Modals of possibility, ability and permission in selected New Englishes

Stephanie Hackert, Ludwig-Maximilians-Universität München
Dagmar Deuber, Universität Münster
Carolin Biewer, Justus-Liebig-Universität Gießen
Michaela Hilbert, Universität Bamberg

Abstract

1. Introduction

Modality is an area of English grammar which not only evidences much synchronic variation but has also seen considerable diachronic change (cf. e.g. Krug 2000; Collins 2009a; Leech et al. 2009). Most work in this area has concentrated on native varieties of English (ENL), where the overall picture is one of decline of the modals and rise of the semi-modals, but this development has been fairly uneven for individual verbs. In recent years, there have also been a growing number of studies on expressions of modality in New Englishes. One tendency in these studies has been to look at aspects known to be subject to variation and change in native varieties to see to what extent the New Englishes follow suit. Another has been to try to identify aspects of modal verb usage that seem to be specific to one or several New English varieties and to discover factors accounting for the patterns found.

The main focus in this endeavour has been on the modals and semi-modals of obligation and necessity. Several studies (e.g. Nelson 2003; Biewer 2009; Collins 2009b; Mair 2009) have compared the frequencies of such modals across individual components of the International Corpus of English (ICE). Among the New Englishes for which complete ICE corpora are available, Indian and Kenyan English in particular tend to be conservative (Collins 2009b); Jamaican English is not conservative with regard to the declining must and the innovative need to but has not really adopted have got to, which could be owed to a reorientation towards American norms, as have got to is traditionally used in British English (Mair 2009). Biewer (2009) also discusses simplification and generalization in second language acquisition as well as substrate influence, but on the whole these factors have been less prominent in research specifically on modal verbs than in the literature on New Englishes generally.

If the main focus of interest is on how the New Englishes can be positioned in comparison to native varieties with regard to conservatism versus innovativeness, it is clear why the modals of possibility, ability and permission have not attracted much attention so far. [1] As Leech et al. (2009: 77) show, can/could, the most common forms in this area of modality, are among the modals whose frequency has not changed drastically if at all in spoken British English. Also, there is no tendency towards increased use of the semi-modal that covers part of the semantic space in question, i.e. be able to (Leech et al. 2009: 99). Only may/might, the second pair of modals expressing possibility, ability and permission, show a notable development, as may has declined so drastically in spoken British English that might is now the more frequent member of this pair (Leech et al. 2009: 77). [2]

The present study aims to contribute to filling this gap in the literature on New Englishes. In what follows, we analyse the quantitative distribution of can/could as well as these modals’ uses and meanings and also contrast them with other forms expressing possibility, ability and permission, i.e. be able to and may/might. This analysis forms part of a two-part comparative project on modal verbs in New Englishes, with the first part (Deuber et al. 2012) focusing on will/would. Accordingly, the varieties investigated and the data and method employed are identical in both cases, as are those parts of Sections 2 and 3 which describe them.

Six New Englishes, all represented in ICE, are considered. Three of them function as a second language (ESL) for most of their speakers (Indian, Singapore and Fiji English), while the other three (Trinidadian, Jamaican and Bahamian English) coexist with an English-based Creole, a situation which has been described as “English as a second dialect” (ESD) (Görlach 1991). British English is taken as a basis for comparison (all countries considered are former British colonies). In selecting the New Englishes for investigation we have placed a special focus on the Caribbean, as this region has so far been given less consideration in corpus-based comparative research on New Englishes than Asia, since until recently most available ICE corpora of New Englishes represented the latter region (cf. e.g. Collins 2009b). ICE-Jamaica became available only in 2009, and ICE-Trinidad & Tobago and ICE-Bahamas are still being compiled (cf. Deuber 2010b and Hackert 2010, respectively, for details of these projects). To further broaden the scope of research, we include data from Fiji, the only New English from the South Pacific region for which an ICE corpus is being compiled (cf. Biewer et al. 2010). Finally, we have selected Singapore and Indian English from the varieties for which complete ICE corpora are currently available.

The data are drawn from the text category “private conversations” in the respective ICE corpora where available; where the conversation component of an ICE corpus has not yet been compiled comparable other data are used. Sections 2 and 3, respectively, provide more detailed information about the varieties selected as well as about our data and method. Section 4 presents and discusses the results. Section 5, finally, offers concluding remarks.

2. The New Englishes selected

Indian and Singapore English are often considered typical ESL varieties, though in the latter variety there are ongoing changes towards ENL status. In both cases, the language was introduced during the colonial period not primarily via face-to-face communication but via the education system and now plays a key role not only in that domain, but also in national government and politics, the judiciary, business, and the media.

Even though, in terms of speaker numbers, Indian English constitutes one of the most important varieties of English worldwide, locally, its status and function are fairly tightly circumscribed. English constitutes “a minority lect, largely restricted to utilitarian functions and certain domains and strata of society” (Schneider 2007: 161). On the one hand, its role is that of a neutral lingua franca in an ethnically, religiously, and linguistically diverse nation; on the other, the language clearly functions as a marker of education and higher social status. Recent work (e.g. Vaish 2008; Chand 2009; Sedlatschek 2009) does suggest that Indian English is becoming increasingly important in the lives of a broadening segment of the urban population, but this development largely postdates the compilation of ICE-India in the early 1990s.

In the city state of Singapore, English “occupies a place, enjoys a status, and performs roles which it does not in any other Asian country” (Tickoo 1996: 431). It is, in fact, a home language for a growing number of speakers (cf. e.g. Lim & Foley 2004). Singapore may thus be considered “an ESL>ENL transition country” (Schneider 1999: 193). It must be noted, though, that many of these speakers have in fact acquired colloquial Singapore English (Lim & Foley 2004), a restructured variety (popularly known as Singlish) that may be considered as creolised (Ansaldo 2004).

