The verb-complementational profile of offer in Sri Lankan English [1]

Tobias Bernaisch
Department of English, University of Giessen

Abstract

1. Introducing Sri Lankan English: history and evolutionary status

Sri Lanka, in line with other countries in South Asia, is characterised by a number of indigenous languages being used on a day-to-day basis. In addition to those indigenous languages, English still continues to be a vital element in the multilingual setting of Sri Lanka even after the end of British colonial rule, which, however, does not find reflection in the amount of empirical research conducted on Sri Lankan English. While other South Asian varieties such as Indian English have already been studied from various linguistic perspectives (cf. e.g. Kachru 1983, Shastri 1988), Sri Lankan English has not yet received equal attention; in particular, corpus-based approaches to Sri Lankan English have been largely neglected so far. With the help of standard corpora and newspaper corpora, the present paper aims at describing one aspect of the lexis-grammar interface of Sri Lankan English. In order to get a better understanding of the socio-historical circumstances under which the Sri Lankan variety of English evolved, which also has implications for its present evolutionary status in Schneider’s (2003, 2007) model, a short overview of the history of Sri Lankan English will be given.

1.1 The history of Sri Lankan English: an overview

The present linguistic situation in Sri Lanka is characterised by three major players: Sinhala, Tamil and English. Sinhala, an Indo-Aryan language with strong historical roots in Sri Lanka, is spoken by 74 per cent of the Sri Lankan inhabitants while Tamil, a Dravidian language linked ethno-linguistically to the southern states of India, is spoken by 25 per cent of the population (cf. Dharmadasa 2007: 116).

After the Portuguese and the Dutch had colonised Sri Lanka in the 16th and 17th century respectively, British forces were subsequently able to establish themselves as the single rulers of the island. Along with the introduction of Western administrative and commercial structures, the British colonists and Anglican missionaries promoted the English language in Sri Lanka from 1815 onwards (cf. Dharmadasa 2007: 119f.).

Owing to the fact that “[t]he serious study of subjects such as science, politics and philosophy was [...] confined to mostly private English-medium education” (Dharmadasa 2007: 125) and only basic reading and writing skills were taught in the indigenous languages, English had a powerful status and was the ticket to well-paid, prestigious jobs almost until the middle of the 20th century (cf. Goonetilleke 2003: 338). In 1956, eight years after Sri Lankan independence, Sinhala was made the only official language of Sri Lanka – a declaration which is one of the main reasons for Sri Lanka’s history of postcolonial civil war (cf. Dharmadasa 2007: 133). As a result of radical and repeatedly violent reactions to this Sinhala-only policy, Tamil was promoted to a national language in the new constitution in 1978 and eventually, Tamil was recognised as “a national official language of Sri Lanka on a par with Sinhala (with English also officially reintroduced as a ‘link’ language)” (Dharmadasa 2007: 136). Although Sinhala is said to have taken over some of the roles formerly ascribed to English, Künstler et al. (2009) show that English continues to fulfil a wide range of core functions in Sri Lanka.

In the following, the evolutionary status of Sri Lankan English will be described on the basis of Schneider’s (2003, 2007) dynamic model of variety-formation (section 1.2). A methodological section (section 2) will introduce the corpus data as well as the syntactic and semantic frameworks according to which the corpus data of offer will be categorised. Section 3 will outline the results of the quantitative analysis based on the categories illustrated in section 2. Section 4 explains the quantitative findings with the help of a more fine-grained semantic perspective on the data while section 5 discusses the results at hand.

1.2. English in Sri Lanka or Sri Lankan English?

Although the present paper focuses on Sri Lankan English as a well-defined variety of English, it is essential to point out that English in Sri Lanka is characterised by a relatively high level of internal variation:

While it is true that for the vast majority of speakers, English in Sri Lanka constitutes an ‘institutionalised second-language variety’ (cf. Kachru 1985), there also exist a comparatively small group of speakers in Sri Lanka for whom English is the first (or native) language on the one hand and a substantial group of speakers that displays a low proficiency in English, which is a truly foreign language to them, on the other. (Künstler et al. 2009: 57f.)

