February 2020 – Volume 23, Number 4
Macquarie University, Sydney, Australia
An approximate replication of Yashima, Zenuk-nishide, and Shimizu (2004), the main focus of this study was on the relationship between L2 WTC and International Posture in an ESL environment. International Posture measures the degree of connectedness to the international community from within an EFL context. However, attitudes may change when there is direct communication with the L2 community. The present study tested the generalisability of IP to an ESL environment. The relationship between IP and L2 WTC was found to be non-significant. This was contrary to Yashima et al. (2004). It indicates that there may be a number of sociocultural factors at play in an ESL environment that may covertly influence a learner’s WTC. Further analysis showed that participants overseas for four months or more presented higher scores in Perceived Competence. This aligns with Yashima et al. (2004), where no relationship was found between Perception of Adjustment and Perceived Competence over a three week sojourn period. This indicates that it may take a longer time to develop language competence while overseas. Intercultural Friendship Orientation had higher scores for participants sojourning for three months or less, highlighting that early sojourners can have a mindset akin to learners in an EFL environment.
Keywords: willingness to communicate; international posture; language learning, overseas L2 learning; language motivation; L2 study abroad
Willingness to Communicate in the second language (L2 WTC) has become an important concept in aiding explanation of communicative behaviour in the target language. It supports the view that in order for language learners to maintain a desire to speak and learn the new language, it is essential for them to engage in meaningful conversation. As WTC is considered the concluding psychological step before L2 communication (MacIntyre, 2007), it follows that higher instances of L2 WTC are linked with higher instances of L2 communicative behaviour (Denies, Yashima & Janssen, 2015). This also aligns with Swain’s (2000) output hypothesis, which states that continued practice in the L2 will lead to successful language acquisition. Contemporary L2 pedagogy has therefore placed significant importance on communicative behavior in the classroom (Munezane, 2013) with the intention of improving learners’ L2 WTC. MacIntyre, Clément, Dörnyei & Noels(1998) even suggest that L2 WTC should be the primary goal of language instruction as to inculcate a desire to seek out and to engage actively in conversation is a necessity in the language learning process. The usefulness of L2 WTC has therefore led Ellis (2008) to describe it as a very auspicious and promising construct within SLA.
The importance of the WTC construct to the field of SLA has resulted in a significant body of research that has investigated the various affective and psychological variables that are understood to be predictors of L2 WTC. Such factors include communication anxiety, L2 competence, motivation, and attitude (e.g. Baker & MacIntyre, 2000; Hashimoto, 2002; MacIntyre & Doucette 2010; MacIntyre & Legatto, 2011; Yashima, 2002). However, with the continued growth and influence of globalising trends, resulting in more people communicating through English, research into sociocultural practices has gained a growing interest in explaining L2 WTC. Recent research into the L2 WTC construct has thus sought to position it within a more globalised context (De Costa, 2014).
One such investigator is Tomoko Yashima. Yashima has been recognised as a pioneer in the field of L2 WTC, especially in regards to learner attitudes, and her work has been cited in numerous studies into the L2 WTC construct (e.g. Aliakbari, Kamangar & Khany, 2016; Cameron, 2015; Ghonsooly, Khajavy & Asadpour, 2012; MacIntyre & Doucette, 2010; Şener, 2014). Gardner (1985) developed the attitudinal construct integrativeness, which describes language learners in terms of integration into a specific target group. Previous research had shown that learners with higher levels of integrative motivation tend to use the language more (e.g. Hashimoto, 2002; MacIntyre & Charos, 1996). However, Yashima (2002) questioned the application of this orientation in a foreign language (EFL) classroom as there is no contact with native speakers of English in this context. Through an investigation on Japanese learners in an EFL context, Yashima (2002) identified an orientation comparable with an integrative orientation and termed it International Posture (IP). It was different in that it highlighted the role of English as a global lingua franca with no specific target community in mind. IP can be considered to encompass attitudes pertaining to both instrumental and integrative orientations and describes a cosmopolitan outlook towards language learning. Ideally, it was proposed as an alternative to Gardner’s (1985) integrativeness construct. According to Yashima (2002), the fundamentals of IP are “interest in foreign or international affairs, willingness to go overseas to stay or work, readiness to interact with intercultural partners, and, one hopes, openness or non-ethnocentric attitude toward different cultures, among others” (p. 57).
Essentially, IP attempts to capture a tendency to have a connection to the international community when learning a language. In Yashima (2002), it was shown through structural equation modelling (SEM) that learners who find intercultural communication appealing (i.e. IP) would have the behavioral intention to communicate in the L2 (i.e. WTC). IP predicted L2 WTC. It was also shown that IP was a predictor of L2 learning motivation, which in turn was a predictor of L2 WTC. Confidence in L2 communication was also shown to be a predictor of L2 WTC.
In Yashima et al. (2004), SEM was again used to examine the relationships between L2 WTC’s associated variables. Additionally, self-reported frequency of communication was added to the model as the learners examined were in an L2 immersion context where use of English was an everyday occurrence. All paths within the model were significant – IP predicted L2 WTC and frequency of communication – L2 WTC predicted frequency of communication. A second part of this study also revealed that learners in a three week study abroad context who registered a high L2 WTC value prior to their sojourn, registered a high frequency of communication while on sojourn. Similar results were also shown in Yashima and Zenuk-Nishide (2008) for learners who had extended exposure to authentic L2 situations in a study abroad context. This aligns with Kang (2005) who indicated that L2 learners demonstrating a high level of L2 WTC would be more likely to use the L2 in authentic communicative situations (i.e. ESL environments).
Context of Communication?
