Tendencies for future research on English Speaking Anxiety in Confucian Heritage Culture (CHC) students and teachers

This paper examines the causes and consequences of Foreign Language Anxiety (FLA) in Second Language Acquisition (SLA). Speaking was the skill that contributed most to a high FLA level and that approximately one-third of students presented a moderate FLA level. The situations experienced by Confucian Heritage Cultures (CHC) students were chosen as the focus of our research because they have higher FLA rates of all ethnic groups. This work aims to shed light on English Speaking Anxiety and observe the tendencies to enable future research in this field. The research is based on a series of papers collected from journals. The findings are related to the influence of emotional factors that contribute to anxiety in the classroom and to teachers’ attitudes. A high anxiety rate may be related not only to a natural propensity to anxiety, but also to factors related to teachers’ activities and attributes. The studies showed that the tone of voice, gender, and the teacher’s dress code affected levels of anxiety experienced by students. Allan Cordeiro da Silveira

For approximately thirty years, scientists have been interested in the causes and consequences of Foreign Language Anxiety (FLA) in Second Language Acquisition (SLA). Horwitz, Horwitz and Cope (1986) found that Speaking was the skill that contributed most to a high FLA level and that approximately one-third of students presented a moderate FLA level (HORWITZ, 2001).
Confucian Heritage Cultures (CHC) students' experiences were chosen as the focus of our research since they have higher FLA rates of all ethnic groups according to Woodrow (2006). CHC students come from countries such as Vietnam, Singapore, Korea, Japan and China, these students are considered submissive and reluctant to speak in class (TRAN, 2013). These countries share a collectivist mindset which leads them to work more efficiently in groups. This characteristic might not necessarily be positive if methods based on Western teaching approaches and concepts are adopted by coursebooks and applied in tasks. According to Phuong-Mai et al. (2005) not all cooperative learning is successful in CHC countries and these approaches might result in learning failure and they might cause some frustration in students.
The choice of subject is justified not only by the FLA rates, but also by the importance of these students in global terms. According to the UK Council for International Student Affairs -UKCISA (2019), the number of Chinese students is higher than any other nationality at 106,530 and more first-year students from China have enrolled at British universities than those first-year students from the whole of the EU combined (GILLMAN, 2015), while there were 16,350 and 7,020 students from Hong Kong andSingapore, respectively, in 2017-18 (UKCISA, 2019).
This paper aims to shed light on English Speaking Anxiety and observe the tendencies to enable future research in this field. The research is based on a series of papers collected from journals. Firstly, some essential concepts regarding the Interaction Hypothesis and Input will be presented briefly. Secondly, three main current research directions (social, pedagogical and cognitive) mentioned by Mackey (2012) will be analysed and we will then examine the general theory of Foreign Language Anxiety. Thirdly, abstracts of the selected papers will be analysed in the light of their relevance to the current research trends in FLA for ten years. Finally, the conclusion is presented.

INTRODUCTION
The Interaction Hypothesis states that interaction facilitates Second Language Acquisition (SLA) since learners who engage in interactional 2. RELEVANT THEORY

INPUT AND INTERACTION IN SECOND LANGUAGE ACQUISITION
situations have more opportunities to communicate and develop their language skills by receiving comprehensible input. Long (1996) linked two approaches in SLA: Hatch's (1978) theory which postulated the role of conversation in the development of grammar and Krashen's (1985) Input Hypothesis, which affirmed that linguistic input must be (i+1), in other words, it has to be a level above the learner's knowledge, but still comprehensible (PICA, 2003).
The interactive processes that occur between interlocutors are the negotiation of meaning, feedback, input, and output (MACKEY; GOO, 2012;MYLES, 2014). Long (1996) believed that what makes the input comprehensible is the modified interaction or negotiation of meaning between interlocutors. This modified interaction consists of various modifications that interlocutors make in order for their input to be comprehensible in an interactive situation (MUHO; KURANI, 2014;GALLAWAY;RICHARDS, 1994;MACKEY, 2002;VARONIS, 1994;PICA, 1994;OLIVER, 1995). Interlocutors must interpret and comprehend the signals in the interaction to negotiate meaning. The aim of this process is to maintain the conversation and to solve problems related to possible breaks in the dialogue (COOK, 2013).
