Varieties of English and variety-status labelling

One of the major achievements of Braj Kachru's (Reference Kachru1991) ‘liberation linguistics’ is that it has squarely placed Outer Circle varieties such as Indian English, Nigerian English and Singaporean English on a par with Inner Circle varieties such as American English and British English – in the face of negative attitudes, ranging ‘from amused condescension to racist stereotyping’ (Bruthiaux, Reference Bruthiaux2003: 160). Following in Kachru's footsteps, many scholars have demonstrated that these Outer Circle Englishes are legitimate varieties of English, with distinct characteristics and with growing numbers of native speakers (e.g. Deterding, Reference Deterding2007; Jowitt, Reference Jowitt1991; Sailaja, Reference Sailaja2009). Indeed these Outer Circle English varieties are increasingly used, in respective countries, not only as the major or default medium of communication but also in the context of important domains such as education, media, government, literature and popular culture. The Kachruvian perspective has also given rise to the ‘egalitarian’ view that Inner Circle English speakers are no longer the only ones who can lay claim to the ownership of English. Outer Circle English speakers are now thought to be as much custodians of English as Inner Circle English speakers are.

One obvious way of recognizing Outer Circle Englishes as legitimate varieties of English is to give them unique identification labels (i.e. based on names of countries or adjective forms thereof), just as Inner Circle Englishes come with their own. For instance, the ‘new’ variety of English spoken in Singapore is called Singapore English or Singaporean English, just as the ‘old’ variety of English spoken in the US is called US English or American English. This labelling practice not only identifies different varieties of English uniquely but also makes it easy to regard Outer Circle Englishes as equal to Inner Circle Englishes in terms of legitimacy, status and suchlike. Unfortunately, the labelling practice has not stopped there, but has extended itself to learner varieties of English (i.e. Learner Englishes) attested in the Expanding Circle countries, e.g. China, Japan, South Korea (Korea hereafter) and non-Anglophone Europe. In Expanding Circle countries, English is generally taught as a school subject – that is, as a foreign language – and used in very restricted domains, typically with foreigners. Nonetheless some scholars are happy to speak of Chinese English (Ahn, Reference Ahn2015; Bolton, Reference Bolton2003; Qiong, Reference Qiong2004; Xu, Reference Xu and Kirkpatrick2010), Dutch English (Edwards, Reference Edwards2010), Japanese English (Ahn, Reference Ahn2015; Stanlaw, Reference Stanlaw2004) and Korean English (Ahn, Reference Ahn2014 & Reference Ahn2015; Hadikin, Reference Hadikin2014; K.-J. Park, Reference Park, Murata and Jenkins2009; Shim, Reference Shim1999). Regardless of their true variety status, the fact that these Learner Englishes have been given identification labels may engender and encourage perceptions of their uniqueness. For instance, by calling it ‘Korean English’ one gives English in Korea a unique identity as if it existed as a functional variety of English, especially because Inner and Outer Circle Englishes are labelled the same way, e.g. American English or Singaporean English. However, English in Korea should not be given such an identification label, because no such variety of English exists in Korea. In fact, ‘Korean English’, to borrow Görlach's (Reference Görlach and Görlach2002: 151) words, ‘is little more than a catchphrase’ – perhaps inspired by Kachruvian liberation linguistics. Some may argue that as long as we understand that we are not talking about New English varieties such as Singaporean English, it does not really matter whether we give Learner Englishes such identification labels. However, we must take issue with this kind of unrigorous labelling practice and demand that it should be discontinued, because it may encourage, if not promote, perceptions, especially among non-specialists or outsiders, that we are dealing with new, instead of learner, varieties of English.

The purpose of this article is to make a plea for the label ‘X English’ to be reserved for New Englishes – and also for the Inner Circle Englishes – and the label ‘English in X’ to be used for Learner Englishes. I will focus on English in Korea as an example of illustration, as I am most familiar with it. But the point to be made here applies to other incorrectly labelled learner varieties of English. Note that I make this plea as an outsider who draws upon World Englishes research for his own work (e.g. language policy) but finds himself having to remind his readers that English in Korea is not a New English variety (e.g. Song, Reference Song2012: 19). The remainder of the article proceeds as follows. First, I will discuss two possible reasons why the distinction between New English and Learner English has not always been maintained as vigorously as it should. Second, I will review a few instances of the label ‘Korean English’ and show that the use of the label is unjustifiable for the learner variety of English in Korea. Third, I will demonstrate that the ‘evidence’ adduced in support of English in Korea as a New English variety does not exit or cannot be accepted as evidence. Fourth, I will propose a simple way of coding New English, Learner English and Hybrid English so that they can be clearly distinguished from each other. Finally, I will bring the article to a close with a brief conclusion.

