiety places foreign language classroom anxiety into a psychological framework of anxiety and coping. Furthermore, it meets the same basic assumptions of foreign language classroom anxiety as proposed in previous studies: namely, that foreign language classroom anxiety is viewed as a situation specific anxiety, which can facilitate or inhibit language learning. Parallel to this, coping is also considered to be a situation-specific process where strategies can result to be effective or ineffective in terms of alleviating anxiety experienced in the language learning situation. Another important point is that as opposed to other existing models, the present one accounts for not only the severity of the vicious circle but also the possibility of language anxiety being alleviated. Finally, the model presented above tries to accommodate the relationships between language anxiety and other individual variables already mapped by previous empirical studies
3.2 Support from Literature: Language Anxiety and Other Individual Variables
The following section intends to clarify how the results of previous studies in applied linguistics support the proposed model of the development of foreign language classroom anxiety. More precisely, evidence from research on individual differences in language learning associated with language anxiety will be presented.
For explanatory purposes of how individual differences variables may interact, the work of Csíkszentmihályi (1997), Higgins (1987, 1999), and Lazarus (1991, 1993) in psychology proved useful and will be referred to at the relevant points.
In demonstrating the role of individual variables in the proposed model, first, the variables related to the stressors are enumerated as elements of discord experienced in the language classroom. These include learners’ beliefs, self-concepts, age, learners’ aptitude, personality traits, learning style, willingness to communicate, and teachers’ beliefs of language learning. Then the cognitive appraisal of Lazarus’
(1993) coping theory will be discussed and its relationship with two other individual variables: learner’s self-efficacy and motivation. Last but not least, a parallel will be drawn between coping strategies and learning strategies, as the final group of individual variables that have been associated with language anxiety.
3.2.1 Stressors
It may appear that research on language anxiety places unjustifiably too much emphasis on the negative side of language learning as stressors, conflicts and worries are discussed extensively. Taking a realistic view, it is unimaginable to expect all learners to experience what Csíkszentmihályi (1997) calls ‘flow’ throughout a language lesson, not to mention throughout a whole language course.
The relevance of flow in the foreign language classroom was investigated in a study by Egbert (2003), where the author proposed a hypothetical model of the relationship between flow and language acquisition. In this model Egbert emphasizes the ‘task’ (cf. Eysenck, 1979) and the learner’s skills and tools (cf. Bandura’s (1988) concept of self-efficacy) that are available to complete the task. What is important to
note is that the experience of ‘flow’ is said to be temporary and it depends on the learners’ ability to focus on a task (be it a language learning task, or any other task).
However, the equilibrium is overturned, and anxiety is experienced when the self is confronted with information that it interprets as threatening. For example, a difficult task that is beyond the abilities of the learner is said to evoke anxiety (Csíkszentmihályi, 1997). Hence, by definition, stressors are events that threaten the learners’ well-being by upsetting the equilibrium that characterizes flow. The model of the development of language anxiety proposes that the equilibrium of flow can be disrupted by stressors which appear due to conflicting points within the network of individual variables involved in language learning. These conflicts can be grouped into two categories: those appearing between the learners’ own expectations and the self, and those appearing between the expectations of others (as perceived by the learner) and the self (see Higgins, 1987).
3.2.1.1 Learners’ Own Expectations
Beliefs and self-concept. Stressors may appear when learners’ own expectations about language learning are in conflict with what is actually experienced in the language learning situation. The first point to look at is the clash between learners’ beliefs about language learning and their self-concept. The most straightforward identification of language learner beliefs is provided by the construct of Horwitz’s Beliefs About Language Learning Inventory (BALLI) (Horwitz, 1988). The instrument was designed to measure language learners’ perception of language learning difficulty, foreign language aptitude, the nature of language learning,
communication and learning strategies, motivations and expectations. As Horwitz (1996) states “[S]ome of these beliefs can be helpful while others can be truly counterproductive for language learning” (p.576). This claim is further supported by the results of an exploratory study conducted by Spielmann and Radnofsky (2001) where learners’ beliefs were shown to negatively affect language learning by way of fostering foreign language anxiety.
What is more, in Horwitz, Horwitz and Cope’s definition (1991) foreign language anxiety is a “distinct complex of self-perceptions, beliefs, feelings related to classroom language learning arising from the uniqueness of the language learning process” (p. 31). Horwitz (1989) in her study confirmed that there was a link between FLA and university Spanish students’ beliefs about their language learning abilities and the perceived difficulty of language learning. She reported that more anxious learners claimed language learning to be more difficult, which was often paired with low self-esteem and negative self-perceptions.
In psychology, the issue of beliefs is treated in relation to self-perceptions and as part of the self-concept. In cognitive psychology, Rogers’ (as cited in Carver &
Scheier, 1998) phenomenological perspective of the self proposed that its constituents can be divided into two parts: the ideal and the actual self. The ideal self is the representation of features the individual would like to have ideally, whereas the actual self is the representation of the features the individual believes to possess at a given moment. Rogers also postulated that incongruence experienced between these two selves leads to anxiety.
Higgins (1987) in his self-discrepancy theory introduces a third component of the self, namely, the ought-to self, which is the “representation of the attributes that someone (oneself or another) believes one should or ought to possess (i.e., a
representation of someone’s sense of one’s duty, obligations, or responsibilities)” (p.
321), and proposes a link between self-cognition and emotion by stating that discrepancy between the actual and the ideal self and dissonance between the actual and the ought-to self both lead to negative emotions. Through experiments and empirical research, Higgins (1987) was able to demonstrate that actual-ideal discrepancies can be linked with depression, whereas actual-ought to discrepancies lead to agitation (tension, anxiety, etc.).
