Introduction

Over the last couple of decades, learning the Korean language has become increasingly popular around the world, due to the global popularity of Korean pop culture (Jee, 2015). According to the Korea Foundation, the number of universities teaching the Korean language worldwide has more than tripled in the last 10 years. By June 2021, the King Sejong Institute Foundation, a public institution supported by the Korean government, had set up more than 230 Korean language institutes in 82 countries. The great enthusiasm for Korean language learning may seem inconsistent with the global status of the Korean language, which is only the 20th most spoken language in the world due to its limited regions of use and relatively small number of speakers (Eberhard et al., 2021). This raises the question of what motivates learners to learn the Korean language despite low social demand for the language.

Second language (L2) learning motivation is a significant predictor of the process and outcomes of L2 acquisition (de Burgh-Hirabe, 2019). The current dominant framework in the language motivation field is the L2 Motivational Self System (L2MSS) model proposed by Dörnyei in 2005, because it is versatile and can accommodate a variety of perspectives from different theoretical orientations (Boo et al., 2015). Empirical studies conducted in various contexts have validated the strong explanatory power of the model for L2 motivation (e.g., Huang, 2019; You & Dörnyei, 2016). However, the majority of these studies have focused on English language learning motivation; learning motivation pertaining to languages other than English (LOTEs), such as the Korean language, has received very little attention (Dörnyei & Al-Hoorie, 2017). In research focusing on German, French and Spanish as target languages, LOTE learners have been found to differ from English language learners in their motivational characteristics and motivated behaviours (e.g., Chen et al., 2021; Huang, 2019). More evidence of such differences is needed in relation to Korean language learners.

In China, a multi-ethnic nation, the population of the Korean ethnic group, as released by China’s National Bureau of Statistics, has reached nearly 2 million by 2020. Meanwhile, Korean has emerged as an important less commonly taught language, with a large number of students choosing it as a university major. Within the L2MSS framework, our study investigated the motivational characteristics of Chinese students learning the Korean language and explored the influence of the different components of the L2MSS on their motivated behaviours. This study helps to enhance the understandings of motivation to learn LOTEs and refine the L2 motivation theory, which has been disproportionately shaped by English language learners over the last decade. In addition, given that the majority of previous research on LOTE learning motivation has been conducted in Europe or the United States (Mendoza & Phung, 2019), the study could shed more light on the LOTE learning dynamics in Asian countries.

Literature Review

L2 Motivational Self System

In the field of L2 acquisition, learning motivation has attracted great interest from researchers. Over the last 60 years, various theories have been used to understand the construct of motivation (Boo et al., 2015). In 2005, Dörnyei proposed the L2MSS theory, which has since become the most commonly used model in motivational studies on L2 learning (Boo et al., 2015). The L2MSS has three component motivators, namely the ideal L2 self, the ought-to L2 self, and the L2 learning experience (Dörnyei, 2005, 2019). Specifically, the ideal L2 self reflects learners’ desire to become proficient L2 users in the future and therefore represents learners’ aspirations and hopes. The ought-to L2 self reflects learners’ sense of responsibility and obligation to learn the L2 to meet the expectations of others and avoid possible undesirable consequences. Centred on the current experience, the L2 learning experience involves several ‘executive’ and situated motives that are associated with learners’ immediate environment in their learning process.

The applicability of the L2MSS model has been examined in various L2 learning contexts, most notably English as a foreign or second language contexts, while LOTE contexts have been relatively unexplored (Boo et al., 2015). However, as English is the global lingua franca, the motivation for learning English might be quite distinct from the motivation for learning LOTEs (Dörnyei & Al-Hoorie, 2017). Research on LOTE learning motivation has consistently indicated the important role of the ideal L2 self in predicting motivated behaviours (e.g., Huang et al., 2015; Kong et al., 2018). Meanwhile, LOTE learners have reported, in general, low levels of the ought-to L2 self compared to English language learners (e.g., de Burgh-Hirabe, 2019; Thompson, 2017). Moreover, the relevance of the ought-to L2 self to motivated behaviours has been found to vary across contexts (e.g., Huang, 2019; Kong et al., 2018). Therefore, some researchers have speculated that LOTE learners’ ought-to L2 self is more fragmented than that of English learners and, in some cases, cannot even be regarded as a unified self-dimension (Dörnyei & Al-Hoorie, 2017). Additionally, empirical studies involving LOTE learners have obtained polarised results regarding the L2 learning experience. For example, the experience of L2 learning was the strongest predictor of the intended effort of LOTE learners in a study conducted in Taiwan (Huang, 2019) but failed to even emerge as a dimension influencing the intended effort of German language learners in a study carried out in Hungary (Csizér & Lukács, 2010).