English in the Fiji Islands, a group of more than 300 islands situated in the South Sea between Vanuatu and New Caledonia to the west and Samoa and Tonga to the east, is again a typical ESL variety. Like India and Singapore as well, Fiji is a multiethnic, multicultural and multilingual country. Beside the two biggest ethnic groups, the Fijians and Indo-Fijians, who comprise 57% and 37% of the population respectively, there are also many other ethnic minorities in the country such as Rotumans, Chinese, other Pacific Islanders, Europeans and Part-Europeans (Fiji Bureau of Statistics). Most people speak English as a second language (Tent 2001: 210) and one important function of English is that it serves as a lingua franca between the different ethnic groups (Tent and Mugler 2008: 236). English is not only the medium of instruction in school and the predominant language in literature and the media (Tent and Mugler 2008: 235), young urban Fijians also tend to use English with their friends and their siblings after school, and business correspondence and business transactions may take place in English depending on the language skills of the customer (Biewer submitted). In the urban centres English is very much part of everyday life. There are different lects of Fiji English that can be distinguished in terms of schooling, social class or rural versus urban upbringing (Biewer submitted; Tent and Mugler 2008: 236). Fiji English may be influenced by various external models such as British English through its history, New Zealand English through its economy and geographical closeness and American English through the media. In addition, second language acquisition, transfer from the mother-tongue and distinct cultural habits shape Fiji English into a recognisably distinct variety of English (Biewer submitted).

Even though the other three New Englishes selected all represent the category ESD, Jamaica, Trinidad and the Bahamas not only have a somewhat different linguistic make-up and different histories but they also represent different settings economically and socially. The most populous state in the Anglophone Caribbean, Jamaica (pop. 2.8 million), builds on tourism and bauxite/alumina as main industries but is currently facing a number of severe economic challenges (cf. Central Intelligence Agency 2010). Ethno-culturally, the country represents the typical situation on the Caribbean islands, with over ninety per cent of the population being of African descent. Trinidad, the larger and more populous constituent of the two-island republic of Trinidad and Tobago (pop. 1.2 million), in contrast, has been described as being “like no other Caribbean island” (Blouet 2002: 352). Reasons for this include its oil and natural gas-based economy, due to which it enjoys considerable prosperity, and the fact that people of African and East Indian descent make up about equal proportions of the population. The Bahamas is one of the wealthiest Caribbean countries, its economy being largely dependent on tourism and offshore banking. The country is heavily urbanized, with roughly two thirds of all Bahamians living in the capital, Nassau. Some 85% of the Bahamian population of ca. 345,000 (see the PDF in the Sources) are black. The 2000 census registered 21,000 Haitians, but some estimates including illegal immigrants put the current number as high as 78,000, or more than 20% of the population.

In each of these three Caribbean countries, English coexists with a lexically related Creole in what has most often been described as a continuum with Creole and English poles and a range of varieties in between. However, whereas in Jamaica a conservative or basilectal variety constitutes the Creole extreme, in Trinidad and in the Bahamas only intermediate or mesolectal Creoles are spoken. Attitudes towards the Creole varieties have been substantially transformed over the past few decades. Whereas formerly they were simply and uniformly designated as “bad” or “broken English”, they are now much more positively valued. However, as language attitude surveys (e.g. Beckford Wassink 1999; Mühleisen 2001) have also revealed, while speakers for the most part now hold positive attitudes towards their vernaculars, perceptions of a functional division between them and English remain fairly strong. As Youssef (2004: 44) observes for Trinidad and Tobago, “the Creole is the language of solidarity, national identity, emotion and humour, and Standard the language of education, religion, and officialdom”. Nevertheless, speakers balance the varieties according to the dynamics of the context, and some degree of “mixing” is characteristic of all but the most formal spoken language situations. Contexts in which such “mixing” increasingly occurs include the media and the classroom, as Youssef (1996: 11) notes. Similar observations have been made by Shields-Brodber (1997) for Jamaica, by Hackert (2004: 54–64) for the Bahamas and by Carrington (2001) for the Caribbean region as a whole. In the context of ICE, the interaction between English and Creole varieties is evident in particular in some of the public dialogues and most especially in the private dialogues (cf. e.g. Sand 1999; Deuber 2009a, 2009b, 2009c, 2010b).

3. Data and method

The following analysis comprises two parts. We first turn to forms and frequencies and look at all modals of possibility, ability and permission in the three New Englishes in our study for which complete ICE corpora exist, i.e. Jamaican, Singapore and Indian English. The quantitative distribution of modal verbs can easily be analysed in large data sets (cf. e.g. Collins 2009b), but the study of uses and meanings necessitates close attention to the context of each token. Both for this reason and because for three of the varieties considered in this study, i.e. Trinidadian, Bahamian and Fiji English, the respective ICE corpora are still being compiled, we have opted for a relatively small sample consisting of 15 ICE texts or equivalent data per variety for the second part of our analysis, which, furthermore, focuses on one pair of modals, i.e. can/could. [3]

The data are spoken, private, dialogic texts. In ICE terms, they make up the category of conversations, with texts coded as S1A-001 to S1A-090. Arguably, this category is the most heterogeneous of all ICE text categories, as conversations can be anything from intimate to fairly formal, i.e. they range from the type of excited exchange that occurs between two best friends cursing their husbands to discussions between colleagues at work or even more interview-like exchanges between relative strangers. Even though the ICE project aims at standard varieties of English around the world, the criteria for the inclusion of texts are based not on a specific language use but on who uses the language. Speakers or writers must be at least eighteen years old, and they must have completed secondary school (cf. Nelson 1996: 28). In native English-speaking contexts it seems natural and even desirable to record informal conversations between people who know each other well (cf. Holmes 1996: 169). In the case of the New Englishes, by contrast, the contexts of use of these varieties tend to lead to the inclusion of mainly relatively formal interactions in this category (cf. Schmied 1996: 185–186; Deuber 2009a, 2009b, 2010b; Hilbert & Krug 2010: 60).