Attitudes towards the learning situation can play a significant role in the language learning process. Yashima’s IP construct, which describes attitudes pertaining to an international or cosmopolitan orientation, has been shown to be a predictor of L2 WTC (e.g. Centinkaya, 2005; Matsuoka, 2006). However, Gallagher (2013) highlights that many investigations into L2 WTC were carried out within EFL classrooms (e.g. Ghonsooly et al., 2012; MacIntyre et al. ,2001; Munezane, 2013; Watanabe, 2013; Yashima, 2002), where contact with the target culture is not an issue. In contexts where English is not spoken as a daily communicative language (e.g. EFL contexts), learners may not develop a clear affective reaction to the target language or L2 group (Aliakbari, Kamangar & Khany, 2016; Dörnyei, 1990). While L2 WTC can be a very good predictor of language behavior, the context of interaction needs to be considered in understanding language learners’ communicative behavior (Cao, 2011). Therefore, when and where an interaction takes place, and with whom, can affect how willing a learner is to communicate in that context.
While full exposure to the target community is believed to benefit a language learner (Kramsch 2008), familiarity with that culture can better facilitate L2 WTC (Bennett, 2006). Furthermore, a learner’s attitude towards a target community can change once communication with that community is immediate and an everyday occurrence (Cameron, 2015; Dufon & Churchill, 2006). Thus, while Yashima’s IP construct successfully predicts L2 WTC in an EFL context, it may not be immediately applicable in predicting L2 WTC in an ESL context. In Zeng (2010), for example, which examined Chinese ESL learners in a Canadian context, there was no correlation between IP and L2 WTC. Lack of cultural understanding was indicated as being the main hindrance to effective communication.
While L2 WTC can be both trait like and situational, the expectation of communication through English may be stronger in the L1 environment than in the L2 environment. Further, the modern learner influenced by many global prospects may not see contact with a local English speaking population of paramount importance – contact with other L2 speakers, remaining within their own cultural peer group, or just experience abroad may be of more interest. Moreover, although L2 WTC has been investigated in several cultural contexts, the variety of these contexts has been rather limited (Cameron, 2015). Studies of L2 WTC in authentic L2 environments are also quite scarce (Zeng, 2010), thus, emphasising the need to extend studies in this area. This paper therefore presents an approximate replication of Yashima et al. (2004) in an overseas ESL environment. Particular focus is put on the L2 WTC and IP constructs and their relationship with associated variables. Further, while WTC is only a measure of L2 communication intent, investigating whether this intent is predictive of actual L2 behaviour in L2 contact situations is a worthy line of inquiry (Yashima et al., 2004). This insight is represented in the voluntary frequency of communication variable.
Conducting replication studies is essential for providing support for a given theory and for examining whether an original investigation can be generalised to other participants and contexts (Chun, 2012; Porte, 2013). Lightbown (2000) also emphasises that the focus of replication should not be on the individual details of a study, but on the general principles that the original study was based upon. Consequently, approximate replication is useful in generalising findings from the original study to a new setting (Porte, 2013). It is therefore assumed that a replication of Yashima et al. (2004) in an ESL environment will yield a deeper insight into the generalisability of the IP construct in contexts other than the one it was developed in.
The present study intended to investigate the interrelations of affective variables influencing L2 WTC among Korean English learners in an overseas ESL environment. By examining these relationships, the study aimed to get more insight into the IP construct when applied to an overseas ESL environment.
It followed a similar methodological approach as was carried out in Yashima et al. (2004), in that, the same hypotheses were assumed regarding L2 WTC’s predictive relationship to its antecedents. See Figure 1 for the hypothesised model.
Figure 1. Hypothesised L2 Communication Model to be Tested
Two relationships were of particular interest to the present study. The first being the relationship between IP and L2 WTC (see Figure 1). The present study was conducted in an overseas ESL context as opposed to an EFL context, which, due to daily contact with native English speakers, adds a degree of uncertainty in suggesting that IP would be predictive of L2 WTC. Secondly, L2 WTC is a predictor of L2 communicative behaviour (see Figure 1). Examining this relationship within an overseas ESL context may further affirm L2 WTC as a reliable predictor of L2 communicative behaviour (Kang, 2005). Three questions focused part one of the study:
- Is L2 WTC among Korean learners of English a predictor of voluntary Frequency of L2 Communication in an overseas ESL environment?
- Is International Posture a predictor of Korean learners of English’s WTC and voluntary Frequency of L2 Communication in an overseas ESL environment?
- What are the relationships among the variables in International Posture and L2 Motivation, L2 Motivation and Communication Confidence, Communication Confidence and L2 WTC?
It is believed that as learners move through the language learning process, the variables underpinning L2 WTC are prone to development (Baker & MacIntyre, 2003). The context of language use, length of exposure, and participation in practical authentic communication situations are strong determining factors in considering this variation (Watanabe, 2013). Although previous studies have shown that L2 WTC and its related antecedents varied significantly over the study period (e.g. MacIntyre, Baker, Clément & Donovan, 2003; House, 2004; Cao, 2006), others have demonstrated no substantial change (e.g. Watanabe, 2013). Accordingly, further analysis of the present data looked at how the length of time spent in the overseas environment influenced the development of L2 WTC and its associated variables. While this was investigated to a certain degree in Yashima et al. (2004), their study focused more on satisfaction in and adjustment to the overseas environment over a short three week period, and how those variables related to L2 WTC and its antecedents. The present study wished to look at a more developmental aspect to the variables over a longer period.
