From learners to users—errors, innovations, and universals

Abstract

This paper looks into the dilemma of what counts as a grammatical ‘learner error’ in ELT on the basis of recent results from English variationist research and English as a lingua franca research. Examples from these studies show that features often perceived as ‘errors’ for EFL speakers also occur in ESL production—where they are called ‘innovations’—and even in English as a native language (ENL) production—where the term ‘spoken language universal’ would seem more appropriate. It is proposed that the traditional categorization of English speakers into EFL, ESL, and ENL actually hinders us from seeing the linguistic realities of the English-speaking world, disfavouring EFL speakers in particular. The paper calls for ELT to see EFL speakers as users of the language rather than learners in order to acknowledge their grammatical creativity as similar to that of other speakers of English.

Introduction

English speakers have traditionally been divided into native (ENL) speakers, second language (ESL) speakers, and foreign language (EFL) speakers. The division between second language and foreign language speakers was originally based on mainly historical developments—ESL varieties having a colonial background (e.g. India or Hong Kong) versus no colonial background in the case of EFL speakers. This difference between the ESL and EFL speech communities was pointed out as a crucial distinction early on: while ESL varieties had an intranational speech community, this was lacking for EFL, rendering EFL speakers ‘norm-dependent’ on native speaker Englishes (Kachru 1985). The presence of an intranational speech community for ESL gradually led scholars to perceive ESL speakers as fully fledged speakers of English, and their Englishes (e.g. Indian English or Nigerian English) to be recognized as varieties in their own right with their own deviations from Standard English. Nevertheless, this did not happen rapidly: the debate over the status of the ESL varieties only really started in the 1990s, and studies into these varieties have grown into a research field of its own only during the past thirty years or so (discussed in, for example, Koch, Lange, and Leuckert 2018). Today, however (as the vast literature in the field of World Englishes evinces), ESL speakers are seen as speakers developing their own Englishes and being creative users of English with their possible deviations from the ENL Standards—they no longer carry the status of ‘erroneous learners’ (see, for example, the discussions in the chapters in Deshors, Götz, and Laporte 2018).

It is interesting, however, that the division into ESL and EFL speakers in English studies was originally based on how the language spread in the world, and therefore, disregarded the essential psycholinguistic similarity between the two speaker groups: their bilingual basis for using English. Because in both ESL and EFL settings, non-native speakers learn English as an additional language on top of their mother tongue(s),¹ similarities between the linguistic outcomes of these speakers are to be expected as the psycholinguistic, bilingual processes, and strategies are also bound to be similar for both speaker groups (see, for example, Paradis 2004 on the processes of a bilingual brain). Although Shridhar and Shridhar (1986) called early on for bridging the ‘paradigm gap’ between EFL and ESL studies, in other words studying both speaker groups together, only recently have scholars in the field of varieties of English (or ‘World Engishes’) begun to scrutinize the possible similarities of the linguistic outcomes of these speakers (especially in grammar) (see, for example, the chapters in Mukherjee and Hundt 2011 and Deshors, Götz, and Laporte 2018). In the research field of English as a lingua franca (ELF), on the other hand, studying ESL and EFL speakers together has been at the core of research all along (e.g. Jenkins 2000). In ELF research, both EFL and ESL speakers are seen as users of English (instead of learners), which instantly opens up a fruitful landscape into the linguistic creativity and strategies used by all non-native speakers in their English-mediated communication.

