The Impact of L2 Motivational Self System and Students’ Perceptions of English Proficiency on Attitudes toward Translanguaging

This mixed method study investigated the relationship between L2 motivational self-system, students' perceptions of English proficiency, and attitudes toward translanguaging among Japanese high school students. Participants completed a questionnaire comprising three parts: background information, motivation factors, and attitudes toward language choice. Structural equation modeling (N = 213) and trajectory equifinality approach (N = 1) were used to analyze the data and chart the process of changes in attitudes toward translanguaging. The results showed that students preferred the use of translanguaging in English classes, and the ideal L2 self was a more important factor than the perceptions of English proficiency in predicting attitudes towards English. In addition, the study found that students' perceptions of English proficiency had a negative effect on the attitude toward English-only communication and a positive effect on the attitude toward translanguaging. Analysis, using the trajectory equifinality


Introduction
The notion that English classes must be taught solely in English, with limited support for students' first language (L1), has been a topic of debate in foreign language education for a considerable time. Consequently, the use of L1 in foreign language classrooms has been the focus of contentious debate and research for several years (Auerbach, 1993). When the grammar-translation method lost its popularity, the systematic use of L1 was discouraged, which led to an increase in research on monolingual approaches. Proponents of this approach assume that L1 has no important role to play in English as a foreign language (EFL) or English as a second language (ESL) classrooms and deprives students of exposure opportunities to target language input from students, thus, impeding progress (Bouangeune, 2009;Auerbach, Japanese to English usage by teachers in the EFL classes was 7:3. Contrary to a monolingual approach, which encourages only English, recent research on translanguaging takes a positive view of the increased use of students' L1 in L2 classrooms.

Translanguaging
Translanguaging has become widespread over the last decade to describe linguistic practices beyond the boundaries of a recognized language. Pedagogical approaches are based on linguistic practices (Cenoz & Durk, 2017;García & Li Wei, 2014).
The term 'translanguaging' (trawsieithu in Welsh) was proposed by Cen Williams. Languages are used in a classroom with planned intentions (Nagy, 2018). An important concept in translanguaging is that students should not use two languages in an uncontrolled manner. García (2009) stated that 'bilinguals translanguaging to include and facilitate communication with others and to construct deeper understandings and make sense of their bilingual worlds ' (p. 45). In addition, García (2009) stated that translanguaging encompassed 'multiple discursive practices in which bilinguals engage in order to make sense of their bilingual worlds' (p. 45). Learners' language choices are related to their ideal L2 selves (Lee & Lo, 2017). However, there has not yet been any research on the relationship among the L2 motivational self-system (L2MSS), students' perceptions of English proficiency, and their attitudes towards translation in the EFL context in Japan.

Complexity Theory
In this study, qualitative analysis, particularly from the perspective of a complexity lens, was used to explore what could not be gathered from statistics alone. When researchers conduct a study related to second language motivation, 'human behavior within a comprehensive theory of motivation will undoubtedly remain elusive, and we shall probably never be able to grasp the whole picture' (Dörnyei & Ushioda, 2013, p. 10). Dörnyei & Ushioda (2013) noted, 'It is clear that a linear approach cannot capture the dynamic and mutually constitutive nature of the relationship between motivation and context' (p. 77). They suggested that for L2 motivation research, a mixed method approach led to invaluable data and that this methodology allowed them to conduct research with a comprehensive view of the complex variables that exist in education and society. Therefore, quantitative and qualitative methods are essential for examining student motivation.
Previous studies of SLA research and foreign language teaching were conducted to identify causal relationships and assumed linear relationships. To reiterate, the theory of second language learning motivation is not yet fully understood. Hence, Dörnyei and Ushioda (2021) critically commented that when studies on motivation are conducted from the perspective of a linear model, limitations are encountered, particularly in cognitive, affective, and contextual variables.
A learners' characteristics, previous experiences, and the environment can have a significant impact. Therefore, it is important to include the entire set of variables around the learner rather than focusing on the body of the relationship with the hypothesized variables. In recent years, several researchers have focused on complex dynamic systems (Dörnyei, 2014), complex systems (Larsen-Freeman & Cameron, 2008), and dynamic system theory (Smith & Thelen, 1993). Larsen-Freeman and Cameron (2008) have conducted SLA research in recent years and attracted considerable attention. The principle of this study is simple. Instead of predicting things in advance, as is often the case in scientific research, they reverse the order of things and pursue predictions based on facts. The field of education is more complex than we believe, with everything intertwined into a single system. From this perspective, conducting research using a complex lens, may lead to more relevant research about the needs of the educational field.

