L2 English Speakers’ Perception of Their English Accent: An Investigation of European and Asian Attitudes

,


Introduction
In the era of globalization, English is used as a common language more so than any other language among people with different linguacultural backgrounds.The spread of English around the globe has created distinctive English varieties (Crystal, 2019).Consequently, those who speak English as a second language (L2) now outnumber those who use English as their first language (L1) (Crystal, 2019).They also more frequently interact among themselves than with L1 English speakers (Jenkins, 2000;Seidlhofer, 2004).In such contexts, L2 English speakers employ their own English varieties that normally have traces of their native language.Another dichotomous categorization is native speakers (NSs) and non-native speakers (NNSs) referring to L1 English speakers and L2 English speakers.(Note 1) Ideologically and pedagogically, NS varieties (e.g., British and American English) are considered legitimate and NNSs should pursue them as a model (Seidlhofer, 2011).
Although it has been scholarly criticized, the binominal label has penetrated among stakeholders including English teaching professionals, learners, and policy makers (e.g., Jenkins, 2006;Seidlehofer, 2011).
Such ideological perception plays a critical role in performing communication: the dichotomy between NSs and NNSs is often referred to in order to evaluate an interlocutor's English and evaluate their own (Shibata, 2021).In particular, accent is considered an influential component of L2 speech perception (e.g., Lippi-Green, 2012).L2 English speakers might have achieved lexicogrammatically accurate forms.On the other hand, phonological features on their own varieties distinguishes speech when compared to L1 English speaker counterparts (Munro, 2003).Such accented speech often results in distinct language attitudes toward L2 English speakers (Riney, Takagi, & Inutsuka, 2005;Giles, 2006;Munro, Derwing, & Morton, 2006).Both L1 and L2 English speakers' language attitudes toward English accents have been extensively investigated: the majority have employed the verbal-guise technique (VGT) which asks participants to evaluate different recorded speech materials of accented English (McKenzie & Gilmore, 2017).Previous studies conducted with a group of Japanese learners of English (JLEs) suggest that they tend to perceive Japanese-accented English negatively as well as other non-native varieties (e.g., Chiba, Matsumoto, & Yamamoto,1995;Fraser, 2006;McKenzie, 2008aMcKenzie, , 2008b)).

