English today is a conglomerate of a vast array of different varieties of English. This linguistic diversity, captured most prominently in the World Englishes paradigm (Kachru, Reference Kachru, Quirk and Widdowson1985), poses a challenge to English language teaching (ELT) in countries where English does not have an official status (i.e. there is no codified local norm) and is learned as a foreign language, such as Armenia or Germany. Learners of English in these countries are norm-dependent on ‘standard’ Englishes spoken as a native language (Kachru, Reference Kachru, Quirk and Widdowson1985) as the models of teaching (Galloway & Rose, Reference Galloway and Rose2015: 196–198; Matsuda and Friedrich, Reference Matsuda, Friedrich and Matsuda2012: 21–22). These ‘Standard Englishes’ are abstract and idealized concepts as they are never fully realized by speakers in their ‘clearly delimited, perfectly uniform, and perfectly stable’ (Milroy, Reference Milroy2001: 543) form. However, they are powerful ideas in the minds of speakers – and learners in particular – as the models of language teaching. Standard British (StBE) and Standard American English (StAmE) and their associated prestige accents Received Pronunciation and General American traditionally serve as the models of language teaching for learners (Kirkpatrick, Reference Kirkpatrick2007: 184-189; Phillipson, Reference Phillipson1992: 136–172).Footnote 1 StBE has long been considered the global prestige accent variety but Bayard et al. (Reference Bayard, Weatherall, Gallois and Pittam2001: 41–43) hypothesize that it is gradually replaced by StAmE due to the global availability of the US media.
To counteract the strong focus on these two ‘native-speaker’ Englishes in ELT, the learners’ own variety of English can potentially serve as a more realistic model of teaching (Kirkpatrick, Reference Kirkpatrick2007: 189–193; Matsuda and Friedrich, Reference Matsuda, Friedrich and Matsuda2012: 21–22): in this case, the local ‘non-native’ varieties Armenian English (ArE) and German English (GE) could serve as models of ELT.Footnote 2 This distinction between ‘native’ and ‘non-native’ speakers and Englishes is highly prevalent in ELT (Phillipson, Reference Phillipson1992: 193–199) but fraught with problems (Galloway & Rose, Reference Galloway and Rose2015: 200–201): (1) It is problematic to clearly define who is a ‘native speaker’ and what exactly defines a ‘native speaker’; (2) the distinction between ‘native’ and ‘non-native’ implies a value judgment; and (3) the term ‘native English’ implies a certain standardness, although most people whose first language is English do not acquire a standard variety of English as their first dialect. The World Englishes paradigm with its inclusion of many institutionalized, standard Englishes provides a certain remedy to this issue but does not do away with the problematic distinction completely. We draw on the terms proposed by the World Englishes framework and summarize StBE and StAmE as English as a native language (ENL) varieties and ArE and GE as English as a foreign language (EFL) varieties. Englishes learned as a second/subsequent language are not discussed in this paper.
Kirkpatrick (Reference Kirkpatrick2007) and Matsuda and Friedrich (Reference Matsuda, Friedrich and Matsuda2012) discuss the pros and cons of using the learners’ own variety of English or an established ENL variety as the model of ELT on a theoretical level.Footnote 3 An ENL variety as a model is advantageous because these varieties are codified, well accepted, and there is ELT material readily available. Such an approach proves mainly beneficial for the American and British ELT industry but devalues ‘non-native’ ELT teachers and their varieties. A shift to a local EFL model, which could prove more realistic and attainable for learners, could empower these teachers, who have experience in learning a foreign language and command the L1 of their students. The main disadvantage of using the learners’ own EFL variety is the lacking codification of these varieties. In conclusion, Kirkpatrick (Reference Kirkpatrick2007) and Matsuda and Friedrich (Reference Matsuda, Friedrich and Matsuda2012) argue for a shift from the traditional way of ELT with its exclusive focus on StBE and StAmE to an inclusion of a wider range of Englishes.