With 82 million speakers, Korean ranks as the world’s 20th most spoken language, far behind other LOTEs, such as Spanish, French, and German (Eberhard et al., 2021). However, over the last few years, Korean has emerged as a language of global influence. In 2020, according to a global language report produced by Duolingo, the world’s foremost language education app, Korean was the second fastest growing language and the seventh most popular language to study worldwide. Previous studies of Korean language learning motivation mostly dealt with either learners from English-speaking countries (Fraschini & Caruso, 2019; Lee, 2019) or international students studying in South Korea (Zhang & Kim, 2014). Those studies have been devoted to identifying different types of Korean language learning motivation and consistently highlighted the significant role of interest in Korean culture in facilitating Korean language learning (Lee, 2018). In terms of the applied motivational theories, the majority of studies have been conducted within Gardner’s socio-educational model and self-determination theory (Jee, 2015; Lee, 2019), and only a few studies were conducted under the L2MSS framework, with an emphasis on L2 selves of Korean language learners (Han, 2021; Wang, 2022).

Of the very limited number of research informed by the L2MSS framework, Wang’s (2022) study with Chinese learners of Korean language found that both ‘ideal Korean self’ and ‘ought-to Korean self’ positively predicted learners’ intended effort, and the former was found to be a stronger predictor than the latter. In Zhang and Kim’s (2014) study with international students studying in South Korea, learners’ motivational characteristics were found to vary with their cultural background. Particularly, Asian learners and non-Asian learners reported significant differences in their perceived levels of the ideal L2 self and the L2 learning experience. These results were also evidenced by recent studies conducted in Australia and Vietnam, in which Australian learners were found to be more likely to learn Korean for enjoyment and leisure (Fraschini & Caruso, 2019), while Vietnamese learners of Korean language were found to be more concerned with careers or high income (Han, 2021).

In summary, despite the growing popularity of the Korean language worldwide, research on Korean language learning motivation remains particularly rare compared to other LOTEs (Mendoza & Phung, 2019), and further studies are needed in a broader context. Considering the large number of Korean language learners in mainland China, it is worthwhile to explore what motivates Chinese students to learn Korean to improve our understanding of LOTE learning motivation.

Engagement and its Relationship with Motivation

Engagement refers to the psychological effort and investment made by learners to learn, understand, or master academic skills or knowledge (Newmann, 1992). It is viewed as a major predictor of academic achievement and has long been a topic of interest in research on educational psychology (Skinner et al., 2009). Engagement is equally important to L2 acquisition, because the active participation of students is key to language learning (Dörnyei, 2019), and there is growing evidence that engagement has a strong influence on L2 learning outcomes (Zhang et al., 2020).

Engagement is generally regarded as a multifaceted construct with three major components, namely behavioural engagement, emotional engagement and cognitive engagement (Fredricks et al., 2005). Behavioural engagement denotes learners’ active participation in their tasks and classroom learning and includes behaviours such as persistence, attention, and effort (Skinner et al., 2009). Emotional engagement focuses on students’ emotional states, such as their enjoyment and curiosity, while involved in learning activities (Skinner et al., 2009). Cognitive engagement reflects students’ strategic learning or self-regulation during learning activities (Fredricks et al., 2005).