Language Attitudes Towards English Accent
Fundamental language attitudes are learned and constructed in the social environment (Bohner & Wänke, 2002;Garrett, 2010).For example, we implicitly and explicitly develop our perceptions through communicating with others and observing the way they interact with different language and cultural groups in our communities (Garrett, 2010).The nature of attitudes is complex as they consist of distinct components.Cargile, Giles, Ryan, & Bradac (1994) claimed three dimensions 'cognitive', 'affective ', and 'behavioral' as described below (p. 221, italic in original): Attitudes are cognitive because they entail beliefs about the world, such as French is a useful language to know, or English people are refined.Attitudes are affective because they involve feelings towards an attitude object, such as a passion for Irish poetry, or an awful taste in the mouth of Georgians when speaking Russian.And lastly, attitudes are behavioural because they encourage certain actions, such as enrolling in a Japanese language course, or hiring a prestige accented speaker for a job.
More critically, language attitudes function as biased social and stereotypical filters when people evaluate others by inferences made from linguistic variations emerged in pronunciation, lexical use, and grammatical patterns (e.g., Riney et al., 2005;Giles, 2006;Munro et al., 2006).Such evaluative reactions can be observed in intercultural communication where L2 English users from different L1 speech communities speak their own accented English.Extensive accent perception research has been conducted with both L1 and L2 English users to investigate their language attitudes toward distinct English varieties.Regarding research conducted with JLEs, Chiba et al. (1995), one of the earlier studies and frequently cited in the literature, investigated the accent perception of 169 university students by having them rate nine male English speakers from different L1 backgrounds (three from Japan, two from America, one from Britain, and three ESL speakers from Sri Lanka, Hong Kong, and Malaysia, respectively).It was revealed that the participants preferred American and British accents, while disapproving of their own accent as well as the other NNS accents.Based on the results of an identification task in which the participants guessed the nationality of the nine speakers, they claimed that familiarity led to an endorsement of NS accents but did not help foster a positive attitude towards Japanese-accented English despite frequent exposure.
In addition, Fraser (2006) examined Japanese high school students' perception of English speakers from six different countries (England, USA, Scotland, Zimbabwe, Taiwan, and Japan) through an accent judgment task and a questionnaire with attitudinal statements.The results revealed that the participants did not highly value Japanese English, despite over 90% reporting that they perceived it as being familiar and easy to understand.Furthermore, they viewed the American accent as the model for teaching and acquisition, despite it being perceived as too fast to understand easily.Their judgments indicate that accent is evaluated along different dimensions.This is evident in McKenzie (2008aMcKenzie ( , 2008b)).His studies found that Japanese university participants favored a Japanese speaker whose English is heavily accented in terms of social attractiveness (i.e., gentle, pleasant, funny, and modest) while positively evaluating the speakers of British and American English in terms of competence (i.e., intelligent, confident, fluent, and clear).Based on such findings, he concluded that JLEs' language attitudes reflect both a social belief-that a native variety of English is prestigious and correct-and an in-group solidarity with those who also speak Japanese-accented English.Other studies have also reported that Japanese L2 English students' have negative language attitudes towards Japanese English (Ishikawa, 2017;Sasayama, 2013).A similar tendency has been reported in other L1 groups, such as Austrians (Dalton-Puffer, Boeckmann, & Hinger, 2019), Iranians (Rezaei, Khosravizadeh, & Mottaghi, 2018), and Norwegians (Rindal & Piercy, 2013).
Most previous studies on L2 English accents have focused on listeners' judgments of accented speech uttered by particular reference groups.However, since interaction is bidirectional, speakers' attitudes toward their own varieties should also affect intercultural communication by influencing their confidence, motivation and willingness to communicate in L2.Nevertheless, introspective research on L2 English speakers' perception of their own accent has seldom been performed.Furthermore, learners' attitudes toward accent often reflect historical, political, and economic situations in individual learning environments (Kang, 2015).Given this, accent perception needs to be examined by comparing multiple groups from different social contexts within a single study.On the contrary, the majority of previous studies have limited participants from a single group from one L1 background.In addition, these studies also combine multiple L1 speakers into a single group.In reality, however, English speakers routinely encounter others from different L1 backgrounds, whose accents they may not be familiar with.Thus, to obtain a global picture of language attitudes toward L2 accents, to advance scholarship and find new pedagogical benefits, L2 English speakers' perception of their own accent should be explored cross-nationally.
Accordingly, Tokumoto and Shibata (2011) investigated L2 English users' introspection by bringing their attention to and having them evaluate their own accents.They compared the evaluative responses from college students in three Asian countries: Japan, South Korea, and Malaysia.The results showed the distinctive perception of their own English varieties.They concluded that the emphasis in English instruction and socio-historical factors in each country appear to influence L2 English speakers' construction of attitudes toward a target language.Their research suggests that L2 English speakers have various schemas and stereotypes that are culturally and socially approved of with respect to native and NNS English varieties.Adopting Tokumoto & Shibata (2011), Monfared & Khatib (2018) compared Indian and Iranian teachers' perception of their own English varieties.The Indian group highly valued their local varieties of English while they favored British English; the Iranian group showed negative attitudes towards their own English variety with strong endorsement for native variety of English, in particular American accent.The authors conclude that such distinctive perceptions should be ascribed to historical and political backgrounds of the two countries: according to Kachru's (1985) three circle model, India is categorized as one of the Outer Circle nations, where nativized varieties are recognized as Indian English and local people identify themselves as native speakers of their own variant.On the contrary, Iran is regarded as an Expanding Circle nation, where English is a foreign language.