Yashima et al. (2004) examined students prior to departure and a second time at the end of three weeks abroad. It was shown that satisfaction gained from the sojourn experience was positively correlated with communication behavior; however, much of this L2 behaviour took place with host families, where the environment was conducive to and invited conversation with host family members. It was also shown that L2 anxiety was negatively correlated with a sense of adjustment to the new environment while perceived competence showed a negligible relationship. This may indicate that while L2 anxiety may wane due to increased exposure and positive feedback in the new environment, the three week study period may have been too short to capture a real sense of increased language competence on the part of the sojourners. Competence in language use may take longer to develop as learners gradually gain a better sense of their language skills over time. Examining learners within a context abroad over a longer period may give more insight into the developmental aspect of L2 WTC, its antecedents, and the relationships between these variables. Two questions were proposed:
- How does length of time residing in an L2 environment relate to the variables International Posture, L2 WTC, L2 Communication Confidence, and L2 Learning Motivation?
- How does the length of time residing in the L2 environment affect the relationships between these variables, including Frequency of L2 Communication?
The current study was aimed at Korean learners of English who were enrolled at various English language institutes throughout Sydney, Australia.
The criteria for the present study was as follows –
- Participants needed to be of Korean descent with the Korean language as their native language.
- To have been raised in South Korea all their lives.
- To be either still in university or have graduated.
- To be between 19 years old and 29 years old and either male or female.
- Students were excluded if they had lived in an English speaking country for over 4 months, or had attended an international school where the medium of education was through English.
When compared to high school students, the participant group in Yashima et al. (2004) whose ages raged between 15 and 16 years old, university study-abroad students within the criteria above can provide a broader range of effective variables and motivations to learn English. Therefore, the widening of criteria for participation may maximise the diversity of L2 WTC. In total, 117 Korean students qualified for the study – 52 male and 65 female with an average age of 24 years old. For SEM, data from the 117 participants were used with no missing values. This is within suggested parameters for reliable calculations (Hair, Black, Babin, Anderson & Tatham, 2006).
A total of nine language schools across the Sydney region took part in the study over a five week period between the months of May and June 2014. A research instrument containing attitudinal, motivational and WTC questionnaires was administered to all willing participants. Participants were instructed to indicate how long they had been in Sydney in number of months. All questionnaires were written in Korean to maximise comprehension. All participants were informed that all questionnaires were anonymous and that all questions should be answered honestly.
Materials and Modifications
The research instrument included questionnaires that were identical to ones used in Yashima et al. (2004). However, the questionnaire Approach Avoidance Tendency, which was included in Yashima et al. (2004), was not included in the present study as its items mainly referred to an EFL context. An additional item was added to the Frequency of L2 Communication questionnaire that represented L2 communication with strangers in out-of-classroom social situations. In total nine questionnaires were used in the present study. See Yashima (2002) and Yashima et al. (2004) for further details on all questionnaires. Cronbach’s alphas (α) reported are from the present study.
- Motivation Intensity (MI)
Consisting of six items measured on a 7-point Likert scale. (α = .78)
e.g. “I really try to learn English.”
- Desire to Learn English (DLE)
Consisting of six items measured on a 7-point Likert scale. (α = .61)
e.g. “I absolutely believe that English should be taught at schools.”
- Interest in International Vocation / Activities (IVA)
Consisting of six items measured on a 7-point Likert scale. (α = .61)
e.g. “I would rather stay in my hometown.”
- Interest in Foreign Affairs (IFA)
Consisting of two items measured on a 7-point Likert scale. (α = .62)
e.g. “I often read and watch news about foreign countries.”
- Intercultural Friendship Orientation (IFO)
Consisting of four items measured on a 7-point Likert scale. (α = .71)
e.g. “I’d like to make friends with foreigners.”
- Communication Anxiety (CA)
Consisting of 12 items measured on a 0% (never feel nervous) to 100% (always feel nervous) scale. (α = .85)
e.g. “Talk in English to a stranger.”
- Perceived Competence in English (PC)
Consisting of 12 items measured on a 0% (completely incompetent) to 100% (completely competent) scale. (α = .88). The same question items used in the CA questionnaire were also applied here.
- Willingness to Communicate questionnaire (WTC)
Consisting of 12 items measured on a 0% (not willing) to 100% (always willing) scale. (α = .89). The same question items used in the CA & PC questionnaires were also applied here.
- Frequency of L2 Communication (FREQ)
Consisting of six items measured on a 10-point Likert scale. (α = .78).
e.g. “I talk with friends or acquaintances outside school in English.”
Data Analysis and Findings
Structural Equation Modelling (SEM) and Goodness-of-Fit (GoF)
SEM seeks to describe relationships among a set of variables based on hypotheses and/or theory (Kim & Bentler, 2006). In keeping to the analytical procedures carried out in Yashima et al. (2004), a similar structural equation model using IBM SPSS AMOS .21 was constructed in order to address the first three research questions. All hypothesised paths in the present model are identical to Yashima et al. (2004). The model with standardised path coefficients is shown in Figure 2. The latent variable FREQ was defined by items 1, 4, 5, and 6, which represent responses that encompass voluntary L2 communication. The latent variable L2 WTC was defined by three indicator variables representing the three situations in which the WTC questionnaire was set. The latent variable IP was defined by the indicator variables IFA, IFV, and IFO. IFO was not included in analysis in Yashima et al. (2004) due to its overlapping operationally with other variables. It remained in the present model as no anomaly was detected during analysis. Remaining latent variables (Motivation in an L2 & Communication Confidence) were defined identically to Yashima et al. (2004).
Once a model has been specified, validity of the model needs to be addressed. Model validity depends on GoF, which is determined by way of model fit indices. Essentially, model fit indices compare the theory to reality as represented by the data (Hair et al, 2006). There are a number of GoF measures to assess a model (Table 1).
Table 1. Goodness-of-Fit Indices
|Model Fit Index||Suggested Cut-Off Criterion||Current Model’s Fit Indices|
|Non-Significant p value
|Significant p value
< 3 (NC = 2.326)
> .08 (RMSEA = .095)
> .90 (CFI = .901)
< .90 (GFI = .850)
Note: RMSEA – Root Mean Square Error of Approximation. CFI – Comparative Fit Index. GFI – Goodness-of-Fit. NC – Normed Chi-Squared.
