Use and acceptability of particle verbs with unique meanings in Sri Lankan English

This article presents the findings regarding the use and acceptability of a selection of particle verbs that have unique meanings in Sri Lankan English (SLE). The article is based on an earlier study conducted by me that aimed at exploring the extent to which certain grammatical and lexico-grammatical features of SLE occur in semi-formal writing, and the level of acceptability associated with them among first language English speakers in Sri Lanka by means of an acceptability study (AJT) based on corpus data. The original study included an analysis of 37 lexico-grammatical features of SLE, but the present article focuses on only seven particle verbs that have interesting localised meanings in Sri Lanka. The corpus data for the study were drawn from the Sri Lankan component of the corpus of South Asian Varieties of English (SAVE) and the Sri Lankan English Newspaper Corpus – 2015. The acceptability study is based on a sample of 80 first language speakers of English in Sri Lanka who responded to an AJT containing sample sentences drawn from the two corpora which included the target particle verbs. The findings revealed that a significant number of these particle verbs with unique meanings do appear in naturally occurring SLE writing, but there is a general lack of a strong consensus regarding the acceptability of a majority of them. On the other hand, the particle verb come down with its unique meaning has a comparatively higher level of acceptance although it appears minimally in the corpora.


Introduction
Particle verbs (pvs) have been identified as a challenging aspect of English grammar that has appeared in early listings of characteristic features of World Englishes (Schneider, 2004).Particle verbs is a collective term for phrasal verbs, prepositional verbs and phrasal prepositional verbs.These verbs operate in the interface between lexis and grammar, an area that has been established as leading to variation in usage (Mukherjee, 2008, Schneider, 2007).Pvs in varieties of English have received significant levels of scholarly attention with valuable contributions, among others, from Schneider (2004), Mendis (2010), Gilquin (2009) and Zip & Bernaische (2012).There have been several studies that have specifically focused on phrasal verbs in SLE including Bernaische (2015), Fernando, D (2007) and Kumara and Mendis (2010).However, there are no published studies that have focused both on the use and acceptability of the pvs such as pass out, come down, bring down, go behind/run behind, taken up with, go as, and make out that have unique meanings in SLE.These verb particle combinations that have acquired unique meanings in the context of SLE present an interesting topic for studying.The following table presents information regarding the pvs focused on in this study as presented by Meyler (2007) and Kumara and Mendis (2010).No equivalent pv in BE The present article attempts to fill the gap in the available literature regarding the use and acceptability of these pvs.It explores and discusses the extent to which these verb particle combinations with unique meanings are used in naturally occurring Sri Lankan news writing and the level of acceptability associated with them among the speech community.The specific research questions that this article attempts to answer are: i.To what extent do the SLE pvs with unique meanings in focus appear in authentic Sri Lankan writing?
ii.To what degree are these pvs considered acceptable by proficient users of English in Sri Lanka?Schneider (2004) defines pvs as "idiomatic combinations of a verb and one or more particles which jointly behave as a single lexicogrammatical unit" and states that these verbs possess distinctive grammatical and semantic properties giving them a unique character (p.230).The particle in these verbs could either be a preposition or a spatial adverb and typically there is a single verb that is synonymous with any pv.Schneider (2004) further explains that phrasal verbs, prepositional verbs and phrasal prepositional verbs all fall under the category of particle verbs.Phrasal prepositional verbs are formed by three elements; a verb with a particle and a preposition and thus can be easily distinguished from the others.The difference between phrasal verbs and prepositional verbs is more nuanced.Phrasal verbs such as call off, hang up, pass out are combinations of verbs and particles and are generally highly idiomatic.In prepositional verbs such as look at and remind of, the original meaning of the preposition is more strongly preserved so that the meaning is semantically transparent.Phrasal verbs can be transitive (e.g.bring up, try out) or intransitive (e.g.pass out, hang up) while prepositional verbs can occur with one noun phrase (e.g.They looked at the picture) or with a direct object followed by a prepositional object/prepositional complement (They invested money in the project).It could be difficult to distinguish between phrasal verbs and prepositional verbs when the verb is followed by a noun phrase.For example, the following two sentences look superficially structurally similar even though bring up is a phrasal verb whereas cope with is a prepositional verb: I will bring up this topic again at the next meeting.

Properties of pvs
Many animal species are struggling to cope with rapid climate change.
The difference between the phrasal verb and the prepositional verb here is that with the phrasal verb, it is possible to move the noun phrase that occurs after the verb in between the verb and the particle as in I will bring this topic up again at the next meaning.It is not possible to do the same with the prepositional verb; *Many animal species are struggling to cope rapid climate change with is grammatically wrong.Bearing in mind the distinction between phrasal verbs and prepositional verbs, the present article focuses on the phrasal verbs pass out, come down, bring down, make out, go as, taken up with and the prepositional verb go behind, all discussed under the cover term particle verbs.