Role of English in the Surveyed Nations
The language situation is complex within the European Union (EU).In addition to 23 officially recognized languages, more than 60 indigenous regional and minority languages are spoken, and many other languages are maintained in migrant communities (European Commission, 2012).The EU encourages all citizens to learn and be skillful in at least two languages besides their native language for mutual understanding and communication.Although it is not officially stated in language and education policy, English is a primary foreign language to be learned and spoken within Europe.According to a 2012 survey conducted by European Commission, 67% of respondents in 27 surveyed countries considered it the most useful language, including, 92% of the Danish, 82% of the German, and 76% of the Austrian respondents.Such English dominance has been criticized with fear of making other foreign languages less attractive to learn, which challenges the promotion of plurilingualism claimed in EU language policy (Busse, 2017;Phillipson, 2001).
Malaysia is a linguistically and culturally pluralistic nation with three major ethnic groups-Malays, Chinese, and Indians.After the Portuguese and Dutch colonialization, the country was under British rule for about two centuries.Under the British administration, English acquired significant status as the official language and was used as the medium of instruction for the elite.However, the National Language Policy in 1967 removed the official status of English and promoted the use of Malay in official settings, including educational institutions.This reform resulted in a decline of English usage and the people were less competent in English.With the recognition of English as a necessity for cultivating human capital to compete in the era of globalization, the government put a priority on English education at primary, secondary, and tertiary levels to empower the nation's English ability (Gill, 2005;Rashid, Rahman, & Yunus, 2017).In particular, the policy of the 'Teaching of mathematics and science in English' was implemented in 2003, whereby the two subjects were taught in English.On the contrary, however, students' mastery of mathematics and science failed, and thus the policy was abolished in 2012.Currently, English is a second or foreign language, while Bahasa Malay is the national language (Rashid et al., 2017).Nevertheless, English is deeply rooted in Malaysian society, recognized as Malaysian English, which encompasses acrolect (Standard Malaysian English), mesolect (dialectal Malaysian English), and basilect (patois Malaysian English) (Hashim, 2020;Lim, 2017;Nair-Venugopal 2000).The people code-switch such varieties along with Malay or other languages in daily conversation (Omar, 2012).
Kazakhstan has been through a similar historical and political fate like that of Malaysia.Besides the two major languages, Kazakh and Russian, 124 languages are spoken by its ethnic groups, which occupy only 13.2% of the population (Aksholakova & Ismailova, 2013).In 1991 with the breakup of the Soviet Union, the Republic of Kazakhstan was founded.Since then, the government has given priorities to the Kazakh language and cultural values of Kazakhs to promote national identity (Turumbetova, 2013).According to a 2009 census, 63.1% of the population is ethnic Kazakh; a majority of them, however, speak Russian as their L1 with very limited use of L2 Kazakh (The Agency on Statistics of the Republic of Kazakhstan 2011).As such, the government requires all citizens to learn Kazakh.At the same time, the language policy recognizes English as a core language along with Kazakh and Russian, realized in the cultural program "The Trinity of languages: Kazakh, Russian, English" (Aksholakova & Ismailova, 2013).All school children and university students are instructed in Kazakh, but many of them complete their education without being skilled in the language.One factor is that many students put a priority on learning English over the national language (Lillis, 2007).
In China and Japan, English is a foreign language and a school subject.English education has been viewed as playing a crucial role in the globalization of nations.Thus, cultivating communicative competence is set as an ultimate pedagogical goal.In China, the Ministry of Education (MOE) requires English education to start from the third grade at all elementary schools.(Note 2) The MOE stipulates nine levels of English proficiency: the second level is required at graduation from elementary school and fifth at graduation from junior high school.English is also a required subject at colleges, in which all university students are required to pass the College English Test before graduating (Hu, 2005).In Japan, emphasizing English as a necessary tool for economic and political success in the era of globalization, the government has aimed to cultivate human capital with English communicative competence in Japanese people.The majority of the population learn English from the 5 th or 7 th grade (first year of junior high school) to the 12 th grade (third year of senior high school) or university.That is, many Japanese have at least six-years of experience learning English in their home country.
Despite that extensive research has been conducted on English users' language attitudes towards L1 (native) and L2 (non-native) speakers of English, attitudinal research is not sufficient regarding the attitudes that L2 English speakers have towards their own English varieties.Consequently, this study should enance understanding of the interaction between language attitudes and socio-historical contexts where L2 English speakers are situated.Moreover, this attitudinal study should yield critical insights in English language teaching to promote a paradigm shift from monolithic native speakerism to pluralistic recognition of Englishes (Shibata, 2021).To conduct this research, the methodology of the current study is explored in the following section.

Research Question
This study aims to deepen our understanding of EFL speakers' language attitudes towards English accents.In order to achieve this aim, the current study investigates what attitudes L2 English speakers from different EFL settings have towards their own accent, incorporating a wide range of nationalities.(Note 3) In light of the literature explored above and the tendency of L2 English speakers' devaluation of NNS accents and favoritism of NS accents, the following two questions are proposed: RQ1: Do EFL speakers negatively perceive their own English accent along with a preference for L1 English?RQ 2: Are there any distinctive tendencies among EFL speakers according to nations?

Participants
In total, there were 290 college student participants from seven countries.All participants had diverse majors of study such as education, engineering, science, and Japanology: 31 Austrians, 27 Germans, 48 Danish, 62 Malaysians, 57 Chinese, 44 Japanese, and 21 Kazakhstan.Their background information is summarized in Table 1.

Procedure
The study employed a 10-item questionnaire adapted from Tokumoto and Shibata (2011) (see Appendix).The participants were requested to respond to 10 statements on a 6-point scale (1 = strongly disagree; 2 = disagree; 3 = moderately disagree; 4 = moderately agree; 5 = agree; 6 = strongly agree).In order to ensure that the tasks were administrated under as identical conditions as possible, specific instructions written in English were audio-taped.The administrators of the questionnaire were asked to play back the audio-taped materials rather than reading aloud.The researcher conducted the survey in class with the Japanese participants, while a university faculty member did so in the other countries.The survey was conducted between December 2013 and July 2014.Prior to conducting it, the research was approved by an ethical committee at the university that the author is affiliated.