Although Sri Lankan English can generally be considered to be an outer-circle variety in Kachruvian terms (cf. Kachru 1985), the sociolinguistic reality is much more complex. Sri Lankan English is thus, in line with the above observation, best perceived as a variety-internal miniature version of the three Kachruvian circles (cf. Mukherjee et al. 2010).

While Kachru’s (1985) three-circle model was designed to capture and categorise all existing varieties of English, Schneider’s (2003, 2007) dynamic model of the evolution of postcolonial Englishes (PCEs) is more restrictive in that it focuses exclusively on New Englishes. Schneider claims that

despite all obvious dissimilarities, a fundamentally uniform developmental process, shaped by consistent sociolinguistic and language-contact conditions, has operated in the individual instances of relocating and re-rooting the English language in another territory [...]. (Schneider 2007: 5)

Schneider (2007: 30) describes this evolutionary process as “five progressive stages”, which he labels “(1) foundation, (2) exonormative stabilization, (3) nativization, (4) endonormative stabilization, and (5) differentiation”. Each of these stages is characterised by historical, political, social and linguistic factors (cf. Schneider 2003: 255). In sum, Schneider’s (2003, 2007) dynamic model of the evolution of PCEs can be seen as a more fine-grained diachronic approach to the ongoing processes in Kachru’s outer-circle varieties.

When assessing the evolutionary status of Sri Lankan English with the help of Schneider’s (2003, 2007) criteria for endonormatively stabilised varieties, Mukherjee (2008: 361) comes to the conclusion that “Sri Lankan English is an institutionalised second-language variety of English which may well be on its way to endonormative stabilisation”. Schneider (2007: 50) highlights the importance of this transition from nativisation to endonormative stabilisation: “the difference between phases 3 and 4 is commonly given symbolic expression by substituting a label of the ‘English in X’ type by a newly coined ‘X English’”. However, in addition to Mukherjee’s (2008) observations, some noteworthy developments with implications for the evolutionary status of Sri Lankan English have set in only relatively recently. For that reason, Sri Lankan English merits an updated assessment concerning the criteria for endonormative stabilisation.

With regard to history and politics, Sri Lanka is undoubtedly an independent and self-dependent country. The criterion as to whether the settlers and the indigenous population are interwoven is not fulfilled in Sri Lanka because there are hardly any settlers left. This is illustrated by statistics showing that the Burgher population, i.e. people of mixed European origin, accounts for less than one percent of the approximately 17 million people living in Sri Lanka (cf. Department of Census and Statistics, Sri Lanka). The concept of a new panethnic identity in Schneider’s (2003, 2007) sense can thus not be attested for Sri Lanka, but given the elimination of radical Tamil forces in May 2009, which was officially declared to end the Sri Lankan civil war, it may be that a new panethnic identity will emerge out of the formerly sharply divided Sinhalese and Tamil ethnic communities. As regards the sociolinguistics of contact, use and attitudes, Künstler et al. (2009: 72) conclude in their recent study that “most functional and attitudinal requirements for the status of Sri Lankan English as an institutionalised second-language variety in its own right have been satisfied at least to some extent”. With regard to literary creativity, the rich collection of Sri Lankan English literature is well-documented in Goonetilleke (2005). While there is still a general lack of empirical data regarding linguistic developments and structural effects (cf. Mukherjee 2008: 361), Meyler’s (2007) A Dictionary of Sri Lankan English represents the first dictionary of Sri Lankan English and marks the beginning of the codification process of the English language as it is used in Sri Lanka (cf. Algama 2008). In addition, the Sri Lankan component of the International Corpus of English (ICE) will soon provide a representative basis for the empirical description of Sri Lankan English. Thus, from a theoretical perspective, one might conclude that Sri Lankan English is beginning to establish itself as an endonormatively stabilised variety of English.

Table 1 summarises the findings just described. In the column on the right-hand side, ‘+’ means that the present situation in Sri Lanka meets the criterion, ‘-’ shows that a criterion is not fulfilled and ‘?’ stands for a general lack of empirical data, which forbids an assessment of the criterion at hand. ‘(+)’ means that there is a tendency towards meeting the criterion.

