2.1. Diffusion patterns of English: elitist vs. grassroots

This entire development is to be set against at least two important frameworks in linguistics: theories of Second Language Acquisition (SLA), which will be addressed below, and the growth of ‘World Englishes’ in general, as a product of colonization and, even more intensely, social changes of the postcolonial period (cf. Schneider, Reference Schneider2011 for an overview of the process itself and its linguistic manifestations and ramifications). I argued earlier that substantial similarities can be observed across the many instantiations of the spread of English through centuries and to many continents. These assumptions have been worked out in the ‘Dynamic Model’ (cf. Schneider, Reference Schneider2007) of the evolution of postcolonial Englishes, which has been widely accepted, adapted and applied so far. While the colonial period ended roughly half a century ago, the expansion of the role of English has not come to a halt since then, quite to the contrary, with English having become the main carrier of globalization. The early 21^st century has seen an even more vibrant expansion of English, which I have called a new ‘phase’ of ‘transnational attraction’ (Schneider, Reference Schneider2014). Buschfeld and Kautzsch (fc. Reference Buschfeld and Kautzsch2016) have proposed a new model which is meant to extrapolate from the Dynamic Model and widen its applicability to countries in which English traditionally had the status of a Foreign Language (EFL), called the ‘Expanding Circle’ in Kachru's (Reference Kachru1992) terms, but where second-language varieties or similar conditions have emerged in recent years. They suggest that in many countries even outside the postcolonial sphere a range of ‘extra- and intra-territorial forces’, many of them driven by globalization, have been causing an increasingly stronger role than just a few decades ago, and often an internal one in their respective nations (as shown convincingly by Edwards, fc. Reference Edwards2016 for the Netherlands, for instance). Tourism is explicitly listed as one of those forces.

A focus on the grassroots diffusion of English introduces and strengthens an innovative dimension insofar as the diffusion of English in the ‘Outer’ and ‘Expanding Circles’ (Kachru Reference Kachru1992) has often been associated with a slightly elitist air, given that it is usually taught in the education system (and higher secondary education in particular), with access to these levels of schooling being largely restricted to more affluent strata of local societies. For example, Bao (2015: 5–6) explicitly classifies New Englishes as ‘scholastic’ but suggests to recast the definition to include vernacular forms. The traditional restriction of associating the acquisition of English predominantly with formal teaching clearly fails to describe current realities adequately. The very strong pull of English in globalization, its ‘transnational attraction’, has increasingly produced novel forms and usage contexts which represent cases of ‘grassroots’ growth. In many countries, even limited proficiency levels in English offer relatively attractive job opportunities, often involving interactions with international visitors. Consequently, there are many speakers of forms and varieties of English nowadays who come from relatively poor backgrounds but nevertheless have striven and managed successfully to acquire a communicatively effective level of competence. Typically they have learned English in direct interactions rather than through formal education, largely disregarding target variety orientations or concerns about linguistic correctness. This is what I mean by the notion of ‘grassroots Englishes’.

Speakers who have acquired English in this way mostly have an exceptionally high instrumental motivation. For example, a guide in the mountains of northern Bali told me, answering my question about his acquisition path, that he had never had any formal instruction in English in school. However, he had gotten hold of a textbook with cassette recordings and spent hours every day to learn what he read and heard there, and he had a friend who spoke English and could be asked for instructions and explanations. This is a pattern, I think, which is not that uncommon, often with ambitious members of poorer strata of many societies – speakers work hard to acquire English in whatever ways possible, mostly because they know that this will offer means of improving their income and new prospects for personal advancement. Of course, instruction in English is increasingly also introduced and its importance recognized in elementary schools in poorer countries. The headmaster of a school for poor children in Varanasi in India tells me that for his pupils, who typically speak Hindi as a home language, learning English is the most important subject and goal, even more important than learning reading or mathematics. Vaish (Reference Vaish2008) is one of the very few studies of such teaching contexts of which I am aware. The problem in these circumstances, e.g. in many rural settings in Africa and Asia, is poor funding of schools, very difficult learning and teaching conditions, and hence generally a low quality of schooling and a relatively low success rate. Another difficulty specifically for language learning is often caused by culture-specific traditions of teaching which tend to emphasize authority, rote learning and chorus repetition rather than the ability to communicate freely. So even in contexts where some elementary teaching of English in the school system is available typically higher proficiency levels are achieved only later, through natural acquisition in communication. But there is no doubt that the strong pull of English enhances its grassroots growth in many countries and contexts. Hence, many speakers of such ‘grassroots’ Englishes, often from relatively poor backgrounds, have managed successfully to acquire a communicatively effective level of competence in English in direct interactions. The goal is an ability to communicate, not to get good grades or to be able to perform ‘correctly’.