Theoretical and empirical studies have suggested that engagement is different from but related to motivation (e.g., Chen & Kraklow, 2015; Skinner et al., 2009). Motivation involves private, unobservable psychological factors and acts as an antecedent of engagement that can be publicly observed (Reeve, 2012). Regarding language classroom contexts, however, studies of the relationship between motivation and engagement are relatively limited (Oga-Baldwin & Nakata, 2017). Chen and Kraklow (2015) found that L2 motivation as conceptualised by self-determination theory explained only a small proportion of English learning engagement. They further speculated that the L2MSS may better explain engagement, as students may be engaged due to their language interactions in and outside the classroom and the need to prepare themselves for global citizenship, the latter of which can in turn be conceptualised as the ideal L2 self. However, as the majority of motivation research under the framework of L2MSS has used intended effort as the primary criterion variable (Al-Hoorie, 2018), studies exploring the relationship between motivation and engagement, especially those pertaining to Korean language learners, are still scarce (e.g., Zhang et al., 2020). Of the particular relevance to the current study is Huang et al.’s (2015) research with LOTE learners in Taiwan, in which the learning experience component out of the L2MSS model for Korean language was found to be significantly related to class involvement (behavioural engagement). To the best of the authors’ knowledge, empirical studies that specifically explore the relationships between motivation and the different subdimensions of engagement have not yet been conducted.

Building on the literature, the purpose of the current study was to investigate the motivation for Korean language learning and its relationship with the multifaceted learning engagement in China. Within the framework of the L2MSS model, the study aimed to answer the following two specific questions:

  1. (1)

    What are the motivational characteristics of Korean language learners in China ?

  2. (2)

    How is Korean language learners’ learning motivation related to their learning engagement?

Methodology

Participants

An investigation using an online questionnaire was carried out from October to November 2020 through the researchers’ contacts with Korean language departments at universities in China. A total of 1,021 Korean major students from 28 universities were invited to participate in the study by using a convenience sampling method. The participants were informed that their participation would be anonymous and voluntary. The study was based on 967 (94.7%) validated questionnaires. As shown in the list of demographic details in Table 1, 858 (88.7%) students were female and 109 (11.3%) were male. As for the type of institutions, 267 students (27.6%) were from 11 key national universities, and 700 (72.4%) were from 17 provincial and municipal universities.

Table 1 Participants’ background information (N = 967)
Full size table

Instruments

The questionnaire consisted of two sections. The first section asked about the participants’ personal backgrounds. The second section included two scales, an L2 motivation scale and an L2 engagement scale (see Appendix). We adapted the L2 motivation scale developed by Taguchi et al. (2009) with three subscales: the ideal L2 self, the ought-to L2 self, and the L2 learning experience. The target language described in the items was replaced with Korean.

The L2 engagement scale consisted of three subscales. The behavioural engagement and the emotional engagement were adapted from Skinner et al.’s (2009) Engagement Versus Disaffection with Learning Scale, reflecting students’ behavioural and emotional participation in learning activities in the classroom. The cognitive engagement was adapted from Fredricks et al. (2005), with two items being excluded regarding the Chinese context. Wording related to general education was replaced with references to Korean language education in all subscales.

The scales were translated into Chinese, with a pilot test conducted among Korean major students who did not participate in the formal survey. Items that were confusing were reworded based on the students’ feedback. For all of the items, we used a 5 point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree).

Data Analysis

SPSS 22.0 and AMOS 22.0 were used to process the collected data. Confirmatory factor analysis (CFA) was used to examine the construct validity of each scale, and Cronbach’s α coefficients were used to test the subscales’ reliability. Descriptive statistics (M, SD) and correlations between the factors were also examined. Finally, structural equation modelling (SEM) was performed to explore the relationships between the learning motivation and engagement of the participating Korean language learners. As previous studies indicated, model fit is excellent when the Tucker–Lewis index (TLI) and comparative fit index (CFI) are above 0.95 and acceptable when TLI and CFI are above 0.90 (Schreiber et al., 2006). For the root mean square error of approximation (RMSEA), values under 0.05 suggest good fit, values as high as 0.08 suggest reasonable fit, while values between 0.08 and 0.10 suggest mediocre fit (Browne & Cudeck, 1993). Effect sizes were calculated for the variables using Gignac and Szodorai’s (2016) guidelines (small = 0.10 −  < 0.20, medium = 0.20 −  < 0.30, large ≥ 0.30).

Results

Construct Validity, Reliability, Descriptive Statistics and Correlation Analysis

The CFA results for each scale indicated that both of the measurement models had good construct validity. The CFA results for the L2 motivation scale revealed that the three-factor L2MSS had a good fit to the data (χ2 = 542.49, df = 97, p < 0.001, TLI = 0.95, CFI = 0.96, RMSEA = 0.069). The factor loadings ranged from 0.52 to 0.93 among the included items, and the Cronbach’s α coefficients of the three factors ranged from 0.89 to 0.92, showing that all of the subscales had good internal consistency (as shown in Table 2).