Feedback (recasts, comprehension checks, clarification requests, selfrepetition or paraphrase, restatement and expansion of NNS statements, and topic switches) has often been an explored topic in the interaction approach (ALJAAFREH; LANTOLF, 1994;LIGHTBOWN;SPADA;RAND, 1993;LONG;INAGAKI;LYSTER;RANTA, 1997;ORTEGA, 1998;WESCHE, 1994). However, given that it does not constitute the focus of this paper, we will not analyse aspects of feedback in detail.
Another concept is input, i.e., the way in which the learner receives information. Such input should be provided at the level of "i + 1" so that new information is added. The Acquisition is attained via comprehensible input which is the new meaning that is gained by the learners. Students would not be able to obtain this new meaning by themselves (SCHÜTZ, 2007). Pivotal points of interaction are considered to be comprehensive input and the negotiation of meaning, which are prerequisites for acquisition.
Second language acquisition, as mentioned in the Introduction, is regarded as a field with more than 100 empirical studies and five analyses examining the validity of the interaction approach (MACKEY, 2012). At present, the implications for teaching have received increasing attention and are one of the trends in future research in this field. External factors contributing to learning have also been addressed, as well as the psycholinguistic processes involved in the construction of learning systems. Some of these processes will be discussed in the next sections, in addition to some other trends.
Researchers have currently moved beyond the initial studies in SLA, which implies interdisciplinarity with other sciences such as Computer Science, Neuroscience, Pedagogy, and Psychology. Mackey (2012) stated that the latest works include research on the social, pedagogical, and cognitive levels.
Concerning the social field, Mackey (2012) states that researchers have developed an elaborate understanding of factors related to the contexts and interpersonal relationships that can interfere with the process of receiving feedback or producing output. "This line of research means recognizing the role that social factors play in influencing cognitive processes.", according to Mackey (2012, p. 136). Mackey, Abbuhl, and Gass (2011) recommended that some replication studies need to be done, but explained that they are rare in this field. The authors suggested that expansion of this area might occur in the direction of cognitive aspects of interaction to include more socio-cognitive characteristics; they cited other authors, such as Bayley and Tarone (2011), andTarone (2009), as scientists who have been studying this line of research. Other trends are the impact of interaction techniques on students' awareness of structures in the foreign language, and the value of interactional activities that promote fluency and accuracy in the L2. Mackey (2012) cited studies about gestures to illustrate these tendencies (BIALYSTOK, 1990;FAERCH;GULLBERG, 1998;KASPER, 1983;KELLERMAN, 1992;KENDON, 2004;SUEYOSHI;HARDISON, 2005).
Thus far, pedagogical applications have not been treated as the central goal of interaction studies. However, many works are focused on task-based and form-focused instructions, and on the effects that different tasks have in terms of negotiating meaning and language production. Such works focus on the relationships among interlocutors and the characteristics of the tasks; there is a tendency to adopt a more socially informed perspective (PHILP, 2009). Another direction for future research is content-based language instruction in which a subject is taught using the learner's target language (LYSTER, 2011). Another topic that is being researched is related to teacher training and focuses on teachers' language awareness and the relevance of reflective teaching. This might allow professionals to offer effective task-based lessons and directions to maximise the effectiveness of focus-on-form techniques (TSUI, 2011).
The cognitive aspects of the current research will be presented. Mackey (2012) affirms that they have become the main focus of investigation especially concerning working memory and the role of attention, as well as the interrelationship between learner-internal cognitive processes and second language learning.
Frequency has been of particular importance in the development of theories related to cognition and interaction. Gass and Mackey (2002)

THREE MAIN CURRENT DIRECTIONS IN SECOND LANGUAGE ACQUISITION
stated that negotiation of meaning requires constructions that occur via repetition. They believe that further research would aid scientists to restructure processes that are typically conceptualised as changes without continuity in the interlanguage that occur in relatively discrete stages.