Previous studies on pvs in SLE
Several studies on the relative frequency of pvs in varieties of English have shown that South Asian Englishes including SLE feature comparatively lower numbers of pvs (Schneider, 2004;Zipp & Bernaisch, 2012;Bernaische, 2015).Bernaische (2015) further showed that SLE exhibits even lesser number of pvs than Indian English and pvs are mostly found in informal genres in SLE.However, it should also be noted that pvs are generally considered informal and they are more frequently found in informal genres than in formal genres in many other varieties of English including Indian English, and British English (Bernaische, 2015).Therefore, associating pvs with informality is not a characteristically Sri Lankan phenomenon.For example, in his study of pvs ending with up, out and off in Sri Lankan, Indian and British Englishes using the relevant subcorpora of the International Corpus of English (ICE) and newspaper corpora from these countries, Bernaische (2015) shows that pvs occur less frequently in academic English in all three varieties of English compared to other genres such as creative writing.Pvs are mostly found in creative writing and instructional writing in SLE and are comparatively less frequent in academic and news writing.In addition to these observations on the frequency of use, he also presents some innovative ways in which pvs are used in SLE.He cites several instances where a particle is redundantly added to a verb, forming a new pv that is similar in meaning to the original single verb such as the use of accelerate up, lurch up, waive off, lease out and detail out where the original meaning of accelerate, lurch, waive, lease, and detail are retained.Although these new coinages have not replaced the one-word alternatives, they are used alongside the single word alternatives (Bernaische, 2015).
Shantha Kumara and Mendis (2010) studied pvs in SLE using the Sri Lankan sub-corpus of the South Asian Varieties of English (SAVE-SL) including data from two Sri Lankan newspapers published between 2001 and 2007.Through the analysis, they have identified 18 distinctively Sri Lankan pvs that appear in SAVE-SL with varying frequencies.These pvs were found to differ from their counterparts in Standard English usage in relation to the form or the meaning.For instance, in certain pv such as usher in, the particle in was dropped and usher was used to convey the same meaning.In certain other instances, another particle was added to an existing pv or one word verb as in the cases of cope up with and settle down.Further, existing pvs were used to convey a different meaning.For instance, they identified that taken up was used to mean 'impressed by' and run behind/go behind was used to mean 'to attempt a relationship or gain an advantage'.Fernando (2007) examined the notions of acceptability associated with a selection of pvs found in SLE among a sample of school English teachers.The study revealed that even though a majority of the participants accepted the use of pvs such bear up, come down and pass out, the teachers' awareness of the distinction between SLE pvs and those of Standard Englishes is limited.However, their responses have shown a shared sense of what is acceptable and not acceptable and she introduces a cline of acceptability in relation to these pvs.The findings revealed that certain grammatical aspects of SLE are gaining acceptance as the local norm even though the acceptance could be based on a lack of awareness of the fact that these aspects are deviations from the standard norm.
However, there are no previous studies that have focused on both the prevalence of the pvs pass out, bring down, come down, go as, go behind, and make out, and the notions of acceptability associated with them.The present article combines a corpus analysis and an acceptability judgment task to study how strongly these pvs have established themselves as features of SLE.

The Corpus Study
The Sri Lankan sub-component of the SAVE corpus (SAVE-SL) containing published data from 2001 to 2007 was used as a reference corpus in this study.SAVE-SL contains texts extracted from two English newspapers; Daily News (texts published between 2001 and 2005) and Daily Mirror (texts published between 2002 and 2007).1.5 million words from each newspaper are included in this SAVE-SL totalling 3 million words.All genres of writing in the online version of the newspapers have been included except advertisements and the "Opinion" column of the Daily Mirror.Further, foreign news reports such as those of the Associated Press and Reuters have been excluded as far as possible as these are not considered representative of SLE (Bernaisch, Koch, Mukherjee, & Schilk, 2011).
To enable comparisons with SAVE data, the Sri Lankan English Newspaper Corpus-2015 (SLENC-2015) was used as it includes the same genres of the online versions of the Daily News and Daily Mirror published in 2015. 1 million words from each newspaper were included in this corpus totalling 2 million words (Hediwaththege, 2021).
Both corpora were searched for the occurrence of the target pvs using Antconc 3.2.4w.Wild card searches were conducted for all conjugations of the pvs.
Concordance lines were analysed to verify the contextual meaning of the feature and for semantic clearance.The frequency of occurrence of the target pvs was studied.