However, if one argues that Sri Lankan English is on its way towards endonormative stabilisation, this implies that it has already completed (at least some of) the processes of structural nativisation that occur in phase 3 of Schneider’s (2003, 2007) framework. Yet, this is still very much open to debate since, as mentioned above, empirical and corpus-based research into Sri Lankan English has largely been neglected until now and needs to add to the relevant sociolinguistic perspectives. Specifically, the lexis-grammar interface of Sri Lankan English has not attracted much interest so far.

In this context, Mukherjee & Gries’ (2009) collostructional analysis demonstrates that the level of verb complementation offers valuable insights into the study of structural nativisation in New Englishes. Also Olavarría de Ersson & Shaw (2003: 138) stress the importance of verb complementation in the assessment of the evolutionary status of a given variety when they say that “[v]erb complementation is an all-pervading structural feature of language and thus likely to be more significant in giving a variety its character than, for example, lexis”. With a focus on structural features of Sri Lankan English, the present paper examines the verb-complementational profile of offer, a polysemous and syntactically variable verb, on the basis of corpus data. [2]

2. Methodology

The present verb-complementational study investigates the ditransitive verb offer. Although the underlying semantic structure of offer and the corresponding semantic roles are not always made entirely explicit in the surface structure, they are always implied in the concrete realisations of offer and can be described as follows. The acting entity (i.e. the agent from a semantic perspective and the subject (S) from a syntactic perspective) offers the goal of the action (i.e. the recipient and the indirect object (O_i)) an affected entity (i.e. the patient and the direct object (O_d)). As some semantic roles tend to be left implicit in given contexts, offer can be realised in various syntactic structures, i.e. ditransitively in the double object construction (pattern I) or in the prepositional dative construction (pattern II), monotransitively (pattern III) or intransitively (pattern IV). An overview of the patterns of offer and their respective passive equivalents is provided from (1) to (8a).

With the type-I pattern, all the semantic roles are realised syntactically. The type-IP pattern is the passive equivalent of the type-I pattern. With the type-IP pattern, the indirect object in active voice is transformed into the passivised subject while the subject in active voice can optionally be realised as a by-agent in the passive. The type-II pattern is also characterised by three semantic roles which are made explicit in the surface structure, but, in contrast to the type-I and the type-IP patterns, the indirect object is always realised as a prepositional phrase. The type-IIP pattern is the passivised equivalent to the type-II pattern with an optional by-agent. With pattern III, the recipient is left implicit and only two semantic roles, i.e. the agent and the patient, are realised syntactically and pattern IIIP, its passive counterpart, is characterised by an explicit patient and an optional by-agent. With the type-IV and the type-IVP patterns, none of the objects is made explicit. Variations of the patterns just introduced such as imperatives or constructions with fronted objects will also be categorised according to these rather wide definitions, but such alternative realisations will not be assigned particular labels. This syntactic framework derived from Mukherjee (2005) forms the basis of the syntactic analysis of the corpus data.

Bresnan & Hay (cf. 2008: 249f.) discuss various covariates which can exert an influence on the choice of a particular syntactic pattern, one of which is semantic class. Here, the patients play a central role since they show the range of semantic meanings with which a given verb can be associated. Empirical studies prove that semantic class is a central factor as regards the choice of syntactic patterns of ditransitive verbs in inner-circle (cf. Bresnan & Hay 2008: 251) as well as outer-circle varieties (cf. Schilk 2009). Based on Bresnan & Hay’s (2008) approach to semantic class, the four central meanings of offer were established by examining the affected (or offered) entities that can be used with offer in five corpus-based dictionaries. [7] The results of this analysis are described in Table 2 and illustrated with examples from the Collins COBUILD Advanced Learner’s English Dictionary (Sinclair 2006).