According to Mackey (2012, p. 140), another research trend in this field is refining the methods used, which are stimulated recall and think-aloud protocols. She suggests the use of split-screen technology to perform close-up images of learner's faces and provide a powerful stimulus. More precise recall and neuro-imaging techniques can also be relevant, such as electroencephalography (EEG), functional magnetic resonance imaging (fMRI) -(Morgan-Short and Ullman, 2013).
Another technique is related to heat maps to explore learner's responses (SMITH, 2012). These maps can be of help to notice recasts or issues related to the impact of individual variation in working memory capacity on noticing and processing interactional feedback (MACKEY; ABBUHL; GASS, 2012, p. 16).
Another topic that has been studied is the affective factors in SLA (DEWAELE, 2007(DEWAELE, , 2008(DEWAELE, , 2010PETRIDES;FURNHAM, 2008;DÖRNYEI;TSENG, 2009). Motivation and attitude can be considered determining facts in SLA (DÖRNYEI, 1994;LAMBERT, 1972) since they are closely related. Students' attitudes towards a language depend on how motivated they feel while another factor is language anxiety, which will be addressed in more detail in the following section.

FOREIGN LANGUAGE ANXIETY
Foreign language anxiety is a type of anxiety that is triggered in situations in which the speakers are exposed to foreign language use and learning (HORWITZ AND YOUNG, 1991). MacIntyre and Gardner (1994) agreed with this definition, affirming that this type of anxiety is related to feelings of tension and may be related to students' beliefs, natural tendency to experience anxiety, the classroom atmosphere, assessments, and the interaction between teachers and students (YOUNG, 1991;HORWITZ, COPE;HORWITZ, 1986). Negative results and demotivating assessments regarding their efforts in the classroom increase students' anxiety levels (MACINTYRE; GARDNER, 1991). A teacher's aim should be to make students more comfortable, open, and free to communicate in the target language. This idea is in line with Krashen's (1981) theory of the affective filter which claims that highly motivated learners are more likely to be successful in learning a foreign language. This theory also states that students with low motivation, low self-esteem, and debilitating anxiety can block their minds from receiving comprehensible input, thus, hindering the Acquisition.

THE ANALYSES
Relevant papers were selected via combinations of keywords to refine this research and two of these papers will be analysed in more depth. These works were chosen due to the similarity of issues. The keywords were: anxiety, foreign language anxiety, Confucian Heritage Cultures (CHC), Chinese, Japanese, and Korean.
In her paper, Woodrow (2006) reported that there are indications that students from Confucian Heritage Cultures (CHC) such as China, Japan, and Korea have higher levels of anxiety than do other ethnic groups. Woodrow (2006) commenting on the theoretical assumptions made by Watkins and Biggs (1996;2001) reported that the majority of the research is situated in a Western context, but that some studies comparing Western students and CHC are available. Woodrow (2006) also stated that a relationship between Chinese cultural issues (values of 'face' and 'silence') has contributed to the high anxiety level.
Similarly, Liu and Littlewood (1997) and Jones (1999) stated that the desire to be correct and perfect, and the fear of losing face, could contribute to this high anxiety level. Nevertheless, Tsui (1996) claimed that students' anxiety levels depend on the teachers' strategies, and that culture is not a factor. In agreement with this idea, Cheng (2000, p. 441) affirmed that, although cultural differences exist, it is "groundless and detrimental" to label Asian language students' reticence as being the result of cultural factors.
After these initial considerations, the first topic to be explored is the factors that contribute to speaking-in-class anxiety. Mak (2011) studied a group of 313 Chinese ESL first-year university students in Hong Kong. The scale used in the experiment was the Foreign Language Classroom Anxiety Scale (FLCAS) which was developed by Horwitz, Horwitz and Cope (1986) as a tool to measure anxiety levels revealed by negative performance expectancies and social comparisons, avoidance behaviours, and psychophysiological symptoms. FLCAS has been criticised on the bases of calculating students' anxiety levels instead of calculating students' language skills and excluding L1 skills (SPARK et al., 2009;TRAN, 2012). Nevertheless, Tran (2012) affirmed that FLCAS has increased its credibility by resolving problems related to inconsistency.
