Keywords

1 Introduction

Why do people learn languages? The reason for each individual undoubtedly depends on a number of complex personal and environmental factors. For some, language learning is done to satisfy a strong internal desire; for others, learning a language will earn them a higher salary at work. However, one aspect of language learning is common for all learners: Context affects the how and the why of the language learning process. Ushioda (2009) stated that “[c]ontext is generally defined as an independent background variable which is theorized to influence motivation, but over which learners have no control” (p. 216) and that “[t]he cross-cultural perspective is undoubtedly an interesting and valuable line of inquiry” (p. 218). Ushioda also discussed the fact that such comparative studies go for “broad brushstrokes only, so that the notion of ‘context’ is often isomorphic with national culture” (p. 218). As Ushioda discussed in relation to the person-in-context concept, it is important to realize that culture is a dynamic entity and that each person in a specific context will react to the interactions with the elements in the environment differently. Mercer (2016) discussed an idea similar to Ushioda’s by stating: “Culture is not monolithic.” Furthering this argument, Mercer explained how “[c]ontext is not an external monolithic variable” but instead is part of the system—a system that is dynamic. Regarding the oft-discussed participants in applied linguistics research, Mercer indicated that People are not  situated in a context. Context is part of who they are. Given that context is an innate part of the language learner’s persona, the impetus of the current study is to address the question: “What is the relationship between the three psychological aspects of self in relation to the context?” We operationalized context as the country of origin; thus, to answer the research question, data were collected from participants learning languages in five different countries (i.e., China, Saudi Arabia, Sweden, Turkey, and the United States) to allow for a comparison of the roles of the psychological aspects of self across multiple contexts. The intent of this study was to look at the overall trends that exist when comparing the psychological aspects of self and the country of origin, while keeping in mind that each individual represents a distinct motivational profile and the fact that culture itself is continuously changing.

2 Literature Review

2.1 The Ideal, Ought-To, and Anti-Ought-To Selves

The motivational framework used in the current study is Dörnyei’s (2009) L2 motivational self system (L2MSS), which consists of two main aspects: (1) the psychological aspects of self (ideal and ought-to), inspired by self-discrepancy theory (Higgins, 1987) and possible selves (Markus & Nurius, 1986) and (2) the learning experience, which shapes individuals’ psychological aspects of self, based on the interactions between individuals and the context. The ideal self has a strong imagery component, as well as a promotional aspect. Someone with a strong ideal self is motivated by an internal desire for language learning. The ought-to self has a preventional aspect, and someone with a strong ought-to self is motivated by external forces with regards to language learning.

Thompson and Vásquez (2015) proposed an additional component to the possible selves aspect of the L2MSS, the anti-ought-to self, which brings together both the “I” and “other” aspects of self-discrepancy theory. This dimension has characteristics of both the ideal and ought-to selves: like the ideal self, the anti-ought-to self has a promotional aspect, but like the ought-to self, external elements are the primary motivating forces. Although the anti-ought-to self is similar to the ought-to self in terms of the importance of the environment, the anti-ought-to self reacts in the opposite way in comparison to the ought-to self. Whereas someone with a strong ought-to self succumbs to the desires of external pressures (social, familiar, or others), someone with a strong anti-ought-to self is motivated to do the opposite of what these external pressures expect or demand, which is the premise of psychological reactance (Brehm, 1966). It should be noted, however, that there is not a one-to-one correlation between the anti-ought-to self and psychological reactance, particularly when reactance is being described in a clinical setting (e.g., De las Cuevas et al., 2014). One of the similarities between reactance and the anti-ought-to self is the goal-setting nature of the two. Dowd and Wallbrown (1993) state that those with high levels of psychological reactance “tend to have a great deal of confidence in their ideas and decisions and would not easily be turned from a goal they consider important” (p. 573). Similarly, those language learners with a strong anti-ought-to self would not be willing to deviate from their goal of learning a specific language, even if it seems like a challenging endeavor and/or is being implicitly or explicitly discouraged by a specific person or general negative attitudes found in a particular context. For example, someone who chooses to study Arabic in the US context (e.g., Thompson, 2017a) or Japanese in the Chinese context (e.g., Liu & Thompson,) might have a strong anti-ought-to self because of the negative connotations of these languages in these specific contexts and/or the perceived difficulty of the specific language. Another example would be for an administrative academic advisor at the university level discouraging a student from taking a language because of schedule constraints (e.g. Thompson 2017b). The student with a strong anti-ought-to self would choose to take the language anyway, even with discouragement from the advisor.

