AN INVESTIGATION OF THE COMPLEXITY OF VERB PHRASES IN THE ICE OF FOUR ASIAN COUNTRIES

Schneider (2003, 2007), theorizing on the development of new Englishes, has proposed that transplanted Englishes in colonial societies go through several phases in their evolution. Consequently, each phase in their evolution impacts the new English, most especially in terms of linguistic structure. This paper builds on the hypothesis that placement in Schneider's dynamic model affects the linguistic development of new Englishes. More specifically, it tests the hypothesis on the complexity of verb phrases in Hong Kong, New Zealand, the Philippines, and Singapore Englishes with reference to their placement in the dynamic model. The study that was reported in this paper is a corpus-based one and the dataset was the national components of Hong Kong, New Zealand, the Philippines, and Singapore in the International Corpus of English. The focal structure in this corpus-based study is the central and most important constituent in the sentence – the verb phrase. Kortmann and Szmrescanyi (2009), and also Schneider (2003, 2007) and other studies that have invoked his developmental theory, have put forward that Englishes that have reached more advanced developmental stages have more complex linguistic structure. This hypothesis-testing should be able to see how these claims take place in the verb phrases of new Englishes.


Introduction
On a historical perspective, the global reach of languages is a phenomenon that has existed due to human's communicative needs.English is one notable language that has undergone unprecedented spread for economic, educational, and political reasons.Crystal (2004) describes the spread of English as a significant social phenomenon.English in paving way for humans to communicate with a universal language.In this regard, several linguistic scholars have become interested in plotting how such linguistic phenomenon, i.e. its spread to different speech communities, and how development of its varieties takes place.One of these is Schneider, (2003is Schneider, ( , 2007) ) who constructed the dynamic model which comprehensively describes how a language, i.e. post-colonial varieties of English (henceforth PCEs), evolves as a process and how certain linguistic features emerge.Influenced by other linguistic theories such as language contact, sociolinguistics, social identity and language evolution, the Dynamic Model explains the uniform process that new Englishes undergo.
The dynamic model outlines five major stages of the evolution of new Englishes, and takes a great deal of consideration on two agents of language spread-settlers and indigenous residents.It is also important to note that the placement of new Englishes in a particular phase of the model considers four underlying principles (i.e., sociopolitical background, identity constructions, sociolinguistic conditions and linguistic effects).The first stage is the foundation stage and it describes the arrival and initial widespread use of English in a previously non-English-speaking location.In this stage, the contact between the English settlers and the indigenous population is initiated.This situation is usually a result of immigration, invasion or occupation.Minor language changes such as pidginization occurs.Next, the exonormative stabilization explains the stable usage of English dictated predominantly by settler native speakers and based on the native variety they brought with them.Language contact between the settlers and the indigenous residents is limited, which results in the state of bilingualism among the latter.It is also worth noting that at this stage, stable and expanded pidgins develop and creolization occurs, and there is code-switching in both strands.The third stage, nativization, takes place when both settler and indigenous realities are reflected in the new hybrid co-existence.Additionally, there are changes on the language use of both strands which are evidenced through linguistic variations and innovations.In this stage, there is a move towards independence from the distant country of origin.The endonormative stabilization stage occurs when an indigenous linguistic norm takes root and is accepted by members of the community, which is marked by political independence.In linguistic terms, instead of English in X where X stands for the country, it is transformed to X English which says that English used in a speech community is a variety in its own right.The localized form of English is aimed at codification and is accepted for standard use.Differentiation, the last stage, marks the emergence of a new variety of language such that there is no question pertaining to its status as a variety.Alongside the community's identity is the birth of nation-internal, group-specific dialects.
Similar to other linguistic theories, the Dynamic Model is supported and/or negated not exclusively as a model in itself but in how Schneider evaluated the different post-colonial Englishes on their placement in the model.For example, in Asian Englishes, Bautista (2010) agrees with Schneider that Singapore English is more advanced than Philippine English by comparing use of subjunctives.Contentions are exemplified by Mukherjee (2007) as he argues that Indian English has moved into phase 4 and that it has steady equilibrium between progression and conservatism.Also, Borlongan (2011a) contends that the Philippines is in the endonormative stage through offering historical and attitudinal evidence.Meanwhile, in Africa, Hoffman (2010) found White Kenyan English to be in the nativization stage as supported by the speakers' production of acoustic vowels.Bekker (2009) claims that South African Black English is in phase 3 and that White South African English appears to be in phase 5.The Dynamic Model was also applied in several studies that analysed new Englishes such as Thusat et al. (2009) and Weston (2011) in assessing the local English of Malta and Gibraltar, respectively.
Collectively, what do these studies on new Englishes say about the Dynamic Model?Despite Schneider's admission that the model may have limitations that for instance concern the omission of other aspects of language evolution and that the model is not a reality in itself but just an abstraction from reality, evidently, it is widely discussed and may be widely accepted, and it appears that it is a powerful tool in analysing the advancement of new Englishes.Hence, the current study adopts this model as framework.