This study revealed five factors that are related to speaking-in-class anxiety: speech anxiety and fear of negative evaluation, discomfort when speaking to native speakers, negative attitudes towards English classes, negative self-evaluation, and fear of failing the class/consequences of personal failure.
Concerning the fear of negative evaluation, Aida (1994) stated that people who are constantly concerned about others evaluating them tend to avoid embarrassing situations, and have a passive attitude regarding potential challenges and interactions. Students who are corrected while speaking feel too inhibited to make new attempts and fluency is lost.
The author stated that not being able to use the L1 in the classroom is a factor in increased anxiety. In agreement with this, McMillan and Rivers (2011) affirmed that the L1 may be used in the classroom to make complex instructions easier and definitions clearer; however, this depends on the students' level of proficiency and preference. Advanced foreign language students do not expect to hear their L1 in the classroom. Nevertheless, an orthodox prohibition of L1 may cause negative effects on the expected fluency result particularly in more basic modules. Thompson and Lee's (2014) main theme was tolerance of ambiguity, which is an emerging theme in the field according to Dewaele and Wei (2013) (as cited in Thompson and Lee, 2014). They focused on Korean learners and how their experiences abroad affected their anxiety profiles. The research tools used were FLCAS and a questionnaire that collected information about their experiences abroad. The findings revealed that students' experiences abroad were closely related to the subfactors in their anxiety scores. This information confirms the fact that experiences in places, where the target language is spoken, seem to encourage students' self-confidence and, eventually, to decrease classroom anxiety (DEWAELE, 2013;MATSUDA;GOBEL, 2004). This is also in accordance with Dewaele and Shan Ip's (2013, p. 173) observation that "students who were more tolerant of second language ambiguity were less anxious in their English classes and they also felt more proficient." Language anxiety in oral examinations is another theme mentioned by Hewitt and Stephenson (2012). The FLCAS was used once more along with the analysis of variance, which was used to evaluate relationships between speaking performance and language anxiety levels. This study confirms Phillips' (1992) findings that affective reactions to oral components affect students' decisions to continue their courses at a more advanced level. It indicates that FLA depends on the attitudes that students have towards a foreign language. Students who have encountered difficult situations tend not to go further with their studies. Phillips (1992) advised teachers to treat FLA as a serious problem to play an active role in encouraging their students' progress in their SLA studies.
This theme is related to Shao, Yu and Ji's (2013) paper on the importance of emotional factors affecting SLA. They examined 510 Chinese students' emotional intelligence and anxiety in English classrooms at three universities in China. The tools used were FLCAS and the Trait Emotional Intelligence Questionnaire-Short Form (TEIQue-SF), which was developed by Petrides (2009). The TEIQue-SF is a 30-item questionnaire designed to measure the trait emotional intelligence (trait EI). Trait EI examines people's abilities to control and address their different emotions (COOPER; PETRIDES, 2010;PETRIDES;FURNHAM, 2006). The findings revealed that FLA has a mediating effect on students' emotional intelligence (EI), as well as on the relationship between self-rated English proficiency and EI. More than half of the students who had a medium or high level of EI had experienced FLA in the classroom, as reported by the authors (2013). Tianjian (2010) investigated the relationship of speaking anxiety (SA) and other variables such as unwillingness to communicate, language achievement, speaking self-efficacy, language class risk-taking, and language class sociability in Chinese students. The data were analysed using SPSS, which is a software programme for statistical data analysis, and Amos, which is a software programme used to design structural equation models (SEM).