The present paper examines whether or not identifiable associations of a particular meaning of offer and the preferred pattern of that meaning are stable across varieties of English. Apart from Sri Lankan English, British English, the historical input variety of Sri Lankan English, and Indian English, which may exert an epicentral influence on the South Asian region (cf. Leitner 1992: 225), are taken into account. The data to be analysed stem from two groups of corpora. The first group consists of the ICE corpora representing the three varieties. Owing to the fact that the Sri Lankan component of ICE (ICE-SL) is still being compiled, a pilot corpus, ICE-SL [W130] – including 130 written texts – is used; the relevant Indian (ICE-IND [W130]) and British (ICE-GB [W130]) corpora follow the same corpus design to ensure maximum comparability. [8] The design of the three ICE [W130] corpora is illustrated in Table 3.

The strength of the ICE corpora consists in the high level of representativeness due to the extensive selection of genres from which corpus texts are drawn. The fact that the ICE project will ultimately represent approximately 25 ENL as well as ESL varieties of English (cf. Mukherjee 2009: 48) enables cross-varietal investigations with a high level of comparability due to a common corpus design and mark-up manual. Nevertheless, it is also critical to have an awareness of the limitations of the ICE project. In order to keep the compilation of regional standard subcorpora feasible, the number of words is limited to one million for each regional variety, which implies that lexical investigations might not be as revealing as e.g. lexicogrammatical or syntactic queries. Furthermore, the emerging time gap between and within the subcorpora as well as diverging understandings of the text categories and the mark-up manual by the various ICE compilation teams do not facilitate the interpretation of ICE-based search results (cf. Mukherjee et al. 2010).

The second group of corpora consists of relatively large newspaper corpora. With the help of an adopted version of the ‘webpage-to-megacorpus’ method (cf. Hoffmann 2007), two South Asian web-derived newspaper corpora were created. [9] Using data from the online archives of the Daily Mirror, the Sri Lankan newspaper corpus (DM) features 1,518,726 tokens. The Indian English equivalent based on The Times of India (TOI) comprises 1,521,388 tokens. The newspaper data from the periodicals section of the BNC (BNC per) contains 8,992,587 tokens. The methodological advantages of the newspaper corpora at hand are their size in comparison to the preliminary ICE corpora as well as their varietal purity since news reports from news agencies have been systematically removed from the respective databases. In addition to that, the study of newspaper data in ESL contexts is particularly interesting for the observation of emerging norms in the respective varieties since newspapers, given the unavailability of dictionaries and grammars for a noticeable amount of New Englishes, may fulfil standardising functions in these varieties (cf. Schilk 2012: 47). Still, the restriction of the corpus data to one (though relatively diverse) written text category and editorial interventions are certainly aspects which have to be taken into consideration with certain investigations.

From each ICE [W130] corpus, an exhaustive concordance with all the realisations of offer was created while from the larger newspaper corpora, 300 randomly selected instances of offer per newspaper corpus were obtained and analysed. Each of the concordance lines was coded according to the syntactic and the semantic frameworks described above. The results of this coding process formed the input for the quantitative data analysis.

3. Quantitative analysis of the corpus data

The aim of the quantitative analysis is to encounter relevant aspects of the verb-complementational profile of offer, which will then be inspected more closely. The first step of the data analysis establishes the preferred pattern of each meaning of offer in the three varieties under scrutiny. The preferred pattern of a given meaning of offer is the syntactic pattern which is used most frequently with that particular meaning in a given corpus. The results of the analysis of the ICE data are illustrated in Table 4. [10]

The data in Table 4 show that the type-III pattern is the syntactic construction in which offer is realised most frequently in each ICE database under scrutiny. In order to statistically test the cross-varietal stability of the associations of a given meaning of offer and its preferred pattern, the chi-square test is the appropriate testing method for the datasets at hand. However, for the ICE [W130] data, it is not possible to make statistically reliable statements on the stability of the associations across varieties as the expected frequencies of each of the three patterns are merely too low because of the relatively low total number of occurrences of offer in the ICE [W130] corpora. [11]

As shown in Table 5, the results of the newspaper corpora are even more consistent with regard to the preferred patterns of each meaning.