Table 2 Variables, descriptive statistics and reliability (N = 967)
Full size table

The fit indices of the L2 engagement scale were also acceptable (χ2 = 888.72, df = 98, p < 0.001, TLI = 0.95, CFI = 0.96, RMSEA = 0.091). The factor loadings of all of the items varied from 0.74 to 0.94. The results also indicated that all of the subscales had good internal consistency, as the Cronbach’s α coefficients ranged from 0.91 to 0.96.

The descriptive statistics, presented in Table 2, indicated that students scored highest for the L2 learning experience (M = 3.68, SD = 0.77). The result of one-way ANOVA (F(1,1932) = 1,712.33, p < 0.001) revealed that the mean score for the ideal L2 self (M = 3.65, SD = 0.81) was significantly higher than that for the ought-to L2 self (M = 2.15, SD = 0.79), with a large effect size (η2 = 0.47). Based on the mean scores for components of learning engagement, the students scored highest for behavioural engagement (M = 4.07, SD = 0.69), followed by emotional engagement (M = 3.99, SD = 0.70) and finally cognitive engagement (M = 3.86, SD = 0.67).

As for the results of the correlation matrix (see Table 3), the ideal L2 self and the L2 learning experience had a positive and significant association with all three components of engagement, with large effect sizes. The ought-to L2 self was significantly associated with the L2 learning experience and cognitive engagement, with small effect sizes, but it was not significantly related to either behavioural engagement or emotional engagement.

Table 3 Correlations among latent variables
Full size table

Structural Equation Model

The relationship between the Korean language learners’ motivation and engagement was explored by constructing a structural equation model using AMOS 22.0, with motivation as the independent variables and engagement as the dependent variables. As shown in Fig. 1, an acceptable fit was found for the structural model of the relationships between motivation and engagement (χ2 = 1972.81, df = 442, p < 0.001, TLI = 0.94, CFI = 0.95, RMSEA = 0.060). The explained variances of the L2 engagement factors, from high to low, were 0.54 (emotional engagement), 0.50 (cognitive engagement), and 0.43 (behavioural engagement).

Fig. 1
figure 1

SEM model results with significant regression paths (N = 967). ***p < 0.001; Goodness-of-fit indices: χ2 = 1972.81, df = 442, p < 0.001, TLI = .94, CFI = .95, RMSEA = 0.060

Full size image

The SEM results indicated that the ideal L2 self was associated positively with all of the three engagement factors, with moderate to large effect sizes and the L2 learning experience was also associated positively with all of the three engagement factors, with large effect sizes. The ought-to L2 self was negatively related to all engagement constructs except cognitive engagement. However, the effect sizes of its association with behavioural engagement (β =  − 0.11, p < 0.001) and emotional engagement (β =  − 0.09, p < 0.001) were very small ( <|0.20|) and have no practical significance.

Discussion

This study adds to research on LOTE learning motivation by revealing the associations between the learning motivation and engagement of Korean language learners in China. Using a sample of Korean major students at Chinese universities, the study indicated that the L2MSS model is a valid tool for understanding and explaining students’ motivation to learn Korean. The SEM results provided insights into the influence of different motivational factors on the three dimensions of engagement.

Motivational Characteristics of Korean Language learners

This study revealed some motivational characteristics of Korean language learners in China. As in most motivation research on LOTEs (e.g., Chen et al., 2021; Huang et al., 2015), the students reported high levels of the L2 learning experience (M = 3.68, SD = 0.77) and the ideal L2 self (M = 3.65, SD = 0.81). This indicated that the students had a strong aspiration to become fluent Korean speakers and that they generally enjoyed their Korean learning experience. This is probably attributable to students’ keen interest in Korean culture, which has been found to enhance learners’ motivation for beginning language learning, their initiative to envision target communities, and their investment in Korean language learning (Lee, 2018). Korea is geographically and culturally proximate to China. Since the 1990s, Korean pop culture has spread to China, growing over the past three decades into a huge cultural phenomenon known as the ‘Korean Wave’. This has stimulated Chinese students’ motivation to learn the Korean language. Imported Korean cultural products such as K-pop music, films and TV dramas offer diverse learning materials that enrich Chinese students’ Korean learning experience.