2.2. The sociolinguistic perspective

The phenomenon in which I am interested here is not new – indirect references to it can be found in writings on World Englishes and sociolinguistics, and it is of course a relevant topic for studies of SLA. It has not been addressed coherently as such, however, spanning these disciplines. In the World Englishes context it is probably best situated in an emerging ‘poststructuralist’ perspective, which looks beyond a traditional notion of ‘varieties’ and refrains from focusing on nations (and the linguistic systems assigned to them) as reference points. Instead, the target here is to understand and describe speakers' attempts to perform successfully in context by using the limited linguistic means available to them profitably, in contexts in which snippets of English, or at least less-than-perfect competence levels, are employed as mobile resources. Comparable frameworks, in which similar observations have been made and similar text samples are documented (albeit in their respective argumentative contexts), are Blommaert's (Reference Blommaert2010) ‘sociolinguistics of globalization’, Meierkord's (Reference Meierkord2012) idea of ‘interactions across Englishes’, or Canagarajah's (2013) concept of ‘translanguaging’.

But clearly the notion of ‘grassroots Englishes’ is not an established one in the sociolinguistic literature. To my knowledge, there is only one context in which the term ‘grassroots English’ as such (in the singular) appears, namely in Blommaert (Reference Blommaert2010: 128), where it is explicitly applied and constrained to the linguistic usage often found in email fraud messages, which are also characterized by low proficiency levels and a complete lack of norms. In general, I think it is justified to state that relatively little attention has been given to this phenomenon in research so far. One reason for this state of affairs may be the fact that while there is an enormous amount of interest internationally in how to learn and teach English most efficiently, this is directed mainly to formal education strategies and the ultimate goal of guiding students to attain norm-oriented linguistic skills. Another may be the low status of speakers of ‘grassroots Englishes’, and the occurrence of such speech forms mainly in informal contexts; the topic may indirectly be considered ‘not respectable’ enough for serious scholarship, at least in some academic circles. And finally, in my view there is also a lack of a suitable theoretical framework to account for grassroots speech forms, no established classification where this kind of performance would fit in neatly. As Görlach once stated, ‘linguists often shy away from cases which do not permit neat classifications’ (Reference Görlach and Görlach1998: 1–2).

2.3. The acquisition and proficiency perspectives

Obviously, speakers of grassroots Englishes are first and foremost native speakers of local languages who strive to acquire English as a second or further language, so they are bilingual or multilingual. Languages may be attributed to functional domains, and in their repertoires English is meant for work, for contacts with visitors. In practice, of course, this produces a wide range of varieties, levels of competence, or realizations.

‘Grassroots Englishes’ therefore offer a very direct connection between the disciplines of World Englishes and sociolinguistics on the one hand and SLA on the other. Note that the term ‘acquire’ is used deliberately here, in Krashen's (Reference Krashen1981) technical sense which explicitly distinguishes unguided, natural second-language ‘acquisition’ in immersive settings from classroom-based, formal and guided ‘learning’. Obviously, the speakers I am considering here ‘acquire’ English in Krashen's sense (with details to be adjusted, given that Krashen's subjects have come to live in the country where the target language is spoken). In many respects their performance (and the samples which I present below) can be classified as ‘interlanguages’ in the sense of Selinker's (Reference Selinker1972) classic paper on SLA. In particular, and in addition to ever-present factors like L1 transfer and employing learning strategies, among the five ‘central processes’ of SLA which Selinker points out, these speakers build on ‘strategies of second-language communication’. Previous interactions with tourists in similar settings have taught them which modes of communicative and structural behavior are successful and sufficient, so their acquisition process operates via specific ‘feedback loops’. The general lack of concern for notions of linguistic correctness is clearly a case in point here: Practically always tourists are equally happy as their local helpers about just understanding each other, i.e. nobody corrects any grammatical errors or cares about linguistic decorum as long as the meaning is clear (cf. Selinker Reference Selinker1972: 220).

