The present section aims at pointing out which factors can influence the use of GET over time as well as across different regional varieties of English. I claim that five factors are of special importance. First, GET has attracted prescriptive resistance, yet this resistance might be waning to a different extent in different varieties. Second, GET is perceived as colloquial and informal in style. This will have effects on its use across mode and genre, and different regional varieties are expected to exhibit different distributions. Third, it can be assumed that in multilingual settings such as the former British colonies where English is spoken today, the respective substrate varieties influence standard language use. Fourth, effects of second language acquisition (SLA), i.e. effects independent of specific substrates, must also be considered as a factor influencing the use of GET in New Englishes. Finally, the two major standard varieties, British and American English, display characteristics in the use of GET, and, depending on their power of acting as models, other varieties will be under their influence in the use of GET to a stronger or lesser extent. The five factors will be expanded on in the following.
Crystal defines the term prescriptive as follows:
A term used by linguists to characterize any approach which attempts to lay down rules of correctness as to how language should be used. Using such criteria as purity, logic, history or literary excellence, prescriptivism aims to preserve imagined standards by insisting on norms of usage and criticizing departures from these norms. (2008: 384, capitalisation removed)
Leech et al. point out the negative connotations of prescriptive because it suggests “doctrinaire linguistic attitudes” and bigotry, and they prefer the more neutral term language prescription to denote “any conscious efforts to change the language habits of English speakers (or more often, writers), whether bigoted or not” (2009: 263). Curzan (2014: 24–25) suggests that four strands of prescriptivism should be distinguished: standardising prescriptivism, stylistic prescriptivism, restorative prescriptivism, and politically responsive prescriptivism. Prescriptive usage guides in the native Anglophone world no longer need to specify what the standard forms are and now mainly put down stylistic rules, which “aim to differentiate among (often fine) points of style within standard usage” (2014: 24). In New Englishes, questions of standardisation may well be an issue, but in the case of the variation of the verb GET, the most relevant strand of prescriptivism is stylistic prescriptivism.
According to Crystal (2004: 523, quoted in Leech et al. 2009: 263), institutional prescriptivism has tapered off since the end of the 20th century at the latest. However, this applies to a different extent to different varieties of English, with American English being still much more influenced by prescriptivism than other inner-circle varieties, for instance in the form of condemnations of the passive and of relative which in restrictive relative clauses. The prescriptive rules are conveyed via style guides, handbooks at universities, editorial rule books of publishing houses and newspapers, glosses in the press, and grammar checking tools on computers (cf. Leech et al. 2009: 264). A famous usage guide is The Elements of Style, first published in 1959 by Strunk and White and targeting young American college students. It identifies “the principal requirements of plain English style” (1959: v–vi), contains rules of usage, principles of composition, and a chapter on commonly misused words and expressions. It has sold more than ten million copies. One can assume that the extent to which varieties worldwide orient towards American English also has an influence on how they deal with prescriptive rules. In Britain, usage guides are in general less popular, but Fowler’s Modern English Usage in its three editions (cf. Busse and Schröder 2010a) has become a household name. It has been commercially successful from the beginning in the 1920s, was even issued in a special impression for the United States, was republished twice and has enjoyed the status of an authority until today (cf. Busse and Schröder 2010a: 47–48). It is organised as a dictionary, but includes issues of grammar, style, and pronunciation. The 1996 version by Burchfield is heavily revised, with a change from pithy, prescriptive statements towards descriptivism (cf. Busse and Schröder 2010a: 50).
It is a general claim that prescriptivism is waning, particularly in innercircle varieties, due to the influence of new forms of communication such as e-mail and text-messaging, and the importance of spoken mass media (cf. Collins and Yao 2013: 499). Outer-circle varieties are probably in general more affected by prescriptive rules than inner-circle varieties because they are less stabilised and therefore more susceptible to putative standards of correctness. While prescriptivism is present in the form of usage guides in inner-circle varieties, in outer-circle varieties it often comes in the form of the complaint tradition (cf. Collins and Yao 2013: 499; Schneider 2007: 43). In the case of Singaporean English, however, the complaints, as voiced, for instance, in the “Speak Good English” campaigns, have led to quite a liberal attitude towards prescriptivism because of a strong backlash against them (cf. Collins and Yao 2013: 491).