We contribute to this discussion of models of ELT by focusing on the perspectives of speakers of English who have learnt English as a foreign language in two different countries. We investigate the attitudes of Armenian and German university students toward StAmE and StBE, the traditional default ENL models of ELT, as well as GE and ArE, the speakers’ own EFL varieties. With the results from this language attitude study, we contribute to the discussion about the hegemony of StAmE and StBE in ELT, the acceptability of EFL varieties as potential models of teaching, and by taking a comparative perspective we aim to show how the different local sociolinguistic situations of English affect the students’ attitudes. In order to establish an empirical basis, we first address the following descriptive research questions:
• How do Armenian and German students evaluate Standard British, Standard American, Armenian, and German English accents in an (indirect) verbal guise study?
• Which attitudes do they hold toward ‘native’ vs. ‘non-native’ Englishes and British vs. American English when asked directly?Footnote 4
• How do the covert and overt attitudes of Armenian and German students toward Standard British, Standard American, Armenian, and German English differ?
We use these language attitudinal insights to discuss the wider and more interpretative research question in the conclusion: what is the status of the two ENLs (StBE and StAmE) in contrast to each other and in contrast to the two EFLs (GE and ArE) as the models of ELT in Armenia and Germany?
The next section addresses previous attitude research on different Englishes with a focus on EFL contexts. We then discuss the differing sociolinguistic situations of English in Armenia and Germany. In the following sections, we present our method, informants, and the results of our study. In the conclusion, we discuss the results and finally return to the question of linguistic models of ELT.
Attitudes towards varieties of English in EFL contexts
Most attitudinal studies on varieties of English have focused on the perceptions of speakers in ENL contexts (Garrett, Reference Garrett2010: 53–69) but there is a growing field of research on the attitudes of speakers who have learned English as a foreign language (Galloway & Rose, Reference Galloway and Rose2015: 182–186). Language attitude research generally distinguishes between direct and indirect methods, which show overt and covert attitudes, respectively (Garrett, Reference Garrett2010: 37–52). In direct research designs, informants are asked directly about their linguistic preferences or beliefs about certain varieties. In indirect attitude studies, informants listen to speech samples of different varieties, which they then rate on several scaled items, such as friendliness or level of education.
In EFL contexts, there is a general preference for ENLs in contrast to EFLs, with StBE and StAmE leading the way. This linguistic deference toward ENLs is consistent across different methodological approaches and origins of learners. For example, in a direct attitude study by Jenkins (Reference Jenkins2007), English teachers from 12 different countries ranked British and American accents as first and second best. Groom (Reference Groom2012) also illustrates a clear preference for ENL over EFL Englishes as linguistic models in a direct attitude study, covering informants from 22 different European L1 backgrounds. In Dalton–Puffer, Kaltenboeck and Smit's (Reference Dalton–Puffer, Kaltenboeck and Smit1997) indirect attitude study, Austrian university students evaluated StBE and StAmE speech samples saliently more positively than Austrian English. McKenzie (Reference McKenzie2010) also used speech samples and shows that Japanese learners prefer ENLs to local Japanese English.
Despite this predominant deference toward ENL varieties among learners, there are also signs of appreciations for EFLs. Individual teachers in Jenkins’ (Reference Jenkins2007) study named their own variety as the best. While Groom's (Reference Groom2012) informants preferred ENLs, they still positively identified with their own ‘non-native’ variety. McKenzie's (Reference McKenzie2010) informants downrated the Japanese English speakers overall but a more fine-grained analysis showed that they even preferred one local speaker to the ENL speakers on items related to solidarity.
In terms of the attitudes of EFL speakers toward StAmE in contrast to StBE, a slight overall preference for StBE seems to linger on among learners of English in contrast to Bayard et al.’s (Reference Bayard, Weatherall, Gallois and Pittam2001) hypothesis of an increasing global dominance of StAmE. In indirect studies from European contexts (Carrie, Reference Carrie2017; Dalton–Puffer et al., Reference Dalton–Puffer, Kaltenboeck and Smit1997), speakers who have learned English as a foreign language rated StBE significantly more positively than StAmE on items related to social status. However, in the same studies StAmE was often preferred for speaker traits relating to solidarity. These differences tie in with the diverging attitudinal profiles of the two varieties shown via qualitative approaches (Carrie, Reference Carrie2017; Garrett, Reference Garrett, Maegaard, Gregersen, Quist and Jørgensen2009): StBE is associated with culture and prestige whereas informants associate StAmE with mass media and informality.