Related to the importance of context to language learning, Thompson (2017a) suggested that the environment may position the learner as submissive to external expectations (the ought-to self) or as dominant in determining what he or she should not do in response to external pressures (the anti-ought-to self). In other words, the anti-ought-to self incorporates the “I” dimension of self-discrepancy theory (much like the ideal self), as the individual pushes back against the expectations of the “other” (as opposed to being overly influenced by them, as in the ought-to self). For the language learner’s anti-ought-to self, thus, the learners themselves would dominate over external pressures (individual = dominant). Understanding the relationship between learner and context is vital to understand language learning, as Ushioda (2009) stated: “We need to understand second language learners as people, and as people who are necessarily located in a particular cultural and historical context” (p. 216). Similarly, Thompson (2021) elaborates on the symbiotic relationship between self and context with concrete examples from language educators located in seven different countries.

2.2 The Role of Language Learning and Motivation in a Variety of Contexts

This study includes data from China, Saudi Arabia, Sweden, Turkey, and the United States, thus representing five distinct contexts with regards to language learning and language learning motivation. The following sections provide an overview of the language learning context in each country, the focus being on English in all countries except in the United States. Also included in these sections is a summary of language learning motivation studies carried out in the various contexts to provide background for the comparisons in each context.

2.2.1 The Chinese Context

In China, English is the most popular foreign language, is a mandatory subject from primary school to university and is regarded as a crucial foreign language (Hu, 2003). Its status bourgeoned after the “reforms and opening-up” (改革开放) policy was implemented in 1978 and was accelerated by the designation of Beijing to be the host of the 2008 Olympic Games (Jin & Cortazzi, 2002). So important is English study that even university students majoring in subjects other than English are required to take minimally 10% of their coursework in English. Nonetheless, Cai (2010) reported that many Chinese college students are not motivated to learn English, in part because of boredom with the test-oriented EFL teaching method lacking in creativity, and the disconnect between the effort invested in learning English and the seemingly slow improvement and low achievement of their language proficiency.

Many motivation studies in the Chinese context used Gardner’s (1985) framework of integrative and instrumental motivation (e.g., Gao et al., 2007; Qin & Wen, 2002; Zhou, 1996). One work specifically relevant to the issue of context was the study by Chen et al. (2005), who found that integrativeness was not a significant motivator for Chinese learners (although it is oftentimes an important factor for studies in Western contexts). As such, Chen et al. (2005) proposed a new factor called Chinese imperative, referring to the “requirements that are internalized within the culturally specific context” (p. 623). Wei (2007) also indicated that Chinese English language learners’ (ELL) instrumental motivation is higher than integrative motivation, which is different from their counterparts in western contexts. Liu (2007) indicated that 202 Chinese university ELL students had much higher instrumental motivation than intrinsic motivation. Similar results were also echoed in Yang et al.’s (2010) study.

More recent studies in China have used the L2MSS to investigate Chinese EFL motivation. For example, Magid (2012) did a comparative study between Chinese high school students and college students. He particularly emphasized that for the Chinese context researchers also need to consider cultural aspects such as familial and social obligations. A large-scale study by You and Dörnyei (2016) indicated that Chinese ELLs’ motivational profiles are different from the established motivational models in Western contexts in terms of regional variations of the ideal self (not for the ought-to self); the authors concluded, however, that the L2MSS is nonetheless a relevant framework to use in the Chinese context. In the specific context of Beijing, Jiang and Dewaele (2015) longitudinally investigated 88 female learners’ ideal and ought-to selves, finding that the ought-to selves fluctuated more than the ideal selves over time. Liu and Thompson (2018) with participants from three different universities in northern China found the anti-ought-to self to quantitatively emerge as a salient latent variable in an EFA, and it was (along with the ideal self), a significant distinguishing construct regarding a variety of learner variables, including the prediction of language proficiency. In the same context, Thompson and Liu (2018) examined how the three psychological aspects of self emerged differently in three different languages—English, French, and Japanese—pointing to the importance of context with regards to the psychological aspect of selves. Another L2MSS study (Zheng, 2013) highlights the importance of context in language learning. The findings indicated that the participants all created native speaker-based images for their ideal L2 selves in English, but they subsequently developed a sense of failure as non-native speakers. The findings suggest the possible conflicts between ideal selves and ought-to selves, which should be managed carefully.

2.2.2 The Saudi Context

Due to historical and geographical reasons, Saudi Arabia’s linguistic landscape is dominated by Arabic (regional dialects), and English is typically the only foreign language offered in the public education system. It is offered in middle and high schools (grades 7 through 12), with students receiving four 45-min lessons a week focusing mostly on grammar and vocabulary (Moskovsky et al., 2016). Early attempts to introduce English to learners at a younger age (e.g., offering language classes to fifth and sixth graders) were blocked by the public fear that English in elementary school would affect the teaching of the Modern Standard Arabic (Al-Thubaiti, 2014). The Ministry of Education, however, started to introduce English gradually at earlier grades, first in the sixth grade, then the fifth, and finally the fourth grade in the academic year of 2013/2014. Beyond the K-12 context, Saudi universities have a mandatory one-year program (preparatory year program, PYP) for students majoring in medicine, engineering, and computer sciences. Half of the hours in these programs are dedicated to teaching English as a foreign language. For example, preparatory year students at Qassim University have EFL instruction for  16 hours a week (for more information see PYP Study Plan, 2013). The program of study for other university majors has less focus on foreign language learning.