Defining the notion of bilingualism, i.e. answering the question of ‘Who is bilingual?’ (quoted from Bialystok, Reference Bialystok2001: 4) by defining a proficiency level which counts as sufficient for that notion to apply, is one of the core issues in bilingualism research. Does it require ‘native-like control of two languages’, as Bloomfield (Reference Bloomfield1933: 56) posited, or are lower proficiency levels sufficient to diagnose ‘relative’ or ‘limited’ bilingualism (Bialystok, Reference Bialystok2001: 226)?

Irrespective of where scholarship draws boundaries, the important point in the present context is that in practice for many speakers proficiency levels are very much functionally determined. As Grosjean put it: ‘Levels of fluency in a language will depend on the need for that language and will be extremely domain-specific’ (Reference Grosjean2008: 14). So for speakers their goal of language acquisition is often not determined by some outside, absolute norm but by their need to function in highly specific contexts, associated with narrowly circumscribed proficiency requirements. And this attitude often has a direct effect on one's acquisition level: In SLA research the phenomenon has been observed that learners display a ‘tendency […] to stop learning as soon as they think they have mastered the TL sufficiently’ (Buschfeld, Reference Buschfeld2013: 58, discussing work and positions by Selinker, Reference Selinker1972 and Corder, Reference Corder1981; cf. Selinker Reference Selinker1972: 217).

Similarly, Blommaert (Reference Blommaert2010) deals with the concept of ‘truncated repertoires’, i.e. partial proficiency levels restricted to specific skills, and notes that ‘no one needs all the resources that a language potentially provides’ (103). It is noteworthy that such a deliberate limitation of one's language learning goals is also relevant in formal education systems: ‘doing ‘well’ at school means doing well by local standards’, not necessarily attaining some sort of context-independent goal; the yardstick often is not a monolithic outside norm but ‘local normativity’ (p. 95). Consequently, with many speakers he observes ‘chequered’ abilities, a ‘patchwork of competences and skills’, determined by an individual's real-life background and needs (p. 133). This is precisely what I wish to illustrate in my next section, with reference to contexts of tourist interactions, but across boundaries of nations or continents.

3.1. Case study I: grassroots idiolectal acquisition in India

The first two texts are from southern India. In (1), a driver explains the use of coconut shells (see illustration 1) to tourists during a stopover at a fruit stall. In (2), a guide comments on illegitimate male-female relations as part of a longer conversation on culture-specific social habits.

1. (1) And this is coconut fibre, this made from coir, rope, coir. This shell, this shell, it is hard, you can use, lot of, this, we are making the spoons, the houses[?] the tradition spoons, yeah, all things is useful.

2. (2) Because you are not by the state, you are not legal-side, this is. But the local peoples, they're[?] do, you have, eh, not married, you have two children. That … , that they put[?], girls, it's not erm, mothers, there's not a, uh, this society is not allowed to, uh, live by the males. That's the male, it is, eh. Without marriage, without marriage they has boy or girl, ne[?] lady, that's not a good habit in India, girl. But, uh, the local peoples, uh, think like this, that is not good way, good way. Who'll take the response, who take the response? If, if the boy or girl, without married, the boy die, what are with the girls? They have nothing. This, what a piece this have, boy's hand, that is going to the boy's home, boy's parents, going back to. This girl and two children, no legal, it's not legal, illegal.

In both texts the basic meaning is comprehensible, though not all details are clear. This applies to sample (2) in particular – perhaps because the subject is more abstract, and perhaps also because it appears to be a bit touchy to the speaker. Both utterance sequences lack coherence in some parts: there are several fragmented, syntactically incomplete utterances, as well as unconnected, ‘dangling’ constituents (this, … ; … ; girl.). Some repetitions (This shell, this shell … ; If, if … ; Without marriage, without marriage …), hesitation forms (eh, uh,) and self-corrections (no legal, it's not legal, illegal) indicate the speakers' insecurity, but also their motivation to communicate successfully and get the message across. Both samples show several grammatical nonstandard features: lack of inflectional concord (all things is … ; they has … ; the boy die; what are with the girls); a redundant plural suffix (peoples); lack of copula or other auxiliaries (this made from … ; who take the response; without married); a progressive for habitual actions (we are making the spoons); and a minimalist negation with pre-adjectival no, found in a similar fashion more often in pidgins (no legal). Missing lexis, words which the speaker needs but doesn't know, is paraphrased, replaced or even freshly coined (are not by the state; live by the males; response ‘responsibility’; legal-side). And there are further, phraseological examples which in a similar vein are indicative of the speaker's creative, successful ability to express himself (that's not a good habit in India, girl; the local peoples think like this, that is not good way).