Yet, because of the traditional categorization, we find ourselves in a peculiar situation: the same non-standard linguistic outcomes, especially in grammar, are labelled as ‘innovations’ or ‘linguistic creativity’ in ESL but ‘errors’ in EFL (see the chapters in Deshors, Götz, and Laporte 2018). This leads us to ask why, and how can we actually distinguish between a ‘learner error’ and a ‘user innovation’. To add to the dilemma, there is another issue that has barely been acknowledged, if at all, in defining an ‘error’ in L2 production: the fact that especially in spoken language both native and non-native speakers may resort to certain similar non-standard shortcuts or simplified constructions, but whereas for natives these features are just normal features of spoken language, for non-native speakers they constitute an ‘error’ because of their non-standardness (Ranta 2013). In these cases, the term ‘spoken language universal’ would often be a more appropriate label (see below). All this disfavours especially EFL speakers. It seems that EFL speakers are often expected to be more ‘correct’ and ‘standard-like’ (when it comes to grammar) than other speakers of English. They are not allowed the same linguistic leeway as others and are penalized for utilizing the same English-intrinsic resources that ESL and ENL speakers are free to capitalize on.

In this paper, we will first look into the terms ‘error’, ‘innovation’, and ‘universal’ from the variationist perspective as they relate to grammar. Next, some examples of non-standard grammatical constructions that are common across different speaker groups but still considered errors when produced by EFL speakers are introduced. And finally, we will discuss how our understanding of ‘errors’ could change for EFL speakers and what the possible implications for ELT would be.

Errors, innovations, and universals

As this paper looks at errors, innovations, and universals from the variationist point of view, we also use the definitions from this field as a starting point. In his seminal work in the 1980s, Kachru laid down a basis for treating deviations from Standard English either as ‘errors’ (or in Kachru’s terms ‘mistakes’) or ‘innovations’ based on who produced them (e.g. Kachru 1985). In effect, commonly occurring divergences from Standard English (gradually) became to be known and treated as ‘innovations’ when produced by ESL speakers (as suggested by Kachru), whereas divergences in the use of EFL speakers continued to be regarded as ‘errors’ (due to the perceived lack of an intranational speech community). This has generally applied for written and spoken language alike, although deviations often start as spoken language features. Due to the perceived ‘norm-dependency’ of EFL speakers, deviations from Standard English have also been traditionally perceived as ‘errors’ in ELT. However, as scholars in the field of variationist studies have pointed out, this kind of division into ‘errors’ and ‘innovations’ between different speaker groups seems to lack a linguistic basis (and evidence) as it is mainly politically motivated (see Gut 2011, and the Discussion Forum in Mukherjee and Hundt 2011). Gut (2011) further notes that the difference between EFL and ESL lies in the norm-orientation and attitudes of the speakers, in other words in the targets set for the learners (and not in their linguistic outcomes).

Therefore, it comes as no surprise that based on recent research (see the next section) it seems that some of the ESL ‘innovations’ are shared between EFL and ESL speakers, presumably because of the shared cognitive bilingual processes, which make both speaker groups ‘innovate’ in similar ways in their non-native English.² In English variationist studies, for a non-standard feature to be called an ‘innovation’, it has to be widely spread and used, thus the quantitative side plays a significant role (although exact definitions of what ‘widespread’ means do not exist) (e.g. Gut 2011). This is also how innovations are seen in this paper: they are recurring non-standard features in L2 speakers’ production. Further, this paper adopts the view that for a feature to be called an ‘innovation’ it must be attested in different L2 varieties (ESL and EFL) in similar ways.

In essence, the above means extending the label of an ‘innovation’ from ESL uses to similar EFL uses. However, it also follows from the above that not everything deviating from the standard language is seen as an ‘innovation’, but there would still also be non-standard features that especially in ELT practices would count as ‘errors’. All speakers of all languages, be they native or non-native, err in their language production every now and then for different reasons (be it due to a change of plan in what one wants to say, fatigue, gaps in linguistic knowledge, etc.). It is good to acknowledge this even if we want to promote the same ‘speaker rights’ for EFL speakers as other English speakers have. Naturally, it is a question of how one defines an ‘innovation’, but especially in ELF research, which strives to highlight the non-native speakers’ right to be users of English (as opposed to perpetual learners), there has been a tendency to overgeneralize any non-standard elements as ‘innovations’ (even in the case of fleeting, one-off deviations). However, it is only natural that especially in real-time speech, so-called performance errors also occur and that is well accepted even if we perceive L2 speakers as users of a language and not learners. But what is suggested here is that deviating features that occur frequently and in the same form in both EFL and ESL production, and which thus seem to result from similar cognitive processes for all non-native speakers, would be better labelled as ‘innovations’—for both speaker groups, not just for ESL speakers.