New English vs. Learner English

The need to distinguish New English and Learner English has been emphasized by a number of scholars (e.g. Görlach, Reference Görlach and Görlach2002; Mollin, Reference Mollin2007; Buschfeld, Reference Buschfeld2013). For instance, Mollin (Reference Mollin2007) provides specific criteria for recognizing X as a New English, as opposed to a Learner English, and she applies them to Euro-English only to conclude that it does not qualify as a New English – Euro-English, as Mullin (2007: 182–183) concludes, ‘remains situated in Kachru's Expanding Circle in which English is used as a foreign language only’. Buschfeld (Reference Buschfeld2013: 60–69) makes a similar attempt to distinguish between English in Cyprus (= a Learner English) and Cyprus English (= a New English) by using Mollin's criteria in conjunction with one additional criterion (i.e. natural vs. instruction-based acquisition). Notwithstanding these attempts to maintain the distinction between New English and Learner English, the incorrect labelling practice persists, as I will demonstrate below with particular reference to English in Korea.

There may be multiple reasons for this unfortunate state of affairs. For instance, Bruthiaux (Reference Bruthiaux2003: 167–168) points to the lack of clarity in defining the concept of the Expanding Circle: it is not clear whether ‘the concept is meant to cover countries, country-based varieties, speakers, or non- (or barely-) speaking learners’. Bruthiaux (Reference Bruthiaux2003: 168) avers that this lack of clarity has resulted in ‘“me-too” calls […] for additional varieties to be admitted to the ever-expanding family of [N]ew Englishes’. One other reason may be that the distinction between Outer Circle English (= New English) and Expanding Circle English (= Learner English) is not as clear-cut as the Kachruvian Three-Circle model would seem to suggest. Recent research has indeed demonstrated that the distinction in question is a fuzzy one and should be regarded as a continuum (e.g. Nesselhauf, Reference Nesselhauf2009; Hundt & Vogel, Reference Hundt, Vogel, Mukherjee and Hundt2011; Buschfeld, Reference Buschfeld2013). For instance, Buschfeld (Reference Buschfeld2013) demonstrates that English used in Cyprus does not fit into either variety category and should instead be considered to be a hybrid of a New English and a Learner English. In view of the fuzzy nature of this distinction, it may be somewhat difficult to refrain from applying to Expanding Circle Englishes the same kind of label used for Outer Circle Englishes, when the difference between these two variety types turns out to be a matter of degree. Nonetheless we should avoid using inaccurate labels such as ‘Korean English’, not least in order to obviate misrepresentation and/or misinterpretation. ‘Korean English’ is a misnomer, because no such New English variety exists, not even as a hybrid kind of English. For Learner Englishes, the label ‘English in X’ must instead be employed.

‘Korean English? You mean, Konglish?‘Footnote ¹

There seem to be at least three different ways in which researchers use the label ‘Korean English’:

• (1) some use the label to imply that English in Korea will eventually become a New English (Hadikin, Reference Hadikin2014; K.-J. Park, Reference Park, Murata and Jenkins2009);

• (2) some use the label and then ask others to define it (Ahn, Reference Ahn2014); and

• (3) some use the label and point to EFL textbooks as evidence of its validity (Shim, Reference Shim1999).

Each of these uses will now be discussed in turn.

First, Hadikin (Reference Hadikin2014: 9) decides to call English in Korea ‘Korean English’ by analogy with Indian English or Singaporean English, arguing that this labelling decision ‘leads one away from the idea that Korean English is a problem to be fixed and more towards the idea that the English spoken in Korea […] can be seen as a legitimate variety’. In fact, Hadikin (Reference Hadikin2014: 16) goes so far as to claim English in South Korea to be a ‘New English’, not ‘just bad English’. Indeed his book is entitled: Korean English: A corpus-driven study of a new English. Hadikin (Reference Hadikin2014: 16) seems to justify his use of ‘Korean English’, because ‘this is the way [he] see[s] the future of’ English in Korea. What this means is that English in Korean is not a New English, but Hadikin still chooses ‘Korean English’ over ‘English in Korea’ because he believes that it will eventually become a New English. This is most unusual reasoning. It is almost akin to calling a tadpole a frog, because a tadpole will eventually turn into a frog. The difference here is that there is every possibility that English in Korea will never evolve into a New English variety, as can be seen in the growing numbers of Koreans who are now switching from English to Chinese as their most important foreign language.

