The Influence of Prior Language Experience on Foreign Language Anxiety: A Study on a Russian-Speaking Sample

Full Text

Introduction Learning foreign languages in the modern world has become an urgent need. Knowledge of a foreign language is crucial to personal and professional growth: it broadens one’s horizons, creates numerous opportunities, and provides cognitive benefits. Succeeding in learning a foreign language appears to be vital for many learners, as it can contribute to career and personal achievements. Language attainment is the ability to repeat linguistic constructions taught and mastered (Prinsloo et al., 2018). Numerous factors influence language achievement: levels of intrinsic (e.g., Harter, Connell, 1984) and extrinsic motivation (e.g., Wen, 1997), learner’s perceived self-confidence (e.g., Clément et al., 1994; Gardner, 1985, 2001), and language learning attitudes (Gardner & Lambert, 1972). Foreign language anxiety (FLA) is one of the most important predictors of language achievement (Onwuegbuzie et al., 2000). At the same time, several studies have shown that the experience of learning a foreign language is often associated with a certain amount of stress and anxiety (e.g., Montgomery, Spalding, 2005; Toyama, Yamazaki, 2021). Foreign language anxiety influences foreign language achievement, emphasizing the relevance of this topic. Therefore, the investigation into the nature of FLA seems to be very important. Understanding the FLA factors will help teachers and methodologists organize language classes more effectively and without anxiety. Theoretical background Foreign language anxiety. FLA is “a form of situation-specific anxiety” (MacIntyre & Gardner, 1994, p. 88). Experiencing this kind of anxiety involves tension in the learning context (MacIntyre & Gardner, 1994). From the emotional perspective, FLA is associated with being upset, worried, and shy (Horwitz et al., 1986). FLA is also linked to physical discomfort, including dry throat and body shaking (Oxford, 2016). In addition, FLA has behavioural consequences, e.g., language learners tend to skip school, remain silent during the class, and avoid communicating in a foreign language (Young, 1991). There are several tools for measuring FLA in the language learning context: the Foreign Language Reading Anxiety Scale (Saito et al., 1999), the Foreign Language Writing Anxiety Scale (Cheng, 2004) and the Foreign Language Listening Anxiety Scale (Elkhafaifi, 2005). However, the most popular tool for measuring foreign language anxiety in the classroom is the Foreign Language Classroom Anxiety Scale (FLCAS) (Horwitz et al., 1986). There is evidence that FLA is associated with language achievement (e.g., Vogely, 1998; Onwuegbuzie et al., 2000), foreign language enjoyment (e.g., Dewaele, MacIntyre, 2014), and academic performance (e.g., Bailey, 1983). For example, Aida (1994) and Rodriguez (1995) found that foreign language anxiety was negatively related to course grades. This type of anxiety could affect learning motivation, interest, and perseverance in the context of foreign language learning (Montgomery, Spalding, 2005). Sources of foreign language anxiety. Factors influencing FLA are being investigated in an increasing amount of literature (review in Toyama, Yamazaki, 2021). These factors include the learner’s academic performance, experience abroad, high school language course experience, expected grade in the current language course, perceived school competence and self-esteem, age, language proficiency, and prior foreign language experience (Onwuegbuzie et al., 2000). Further, there are conflicting data about the role of age in FLA. For example, Dewaele (2007) found the age of students to be a significant predictor of their language anxiety. Older students tended to express greater language anxiety. These findings, however, were only significant for interactions with strangers. In contrast, another study found that older students were less anxious when they had to communicate in a foreign language (Dewaele et al., 2008). Moreover, some researchers reported no relationship between students’ age and anxiety associated with learning a foreign language (Onwuegbuzie et al., 2000; Saito, Samimy, 1996). Native and foreign language proficiency is also a significant predictor of FLA. Several studies have shown that mother tongue proficiency can affect the levels of foreign language anxiety (Ganschow & Sparks, 1996; Sparks et al., 1998). Some other studies have shown that participants with lower levels of foreign language proficiency scored higher on FLA (e.g., Liu, 2006; Marcos-Llinás, Garau, 2009). For example, Erzhanova, Kharkhurin, and Koncha (2022) found that foreign language learners with higher proficiency in all their languages had lower scores on FLA. Another factor that appears to be central to the present study is prior language learning experience. Knowledge of more than one foreign language makes it possible to investigate the interrelationships of these languages (e.g., Westergaard et al., 2017; Arıbaş, Cele, 2021; Bardel, Falk, 2021). According to Westergaard (2021), both the first language (L1) and the second language (L2) have an impact on the target language