Table 1 depicts five of the most common GoF indices, their suggested criterion, and the current model’s fit indices’ value. Three of the five GoF indices did not pass the suggested cut off criterion – Chi-Square, RMSEA, and GFI. However, NC and CFI were within their respective limits. Although overall this may indicate that the model does not fit the data well, CFI is among one of the more widely used and reliable GoF indices, thus suggesting that the model is a good fit for its data. Moreover, Kline (2011) states that there is no statistical gold standard in SEM analysis that should automatically lead to the rejection or retention of a model (see also Hair et al., 2006; Kim & Bentler, 2006). He further contends that fit indices provide a general rule of thumb in that they do not adequately cover the whole range of variances within a model, and to solely rely on statistical values removes subjectivity in the decision process (Huberty & Morris, 1988, as cited in Kline, 2011). Hair et al. (2006) similarly asserts that “SEM is not used to get a good fit; it is used to test theory” (p. 751). In light of this and also upon recommendation by a statistician, it was decided to move forward with the study. There were no substantial or noteworthy changes made to the model, and all hypothesised paths remained identical with the same theories being tested in Yashima et al. (2004).
Research Questions 1-3
Figure 2 depicts the structural model used to answer the first three research questions. A correction value of r = .2 indicates a significant relationship. There was neither a significant direct nor in-direct path detected between IP and L2 WTC. This is contrary to Yashima et al. (2004). All other paths in the model were significant. L2 WTC was a predictor of voluntary L2 communication but with a stronger path coefficient than in Yashima et al. (2004). A significant path was also identified between IP and voluntary L2 communication although weaker than in Yashima et al. (2004). IP was a predictor of Motivation in an L2, and Communication Confidence was a predictor of L2 WTC, both with stronger path coefficients than in Yashima et al. (2004).
Figure 2. L2 Communication Structural Equation Model.
N = 117. * P < .05. **P < .01.
Note: IVA = Interest in International Vocation / Activities. IFA = Interest in Foreign Affairs. IFO = Intercultural Friendship Orientation. CA = Communication Anxiety. PC = Perceived Competence. MI = Motivation Intensity. DLE = Desire to Learn English. WTC = Willingness to Communicate. Acqt. = Acquaintance. Item 1, Item 4, Item 5, and Item 6 = volunteer Frequency of L2 Communication items from questionnaire.
Research Questions 4 & 5
In order to test the hypothesis that time is a mitigating factor in describing L2 WTC and its associated variables, and the relationships between these variables, two groups were identified depending on the length of the time participants spent in the L2 environment. Streiner (2002) indicates that findings are generally more accurate when the scaling of continuous variables is retained. However, Streiner also explains that there are situations in which it is prudent to divide a continuous variable into a dichotomous variable. One such situation is found when the distribution of a variable is somewhat J-shaped, i.e. most of the subjects are clumped at one end, while the rest trail off in the opposite direction. The distribution of the participants’ length of time residing in the L2 environment in the present study presented such a situation and therefore necessitated the placing of participants into two respective groups (see Appendix C).
Group One (n = 57) was identified as three months or less and represented the early stages of L2 immersion. Group Two (n = 60) was identified as four months or more and represented the later stages of L2 immersion. Such grouping resulted in the present study being cross-sectional, i.e., Group One and Group Two consisted of different sets of participants. Consequently, a new variable was created in SPSS identifying Group one with a 0 and Groupe Two with a 1. This variable was termed Categorical Time (CT) and was used to examine how time spent in the L2 environment related to the variables across the two groups. Results with a positive or negative value represented Group Two or Group One respectively.
IBM SPSS Statistics 21.0 was used to conduct Pearson correlation analysis to investigate research question four, which sought to understand how length of time in the L2 environment related to the latent variables International Posture, L2 WTC, Communication Confidence, and Motivation in an L2. Table 2 shows the interaction between CT and the aforementioned latent variables.
Table 2. Categorical Time and Latent Variables
|1 Categorical Time
2 International Posture
3 L2 WTC
4 Communication Confidence
5 Motivation in an L2
* p < .05; ** p < .01
Looking at Table 2, positive values between CT and the latent variables indicate that Group Two participants registered a higher score on each variable compared to Group One participants. This means that participants who had stayed longer in the L2 environment scored higher on the variables than did the participants who had stayed a shorter time. However, all correlations are non-significant, meaning, data is inconclusive.
Table 3 depicts the interaction between CT and the indicator variables. A positive significant correlation can be seen between PC and CT. This indicates that participants who had stayed longer in the L2 environment had more PC in their speaking. A negative significant relationship between IFO and CT is also shown. This implies that participants who had stayed a shorter time in the L2 environment demonstrated a higher degree of IFO. IFO is one of the primary theoretical concepts behind the IP construct.
Table 3. Categorical Time and Indicator Variables
|1 Categorical Time
* p < .05; ** p < .01
Note: CA – Communication Anxiety. PC – Perceived Competence. IFO – Intercultural Friendship Orientation. IVA – Interest in International Vocation. IFA – Interest in Foreign Affairs. MI – Motivational Intensity. DLE – Desire to Learn English.
In regards to research question five, effect moderation via multiple regression (Glass & Hopkins, 1996) was applied in IBM SPSS. This examined if the relationship between two variables remains the same or changes across the length of time residing in the L2 environment. To achieve this, a new variable was created – independent variable x CT. The significance (p value) of this new variable would indicate that a change has occurred between the examined variables across the two groups. Table 4 shows the variables being measured and the resultant p value. For example, the dependent variable L2 WTC and the independent variable IP x CT is testing the relationship between L2 WTC and IP across the two groups.