We turn now to Ahn (Reference Ahn2014), who adopts the label ‘Korean English’, while acknowledging that English in Korea is an Expanding Circle variety. In point of fact, she (2014: 214) suggests that it is ‘still situated between foundation and exonormative stabilization’, the first two stages in Schneider's (2007) five-stage evolutionary model of (postcolonial) New Englishes.Footnote ² Schneider's model is designed to deal with postcolonial New Englishes such as Singaporean English. But let us set aside the issue of Korea being a country never colonized by an English-speaking country. The point is that Ahn does not seem to be overly concerned about her use of ‘Korean English’, in spite of the fact that Schneider (Reference Schneider2007: 50) himself highlights the need to distinguish between ‘English in X’ (= Learner English) and ‘X English’ (= New English). Having decided to call English in Korea ‘Korean English’, Ahn proceeds to investigate native and non-native English teachers’ attitudes towards English in Korea without explaining to her interviewees what she means by ‘Korean English’. In fact, Ahn's interview begins with the question ‘Have you heard of Korean English?’ and includes subsequent questions such as ‘How would you define Korean English and Konglish?’ (Ahn, Reference Ahn2014: 203). Note that Konglish is a term that is used, usually derogatively, to refer to Korean expressions that make use of English words or phrases – sometimes in conjunction of Korean words or phrases – in an idiosyncratic manner to the effect that their meanings differ greatly or radically from the original meanings in English.Footnote ³ So much so that Konglish ‘might rather be conceptualized as a sub–variety of Korean, in the form of words and phrases’ (Lawrence, Reference Lawrence2012: 73). For example, eye shopping (window shopping), hostess (barmaid), talent (celebrity) and menbwung (‘brain fade’, based on English mental and Korean bwungkoy ‘collapse’) are Konglish expressions. As such, Konglish is part and parcel of Koreans’ linguistic repertoire. By calling English in Korea ‘Korean English’, however, Ahn has effectively identified it as a unique variety of English. Korean English does not exist, however. Thus it comes as no surprise that ‘the overwhelming majority of [Ahn's] interviewees interpreted the term Ko[rean]E[nglish] as Konglish’ (Ahn, Reference Ahn2014: 203). In this context, one of her interviewees is worth quoting:

Nonetheless Ahn (Reference Ahn2014) refers to English in Korea consistently as ‘Korean English’ throughout her article.

Thirdly and finally, we turn to the case of Shim (Reference Shim1999), who adduces evidence for ‘Korean English’ from EFL textbooks used in Korea. While considering Korea to be a member of the Expanding Circle, with no institutionalized variety of English used for intranational communication, Shim (Reference Shim1999: 256) claims that Korean English exists in codified form, namely in the context of EFL textbooks used in Korean schools, to the effect that codified Korean English ‘now serves as the endonormative standard for Korean English education.’ Recall that Bruthiaux (Reference Bruthiaux2003: 168) warns against this kind of ‘me–too’ call for additional varieties of English to be added to the ever–expanding family of New Englishes. In fact, he cites Shim's claim as one such ‘me–too’ call. Bruthiaux's (Reference Bruthiaux2003: 171) criticism bears quoting at length:

What counts as evidence for Korean English?