On the other hand, there are also the proposed English-specific universals—research into which has not been given much attention in applied linguistics. In short, these are grammatical constructions that deviate from the standard but seem to occur across native and non-native Englishes alike (e.g. Kortmann 2004). They surface especially in spontaneous spoken language presumably due to the same short-term-memory processing limitations and time constraints that apply to all speakers whether native or non-native. Such phenomena could potentially be detected in any language, but our scope is, of course, limited to English, and thus to English-specific universals. These have also been termed ‘angloversals’ (originally coined by Mair 2003). The essential difference between an ‘innovation’ and an ‘angloversal’, then, is that while ‘innovations’ denote only non-standard non-native features, ‘angloversals’ cover non-standard features that occur in both non-native and native production. However, the similarity between ‘innovations’ and spoken language universals lies in the fact that both seem to be attributed to cognitive processes that are (respectively) similar across speaker groups. Thus, it is suggested here that spoken language universals, too, should be distinguished from ‘errors’. Especially in ELT, the universals have barely been considered when deciding what counts as an error in learner output, again leading to a situation where EFL speakers have been penalized for something that might be normal language use for other native and non-native speakers of English.

Examples of innovations and universals

To clarify the above theoretical considerations, let us have a brief look at some examples of innovations and angloversals that suggest themselves in research data. It has to be borne in mind, however, that research in both of these fields is still in its early stages and that data have to accumulate from a number of corpora to enable more definitive conclusions. However, the constant development of corpus methods has made promising progress in detecting similarities between the EFL and ESL uses of English (see, for example, the chapters in Deshors, Götz, and Laporte 2018) as well as between EFL/ESL and ENL uses.

As regards innovations, Schneider and Gilquin (2018) compared written EFL and ESL data in the ICLE learner corpus and the ICE ESL corpus, respectively, to find possible similarities in the ways EFL and ESL speakers form non-standard verb/adjective + preposition combinations—an area of grammar that is often cited as problematic for EFL speakers but where ESL speakers are allowed creativity. They found that, indeed, similar non-standard constructions occurred with a similar ratio in both datasets. Some examples of such compositions were: discuss about sth. (instead of discuss sth.), superior than (instead of superior to), conscious about (instead of conscious of), aspire for (instead of aspire to), study about sth. (instead of study sth.), and depend from (instead of depend on). The combinations occurred independently of the speakers’ L1s in both (EFL and ESL) corpora. Thus, the writers attribute the similarities first and foremost to analogies that non-native speakers draw in their non-native language (e.g. superior than by analogy with better than) or iconicity that non-native speakers exploit to organize the language data (e.g. in depend from the preposition duplicates the directionality provoked by the verb, as opposed to on which does not). In other words, it seems that these deviations from Standard English are due to general psycholinguistic processes common to both EFL and ESL speakers—and not, for example, transfer from one’s mother tongue. Therefore, such combinations could be seen as innovations and not errors—not only in ESL but in EFL, too.

