In the last two decades, positive psychology (PP) has caused an everlasting revolution in mainstream psychology, deviating from an exclusive focus on problems to more positive dimensions of humans’ minds (Dewaele et al., 2019). Before the advent of PP, psychology was regarded as a science to repair the damaged. It was replete with studies on “psychological disorders and the negative effects of environmental stressors such as parental divorce, death, and physical and sexual abuse” (Gallagher and Lopez, 2009, p. 4). Around the turn of the millennium, psychologists realized that psychology is not just about mental illnesses and how to heal them. With the rise of PP, these researchers started investing in positive aspects of human experience, such a hope, love, joy, optimism, etc. (Dewaele et al., 2019). Seligman and Csikszentmihalyi (2000), as the pioneers of the field, explain that PP aims to “catalyze a change in the focus of psychology from preoccupation only with repairing the worst things in life to also building positive qualities” (p. 5).

Throughout the years, a cognitive perspective dominated the SLA and applied linguistics resulting in sheer ignorance of the human psyche (Dewaele et al., 2019). With the flourishing of humanistic approaches such as the whole-person approach in the SLA domain, affective variables were brought into the spotlight. However, researchers believe that even after thriving of scholars’ interest in affective variables, positive emotions have been overshadowed by negative emotions (Seligman et al., 2006). Therefore, little attention has been given to the role of positive variables in second and foreign language teaching and learning. Under the umbrella of PP, there has also been a shift from a deep-rooted focus on negative emotions, especially anxiety (Jiang and Dewaele, 2019) and cognitive perspectives (Li, 2020) in the SLA, to a more holistic analysis of emotions. Scholars have started to recognize the positive and negative emotional dimensions of language learning as significant assets. Before that, emotions were “poorly studied, poorly understood, seen as inferior to rational thought” (Swain, 2013, p. 11).

In accordance with positive psychology, researchers have started a positive renaissance or, in other words, a flowering of positive psychology in the SLA, too (Lee, 2020). Based on this view, MacIntyre and Gregersen (2012), pointed out that L2 teachers can help learners increase the degree of students’ positive emotions in the classroom, thereby “broaden a person’s perspective [and] opening the individual to absorb the language” (p. 193). PP in the SLA domain has been studied in relation to a variety of perspectives such as academic achievement, positive L2 self (e.g., Kikuchi and Lake, 2021), motivation (e.g., Martin, 2005), emotional intelligence (Li, 2020), WTC (e.g., Mystkowska-Wiertelak, 2021), and classroom enjoyment (Jin and Zhang, 2018). With the advent of PP, another variable gardening attention in the SLA domain is grit, a positive, non-cognitive trait which has been discussed in detail in the next section.

As the pioneers of the grit construct, Duckworth et al. (2007) defined the notion in a broad sense concerning success in the workplace, military, education, and sports. Grit gains immediate relevance in the SLA field, especially with the blooming of PP in this domain, since successful proficiency of an L2 is highly dependent on learners’ sustained effort (Dörnyei and Ushioda, 2013). Researchers have established that grittier EFL students succeed in achieving higher scores in English (Strayhorn, 2014; Khajavy et al., 2021), employ persistent effort during learning an L2 (Lake, 2015), and experience more positive classroom emotions (Wei et al., 2019).

Grit as a higher-ordered variable comprises two lower-order constructs: Perseverance of Effort (POE), which refers to a person’s continued investment of energy in long-term pursuits; and Consistency of Interests and passion for long-term goals (COI), which refers to a person’s durable passion for high-order goals over a long period regardless of failures, disappointments, or challenges (Lee, 2020; Teimouri et al., 2020). Since grit is not a fixed personality trait and can be malleable and teachable, the importance of measuring and estimating grit is gaining unprecedented attention. Duckworth et al.’s (2007) grit is a domain-general variable that could be applied to various fields such as military, workplace, education, and sports. As Teimouri et al. (2020) explained, there are two main reasons behind the incongruent findings of research on personality variables such as grit in the SLA field: “(a) its focus on super-traits instead of lower-traits with more relevancy to L2 learning and (b) its use of general personality measures instead of the situation- or domain-specific measures that assess students’ traits in L2 situations” (p. 32). As a result, L2 researchers have recently extended the concept to L2 learning and developed L2 domain-specific grit notions.