The sociolinguistic situations of English in Germany and Armenia
English is by far the most widely taught foreign language in Germany. ELT starts in first grade in many federal states and lasts until the end of schooling. In higher education, English takes on more and more functions and has become the medium of instruction for many degree programs. This change has been pushed by German and European educational policies for internationalization of higher education (Hilgendorf, Reference Hilgendorf2005: 58–63). The ubiquitous status of English in Germany has also been enhanced by the role of English as the dominant lingua franca in Europe even after a potential Brexit (Bolton & Davis, Reference Bolton and Davis2017). These developments have resulted in very high proficiency rates of English. Therefore, Hilgendorf (Reference Hilgendorf2005: 64) proposes that the status of English is changing from a foreign to a second or additional language.
In contrast to this long-standing and established presence of English in Germany, the status of foreign languages in Armenia is traditionally seen against the background of the country's Soviet past, when Russian functioned as a second language (Khachikyan, Reference Khachikyan, Cohen, McAlister, Rolstad and MacSwan2005). After the declaration of independence in 1991, the status of Russian has been decreasing in favor of Armenian, but also because there was a great urge for an increased inclusion of English in various spheres of Armenian society, such as education, commerce, and mass media. The establishment of the American University of Armenia in 1991 has also contributed to the strengthened status of English. Another strong factor for the increasing role of English is the large Armenian diaspora in the USA, which is one of the most politically and economically influential diaspora communities for Armenia. In addition, through initiatives such as Peace Corps or Repat Armenia, there are many native speakers of American Englishes visiting Armenia. Thus, the sociolinguistic situation of English in Armenia is distinct from Germany due to the persistent importance of Russian and the strong presence of American Englishes, including StAmE.
Methods and informants
For the assessment of attitudes toward ENL (StAmE and StBE) and EFL (GE and ArE) varieties, we used a questionnaire study, which combines an indirect verbal guise test (VGT) with direct questions. In the VGT, the informants listened to short (47–60 seconds) recordings of eight different speakers who all read the same text (Comma gets a cure). The eight speakers are gendered pairs who represent StAmE, StBE, GE, and ArE. The voice samples for StBE and StAmE are taken from the International Dialects of English ArchiveFootnote 5, whereas the speech samples of the four EFLs were recorded by the first author during fieldwork in Germany and Armenia. The speech samples are all from speakers in their early twenties, balanced for paralinguistic variation, and were chosen to represent a somewhat ‘relaxed’ version of each variety. The informants had to rate these eight speakers on 13 items, which cover the attitudinal dimensions of social status (competent, educated, intelligible, standard, knowledgeable, proper) and solidarity (global, friendly, elegant, pleasant, organized, cool, confident) with six-point scales (e.g. 1: not at all competent; 2: not competent; 3: rather not competent; 4: rather competent; 5: competent; 6: very competent).
The questionnaire also contains direct questions in the form of statements to which the informants agree/disagree with six-point rating scales. The statements address the importance of English in communication with ‘native’ and ‘non-native’ speakers, attitudes toward ‘native-like’ pronunciation, and preferences for ‘native’ or ‘non-native’ English teachers. The informants also indicated which variety they preferred as the model for ELT. Informants had to choose one variety from a list (American, Australian, British, Canadian, and New Zealand English, English spoken by Armenians/Germans, other) and were asked to openly describe a reason for their choice. In this direct part, we utilized the problematic distinction between ‘native’ and ‘non-native’ and the generalizing concepts American English (AmE) and British English (BE), as previous research (e.g. Carrie, Reference Carrie2017; Groom, Reference Groom2012; Jenkins, Reference Jenkins2007) has shown that these terms are well established ideological constructs for the informants. Thus, despite their vagueness they are useful abstractions for direct questions on the students’ language attitudes towards different Englishes.