The findings were that over 50% of the students presented moderate or high anxiety levels, that there were no differences in speaking anxiety (SA) levels based on gender. Although, there was a difference in language proficiency, that personality factors were the primary factors in SA and that SA does influence language achievement. Valls Ferrer (2010) affirmed that SLA research has been focused on the extraversion-introversion dichotomy with regard to the topic of personality factors; however, the results reflecting the relationship with learning a language successfully have been inconclusive.
Two papers will be presented as proposed in the paper in the next section.

TEACHERS' PERSPECTIVES AND TEACHING AND LEARNING ENGLISH IN JAPAN FROM STUDENTS ANALYSES
Two papers selected for analysis will be presented, it was decided to call them Paper A (Getting Them Speaking: Classroom Social Factors and Foreign Language Anxiety) by Dr. Okon Effiong and Paper B (Japanese elementary school teachers and English language anxiety) by Dr. Tomohisa Machida, both published in the TESOL Journal in 2016. In addition to the similarities with regard to the reputation of the journal and the year of publication, they are also closely related due to the fact that they focus on the teacher's perspective.
With regard to Paper A, Effiong (2016) explored causative agents in classrooms in order to observe the level of Foreign Language Anxiety (FLA) experienced by Japanese students at four universities. The research was guided by qualitative procedures for data collection (interviews) over a period of four months.
The survey questions were related to social factors (causative agents) that may influence the degree of FLA in the classroom. There was a focus on teachers' dress codes as this is important in Japan. Formality can be seen as a barrier to foreign language learning, an element that contributes to an increase in anxiety levels because it is considered to be intimidating. Effiong (2016) affirmed that teachers should dress more casually to decrease FLA.
Paper A explores the teacher's age factor, which is a factor to be debated. Older teachers tend to inspire more respect, thus raising students' anxiety levels. However, young teachers tend to lack the ability to discipline younger students or, may, convey insecurity to older students. The tone of voice was also discussed in Paper A, teachers with a more severe tone tended to increase students' FLA and decrease their participation in classes. Students felt too repressed to participate and were afraid to make mistakes and, consequently, did not speak in the classroom. Horwitz, Horwitz and Cope (1986) affirm that language use situations in the classroom are possible sources of anxiety. Subjects related to the classroom atmosphere, such as quietude, familiarity, and friendliness, were also mentioned in the paper. A quiet environment was deemed positive for skills such as listening but could be negative for Speaking. Teachers should have a friendly and open attitude towards their students and should use an appropriate tone of voice; the converse could increase students' anxiety levels. As an example, a work by Gregersen (2006) reported cultural differences between Chilean and Russian students that revealed how cultural issues and teachers' attitudes in the classroom are crucial concerning teaching. According to the author, Chilean students are members of a highly expressive and immediate Spanish-speaking culture while Russians represent a less-expressive, restrained northern European culture. Chilean teachers lead their classes in a group-oriented format, which implies that students participate more democratically, whereas Russian teachers maintain a formal relationship between the teacher and the students (teacher-centred) in which it is expected that students respect their teachers and ask permission to participate; moreover, students perceive any change in this hierarchical model as anarchy (SMITH, 2000).
Another factor to be considered is the tolerance of ambiguity as in Thompson and Lee's work (2104). Teachers and students should be flexible regarding their mistakes; it makes the learning process easier and results in decreasing the affective filter (KRASHEN, 1981). Students should be instructed to be tolerant as part of their linguistic education and should understand that linguistic inadequacies are part of the process of SLA; accordingly, and it is the teacher's role to provide the students with this information.
Effiong (2016) also discussed competition, which can be simultaneously good and harmful for FLA. For instance, it is productive if it aids the teacher to engender challenging situations in order to inspire students' participation. However, it can be extremely dangerous if the competitive spirit is lost and the competition itself gives way to student bullying and humiliating each other. This type of procedure should be well designed by teachers to avoid losing control of the situation.