GET has attracted the attention of prescriptivists, the reason being its versatility, which entails that it is not particularly well-defined semantically and, without context, vague in meaning – but not before the 20th century. Although the 19th century is known as the prescriptive century and saw a strong increase in the use of GET overall, GET-constructions were not yet commented on in British or American grammars of the time (cf. Anderwald 2012a: 31–32). In Strunk and White (1959: 39), one only finds the recommendations for avoiding have got in writing and for using the participle got, not gotten. According to Burchfield in The New Fowler’s Modern English Usage, his “teachers were hostile to this small and useful verb” (1996: 329), asking the students to substitute GET with other verbs. While Burchfield points out that lexicographers mark many uses of GET as colloquial or informal, he makes it clear that many idiomatic phrases would not exist without GET, e.g. get along with, get away with, get down to, etc., and that in many combinations, GET is naturally used and goes unnoticed, e.g. in get a job, get my book for me, get rich, etc. (1996: 330).
Researchers who attempt to measure prescriptive influence in the history of English look for interdependencies between comments in prescriptive grammars and actual language use in corpora. With the help of a precept corpus and a usage corpus, Auer and González-Díaz (2005) demonstrate that 18th-century prescriptivism has been overestimated as a factor in language change. Although grammarians’ prescriptive comments had an influence on the use of the subjunctive in British English, this happened only with a certain time lag and, ultimately, the decline of the subjunctive was not prevented (cf. Auer and González-Díaz 2005: 322–324).
Anderwald (2013, 2012b), in her study based on data from the Corpus of Historical American English and on analyses of over 250 historical grammar books, finds that the regularisation of selected irregular past-tense verb forms and the irregularisation of selected regular past-tense verb forms in American English from the 19th century to today are not the result of grammarians prescribing a certain usage but occurred as changes from below. With a certain time lag, these changes were then documented in the grammar books of the time, which were quite liberal, allowing the use of several alternative past tense forms rather than promoting the use of one specific form. Anderwald concludes that there is “strong counterevidence of the presumed general success of prescriptivism in the 19th century” (2013: 168).
Bohmann and Schultz (2011: 96–98) detect a correlation between prescriptive dicta on the use of the relative pronoun that in non-animate restrictive relative clauses and the actual occurrence of the relativiser in diachronic data from the 20th century, particularly in American English. However, high frequencies of that also correlate with high frequencies of stranded prepositions, which runs counter to predictions and leads Bohmann and Schultz to infer that different prescriptive rules do not have equally strong bearing and that text types need to be more strongly considered in the discussion on prescriptivism and language change.
To sum up, although correlations can be found, current research shows that the influence of prescriptive dicta on actual usage in historical language corpora has been overrated. Modern usage guides, e.g. The Cambridge Guide to English Usage (Peters 2004, quoted in Busse and Schröder 2010b: 92–93), are largely descriptive, based on different kinds of evidence and authority, and no longer reflect the personal preference of the editor. While the traditional prescriptivist should not be written off, particularly not in the US, the influence of electronic media and spoken mass media has grown, and this means that more dominance is given to spontaneous uses of language, which are less within the reach of prescriptivism (cf. Leech et al. 2009: 264). Prescriptivism will in general have more influence on written language because spoken language is being less monitored. As will be discussed in detail in the respective individual chapters, prescriptivism becomes a special focus when considering overall token frequencies of GET, mode and genre differences, the GET-passive, GET-PVs, and whenever American English influence is assumed.
Many uses of GET are marked as colloquial and informal by lexicographers (cf. Burchfield 1996: 330). The Oxford Advanced Learner’s Dictionary of Current English (2005: s.v. “get”), for instance, notes that some people try to avoid GET in formal writing. This is only natural because the versatility and flexibility of GET entail a certain vagueness, which is less acceptable in carefully formulated written language than in spontaneous speech. The colloquial character of GET will therefore have strong effects on how it is distributed across spoken and written language and across more formal and more informal text types within either mode. In the case of very versatile verbs, such as GET and SHOW, register dependence is particularly great (cf. Schilk 2011: 32), which is why data analyses sensitive to mode and text type will play a major role in the present study. Furthermore, although it has been claimed that the stratifications of register are relatively uniform across regional varieties of English (cf. Zipp and Bernaisch 2012: 169), it should be considered that quantitative differences between World Englishes might also be due to style. Collins and Yao’s analysis (2013: 479, 500), for instance, confirms a tendency towards stylistic homogeneity in outer-circle varieties noted by some previous scholars, reflected in relatively small differences between text types.
Following Collins and Yao (2013: 480), I will use the term colloquialisation as a process-oriented concept for diachronic data, and colloquialism as a stateoriented concept for synchronic data. Colloquialisation has been traditionally defined as a “general stylistic shift in English that has been operating to make written genres more like spoken ones” (Collins and Yao 2013: 480; also cf. Leech et al. 2009: 239). Colloquialisation is a trend claimed to be affecting the English language in general, and in the case of GET and the present study, it would have to be reflected in an increase in use from LOB to FLOB in informal written genres.