Several studies have examined the L2 motivation of English language learners in the Saudi context. Eusafzai (2013) conducted a study of EFL motivation among Saudi students in the PYP to account for the variability in learners’ motivation. Results showed that the strongest predictors of motivation were attitudes towards learning English and instrumentality-promotion. Daif-Allah and Alsamani (2014) also studied the English learning motivation of PYP students and found that most demotivating factors were related to difficulties in coping with classroom environments and classroom assignments.

Regarding the L2MSS, Moskovsky et al. (2016) explored whether the components of the L2MSS would be good predictors of intended learning effort and actual linguistic gains with the findings showing that the L2MSS components were good predictors of the intended learning efforts but failed to predict actual gains. The authors suggested that possible extraneous factors were mediating the relationship between the L2MSS and the actual linguistic gains. In another recent study with Saudi ESL learners in the US context (Alharbi, 2017), the anti-ought-to self stood out as the main factor that accounted for most of the explained variance of the intended learning effort. Other factors, including the ideal self, were found to be less prominent. These findings suggest that ESL learners with a strong anti-ought-to self have a stronger intended learning effort and were consequently more likely to engage in learning activities. Also, in a study that examined the motivation of English language learners from Saudi Arabia, Al-shehri (2009) explored the relationship between the imagery capacity of the learners and their L2 future selves. The results confirmed the hypothesis that there is a positive relationship between these two constructs, meaning that L2 learners with stronger visual imagery and imagination had a stronger sense of EFL ideal selves.

2.2.3 The Swedish Context

Like China and Saudi Arabia, English is the most widely used foreign language in Sweden. Different from these other contexts, though, is the fact that there are ongoing heated debates as to whether it should be referred to as a second rather than a foreign language (Hyltenstam, 2004; Josephson, 2004). English is the only mandatory language to study at school and is often introduced as early as first grade. It is encountered on a daily basis in everyday life (Medierådet, 2010): English-medium TV-shows and films are subtitled, rather than dubbed, and English is used to a great extent in advertising. The level of English proficiency among Swedes is high (National Agency for Education, 2012). One factor to explain the high proficiency is the intensive exposure to English outside of the language classroom, so-called extramural English (Sundqvist & Sylvén, 2016; Sylvén & Sundqvist, 2012). This makes English relatively unique in the Swedish context, and English is more to be seen as a societal prerequisite rather than merely a foreign language, in line with Ushioda’s (2014) discussion about the emerging view on English as a twenty-first century basic skill (see also Graddol, 2006). Motivation to study English is, in general, very high among Swedes, so studies comparing the motivation of English compared to other languages are important. In an attempt to investigate possible effects of English studies on learners’ motivation to study other foreign languages, Henry (2014) interviewed learners (N = 21) of L3 Spanish and German. The study clearly shows the need for L3 learning pedagogies to take into account the strong, and sometimes detrimental, influence L2 English may have on L3 studies because of strong preference given to English language learning.

In an effort to boost motivation for language learning of languages other than English among students in high school, in 2006 the Swedish government decided to reward studies of a third language (L3) with extra credit points, which in turn would have a positive effect on the possibilities of entering higher education. Henry (2017) investigated the effects of these extra credits on learners’ (N = 6) motivation to study French. The results of this interview-based study indicated that for three of the students, whose basic motivation was intrinsic and/or self-determined extrinsic, the extra credits had little or no impact. For the other three, however, the extra credits were the sole source of motivation for their continued French studies. It was clear, though, for these three students, that upon graduation from high school, they would not continue their study of French.

In studies related to content and language integrated learning (CLIL) and motivation in the Swedish context, Sylvén and Thompson (2015) investigated the role of L2 English in CLIL and non-CLIL contexts. It was shown that the CLIL students were significantly more motivated than the non-CLIL students on a number of motivational factors, as measured by the Motivational Factors Questionnaire (MFQ; Ryan, 2009) before they started CLIL. A repeated administration of the MFQ almost three years later showed that the only significant interaction for group (CLIL) and time was with anxiety. As had been shown previously, CLIL students exhibited lower levels of anxiety already at the start of their CLIL experience (Thompson & Sylvén, 2015), and they continued to be so after their three years of CLIL. Furthermore, it was seen that CLIL students were less ethnocentric, and, interestingly, also less fond of English than they were at the beginning of CLIL. The fact that they liked English less at the end of high school is explained by the fact that it had become an everyday occurrence, and the novelty of using English had worn off.