However, a considerable range including some rather more informal conversations may still be represented. For example, the Bahamian texts analysed in this study consist of three types. There are, first, sociolinguistic interviews, i.e. interactions between a fieldworker and an interviewee designed to elicit casual speech (cf. Labov 1984: 32–33). Another set comprises interviews with linguistically sensitive professionals, such as teachers or journalists, about language use and attitudes in the community. Both sets of interviews were conducted by one of the authors of this paper, Stephanie Hackert, who was a student researcher at the time. Obviously, such interactions with outsiders to the community represent precisely the kind of situation in which recourse to the standard variety or the upper lectal levels of one’s linguistic competence is called for. Finally, two conversations involved only community members, with the researcher, SH, as participant observer. These conversations obviously were the most private, informal ones and produced the most vernacular speech. A similar range of data is found in the conversation category of the two other Caribbean components of ICE, namely those for Jamaica and Trinidad and Tobago (cf. Deuber 2009a, 2009b, 2010b).

For four of the six varieties selected, namely Singapore, Indian, Jamaican and Trinidadian English, a fully or partially complete ICE conversation category was available at the time of the study. In these cases, we have used the first fifteen texts (codes S1A-001 to S1A-015). The amount of data is in each case about 31,000 words. [4] The Bahamian data described above, which are yet to be integrated into ICE-Bahamas, also amount to about 31,000 words. In the case of ICE-Fiji, conversation data are not available for research at the present stage. The analysis of can/could in Fiji English in this paper is based on 14 interviews with Fijians which were recorded in Suva in 2007 as part of field research on South Pacific Englishes (Biewer submitted). [5] As each interviewee also filled in a questionnaire giving sociolinguistic background information, their age, ethnicity and regional upbringing is known as well as their highest education and current occupation. The 14 interviews were chosen out of 28 to have a more or less equal distribution of younger Fijians still in school or studying at university and older Fijians working in a job or enjoying retirement. Also a more or less equal distribution of male and female participants was envisaged. As a result eight of the interviewees are women, nine are men. While two of the interviewees are pupils, another seven are students from the University of the South Pacific, and the remaining eight are adults who finished their education and are part of Fiji’s work-force – apart from one woman who was already retired at the time the interviews were recorded. The interviews consist of 37,519 words; in order to achieve comparability with the data retrieved for the other varieties in question the results were normalised to 31,000 words.

We consider all forms of the modals in question in the first part of the analysis but in the second part we restrict our attention to positive, non-perfect forms, i.e. can/could followed by a verb in the base form. This has to do with asymmetries in the verb systems of Caribbean Creoles, which, at least potentially, influence Caribbean Standard English as represented in ICE-Jamaica, ICE-Trinidad & Tobago and ICE-Bahamas. In the case of negative forms, in both Trinidadian and Bahamian Creole, the equivalent of could is not couldn’t but kyaan (cf. Winer 1993: 35; Solomon 1993: 114; Holm & Shilling 1982: 35, 51), which is, of course, a reflex of English can’t. Not unexpectedly, a pilot study on Trinidadian Standard English (Deuber & Youssef 2007) revealed that the negative forms of can/could pattern differently from the positive ones. Perfect forms, i.e. can/could have followed by a past participle, were also excluded, as they have a different range of meanings and uses and are influenced by different Creole forms than can/could, i.e. coulda/woulda plus base form, as in example (11) below. Not included in either part of the analysis are tokens of the modals that occur in uncertain parts of transcriptions (text enclosed by ICE mark-up symbols <?> </?>), quotations (text enclosed by ICE mark-up symbols <quote> </quote>), extra-corpus text (text enclosed by ICE mark-up symbols <X> </X>) and repetitions or self-corrections (text enclosed by ICE mark-up symbols <-> </->).

The framework for the analysis of the uses and meanings of can/could has been adopted from Deuber (2010a); it is informed by descriptions of the meanings of the modal verbs in the major reference grammars of English as well as by specialized studies such as Coates (1983) and Palmer (1990) but adapted to the research issue at hand. Accordingly, a basic distinction is drawn in the categorization between the past or hypothetical uses associated with could in British and American Standard English and non-past, non-hypothetical uses. A separate category was further set up for pragmatically specialized uses, i.e. requests, instructions, suggestions and offers.

In the non-past, non-hypothetical category, three basic types of meaning were distinguished: possibility, ability and permission. Whereas ability and permission are associated exclusively with can in British and American Standard English, there is variation between can and could in the case of possibility. If could occurs, it may be either epistemic – a meaning which is very rare for can in assertive contexts, though a few examples have been found in ICE-GB (Facchinetti 2002: 235; Collins 2007: 479) – or, if dynamic, more tentative than can (cf. Huddleston & Pullum 2002: 200). The following are typical examples of epistemic and tentative dynamic could, respectively.