and might either promote or hinder third language (L3) acquisition. There is evidence that linguistic similarity between L1 or L2 and target L3 affects L3 acquisition (Li & Shirai, 2000). For example, Cenoz (2001) argued that a positive transfer from L2 was the most significant source of influence on L3. The results obtained by Chin (2009) and Bayona (2009) also suggest that linguistic similarity plays an important role, but only between L2 and L3. In the production of L3, cross-language interference mostly comes from a typologically closer language. L2 is expected to predominate during cross-language transfer into L3 if L1 and L2 are relatively close to L3 (e.g., Cenoz, 2001, 2003; Hammarberg, 2001). Moreover, speakers may unintentionally use the same methods they used to learn their first non-native language (L2) when learning their second non-native language (L3). These findings converge on the idea that L2 may become the primary source of transfer during L3 processing (Hammarberg, 2001). Thus, it is possible that prior experience of learning a similar language could be an advantage for further language acquisition. On the other hand, some researchers reported that prior multilingual experience had either negative or no consequences for L3 acquisition (e.g., Wagner et al., 1989; Gibson et al., 2001; Okita, Guo, 2001). Some linguistic characteristics may be the basis for cross-language similarity. These characteristics include phonology, morphology, semantics, syntax, and pragmatics (Bloom, Lahey, 1978). The first set of characteristics is related to phonology. For example, cross-language similarity predicted the ease of learning L2 phonemes (Sturman et al., 2016). The sounds common to the languages were recognized and used in learning. Understanding the common and uncommon sounds in native and foreign languages can be helpful for language learners. The second set of characteristics is related to morphology. For example, Spanish and English have many similar words. Reading accuracy, fluency, and comprehension can benefit from a clear understanding of word components and morphemes (Ramírez, 2017). The third set of characteristics is related to semantics. Semantics involves understanding the meaning of words. Learners’ understanding of the specific word meanings affects their ability to understand both spoken and written language. Successful comprehension is based on understanding the meanings of individual words (Koda, 2007). The final set of characteristics is related to syntax. Grammatical features can be compared across languages.[3] Nouns, verbs, and adjectives are present in all languages. The order and grammatically proper usage are subject to change. Understanding how languages are used in social and academic contexts is one of the pragmatic linguistic aspects. Studies on foreign language classroom anxiety. Numerous studies have analyzed foreign language classroom anxiety (FLCA) among native speakers of different languages (e.g., Amengual-Pizzaro, 2018; Darmi, Albion, 2014; Bergström, 2017), in which each component of FLA has been studied separately. For example, Amengual-Pizarro (2018) found that communication apprehension was the primary source of FLA among Spanish speakers. Darmi and Albion (2014), on the other hand, suggested that fear of negative evaluation was the most common factor in FLA among Malaysian students. Several studies have been focused on the problems of FLA among Russian speakers (Eremeeva, 2012; Sanakoeva, 2022; Shurygina et al., 2022; Antropova et al., 2022). For example, Iksanova, Krasilnikova and Vorobieva (2021) studied language anxiety among Russian learners of English. The results of their study revealed a greater degree of communication apprehension and fear of negative evaluation in male than in female respondents. Andryushkina (2019) highlighted that communication apprehension scored the highest in a group of Russian learners of French and Spanish. At the same time, their peers learning German showed the lowest level of communication apprehension. A similar pattern was found for fear of negative evaluation. Interestingly, Andryushkina reported that test anxiety levels were higher among learners of Romance languages than among learners of Chinese. Unfortunately, it must be recognized that very few studies have been conducted on Russian-speaking samples. For this reason, there are no empirical data on which factors and to what extent FLA develops in Russian speakers. At the same time, some factors make it extremely important to study FLA in a culture-specific context. For example, Ter-Minasova (2005) suggested that the Russian education system is oriented towards perfectionism and theoretical approaches and does not pay due attention to communication skills. Hubenthal (2004) studied Russian-speaking immigrants in the USA and concluded that they were afraid of making mistakes when talking to their peers. She also suggested that this was to the fact that in Soviet schools the main emphasis was placed on grammar rules. Present study. The above literature analysis showed that prior language experience could impact FLA. Moreover, the foreign language acquisition age and proficiency level could also influence