2.2.4 The Turkish Context

Historically, what is now the country of Turkey has always been marked by cultural and linguistic diversity, and is part of what is known as “the cradle of civilization” and as a centralized part of the former Ottoman Empire. That being said, today, most people in Turkey speak Turkish as their mother tongue, and Turkish is the official language of the country (Topbaş, 2011). English is the most popular foreign language to study in the Turkish context, with German being a far second (Eskicumali & Türedi, 2010). As in China, English is used as a link to those outside of Turkey: “On an interpersonal level, it is used as a link language for international business and for tourism while also providing a code that symbolizes modernization and elitism to the educated middle classes and those in the upper strata of the socioeconomic ladder” (Doğançay-Aktuna, 1998, p. 37). The increased focus on language education in recent years in Turkey is in part because of increased efforts to abide by the standards of the European Union and gain membership in the future (Alptekin & Tatar, 2011). Currently, compulsory English language instruction begins in the fourth grade, and there are several one-year preparatory intensive English instruction programs that are mandatory to gain admission to the competitive English-medium undergraduate programs. There are several highly-ranked English medium universities in Turkey (Selvi, 2011), but English is not used prevalently in society on a daily basis, as is the case in Sweden.

Regarding language learning motivation in the Turkish context, several studies have shown the relationship between motivation and Turkish EFL learners’ language learning achievement (e.g., Engin, 2009; Şakiroğlu & Dikilitaş, 2012). Engin (2009) found learners to have both instrumental and integrative motivation and stated that those with integrative motivation tended to put more effort into their coursework, concluding that this internal desire corresponds to the ideal self component of the L2MSS. Bektaş-Çetinkaya and Oruç (2010) made a similar comparison of Gardner’s model and the L2MSS: the reinterpretation of integrative motivation as the desire to “integrate into the international community” (ideal self) and instrumental motivation as an “investment in global market” (ought-to self) (p. 4664). The impact of various individual differences on motivation and achievement (Engin, 2009; Gömleksiz, 2010; Şakiroğlu & Dikilitaş, 2012; Şen & Şen, 2012; Üstünel & Samur, 2009) was previously a common theme of motivation studies in the Turkish context. Studies involving teachers’ perspectives of language learning motivation (Deniz, 2010; İnceçay, 2011) was also a common theme, the most recent study being Erdil-Moody (2016), who worked with English teachers in Turkey to teach motivational strategies to be implemented in the classroom by the teachers themselves. The teachers who took part in the project volunteered to participate in two workshops. The first was an overview of the theoretical aspects of the L2MSS and the second was an overview of Magid and Chan (2012), discussing the ideas of how activities formed from the tenets of the L2MSS can be successfully integrated into the language learning classroom. Her results indicated that the workshops increased the teachers’ motivational strategy use, and that the students in their English classes were overall more motivated and happier.

In a more recent study using the L2MSS, Thompson and Erdil-Moody (2016) collected data from 159 bilingual and multilingual L1 Turkish EFL learners, using two different operationalizations of multilingualism: (1) experience with more than one FL, irrespective of the amount (2) perceived positive language interaction (PPLI), an emic perspective of multilingualism, operationalized by the learner perception of the positive interactions between second or foreign languages (see Thompson, 2016a, for details). Both operationalizations of multilingualism resulted in significant group differences for the ideal self, but not for the ought-to self.

2.2.5 The US Context

Unlike the other contexts in this study, English is the primary language used on a daily basis in the US, although the 2011 Census Bureau survey indicated that about 21% of people in the US speak a language other than English at home. Whereas English is by far the most commonly studied foreign language in many countries around the world, the US does not have a default foreign language to study. Taking that into consideration, Spanish is currently by far the most commonly studied foreign language in this context, perhaps because of the large Spanish-speaking population in this context (about 62% of those who spoke a language other than English at home indicated that they were Spanish-speakers). According to the MLA database, 790,756 post-secondary students were studying Spanish in fall 2013, whereas 771,423 students were studying all other languages combined (Goldberg et al., 2015). However, there is a wide range of languages to choose from in terms of foreign language study, especially at the university level.