Perception and cognition verbs involve a subtype of the ability meaning with a low degree of modality, where the ability is currently actualized as opposed to theoretical (cf. Coates 1983: 90–91; Huddleston & Pullum 2002: 185). Again, in British and American Standard English, only can occurs with non-past reference; in our data, there is variation in this area of modality, too.

Some tokens had to be classified as unclear or indeterminate. “Unclear” means that there was not enough context to determine the meaning of the token in question, which happens either when speakers break off in the middle of a text unit or the surrounding words are unintelligible. Indeterminate instances are those where there is in principle enough context but where the token still cannot be assigned to a single category, either because of ambiguity or because the categories themselves are not discrete and the token falls on the gradient between them. Example (3) contains a token in which both inherent properties of the subject referent (i.e. ability) and external circumstances (i.e. dynamic possibility) influence the verb situation.

4. Results and discussion

This section presents details on the quantitative distribution of the modals of possibility, ability and permission in six New Englishes as well as on their uses and meanings. We first present and discuss the general findings and then comment on some variety-specific patterns.

4.1 General findings

An overview of the frequency of the various forms in conversations in three new Englishes and British English is shown in Figure 1. The varieties represented are those for which complete ICE corpora exist.

If a high frequency of may in relation to might is taken as a conservative feature (cf. Section 1), then the results for the three complete ICE corpora in our study are as follows: Indian English is very conservative with a high frequency of may, Singapore English less so, and Jamaican English not at all. As regards can/could, the New Englishes – and among these especially Singapore English – show a tendency towards greater use of can. Chi-square tests resulted in the following statistically significant oppositions with regard to can vs. could: ICE-GB vs. ICE-Jamaica, ICE-GB vs. ICE-Singapore and ICE-GB vs. ICE-India. Only the latter two oppositions emerged as statistically significant with regard to may vs. might. The semi-modal be able to, finally, is not a differentiating factor. Because may and might are too infrequent to be profitably investigated in a small data set, in the following we focus on can/could.

As Figure 2 shows, even though the number of forms is now restricted to two, the inclusion of more varieties leads to a rather diverse picture. In four New Englishes there is a clear preference for can, but this is more pronounced in Singapore and Indian English than in Jamaican and Fiji English, and the data from ICE-Trinidad & Tobago and ICE-Bahamas show a reverse tendency.

The distribution of can vs. could has so far been investigated mostly for ENL corpora, in which can is usually at least twice as frequent as could in speech and also more frequent in writing, though to a lesser extent. According to Collins (2007: 488), for example, the ratio of can to could in ICE-GB is about 2.2:1 in the spoken and 2.1:1 in the written component, while the corresponding figures for ICE-Australia are 2.3:1 (spoken) and 1.8:1 (written). In American English, as represented by the Santa Barbara Corpus (spoken) and Frown (written), the ratios are 2.6:1 and 1.4:1 respectively. There are exceptions to this general pattern. In speech, they concern the category of broadcast news, with about equal frequencies of can and could. In writing, could is favoured over can in news reports, novels and short stories, which suggests that in native varieties of English high frequencies of could are a function of frequent past reference in certain genres, especially in those where the narrating or reporting of events figures prominently (Deuber 2010a: 116–117).

The results in Figure 2, which is based on Table 1, to be discussed more fully below, show that private conversations display the “regular” pattern of high can-to-could ratios in all New Englishes except Trinidadian and Bahamian English. These ratios range from 2.4:1 (ICE-GB) over 3.7:1 (ICE-Jamaica) and 4.4:1 (Fijian data) all the way to 8.6:1 (ICE-India) and 8.8:1 (ICE-Singapore). Compare this to 0.5:1 and 0.6:1 for ICE-Trinidad & Tobago and ICE-Bahamas respectively. Chi-square tests resulted in the following statistically significant oppositions: ICE-GB vs. ICE-Trinidad & Tobago, ICE-GB vs. ICE-Bahamas, ICE-GB vs. ICE-India and ICE-GB vs. ICE-Singapore. In other words, whereas private conversations from Trinidad and the Bahamas evidence an exceptionally frequent use of could, the same text category shows very high figures for can in Singapore and India. It seems, thus, that at least some Creole-influenced varieties show a strong preference for could over can, whereas at least some ESL varieties clearly prefer can over could.

A look at the individual categories of meanings and uses of can/could as displayed in Table 1 reveals other interesting patterns.

The category of past or hypothetical uses does not really present any surprises in that it is covered almost exclusively by could, with which it is generally associated in British and American Standard English. There is an individual hypothetical token of can in ICE-Singapore, ICE-India and the Fijian data each; example (4) illustrates.

In the Fijian dataset, there is also a stray token of can with past temporal reference. Three such tokens occur in ICE-Bahamas as well.

As for non-past, non-hypothetical uses of can/could, it was noted above that the dynamic meaning of ability, including currently actualized ability with perception and cognition verbs, and the deontic meaning of permission are ruled out for could in British and American Standard English. Except for another stray token in Fiji English, such uses do not occur in our corpora, either – except for ICE-Trinidad & Tobago and ICE-Bahamas, as examples (6) to (8) illustrate.