Pattern III is the preferred syntactic choice of each meaning of offer in all the newspaper corpora investigated. [12] Consequently, the absolute frequencies of the type-III pattern as illustrated in Table 6 are used as input data for the chi-square test. In Table 6, the absolute frequencies of pattern III have been re-arranged according to the meanings of offer and the corpora analysed in order to test the cross-varietal stability of the associations.

The quantitative analysis shows that there is a statistically highly significant, but relatively weak correlation (as expressed by Cramer’s V) between the associations of the meanings of offer and their preferred pattern (type-III) across the varieties in the newspaper corpora. The association plot in Figure 1 draws attention to particularly interesting associations of the meanings of offer and the type-III pattern. The black and white boxes match positive and negative Pearson residuals and represent observed frequencies which are greater and smaller than the expected frequencies respectively. The size of the boxes behaves proportionally to the differences in observed and expected frequencies (cf. Gries 2009: 198). Thus, the associations that stand out in Figure 1 are the associations of the abstract meaning of offer and the type-III pattern in DM and TOI and the associations of the informational meaning of offer and pattern III in DM and TOI. A post-hoc test investigating the contributions to chi-square shows that the statistically significant effect originates from the fact that Sri Lankan English does not frequently use the type-III pattern in connection with the meaning inf while Indian English uses the type-III pattern with the meaning inf significantly more often. [13] [14]

Before delving more deeply into the analysis of the origin of the statistically significant association of informational offered entities and monotransitive realisations of offer, one can summarise the findings so far as follows: Although there is one exception in the ICE [W130] data, the syntactic pattern III is generally found to be the preferred pattern for each individual meaning of offer in each of the corpora investigated. [15]

4. Semantic subsets of informational offered entities in monotransitive patterns of offer

The statistically significant effect of the association inf & III in the newspaper corpora can be explained by a more fine-grained look at the data. Fulfilling the premise of being verbal messages in the spoken or written medium, the offered entities used with this association can be grouped into three semantic subsets. These subsets are ‘religious’ (offering verbal messages to a deity (e.g. prayers <TOI 37330>)), ‘educational’ (offering instruction (e.g. courses <TOI 37459>)) and ‘various’ (offering personal views, explanations and arrangements (e.g. regrets <TOI 38094>)). Figure 2 shows the results of the categorisation of the offered entities used with inf & III according to the above semantic subsets.

Looking at the semantic subset ‘religious’, one realises that ‘religious’ offered entities are used eleven times in TOI with the association inf & III while members of this semantic subset in connection with the association concerned cannot be found in DM or in BNC per. [16] The impression that the semantic subset ‘religious’ is of particular importance for Indian English – and consequently for the quantitative differences in comparison to the other varieties covered – is supported by the fact that, independent of the syntactic realisation, offer was not used at all with prayer or prayers in the DM or BNC per data investigated. Thus, it seems reasonable that the use of the semantic subset ‘religious’ with the given association is one phraseological factor in Indian English that has led to differences in the varieties under scrutiny.

Furthermore, the semantic subset ‘educational’ also provides some interesting insights. In relative frequencies, DM (50%) and TOI (33%) show offered entities of this semantic subset more frequently than BNC per (4%). As a result, it might be argued that the South Asian varieties are more likely to use offers in the context of instructions than British English.

In conclusion, then, the comparatively frequent use of inf & III in TOI can be explained by the range of semantic subsets speakers of Indian English have available to them. Only the data in TOI attested the usage of inf & III with all the semantic subsets established. Thus, the increased range of semantic subsets might result in an increased usage of the given association, which could serve as an explanation of the quantitative differences between the varieties concerned.

5. Discussion of the results of the data analyses

In the light of the above findings, it is valid to argue that Sri Lankan English displays certain verb-complementational features which distinguish it from other varieties of English and provide Sri Lankan English lexicogrammar with a distinct variety-specific character. The results of the present study also call attention to that the fact that the interface between lexis and grammar in general and verb complementation in particular may prove a valuable object of investigation for future studies on Sri Lankan English to come. Although a sociolinguistic perspective on the present-day usage of English in Sri Lanka in Schneider’s (2003, 2007) framework suggests that Sri Lankan English could potentially display what Schilk (2012: 172) labels “verb-complementational nativisation” (see section 1.2.), it is clearly beyond the scope of this paper to articulate claims of this nature due to its narrow focus.