Data Analysis
Distribution patterns of responses to individual items were observed in percentages, followed by a binominal test to determine whether positive and negative evaluations would be significantly different in each group.In order to divide the 6 scalar responses into a binominal framework, 1, 2, and 3 points are merged as negative and 4, 5, and 6 as positive.Previous accent perception work indicates that attitude falls into distinct dimensions such as social status and solidarity (Edwards, 1999;McKenzie, 2008aMcKenzie, , 2008b).As mentioned above, language attitudes presumably influence speakers' involvement and behaviors in interaction.This study followed three dimensions (i.e., cognitive, affective, and behavioral) of attitudes from Cargile et al. (1994): Items 1 and 4 as affective (feelings towards the object), Items 3, 8, 9, and 10 as cognitive (beliefs about an attitude object), and Items 2, 5, 6, and 7 as behavioral (actions encouraged).

Results
The results are presented in order of affective, cognitive, and behavioral components in this section.

Affect for the Participant's Own Variety of English
Items 1 and 4 requested participants to judge whether they were affectively attached to their own English pronunciation.Item 1 measured the degree of confidence in their English pronunciation.As shown in Figure 1, the Japanese participants showed much less confidence than other groups, with 72.7% of negative responses, which was significantly higher than the positive responses, at p = .004.On the other hand, a total of 98.0% of the Danish responded positively, followed by 88.8% of the Germans, 87.1% of the Austrians, 85.8% from the Kazakh speakers, 85.4% from the Malaysians, and 75.4% from the Chinese, which were all found significant between the positive and negative binominal responses, at p < .0001.
Figure 2 indicates the results of Item 4 (I am happy with my accent).The patterns followed those revealed in Item 1: More negative responses were observed than positive with the Japanese group; 65.9% and.34.0%, which the binominal test found a significant difference, at p = .049.On the other hand, 82.2% of the Malaysians positively responded to the statement, followed by 70.3% of the German, 68.8% of the Danish, 63.2% of the Chinese, and 61.9% of the Kazakh.However, the binominal test revealed significant differences with the Malaysians and the Danish (p < .0001and p = .013respectively).The results show that the Japanese group was least confident in and happy with their own accent among the seven groups.

Cognitive Constituents
Items 3, 8, 9, and 10 uncovered participants' claims about their accent beliefs.Figure 3 shows the distribution patterns of Item 3 (I have a non-native accent).
Figure 3. Percentages of Evaluative Responses for Item 3: Carrying Non-native Accent Among the groups, only the Danish group denied the non-native accent on their variety with 70.9%, whereas 35.4% claimed to believe so.On the other hand, the remaining six groups accepted their non-native accent with different degrees: overall 88.6% of Japanese believed that their English had non-native accent, followed by 85.7% of Kazakh, 77.5% of Malaysians, 64.6% of Austrians, 62.9% of Germans, and 59.7% of Chinese.The binominal tests revealed significant difference between positive and negative responses with the Danish (p = .006)and three Asian groups (p < .0001for both the Malaysian and the Japanese, and p = .001withthe Kazakh).
Items 8 and 9 asked about the intelligibility of their English by NSs and NNSs of English.As shown in Figure 4 (Item 8), a combined total of 'strongly agree' and 'agree' accounted for 90.4% of Austrians, 88.8% of Germans, 93.8% of Danish, and 90.4% of Kazakhs, which were significantly different from the negative responses at p < .0001.In particular, 68.8% of the Danish strongly believed that their own accented English was intelligible to NSs.A total of 82.3% of Malaysians and 63.1% of Chinese responded with either 'agree' or 'moderately agree,' by which the binominal test found significant difference between positive and negative responses at p < .0001for the former and p = .016for the latter, respectively.On the other hand, the responses from the Japanese are indistinct with 52.3% positive and 47.7% negative.In addition, their evaluation was relatively reserved: 43.2% of moderate agreement and 29.5% of moderate disagreement.The binominal test did not find a significant difference (p = .88).Most noticeably, the Japanese group was more positive about intelligibility of their English to NNSs (70.4%) than to NSs (52.3%) in the previous item.Only 9.1% of the participants perceived their English intelligible to NSs, but it increased to 31.8% in case of NNSs.On the other hand, the Austrian, German, Danish, and Kazakh groups who responded with strong agreement in case of NSs, evaluated their own variants less intelligible to NNSs.Specficially, as shown in the table, less strong agreement and more moderate agreement compared to the response distribution for NSs.The binominal test found that all groups were significantly positive about intelligibility of their own English at p = .01for the Japanese and p < .0001for the remaining groups.
Regardless of NSs or NNSs, positive evaluations by the Japanese group tended to be lower than other groups: 52.3% with NSs and 70.4% with NNSs.The Chinese group followed with 66.6% with NSs and 75.5% with NNSs.On the other hand, around 90% of the remaining groups claimed that their own accented English should be intelligible to both NSs and NNSs (see Figures 4 and 5).
Figure 6 indicates the results of Item 10 as to whether everyone should speak English with a native-like accent in intercultural communication.Among the groups, only the Chinese group evaluated the statement with more positive (63.2%) than negative responses (36.9%), although the binominal test did not find a significant difference between binominal responses (p = .063).Noticeably, 26.3% of strong agreement was found.On the other hand, a total of 74.0% of German, 65.9% of Japanese, 62.9% of Malaysia, and 61.9% of Kazakh denied the statement.The binominal test found that the German group significantly disagreed with such a view (p = .019)and that the Japanese group tended to be more negative than positive (p= .049).Among the groups, the Austrian and Danish groups showed little inclination between positive and negative responses: 48.4% and 51.7% with the former group, and 45.9% and 54.2% with the latter group.The results indicate that the majority incline to accept non-native accented English while communicating with people from different linguacultural backgrounds.