FLA levels. In the present study, we tested whether the similarity of prior language experiences would affect the level of anxiety in the process of learning a foreign language and whether this relationship could be affected by the language acquisition age and proficiency level. To this end, we formulated the following hypothesis: “prior language learning experience influences FLA”. Under this hypothesis, we expected to find that the participants who were learning a foreign language similar to the previously learned one would show lower levels of FLA. We also expected that the acquisition age and proficiency level of the known language would change the influence of prior language experience on FLA. Methods Participants. The study involved 152 respondents (21 males and 131 females) aged 16-45 (M = 25.54, SD = 7.75). They were recruited through social media. The online survey was conducted using a reliable survey platform.[4] We posted general information about the research on these social media and invited the audience to participate via the link. These messages were posted in social media groups for people learning a foreign language. The respondents were informed that participation was voluntary and charge-free. The respondents were Russian speakers from 23 countries: most of them were from Russia (96). There were also participants from Kazakhstan (10), Germany (7), Turkey (6), Italy (6), Egypt (3), Japan (3), etc. Since the survey was conducted in Russian, we made sure that all the participants were fluent in this language: their self-reported proficiency in Russian was M = 4.86 out of 5 (SD = .41); see language proficiency assessments below). In addition to Russian, the participants reported proficiency in two (93.42%), three (51.31%), and four or more (17.7%) foreign languages. In addition, they reported L2 proficiency (M = 4.07, SD = 1.24), L3 proficiency (M = 3.00, SD = 1.24), L4 proficiency (M = 2.69, SD = 1.08), and L5 proficiency (M = 3.01, SD = 1.15). Procedure. The data were collected online on a reliable survey platform.[5] The survey took approximately 20 minutes to complete. The questionnaire included socio-demographic questions, a language learning experience questionnaire, and the Foreign Language Classroom Anxiety Scale (FLCAS). Before the survey, the informed consent of the participants was obtained according to the form approved by the HSE University Ethics Committee. The respondents were aware that participation was voluntary and charge-free. Instruments. Foreign Language Anxiety assessment. The Foreign Language Classroom Anxiety Scale, FLCAS (Horwitz et al., 1986) determined the participants’ levels of FLA. The questionnaire included 33 questions on a 5-point Likert scale, grouped into three categories: test anxiety (e.g., “I am usually at ease during tests in my class”), communication apprehension (e.g., “I get nervous when I do not understand every word the teacher says”), and fear of negative evaluation (e.g., “I feel confident when I speak in the class”). The scores ranged from 16 to 68 for fear of negative evaluation, from 14 to 68 for communication apprehension, and from 3 to 15 for test anxiety. The higher scores indicated a higher level of anxiety on each scale. The FLCAS is the most accurate and valid tool for evaluating FLA (Aida, 1994). The FLCAS showed high internal consistency (α = .93; Horwitz, 1986). We used the Russian version of the survey adapted by Kalganova and Mardanshina (2015). Language learning experience. The respondents also answered questions regarding their language learning experiences: about the number of acquired languages, the age of their acquisition, and the level of proficiency in these languages. The latter was assessed on a 5-point Likert scale, on which participants reported their speaking, listening, writing, and reading skills. For each language, the respondents were asked the following question, “On a scale from 1 (poor) to 5 (fluent), how do you assess your level of proficiency in speaking, listening, reading, and writing in these languages?”. Proficiency in each language was assessed as the average of the four language skills. A similar approach was used by Dewaele and Wei (2013) and Dewaele and van Oudenhoven (2009) for assessing the linguistic abilities of the respondents. In addition, the participants indicated which language they were learning at the time of the survey. Based on this information, we divided the participants into two groups (a family-based binary division): one group learned a foreign language similar to the one previously learned; the other group learned a foreign language different from the one previously known. We identified similarities/differences based on the language families of the respondents’ listed languages. The respondents were learners of languages from the following language families: Turkic (8), Sino-Tibetan (15), Japonic (15), Semitic (16), Altaic (22), and Indo-European (76). The languages representing these groups were Turkish, Chinese, Arabic, Korean, and English. For example, if a respondent reported learning German and having learned English before, he/she was placed in the group with a similar prior language experience. But a