Table 4. Testing Relationships Between Variables Across the Two Groups
|Dependent Variable||Independent Variable||p value|
|IP x CT
IP x CT
IP x CT
L2 WTC x CT
CC x CT
L2 Mot x CT
Note: L2 Mot = Motivation in an L2. Freq = Frequency of L2 communication. CC = Communication Confidence.
As shown in Table 4, there are no significant relationships. All p values are > .05. This means that the relationships between the measured variables remain the same for Group One and Group Two. For example, the relationship between IP and L2 WTC is the same for Group One and Group Two. There is no significant variation that can be attributed to time spent in the L2 environment. In other words, time spent in the L2 environment was not a factor in describing the relationships between the variables across the two groups.
Relationships Between Variables
Results will be discussed in light of Yashima et al. (2004) and other related studies. It was shown that L2 WTC was a predictor of voluntary L2 communication in and out of the classroom. This was also shown in Yashima et al. (2004). However, the path correlation in the SEM model was stronger than in Yashima et al. (2004) and in other immersion context studies (e.g. Baker & MacIntyre, 2000; Clement, Baker & MacIntyre, 2003). It can therefore be suggested that learners who are more willing to communicate will use the L2 more when in an environment that it is more conducive to do so (MacIntyre & Charos, 1996); there is increased opportunity within varied social situations to use the L2 (see also Windle, 2006). In an immersion context, although frequent, L2 use is limited to that environment, where situations for L2 communication might also be repetitive. Additionally, students who were more internationally orientated were more inclined to engage in voluntary L2 behaviour. This was also observed in Yashima et al. (2004) albeit slightly stronger. A weaker correlation here may be due to social factors in an overseas context that are not present in an EFL or immersion setting. Furthermore, similar to Yashima et al. (2004), it was also demonstrated that participants who were more interested in international affairs (IFA) had higher levels of L2 behaviour (see Appendix A – Correlation Matric for Indicator Variables). This significant relationship between these two variables found in both studies may be indicative of the contemporary global world, in that, regardless of context, L2 learners are interested in and are willing to discuss global matters.
It was hypothesised in Yashima (2002) and Yashima et al. (2004) that a more positive disposition towards the international community (IP) would result in a stronger L2 WTC with that community. Both studies show significant relationships. However, in the present study IP was not a predictor of L2 WTC, neither directly nor indirectly. For learners in an overseas L2 environment, an attitude that embodies an outwardly look and desire to interact with an international community (IP) may not be apt to describe an intercultural contact situation. A similar result was also found in Zeng (2010) where learners’ L2 WTC was not closely connected to IP in an overseas ESL environment. Cultural issues were highlighted as the main communicative hindrance. Zeng surmised that learners can feel disconnected from a local population if they are not experienced enough in how to conduct themselves within it. In typical language classes in South Korea, little emphasis is put on cultural learning and more on rote learning and the passing of exams (Kim, 2004; Liu & Park, 2012). This can result in students’ cultural learning lagging their language competence, which may lead to a lack of interest or willingness to engage with a local population. Language use cannot be separated from its context of use (Liddicoat, Papademetre, Scarino & Kohler, 2003); therefore, competent L2 communication requires competent understanding of the cultural context of use. This was observed in Windle (2006) where Korean ESL students in Canada would predominantly socialize with other Korean students – they were drawn to and more interested in the cultural familiarity of their peers. Windle stated the students lacked the appropriate skills to interact with the local population effectively, which discouraged an overall willingness to communicate. However, IFA, a variable that describes interest in international news and events was the only indicator variable to be correlated with L2 WTC (see Appendix A – Correlation Matric for Indicator Variables). Along with its correlation with FREQ, this may be again indicative of the global nature of contemporary learners. They may be willing to talk about current events that have a global multicultural audience, and may not be particularly keen to participate in specific cultural activities within the target L2 community.
This global perspective brings into focus the personal relevance English has on contemporary language learners. Through continuing global connectedness, exposure, and opportunity, English learners can cultivate and develop very personal orientated goals, which may result in them being drawn to activities that encompass a wider global context. They may be more concerned with how they relate themselves to the world rather than to specific English speaking contexts. Dörnyei’s (2001) goal-setting and self-efficacy theories would need to be considered more to understand how the global context is becoming more relevant in the lives of contemporary English learners. Additionally, Dörnyei’s (2005) Ideal L2 Self model, would also be helpful to describe the attitudes of modern language learners, especially in ESL contexts. Instead of trying to isolate desires of integration into external communities, the Ideal L2 Self model puts more focus on the learner and how they visualise their English using selves. Learners who can clearly visualize possible or ideal English-using selves are perhaps more likely to develop positive WTC strategies and engage in L2 communication (Yashima et al. 2004).
As shown in Yashima (2002) and Yashima at al. (2004), a significant indirect path was observed from Motivation in an L2 through CC to L2 WTC. This highlights that regardless of context of language use, a learner needs more than just motivation in order to be willing to communicate; he/she needs confidence. This perspective is also supported in MacIntyre et al.’s (1998) heuristic model where it is suggested that L2 motivation in Layer IV of their model has indirect relation through L2 confidence in Layer III to L2 WTC in Layer II (see Appendix B). The significant path from IP to Motivation in an L2, which was also observed in Yashima et al. (2004), indicates that learners who see the relevance of learning English in the global context are more motivated to learn it. Further, the relationship between CC, a higher order construct defined by CA and PC, and L2 WTC was the strongest in the model. A similar relationship was found in the aforementioned Yashima studies, albeit slightly weaker (see also, Clément et al., 2003; Peng & Woodrow, 2010) and further supports the inclination that PC and CA are considered two of the most immediate predictors of L2 WTC (Baker & MacIntyre, 2003).