There are surprisingly few motivation studies carried out with learners of languages other than English (LOTEs) in the context of the US. Those studies that have been carried out oftentimes have a focus on heritage language learners (Oh & Nash, 2014; Xie, 2014), language learning technology (Cai & Zhu, 2012), or study abroad (Martinsen et al., 2014). The aforementioned Thompson and Vásquez (2015) study in which there was the inception of the anti-ought-to self, analyzed the motivation of adult learners of Chinese, German, and Italian in a US setting; however, these participants were not currently enrolled in language courses at the time of the study. Using the L2MSS framework, Thompson (2017b) investigated university language learners in the US context and found different motivational profiles for different languages studied. For example, students of Spanish had the lowest level of anti-ought-to self motivation; the explanation of the relatively prevalent nature of Spanish in this context was given for this finding. In this same context, Huensch and Thompson (2017) looked at the relationship of the psychological aspects of self and attitudes towards pronunciation. Among other findings, these authors found correlations between the ideal and anti-ought-to selves and the desire to improve pronunciation. Given that many universities in the United States require a minimum amount of language study, there are many opportunities for research in this context.

3 The Study

This study is an inquiry into a cultural comparison of the psychological aspects of self of the L2MSS, and the overarching research question is “What is the relationship between the three psychological aspects of self and context?” Data from participants from China, Saudi Arabia, Sweden, Turkey, and the United States were used for the ideal and ought-to self analyses, and data from the American and Chinese contexts were used in the anti-ought-to self analysis. The anti-ought-to self was first proposed by Thompson and Vásquez (2015), and later elaborated in several of Thompson’s later works. As this newly proposed latent variable did not exist at the beginning of the data collection period of some of these contexts, there is only anti-ought-to self data from the participants in the Chinese and US contexts. As such, an additional and separate analysis of the Chinese and U.S. data were performed. These results are an overview (i.e., broad brushstrokes) of the different contexts, which is the first step to further comparative studies of the L2MSS from a cultural standpoint.

3.1 Participants, Data Collection, and Instruments

Data for the current study were collected from 2010 to 2016 and data collection in other contexts is ongoing. This manuscript is one of several as part of a larger research project on language learning motivation, classroom anxiety, attitudes towards pronunciation, multilingualism, and a variety of other learner variables. Thus, for each context, the overall data collection instrument varied somewhat; however, each instrument contained a motivation questionnaire component. For the current analysis, only the data from the motivation questionnaire (items regarding the ideal, ought-to, and anti-ought-to selves) were used. In most cases, the entire survey took about 30 min to complete via Survey Monkey (www.surveymonkey.com). In all cases other than for data collection in the US, the entire survey was translated into to the L1 of the context—Arabic, Chinese, Swedish, and Turkish—and was presented to the participants as a bilingual questionnaire (i.e., the items were presented in both English and the first language, and the participants could use either language for the open-ended items). The forced-choice motivation items were primarily 6-point Likert-scale items (1 = strongly disagree; 6 = strongly agree). There were also open-ended items at the end of the motivation questionnaire where the participants could further elaborate on their language learning motivation, although responses to the open-ended items were not analyzed for the current study. The ideal (N = 10) and ought-to self (N = 10) items are from Dörnyei and Taguchi (2010) and the anti-ought-to self items (N = 11) from Thompson and Liu (2018)/Liu and Thompson (2018). All participants responded to ideal and ought-to self items. Participants in the Chinese and US contexts additionally responded to the anti-ought-to self items. Table 1 provides samples of the questions used; all motivation questions can be found in the appendix. The italicized words in the items were not italicized in the original questionnaire but are here to emphasize the key components of the selves.

Table 1 Sample items from the motivation questionnaire
Full size table

All participants were enrolled in university courses at the time of the study. The only requirements to participate were that the students had to be over 18 years old and had to have studied at least one foreign language before or during the data collection period. For all five contexts, there were two to three times more females than males who filled out the survey. Additionally, the majority of the participants in each context were in their late teens and early twenties; however, there were a few participants in each context who were 30 and older. The context with the lowest number of older participants was the Saudi context with only 2.5% of the participants 30 or over; the context with the highest number of participants 30 and over was the Swedish context with 23.6%. The other three contexts had approximately 10% of the participants age 30 and older.Footnote 1 Table 2 summarizes the number of participants in each context.

Table 2 Summary of participant numbers
Full size table

Two separate groups of analyses were performed for the current study, both of which included exploratory factor analyses (EFAs) for factor determination and ANOVAs for the comparison of the context. The EFAs were performed using the maximum likelihood extraction method and the oblique direct oblimin rotation method. Because anti-ought-to self data were only collected in Chinese and US contexts, two separate analyses were necessary. The first EFA was performed with data from the ideal and ought-to self items from the 1209 participants from the five contexts and the second EFA was performed with ideal, ought-to, and anti-ought-to self data from the 663 participants from the Chinese and US contexts. The 20 ideal and ought-to self items were adopted from Dörnyei and Taguchi (2010), and have been used in a number of studies. The 11 anti-ought-to self items were written based on the concept psychological reactance (Brehm, 1966). Interview data from Thompson and Vásquez (2015), focus groups with students, and conversations/suggestions at conferences were all used to inform item creation. The items were piloted with small samples with discussions of the items afterwards, per the rigorous norm of questionnaire item development (Dörnyei & Taguchi, 2010). For both EFA analyses, the internal reliability of the factors (Cronbach’s alpha) were 0.8 or higher, indicating high internal reliability of the factors. The exact Cronbach’s alpha values are indicated in the visual representations of the scree plots. The ANOVA for the five group comparison violated the assumption of homogeneity of variance for both the ideal and ought-to self (i.e., Levene’s statistic was significant); thus, Dunnett T3 post-hoc tests were carried out. For the Chinese and US group comparison for the three selves (ideal, ought-to, and anti-ought-to), the assumption of homogeneity of variance was met.