Research Design and Participants
This study described the relationship among the L2MSS, students' perceptions of English proficiency, and the attitude towards translanguaging. Data for this study were extracted from a questionnaire survey administered to high school students enrolled in a single private high school in Japan. The study consisted of six variables -attitude toward translanguaging, attitude toward English-only, and the L2 motivational self-system (ideal L2 self, ought-to L2 self, L2 learning experience, and English proficiency). The relationships between these variables were investigated using SEM. Therefore, as a qualitative method, the TEA was used to chart the change in individual learners' attitudes towards the English-only class. The diagram was checked and analyzed along with the research collaborator, and the process of change in attitude toward the English-only class was detailed and analyzed. A total of 213 students from a private high school in Sapporo participated in the study. The participants were of mixed genders who were Japanese natives learning EFL. Their English classes were divided into two types: communication classes, which emphasized listening and speaking, and expression classes, which focused on writing and reading. The participants had an English proficiency level of approximately CEFR A1.

Instrument
The questionnaire comprised three parts. The first part elicited background information about the students, such as their gender, age and grade. The second part consisted of 12 items targeting the L2MSS factors. The purpose of the motivation portion of the questionnaire was to investigate the strength of motivated learning behaviour and variables which predict motivation, such as ideal L2 self, ought-to L2 self and L2 learning experience. These variables were adapted from Aubrey (2014) and Taguchi et al. (2009). The third part consisted of seven items, which targeted learners' attitudes towards language choice in an English classroom, such as classroom codeswitching and English-only education. These variables were adapted from Lee and Lo (2017). As mentioned earlier, codeswitching and translanguaging are similar linguistic phenomena in which a multilingual speaker naturally and comfortably moves back and forth between languages. Therefore, attitudes towards classroom codeswitching and translanguaging were considered as a single item in this study.

Trajectory Equifinality Approach (TEA)
Life course studies tend to use these correlations efficiently. There are more complex methods, such as SEM; however, their basic properties are the same, and SEM relies entirely on correlations (Sato et al., 2009). Despite this complicated model, in this study, the factors influencing the attitude towards translanguaging and English were fixed in advance. From the perspective of that study, it cannot be determined whether other factors influence attitudes toward language choice. A research method that addresses the limitations of this research is Dörnyei (2014), which introduces Retrodictive Qualitative Modelling (RQM) as a research method in the context of complexity theory. He stated that the idea behind the RQM was that by identifying the main emerging system prototypes, we could work 'backwards' and pinpoint the principal factors that have led to the specific 'settled states' (p. 85). A research method that can be used to incorporate these perspectives, the TEA belongs to the qualitative method category. There are three components of TEA -trajectory equifinality modeling (TEM), historically structured inviting (HSI), and a three-layer model of genesis (TLMG) (Sato et al., 2014). The TEM is a model for charting and capturing a path to a destination based on individual choices and decisions, along with the flow of time. To understand the trajectory of multiple decisions, the basic concepts of the TEM are illustrated in Figure 1 from Marsico & Valsiner (2018, p. 57). The HSI is a procedure in which researchers invite interested and concerned people to talk to them. The TLMG is a model for capturing the transformation and emergence of beliefs or values.  Bifurcating point to achieve EFP Social Direction (SD) The social power of inhibition that leads to EFP Social Guidance (SG) The social power of promotion to proceed to EFP

Data Collection and Procedure
Permission to conduct the study was obtained in writing from the high school before data collection (see Appendix A). The researcher conducted a pilot study prior to the main survey to check the data collection methods and detect problems which may compromise the data on a large scale. Following the results of the pilot study, all items in the questionnaire were written using Google Forms.
The pilot test was completed on 7 July 2021, and the data collection required for this study was completed on 20 August 2021. To collect the data, the link to the Google Form created by the researcher was shared with the classroom teachers of the three participating classes through Google Chat. The researcher connected with the students through Google Classroom. On the first day of data collection, the researcher visited participants' classes before completing the questionnaires. During this visit, participants were briefed on the content and length of the questionnaire. Additionally, the researcher informed the participants that they could choose to participate or withdraw from the study at any time; this would not have any negative impact on the students, and the data obtained would be managed appropriately to maintain their anonymity.