The questionnaire was distributed to small groups of university students who studied English philology at the Yerevan State University and the Yerevan Brusov State University of Languages and Social Sciences, Armenia, and the University of Münster, Germany, in August and September/October 2015. These university students are a convenient sample, have sufficient English proficiency to do the complex questionnaire tasks, are a young and well educated part of society in both countries, and allow for a sound cross-country comparison. The majority of students from both countries aimed at becoming English language teachers and thus their attitudes as future norm-providers are particularly important for a discussion of models in ELT. The Armenian sample consists of 100 (five males; 95 females) students and the German sample of 107 students (12 males; 95 females). The two informant groups are of roughly the same size, are both biased toward female informants, and are of similar age (MeanAgeArmenians=20 years; MeanAgeGermans=21 years).
Results: Covert and overt attitudes of German and Armenian informants
Figures 1 and 2 show the results of the VGT for the Armenian and German informants, respectively (1–3: negative ratings; 4–6: positive ratings). For both groups there are salient differences in the ratings for the eight speakers. There is also some variation in the ratings along the 13 items but the pattern of speaker ratings is fairly consistent across the items. The Armenian informants rated the two American speakers the most favorably across all items, followed by the British speakers. The four EFL speakers tended to be rated less positively than the ENL speakers with the Armenian male speaker being rated the least positively. For the German informants’ ratings, there is a clear division between ENL and EFL: the British and American speakers were rated saliently more positively than the German and Armenian speakers across all items. The British female speaker leads in the ratings for most of the items.
Speaker ratings by Armenian informants
Speaker ratings by German informants
The homogenous structure of the data with regard to the different items is verified by a factor analysis, which clusters items to meaningful bundles (i.e. attitudinal dimensions): all 13 items load onto one factor (Eigenvalue 7.91; variance explained 60.88%) with a factor weight above 0.6. Based on this result we calculated an overall mean of all items for each informant. We used this overall value to analyze the differences in the ratings for the eight speakers and between the two different informant groups with a repeated measurement ANOVA. Table 1 shows the overall mean scores and standard deviations for the eight speakers in ranked order. Dotted lines indicate significant differences in the ratings based on a post hoc test of the ANOVA with Bonferroni correction.
Overall mean ratings of the eight speakers
The repeated-measurement ANOVAs, which were carried out for both groups separately (overall mean as dependent variable and speaker as predictor variable), verify the descriptive tendencies. The ANOVAs show that the ratings differ significantly for the eight speakers for both informant groups: Armenian F(5.7, 564.6) = 36.4, p < 0.001,η2p=0.27; German F(5.3, 564.8) = 51.0, p < 0.001, η2p=0.53. Post hoc test with Bonferroni correction shows that the Armenian informants rated the two American speakers significantly more positively than all other six speakers. The two British speakers were also rated significantly more positively than the four EFL speakers. For the ratings of the German informants, post hoc test shows that there is a clear separation between ENL and EFL speakers: the American and British speakers were rated significantly more positively than the German and Armenian ones.
These rating differences show that both informant groups make a clear distinction in their attitudes between ENL and EFL varieties. Furthermore, the Armenian informants also show a greater deference toward StAmE than StBE. The ratings of the German informants do not show a covert attitudinal preference for either StBE or StAmE. MANOVA (overall means for the eight speakers as dependent variables and nationality as predictor variable) shows that nationality has a significant effect on the ratings F(8; 198) = 14.7, p < 0.001: the Armenian informants rated the Armenian male speaker significantly less positively than the German informants (3.33 vs. 3.65).Footnote 6 This difference seems to indicate a certain linguistic insecurity among the Armenian students about their own English accent. Furthermore, the Armenian students also rated the British female (3.74 vs. 4.84) and male speaker (3.93 vs. 4.42) significantly less positively than the Germans.Footnote 7 Thus, the Armenian students are less inclined toward StBE than the Germans.
Table 2 shows the results for the direct attitudinal questions for the two informant groups. The German students allocate a significant higher importance to English for communication with ‘native’ and ‘non-native’ speakers than the Armenian informants. The Armenians show stronger linguistic insecurity about their own ‘non-native’ accent than the Germans as they agree significantly stronger to the importance of ‘native speaker-like pronunciation’. Despite this linguistic insecurity, both the Armenian and the German students tend to disagree similarly to the sole reliance on ‘native speakers’ as English teachers. Both groups agree strongly to the statement that local teachers can teach English grammar and speaking effectively.