With regard to possibility, it was noted above that in British and American Standard English, this category is typically covered by can, with could only occurring in epistemic or dynamic but more tentative contexts. Keeping this in mind, the quantitative distribution of can vs. could in the possibility category in Table 1 is remarkable. Whereas in ICE-Great Britain, can occurs about three times as frequently as could (3.2 : 1), this ratio rises to 6.5 : 1 in the case of ICE-Singapore, 7.6 : 1 in ICE-India and 8 : 1 in ICE-Jamaica. In the Fijian data, finally, we find 27 tokens of can but only a single one of could. ICE-Trinidad & Tobago and ICE-Bahamas, on the other hand, both evidence a ratio of can to could of 0.6:1. Chi-square tests resulted in the following statistically significant oppositions: ICE-GB vs. ICE-Jamaica and ICE-GB vs. ICE-Bahamas. At p = 0.05, the opposition ICE-GB vs. Fijian data just about cleared the mark of statistical significance. In other words, whereas ESL varieties evidence elevated rates of can in contexts expressing possibility, Trinidadian and Bahamian English show a strong preference for could. The frequency with which such contexts occur in our corpora, in turn, accounts for the strong preference of could which so clearly emerges in the overall quantitative distribution of can vs. could shown in Table 1 and Figure 2. We now turn to variety-specific patterns.

4.2 The Caribbean

Since can and could are associated with distinct subtypes of possibility – at least in British and American Standard English – it might, of course, be the case that there simply happen to be more contexts requiring or favouring could in the Trinidadian and Bahamian data than in the other datasets. At least for the Trinidadian conversations, a look at the possibility uses of could in context revealed that this could indeed be a contributing factor, as the material analysed for this study consists of fifteen conversations or discussions between teachers, and a frequently occurring topic is language use in school. In several of the conversations, the teachers discussed what they thought were possible ways of improving English language teaching, and there are a number of tokens of could in what can be interpreted as tentative expressions of an opinion on this issue. Example (2) above contains precisely such a token. For the Bahamian data, however, this explanation does not hold, as a topic restriction of this sort did not apply. Moreover, in both Trinidadian and Bahamian English, could often occurs not with an epistemic or dynamic tentative possibility meaning, but with a dynamic meaning which involves no tentativeness whatsoever. Example (9) illustrates.

There is, in fact, a more plausible explanation for the observed pattern of elevated frequencies of could in the Trinidadian and the Bahamian data. Both Trinidadian and Bahamian Creole possess systems of modal verbs in which could is actually a non-past form, the corresponding past or hypothetical forms being coulda and woulda, as illustrated in examples (10) and (11) from Bahamian Creole.

It seems highly likely that the distinctive preference for could in both ICE-Trinidad & Tobago and ICE-Bahamas is attributable to the continuum situation in which Creole and Standard English co-exist in the two countries. As already sketched briefly, conversations represent a situation that favours the more Creole varieties of the continuum even among Trinidadians or Bahamians with a high level of formal education. What we seem to be witnessing in the case of could in Trinidadian and Bahamian English may thus be described as a case of indirect Creole influence in that a form which is also present in British and American Standard English, could, is frequently used with Creole meanings, which comprise possibility, ability and permission with non-past reference.

This explanation may, in fact, also account for the distribution of can vs. could in the category of pragmatically specialized uses, which comprises requests, instructions, suggestions and offers. In conversations, such uses are often absent or rare, as is attested most clearly by the Jamaican data. Nevertheless, if they do occur, could is generally employed in British and American Standard English as a more polite form of can. The Trinidadian conversations contain only a single token of can and could each in a request. What is interesting is the Bahamian pattern, which, even though there are also only a few tokens, with the exception of one of them, all of these tokens are tokens of could rather than of can. Even though politeness conventions differ across cultures, of course, in a number of cases, there were indicators in the linguistic context which suggested that no special degree of politeness was intended in the use of could in the Bahamian data. What follows is a particularly striking example.

The Creole explanation, finally, also accounts for the distribution of can vs. could in ICE-Jamaica, which, as Table 1 shows, bears a striking resemblance to ICE-Great Britain. While Jamaican English co-exists with an English-based Creole in exactly the same way as Trinidadian and Bahamian English do, Jamaican Creole has a different set of modal verb forms. The past form is coulda as well (cf. Section 3), but the corresponding non-past form is kyan, which is simply the equivalent of English can. In other words, in Jamaica the Creole and English forms of the non-past modal expressing possibility, ability and permission are identical (except in details of phonology), which precludes Creole influence in this area.

4.3 India and Singapore

The specific features of the use of can and could in the two Asian varieties are mainly quantitative, not categorical. Generally, Indian and Singapore English display a substantially higher overall frequency of can than the non-Asian varieties, and a considerably lower overall occurrence of could, and consequently a higher ratio of can to could in comparison to the other New Englishes represented (cf. Table 1). This is in line with a previous study on will/would in the same varieties (Deuber et al. 2012), in which the Asian varieties also showed a higher overall frequency of will/would as well as a particularly high ratio of will to would. Besides more frequent overall modal usage, Asian Englishes thus apparently prefer the non-past modals can and will over the past-tense modals could and would.

Upon closer inspection, the elevated rates of can in the two Asian varieties are noticeably owed to a preference for using can in the expression of ability, and, to a lesser extent, the expression of permission, the latter particularly in Singapore English. The infrequent occurrence of be able to does not seem to play a role here, as this form also occurs relatively infrequently in the other varieties under study (cf. Figure 1). Possibility, on the other hand, is among the meanings for which can is used less frequently in the Singaporean data than in the other datasets employed here.

Additionally, can is frequently used to express suggestions, seemingly at the expense of could, which is hardly used at all for suggestions. We need to take into account here that, overall, we are dealing with relatively low absolute frequencies, which might be influenced by text-specific factors. Some of the conversations in the corpus sample of both Asian varieties involved lengthy passages on seeking and giving advice in the context of personal problems and (in the case of Singapore English) suggestions with regard to holiday activities, which are responsible for the comparatively high number of can tokens in this pragmatic context.