However, I will follow Collins and Yao in their claim, based on their analyses of several colloquial features across a number of varieties of English, that colloquialisation is
[…] a process involving a spreading of colloquial features from casual face-to-face conversation to other genres, not only written (e.g. personal letters, examination scripts, newspaper reportage, fiction, academic writing) but also spoken (e.g. public dialogues, scripted and unscripted monologues). (2013: 480)
This means that the extent of colloquial grammatical features, i.e. colloquialism, can also be determined for spoken text types. Private dialogues have emerged as “typical speech” in Collins and Yao’s analysis (2013: 487) and are sharply differentiated from all other genres, with informational writing constituting “typical writing” and exhibiting the least colloquial features. In between the two extremes, fiction has turned out to be the second most colloquial genre before monologues and non-printed writing.
Collins and Yao’s analysis (2013: 498–500) indicates that, in general, outercircle varieties display less colloquialism than inner-circle varieties, measured with the help of the parameters contraction of verbs and negatives, GET-passives and semi-modals. However, Singaporean English displays a relatively advanced stage of colloquialisation, while British English is very conservative. To be more precise, Singaporean English and British English represent the most colloquial and the least colloquial members of their respective circle, so that the two varieties can be said to be on a par in regard to colloquialism. Jamaican English was not part of Collins and Yao’s investigation. As far as the influence of American English on other varieties is concerned, it remains to be seen how colloquialism and prescriptivism interact. On the one hand, American English leads many changes towards colloquialism (cf. Collins and Yao 2013: 481). On the other hand, prescriptivism can have an inhibiting effect and reduce the amount of colloquialism. The quantitative variational analysis of GET will provide more information on how tolerant or intolerant of colloquialism and informality British, Jamaican, and Singaporean English are in their use of the individual GET-constructions.
In Schneider’s Dynamic Model of the spread of English (2007: 56), substrate influence surfaces in phase 3 (nativisation) and can lead to a growing distinction between the input variety and the newly emerging variety. As far as the New Englishes focussed on in the present study are concerned, in Jamaica, influence from Jamaican Creole on Jamaican English can be postulated. Sinitic and Malay varieties are the substrates of Colloquial Singapore English and thus likely to influence Standard Singaporean English, at least indirectly via Colloquial Singapore English.
A look at the entries for GET in dictionaries of Jamaican Creole and Colloquial Singapore English might provide a first hint about substrate influence on GET. The Dictionary of Jamaican English (cf. Cassidy and LePage 1980: 196) only lists intransitive uses and differentiates two meanings, viz. ‘to arrive (at a place)’, as in […] till at last them get, and ‘to succeed (in doing sth)’, as in […] im jos kyaan get at aal. A second entry lists the meaning ‘to be begotten’, as in mi get hie ‘I was conceived here’. The Dictionary of Caribbean English Usage aims to depict “[t]he literate English of educated nationals of Caribbean territories and their spoken English such as is considered natural in formal social contexts” (Allsopp 1996: lvi) and thus contains much more information on the general use of GET. Apart from its use as a transitive verb meaning ‘to have, to own, to possess’ or ‘to acquire specially’, and its use as an intransitive verb to form passives with a following verb in the base form, the bulk of the listings consists of various idiomatic phrases, which range from PVs to fixed idioms (cf. Allsopp 1996: 253–254).
The Times-Chambers Essential English Dictionary (1997) is a prescriptive learner’s dictionary containing 1,000 items from the English used in Singapore and Malaysia, and intended to provide practical guidance to speakers of ESL in those places, for instance in the form of usage notes or spelling help. This dictionary is not available, but it can be assumed that the item GET is not listed since the dictionary mainly contains distinctively Singaporean lexemes (cf. Kubota 1998: 160–161; Schneider 1999: 201–203). There is no other printed dictionary of Singaporean English, and dictionaries of Singlish available on the Internet do not list GET.
It becomes clear that in order to trace substrate influence, the individual constructions in which GET occurs must be considered. Thus, whenever substrate influence on a GET-construction can be assumed, I will, in the respective chapter, describe the situation in the substrates or local varieties used in Jamaica and Singapore. If substrate influence becomes apparent, I will determine the precise origin. In this context, it should be kept in mind that in the standard language corpora of ICE, hardly any direct substrate influence will become noticeable in the use of GET. Rather, the aim will be to trace indirect influence from local languages in two ways. First, substrate influence will be reflected in quantitative differences of standard features. Second, there will be substrateinfluenced features that point to innovation and nativisation but which cannot yet be said to be entrenched, the reason being that they occur in low numbers. In the present study, substrate influence will turn out to be especially relevant for monotransitive GET, GET-passives, GET-existentials, possessive (HAVE) got, semi-modal (HAVE) got to, and catenative GET.