Linguistic Aspects of Nativization
An important discussion in the Dynamic Model is the structural and linguistic properties of PCEs and their emergence.Structural nativization, according to Schneider (2004), takes place when varieties of English select and adopt distinctive linguistic features or structural innovations of their own.The process is said to be concerned with 1) the features that may occur in PCEs amidst the extrinsic factors on language acquisition; 2) the similarities with the conceptualization of varieties; 3) language evolution principles; 4) linguistic process accountable for the individual structural properties in PCEs; and 5) the process of structural nativization itself; all of which are exclusively highlighted in a variety of studies in the area of World Englishes (e.g.Platt, Weber, & Ho, 1984;Kachru, 1986;Trudgill & Hanna, 2002).Schneider (2004) outlines the features of PCEs based on phonology, vocabulary, and grammar features of PCEs, which according to him are distinctive features difficult to detect since structural nativization is subtle and it affects only selected linguistic phenomena, thus, there is a need for a systematized method in observing the said features.
The current study delves into complex verb phrases in various registers of several PCEs vis-a-vis their placement in the dynamic model which calls for a brief discussion on grammar as an aspect of nativization.Quirk et al. (1985) note that linguistic nativization on the level of grammar may also refer to differences between varieties of English in the use of forms and structure shared by the PCEs aside from merely new forms and structures.Therefore, the evolution of a language variety, shown through the identity construction that results from the interaction between settlers (STL) and indigenous (IDG) strands, coincides with the language's syntactical changes and evolution.
Pattern distribution among language varieties is found in the majority of nonstandard dialects in different speech communities.Schneider (2004) points out those PCEs possess distinctive rules and patterns making their variety distinct from other varieties.Additionally, he noted that through simplification, sketching a number of grammatical phenomena of several PCEs may be possible.Additionally, linguistic variability in this sense is strongly related to sociological aspects which stress the features of non-standard and standard forms of language.In this regard, it can be deduced through empirical studies (e.g.Kortmann & Szmrecsanyi, 2004;Platt, Weber, & Ho,1984) that language varieties are distinct mostly on the surface level that structural properties remain the same due to association with other varieties.
Looking into the grammar as an aspect of nativization also takes into consideration the manner how a variety originates and how it is disseminated.Schneider (2004) describes the process as a product of language contact.When a language which has a rigid structural mechanism in a speech community is used in another speech environment, linguistic features are adopted by those who take part in the discourse.
Grammar also interacts with the innovation of new lexis given that the former is more stable.In the process of structural nativization, however, grammar may demonstrate flexibility since forms change as new word forms emerge.Lexico-grammatical innovations may be established gradually up to the point that new and old forms are interchangeably used, which results in the variability of forms.Overall, Schneider (2004) stresses that PCEs are described by meaningful new forms of a speech community which he says cannot be accounted for since the innovations could be more related to issues on language variation and language change and not on PCEs.