Still, the present study hopes to have demonstrated that the investigation of verb complementation in connection with covariates influencing the selection of a particular syntactic pattern is a beneficial means to explore structural aspects in New Englishes. However, the focus of this study was limited in that it took only one covariate, namely the meaning of a ditransitive verb, into consideration. Thus, exploring the effects of other (or optimally all) covariates such as e.g. syntactic complexity, animacy or discourse accessibility (cf. Bresnan & Hay 2008) would allow researchers to approach verb-complementational profiles in varieties of English from a much broader perspective.

It is also important to stress that the complementary use of a set of balanced ICE [W130] corpora and a collection of relatively large newspaper corpora has proved to be useful and necessary. Although Greenbaum & Nelson (cf. 1996: 13) put forward that the ICE databases can be used to yield interesting lexical results, their limitations in size became obvious as no statistically viable results could be obtained from the ICE [W130] corpora, which, however, were only smaller (preliminary) text collections of already existing or soon-to-be-finished ICE corpora. However, using newspaper corpora as complementary text sources is certainly attractive as validating results in two relatively independent groups of corpora significantly adds to the reliability of the results and might even be an appropriate approach for future studies of New Englishes for which large and standard corpora are not available.

As has already been alluded to above, there is a need for further empirical studies on Sri Lankan English. A comprehensive description of Sri Lankan English would warrant analyses of many more aspects of the lexis-grammar interface (such as e.g. particle verbs or light-verb constructions) as well as of other structural levels of language in order to discover and accurately describe what is typically Sri Lankan in Sri Lankan English.

Notes

[1] I would like to express thanks to Joybrato Mukherjee, Marco Schilk and the reviewers of this paper, who provided me with helpful feedback and very useful suggestions. I am also grateful to Stefan Th. Gries for his invaluable support regarding the statistical tests and the visualisation of their results. For proofreading the original manuscript, I wish to thank Rosemary Bock. I alone remain responsible for the text.

[2] The abstract lemma of a verb will be given in capital letters. The word-forms of the lemma will be given in lower case and italics.

[3] The subjects in the patterns illustrated in (1), (3), (5) and (7) are given in round brackets in order to subsume imperative structures with implicit subjects under the respective patterns as well.

[4] Strictly speaking, the syntactic [S < O_d active] in example (4a) is which. Via anaphora resolution, it is possible to determine that, in this case, which refers to trays of food.

[5] In this example, offer is used in a zero participle clause. Only through the resolution of a zero object can the Bluetooth version be identified as the [S < O_d active] of example (6a).

[6] Passivisation of an intransitive construction is certainly counter-intuitive since the surface structure of the intransitive complementation pattern does not feature any objects that could function as a subject in passive voice. Still, it may be argued that any usage of a ditransitive verb cognitively evokes a ditransitive construction consisting of three argument roles, i.e. the agent, the recipient and the patient, independent of how many argument roles are made explicit (cf. Mukherjee & Hoffmann 2006: 151). For the type-IV pattern, this means that the objects of the ditransitive verb are cognitively implied, although they may not be syntactically realised, which, in turn, allows for passivisation of the type-IV pattern.

[7] The five corpus-based dictionaries used to establish the central meanings of offer were Collins COBUILD Advanced Learner’s English Dictionary (Sinclair 2006), MacMillan English Dictionary for Advanced Learners (Rundell 2002), Collins COBUILD English Language Dictionary (Sinclair 1987), Longman Dictionary of Contemporary English (Bullon 2007) and Collins Dictionary (Anderson 2007).

[8] ICE-SL is currently being compiled at Justus Liebig University Giessen, Germany, in collaboration with the University of Colombo, Sri Lanka.

[9] The web-derived newspaper corpora of Sri Lankan and Indian English were compiled at Justus Liebig University Giessen, Germany, in the context of the research project on “Verb-complementational profiles in South Asian Englishes” funded by the Deutsche Forschungsgemeinschaft (MU 1683/3-1).