Comparatively, there was a large gap between the low mean score for the ought-to L2 self (M = 2.15, SD = 0.79) and that for the ideal L2 self, indicating that the students generally learned Korean of their own volition, rather than out of a sense of obligation to others such as parents or friends. This result is consistent with the findings of Zhang and Kim’s (2014) study, in which Korean language learners, regardless of background, were found to perceive higher levels of the ideal L2 self than the ought-to L2 self. Likewise, in studies involving LOTE learners in New Zealand and the United States, the ought-to L2 self has also been found to be the weakest component of the L2MSS (de Burgh-Hirabe, 2019; Thompson, 2017). This is probably due to the lower demand for LOTEs than for English at the societal level, leading to inadequate social support for LOTEs (Dörnyei & Al-Hoorie, 2017). In China, there is a large Korean ethnic group whose members speak both Korean and Chinese, with the former being their first language. With bilingual skills, members of the Korean ethnic group undertake a variety of jobs relating to business with South Korea (Tu, 2018), leading to the relatively low demand for Korean majors in the job market. Therefore, for Korean language learners in China, external expectations to learn the language is low.

The Relationship Between Korean Language Learners’ Motivation and Engagement

The SEM results revealed that among the Korean language learners, the ideal L2 self and the L2 learning experience had a positive association with all three dimensions of engagement, and the ought-to L2 self was not significantly associated with all of the three engagement factors. The significant positive correlation between engagement and the ideal self suggested that the students who more clearly visualised their future selves as fluent Korean speakers had a stronger intention to behaviourally, emotionally, and cognitively implement learning activities related to the Korean language. The results confirmed the findings of previous studies in which the ideal self was found to be positively associated with L2 persistence and effort (Feng & Papi, 2020; Zhang et al., 2020). Meanwhile, learners with strong ideal selves were found to be more likely to feel joy while learning L2 (Papi & Khajavy, 2021), leading to higher levels of emotional engagement. The positive relationship between the ideal self and cognitive engagement was in line with the finding of Csizér and Kormos’ (2014) study with Hungarian learners of English in which the ideal self was identified as a positive predictor of self-regulation. The ideal self serves as an important driver of self-motivated engagement in L2 learning as learners would like to reduce the discrepancy between their ideal selves and actual selves (You & Dörnyei, 2016). Therefore, learners with high levels of the ideal self would like to put more effort and display more interest and curiosity, as well as conduct more effective strategic learning to improve their language skills. Our study added empirical evidence to support that motivated behaviour had a strong link with the ideal L2 self, regardless of language types.

This study also revealed positive relationships between the learning experience of Korean language learners and all three engagement dimensions, with the largest effect sizes. This is consistent with the finding of Huang et al.’s (2015) study with Taiwanese learners of Korean language, in which students who perceived higher levels of the L2 learning experience were found to be more likely to participate in classroom activities. As the L2 learning experience captures learners’ perceptions of the immediate learning environment (Dörnyei, 2005), it may directly influence learners’ engagement in immediate learning activities. A recent survey with French language learners (Sulis & Philp, 2021) also highlighted the role of positive classroom atmosphere shaped by teachers and peers in students’ behavioural, emotional, and cognitive engagement. Specifically, teachers’ provision of equal opportunities for participation and monitoring of group activities were found to promote students’ active participation in the classroom, and their feedback during class or through assessments helped students evaluate their competence and monitor the effort to be exerted in the following learning activities. Moreover, the support and mutual aid offered by peers also contributed to fostering persistence and enjoyment.

It should be noted that, of the three motivational factors that make up the L2MSS model, the L2 learning experience is not the theoretical emphasis (Dörnyei, 2005), and has been omitted from the majority of previous studies of Korean language learning motivation, which have frequently focused on the two future self-guides (e.g., Fraschini & Caruso, 2019; Wang, 2022). However, the large magnitude of its path relationships with motivated behaviours, especially in the context of English language learning, has been demonstrated in many studies (e.g., Taguchi et al., 2009; You & Dörnyei, 2016). Recent motivational studies of LOTEs have also revealed that the L2 learning experience is the strongest predictor in the L2MSS (e.g., Kong et al., 2018), and speculated that for LOTE learners who rarely have the opportunity to use the language in their daily lives, interaction with teachers and peers dedicated to these LOTEs remain the main source of influence on their behaviours (Huang, 2019). In China, a number of Korean language teachers are of Korean ethnic. This has created an immersive language learning environment for students, and thus might make them more engaged to learn the Korean language. Our results further indicated the significance of the L2 learning experience to engagement in learning activities, highlighting the need for more investigation of LOTE learners’ learning experience.