Complex Verb Phrases
It is said that the verb phrase and the noun phrase are the core elements in a clause.In particular, the verb phrase, according to Quirk et al. (1985), is "the most important phrasal category for functional reasons"(p.61)and that it is "the most 'central' and indispensable part of the clause" (p.61).Additionally, Biber, Johansson, Leech, Conrad, and Finegan, (1999) emphasize on and describe the role of the verb phrase in clauses as essential.This could be strengthened by its ubiquity in various corpora such as the Longman Spoken and Written English (LSWE) and International Corpus of English (ICE).Undeniably, perhaps due to the seemingly significant role that the verb phrase plays in the English linguistic structure, it is not surprising that a number of scholars have examined in detail this feature.
In the World Englishes domain, scholars such as Collins (2009) and Schneider (2011) have investigated various areas of the verb phrase across several English varieties.The same is true for Philippine English domain.For example, Bautista (2000aBautista ( , 2000b)), Borlongan (2011bBorlongan ( , 2013) ) and Friginal (2011) have emerged to conduct remarkable studies in exploring various areas of research on the verb phrase.Tay (1991) acknowledges this development and states that the Philippine English may be considered to have the most comprehensive scholarly outputs on indigenous English varieties.Clearly conducting such studies have been useful in the Philippine English and in general, the World Englishes domains as a variety of features can describe the grammar of advancement of emerging and existing language varieties.However, as Borlongan said, there is still so much to know about the verb phrase.Particularly, he suggested that there may be a need to look into and compare verb phrases in other language varieties through corpus-based studies.The current study responds to Borlongan's proposal in order to maximize the various corpora available for conducting such studies and ultimately, to further explore aspects of verb phrases that need comprehensive analysis.Quirk et al. (1985) posit that complex verb phrases are formed by several words through the following types of construction: 1. Type A: Modal-a modal auxiliary verb + the base of a lexical verb as in will go 2. Type B: Perfective-the perfective auxiliary HAVE + the -ed participle of a lexical verb, as in has gone 3. Type C: Progressive-the progressive auxiliary BE+ the -ing participle of a lexical verb, as in is going 4. Type D: Passive-the passive auxiliary BE/GET + ed participle of a lexical verb, as in is studied Hence, the following combinations could be formed resulting from the said constructions types:

The English varieties
Admittedly, Singapore English is already in phase 4 of the cycle.Several reasons can be used to evidence this particular claim.For instance, Singaporeans show communion in their local variety, which for them is an expression of the pride concerning their nation.Singaporeans appear to make it certain that their English remains intelligible in international contexts and somehow maintain their ideological distance from Western culture.It can also be traced that since then, scholars like Pakir (1991) found it distinct that Singapore English gives a sense of identity to the users as belonging to the Singaporean speech community.
The Philippine English, with influence from American English, is currently at the third stage of the Dynamic Model.Schneider believes that the Philippines has been stuck in the said stage for decades since the Marcos administration, which prompted the deteriorating use of English among Filipinos.However, studies like that of Borlongan (2011a), as earlier mentioned, contend that the Philippine English already possesses characteristics of New Englishes in the fourth stage.This argument is supported by several historical incidents such as the studies of Borlongan (2011) and Schneider (2007), which signal a definitive separation between the settlers (STL) and the indigenous (IDG) groups.These studies contributed to the sense of separation from the United States.
Meanwhile, Hong Kong is one of the last outposts of British colonialism, having received independence only in 1997.Schneider asserts that this recent contact with colonialist British English does not seem to be advantageous for Hong Kong English in its advancement in the Dynamic Model cycle.Hence, it is still in phase 3.Evans (2009) confirms through historical evidence that the said variety possesses characteristics of varieties belonging to stage 3. Setter et al. (2010), however, remain positive in that Hong Kong English has potential in moving to phase 4.
Among the Englishes discussed in the current study, New Zealand is the variety that has completed the cycle albeit at different historical periods and under divergent political and linguistic circumstances.Schneider claims that New Zealand English has diversified under the influence of immigration which brings other language backgrounds into the country.Its foundation stage was said to have begun in the mid-19 th century under the colonization of the British.By 1985, the New Zealand English dialect has achieved its differentiation stage after a period of consolidation and homogenization.