4 Results

4.1 Comparison of Five Contexts

Before doing a context comparison, an EFA with ideal and ought-to self data from participants from China, Saudi Arabia, Sweden, Turkey, and the US was performed to investigate whether or not the motivational structure of foreign language students from five countries would result in two factors of the ideal and ought-to selves. The sample size was satisfactory, as demonstrated but the Kaiser–Meyer–Olkin (KMO) value of this EFA (0.930). The items included in the final factor solution had eigenvalues greater than one and loading values of 0.3 or greater, which are the desired values for EFA analyses. In the first EFA, two items, that is, “The things I want to do in the future require me to use English” (ideal self) and “My parents believe that I must study English to be an educated person” (ought-to self) did not load with a value of 0.3 or higher and were thus removed. During the second EFA, the ought-to self item, that is, “Studying English is important to me because an educated person is supposed to be able to speak English” did not load at a 0.3 value and was eliminated. The final factor solution resulted in 17 items, explaining 66.82% of the total variance. Percentage variance and Cronbach’s alpha values are shown on the scree plot in Fig. 1. The final factor solution is in Table 3.

Fig. 1
A scree plot of eigenvalue versus factor number. In it, a line that first descends linearly with a higher slope value and then with a lower slope value is plotted. The first 2 points on the higher slope part are labeled F 1 and 2.

Scree plot with factors, variance, and Cronbach’s alphas

Full size image
Table 3 EFA for ideal and ought-to selves for China, Saudi Arabia, Sweden, Turkey, and the United States
Full size table

The items that loaded onto the ideal and ought-to self factors were averaged to obtain a factor score for each participant. Using nationality as the group variable, one-way ANOVAs were performed with the ideal and ought-to self factors, using Dunnett T3 post-hoc tests. The ANOVA results were significant for both self factors: ideal self (F[4, 1204] = 46.44, p < 0.001); ought-to self (F[4, 1204] = 54.26, p < 0.001). Table 4 provides the means and standard deviations to help with the interpretation of Fig. 2. Tables 5 and 6 provide the summary results of the post-hoc tests for the ideal self (Table 5) and the ought-to self (Table 6).

Table 4 Means and standard deviations of the ideal and ought-to selves
Full size table
Fig. 2
A multi bar graph of the means for the ideal and ought to self-case. Under each case, bars for China, Saudi Arabia, Sweden, Turkey, and the United States are plotted. In both cases, the mean for Saudi Arabia is the highest.

Pictorial representation of the ideal and ought-to self means

Full size image
Table 5 Summary post-hoc ANOVA results for ideal self
Full size table
Table 6 Summary post-hoc ANOVA results for ought-to self
Full size table

4.2 Comparison of China and the US

Before doing a context comparison, an EFA with ideal, ought-to, and anti-ought-to self data from participants from China and the United States was performed to investigate the emergent factors. The KMO value of this EFA was 0.931, indicating a satisfactory sample size. The items included in the final factor solution had eigenvalues greater than one and loading values of 0.3 or greater. After the initial EFA five items were removed because of loading value and cross-loadings: three anti-ought-to self items, that is, “I want to prove others wrong by becoming good at English,” “I want to speak English because it is not something that most people can do,” and “I am studying English because I want to stand out amongst my peers and/or colleagues,” and two ought-to self items, that is, “I study this language because close friends of mine think it is important” and “Studying English is important to me because other people will respect me more if I have knowledge of English.” The ideal and ought-to self factor results were very similar to the five country comparisons. As can be seen in Table 7, there was also an emergent anti-ought-to self in this analysis. The final factor solution resulted in 26 items, explaining 53.53% of the total variance. Percentage variance for each factor and Cronbach’s alpha values are shown on the scree plot in Fig. 3. The final factor solution is in Table 7.

Table 7 EFA for ideal, ought-to, and anti-ought-to selves for China and the United States
Full size table
Fig. 3
A scree plot of eigenvalue versus factor number. In it, a line that first descends linearly with a higher slope value and then with a lower slope value is plotted. The first 3 points on the higher slope part are labeled as follows. F 1, ideal L 2 self. F 2, ought to L 2 self. F 3, anti ought to L 2 self.