As for spoken language angloversals, Ranta (2013) delved into the possible similarities in attested non-standard constructions among ELF speakers in the ELFA corpus, and among ENL speakers in the comparable MICASE corpus. The ELFA data comprised speakers from both EFL and ESL backgrounds. She found two grammatical features that were used in similar, non-standard ways in both non-native and native data: there is/there’s used with a plural noun (instead of there are), as in There’s a lot of people in the street, and using the inverted word order in indirect questions (instead of the direct word order), as in I am just wondering what’s the right answer (instead of what the right answer is). The latter phenomenon is also called ‘embedded inversions’ in the research literature. In the case of embedded inversions, it turned out that the construction is more common among non-native speakers than among native speakers, but that the linguistic context where these kinds of non-standard word orders occur are practically identical for both speaker groups (with, for example, the cliticized what’s triggering the non-standard use most often for both groups, and the verbs ask, wonder, and know as the preceding verbs of the indirect question provoking the most non-standard word orders for both). As regards the use of the existential there construction with plural nouns, it is striking that the non-standard there’s + plural was even more common among native speakers than among non-native speakers, while it is well attested in both corpora. Both of the findings above would seem to point to some kind of universal processing effects in English—independent of whether one speaks the language natively or non-natively. As Ranta (2013) reasons, the explanation most likely derives from the use of grammaticalized chunks in spoken English that reduce processing costs for the speakers and thus add fluency in spontaneous speech. The curious fact just persists that while for native speakers these non-standard uses are everyday language, in non-native speaker production they continue to be frowned upon.

From learners to users—implications for EFL and ELT

As the above examples show, there are non-standard grammatical uses in English that are ‘allowed’ for ESL and ENL speakers but not so readily for EFL speakers. The situation seems questionable. Even if not all non-standard features should be called, for example, innovations, it seems equally true that not all deviations from the standard can be labelled and treated as ‘errors’ either, just because they were produced by an EFL speaker. On the contrary, it seems more likely that even the EFL speakers are using the same affordances of the English language as other speakers to be creative and find ways to streamline and facilitate their communication. Indeed, the fact that EFL speakers exploit similar linguistic means (albeit non-standard) goes to show their mastery of the English language resources and their ability to be creative within the language system itself—rather than their inability as language users. The key question then is: how should this be addressed in ELT?

To answer the question, it is helpful to refer to Melchers and Shaw’s (2003) model of what a ‘school variety’ of English consists of (see Ranta 2013 for an earlier discussion of this). The researchers distinguish between ‘exposure’, ‘model’, and ‘target’, where ‘exposure’ means all the English learners read and listen to in classroom settings, ‘model’ refers to the kind of English the learners are supposed to imitate, and ‘target’ is the outcome that we expect learners to produce themselves. If we apply this division to ELT (which sees English as a global language with all its variation) for teaching lexico-grammar this could mean that we, first of all, expose learners to many different kinds of varieties of English (to prepare them to encounter different kinds of Englishes globally). Yet, the ‘model’ for teaching lexico-grammar in EFL settings would still have to be Standard English because there are currently no other feasible, endonormative models of English in EFL contexts (unlike in ESL contexts). Thus, the point is not to start teaching the non-standard forms per se. However, when it comes to the ‘target’ (i.e. what we expect learners to produce themselves), there the naturally occurring innovations and angloversals described above could be taken into account.

As Gut (2011; see above) notes, it is the norm-orientation and attitudes of a speaker or speech community (e.g. ESL speakers orienting towards a local norm, and EFL towards Standard English) that ultimately decide whether a feature counts as an ‘error’ or something else. These sociolinguistic realities cannot be easily altered (as was seen in the case of ESL varieties for which it took over thirty years of scholarly debate to gain acceptance). However, on purely cognitive/psycholinguistic grounds it looks like the expectations for the linguistic outcomes of EFL speakers could be adjusted to correlate with the uses visible in other speaker groups. If a feature has its roots in shared, cognitive processes, why would it be an ‘error’ for an EFL speaker, but when produced by an ESL speaker considered a legitimate innovation, and when detected in ENL speech, be a natural adaptation of language? In ELT assessment and testing, for example, the gravity of a non-standard feature could be assessed against its prevalence in ESL and ENL settings. If the feature is frequent in other varieties and accepted there, it probably does not cause problems in communication globally, and could be accepted in EFL production as well. Of course, it is also a question of which yardsticks the EFL speakers themselves wish to be measured against and for what purposes. But at the very least the accumulating results on non-standard but cross-varietal features could be considered in determining different competence levels in EFL. They should also be brought to teachers’ and curriculum developers’ attention so that they can better judge what features are worth honing and what less so, for example from the communicative point of view. But even more importantly, testing practices and test development should be informed of such features (especially in tests that claim to assess the speaker’s ability to manage in ‘international settings’), so that EFL speakers would receive a fair evaluation against the linguistic realities of the English-speaking world.