[10] The preferred pattern of each meaning of offer in the respective corpora is highlighted in Table 4 and in Table 5.

[11] Despite the low absolute number of occurrences of offer in the ICE [W130] corpora, the data also yield interesting tendencies regarding the dative alternation of offer. While the double object construction (pattern I and IP) occurs more frequently than the prepositional dative construction (pattern II and IIP) in the British English (22 vs. 8) and Indian English data (14 vs. 3), the prepositional dative construction is more frequent than the double object construction in the Sri Lankan English data (14 vs. 11). Passivisation in this context occurs frequently with the double object construction in the British data while passivised double object constructions are virtually absent in the South Asian data. Passives are comparatively frequent with the prepositional dative in the Sri Lankan component of ICE [W130].

[12] As regards the dative alternation of offer in the newspaper data, the double object construction occurs more frequently than the prepositional dative construction in the British English data (60 vs. 42), but, in contrast to the ICE-based findings, the Indian newspaper data show a tendency similar to that of the Sri Lankan English data since the prepositional dative is more frequent than the double object construction in TOI (67 vs. 42) and in DM (70 vs. 52). The preference of the prepositional dative over the double object construction in Indian English may be related to the fact that Hindi, an indigenous language of India with official status, morphosyntactically marks the recipient in transfer events (cf. Haspelmath et al. 2005: 426ff.). In the British data, passivisation occurs more frequently with the double object construction as opposed to the prepositional dative (34 vs. 5) while the reverse is true for the Indian (13 vs. 10) and Sri Lankan newspaper texts (11 vs. 8). Thus, in particular with regard to the dative alternation in the Indian English data, differences between the ICE [W130] data and the newspaper data occur. These differences, however, may be attributed to the low overall frequency of offer in the ICE [W130] data and, consequently, the interpretation of the results will have to rely more strongly on the newspaper texts.

[13] The post-hoc test used squares the Pearson residuals of the dataset to contributions to chi-square and compares the values obtained to the corrected p-value for multiple post-hoc tests.

[14] Note that the association of the abstract meaning of offer and the type-III pattern in DM and TOI failed the post-hoc test.

[15] If a particular association is under discussion, it will hereafter be referred to in abbreviated form. The meaning of the association will be given in italics first and the syntactic pattern of the association follows so that the association of the type-III pattern with the meaning inf would read ‘the association inf & III’.

[16] In his analysis of Indian newspaper data, Schilk (cf. 2012: 133ff.) establishes that prayers, a member of the semantic subset ‘religious’, is an important collocate of offer in Indian English.

Sources

Department of Census and Statistics, Sri Lanka. 2001. Census of population and housing 2001: Number and percentage of population by district and ethnic group. http://www.statistics.gov.lk/PopHouSat/PDF/Population/p9p8%20Ethnicity.pdf

References

Algama, Dilini. 2008. “Who’s collecting ‘bits and bobs’: Meyler or Dylan?”. Sunday Times 23 Mar 2008. http://mirisgala.net/Dilini_Algama_review.html

Anderson, Sandra, ed. 2007. Collins Dictionary. Ninth Edition. Glasgow: Collins.

Bresnan, Joan & Jennifer Hay. 2008. “Gradient grammar: An effect of animacy on the syntax of give in New Zealand and American English”. Lingua 118(2): 245–259.

Bullon, Stephen, ed. 2007. Longman Dictionary of Contemporary English. Fourth Edition. Harlow: Pearson Longman.

Dharmadasa, K.N.O. 2007. “Sri Lanka”. Language and National Identity in Asia, ed. by Andrew Simpson, 116–138. New York: Oxford University Press.

Goonetilleke, D.C.R.A. 2003. “The interface of language, literature and politics in Sri Lanka: A paradigm for ex-colonies of Britain”. The Politics of English as a World Language: New Horizons in Postcolonial Cultural Studies, ed. by Christian Mair, 337–358. New York: Rodopi.

Goonetilleke, D.C.R.A. 2005. Sri Lankan English Literature and the Sri Lankan People 1917–2003. Colombo: Vijitha Yapa.