Illustration 1:

How to use a coconut shell

3.2. Case study II: grassroots idiolectal acquisition in Indonesia

My first example from Indonesia, not based on recordings but on my notes from immediate memory, is the idiolect of ‘Mama M.’, as she refers to herself. She is a middle-aged small-scale vendor who lives on the Indonesian island of Lombok, commutes regularly with a local boat to the adjacent tourist island of Gili Meno, and walks arounds there, talks to people, and sells Mangos, Papayas, bracelets, and massages. Her speech shows some features which are conventionally associated with pidgins (lack of copulas, articles, subordination, or complexity; third-person self-reference) but which could also be explained as direct transfer from her native language;Footnote ³ they are illustrated in the following examples:

1. (3) ‘Mango good!’

   ‘You pay ten, Mama M. not happy. You pay fifteen, Mama M. happy!’

Despite her evident proficiency limitations, Mama M. is extremely communicative. The topics of her conversations span a fairly wide range, including, naturally, her business needs (praising her goods, negotiating prices), but also covering basic facts of life, both concerning herself and information asked for from the tourists (work and life patterns, family relationships, one's children and what they do in life, etc.). Her English is fully self-acquired in contact with tourists; she has never had any formal tuition in it. But clearly it is fully sufficient for her purposes, and in some ways quite versatile, in fact (consider the implied conditional without conjunctions in the second sentence quoted above); it is simple but efficient.

The second example from Indonesia is a recording made with a middle-aged, male driver in Bali, who says he speaks Balinese, Bahasa Indonesia, and English. He explains rites and habits after a death in the neighborhood and, subsequently, mobility constraints:

1. (4) If de people going to de cem-, if now die, looking for good -, looking, eh, Balinese calendar, when I can bring to cemetery, but if we do cemetery, around four o'clock, five o'clock, before de, uh, sun down, maybe around five o'clock, and den, you give in[?], with many people. If here, if, uh, if I always come when the neighbor co- [?] dead, the neighbor have, uh, something, and then I hel(e)p him, evry[?],evry[?], uh, neighbor dead, yeah, neighbor, I have, dead, I come. When I die also come. When me lazy, only my family. There is dis[?] difficul' here in, uh, in Bali. Also, I am not from Lovina, when in Lovina I live with, rent a house, yeah, but I'm from de mountain here. If my neighbor in de mountain dead I come with motorbike go home. Sometime alone. If I have ceremony, one bike, four people: my daughter, my son, my wife, and suitcase behind. My daughter fourteen years, my son ten years, and small motorbike. But I have only one motorbike. But now I must try another motorbike.

Again, in my view this is an interesting linguistic sample which illustrates a number of phenomena. The speaker's proficiency is clearly limited, presumably due to his informal, natural acquisition (I did not ask in this case). There are hardly any complete sentences. There are many false starts and repetitions (going to de cem-, … , bring to cemetery, but if we do cemetery, … ; from Lovina … in Lovina). Grammatically deviant patterns could be analyzed as nonstandard behavior or learner errors, or they could also be explained as representing instances of first-language transfer (given that in indigenous languages, notably spoken Bahasa Indonesia, the omission of copulas, articles and clause subjects is also common and singular-plural agreement is unknown). They include the lack of copulas and (some) articles or determiners (e.g. my son ten years; with motorbike), agreement errors (the neighbor have …), the lack of obligatory syntactic constituents (e.g. the subject of the main clause: When I die also come.), and the lack of conjunctions and clause-connecting devices (neighbor dead, … , I come). It is noteworthy, however, that the latter example, just like the one in (2), expresses a cognitively rather complex conditional relationship ('When a neighbor is dead I come / go there.'). Again, many hesitation phenomena can be found (if, uh, if …), and there are simplistic word choices, esp. for verb collocations with nouns and adjectives (do cemetery; have ceremony; have dead).