Among these scholars, Teimouri et al. (2020) and Alamer (2021) proposed L2 domain-specific grit scales to help researchers in addressing grit in the language studies domain. The former examined the 12-scale validity through principal component analysis (PCA), which resulted in the reduction of the items from 12 items in the original scale to 9 items. They explained that grit as a higher-order factor consisting of two lower-order elements, namely POE and COI, should be examined separately. The Latter, however, retained the 12 items and used bifactor-CFA to account for the specific constructs, COI and POE, and the general construct, Grit. The study examined the effect of L2 grit on future attainment of vocabulary. The study illustrated a weak predictive power of initial levels of grit on subsequent L2 learning. However, the author found that only students who sustained their levels of grit over time showed later language achievement. For this study that was conducted in Saudi Arabia, the author developed a 12-item L2 domain-specific grit scale. In the present study, we utilized Teimouri et al.’s (2020) grit scale.

Motivation is of great significance within the field of SLA. To highlight the importance of motivation in language learning, Dörnyei (2005) argues that motivation is the key impetus to start learning an L2 and to carry on through the long and arduous journey. He also points out that “all the other factors involved in SLA presuppose motivation to some extent” (p. 65). While early studies on L2 motivation investigated the issue through Gardner’s (1985) socio-educational model, the concept was soon challenged by researchers (e.g., Dörnyei, 2005, 2009) who questioned the model’s efficiency for addressing the expanding FL learning environment. As an answer to a call for a reconceptualization of motivation theory and inspired by self-discrepancy theory (Higgins, 1987) and possible selves theory (Markus and Nurius, 1986), Dörnyei (2005, 2009) presented a prominent model of learner motivation that is L2 motivational self-system. L2MSS has three components; ILS is the L2-specific facet of one’s ideal self; OLS refers to the attributes that one believes one ought to possess to avoid possible negative outcomes (Dörnyei, 2009, p. 29). And L2 Learning Experience focuses on the learner’s present experience, covering a range of situated, “executive” motives related to the immediate learning environment (Dörnyei and Ryan, 2015, p. 88).

Over the past years, multiple empirical studies have been conducted to test and validate L2MSS in various learning environments (e.g., Csizér and Kormos, 2009; MacIntyre et al., 2009; Ryan, 2009; Taguchi et al., 2009; Kim and Kim, 2014). All the studies in the literature report solid confirmation for the proposed L2MSS. Inspired by Dörnyei’s (2005) argument regarding a positive relationship between the ILS and L2 WTC, a new strand of research has focused on the relationship between L2 WTC and L2MSS. Munezane (2013) was the first researcher who tested the path between the ILS and L2 WTC and approved Dörnyei’s (2005) hypothesis. Ever since, various studies have lent further support to the relationship between L2MSS components and L2 WTC (see Munezane, 2015, 2016; Darling and Chanyoo, 2018; Zulkepli, 2020).

Verbal communication holds a central place among humans. The importance of talk in human interpersonal communication is an undeniable fact. As Khajavy et al. (2018) put it, the plethora of investigations on communication is because having communicative competence does not necessarily imply a willingness to use the language for authentic communication by itself. This can explain why some people with an acceptable linguistic competence may avoid communication and stay silent. In contrast, some with partial competence seek every opportunity to speak, which brings us to the issue of WTC.

Throughout the last decades, Considerable research has been conducted on the various dimensions involved in WTC as an influential aspect of communication in L2. MacIntyre et al. (2003, p. 590) define WTC as “the probability of initiating communication, specifically talking when the opportunity arises.” In its early stages, WTC was considered as a personality variable that is stable across time and in different situational contexts. More recently, however, L2 WTC has been presented not only as a stable tendency toward communication but also as a situational trait. Nowadays, researchers believe that internal variables such as L2 motivation and attitude toward L2 and external variables such as inter-group climate and social support can influence learners’ L2 WTC (Zhang et al., 2019).

In their groundbreaking model of L2 WTC known as the Heuristic model of WTC in L2 (Figure 1; MacIntyre et al., 1998) emphasized that L2 WTC is dependent on complex interactions of varying factors such as context, personality traits, cognition, and emotion. This model illustrates the multifaceted effects of different individual and situational variables, including the variables of interest to the present study (i.e., grit and L2MSS), in each layer on L2 communication. This model consists of six layers which include 12 constructs. In this model, the above layers (I, II, and III) include “more changeable, and context-dependent” (state) variables, and at the bottom layers of the pyramid (IV, V, and VI) more “stable social individual context” (trait) variables are placed (Dewaele and Pavelescu, 2021, p. 3).