Behavioral Judgment of the Participant's Own Variant
Items 2, 5, 6, and 7 attempted to uncover the participants' behavioral intentions in speaking English.Regarding speaking native-accented English (Item 2), a total of 88.6% of the Japanese denied the statement: among negative responses, 31.8%indicated strong disagreement, followed by 38.6% of disagreement (see Figure 7).The denial appeared with 77.4% of the Malaysian group and 71.5% of the Kazakh speakers.On the other hand, a total of 81.3% of the Danish respondents were positive with 43.8% of moderate agreement.The binominal test revealed, at p < .0001,that the Danish group was significantly more positive than negative, whereas both the Japanese and Malaysian groups were significantly more negative than positive.That is, there are significantly more Danish participants who declared that they speak native-accented English than those who did not, and vice versa with the Japanese and Malaysian.
Figure 7. Percentages of Evaluative Responses for Item 2: Speaking Native-accented English Items 5 and 6 evaluated participants' desire for maintaining their own accent and achieving a native-like accent.
As for Item 5 (I would like to keep my accent), Figure 8 shows that a total of 95.5% of Japanese respondents were negative about maintaining their accent, which the binominal test found statistically significant, at p < .0001.The Austrian and Kazakh groups responded more negatively than positively: 66.7% of negative and 33.3% of positive with the Kazakh, 64.5% of negative and 35.5% of positive with the Austrian, which was not significantly different (p= .150and p= .189respectively).On the other hand, significantly more positive responses were found with the Malaysian group (66.1% and.33.9%), at p = .015.The remaining groups' (German, Danish, and Chinese) responses did not indicate a distinctive tendency.Item 6 asked about the degree of desire to achieve a native accent.As shown in Figure 9, all groups had a desire: Among them, 98.0% of the Danish, 95.4% of the Japanese and 90.5% of the Kazakh speakers responded positively.In addition, the most frequent response was a strong agreement.Noticeably a strong desire was found with 68.8% of the Danish and 56.8% of the Japanese, and the rest indicated with over 40.0%except the Malaysians (16.1%).The binominal test found significant difference in the denial responses (p = .011with the Austrians, p = .002with the Germans, and p < .0001with the rest).
Figure 9. Percentages of Evaluative Responses for Item 6: Desire for Achieving Native Accent Item 7 examined the degree of participants' hesitation in speaking accented English.As shown in Figure 10, all groups denied hesitation: 89.6% of the Danish, 87.1% of the Malaysians, 83.9% of the Austrians, 77.8% of the Germans, 76.2% of the Kazakh speakers, 73.6% of the Chinese, and 61.3% of the Japanese.The binominal test found that there was more denial than acceptance of the statement at the significant level with all groups except the Japanese: p < .0001with the Austrians, Danish, Malaysians, and Chinese, p = .006with the Germans, and p = .027with the Kazakh speakers.It should be noted that the Austrian, German, Danish, and Kazakh groups marked strong disagreement as the most frequent response: 63.0% of the Germans, 54.2% of the Danish, 51.6% of the Austrians, and 47.6% of the Kazakh speakers.
Figure 10.Percentages of Evaluative Responses for Item 7: Reluctance of Speaking Accented English