Buschfeld (Reference Buschfeld2013: 4) laments the fact that most World Englishes ‘studies seem to presuppose variety status for the type of English under investigation on the basis of impressionistic observations […] and hence do not ask the variety status question at all [emphasis added].’ This is not the place to review criteria that can be used to distinguish between New English and Learner English. However, it is worth mentioning one such criterion here, because Buschfeld's assessment applies to the previous studies on English in Korea. Mollin (Reference Mollin2007: 170) provides a set of three developmental criteria for regarding X as a New English, as opposed to a Learner English: Expansion, Nativization and Institutionalization. I will focus on Expansion, the initial stage of development, as it is presupposed by the other two criteria. By Expansion is meant that English must first be used in many different functional domains in the community.Footnote ⁴ In other words, English must significantly occupy different domains of interpersonal communication within the country (Mollin, Reference Mollin2007: 170). Mollin further points to widespread competence in English as a prerequisite for Expansion. Needless to say, unless most people in the country have competence in the language, English is not going to be used in many functional domains, because speakers are unable to express themselves in it. Does English in Korea meet this criterion for New English? The answer is no, and I will explain why by evaluating the ‘evidence’ alluded to by those who have promoted the incorrect label of ‘Korean English’.

Shim (Reference Shim1994) paints a picture of Korea as a country with a high level of bilingualism. In fact, Shim (Reference Shim1994: 237) estimates that at least half of the younger generation (aged 15–40) are Korean–English bilinguals, since ‘all Korean children are exposed to at least three years of formal English education during the mandatory education period, and since more than sixty percent of middle school graduates also finish high school education, during which English education is even more strongly emphasized.’ Since 1997, English has been taught from Year Three at the primary school level, and the retention rates from primary to middle school and from middle to high school are actually nearly 100 per cent (Song, Reference Song2012: 28, 34). In view of these developments, one would expect Koreans to have a higher level of competence in English than they did when Shim (Reference Shim1994) was published. Note, however, that Shim (Reference Shim1994: 237–238) includes Koreans with even a limited proficiency in English under her category of Korean–English bilinguals. Thus her claim that at least half of the younger generation are Korean–English bilinguals must be taken with a grain of salt. As evidence for her claim, Shim (Reference Shim1994: 234) argues that ‘[t]he increase in the number of English words in Korean naturally resulted in code–mixed sentences in discourse.’ She also observes that some Koreans code–mix beyond the level of English words to the extent of code–switching. However, her evidence falls far short of her argument, as it comes only from the mass media. Shim (Reference Shim1994: 235) is well aware of this problem when she writes that ‘[o]ne can argue that the examples that I have presented so far are all from popular media of advertisements, and television shows, and these cannot be used as concrete evidence for linguistic changes in everyday discourse patterns.’ Song (Reference Song1998: 267–268) calls Shim's (Reference Shim1994) claim into question by pointing out that code–mixing and –switching cannot be substantiated unless evidence comes not only from the mass media but also from other domains such as home, school and work, because (Shim's) mass media data are representative of what Haarmann (Reference Haarmann1986) refers to as ‘impersonal bilingualism’, i.e. the use of English for appealing to the public's positive feelings, not for practical communication. In their joint reply to Song's criticism, Baik & Shim (Reference Baik and Shim1998: 277) provide a few more examples of Korean–English code–mixing and –switching that they observed in their ‘chance meetings in public places with people who may or may not be proficient bilinguals [emphasis added].’ Baik & Shim (Reference Baik and Shim1998: 277–288) conclude: ‘[h]opefully, the above observations are sufficient evidence that code–mixing is a widespread phenomenon in Korea.’ (Note that Baik & Shim (Reference Baik and Shim1998) do not extend their additional observations to Shim's (Reference Shim1994) claim for Korean–English code–switching.)

Unfortunately, these examples count as no more than the kind of impressionistic observation that Buschfeld (Reference Buschfeld2013: 4) speaks of. Indeed recent research (Moon, Reference Moon2009; J. S.-Y. Park, Reference Park2009) indicates that the use of English words – let alone code–mixing and –switching – in Korea can only be attested to a considerably lesser extent than Shim (Reference Shim1994) and Baik & Shim (Reference Baik and Shim1998) lead one to believe. As J. S.-Y. Park (Reference Park2009: 34, 35) points out:

Further (obvious) evidence against Shim's claim for widespread Korean–English bilingualism comes from the fact that Koreans spent over US $18 billion on private tuition in 2013, a third of which went towards private English lessons (Allkpop, 3 September, 2014). Koreans annually spend as much as US $752 million on EFL proficiency tests (Card, Reference Card2006); Korea is the world's largest market for TOEFL (Card, Reference Card2006). This heavy investment notwithstanding, Korea ranked a disappointing 65th, jointly with eight other countries, out of 169 countries in 2014 TOEFL iBT scores (Educational Testing Service, 2015). These facts simply do not support Shim's depiction of Korea as having a high level of Korean–English bilingualism; rather, they indicate the status of English as that of a foreign language to be learned, unfortunately, at extremely high costs and with relatively poor results. The ineluctable conclusion is that there is no widespread competence in English and, consequently, no evidence for Expansion in Korea.