Lee (2020) examined the effects of grit and classroom enjoyment on Korean FL students’ WTC. To obtain the data, he utilized Duckworth et al.’s (2007) grit scale, Dewaele and MacIntyre’s (2014) FLE scale, and a five-item WTC scale (Lee et al., 2020). Lee realized that POE and classroom enjoyment (CE) correlate positively with L2 WTC among the participants. He explained that students with higher levels of POE who experience a positive environment in the classroom are more likely to show higher levels of WTC. On the other hand, COI was not observed as a significant predictor of L2 WTC. Another important finding of this study was the positive relationship between L2 WTC and the length of time devoted to learning English. In other words, students with more experience learning English tend to communicate more than others; on the contrary, Lee found out that there is a negative correlation between the age of students and their level of L2 WTC.

Cheng (2021) investigated the effect of language-specific grit and future self-guides on Taiwanese college English students’ L2 WTC. In his study, L2 grit was measured using the 9-item scale developed by Teimouri et al. (2020), and L2 selves were estimated by Tseng et al.’s (2020) 16-item scale measuring four different L2 self-configuration (i.e., ILS/own, ILS/other, OLS/own, OLS/other). The data regarding the L2 WTC was collected using a questionnaire developed by Lee and Hsieh (2019). In disagreement with Lee (2020), the results of this study indicated that both POE and COI could predict L2 WTC. Yet, Cheng (2021) realized that COI showed a far less predictive power compared to POE. Besides L2 grit, future L2 self-guides exerted a predictive power on L2 WTC. While both ILS, own and other, played a predictive role, only OLS/own could affect WTC, and OLS/other failed to do so.

Bursali and Öz (2017) focused on one of the components of L2MSS, that is, the ILS, and explored its relationship with L2 WTC of Turkish EFL learners. The data collected through a questionnaire from 56 students showed a positive and significant relationship between learners’ ILS and their L2 WTC inside the classroom. The findings indicated that 32.1 percent of the participants had high, 30.4 percent had moderate, and 37.5 percent had low L2 WTC inside the classroom. However, they have noted that due to this study’s limited sample, a generalization cannot be made concerning the relationship between L2MSS and L2 WTC at large.

Ghasemi (2018) set out to probe deeper into the WTC of Iranian English language major students and determine the variables contributing to L2 WTC in a single mode. The interrelationships between motivation, perceptual learning styles, cognitive variables, and language proficiency were under focus. Specifically, the role of the ILS, visual learning styles, and L2 confidence in predicting learners’ language proficiency and WTC in real life were analyzed. 150 university freshmen students answered a five-section questionnaire. The results revealed that visual learning styles positively impact the ILS and are indirectly connected to L2 proficiency through the mediating function of motivation. These relationships support the fact that learners with more vivid images of the distance between their actual and ILS find language learning a goal-oriented activity and, consequently, are encouraged to attain higher levels of language proficiency. The desire to fill the gap between their current and ideal selves increases learner self-confidence, leading to higher levels of WTC.

To better understand the link between the ILS and L2 WTC, Lan et al. (2021) conducted a study that proposed a moderated mediation model incorporating grit and psychological shyness. To this end, the data were collected through a four-section questionnaire from 842 undergraduate students in three Chinese universities. Their findings revealed both direct and mediated links between the ILS and L2 WTC. ILS was positively associated with L2 WTC, and grit played a mediating role in this relationship. The authors argue that language learners with clear ILS images are grittier in achieving their language learning goals. Furthermore, learners with higher grit levels are more successful in facing learning tasks, such as participating in communicative activities.

In the present study, the aim is investigating the role of L2MSS and grit in EFL learners’ L2 WTC in two different educational settings of public schools and private language institutes. To this end, the following research questions have been posited:

A total of 308 Iranian EFL learners from both public schools and private institutes completed a survey. Based on convenience sampling, the participants selected from the public schools were studying in the 11th and 12th grades at the time. As for the institute sample, language learners with elementary (A1) and pre-intermediate (A2) levels of English proficiency were selected so that the two groups of participants were at the same level of language proficiency. These levels were based on the Common European Framework of Reference for Languages (CEFR), which describes language ability on a six-point scale, from A1 for beginners to C2 for advanced students. Gender was not considered a variable in our study; hence, the inequality of gender distribution was not an issue. Table 1 represents the demographic characteristics of each group that took part in this study.

The data for each variable of interest were collected using an online survey consisting of four parts, viz. demographic information, L2 WTC questionnaire (Peng and Woodrow, 2010), grit questionnaire (Teimouri et al., 2020), and L2MSS questionnaire (Taguchi et al., 2009). In the first part of the survey, demographic information, two items in the public schools’ questionnaire version were added to detect and remove those public-school students who were simultaneously attending private institutes and to identify students’ grades. The demographic section of the questionnaire administered to the learners in the private institutes also contained an additional item regarding their current English proficiency level. Moreover, to prevent age differences from affecting our results, the data collected from those institute learners older than 18 and younger than 15 were excluded from the analysis. The details of the different parts of the questionnaire are detailed below.