ELF research has been calling for a change in ELT practices in seeing L2 speakers (whether from EFL or ESL background) not as learners, doomed to continuous error production, but as users making the most out of the language they employ to communicate globally. The above examples of specific grammatical features that L2 speakers use in similar non-standard ways across varieties and/or in consonance with native speaker uses further substantiate these claims. The traditional categorizations of different types of speakers have rather obscured us from seeing the similarities between the speakers and from acknowledging the fact that we all operate under similar constraints in communication and therefore also produce similar solutions to these constraints. It seems, therefore, questionable to hold EFL speakers accountable for ‘errors’ that are not errors by other speaker groups’ standards.

It is true, as Mauranen (2012) argues, that we can distinguish between (EFL) learners and (ELF) users of English by the context in which the speakers are situated (learners if in classroom settings versus users if in ‘real world’ settings). However, the cognitive processes for the speakers will be the same regardless of the settings, and therefore there is an intrinsic link between the two. As far as lexico-grammar goes, ELT could proactively start seeing their learners as users of the language, aligning EFL speakers with other English speakers in the world. By adopting this kind of view, there would be even less justification for penalizing EFL speakers for discuss about or for there’s differences in their production.

Conclusion

Seeing EFL speakers’ deviations from Standard English as errors does not always seem to be justified. This paper has brought forth examples of grammatical features that could be better considered linguistic innovations or general spoken English universals on the basis of them occurring across ESL and ENL varieties of English, respectively. Holding only EFL speakers accountable for such non-standard features appears to reflect more the traditional categorizations of English speakers and attitudes towards ‘learners’ than the linguistic realities of the English-speaking world.

The fact that EFL speakers lack an intranational speech community is not a strong enough reason to put them at a disadvantage compared to other speakers of English. Their psycholinguistic non-native processing is still similar to that of ESL speakers, and the constraints of real-time communication make them adjust linguistically in the same ways as ENL speakers do. These facts often lead to similar, non-standard linguistic outcomes independent of the speaker type. Further, the reference point for most EFL speakers would, indeed, be the international speech community, and therefore they should also be taught and evaluated with that reference point in mind. Seeing the EFL speakers as users of English, as ELF research calls for, could help to see some of the traditional ‘errors’ in a new light also in ELT. Yet, judging by how long it took for the ESL speakers to acquire ‘full rights’ as English speakers, it may still take some time for EFL speakers to achieve the same.

Nevertheless, building a stronger bridge not only between the studies of ESL and EFL but also between the English variationist studies and ELT in general would give English teaching a new perspective on EFL speakers. As the examples in this paper suggest, it seems that EFL speakers often are actually utilizing the grammatical affordances of the language in much the same ways as ESL and ENL speakers—because all non-native speakers alike are susceptible to similar cognitive bilingual processes, and all speakers to similar production constraints, and on the other hand, all have access to similar creative sources of the language. Thus, if EFL speakers’ creativeness points in the same direction as that of other English speakers, it actually goes to demonstrate their capability in the language, not their incapability, and thus the conceptualization of a grammatical error should be rethought.

Final version received May 2022

The author

Elina Ranta currently works as a university teacher of English at Tampere University (Finland), where she has also worked as a researcher and a lecturer in language teacher education. Her research work focuses on English as a Lingua Franca (ELF) and particularly the grammar of (spoken) ELF, on which she has published widely. She was also centrally involved in compiling the first one-million-word spoken ELF corpus, ELFA (English as a Lingua Franca in academic settings), in a research project with the same title.

Email: [email protected]

Footnotes

References