Our study also revealed that the ought-to self could not account for the learning engagement of Korean language learners. This is in line with the finding of a previous study suggesting that the ought-to L2 self is not a significant predictor of classroom engagement (Zhang et al., 2020). As ought selves are usually regarded as short-term goals that do not need persistent efforts, they may not influence learners’ behavioural engagement like persistence (Feng & Papi, 2020). Meanwhile, learners with strong ought-to L2 selves tend to learn the L2 in order to meet the expectations of the external world and avoid negative outcomes, which may direct them to take L2 learning as an obligation and thus less likely to undertake strategic learning during learning activities. Moreover, as the ought-to L2 self has also been found to be unrelated to enjoyment (Papi & Khajavy, 2021), an emotional state representing the enthusiasm and interest in the learning process, this may have led to its inadequate explanatory power for emotional engagement.

Additionally, the ought-to L2 self has generally been found to have a positive effect on the motivated behaviours, especially in East Asian cultural contexts (Kong et al., 2018); this has typically been attributed to the influence of Confucian heritage culture or social obligations (Huang et al., 2015). In some recent studies, however, the association between Chinese students’ ought-to self and engagement or intended effort in L2 learning has not been found to be significant (e.g., Huang, 2019; Zhang et al., 2020). A study with Chinese learners revealed that they were more self-motivated to learn LOTEs and their parents rarely interfered in their academic studies, leaving the ought-to L2 self unrelated to their motivated behaviours (Chen et al., 2021). Our study provides additional evidence that in the Chinese context, the ought-to L2 self is not always a significant latent dimension.

Some limitations of the study need to be noted to shed light on future research directions. First, as this study was designed as a cross-sectional investigation, it could not identify the potential causal relationships between motivation and engagement. A longitudinal research design should be used in the future to examine the directionality of the regression paths. Second, the findings of the current study were mainly based on self-report measures, which may have led to inflated results for the relationships between motivation and engagement. Future research could be conducted using a mixed-methods design to add new insights into the relationships between the variables. Third, as this study used a variable-centered approach, the findings corresponded to a statistically defined average participant who might not be representative of any actual participant in the study (Roeser et al., 1998). Future studies may consider using a person-centered approach, such as latent profile analysis, to enable a richer description of the different subgroups in a particular sample. Last but not least, as a preliminary study based on questionnaire survey, the single source of data may not provide sufficient information to comprehensively and deeply understand the subject of the study, multiple types of data are expected to be used to triangulate the findings in future studies.

Implications for Practice

This study enriches understanding of learners’ motivation to learn LOTEs by shedding light on the relationship between motivation and engagement among Korean language learners in China. The findings of the study have some significant pedagogical implications related to understanding and improving LOTE learning in higher education institutions.

First, the strong positive relationship between learning experience and engagement indicates that teachers need to provide a conducive learning environment that helps students be more involved in language learning (Dörnyei, 2001). It is advisable for teachers to ensure equal participation, design cooperative learning activities and provide timely feedback in language teaching. Meanwhile, given the rapid development of digital technology nowadays, teachers may help increase students’ learning enjoyment by encouraging informal absorption and use of L2 through various digital resources, such as language learning apps and social media (Lee & Lee, 2021). Such strategies stimulate students’ interest in language learning and improve their initiative and self-learning ability, leading to enhanced engagement.

Second, the significant relationship between the ideal L2 self and engagement revealed in the current study indicates the importance of promoting vivid ideal L2 selves among students. Diverse approaches (see Dörnyei & Kubanyiova, 2014) are needed to help enhance students’ vision of ideal L2 selves, thus stimulating them to engage more actively and enthusiastically in classroom learning and encouraging them to initiate self-regulation strategies to resolve the discrepancy between their actual and ideal selves. For example, teachers could use guided imagery to design motivational activities, show videos of outstanding peers participating in L2 speech competitions, and provide volunteering opportunities where students can interact with native speakers.