To collect the required data for the present research, the following procedure was adopted. First, the questionnaires were translated into Persian by two professional translators. Google Forms was chosen to design an online survey. A pilot study was conducted with 50 students (25 public school students and 25 private language institute learners) from the target population. Based on the participants’ comments, the translation of some of the items were edited to avoid confusion. After obtaining the schools (N = 5) and institutes (N = 6) board approval, the survey link was sent to the participants. The participants were assured of anonymity, and their participation was entirely voluntary.

Cronbach’s alpha was used to estimate the reliability of each questionnaire comprising the survey. As seen in Table 2, the questionnaires enjoyed high levels of reliability to elicit L2 WTC, Grit, and L2MSS from all the respondents in public schools and private institutes.

For the first research question, the mean and standard deviation for each group were calculated (Table 3). The results of the descriptive analysis revealed that the mean of L2 WTC in public schools was 2.56, and the standard deviation was 1.29. Table 3 displays the descriptive statistics for WTC, L2 Grit (POE and COI), and L2MSS (ILS and OLS) in public schools and private language institutes. The respondents in public schools mean performance regarding these variables was above 2, which indicates a medium performance on L2 Grit (POE and COI), and L2MSS (ILS and OLS).

The mean of L2 WTC in private institutes was 3.82, and the standard deviation was 1.22, revealing that the participants’ level of L2 WTC was also medium. The mean for POE, COI, ILS, and OLS, which were 3.90, 2.33, 2.61, and 3.31, respectively, indicated that the participants’ POE, COI, ILS, and OLS were at medium range for private institute learners. The POE has the highest mean (3.90) compared with that obtained for COI, ILS, and OLS.

To investigate how public school students and private institute learners differ in their level of L2 WTC, Kolmogorov-Smirnov and Shapiro-Wilk tests were first conducted to see if the data were normally distributed. The results showed the data did not enjoy normal distribution; therefore, the Mann-Whitney U-test, the non-parametric equivalent of the independent t-test, was used. As shown in Table 4, the Mann-Whitney U-test showed a significant difference between the level of L2 WTC in private and public school learners (U = 6789.50, p < 0.001).

The Spearman correlation was employed to investigate the relationship between L2 WTC, L2 Grit (POE and COI), and L2MSS (ILS and OLS) among the EFL learners in public schools and private institutes (questions two and three). The results revealed that, in public schools, there was a significant positive association between POE and WTC, [r(132) = 0.588, p < 0.001] (Table 5). In the case of L2MSS, the results indicated a significant positive correlation between OLS and L2 WTC, [r(132) = 0.679, p < 0.001] and, between ILS and L2 WTC, [r(132) = 0.340, p < 0.001]. Grit (COI), on the other hand, did not show a significant correlation with L2 WTC, [r(132) = 0.110, p > 0.05]. Regarding the private institute learners, the results indicated significant positive associations between POE, [r(176) = 0.352, p < 0.001], and ILS, [r(176) = 0.372, p < 0.001], and L2 WTC in private institutes’ context. COI, [r(176) = 0.135, p > 0.05], and OLS, [r(176) = –0.117, p > 0.05], on the other hand, did not show significant correlations with private institutes’ EFL learners’ L2 WTC.

Two separate regression analyses were also performed to estimate the predictive power of L2 Grit (POE and COI) and L2MSS (ILS and OLS) for L2 WTC. Table 6 shows information about the regression model as a whole. The four variables together positively correlated with the total score at 0.90, which is high. The adjusted R² indicated that the model significantly predicted 81 percent of the variance in the population.

The results of the ANOVA test (Table 7) revealed that this model differed significantly from the hypothetical model (4,127) = 134.716, p <0.001).

Moreover, the coefficients table (Table 8) showed that POE as the Grit’s first component (Beta = 0.589, p <0.001), OLS (Beta = 0.439, p < 0.001), and the ILS (Beta = 0.143, p = 0.001) wielded significant predictive power over public schools’ students L2 WTC. On the other hand, COI did not show any significant predictive power on L2 WTC among learners (Beta –0.026, p 0.05).

Table 9 demonstrates that the four variables positively correlate with the total score at R = 0.508, which is high. R-squared (R²) shows 26 percent of the variance. The adjusted R² in the model reveals that L2 Grit (POE and COI) and L2MSS (ILS and OLS) could predict 24 percent of EFL learners’ L2 WTC.