In Paper B, Machida (2016) examined whether Japanese elementary education teachers had experienced FLA and the factors that contributed to this anxiety. The participants in the study were 133 teachers (71 males and 62 females) who were to teach in English despite the lack of preparation and insufficient language knowledge. The data collection was done via a questionnaire and the Teacher Foreign Language Anxiety Scale (TFLAS), which measures teachers' anxiety levels regarding their proficiency in a foreign language (HORWITZ, 2008).
The results revealed that a considerable number of teachers showed linguistic anxiety. The main sources of anxiety were the lack of confidence when communicating in English and the lack of preparation when teaching English. The fact that teachers had to team-teach with an English-speaking assistant certainly raised their anxiety levels, as any linguistic inadequacy could be a reason to expose them in front of the classroom. Consequently, they would feel embarrassed and would eventually lose their authority. Teachers and students must understand that linguistic inadequacies are part of foreign language learning, and the obsessive attempt to make a good impression when speaking may draw attention to their flaws instead of their positive points (MACINTYRE; GARDNER, 1989).
It is noticeable that the lack of confidence regarding communication skills stems mainly from perfectionism. Pishghadam and Akhondpoor (2011) stated that individuals who are perfectionists set goals that are too difficult to reach, do not allow themselves to make mistakes, and want to speak without any pronunciation deviation or grammatical inadequacy. They concluded that the anxiety levels of perfectionists may be a factor contributing to lower performance when speaking. Similarly, Gregersen and Horwitz (2002) claimed that perfectionists expose themselves to other people's bad judgment and focus too much on avoiding making linguistic mistakes rather than on communicating and learning from experience.
Another factor mentioned in Paper B is the lack of preparation when teaching English. Thus, governmental projects should be well planned by competent linguists as it is not possible to transform monolingual teachers with limited English-speaking skills into trained bilingual teachers to teach a subject in another language in a short period; it also requires significant investments in terms of funding and time investments. This preparation and training may be done via extra courses that could be offered to teachers, as well as by preparing them to sit international exams such as TOEFL, IELTS, and CPE, among others. Japan also has an added incentive in the form of the 2020 Olympics.

CONCLUSION
This paper aimed to conduct a literature survey of recent research on Foreign Language Anxiety and observe tendencies for future research based on Confucian Heritage Culture students and teachers.
In methodological terms, FLCAS was extensively used to measure students' anxiety and TFLAS to measure teachers' anxiety. Other tools, such as TEIQue-SF, SPSS and AMOS software, were also applied, despite the criticism related to its credibility regarding its inconsistency.
The findings in the papers are related via a network of relationships. The influence of emotional factors that contribute to anxiety in the classroom and teachers' attitudes are the central themes in the results. Students who experience FLA in oral examinations are afraid of receiving a negative evaluation from teachers and peers. A high anxiety rate may be related not only to a natural propensity to anxiety but also to factors related to teachers' activities and attributes.
The studies showed that the tone of voice, gender, and the teacher's dress code affected levels of anxiety experienced by students. Excessively formal clothing can inhibit and repress feelings of friendliness, thus hindering students' acquisition. On the other hand, the teachers who were interviewed reported a lack of knowledge and preparation when teaching in English. Having to communicate in the target language, combined with the constant search for linguistic perfection, can cause anxiety for teachers and eventually for students; teaching in the L2 requires a level of oral competence that teachers themselves lack. It is the teacher's role to promote the concept of tolerance of ambiguity by focusing on students' communicative aspects.
The consequences of poorly planned foreign language teaching are dramatic. In a globalised world, professionals who do not master foreign languages are not well placed in the labour market. Furthermore, it can potentially be disastrous to the economy of a developing country, as in the case of China, and it can hinder growth in a developed nation such as Japan.
This lack of linguistic proficiency from the population affects Brazil to the same extent because the majority of the population does not have access to the English language as a cultural asset. The upper-classes that have access to this asset and tend to have the best opportunities in the labour market.
To conclude, further research incorporating the same design would be of value if replicated in other developing countries, such as Brazil or Russia. It would be relevant to observe whether the same factors affect other cultural groups and to compare the levels of anxiety among various communities within the same country since these countries have continental dimensions and contrasting realities.