respondent who reported learning Korean and having learned English before was placed in the group with a different prior language experience. Results Table 1 presents descriptive statistics (gender, age, number of languages, language proficiency, FLA components) for the entire sample as well as for the groups with the same and different prior language experiences. Descriptive statistics for the groups with the same and different prior language experiences (N = 152) Variable Total sample Similar prior language experience Different prior language experience N 152 67 85 Male 21 8 13 Female 131 59 72 Mean SD Mean SD Mean SD Age 25.54 7.75 25.46 7.84 25.60 7.72 Number of languages 3.75 .80 3.88 .83 3.65 .78 Language proficiency 2.79 1.03 2.71 1.12 2.86 .96 Fear of negative evaluation 39.23 11.45 37.24 10.80 40.80 11.77 Communication apprehension 38.69 13.17 37.16 13.30 39.89 13.02 Test anxiety 8.04 2.79 7.99 2.85 8.08 2.76 The table shows that the participants of the study were predominantly females (86.18%). No significant gender differences were found in the studied variables. The mean age for both groups was comparable (for the group with a similar language experience: M = 25.45, SD = 7.84; for the group with a different language experience: M = 25.60, SD = 7.72). There were also no significant differences between the groups in terms of the number of languages (for the group with a similar language experience: M = 3.88, SD = .83; for the group with a different language experience: M = 3.65, SD = .78) and language proficiency (for the group with a similar language experience: M = 2.71, SD = 1.12; for the group with a different language experience: M = 2.86, SD = .96). To test our hypothesis, we performed a MANCOVA using the groups with similar and different prior language experiences as independent variables, the three components of FLA (test anxiety, communication apprehension, and fear of negative evaluation) as dependent variables, and the age of acquisition and proficiency in the language being learned as covariates. Before proceeding to the analysis, Box’s Test for Equality of Covariance Matrices and Levene’s Test for Equality of Error Variances were performed. The results showed that the model satisfied all the assumptions of normality, homogeneity of variance-covariance matrices, linearity, and multicollinearity (Box’s M = 6.45; F(6, 140763.535) = 1.051, p = .39). Since the p-value is greater than .05, it allows us to assume the covariance is homogeneous. As for Levene’s Test for Equality of Error Variances, all the p-values for the three dependent variables are greater than .05 (p = .570, .794, .483, for fear of negative evaluation, communication apprehension, and test anxiety, respectively). This suggests the homogeneity of variance. Since both tests were passed successfully, we were able to continue with MANCOVA. Figure 1. Differences between the groups with prior language experience regarding Fear of Negative Evaluation Figure 2. Differences between the groups with prior language experience regarding Communication Apprehension Figure 3. Differences between the groups with prior language experience regarding Test Anxiety First, we found a main effect for the proficiency in the language being learned (Pillai’s trace: F(3, 146)=4.001, p < .05, η2 = .076), but not for the age of recent language acquisition. Second, we found no main effect for the groups with prior language experiences (Pillai’s trace: F(3, 146) = 2.128, p = .099, η2 = .042). However, we found a significant difference between these groups regarding fear of negative evaluation (F(1, 148) = 4.368, p < .05, η2 = .029). In particular, the group with similar language experience scored significantly lower on this FLA component than the group with different language experience (ΔM = 3.56, SE = 1.79). No other significant differences were found. Figures 1-3 illustrates our findings. Discussion The current study is aimed at determining the role of prior language learning experience in FLA. The results of the study showed that when people learn a language similar to the one they knew before, their fear of negative evaluation becomes lower. According to Ramírez (2017), it is easier to achieve reading accuracy and fluency in a foreign language if there is prior experience of learning a similar language. We hypothesize that knowing a similar foreign language helps learners become more communicable and less anxious about making mistakes or being misunderstood. These findings are consistent with previous studies (e.g., Nayak et al., 1990), which suggested that prior experience of learning a similar language facilitates learning a new language. For example, Yang, Chen, and Xiao (2022) reported that mastering the phonological contrast of a foreign language could be difficult in case of its absence in the native language. It means that the role of “cross-linguistic” similarity in mastering a foreign language is of great importance. Learning a language, the sounds of which are identical to those of the previously learned language creates fertile ground for language transfer. If the language being learned differs from the previously learned one, the success of its acquisition will depend on the