Discussion
The survey explored EFL speakers' perception of their own accents from different countries through self-assessment questionnaire.Individual nations are diverse in terms of their socio-historical, socio-political (e.g., language policy and English education), and socio-structural (e.g., demographic situations) contexts.The evaluative responses from seven EFL nations (Austria, Germany, Denmark, Malaysia, China, Japan, and Kazakh) were compared.Table 2 shows the ranking of nations based on the degree of positiveness (a total of percentages) for each survey item in three dimensions.
The table shows that the Japanese group is ranked the lowest for 7 items out of 10.However, the ranking of items 3 and 7 needs to be interpreted with caution: item 3 was highly scored, meaning that they admitted non-native accent on their own English, and item 7, implying their reluctance of speaking accented English.In other words, the highly scored items indicate their negative perception.Thus, their lowest positiveness appears with 9 items.The first research question was whether EFL speakers negatively perceive their own English accent along with a preference for L1 English.Overall, Japanese respondents' negativity was significant when compared to other groups, whereas Danish, Kazakh, and Malaysian respondents perceived their own English more positively.On the other hand, the majority across the groups had a desire to achieve a native-like accent.In particular, the Danish and Japanese groups showed a strong aspiration, whereas the Malaysian group indicated a much lower desire.The issue will be further argued later.
Regarding the second research question, the specific tendencies among the groups are described below: (1) The Japanese participants showed strong negative attitude toward their own accent.For example, they were neither confident in nor happy with their accent (affective component with items 1 and 4).They were not willing to maintain their accented English in its current form (behavioral component with item 5).They perceived their accented English less intelligible to both native and non-native interlocutors than other groups (behavioral with items 8 and 9).
(2) The Malaysian group showed more positive attitudes on the affective components (i.e., confidence and satisfaction) of their accents compared to other groups.Their affection for their accented English appears in their willingness to maintain it and suggested little hesitation in speaking it, despite that they recognize their accented-English is non-native.
(3) All Kazakh participants claimed that their accented English is intelligible to both L1 and L2 English speakers, and the majority did not hesitate to speak their own variety while acknowledging that they did not speak native-accented English.
(4) European groups perceived their English positively.Among them, the Danish group showed strong positive attitudes toward their own variant of English.They were the only group who claimed that a non-native accent was not recognized in their variety.The majority had confidence in their English and suggested little hesitation in speaking it with a strong suggestion that L1 English speakers could understand it.
(5) The Chinese group judged their accented English less intelligible than other groups except the Japanese.They strongly claimed a belief that everyone should speak native-like accented English.
No peculiar disposition emerged within three European nations, although the positive perception was more significant with the Danish group than two other groups.On the other hand, noteworthy variances are revealed in the Asian region: both Malaysian and Kazakh groups perceive their English accent more positively, compared to Japan and China.Tokumoto and Shibata (2011) suggested that historical and political backgrounds in the societies might have impacted on the process of constructing L2 English speakers' language attitudes.Comparing three Asian countries (Japan, Malaysia, and South Korea), they found that the Japanese college students showed the strongest negativity toward their own accent, whereas the Malaysian were most positive.They claimed that their distinct trend could depend on the history, socio-political environment, and pedagogical goals of each country.
The current study provides further evidence for their claim with a wider range of EFL countries: distinctive language attitudes are interwoven with socio-historical and socio-political situations in the nation.The self-assessed perception traits will be explored within the socio-cognitive framework below.
Noticeably, Malaysian and Kazakh respondents were positive about their own varieties of English, while wishing to obtain a native-like accent.Their positive attitudes may relate to the demographic reality: both are multiethnic and multilingual nations.Yet, the social status of English is different in the two countries.Although English has lost the formal status it once had in Malaysia, it still plays an important role in the community (Crystal, 2003): It has been a lingua franca among linguistically diverse ethnic groups; 55% of the population speaks Malay, 35% Chinese (Cantonese, Hokkien, Hakka, and Mandarin as predominant dialects), and 9% Indian (Tamil by the majority) (Preshous, 2001).Through intra-ethnic use, English is indigenized as Manglish (Gill, 2005;Nair-Venugopal, 2000;Preshous, 2001).Consequently, it is reasonable to speculate that the availability and exposure to indigenized English has influenced the Malaysians' perception of native and non-native accents, which appears in their overall high positiveness: 82.2% of them are happy with their accent; 85.4% are confident in their English; 77.4% denied their enactment of speaking native-accented English; 66.1% would like to keep their accent; 87.1% denied hesitation of speaking accented English.On the other hand, English is a foreign language in Kazakh.Despite the different role of English in each country, the social structure-multiethnic and multilingual-can count as attributes to provide the people opportunities to be aware of language diversity in society.Consequently, their awareness may help them cultivate tolerance towards and acceptance of different accents.
The demographic situation in the surveyed European nations is diverse due to the influx of migrants and the impact of globalization, factors which increase the flow of people across the EU's borders (Dalton-Puffer et al., 2019;Salö, Ganuza, Hedman, & Karrebaek, 2018).Such multilingual contexts may have provided people with opportunities to be exposed to linguistic and ethnic diversity, plausibly leading to their awareness of plurilingualism (Busse, 2017).English is the most popular language in European nations (Busse, 2017;Phillipson, 2001).Among the three European nations, the Danish group had the most confidence in English pronunciation and claimed that they speak with a NS accent.Consequently, their accent might not discourage them to speak, as they further claimed that NSs should understand their own accented English.The 2012 survey conducted by European Commission reported that 68% of Europeans who are able to speak English claim to believe that they have relatively good skills (21% with very good and 47% with good) (European Commission, 2012).In particular, 44% of the Danish evaluated their English as "very good".In addition, 86% of Danish participants claimed that they spoke English well enough to hold a conversation, followed by 73% of the Austrian, and 56% of the German.In addition, according to the survey, 71% of the Danish group watched films/TV programs or listened to the radio in English, and 69% of them used the Internet in foreign languages (European Commission, 2012).(Note 5) The access to different sources should enhance exposure to authentic English (Rindal and Piercy 2013).Moreover, another study reported that the Danish are the second most proficient L2 English speakers among 72 countries according to the EF English Proficiency Index (Education First, 2015).Such statistical data validates the Danish college students in this study, and their positive evaluation of their own accent in this study.
On the other hand, L2 English learners in China and Japan may have very limited exposure to diversified Englishes and other foreign languages outside of the classroom as frequently as those in other surveyed countries.
According to the national survey conducted in 2006 [cited in Wei & Su (2016)], only 7.3% and 23.3% claimed that they used English 'often' and 'sometimes', respectively.In another survey conducted with 260 parents of primary and secondary students in Shanghai, only 23% to 33% of them knew English, yet only 15% to 24% used English when reading, watching TV, or listening to radio despite the accessibility of TV and radio programs, online or printed materials in English.In other words, most Chinese have limited exposure to authentic English in their daily lives, and their main linguistic input is L1 (NS) varieties in English lessons.Such foreign language situations may have strengthened their endorsement of native accents with a monolithic view of English (i.e., there is a single native English).Their responses to Item 10 reflect this claim: 63.2% claimed that the native-like accented English is appropriate among participants in intercultural communication regardless of L1 and L2 users, whereas the rest denied the claim.
The current study further confirmed the Japanese participants' strong negativity to their own accented English along with endorsement for the native variety of English.Their evaluative performance could be ascribed to the social and historical traits in society.Regarding the social factor, the majority claim to believe that there is no other nationality that insists on linguistic and political autonomy in its own right, although the language rights of the Ainu and Ryukyu languages have been argued for (Hendrich & Galan, 2011).Historically, Japan has seldom been threatened in terms of language choice and use.According to Suzuki (2006), the very limited experience with foreign powers might have eased the acceptability of Westerners and their culture in Japan.Furthermore, the Japanese tend to evaluate their own and other cultures in comparison with Western culture, and they simply classify them into either Western or Other (i.e., non-Western) (Kubota, 1998).Based on such a dichotomous view, many Japanese believed that all Caucasians speak English (Kubota & McKay, 2009).Consequently, the alternative judgment can enhance Japanese L2 speakers' uncritical and unconscious favoritism for the NS English and devaluation of NNS varieties of English.The inclination observed with Japanese participants in this study reflects such prejudiced monolithic views.
Note that despite such dissimilarities, a desire to acquire native-like accented English was observed with all groups.Such an inclination might have been promoted through emphasis on native speaker norms.The monolithic view of native English as correct and standard has prevailed in EFL teaching and in society (Seidlhofer, 2004;Jenkins, 2006).Seemingly, such perceptions lead to an ultimate educational goal of acquiring native-like competence.Reluctantly, the pedagogical and ideological pressure has enhanced EFL learners' obsession of native English.This cognitive notion was reflected in the participants' endorsement of NS accent in the current study.Moreover, previous studies reported a link between L2 English speakers'/learners' strong desire of achieving native accent and their tendency of evaluating less NNS and NS accented English at the perceptional level (e.g., Tokumoto & Shibata, 2011).However, the self-assessment of their own accent revealed that they do not necessarily devaluate their own English accent.That is, ELF speakers' aspiration for a native accent does not necessarily correlate with the denial of their own accented English.The gap between the perception task and the retrospective task needs to be further investigated in future.
To sum, the current study revealed that language attitudes are distinct due to macro-contextual attributes such as national history, demographic diversity, and language policy, which could influence the accessibility to other resources (e.g., media and internet) at the personal level.More critically, the study results demonstrate that it may not be adequate that L2 English learners/speakers are lumped as an NNS group as opposed to an NS group in Applied Linguistics and English language teaching fields.