On the other hand, despite some restricted coherence of the text the speaker is nevertheless largely communicatively successful for the purposes required. He gets his message across, and he is also aware of pragmatic needs, e.g. by seeking backchannels which he waits for before he continues speaking. So, while his proficiency is obviously fairly limited, by the standards called for he qualifies as a successful speaker of English; he fulfills all prerequisites for his job - of which being able to speak some English is a core one.

3.3. Case study III: grassroots idiolectal acquisition in Tanzania

A similar situation obtains in northern Tanzania, where tourism (supporting and accompanying tourists who wish to climb the Kilimanjaro and other mountains, cf. Illustration 2; organizing and guiding safaris through the National Parks of the region) is a major business and source of income for many. Language, and notably the ability to speak English, is the prime component for social and economic advancement: To achieve promotion from the low-prestige, demanding and relatively badly paid role of a porter to the respectable, reasonably well-paid functions of (chief or assistant) guide or driver, the main qualification is the ability to speak English, i.e. to be able to communicate with clients. Most porters speak Swahili (and possibly other local languages) but no English. All guides usually worked as porters for a number of years first and worked their way up through some education (but of course all the locals know the mountains and the animals almost equally well) and, most importantly, the learning of English.Footnote ⁴ As in the cases discussed before, there is no need for any ‘respectable standard’, for ‘international norms’, or whatever (which most tourists also do not command or care about anyhow) – the only yardstick is intelligibility, the prerequisite for communicative success.

Illustration 2:

Porters during a traditional dance and song after a successful tour up Kilimanjaro (in the back)

In these cases the speakers' background is usually slightly different from the aforementioned ones because most of them would have had some but highly limited schooling, mostly including some teaching of English. In government schools in Tanzania, English is a subject of education in primary schools and the medium of education in secondary schools (Ochieng, Reference Ochieng2015) – but classes tend to be huge, and the effectiveness of teaching English is ultimately low; consequently, many Tanzanians have practically no or only a very limited command of English. Real, effective acquisition then comes largely naturally, in contact with clients. And again, for most individuals a huge interest in learning English and an enormous demand for it can be observed, often driven by an extremely high personal motivation.Footnote ⁵

The speakers in the samples below are two different guides. In the first text the guide gives the briefing on the upcoming hike which is customary every morning on the mountain, explaining the conditions of the path to come and commenting on technical arrangements, as in this case when and how to get food on the way. The topic of the second text is the Senecio plant (see illustration 3), which is endemic to that region, and has quite unusual properties, which are described by the guide.

1. (5) After we reaching Saddle we may finding the cooker is already there, we gonna maybe you have a lunch dere, or we can, we're going to ask him the cooker if you will gonna have a lunch, pocket lunch, on the way, or we're gonna have hot lunch, right? It just depending, with the cooker, how, and we gonna see how you, how you, how, how you fitness you are, right? If you walking very slowly we can decide for all of us, we can make me pocket lunch for you, but if you have a go with the hard time, is better for us to go to have a lunch in, in the Saddle Hut, hot lunch, right? Because sometime we will gonna walking slowly, sometime, just depending how our body [ac]climatization [xx] is, all right? Yeah.

2. (6) And it's among of the leaves, the leaves are growing from 3500 meters, below that you can't found this tree or this leaves. And this leaves, once their, their leaves are die, can't grow, because, need to cover the body, because the tree inside here, it doesn't have in anything inside you need, because inside it just a hole.

Broadly, these speakers are more fluent than the Indian or Indonesian ones quoted above, presumably because their level of education and experience is higher. Similarly to the previous samples, some structural phenomena are worth observing and analyzing. Texts (5) and (6) also show a lack of the copula (we reaching; you walking) and some missing number concord (this leaves). Lexically, there is an interesting case of creative derivation (cooker)Footnote ⁶ . Syntactically, interesting and mostly recurrent patterns include the use of -ing verb forms rather than infinitives after modals (we may finding; we will gonna walking), the progressive for stative verbs (leaves are growing), word class reassignments which might be analyzed as instances of conversion (how you fitness you are; leaves are die), redundancies in constituent doubling (ask him the cooker; we can make me pocket lunch) and preposition doubling (among of).

Illustration 3:

A Senecio plant

Again, despite some idiosyncracies these are clearly highly successful and sufficiently qualified speakers, certainly so by local standards. Their proficiency levels are perfectly functional, and in fact comparatively high. They are able to truly build bridges between speakers from different cultural and linguistic backgrounds.