learner’s ability to distinguish between these phonetic differences (Flege, 1987, 1988). We also found that the effect of the previously learned language is influenced by the degree of proficiency in the currently known language. Prior language experience creates an advantage for those who are learning a similar language. These findings are consistent with previous research. For example, according to Cummins (1979), the benefits of being bilingual can only become apparent once a basic level of proficiency in both languages has been achieved. Similarly, in a study on bilingual adults with high, intermediate, and low proficiency, Xie (2018) found that those who were more proficient in both languages performed better on cognitive control. At the same time, it is also suggested that cognitive control training may alleviate anxiety symptoms (Edwards et al., 2022). This may explain why more advanced levels of language proficiency affect the fear of negative evaluation. There are some recommendations for further studies. First, it is advisable to explore in more detail the differences in language anxiety depending on the particular language. Second, it must be taken into account that the cultural background can potentially be a significant factor in FLA. Djafri and Wimbarti (2018) found that learners of Japanese experienced the highest FLA level, followed by those who learned Arabic, French, Korean, and English. In addition, FLA levels differ between learners of non-Latin languages (Japanese, Korean, and Arabic) and Latin languages (English and French), with the non-Latin language learners having a higher level of anxiety. According to Toyama and Yamazuki (2022), there is a significant correlation between FLA and the collectivist and individualistic cultures. Lim’s study (2009) also supports this idea, suggesting a relationship between the geographic areas of learners and FLA levels. Conclusion This study had some limitations. First, the sample was relatively small. One hundred and fifty-two participants may not be enough to draw conclusions, as this increases the risk of false positive results. We intend to recruit a more significant sample to test the hypothesis in further studies. Second, the sample is gender-imbalanced (with 85.71% female respondents). Although there were no significant gender differences in our research variables, this imbalance might have made it impossible to generalize the results. Future studies on the topic should focus on creating a more gender-balanced sample with an equal number of male and female participants. Summarizing our findings, we can say that prior experience of learning a foreign language similar to the one previously learned has advantages in what concerns fear of negative evaluation, but only for those who have mastered this language to a higher degree. In other words, prior experience of learning a similar language affects FLA at a certain level of language proficiency. Today, the topic of FLA is of great importance. According to Philipps (1992), language anxiety in the classroom can lead to adverse affective responses, unfavourable attitudes, and a lack of interest in language learning. This inevitably leads to difficulties learning a foreign language and loss of motivation. Examining the factors that influence the levels of FLA might help teachers and learners reduce it. Teachers should be aware of the sources of FLA and, based on that knowledge, try to create an anxiety-free environment (MacIntyre, Noels, 1996). The fact that linguistic similarity affects fear of negative evaluation may make teachers think about putting more effort in creating an inclusive and open environment for students without such background. First, based on the results, the teacher should be aware of the student’s prior language experience. We suggest asking students what language they speak before starting the learning process. Understanding students’ prior language experience will help teachers plan their lessons. Second, one of the primary steps in teaching of a foreign language should be teaching its phonology. Teachers can start classes by explaining to students the similarities between the sounds of the target and previously known languages to make it easier for them to learn a new foreign language.[6] Third, it can also be useful to highlight common morphemes in the languages spoken by students. The morphological overlap between languages may influence the correlation between language skills (Pasquarella et al., 2014). To sum up, making these cross-linguistic links will improve language learners’ understanding of words in a foreign language. Thus, studies on language anxiety will significantly affect the teaching and learning of foreign languages.

About the authors

Asiia Erzhanova

HSE University

Author for correspondence.
Email: aerzhanova@hse.ru
ORCID iD: 0000-0003-3871-3818

MA in Psychology, PhD Student, Department of Psychology

20 Myasnitskaya St, Moscow, 101000, Russian Federation

Anatoliy V. Kharkhurin

HSE University

Email: akharkhurin@hse.ru
ORCID iD: 0000-0001-5970-5003

PhD in Psychology, Chief Research Fellow, Department of Psychology

20 Myasnitskaya St, Moscow, 101000, Russian Federation

References