Examining the ANOVA results (Table 10), one can realize that this model differed significantly from the hypothetical model [F(4, 171) = 14.855, p < 0.001].

The table of coefficients (Table 11) displays that POE as the Grit’s first component (Beta = 0.265, p = 0.003), and the ILS (Beta = 0.266, p = 0.007) exerted a significant predictive power over learners’ L2 WTC in the private institutes setting. COI and OLS, however, did not wield a significant effect on L2 WTC.

The descriptive data and the Mann-Whitney U-test demonstrated higher levels of L2 WTC among private institutes learners than public school students. From a socio-educational perspective, several reasons can be highlighted concerning this phenomenon.

In line with Teimouri et al. (2020), Cheng (2021), and Khajavy et al. (2021), who confirmed the superiority of a two-factor model over a single-factor model of language-specific grit, in the present study, the researchers examined grits’ components (POE and COI) separately. This is in disagreement with Duckworth et al. (2007), who saw grit as a single variable. The results of Spearman’s rho and multiple regression analyses revealed that grit (POE) has a positive correlation and significant predictive power with/over L2 WTC in both contexts, whereas grit (COI) showed no significant correlation or predictive power with/over EFL learners L2 WTC in either context. In a similar vein, Credé et al. (2017) proposed that “perseverance is a much better predictor of performance than either consistency or overall grit.” (p. 502).

The positive relationship between grit (POE) and WTC can be reviewed in light of Fredrickson’s (2003) “broaden and build” theory: The “broaden” side of this theory explains in what way positive emotions play a motivating role in encouraging the learners to explore new experiences and seek more learning opportunities. This theory expounds why learners who experience positive emotions in the classroom are more willing to participate in communicative activities. Therefore, we can infer from the analysis that grittier students who show determination and hard work are more likely to engage in communication despite setbacks. This is in agreement with Dörnyei and Ushioda’s (2013) claim that success in an L2 is at the mercy of learners’ sustained effort. These findings align well with those of the prior studies that confirmed a positive relationship between positive emotions and L2 WTC (Peng, 2015; Khajavy et al., 2018, 2021; Jiang and Dewaele, 2019).

Regarding grit (COI), during the lengthy and tedious language learning process, learners’ interest in language learning may suffer some ups and downs due to the numerous setbacks and failures. Nonetheless, learners’ hard work and sustained effort are not reliant on unwavering interest in the process. Similar findings in other contexts such as Korea (Lee, 2020), Switzerland (Karlen et al., 2019), and China (Feng and Papi, 2020) further confirms the results of this article.

According to the Spearman correlation, both ILS and OLS showed a significant relationship with L2 WTC in public school setting. However, in the case of the learners in the private institutes, only ILS correlated with L2 WTC. Additionally, the multiple regression revealed that OLS was the second predictor of WTC in public school, while it failed to predict WTC in private language institutes. On the other hand, the ILS was the second predictor of L2 WTC in private language institutes and the third in public schools.

As noted, between the two components of L2MSS, OLS plays a more vital role on L2 WTC of public school students, which suggests that these students tend to become motivated to learn English to meet the expectations of their significant others, such as parents and teachers. Although the OLS usually appears as the weakest component of L2MSS (Csizér and Kormos, 2009; Aubrey, 2014), research in collectivist cultures has proved the opposite. It has been claimed that learners with collectivist cultures, such as Thai and Chinese EFL learners, tend to have stronger OLS (Patterson and Smith, 2003; Taguchi et al., 2009; Lanvers, 2016; Darling and Chanyoo, 2018). In concordance with these findings, the results of the present study invite the possibility that Iranian collectivist culture has led to a significant relationship between OLS and school students’ L2 WTC.

The strong impact of the ILS on the L2 WTC of private language institute learners denotes that in this group of learners, an ideal future image rather than external expectations and obligations aid language learners to put aside any fears they might have of communication and engage in the communicative act. The general tenor of previous studies on L2MSS has introduced the ILS as the strongest dimension of the model (Csizér and Kormos, 2009; Dörnyei and Ushioda, 2009; Papi, 2010; Kim, 2012). Thus, the present study’s findings tie in well with the literature. Furthermore, the results of the current research are in line with studies that have reported a positive relationship between ILS and L2 WTC among Turkish (Bursali and Öz, 2017, 2018; Sak, 2020), Thai (Darling and Chanyoo, 2018), Japanese (Munezane, 2013, 2015, 2016), Chinese (Peng, 2015; Lan et al., 2021; Li and Liu, 2021), and Iranian (Ghasemi, 2018) EFL learners.