The stronger relationship in this context may also suggest that communication in intercultural contact situations may have a more positive bearing on a learner’s confidence to speak the L2 than it does in an immersion setting, where there may be higher instances of peer evaluation. An L2 environment provides many varied contexts in which to use the L2, which excludes communication with fellow peers as might be the case in an immersion setting. Furthermore, the strong relationship between L2 WTC and PC (see Appendix A – Correlation Matric for Indicator Variables) also supports the view that language learners tend to initiate conversation based upon how they feel about their competence (McCroskey & McCroskey, 1988). Liu & Park (2012), for example, indicate that because social evaluation plays an important role in one’s self value in Korean culture, Korean L2 learners can be more forthcoming in speaking with strangers, as they are less concerned with losing face when chatting with somebody they do not know. The L2 environment, therefore, may have provided a more active atmosphere that allowed the Korean students in the present study to feel more assured and liberated with their communicating and thus could express themselves more competently. The correlation between L2 WTC and PC in the present study, an overseas context, was significantly stronger than in Yashima et al. (2004), an immersion context. This possibly highlights again how increased varied opportunities to use the L2 coupled with positive and/or active feedback through interactions with locals, who would have no reason to negatively evaluate the L2 learners, can positively affect the learners’ attitude towards their language competence. Kim (2001) underscores the importance of communicative competence as it is through competent communication that individuals learn to relate effectively and suitably to the new social environment.
Development Aspects of Variables
MacIntyre et al. (2003), proposed that the underlying variables of the L2 WTC construct tend to vary over time as language learners come upon new challenges and experiences. However, in the present study, while correlation coefficients had a preference towards Group Two, indicating slight developmental change, it was shown that there was no statistically significant change in the individual variables from Group One (< 3 months) to Group Two (> 4 months). For example, L2 WTC did not show a developmental aspect from Group One to Group Two. It was also shown that the relationships between the variables did not change across the two groups. For example, the participants who were in the L2 environment three months or less demonstrated the same relationship between IP and Motivation in an L2, as did the participants who were in the L2 environment 4 months or more. This is contrary to what has been indicated in previous studies, which highlights the variable nature of the antecedents of L2 WTC given the different experiences L2 learners go through (e.g. Yashima et al., 2004; Yashima & Zenuk-Nishide, 2008).
However, this study was cross-sectional and did not track the progression of the same group of participants from an early point to a later point. The two groups consisted of two different sets of students. In studies such as Yashima et al. (2004), and Yashima & Zenuk-Nishide (2008), the same participants were observed from an earlier point to a later point making observation of the developmental nature of variables more apparent. Given the nature of the two identified groups in the present study, a development aspect may appear less obvious.
However, IFO, an indicator of IP, showed higher scores for Group One (< 3 months). IFO is a variable that has close ties to Norton’s (2001) concept of Imagined Communities. It was developed with regard to EFL learners having a more global integrative orientation and is affiliated with IP – a construct conceptualised in an EFL context (see Yashima, 2002). EFL learners imagine situations where they might use their L2 for international communication. It is reasonable to assume that L2 learners who have been in an L2 environment for a short period of time might still exhibit tendencies similar to those of EFL learners. Therefore, it can be expected that this would be more pronounced in Group One participants rather than Group Two participants.
Additionally, PC showed higher scores for Group Two. PC describes how competent a learner perceives him/her self to be. Therefore, this higher score for Group Two may be indicative of longer time spent in the L2 environment. A language learner can gain confidence over time through everyday usage of the language, and can therefore gradually perceive him/her self to be more competent in his/her language use. In Yashima et al. (2004) it was shown that sojourners who felt more adjusted to the overseas environment exhibited less anxiety and engaged in L2 behaviour more; however, there was no correlation observed between Adjustment and PC. The sojourners were examined over a period of three weeks, and while this may be sufficient time to acclimatize and reduce anxiety levels when speaking, it may not be sufficient time to develop a sense of language competence. A learner who has been in an L2 environment a shorter length of time might still be unsure of his/her language competency due to not yet being fully acculturated to the dynamics of the new environment. Therefore, it would be safe to assume that more exposure to interaction and language usage over a longer period of time lends itself to the betterment of a learner’s sense of language competence.
The results from this study lend empirical support to the claim that the WTC construct is a reliable predictor of L2 communication behaviour in an ESL context. The results also show that motivational and communication confidence variables were related to the L2 WTC of Korean ESL learners. These relationships further confirm associations outlined in MacIntryre et al.’s (1998) WTC heuristic model. While no developmental aspects were directly observed for the variables, it is of note that IFO had scores more associated with Group One participants, and PC had scores more associated with Group Two participants. This highlights that continued varied experiences in the L2 environment can have a positive effect on a learner’s competence, and that imagined perceptions cultivated in a learner’s EFL context can be gradually reshaped by the new L2 environment – the reality of the new environment soon replaces the imagined one. How a learner interprets the new environment and its effects on his/her learning experience would be an interesting line of inquiry.
As hypothesised at the outset, the attitudinal variable IP was not a predictor of L2 WTC in an ESL context, which was contrary to Yashima et al. (2004). This result does not outrightly question the applicability of the IP construct in accounting for English learners’ willingness to engage in conversation; however, it does highlight the possible more global nature of contemporary English language learners. IFA being the only variable correlated with L2 WTC, which describes interest in international affairs, gives credence to this premise. A line of inquiry, if pursued, could investigate the more global dispositions and inclinations of language learners and how these self-perceptions influence their communicative behaviour or level of engagement in the L2 when in various contexts.