In the description of characteristic features of New Englishes, so far, many researchers have focussed on substrate effects, and insights from SLA (second language acquisition),5 used here to refer to “the acquisition of any language after the acquisition of the mother tongue” (Ellis 2008: 6), have not taken centre stage. However, New Englishes are clearly used as second languages in the majority of cases, so that effects of SLA, by which I understand learning effects that are independent of specific substrates, need to be addressed more prominently.6 This applies especially to an analysis of GET-constructions, for which I claim that general second-language learning strategies will influence variation at least as much as substrate effects. In this context, it should be considered that substrate influences affecting grammar and lexico-grammar are more difficult to trace than those affecting phonology and the lexicon.
Unlike the case for substrates, where the specific multilingual setting has to be taken into account, effects of SLA are assumed to be universal, but they can, of course, also apply to different degrees to different varieties. Williams (1987: 165–168) points out that some subsystems of the grammar of English, so-called weak links, seem to pose inherent difficulties for learners, and that psycholinguistic production and perception principles govern language acquisition in New English contexts just as in other learner contexts. Language contact can magnify these problem areas, ultimately leading to a development away from them (cf. Sand 2008: 200). Examples of principles of SLA are regularisation, overgeneralisation, maximisation of salience, and simplification (cf. Williams 1987: 169, 188; also cf. Hoffmann 2011: 104–105), but also the overuse7 of known forms. Since the two latter processes, viz. simplification and overuse, will turn out to be particularly relevant to the present study, a more detailed explanation is in order at this point.
When one speaks of simplification, a standard of comparison has to be referred to. In the case of New Englishes, this is British English, which is putatively more complex than the second-language variety. There are two measures for linguistic complexity: measures of absolute complexity are objective and refer to the number of parts in a system, while measures of relative complexity are subjective and refer to the processing difficulty of a language and the production cost to language users (cf. Szmrecsanyi and Kortmann 2012: 10). The measures cannot be strictly separated, however, and while the present study is oriented towards relative complexity, it combines the two. I quantify the uses of GET-constructions in the different varieties, but the overall range of GET-constructions – or the number of constituents in the system – is equal in all varieties (cf. Figure 5.2), if one leaves aside the range of combinations that certain constructions allow, e.g. GET-PVs, or special uses, e.g. the zero-subject-got-existential. What is of prime interest in the present study is the question to what extent speakers of New Englishes avoid or underuse certain GET-constructions and possibly prefer others because they are second-language learners and perceive certain constructions to be difficult, in contrast to first-language speakers of English. The kind of complexity in focus here is L2-acquisition complexity, “the degree to which a language or language variety (or some aspect of a language or language variety) is difficult to acquire for adult language learners” (Szmrecsanyi and Kortmann 2012: 12; also cf. Trudgill 2001: 371; Kortmann and Szmrecsanyi 2009: 269).
Thus, a simple structure or use is here defined as one which requires little cognitive processing effort of the learner. This can manifest itself in various ways. For instance, simplification is apparent in the reduction of the number of units in a system (cf. Schneider 2007: 102) or in shorter structures. Yet it has also been claimed that greater simplicity can be achieved by the so-called isomorphism principle, i.e. the transparent and iconic one-to-one mapping of conceptual structure and surface form (cf. Steger and Schneider 2012: 156). This principle, however, can have the effect that in a second-language variety more language material is used to render a notion than in the parent variety (also cf. Mesthrie 2012). New Englishes are assumed to be particularly affected by iconicity effects because iconic structures are easier to learn and iconicity-destroying effects such as lexicalisation have had less time to exert their power (cf. Steger and Schneider 2012: 163). In the individual sections of the present study, the precise effects will be explained in more detail.