Direct questions on English and ELT*
* 1: strongly disagree; 2: disagree; 3: slightly disagree, 4 slightly agree; 5: agree; 6 strongly agree
Despite these positive attitudes toward local teachers, both groups show a very strong orientation toward ENL varieties as the model of ELT. Table 3 shows the results to the questions which variety of English the informants prefer their teacher to use. Chi-square test shows that both groups do not differ significantly from each other: χ2=3.07, df = 3, p = 0.38. The German and Armenian students show a strong preference for StBE to StAmE. Other ENLs were hardly selected. Only two Armenian informants chose their own variety. For the German sample, two students did not choose a variety and one student expressed that he saw ‘no difference’ between the varieties. Thus, ENL varieties are the clearly preferred models with a salient preference for StBE.
Preferred model of ELT
In the follow-up to this selection of a teaching model, we asked the students to indicate a reason for their choice. The evaluative profiles illustrated for BE and AmE through these open-ended questions are identical for the German and Armenian students. Students from both groups who opted for BE explained their choice by alluding to the variety's high social status: they described BE as ‘well educated’, ‘formal’, ‘most sophisticated’, and ‘polite’:
(1) I preferred the British one, cause I think that it is the most beautiful and at the same time formal English (Armenian19)
Students also described it as ‘royal’ and mentioned ‘the Queen's English’ as an explanation. In addition to this cultured/social status profile, a discourse of heritage and authenticity is evident in the students’ comments: they described BE as ‘real English’, ‘right version’, ‘original’, ‘classic way of English’, and ‘pure’. Some students also explained their choice for BE by describing it as the default option for ELT at school and at university:
(2) British English sounds more educated than other types of English and should be a standard to use at university (German11)
Many Armenian and German students who chose AmE as the preferred model of teaching commented on its wider global outreach. Students described it as ‘global’ and ‘more widely used than BE’:
(3) American English is spoken more than the other languages, and also it is an international language; everybody should know it (Armenian15)
Students from both groups also associated AmE with modernity and certain informality: they described it as ‘English for nowadays’, ‘not formal’, ‘sounds cool’. Some also described it as easier to understand: for example, students wrote ‘simplified version of British English’ or ‘more comprehensible’. Many students explained their choice for AmE by referring to the strong presence of AmE in mass media:
(4) It is most common to hear, because most TV shows and movies use American English (German32)
An attitudinal perspective on linguistic models of ELT
Overall, these attitudinal results show that there is a persisting deference toward StAmE and StBE, while the speakers’ own varieties are downgraded. However, the results of the indirect and direct approach show somewhat diverging attitudinal patterns for Armenian and German university students toward ENL and EFL varieties. This final section first discusses the results of the two approaches in relation to previous language attitude research and the sociolinguistic situations of English in Armenia and Germany, and then interprets them with regard to issues of models of ELT.
On the one hand, the results of the VGT differ from previous research (Dalton–Puffer et al., Reference Dalton–Puffer, Kaltenboeck and Smit1997; McKenzie, Reference McKenzie2010; Carrie, Reference Carrie2017) as there is no distinction between social status and solidarity: both informant groups rated the eight speakers along one attitudinal dimension that encompasses aspects of both dimensions. On the other hand, the results also corroborate these earlier attitude studies in EFL contexts as there is covert linguistic deference toward ENL varieties for both informant groups. The German students value StBE and StAmE similarly, whereas the Armenians rate StBE less positively than StAmE. In addition, the Armenian students downrated their own accents in contrast to the German reference group. These differences map onto the different sociolinguistic situations of English in the two countries: the stronger presence of (St)AmE and the relative short history of English in Armenia relate to the stronger deference toward StAmE (in contrast to StBE) and the greater linguistic insecurity. English in Germany is more established but at the same time, there is still an orientation toward StBE, which is in close proximity, as well as StAmE, which is more readily available through mass media. Thus, despite the changing status of English in Germany hypothesized by Hilgendorf (Reference Hilgendorf2005) there is a clear orientation to foreign norms.