Data Analysis
This study examined students' perceptions of English proficiency, L2MSS, and their attitude toward classroom language choice. The collected data were analyzed using IBM SPSS Statistics version 26. The maximum likelihood method was used to estimate the parameters. The SEM had two parts -the measurement model and the structural model (Kunnan, 1998). The measurement phase involved confirmatory factor analysis, which aimed to test validity. Thus, factor analysis was conducted on all the items to examine the variable relationships, and 19 latent factors were identified. The results revealed five latent factors. Finally, the data were analyzed by SEM to answer the research question of this study.

Descriptive Analysis of the Survey
The minimum, maximum, mean, and standard deviation of each question and their skewness and kurtosis values are listed in scale. For the items of the L2MSS, 1 indicated 'strongly disagree' and 6 indicated 'strongly agree'. These descriptive statistics show that L2 learning experience had the highest mean in this study (M = 3.63) and ought-to L2 self had the lowest mean value (M = 2.64). For Japanese high school students, the ought-to L2 self is often high since English is required for university entrance exams. However, L2 learning motivation was the strongest component of the L2MSS. This indicates that the study participants were relatively motivated. Additionally, the mean value of the ideal L2 self was 2.81. Zentner and Renaud (2007) argued that the ideal self in adolescents is still developing, and it is difficult to set an ideal image, an argument that supports our results.
The results indicated that the attitude towards the translanguaging construct had a high mean value (M = 4.38). A few participants expressed negative attitudes toward translanguaging, whereas attitudes toward English-only communication had a low mean value (M = 2.62). When comparing the constructs, it is clear that the attitude towards translanguaging had a higher mean value than English-only communication.
To estimate the relationships among the L2MSS, the intercorrelations between English-only communication and the attitude towards translanguaging variables were examined. As seen in Table 3, for the L2MSS variables, the ideal L2 self had a moderate positive correlation with the L2 learning experience. Therefore, this correlates to a lesser degree with the ideal L2 self and ought-to L2 self. Furthermore, the ought-to L2 self and the L2 learning experience had a weak positive correlation. These results substantiate the statement that 'correlations among adjacent scales would be positive' (Noels et al., 2000, p. 71). One unexpected pattern of interest is the negative correlation between the ideal L2 self and students' perceptions of English proficiency (r = -0.407).

The SEM Analysis of the Quantitative Data
This section presents the results of the SEM analysis of the L2MSS, attitude towards translanguaging, and attitude towards English-only communication in high school students. Figure 2 presents the SEM model of the data and the final model with standardized path coefficients. It was significant that students' perceptions of English proficiency had a negative impact on the attitude towards English-only communication and a strong positive impact on the attitude towards translanguaging. Given the negative impact of the ideal L2 self on students' perceptions of English proficiency and attitudes towards translanguaging and English-only communication, it appears that their ideal L2 self mainly mediates the role of high school students' attitudes toward classroom language choice. Following the conventional procedures in SEM studies, chi-square statistics were calculated. The p-value should not be dominant in the chi-square statistics; however, since the sample was relatively large (N = 213) in this study, it was judged by other model fit indicators. In this study, five models were used -CFI, GFI, AGFI, TL1, and RMSEA. The SEM models were simultaneously fitted using a multigroup procedure. The results of the procedure indicated goodness of fit (Table 4).

TEM Analysis
The purpose of this analysis was to investigate the data qualitatively. An unexpected result is a positive correlation between students' perceptions of English proficiency and the attitude towards translanguaging and a negative impact on the attitude towards English-only communication. Prior to the analysis, the researcher hypothesized that participants who perceived their English proficiency to be low would be more likely to prefer translanguaging. However, the participants used both Japanese and English in their English classrooms. Quantitative research alone cannot capture the process of students' attitudes toward translanguaging in the classroom. This section describes the results of the process favouring translanguaging in English classes. The social power of inhibition leads to the use of Japanese and English in English classes Social Guidance (SG) The social power of promoting the use of English only in English classes