Methodology
The International Corpus of English (ICE), a database of written and spoken English texts from 24 different countries, played a key role in the data gathering of the current study.Specifically, the ICE-PHI (Philippine English), ICE-SIN (Singapore English), ICE-HK (Hong Kong English) and ICE-NZ (New Zealand English) had to be obtained to successfully carry out the research objectives.Each corpus is composed of about one million words distributed almost evenly across 500 texts with specified categories; therefore, there are approximately 2000 words per text with some being composite to reach the 2000-word minimum.Also, the texts were sampled from the English spoken or written by adults aged 18 and above and who received formal education through the medium of English up to the post-secondary level.The texts of the corpora are divided into spoken and written texts, the major text categories.Overall, the texts include private and public dialogues, unscripted and scripted monologues, and nonprinted and printed written materials.Table 1 shows the details of the text categories including the approximate number of words.
All data have been utilized in the researcher's previous studies which provided ease of access to the corpora.Corpus analysis was carried out using the WordSmith Concord Tool, with the nine modal verbs (i.e.should, would, could, will, shall, must, can, may, might) specified as context words for the first six combinations, i.e.AB, ABC, ABD, ABCD, AC, ACD, AD.This was done based on the discussion of Quirk et al. on the combinations that one of the nine modal verbs is expected to be present in a complex verb phrase.For the last four combinations i.e.BC, BCD, BD, and CD, the has/had/have and is/are being were used as content word for the concordance.Texts analysed were Face-to-Face Conversations (C) (file names S1A-001 -S1A-090), academic texts (AT) specifically in the field of humanities, social sciences, natural sciences and technology (file names W2A-001 to W2A-040), press news reports (PA) (file names W2C-001 to W2C-020), and novels and stories (F) (file names W2F-001 to W2F-020).It has to be justified that the said texts were chosen for analysis among others because they are considered to be the major text types necessary for studies of this nature (Biber et al., 1999).
After each corpus was searched for the retrieval of lines that contain the modals, analysis had to be conducted if the phrase indeed is a verb in function and if it falls in any of the combinations.If so, a phrase was then considered a valid token; hence, frequency counting took place.The retrieved verb phrases were then tabulated and categorized.The data counting took place for more than a week at a university in Manila.Re-count was conducted by a colleague two days after the frequency counting.Lastly, the frequency count of the tabulated verb phrases were computed for quantitative comparison (i.e.number of tokens divided by the total number of words found in the text type multiplied by one million).Understandably, this has to be performed to normalize the data since the texts analysed were not equal in number. .