Greenbaum, Sidney & Gerald Nelson. 1996. “The International Corpus of English (ICE) project”. World Englishes 15(1): 3–15.

Gries, Stefan Th. 2009. Quantitative Corpus Linguistics with R: A Practical Introduction. New York: Routledge.

Haspelmath, Martin, Matthew S. Dryer, David Gill & Bernard Comrie, eds. 2005. The World Atlas of Language Structure. Oxford: Oxford University Press.

Hoffmann, Sebastian. 2007. “From webpage to mega-corpus: The CNN transcripts”. Corpus Linguistics and the Web, ed. by Marianne Hundt, Nadja Nesselhauf & Carolin Biewer, 69–85. Amsterdam: Rodopi.

Kachru, Braj B. 1983. The Indianization of English: The English Language in India. Delhi: Oxford University Press.

Kachru, Braj B. 1985. “Standards, codification and sociolinguistic realism: The English language in the outer circle”. English in the World: Teaching and Learning the Language and Literatures, ed. by Randolph Quirk & Henry G. Widdowson, 11–30. Cambridge: Cambridge University Press.

Künstler, Viktoria, Dushyanthi Mendis & Joybrato Mukherjee. 2009. “English in Sri Lanka: Language functions and speaker attitudes”. Anglistik – International Journal of English Studies 20(2): 57–74.

Leitner, Gerhard. 1992. “English as a pluricentric language”. Pluricentric Languages: Differing Norms in Different Nations, ed. by Michael G. Clyne, 179–237. Berlin: Mouton de Gruyter.

Meyler, Michael. 2007. A Dictionary of Sri Lankan English. Colombo: Mirisgala.

Mukherjee, Joybrato. 2005. English Ditransitive Verbs: Aspects of Theory, Description and a Usage-based Model. Amsterdam: Rodopi.

Mukherjee, Joybrato. 2008. “Sri Lankan English: Evolutionary status and epicentral influence from Indian English”. Anglistentag 2007 Münster: Proceedings, ed. by Klaus Stierstorfer, 359–368. Trier: WVT.

Mukherjee, Joybrato. 2009. Anglistische Korpuslinguistik: Eine Einführung. Berlin: Erich Schmidt.

Mukherjee, Joybrato & Stefan Th. Gries. 2009. “Collostructional nativisation in New Englishes: Verb-construction associations in the International Corpus of English”. English World-Wide 30(1): 27–51.

Mukherjee, Joybrato & Sebastian Hoffmann. 2006. “Describing verb-complementational profiles of New Englishes: A pilot study of Indian English”. English World-Wide 27(2): 147–173.

Mukherjee, Joybrato, Marco Schilk & Tobias Bernaisch. 2010. “Compiling the Sri Lankan component of ICE: Principles, problems, prospects”. ICAME Journal 34: 64–77.

Olavarría de Ersson, Eugenia & Philip Shaw. 2003. “Verb complementation patterns in Indian Standard English”. English World-Wide 24(2): 137–161.

Rundell, Michael, ed. 2002. Macmillan English Dictionary for Advanced Learners. Oxford: Macmillan.

Schilk, Marco. 2009. “Complementation of ditransitive verbs in South Asian Englishes: A multifactorial analysis”. Paper presented at the 30th ICAME Conference, Lancaster, May 2009.

Schilk, Marco. 2012. Structural Nativization in Indian English Lexicogrammar. Amsterdam & Philadelphia: John Benjamins.

Schneider, Edgar W. 2003. “The dynamics of New Englishes: From identity construction to dialect birth”. Language 79(2): 233–281.

Schneider, Edgar W. 2007. Postcolonial English: Varieties around the World. Cambridge: Cambridge University Press.

Shastri, S.V. 1988. “The Kolhapur Corpus of Indian English and work done on its basis so far”. ICAME Journal 12: 15–26.

Sinclair, John, ed. 1987. Collins COBUILD English Language Dictionary. London: Collins.

Sinclair, John, ed. 2006. Collins COBUILD Advanced Learner's English Dictionary. Fifth Edition. Glasgow: HarperCollins.