A difference between the two Asian varieties with regard to these pragmatically specialised uses is the expression of requests. In Singapore, can is also preferred for requests, again at the expense of could. In Indian English, on the other hand, can is only infrequently used in requests, and could is preferred. Requests are, in fact, the only context in which could is used more frequently in Indian English than in the other varieties under study here. In all other contexts, could is, as mentioned above, comparatively rare. The semantics of possibility are largely taken over by can, and likewise contexts of suggestions, as described above. In past or hypothetical uses could hardly occurs at all. Likewise, in Singapore English the occurrence of could in contexts of possibility, past time and hypothetical meanings is also reduced significantly, although not as dramatically as in Indian English. On the other hand, it hardly ever occurs at all in pragmatically specialised contexts, so that the absolute frequency of could is again very low in comparison to the other varieties and on a similarly low level as in Indian English.

All in all, the two varieties display a range of parallels: first, the high overall frequency of can; second, a focus with regard to the meaning of can on the expression of ability, permission (to varying extent) and pragmatically specialised uses like suggestions; and third, the low overall frequency of could, which is based on a lower occurrence in all potential semantic contexts. The main difference between the two varieties seems to be that these tendencies are more pronounced in Singapore English than in Indian English, a result that parallels the findings for will and would in these varieties (Deuber et al. 2012).

Finally, it is worthwhile mentioning that in Singapore English can has a relatively high idiomatic status and various different functions. Among the frequent invariant tags used in this variety are can or not and can cannot, which apparently have their sources in or are at least influenced by Chinese and Cantonese (cf. Gupta 1994: 127). Expressions like can lah ‘yes’ (cf. Gupta 2006: 253) are another context of occurrence for can. Such special uses of the modal are not part of this study and in fact occur rather infrequently in the data under analysis here, but they demonstrate the multifunctionality and idiomaticity of can in Singapore English, which may have a reinforcing effect on the form in general, thus marginalising could.

4.4 Fiji

In contrast to the other New Englishes, Fiji English shows a high relative frequency of can with the meaning of permission, which is only topped by Singapore English. In the case of Fiji English, this seems to be a socio-culturally motivated phenomenon. For Fijians it is very important to respect rank and hierarchy, and it is vital to know what is allowed and what is not according to customs, e.g. where to sit down in a house or around the kava bowl or what to wear in a village. [6] The Fijian participants in the interviews kept referring to such customs, partly because they wanted to explain their culture to a Western outsider but also because this is a fundamental part of their daily lives. Examples (13) and (14) illustrate.

This finding is in line with the results obtained in a study on modals of obligation and necessity in Fiji English (Biewer 2009), which showed that the deontic usage of should and must appears particularly frequently in comparison to native varieties of English. As the present data indicate, can is another modal auxiliary with which the community’s decisions on acceptance and prohibition may be expressed.

Two more unusual cases of can where British Standard English requires could occur in the Fijian data. They were produced by the same person, a young Fijian male student, who strongly identified with his rural upbringing. People living in rural settings in Fiji are usually less exposed to English; their use of English is identified by urban Fijians as “more broken”. This usage of can for could here either displays a lower proficiency in English or marks a rural identity. Example (4) above and (15) illustrate this use of can.

These examples show two things. First, socio-cultural background information on the society in question and sociolinguistic background information on individual speakers are vital in order to explain the origins of certain usages; second, learner errors may also play a role in the usage of can/could at least in some New Englishes.

5. Conclusion and outlook

This study has compared six New Englishes from three different world regions, i.e. the Caribbean, Asia and the South Pacific, with regard to modals of possibility, ability and permission. Since be able to and may/might occurred too infrequently in our data, we focused on can/could. With regard to this pair of modals, which we investigated in terms of sheer overall frequencies as well as in terms of meanings and uses, the varieties under study behaved in complementary fashion. Whereas in Trinidadian and Bahamian English could occurred considerably more often than can, the Asian Englishes under study employed can in about 90% of all cases.

The frequent occurrence of could in ICE-Trinidad & Tobago and ICE-Bahamas can be explained by the influence of the local Creole languages, in which could is used with meanings which parallel those found for British and American Standard English can. Jamaican Creole does not have that feature, which explains the comparatively lower frequency of could in ICE-Jamaica. The tendency of the Asian Englishes, and also Fiji English, to show elevated rates of can appears to be owed to different factors: socio-cultural phenomena in Fiji; topic constraints and the idiomatic usage of can in Singapore English; and finally, but possibly marginally, learner errors in Fiji English. Clearly, however, SLA is an important factor in the development of both Fiji English and Asian Englishes.

In sum, varieties of English spoken as a second language or dialect appear to show greater variability in the usage of can/could than native varieties of English. Whether and to what extent this holds for other linguistic features is not entirely clear as of yet. It is with the help of the extension of the ICE framework to more ESL and ESD contexts and of small-scale qualitative and quantitative studies such as the present one that this variability will eventually be appreciated fully.

Notes

[1]   Apart from Deuber (2010a) on can/could and will/would in Trinidadian English, one just finds the odd comment here and there on the use of can/could in various New Englishes (e.g. Trudgill & Hannah 2002: 132; Alo & Mesthrie 2004: 815; Huber & Dako 2004: 856).

[2] Cf. also the findings for written British and American English (Leech et al. 2009: 74, 97).

[3] All the data from ICE-Trinidad & Tobago analysed in this study are from Trinidad.

[4] The standard text length in ICE is approximately 2,000 words but many texts are slightly longer so that the end of the text occurs at an appropriate point in the discourse (cf. Nelson 1996: 27).