Results
A particular text type calls for a specific register; hence, a considerable difference in the forms of verb phrases could be found in certain text types.Conclusively, in fulfilling the aims of the current study, it is seen relevant to look into the complexity of verb phrases vis-à-vis registers of the New Englishes.As seen in Table 4, similar to the ICE-SIN and the ICE-PHI, the AD combination in ICE-HK is commonly used in academic texts and newspapers.Hong Kong English speakers appear to use the BD combination the most, especially in academic texts and press articles.Additionally, the AB verb form is also commonly used in fictions and press articles.In the same way, the AC combination is usually employed in conversations and press articles.What is evidently noticeable is the nonexistence of the ABC combination, which was present in Singapore and Philippine Englishes.The ICE-NZ Corpus has an almost identical number of frequencies of verb phrases with the other corpora of the current study.One remarkable finding is the highest numbers of AB combinations in the written texts, in contrast to the other language varieties.Also, the AD combination is also dominantly present mainly in academic texts and press articles.In the same way, a fairly frequently used combination is the AB verb form, which is commonly used in the four registers investigated.Similar to the ICE-SIN, the ACD combination was also found in this corpus, this time, in the spoken discourse.
Among the post-colonial Englishes investigated, it may be safe to say that the Singapore English and New Zealand corpus appear to have salient findings which could provide evidence to the sophistication of the language variety.Firstly, the New Zealand corpus has the highest number of complex verb phrases across all text types (i.e.14271 complex verb phrases).Secondly, the ICE-SIN and the ICE-NZ corpora contain the most varied complex verb phrases among other English varieties in the current study.To illustrate particularly the second finding, nine verb combinations can be found in the Singapore English and New Zealand English, particularly in academic texts.Specifically, Singapore English users frequently utilize the AD combination in academic texts and in press articles with a relatively large margin over the other varieties.In addition, users of Singapore English make use of the ABC and ACD combinations in academic texts, which is not typical in the other text types of the other corpora, except for conversations of New Zealand English.This could manifest the awareness of academic scholars and even press writers on the complexity of verbs needed to be employed in such texts.
Further, New Zealand English, being identical to Singapore English with nine combinations found in their respective corpus, could be a clear indication of the indistinguishable awareness of both groups of users concerning the complexity of verb phrases.The similarity, however, is not absolute since there were nine verb phrase combinations found in conversations in the ICE-NZ corpus.This may denote that users of New Zealand English tend to be more sophisticated in informal conversations compared to Singlish users.Likewise, AB combination appears to be relatively common in the New Zealand corpus across all text types.It may be said then that the similarity and differences of the two Englishes may reflect the results of the differing sociolinguistic conditions that the speakers of both speech communities are surrounded with.
Complexity of verbs is also observed in the Philippine English data.Identical to the other varieties in the current study, eight combinations are found in the Philippine English corpus, though not too salient.It has to be noted, however, that conversations and fictions are the discourse genres which have eight combinations.What can be considered distinct about it is that a number of speakers of this language variety appear to be have some level of sophistication in their conversations.A minor finding that may be mentioned is that some users of Philippine English employ ABD combinations in conversations.Though, this may still be inconclusive since only a few tokens of the ABD combinations were retrieved from the data.
The HK English appears to have the least complicated structure among the New Englishes investigated in the present study.This is said due to the seven variations of complex verb phrases which were found in the corpus, evidently lesser and the least compared to the other language varieties in the present study.Only six combinations were found in conversations and five combinations were retrieved in fictional texts in the said variety.Fictions were probably considered by users of the other Englishes as texts that need complexity in constructions.Interestingly, perhaps due to the poetic license employed by literary writers in Hong Kong, only the AB, AC AD, BC and BD are used in Hong Kong English fictions.Consequently, the ABC and the CD verb combinations, which were found in the rest of the corpora, were not retrieved in the Hong Kong English corpus.It can also be noted that only a few AB combinations were found in the field of academic writing in Hong Kong English.
Overall, similarities between and among the New Englishes featured in the study were also observed.
For instance, none of the Englishes use the ABCD and the BCD combinations.Additionally, academic texts, seconded by press articles, are the text types which appear to need sophistication in construction as seen in the complexities of verbs employed in the text type among all Englishes.More specific findings include the similarity of the Philippine English and New Zealand English in terms of the ABD combinations found in all discourse types.In addition, the use of AC combination also stood out to be common in press articles of the Philippine English and New Zealand English.Convincingly, these identical characteristics could signify possible similarities in the linguistic structure of both Englishes vis-à-vis their placement in the Dynamic Model.

Discussion
Does placement in Schneider's dynamic model really affect the structural development of new Englishes?The current study attempted to answer this question, with a hypothesis-testing exercise of some sort.The basis has been the analysis of the complexity of verb phrases found in the corpora of four postcolonial Englishes, namely Hong Kong English, New Zealand English, Philippine English, and Singapore English.Corpus findings reveal the New Zealand English and Singapore English to be having the widest repertoire of complex verb phrases.It also demonstrates internal stylistic heterogeneity, being able to have specialized uses of certain combinations in certain text types.This finding may imply greatly on the placement of two Englishes in the dynamic model as they approach the differentiation stage.The Philippine English trails behind these two Englishes, being less developed than them.Furthermore, it was earlier mentioned that Philippine English is moving on from Phase 3 to Phase 4 (Borlongan, 2011) and Philippine English trailing behind these two Englishes should be expected.Schneider mentioned that the Philippines has been stuck in the nativization stage for decades and this occurrence may reflect the less complex linguistic structure of the Philippine English.Lastly, Hong Kong English is the least complex and least stylistically heterogeneous and that is also because of Hong Kong being the least developed among the four Englishes.Clearly, as Schneider pointed out, there is an existing variety such as Hong Kong English.
However, what Setter et al. (2010) appear to contest is that the variety is in the endonormative stage based on several parameters.The present study suggests otherwise; that the linguistic effects of Hong Kong English are not as advanced, at least in terms of verb structures, compared to known varieties in the fourth stage of the dynamic model.
However, it seems unexpected that Singapore English and New Zealand English advance identically in terms of frequencies and complexity of verb phrases.
"One form of variability that is definitely to be expected concerns its chronology, the existence of unequal duration times and overlapping characteristics of individual phases.In every developmental process the boundaries and succession of stages may be realized fuzzily.There are both dynamic periods when certain phenomena change very rapidly and periods of inertia when things stay rather stable for a long while.Similarly, not all characteristics of a given phase occur exactly simultaneously: it is normal for individual characteristics of a certain stage to appear earlier than others, or to be delayed, as well as for elements of subsequent but distinct stages to overlap and co-occur in time.To some extent the linearity of the model is also an abstraction from what in reality is a multidimensional interplay of dynamic processes."[emphasis added] (Schneider, 2007, p. 57) So as answer to the question posed earlier whether placement in Schneider's dynamic model really affects the structural development of new Englishes, it generally has an effect.Placement in the model will generally be able to predict the structural development of a certain English; however, there can be a few exceptions.In the case of verb phrase complexity, some aspects of Singapore English are more complex than the purportedly more developmentally-advanced New Zealand English and vice-versa.This is also evident in the use of some outstanding frequency of combinations of less advanced Philippine English and Hong Kong English.