Participants, and SEM Analysis of the Quantitative Data
Amy was a senior high school student at a high school, which taught English. She was selected to clarify the patterns and processes of positive attitudes toward translanguaging in English classes. Participant information is presented in Table 6. To help Amy recall her experiences and identify the social factors, which influence her language use, the researcher explained the theme of the relationship between students' perceptions of English proficiency, the L2 motivational self-system, and the attitude towards translanguaging. Next, the researcher described the TEM in simple language. The researcher asked Amy to describe the theme and the TEM in an easy-to-understand manner. Once Amy understood and was comfortable with the themes and TEM, the researcher began the study in a closed classroom, with no other people present, to help her relax and speak freely. The researcher collected data twice during the interview process. The researcher interviewed Amy about her attitude toward translanguaging and how she had changed and then drew a TEM diagram to show her when and how social factors influenced her attitude towards translanguaging. The first and second interviews lasted approximately 60 and 30 minutes, respectively.

Summary of Findings (SEM)
To examine the relationships among L2 motivational self system variables, students' perceptions of English proficiency and their attitude towards classroom language choice and their intercorrelations were examined. In this study, students preferred translanguaging in their English classes, which corroborates previous findings in Japan (Aoyama, 2020;Turnbull, 2019). Students' preferences for translanguaging in English classes were not unique to the study participants. The results show that it is realistic to allow students to use L1 in English classrooms. In addition, the ideal L2 self only significantly contributed to the attitude towards English-only communication. Therefore, students' perceptions of English proficiency showed a negative effect on their attitude toward English-only communication and a positive effect on their attitudes toward translanguaging. Thus, this study concludes that the ideal L2 self is a more important factor than students' perceptions of English proficiency in predicting attitudes toward English-only communication. This result can be explained in terms of the 'cognitive' and 'affective' aspects of language learning motivation and attitudes toward classroom language choices (Lee & Lo, 2017). However, English proficiency may be external and less controllable by the learner since it represents a 'sense of achievement' or 'performance'. Therefore, proficiency may play a less important role, even though it may be related to the attitude toward classroom language choice. However, the ideal L2 self and attitude toward English showed weak associations. Lamb (2004) suggested that learners were motivated to learn by imagining themselves using English in the future in academic and business activities in the international community. However, in Japan, imagining an ideal L2 self is difficult. Therefore, English teachers must allow students to enhance their ideal L2 selves. In addition, this study indicates that if students' ideal L2 self could be increased, their attitude towards English-only classes would improve, and it would be more likely that English classes would be taught in English.
The results show that students' perceptions of English proficiency have a negative effect on their attitude toward English-only communication and a positive effect on their attitude toward translanguaging. This implies that their perceptions of English proficiency were not related to their output skills, such as speaking, but to their input skills, such as their knowledge of grammar or vocabulary. Consequently, it can be concluded that even though students' perceptions of English proficiency were high, they were not confident in their output skills, such as speaking, and thus showed a positive relationship with attitudes toward translanguaging. These findings show that the use of L1 in L2 classrooms serves necessary functions in communication. Students may use L1 intentionally and tactfully, even when they perceive and self-assess their speaking abilities as high. Finally, the current study found that students' perceptions of English proficiency had a stronger influence than their ideal L2 self as a variable explaining their attitudes toward language choice. As mentioned above, it can be difficult to imagine a situation in which English is used as a communication tool in the Japanese context. This is thought to be the result of the emphasis on English being considered a tool for passing exams. Nevertheless, it must not be forgotten that learner motivation is a multifaceted concept influenced by many variables. Therefore, many variables in motivational research cannot be examined by quantitative research alone, and it is necessary to use a variety of statistical methods to conduct research.