Scree plot with factors, variance and Cronbach’s alphas

Full size image

The items that loaded onto the ideal, ought-to, and anti-ought-to self factors were averaged to obtain a factor score for each participant. As the ANOVA results of the ideal and ought-to selves were discussed in the previous section, only the anti-ought-to self results are reported in this section. Using nationality as the group variable, a one-way ANOVA was performed. The ANOVA result was not significant for the anti-ought-to self when comparing language learners from China and the US: F(1, 661) = 0.011, p = 0.981. Table 8 provides the means and standard deviations to help with the interpretation of Fig. 4. The ideal and ought-to selves are included in Fig. 4 to provide a more complete picture of the comparison between the sense of selves in these contexts.

Table 8 Means and standard deviations of the anti-ought-to self for the American and Chinese contexts
Full size table
Fig. 4
A multi bar graph of the means for the ideal, ought to and anti ought to self-cases. Under each case, bars for China and the United States are plotted. In the ideal self-case, U S has a higher mean value, in the ought to self-case, China has a higher mean and in the anti ought to self-case both have the same mean of 3.5.

Pictorial representation of the ideal, ought-to, and anti-ought-to self means

Full size image

5 Discussion

5.1 Implications for the Ideal Self

When looking at the different results of the psychological aspects of self in the various contexts, a variety of cultural differences can help explain the group differences and non-differences. The ideal self of language learners in Turkey is similar (almost identical) to that of learners in China. This could be explained by the sharing of an Eastern culture, in which family responsibilities are similar (e.g., Hassan et al., 2010). At the same time, however, the ideal selves of the students in Turkey are similar to those of students in the United States. Although this result might seem contradictory, Turkey is commonly known as a place where “East meets West.” The similarity to the ideal selves of students in the US context could have to do with the “Western” thinking promulgated by Ataturk, which influences the ideals of most educated Turks. The difference between the ideal selves of the Turks and the Swedes could potentially be because of the overall lack of interaction with foreigners in the Turkish context. This rationale could also be used to explain the difference in the ideal selves of the students in the Swedish and US contexts.

Interestingly, the ideal selves of the Swedes and Saudis are similar but, ostensibly for different reasons. Swedish students have experienced many of the items firsthand because of the prevalence of English in society (“I imagine myself using English to talk to foreigners”). The Saudi students in this participant pool were mostly English majors, which is different from the participants of other nationalities. Because of this, they could also perhaps imagine these situations. Additionally, Saudi society is characterized as having a collective culture, in which individuals are pressured to conform to social norms. As pointed out by Dwairy and Achoui (2006), “the collective and authoritarian (tight or uncertainty-avoidant) culture does not appreciate autonomy but rather considers it a threat to the harmony of the collective” (p. 223). According to these authors, Saudi Arabia “has remained more collective and authoritarian and seems to be discouraging Western influences on the people’s social and political life” (p. 224). Participants in this culture feel obliged to project a “sanctioned” self-image. In this mono-cultural conservative society (Al-Zahrani, 2016), where these projected self-images are orally circulated in the active Saudi social life, individuals tend to assign greater significance to their social images and create vivid and elaborate future selves, and as cited earlier, Al-Shehri (2009) found that individuals with stronger imagery capacity were likely to have stronger ideal selves. This relationship with imagery and culture in Saudi Arabia could explain the high ideal selves of the Saudi students.

Differences between the ideal selves of learners in the Chinese and US contexts might be explained by differences in language learning foci in the classroom. For example, Cai (2010) stated that many Chinese students are frustrated by a lack of creativity in the English classroom in which there is typically more a focus on grammar exercises. While the grammar knowledge obtained is useful, these types of activities do not allow for future-based imagery with regards to the target language. The Chinese students are different from Swedish students possibly because of the same reason as with the students in the Swedish versus Saudi contexts; similarly to the Saudi context, those in the Chinese context could lack firsthand experience with the questionnaire items in real life.

5.2 Implications for the Ought-To Self

Turkish and Chinese students have almost identical mean scores for the anti-ought-to self. This result is perhaps not surprising as both cultures have strong ties with those in the environment, especially with regard to family expectations. Both Turkey and China could be described as “We” (collective) cultures versus “I” (individualistic) cultures of the West. But why would the Chinese and Turkish ought-to selves be different from those of the Saudi students? Lamb (2012) stated that the ought-to self “might be more relevant in Asian or Arab cultures where young people have shown themselves to be more susceptible to the influence of significant others” (p. 1002), which is a sentiment also discussed in Kormos and Csizér (2008) when they did not find significant results for this construct with their learners in Hungary.