Conclusion
The current study attempted to investigate the complexity of verb phrases in four Englishes across four text types.It hypothesized that placement in Schneider's dynamic model affects the linguistic development of new Englishes.Using the ICE for each type of English, the study quantitatively demonstrated how the verb phrases in each language variety differ.This corpus-based exercise confirms how empirical data (i.e.corpora) can enrich theory and in this case, Schneider's dynamic model.Indeed, the Englishes go through stages of development, and Schneider rightly so identifies these stages.Conclusively, this paper has demonstrated how his theory is realized structurally in complex verb phrases.
Additional things can be said about the study.It was noticeable that there is scarcity or even absence of some combinations (i.e.ABCD and BCD combinations) across all language varieties.What can be said about this phenomenon?The new Englishes in the study, all Asia-Pacific varieties, or at least some or most of their users, may still be alienated with the perfectives which are new developments in English grammar.This results in the preference, or perhaps implicit knowledge in using the alternative variant, which is the simple past.As for the salient findings, indeed, tracing the advancement of an English variety through analysis of verb phrases is plausible.It was seen that the new Englishes in the advanced stage of the Dynamic Model have more complex verb structures.
Overall, it can be said that the current study yielded results on the verb phrase and its relationship with the Englishes that make use of it and its placement in the Dynamic model.However, a plethora of areas can still be investigated in the World Englishes domain and linguistically speaking, the verb phrase itself.Other emerging Englishes can be used for analysis.This procedure may be conducted as part of Schneider's transitional attraction and include the Englishes placed in the expanding circle.In order to achieve this goal, corpora building and/or expansion may be more than necessary with respect to other Englishes.
of the model affects the linguistic development of new Englishes.Specifically, it investigates the complexity of verb phrases found in conversations, academic texts, press articles and fictions among Hong Kong, New Zealand, Philippine, and Singapore Englishes with reference to their placement in the dynamic model.
Due to the importance of looking into complex verb phrases as linguistic basis of an English variety's placement in the Schneider Dynamic Model, the current study tests the hypothesis that placement in a certain stage

Table 1 .
Text categories of the ICE for the current study

Table 2 .
Frequency of Complex Verb Phrases of ICE-PHI PHI as seen in Table2indicates high frequencies in the AD combination of complex verb phrases.It is notable that it is in Academic texts and press related texts where most of the AD combination is used.Another dominant finding is that the BD combination is common in informal combinations.In addition, the AB combination is somewhat frequently used by speakers of the Philippine English.It is in fictional and press articles where the combination is used most frequently.Worth mentioning is the relative frequency of the AC combination in conversations and press articles.Table3shows the prevalent use of complex verb phrases in Singapore English, most particularly the AD combination.A huge portion of the frequencies can be found in academic texts and press articles.Speakers of Singapore English also commonly employ the BD combination in academic texts.Meanwhile, fictions have the most number of AB combinations.Moreover, AC combinations are frequently used in conversations and press articles.Interestingly, academic texts in the ICE-SIN contain the ABC combination in academic texts, which is absent in other varieties of the current study.

Table 5 .
Frequency of Complex Verb Phrases of ICE-NZ