Unfortunately, for a study that was investigating how learner perception may influence language behavior, the quantitative methodological approach employed here was limited in that it could not fully capture the nature and complexity of the Korean students’ attitudes toward their learning situation. A qualitative component would have provided better insight into some of the contextual, sociocultural, and global factors influencing the participants’ language behaviors and personal motivations. Also, the participants assessed their own voluntary frequency of L2 behavior leaving it susceptible to exaggeration. Addition of an observational component would better gauge amount of L2 communication in terms of behavioural manifestation, e.g. class observation.
The findings in this study shed light on the dynamic and capricious nature of the global world in which language learners are increasingly interacting, and how such interactions and experiences can influence learner dispositions and language practices. Subsequently, future research in this area needs to be more mindful of the increasing interrelatedness of the global world, which situates language learners in dynamic international contexts. Contemporary language learners are more likely to have increased intercultural contact and experiences with people from many social backgrounds, both online and offline. More focus, therefore, ought to be put on the sociocultural context in which learners use English. Unique contextual variables present in one context and not in another that can influence language behavior need to be more closely investigated. Exchanges within local and more global contexts, especially in terms of native and other non-native speakers, should be interpreted with particular focus on the individual learner and their unique and personal attitudes, anxieties, motivations, and interests. To get a clearer picture of this paradigm and of the developmental and interactive aspects of L2 WTC, a combination of interpretive and functional approaches need to be utilized. Understanding unique individual interactions rather than trends can give better insight into contemporary L2 intercultural communication and L2 behaviour.
About the Author
Colum Ruane completed his PhD in Applied Linguistics at Macquarie University Sydney Australia. He currently holds a research role at the Australian Institute of Health Innovation where his skills in qualitative methodologies are spread across a number of projects that deal with the evaluation of health service initiatives. His main interest areas range widely and include L2 learner identity, L2 learner attitudes and self-perceptions, globalisation & cosmopolitanism, World Englishes, L2 motivations, linguistic imperialism, out-of-class L2 learning, and willingness to communicate.
Aliakbari, M., Kamangar, M. & Khany, R. (2016). Willingness to communicate in English aiming Iranian EFL students. English Language Teaching. 9(5), 33–45.
Baker, S. C. & MacIntyre, P. D. (2000). The role of gender and immersion in communication and second language orientations. Language Learning, 50(2), 311–341.
Baker, S. C. & MacIntyre, P. D. (2003). The role of gender and immersion in communication and second language orientations. Language Learning, 50(2), 65–96
Bennett, C. I. (2006). Comprehensive multicultural education: theory and practice (6th edn). Boston: Allyn & Bacon.
Cameron, D. (2015). ‘In New Zealand I feel more confidence’: The role of context in the willingness to communicate (WTC) of migrant Iranian English language learners. IJES. 15(2), 61–80.
Cao, Y. (2006). Temporal fluctuations in situational willingness to communicate in a second language classroom. New Zealand Studies in Applied Linguistics, 12(2), 1–16
Cao, Y. (2011). Investigating situational willingness to communicate within second language classrooms from an ecological perspective. System, 39(4), 468–479.
Cetinkaya, Y. (2005). Turkish college students’ willingness to communicate in English as a foreign language. (Doctoral dissertation). The Ohio State University, Columbus, OH.
Chun, D. (2012). Replication studies in CALL research. CALICO Journal, 29(4), 591–600.
Clément, R., Baker, S. C. & MacIntyre, P. D. (2003). Willingness to Communicate in a Second Language: The effects of context, norms, and vitality. Journal of Language and Social Psychology, 22(2), 190–209
De Costa, P. I. (2014). Bridging the socio-cognitive divide: Rethinking the willingness to communicate concept from a conversation-based ELF perspective. Novitas Royal: Research on Youth and Language, 8, 11–29.
Denies, K., Yashima, T. & Janssen, R (2015). Classroom versus societal willingness to communicate: Investigating French as a second language in Flanders. The Modern Language Journal, 99(4), 718–739.
Dörnyei, Z. (1990). Conceptualizing motivation in foreign language learning. Language Learning, 40, 45–78.
Dörnyei, Z. (2001). Teaching and researching motivation. Harlow: Pearson Professional Education.
Dörnyei, Z. (2005). The psychology of the language learner: Individual differences in second language acquisition. Mahwah, NJ: Lawrence Erlbaum Associates Inc.
Dufon, M. A. & Churchill, E. (2006). Language learners in study abroad contexts. Clevedon, England: Multilingual Matters.
Ellis, N.C. (2008). The dynamics of social language emergence: cycles of language use, language change, and language acquisition. The Modern Language Journal. 92(2), 232–249.
Gardner, R. (1985). Social psychology and second language learning: The role of attitudes and motivation. London: Arnold.
Gallagher, H. C. (2013). Willingness to communicate and cross-cultural adaption: L2 communication and acculturative stress as transaction. Applied Linguistics. 34(1), 53–73.
Ghonsooly, B., Khajavy, G. H. & Asadpour, S. F. (2012). Willingness to communicate in English among Iranian non-English major university students. Journal of Language and Social Psychology, 31(20), 197–211.
Glass, G. V., & Hopkins, K. D. (1996). Statistical Methods in psychology and education (3rd ed.). Needham Heights, MA: Allyn & Bacon.
Hair, J. F., Black, W. C., Babin, B. J., Anderson, R. E., & Tatham, R. L. (2006). Multivariate data analysis (6th edn.). Upper Saddle River, New Jersey: Prentice Hall.
Hashimoto, Y. (2002). Motivation and willingness to communicate as predictors of reported L2 use: The Japanese context. Second Language Studies, 20(2), 29–70.
House, A. (2004). Learner perceptions of willingness to communicate. (Unpublished master’s thesis). University of Auckland, New Zealand.