Conclusion
To conclude, the seven EFL countries surveyed have a common perspective that English proficiency is perceived essential while emphasizing the need for learners to achieve communicative competence as human capital for global competition; on the other hand, the socio-historical and socio-political backgrounds in individual nations could affect L2 speakers' language attitudes toward their own accents.The multiethnic and multilingual environment (i.e., Malaysia and Kazakhstan) and similar situations (i.e., the European countries) appear to be critical to promote a greater degree of awareness of the social and geographical diversity within and between English.On the other hand, little exposure to different English varieties and a lack of experience in interacting with other L2 English speakers and speakers of other languages promotes a monolithic view of looking toward correctness in native English, often tied up with denial of their own accent.In monoethnic and monolingual EFL countries (i.e., China and Japan) (Note 6), pedagogical emphasis on native speaker norms, and language ideology, have far more substantial influence on such narrow and inflexible attitudes toward the English language, compared to plurilingual EFL contexts.
Furthermore, considering the current movement of English-medium instruction (EMI) in higher education, such distinctive perceptions may affect students' engagement in EMI classrooms.Universities around the world keenly offer EMI courses and programs to attract inbound students from abroad.Despite such governmental and institutional ambitions, EMI research has problematized insufficient English proficiency of domestic students to fully understand the content (e.g., Bradford, 2016;Macaro, 2018).More critically, in addition to their proficiency, the negative perception of their own English could also impede their activeness in EMI classroom.Given this, EMI practitioners need to be aware that students from different EFL settings bring distinctive language attitudes to EMI classrooms.As scholary mobility will be continuously accelerated and EMI implementation will advance, diversity in the EMI context should demand more cross-national research like the present study.
Finally, the results should be cautiously interpreted since the socio-contextual factors may not be directly linked to distinctive language attitudes; on the contrary, they could interweave with and mediate other factors.Another limitation is that participants' English proficiency was neither self-claimed nor measured by means of objective measurement.Plausibly, the proficiency level has influenced their attitudes towards and perception of their own accented English.Despite these shortcomings, the current cross-national study should make insightful contributions by demonstrating that historical, ideological, and demographic backgrounds better account for L2 English speakers' language attitudes toward their own accent.