The results of the direct questions show that there is also linguistic deference toward ENLs with regard to overt language attitudes similar to Jenkins’ (Reference Jenkins2007) and Groom's (Reference Groom2012) results. In contrast to the VGT, both informant groups prefer StBE to StAmE. The results of the open-ended questions reveal that these two ENLs have distinct attitudinal profiles for both informant groups, which are in line with Garrett's (Reference Garrett, Maegaard, Gregersen, Quist and Jørgensen2009) findings: StBE holds strong value as a traditional model of teaching as it is perceived as the original English with high cultural value. StAmE is seen as a modern and more global model of ELT.
These attitudinal results allow refining Bayard et al.’s (Reference Bayard, Weatherall, Gallois and Pittam2001) prognosis of a rising global dominance of StAmE at the expense of StBE. From the Armenian and German perspective, StBE still dominates as a traditional model on an overt ideological level, whereas StAmE seems to be equaling StBE as the prestige variety on a more covert level – most likely due to its strong presence in mass media. Despite this strong orientation to British and American norms of English, Armenian and German students do accept and value local English teachers. As many of the informants will be English language teachers, they are caught in a discrepancy between ENL norms and their own status as speakers who have learned English as a foreign language.
This paper has presented the first attitude study on varieties of English in Armenia, has taken a comparative perspective with a mix of direct and indirect methods, and thus adds to the field of attitude research on varieties of English and highlights the perspective of students and future teachers of English on different potential varieties as models in ELT. The current attitude study has shown that StAmE and StBE are both very firmly established as norms on an ideological level. Thus, from an attitudinal perspective, it makes sense to keep them as benchmark varieties in ELT. According to Kirkpatrick (Reference Kirkpatrick2007) and Matsuda and Friedrich (Reference Matsuda, Friedrich and Matsuda2012) this ELT practice reinforces linguistic deference and potentially devalues Armenian and German ELT teachers and the varieties spoken by many of them and their learners. On the one hand, the persistent covert linguistic deference of both groups to ENL norms and the more pronounced linguistic insecurity among the Armenian informants about their own variety corroborates this conclusion. However, on the other hand, the results show that this does not necessarily seem to be the case for the overt perceptions of the Armenian and German students: both groups disagree that only ‘native speakers’ of English should teach English, and they view Armenian and German teachers as very suitable English educators. In addition, the almost neutral ratings of the importance of native-like pronunciation among the German informants seems to indicate some sort of easing of the pressure to adhere fully to the ENL ideals. Thus, the students do look to ENL norms but do not fully commit to a native-speaker fallacy: they have a somewhat relaxed relationship to the implementation of the ENL norms and very importantly do not necessarily perceive native speakers as the better ‘embodiment of the target and norm for learners’ (Phillipson, Reference Phillipson1992: 194) than Armenian and German teachers.
While such language attitudinal results are crucial to the discussion of models of ELT as they take the perspective of the actual users into account, they need to be treated with some caution: the attitudes of the informants have been shaped by the hegemony of StAmE and StBE in ELT and the results show that they reproduce this ideology but also adapt it to their needs. In order to tap closer into the complexities of the ideological dispositions of EFL speakers – and especially teachers and their students – future attitudinal research could use open-ended interviews to give the informants more room to express their perspectives on different models of ELT in more detail.
SIRANUSH SEYRANYAN completed a BA and an MA in Translation Studies at Yerevan State University, Armenia. She also obtained an MA in National and Transnational Studies at the University of Münster, Germany. Her research interests include English language teaching in Armenia at the levels of secondary and tertiary education. She has worked for several educational institutions in Armenia, being involved in teaching and educational management. Email: siranushss@yahoo.com
MICHAEL WESTPHAL is a post-doctoral researcher at the chair of Variation Linguistics at the University of Münster, Germany. His research interests include World Englishes, variational pragmatics, language in the media, as well as language attitudes and ideologies. His PhD thesis is about language variation on Jamaican Radio and was published at John Benjamins in 2017. In his post-doc project, he investigates pragmatic variation across World Englishes. Email: michael.westphal@wwu.de
Open access