Kang, S. J. (2005). Dynamic emergence of situational willingness to communicate in a second language. System, 33(2), 277–292.
Kim, K. H. & Bentler, P. M. (2006). Data modeling: Structural equation modeling. In J. L. Green, G. Camilli & P. B. Elmore (Eds.), Complete methods in education research (pp. 161-175). Washington D. C: Lawrence Erlbaum Associates, Inc.
Kim, S. J. (2004). Exploring willingness to communicate (WTC) in English among Korean EFL (English as a foreign language) students in Korea: WTC as a predictor of success in second language acquisition. (Doctoral dissertation). Retrieved from ProQuest Dissertations & Theses Global. (UMI 305139680).
Kim, Y. Y. (2001). Becoming intercultural. Thousand Oaks, CA: Sage.
Kline, R. B. (2011). Principles and practices of structural equation modeling (3rd ed.). New York, NY: Guilford Press.
Kramsch, C. (2008). Ecological perspectives on foreign language education. Language Teaching, 41(3), 389–408.
Liddecoat, A. J., Papademetre, L., Scarino, A & M. Kohler (2003). Report on intercultural language learning. Australia government, Department of Education, Science and Training. Retrieved from http://www1.curriculum.edu.au/nalsas/pdf/intercultural.pdf
Lightbown, P. (2000). Anniversary article: Classroom SLA research and second language teaching. Applied Linguistics, 21(4), 431–462.
Liu, Y. & Park, H. (2012). A study of Korean EFL learners’ WTC and motivation. Journal of Pan-Pacific Association of Applied Linguistics, 16(2), 35–58.
MacIntyre, P. (2007). Willingness to communicate in the second language: Understanding the decision to speak as a volitional process. The Modern Language Journal, 91(4), 564–576.
MacIntyre, P. D. & Charos, C. (1996). Personality, attitudes, and effect as predictors of second language communication. Journal of Language and Social Psychology, 15(1), 3–26.
MacIntyre, P. D. & Doucette, J. (2010). Willingness to communicate and action control. System, 38(2), 161–171.
MacIntyre, P. D. & Legatto, J. J. (2011). A dynamic system approach to willingness to communicate: Developing an idiodynamic method to capture rapidly changing affect. Applied Linguistics, 32(2), 149–171.
MacIntyre, P., Clément, R., Dörnyei, Z & Noels, K. (1998). Conceptualizing willingness to communicate in a L2: A situational model of L2 confidence and affiliation. Modern Language Journal, 82, 545–562.
MacIntyre, P., Baker, S., Clément, R. & Conrod, S. (2001). Willingness to communicate, social support, and language-learning orientations of immersion students. Studies in Second Language Acquisition, 23(3), 369–388.
MacIntyre, P. D., Baker, S. C., Clément, R & Donovan, L. A. (2003). Talking in order to learn: Willingness to communicate and intensive language programs. Canadian Modern Language Review, 59(4), 589–607.
Matsuoka, R. (2006). Japanese college students’ willingness to communicate in English, (Doctoral dissertation). Retrieved from ProQuest Dissertations and Theses Global. (UMI 304982405).
McCroskey, J. C. & McCroskey, L. L. (1988). Self-report as an approach to measuring communication competence. Communication Research Reports, 5(2), 108–113.
Munezane, Y. (2013). Attitudes, affect and ideal L2 self as predictors of willingness to communicate. EUROSLA Yearbook, 13, 176–198.
Norton, B. (2001). Non-participation, imagined communities and the language classroom, In M. Breen (Ed.), Learner contributions to language learning: New directions in research (pp. 159-171). Harlow, UK: Longman.
Peng, J., & Woodrow, L. J. (2010). Willingness to communicate in English: A model in Chinese EFL classroom affordances. Language Learning, 60(4), 834–876.
Porte, G. (2013). Who needs replication? CALICO Journal. 30(1), 10–15.
Streiner, D. L. (2002). Breaking up is hard to do: The heartbreak of dichotomizing continuous data. Canadian Journal of Psychiatry, 47(3) 262-266.
Şener, S. (2014). Turkish ELT students’ willingness to communicate in English. ESL Research Journal, 3(2), 91–109.
Swain, M. (2000). The output hypothesis and beyond: Mediating acquisition through collaborative dialogue. In J. P. Lantolf (Ed.). Sociocultural theory and second language learning (pp. 97-114). Oxford: Oxford University Press.
Watanabe, M. (2013). Willingness to communicate and Japanese high school English learners. JALT Journal, 35(2), 153–171.
Windle, S. (2006). The English language acquisition process as experienced by Korean university students: A grounded theory, (Doctoral dissertation). University of Windsor, Ontario, Canada.
Yashima, T. (2002). Willingness to communicate in a second language: The Japanese EFL context. The Modern Language Journal, 86(i), 54–66.
Yashima, T. & Zenuk-Nishide, L. (2008). The impact of learning contexts on proficiency, attitudes, and L2 communication: Creating an imagined international community. System, 36, 566–585.
Yashima, T., Zenuk-Nishide, L. & Shimizu, K. (2004). The influences of attitudes and affect on willingness to communicate and second language communication. Language learning, 54(1), 119–152.
Zeng, M. (2010). Chinese students’ willingness to communicate in English in Canada (Doctoral dissertation). Retrieved from ProQuest Dissertations & Theses Global. (UMI 755053544).
Correlation Matrix for Indicator Variables
* p < .05; ** p < .01
[back to article]
MacIntyre’s et al.’s (1998) Heuristic WTC Model
Time Spend by Participants in Australia – J Shaped Distribution
[back to article]
|Copyright rests with authors. Please cite TESL-EJ appropriately.
Editor’s Note: The HTML version contains no page numbers. Please use the PDF version of this article for citations.