A similar objection can be made about Hadikin's (Reference Hadikin2014) ‘corpus–driven’ study. Hadikin's interviewees consist of 40 Korean–English bilingual Koreans from Seoul (Korea) and 30 Korean–English bilingual Koreans from Liverpool (UK). Hadikin recruited volunteers by using university (e–)noticeboards, Facebook, already recruited volunteers’ social networks and the like. Hadikin's semi–structured interviews with these bilingual Koreans constitute his ‘Korean English’ corpus data. It is, however, unreasonable to think that these data can somehow be representative of English in Korea, because his interviewees were not at all sampled and stratified in terms of age, gender, education, socioeconomic class and the like. While he does not provide any background information on them, it is likely that Hadikin's interviewees were mainly university students or, more generally, young Korean–English bilingual Koreans – Hadikin (Reference Hadikin2014: 16) calls them ‘educated users’. Thus Hadikin's study is not about English in Korea, let alone ‘Korean English’, but about English spoken by a very small number of educated bilingual Koreans. However, when we recognize X as a New English, we must do so on the basis of societal bilingualism in the country under investigation, not on the basis of a small minority of bilingual speakers in that country. What is needed is sociolinguistic domain research, that is, the kind of research that involves Koreans from all walks of life and systematically investigates their linguistic practices in different domains such as home, school and work. If one claims, as Hadikin does, that a given variety of English is a New English on the basis of a small minority of (educated) bilingual speakers, then one will have to say that virtually every country outside the Inner Circle speaks a New English.

Variety–status coding: a proposal

Schneider (Reference Schneider2007: 50) proposes that when a local variety of English is accepted as providing the linguistic norm for its speakers, that variety should be ‘given symbolic expression by substituting a label of the ‘English in X’ type by a newly coined ‘X English’’. This entails that a New English is to be called ‘X English’ while a Learner English is to be given the label of ‘English in X’ (X being the name of the country or the adjective form thereof). This is intended to maintain a strict distinction between New English and Learner English. Unfortunately, ‘English in X’ is not always easy to use unambiguously as a Learner–English label, because it can also be understood as a generic reference to all variety types of English found in X: Learner English of the majority of speakers, English spoken by a minority of bilingual speakers, and English spoken by a small number of residents or visitors from the Inner or Outer Circle countries. To address this potential problem, I propose a simple coding convention. For New Englishes, a two–letter ISO country code and ‘E’ (short for English) will be juxtaposed in that order (a complete list of ISO country codes is available at https://countrycode.org). For instance, Indian English will be coded as INE, Nigerian English as NGE and Singaporean English as SGE. For Learner Englishes, E and a two–letter ISO country code will be linked, in that order, with the lower case ‘i’ (short for in) intervening between E and the ISO code. For instance, the Learner English variety in Korea will be coded as EiKR, the Learner English variety in China as EiCN, the Learner English variety in Japan as EiJP, and so on. Unlike the label ‘English in Korea’, EiKR will not be understood ambiguously, because this code refers to the Learner English variety in Korea only. Following Buschfeld (Reference Buschfeld2013), I also suggest that Hybrid Englishes – i.e. varieties bearing a resemblance to both New English and Learner English – be double–coded, with a slanting slash between the two codes. For instance, Hybrid English in Cyprus will be coded as EiCY/CYE. Moreover, if Hybrid English in XY is more akin to one than the other end of the continuum, the coding convention can make use of a greater–than sign ‘>’ to reflect this complexity. For instance, English spoken in Cyprus can be coded as CYE>EiCY, because it is thought to be closer to the New English end than the Learner English end (Buschfeld, Reference Buschfeld2013: 202). If one does not wish to commit oneself to saying to which end of the continuum Hybrid English spoken in Cyprus lies closer, one can simply code it as EiCY/CYE. If Hybrid English in XY is more akin to the Learner English end than the New English end, it will be coded as EiXY>XYE or simply as EiXY/XYE.