It seems that negative self-perceptions are likely to arise in learners who tend to “set excessively high standards for performance” (Gregersen & Horwitz, 2002, p.
563), are often overcritical of themselves, and possess a fear of failure. Gregersen and Horwitz (2002) refer to these learners as ‘perfectionists’. The authors in their interview study assumed that behavior manifested by perfectionist learners parallels that of language learners with high levels of anxiety. A group of anxious and non- anxious students were asked to participate in oral interviews and comment on their own performance. Indeed, results showed that highly anxious learners’ self-reports include perfectionist characteristics in that they have “higher standards for their English performance, a greater tendency toward procrastination, greater worry over the opinions of others and higher level of concern over their errors” (p. 568) than non- anxious learners.
Age. Similarly, in case of adult learners, self-perceptions and beliefs can come into conflict with the actual experience of language learning. This discrepancy is likely to be rooted in the fact that adult learners view their ought-to and ideal selves capable of expressing complex views not only as native speakers but also as speakers of a foreign language. Hence, beginner adult learners whose actual self is not yet capable of expressing complex ideas enters into conflict with the ideal and/or ought-to
self characterized by well-defined high expectations. As a result, adult learners may possess a stronger subjective feeling of inhibition than a child who is still learning about the world around him, causing adult learners to feel more vulnerable when learning a foreign language (Horwitz, Horwitz & Cope, 1991; Spielmann &
Radnofsky, 2001).
This idea is further supported by the results of studies where learners’ age and level of FLA show correlations. Onwuegbuzie, Bailey and Daley (1999) in their study of factors associated with foreign language anxiety found that the demographic variable of age is also related to FLA. According to the authors, some of the possible causes that may lie behind adult learners’ foreign language anxiety are their beliefs of inability to ultimately acquire native-like pronunciation, their seemingly slower cognitive processing due to cautiousness, and their negative beliefs about language learning in general. Findings of this study also suggest that learners with high levels of FLA have negative self-perceptions as they view their course achievement, their self-worth, and their scholastic competence rather negatively.
Upon repeating their study with MacIntyre and Gardner’s (1994) anxiety instruments, Bailey, Onwuegbuzie and Daley (2000) confirm that “students with the highest levels of anxiety at the input, processing, and output stages tend to be older”
(p. 474). Thus, age theoretically can be indirectly linked to the development of foreign language anxiety. As seen in the line of argumentation above, with age, self- perceptions gain more importance and seem to be sources of discrepancy between the ideal and actual self of the language learner.
3.2.1.2 Expectations of Others as Perceived by the Learner
Teacher’s beliefs. Returning to Higgins’ (1987) self-discrepancy theory, among the different possible discrepancies between the ought-to and actual self, it is the actual/own - ought/other discrepancy that can be significantly linked with anxiety (more precisely, social anxiety). This means that the individual’s perception of his own actual self is incongruent with his perception of what others believe he should be (or ought to be) like, an important aspect of language anxiety that is accounted for in Horwitz, Horwitz, and Cope’s (1991) construct of language anxiety, namely as the fear of negative evaluation.
Exploratory studies (Price, 1991; Spielmann & Radnofsky, 2001; Yan &
Horwitz, 2008) in the field of foreign language classroom anxiety established that there seems to be discrepancy between what students perceive their instructors’
expectations to be (ought-to/other) and what they can actually perform (actual/own).
Price (1991) in her study reported that some students possessing high levels of anxiety had teachers “who criticized student’s accents”, making “classroom time a performance rather than learning time” (p. 106). The outcome of Price’s study indicates that indeed the instructor, as a significant other, can play an important role in developing language learners’ anxiety.
Negative beliefs of the language learning process and incongruent beliefs of the language learner and/or the language teacher about language learning and teaching (see Bailey, 1983; Horwitz, 1988, 1996; Spielmann & Radnofsky, 2001) are also said to foster FLA. Horwitz (1988) suggests that negative beliefs or learner beliefs that are not consistent with teacher beliefs of language learning may cause difficulties in the foreign language learning process: “clash of expectations between students and the
teacher about language learning can lead to a lack of student confidence in and satisfaction with the language class” (p. 290). She especially emphasizes “beliefs stressing the importance of target language accuracy are a contributing factor to anxiety reaction in foreign language learning” (Horwitz, 1988, p. 292).
Aptitude. Discrepancies may also arise from the fact that expectations of what learners ought-to be able to learn at a given pace are different to what they can actually learn, often due to difficulties. Such difficulties in learning a foreign language seem to be related to language learning aptitude (Sparks & Ganschow, 1993a).
Language learning aptitude is an important individual difference that affects the input, process, and output stages of language learning via auditory, linguistic, and memory abilities (Skehan, 1998).
Sparks and Ganschow (1993a, 1993b) proposed that language anxiety is a consequence rather than a cause of language learning difficulties, which in turn are mainly due to coding problems related to language learning aptitude. The authors termed this the Linguistic Coding Deficit Hypothesis (LCDH, later modified to Linguistic Coding Differences Hypothesis (Sparks & Ganschow, 1995)). The hypothesis posits that a language learner experiencing problems of coding (especially phonological coding) in L1 is highly likely to experience such deficits in L2, which causes difficulties in foreign language learning, consequently causing language anxiety. As such, according to the authors’ view, it is phonological, syntactic, and semantic coding capabilities (especially the first two) and thus language learning aptitude that ultimately determine foreign language learning success rather than affective factors. This hypothesis was further investigated in a theoretical paper where the same authors noted that written and spoken phonological processing and reading