GET could also be a candidate for systematic overuse by advanced learners because it is one of those words that learners can feel safe with: it remains vague enough in any context. Hasselgren explicitly refers to GET as a favourite core item, “overused at the expense of such words as gain, where considerations of collocation or style are deciding factors as to whether a wrong choice has been made” (1994: 242). She has found that advanced Norwegian learners of English choose core verbs, among them GET and GIVE, more frequently than native speakers. Other studies, too, have shown that learners favour high-frequency general verbs that are usable in many contexts over more specific verbs (cf. Nesselhauf 2005: 86; Mukherjee 2009: 131n9). What adds to the overuse of highly frequent lexemes even among advanced language users is that these lexemes are not perceived or marked as errors (1994: 250). As Hasselgren formulates it, speakers “get away with them” (1994: 251). Polysemous high-frequency words that are overused can be called lexical teddy bears, defined by Hasselgren (1994: 237) as the words one clutches for and feels safe with when using a second language. In more general terms, any overuse of well-known forms can be called teddy bear effect (cf. Hasselgren 1994; also cf. Gilquin 2009, quoted in Zipp and Bernaisch 2012: 170) because language learners cling to familiar forms like children to their teddy bears. This means that the phenomenon of overuse might only affect certain uses or constructions into which a lexeme enters, which need not necessarily be reflected in a general overuse of this lexeme.
Issues of SLA, especially simplification, will play a prominent role in the chapters on token frequencies, monotransitive and ditransitive GET, GET-PVs, GET as a verb of motion, possessive (HAVE) got, semi-modal (HAVE) got to, catenative GET, and GET-chunks. The teddy bear effect will turn out to be particularly relevant in the use of monotransitive GET, GET-PVs, and GET-idioms in Jamaican English.
As a consequence of British colonisation, British English is the parent variety and original base of almost all Englishes spoken in ESL countries today. In the process of nativisation, the new varieties develop their own features, for instance due to the influence from local languages. Yet the orientation towards Standard British English can function as an important factor preventing or counteracting increasing endonormative stabilisation. In every speech community, there will be speakers who are still more oriented towards the (former) exonormative target model and retain as many features of British English as possible, and others who more readily embrace features that lead to endonormative stabilisation (cf. Schneider 2007: 42). What must also be considered is that changes that have been going in the historical input variety might not be reflected in the use of English in the ex-colonies or reach it only after a certain time lag (cf. Schilk 2011: 15).
While British English is the input variety of both Jamaican and Singaporean English, there is widespread agreement that the second major standard variety of English, viz. American English, is increasingly exerting its influence on English spoken around the world, with the influence of the US economy and mass media as commonly cited reasons. Leech et al. (2009: 253) identify several patterns of diachronic variability and apply them to Americanisation: regionally specific change, convergent or divergent change, parallel change, different or similar rates of change, and follow-my-leader. These patterns of change are not mutually exclusive and can be combined. Leech et al. emphasise that direct dialect contact is not a prerequisite for supposedly American English forms to surface in British English or another variety. For instance, the follow-my-leader pattern, which means that American English is in the lead or changing more quickly than British English or another variety, occurs frequently, but one should not be tempted to explain it all too quickly by Americanisation. The pattern may be the result of a general trend which is already more advanced in American English than in British English or another variety, so that what looks like Americanisation is in fact only a parallel change occurring at different rates of change (2009: 254).
Although it can, of course, never be completely excluded that similar language use is due to independent regional developments rather than Americanisation, there are many reasons to claim that the cultural and economic influence of the US also has linguistic repercussions. Countries in close geographical vicinity and with strong links to the US, such as Jamaica, can be assumed to be more heavily influenced by it than others. Mair and Sand (n.d.) think that in Jamaica, American English influence is almost as powerful as that of British English today, given the influence of US business and media in the region as well as massive emigration, so that British English has become a norm more often proclaimed than followed. In the case of Singapore, the two inner-circle varieties might interestingly exert their influence in opposite directions. Lim (2012: 288) explicitly refers to American English influence on Singapore, particularly in the domain of the media, because the majority of Singapore’s TV programmes are American productions (also cf. Schneider 1999: 196–198; Lim and Ansaldo 2013b).
At the time it began to emerge, American English was sought as a variety for democratising the country (cf. Schneider 2007: 278). It has been shown that American English is leading the way in many recent developments of grammatical variation (cf. Collins 2009b: 284), such as the rise of semi-modals, and some of these, for instance the latter, can be seen to be in line with such a trend of democratisation and the lessening of an authoritarian tone (also cf. Leech et al. 2009: 259). British English, by contrast, has been found to be more conservative than American English in a number of recent developments. It goes without saying that calling British English a conservative variety and American English an innovative variety simplifies the more complex reality and neglects individual phenomena such as the revival of the mandative subjunctive in American English (cf. Leech et al. 2009: 69). The present study will therefore consider each type of GET-construction individually in order to determine whether the British English model is stronger or the power of American English to function as a linguistic epicentre (cf. Hundt 2013). The issue of the respective influence of the two major standard varieties will come into focus in the analysis of the following phenomena: the word-forms gotta and gotten, the GET-passive, and the surface form (HAVE) got (to), which means GET-existentials, possessive (HAVE) got, and semi-modal (HAVE) got to.