The students in Swedish and US contexts have a similar ought-to self, but likely for different reasons. Relatively speaking, when comparing the US and Swedish contexts, it is reasonable to say that whereas learners of English in Sweden are more likely to need English to be successful in careers or to travel for work or pleasure, it is less likely that learners of foreign languages in the US would need those languages to be successful in those areas. The general sentiment in Sweden is often the idea that so few people speak Swedish; of course Swedes should speak other languages to communicate internationally, to work with international colleagues, or to travel to warmer climates during the long, dark winter. There is no sense of “obligation” (i.e., ought-to self) if learning English is a regular part of life.

The Saudi students’ ought-to self is significantly higher than those in all other contexts. The reason is likely multifaceted. There is a relatively new (and strong) governmental push for everyone to learn English (Al-Thubaiti, 2014). Also, many private sector jobs require their employees to speak both English and Arabic. This is similar to the situation in Sweden, but the English-learning tradition is older in Sweden than in Saudi Arabia. Perhaps in the future, English language learning will be as integrated into the Saudi culture as it is into the Swedish. Finally, enrollment in prestigious university programs in Saudi Arabia requires a functioning knowledge of English, which could also add to the pressure to learn.

5.3 Implications for the Anti-Ought-To Self

Quantitatively, there have only been two contexts investigated with regards to the anti-ought-to self: China and the US. A comparison of these contexts found no significant difference in the mean scores of the anti-ought-to self; further examining the contexts and the construct could provide some speculative answers. For example, it is possible that the anti-ought-to self exists in most, if not all, language learners because of the challenge involved in the language learning process. Further research in additional contexts is needed to test this idea, including incorporating more subtleties when looking at the data; for example, the anti-ought-to self acts differently in the models predicting proficiency for the learners in Chinese and US contexts. Liu and Thompson (2018) showed that both the ideal self and the anti-ought-to self positively predicted proficiency in language learners in China (the ought-to self has a significant inverse relationship). However, Thompson (2016b) found that only the ideal self predicts proficiency in foreign language learners in the US. These diverse results in terms of the predictive power of the anti-ought-to self illustrate the different roles it might have in these two contexts.

Alternatively, there could be a different explanation for the similar anti-ought-to self mean scores in both contexts. For example, Americans are known for operating with more of an individualistic mindset; thus, the existence of an anti-ought-to self (i.e., going against what is expected) is not surprising. In the Chinese context, one possibility is that the relatively strong ought-to self could have triggered the anti-ought-to self. In other words, Chinese students are expected to react in concordance with the expectations of family and society. In terms of language learning, it is possible that exposure to English-speaking cultures via the language could have awakened a sort of rebellious attitude in these Chinese language learners, although further exploration of this idea is needed. That being said, the overall mean scores for learners in both contexts are relatively neutral, leaning towards positive. It is possible that it is not as useful to look at overall means for the anti-ought-to self construct, as it is more susceptible to subtle environmental interactions. The results do have a concrete conclusion: The similarity found with the anti-ought-to self in the Chinese and US contexts goes again the idea of psychological reactance existing primarily in Western societies (Laurin et al., 2013).

6 Conclusions

As in any study, there are limitations of the current work. In a large-scale quantitative study, such as this one, the broad strokes, or trends, are presented and discussed. A nuanced discussion of participant differences was not included, due to space limitations. More information on the qualitative short answer data is provided in context-specific publications using this data, and readers are encouraged to refer to those publications for more information on the specific contexts and other data collected and analyzed. Additionally, it is important to remember that generalizations made from the participant data in this study should not be assumed to account for all individuals who reside in the contexts in question. With a combination of smaller-scale qualitative research that provides details about contexts and participants and large-scale quantitative data that reports on trends, a more complete picture of motivational profiles in specific contexts can be realized.

These results are not meant to be the definite answers to the roles of the psychological aspects of self in these five contexts. Understanding the relative strength of these selves in a variety of contexts would help language instructors alter instructional activities accordingly. The results related to the anti-ought-to self only included two contexts; thus, this feature in particular requires further exploration. As the anti-ought-to self is a newer “self,” understanding how the desire to go against social expectations varies from setting to setting (i.e., East versus West) is an important aspect of language learning motivation. A variety of contexts need to be more thoroughly investigated in order to better understand the functionality of the L2MSS around the world. A specific focus on motivation to learn languages other than English (LOTEs), is an area ripe for exploration, as much of the current motivation literature focuses on English as the point of inquiry. Further exploration and development of the anti-ought-to self would particularly help to understand why speakers of English as a first language choose to study the LOTEs that they do (Dörnyei & Al-Hoorie, 2017). As Ushioda (2009) pointed out, the results of the current study give us “broad brushstrokes” in terms of the idiosyncratic nature of the different contexts, but broad brushstrokes are needed to first cover the canvas before filling in the details of the painting.