Summary of Findings (TEM)
Amy, in this study, was characterized by very high proficiency in English and very high motivation to learn it. The period under analysis was between her first and second years of high school, but there were two periods during which her motivation to study English dropped. However, a major factor in this move towards wanting to study English was that Amy found role models and wanted to be like other students and teachers. Then, with repeated advice (social guidance: their teacher recommended that Amy might be suited to study at a university abroad), Amy began to imagine herself studying in an English-speaking country. Consequently, Amy imagined herself speaking English abroad and started to see herself as a person who speaks English. This led Amy to discover her ideal L2 self through these transitions.
Her ideal L2 self increased, and her image of her ideal English speaker became clearer. Simultaneously, her English motivation and grades also improved. Her goals became clear, from the abstract goals of improving her English to wanting input and strengthening her English listening skills. A very interesting point in this study was that Amy felt negative about English classes being taught in English-only but did not say that English classes should be taught using Japanese and English; Amy has already recognized that classes in English-only are necessary to improve her English proficiency. Amy had very high English language skills and was highly motivated to study English; thus, the reason given for why English classes should be taught using English only was 'interaction with other students'.
The results of the SEM analysis showed that students' perceptions of English proficiency had a negative effect on their attitudes toward English-only and a positive effect on their attitudes toward translanguaging. The reasons for this relationship, identified in the quantitative SEM analysis, became slightly clearer in the TEM analysis, although they cannot be generalized. The reason for this is that even if Amy understands the English spoken by the teacher in an English class, she cannot interact with other students in the class who did not understand equally well. In other words, Amy was aware that her actual class would not be able to foster interactions with other students because there would be a gap between her and the students around her. Amy was more motivated to learn English through interactions with other students than in class. Therefore, in the case of Amy in this study, even though she had a high level of English and was very motivated to study English, she believed that English should only be taught using Japanese and English because she wanted the opportunity to interact with other students.

Pedagogical Implication of the Study
The results of this study have pedagogical implications for teaching English that differ from how English has been taught in the past. The results suggest that to make teaching in English is viable. It is crucial for learners to be able to imagine their ideal L2 self than to be aware that they are good at English. Therefore, it is important to create an environment in an English classroom where students can imagine their ideal L2 selves.
Dörnyei (2009) stated that a prerequisite for the ideal self to have a real and beneficial impact on language learning is that it is imagined and perceived with a sense of reality and that it is used in conjunction with appropriate planning and self-regulatory strategies. As mentioned earlier, in an environment such as Japanese society, where there are few opportunities to use English in everyday life, it is difficult to imagine using English in the future with a sense of reality. This is particularly difficult for learners who are not proficient in English. Yashima (2009) reported that the learning experience of active participation in a virtual international community through the teaching of English helps learners shape their future ideal selves. Murray (2011) argued that even without this high level of learning activity, Japanese English teachers who teach with a high level of English language proficiency could help students develop their ideal selves. In any case, there is no arguing that daily lessons play an important role for Japanese learners of English to develop their ideal selves. Furthermore, they also studied other subjects and were involved in various activities outside the classroom. For them, learning English is their way of life. Their motivation to learn English was not independent of their daily lives or the motivation to learn about other subjects. This is something that English teachers must be aware of. It is also necessary to take a relative view of English learning (Ushioda, 1998).
Another finding of the study, which was only a preliminary survey of participants of a particular age and group, was that high school student had positive attitudes toward translanguaging in English classrooms. As mentioned, this suggests that English should be used with scaffolding from Vygotsky's zone of proximal development zone of approximal development (ZPD) in English classes until students' ideal L2 self becomes higher and their attitudes towards English become more favorable. As mentioned previously, English teaching in Japan focuses on communication activities. Simultaneously, students are expected to participate in English only.
However, as the results of this study show, few students have positive attitudes toward English. Encouraging English classes to be conducted exclusively in English can be a major factor in making it impossible to establish communicative activities. Therefore, it is worthwhile to use students' L1 positively, such as through translanguaging.

Limitations of the Study
This study was conducted within several constraints, and caution must be exercised when discussing the generalisability of the results. First, in the present study, all the data were collected from a single private high school. Collecting data from multiple high schools may have provided different results because of possible differences in participants' social backgrounds. Furthermore, explaining and predicting students' attitudes toward classroom language choices can be influenced by multiple variables. Future studies should include additional variables to explore other possibilities. In the qualitative analysis stage, it is not feasible to draw generalizations, as the present study analyzed only a single individual.
Consequently, a larger sample size must be analyzed, and various transformation processes must be observed to form more accurate conclusions.
Lastly, Nguyen (2022) argues that the unsystematic use of translanguaging is common among EFL teachers and students, which is considered unfavorable. To address this, there is a need to standardize definitions since translanguaging is often perceived negatively.