[5] The 14 interviews are raw transcriptions which have been checked once but still await their final correction; they are therefore liable to minor changes.

[6] Kava is a drink made from the roots or the stem of the yaqona plant; it is served in Fiji in a bowl on special occasions (cf. Geraghty et al. 2006: 637).

Sources

Bahamian population statistics: http://www.bahamas.gov.bs/wps/portal/public/Key.

Fiji Bureau of Statistics: www.statsfiji.gov.fj.

References

Alo, M.A. & R. Mesthrie. 2004. “Nigerian English: Morphology and syntax”. A Handbook of Varieties of English. Vol. II: Morphology and Syntax, ed. by B. Kortmann & E.W. Schneider,813–827. Berlin: Mouton de Gruyter.

Ansaldo, U. 2004. “The evolution of Singapore English: Finding the matrix”. Singapore English: A Grammatical Description, ed. by L. Lim, 127–149. Amsterdam: Benjamins.

Beckford Wassink, A. 1999. “Historic low prestige and seeds of change: Attitudes toward Jamaican Creole”. Language in Society 28: 57–92.

Biewer, C. 2009. “Modals and semi-modals of obligation and necessity in South Pacific Englishes”. Anglistik 20(2): 41–55.

Biewer, C. Submitted. South Pacific Englishes. The Dynamics of Second-language Varieties in Fiji, Samoa and the Cook Islands. Postdoctoral thesis, University of Zurich.

Biewer, C., M. Hundt, & L. Zipp. 2010. “‘How’ a Fiji corpus? Challenges in the compilation of an ESL ICE component”. ICAME Journal 34: 5–23.

Blouet, O.M. 2002. “The Caribbean”. Latin America and the Caribbean: A Systematic and Regional Survey, ed. by B.W. Blouet & O.M. Blouet, 4th ed., 311–366. New York: Wiley.

Carrington, L.D. 2001. “The status of Creole in the Caribbean”. Due Respect: Essays on English and English-Related Creoles in the Caribbean in Honour of Professor Robert Le Page, ed. by P. Christie, 25–36.Mona: University of the West Indies Press.

Central Intelligence Agency. 2010. “Jamaica”. The World Factbook. https://www.cia.gov/library/publications/the-world-factbook/geos/jm.html

Chand, V. 2009. “[v]at is going on? Local and global ideologies about Indian English”. Language in Society 38: 393–419.

Coates, J. 1983. The Semantics of the Modal Auxiliaries. London: Croom Helm.

Collins, P.C. 2007. “Can/could and may/might in British, American and Australian English: A corpus-based account”. World Englishes 26: 474–491.

Collins, P.C. 2009a. Modals and Quasi-Modals in English. Amsterdam: Rodopi.

Collins, P.C. 2009b. “Modals and quasi-modals in world Englishes”. World Englishes 28: 281–292.

Deuber, D. 2009a. “Caribbean ICE corpora: Some issues for fieldwork and analysis”. Corpora: Pragmatics and Discourse. Papers from the 29th International Conference on English Language Research on Computerized Corpora (ICAME 29), ed. by A.H. Jucker, M. Hundt & D. Schreier, 425–450. Amsterdam: Rodopi.

Deuber, D. 2009b. “’The English we speaking’: Morphological and syntactic variation in educated Jamaican speech”. Journal of Pidgin and Creole Languages 24: 1–52.

Deuber, D. 2009c. “Standard English in the secondary school in Trinidad: Problems – properties – prospects”. World Englishes – Problems, Properties and Prospects: Selected Papers from the 13th IAWE Conference, ed. by T. Hoffmann & L. Siebers, 83–104. Amsterdam: Benjamins.

Deuber, D. 2010a. “Modal verb usage at the interface of English and a related Creole: A corpus-based study of can/could and will/would in Trinidadian English”. Journal of English Linguistics 38: 105–142.

Deuber, D. 2010b. “Standard English and situational variation: Sociolinguistic considerations in the compilation of ICE-Trinidad and Tobago”. ICAME Journal 34: 24–40.

Deuber, D. and Youssef, V. 2007. “Teacher language in Trinidad: A pilot corpus study of direct and indirect creolisms in the verb phrase”. Proceedings from the Corpus Linguistics 2007 Conference. http://www.birmingham.ac.uk/documents/college-artslaw/corpus/conference-archives/2007/31Paper.pdf

Deuber, D., C. Biewer, S. Hackert, and M. Hilbert. 2012. “Will and would in selected New Englishes: General and variety-specific tendencies”. Mapping Unity and Diversity in Englishes Around the World. Corpus-Based Studies of New Englishes, ed. by M. Hundt & U. Gut. Amsterdam: Benjamins.

Facchinetti, R. 2002. “Can and could in contemporary British English: A study of the ICE-GB corpus”. New Frontiers of Corpus Research: Papers from the Twenty First International Conference on English Language Research on Computerized Corpora Sydney 2000, ed. by P. Peters, P.C. Collins & A. Smith, 229–246. Amsterdam: Rodopi.

Geraghty, P, F. Mugler & J. Tent. 2006. Macquarie Dictionary of English for the Fiji Islands. Sydney: Macquarie Library.

Görlach, M. 1991. Englishes: Studies in Varieties of English 1984–1988. Amsterdam: Benjamins.

Gupta, A.F. 1994. The Step-Tongue: Children’s English in Singapore. Clevedon: Multilingual Matters.

Gupta, A.F. 2006. “Epistemic modalities and the discourse particles of Singapore”. Approaches to Discourse Particles, ed. by K. Fischer, 243–263. Amsterdam: Elsevier.