Notes
Note 1.An L1 English speaker refers to those who have acquired English as their first or native language, whereas an L2 English speaker refers to those who learn English after acquiring their first or native language.Such categorization is often equated with the dichotomy of native and non-native speakers of English.In this paper, I use the former, a numerical order of languages acquired and/or learnt, to connote neutrality, while using the latter to problematizes unequal power relation between NSs and NNSs.
Note 2. There is a gap between developed cities such as Shanghai and rural regions where ethnic minorities are resident and use their own native languages.In such areas, Chinese must be taught as a second language rather than English (Wang, 2015).
Note 3. The current study is a part of large cross-national project.
Note 4. A = Austrian, G = German, D = Danish, M = Malaysian, C = Chinese, J = Japanese, K = Kazakh Note 5.The media is the least common way of foreign language use in Austria (22%) and no percentage is reported about Germany.
Note 6.In fact, China is a multiethnic and multilingual country with 55 minority groups and 120 minority languages.Such demographic diversity might be unperceptive since Han is the dominant group comprising 92% of the total population and Mandarin Chinese as a national language has been promoted nationwide, which has endangered minority languages (Wang & Phillion, 2009).

Figure 2 .
Figure 2. Percentages of Evaluative Responses for Item 4: Satisfaction with Pronunciation

Figure 4 .
Figure 4. Percentages of Evaluative Responses for Item 8: Intelligibility of Accented English to Native SpeakersFigure5shows the responses to interaction with non-native interlocutors.

Figure 5 .
Figure 5. Percentages of Evaluative Responses for Item 9: Intelligibility of Accented English to Non-native Speakers

Figure 6 .
Figure 6.Percentages of Evaluative Responses for Item 10: Native-accented English by Everyone

Figure 8 .
Figure 8. Percentages of Evaluative Responses for Item 5: Willingness to Maintain Accent

Table 2 .
Ranking of Nations According to Positive Responses