Hackert, S. 2004. Urban Bahamian Creole: System and Variation. Amsterdam: Benjamins.

Hackert, S. 2010. “ICE Bahamas: Why and how?” ICAME Journal 34: 41–53.

Hilbert, M. & M. Krug. 2010. “The compilation of ICE-Malta: State of the art and challenges along the way”. ICAME Journal 34: 54–63.

Holm, John A. & Alison Shilling. 1982. Dictionary of Bahamian English. Cold Spring, NY: Lexik House.

Holmes, J. 1996. “The New Zealand spoken component of ICE: Some methodological challenges”. Comparing English Worldwide: The International Corpus of English, ed. by S. Greenbaum, 163–181.Oxford: Clarendon.

Huber, M. & K. Dako. 2004. “Ghanaian English: Morphology and syntax”. A Handbook of Varieties of English, vol. II: Morphology and Syntax, ed. by B. Kortmann & E.W. Schneider, 854–865. Berlin: Mouton de Gruyter.

Huddleston, R. & G.K. Pullum. 2002. The Cambridge Grammar of the English Language. Cambridge: CUP.

Krug, M. 2000. Emerging English Modals: A Corpus-Based Study of Grammaticalization. Berlin: Mouton de Gruyter.

Labov, W. 1984. “Field methods of the Project on Linguistic Change and Variation”. Language in Use: Readings in Sociolinguistics, ed. by J. Baugh & J. Sherzer, 28–53. Englewood Cliffs, NJ: Prentice-Hall.

Leech, G., M. Hundt, C. Mair & N. Smith. 2009. Change in Contemporary English: A Grammatical Study. Cambridge: CUP.

Lim, L. and J.A. Foley. 2004. “English in Singapore and Singapore English: Background and methodology”. Singapore English: A Grammatical Description, ed. by L. Lim, 1–18. Amsterdam: Benjamins.

Mair, Christian. 2009. “Corpus linguistics meets sociolinguistics: The role of corpus evidence in the study of sociolinguistic variation and change”. Corpus Linguistics: Refinements and Reassessments, ed. by Antoinette Renouf & Andrew Kehoe, 7–32. Amsterdam: Rodopi.

Mühleisen, S. 2001. “Is ‘Bad English’ dying out? A diachronic comparative study of attitudes towards Creole versus Standard English in Trinidad”. PhiN 15: 43–78. http://web.fu-berlin.de/phin/phin15/p15t3.htm

Nelson, G. 1996. “The design of the corpus”. Comparing English Worldwide: The International Corpus of English Greenbaum, ed. by S. Greenbaum, 27–53.Oxford: Clarendon.

Nelson, G. 2003. “Modals of obligation and necessity in varieties of English”. From Local to Global English: Proceedings of the Style Council 2001/2, ed. by P. Peters, 25–32. Sydney: Macquarie University, Dictionary Research Centre.

Palmer, F.R. 1990. Modality and the English Modals. 2nd ed. London: Longman.

Sand, A. 1999. Linguistic Variation in Jamaica: A Corpus-Based Study of Radio and Newspaper Usage.Tübingen: Narr.

Schmied, J. 1996. “Second-language corpora”. Comparing English Worldwide: The International Corpus of English, ed. by S. Greenbaum, 182–196. Oxford: Clarendon.

Schneider, E.W. 1999. “Notes on Singaporean English”. Form, Function and Variation in English: Studies in Honour of Klaus Hansen, ed. by U. Carls & P. Lucko, 193–205. Bern: Lang.

Schneider, E.W. 2007. Postcolonial English: Varieties around the World. Cambridge: CUP.

Sedlatschek, A. 2009. Contemporary Indian English: Variation and Change. Amsterdam: Benjamins.

Shields-Brodber, K. 1997. “Requiem for English in an ‘English-speaking’ community: The case of Jamaica”. Englishes around the World, vol. II: Caribbean, Africa and Australasia. Studies in Honour of Manfred Görlach, ed. by E.W. Schneide, 57–67. Amsterdam: Benjamins.

Solomon, Denis. 1993. The Speech of Trinidad: A Reference Grammar. St. Augustine, Trinidad: University of the West Indies School of Continuing Studies.

Tent, J. 2001. “A profile of the Fiji English lexis”. English World Wide 22: 209–245.

Tent, J. & F. Mugler. 2008. “Fiji English: phonology”. Varieties of English, vol. III: The Pacific and Australasia, ed. by K. Burridge & B. Kortmann, 234–266. Berlin: Mouton de Gruyter.

Tickoo, M.L. 1996. “Fifty years of English in Singapore: All gains, (a) few losses?” Post-Imperial English: Status Change in Former British and American Colonies, 1940–1990, ed. by J.A. Fishman, A.W. Conrad & A. Rubal-Lopez, 431–455. Berlin: Mouton de Gruyter.

Trudgill, P. & J. Hannah. 2002. International English: A Guide to Varieties of Standard English. London: Arnold.

Vaish, V. 2008. Biliteracy and Globalization: English Language Education in India. Clevedon: Multilingual Matters.

Winer, Lise. 1993. Trinidad and Tobago. Amsterdam: Benjamins.

Youssef, V. 1996. “Varilingualism: the competence underlying codemixing in Trinidad and Tobago”. Journal of Pidgin and Creole Languages 11: 1–22.

Youssef, V. 2004. “’Is English we speaking’: Trinbagonian in the twenty-first century. Some notes on the English usage of Trinidad and Tobago”. English Today 80: 42–49.