Acceptability Judgment Task (AJT) as a research tool
Grammaticality/Acceptability Judgment Tasks involve exposing the respondents to sentences/strings of words and requesting them to judge/rate the well-formedness of the sentence/string of words (Schütze & Sprouse, 2013).There has been a considerable amount of discussion over naming this test instrument (Schütze & Sprouse, 2013).For instance, Chomsky (1965) has argued that the idea of "grammaticality" should not be confused with "acceptability" as grammar is a mental construct that is not accessible to conscious awareness.He argues that grammaticality is related to the "competence" of a speaker whereas acceptability is related to "performance".Thus, respondents may not be able to comment on the grammaticality of a sentence whereas they could make comments on its acceptability, and grammaticalness is just one factor that determines the acceptability of a feature (Chomsky, 1965, as cited in Schütze & Sprouse, 2013 ;Tremblay, 2005).Schütze and Sprouse (2013) identify AJT as playing a "crucial role" in the field of linguistic research as it enables the collection of otherwise unavailable data such as data related to the acceptability status of rare linguistic features (p.131).They mention that the non-occurrence of a rare feature, even in a very large corpus, would not provide negative evidence for grammaticality, whereas an AJT could enable making comments on the status of that feature.Another advantage of AJTs compared to corpus studies or spontaneous usage data is that naturally occurring data may include unintentional mistakes which the producers themselves may not accept as correct.The occurrence of such phenomena will not say anything about acceptability.AJTs on the other hand will bring these features to the conscious attention of the respondents and obtain their comments.

The AJT
The AJT was conducted through an online worksheet in which respondents had to rate the grammatical acceptability of given sentences using a grading scale.Sentences containing the target pvs with unique SLE meaning which had occurred in both corpora were included in the worksheet.Therefore, only sentences containing pass out, come down, bring down, and go behind were included in the worksheet while the pvs make out, go as and taken up with which occurred in only one corpus were not included in it.An equal number of sentences with the same pvs that carried the BE meanings were also included in the worksheet as distractors except in the case of the pv "come down" which does not have a BE counterpart.BE counterparts were used as a reference point as British English is the input variety.Sentences including the target pvs with both the unique SLE meaning and the BE meaning were included in the worksheet in order to minimize the potential extragrammatical influence of having only potentially grammatically doubtful sentences in the context.This also minimizes scale bias, a condition that occurs when respondents have to use only one end of a rating scale continuously or when they limit their responses to a few options on a rating scale.
The sentences that were extracted from the SLE corpora were edited by the researcher in order to reduce parsing difficulty which is an extragrammatical factor that could influence the judgment of a respondent.Moreover, grammatical errors in all sentences were corrected in order to make sure that a sentence is not judged based on a grammar feature which is not a focus of interest in this research.When choosing sentences, those that sound awkward or incomplete in meaning when taken out of the larger context were not included in the worksheet in order minimize the effects of the extragrammatical factors and pragmatic considerations.
The importance of providing the respondents with clear and detailed instructions as to what they are supposed to do and the basis on which they have to judge a sentence has been emphasized by Schütze (1996).The respondents of the present study were specifically instructed to rate the sentences only in relation to grammar, and not to consider the content, meaning or the political correctness of the sentences.The respondents were requested to write down any errors or awkward expressions that they see in a sentence in order to make sure that a sentence is not judged based on a feature other than a grammar feature in focus.
The rating scale of the AJT was designed following Schütze's recommendation (1996) that grammaticality be treated as a gradient concept.Therefore, the respondents were requested to rate the sentence on a Likert Scale of six points as follows: • "Completely unacceptable" • "Sounds awkward" • "Slightly unnatural but okay" • "Sounds okay" • "Completely acceptable" • "Not sure" Even though Likert Scales usually use numerical values, the above labels were used to define each point on the scale.This was done in order to make it easier for the respondents to rate the sentence as they may not be familiar with the concept of grammaticality being a gradient concept and self-explanatory labels may be therefore more useful than impersonal numbers.The respondents were informed that they have the option of choosing the "not sure" option if they felt uncertain about the grammatical un/acceptability of a sentence.
The respondents were given up to 2 ½ minutes to evaluate each sentence.

The Selection of the Sample, Sample Size and Distribution of the worksheet
The respondents were a sample of 80 Sri Lankans who used English as their first or co-first language.A significant number of the respondents were teachers of English in tertiary institutions or leading international schools in Colombo.The worksheet was sent to 57 respondents initially.The sampling method was similar to the snowball sampling method.The first 57 respondents to whom the worksheet was sent are proficient speakers of English known to the researcher.They participated in the study with informed consent and shared the worksheet with other proficient speakers of English that they knew.Altogether 156 responses were received, of which only 111 were complete.In addition, 31 of the respondents had mentioned either Sinhala or Tamil as their first language/mother tongue.These respondents were not included in the sample as there was no way of knowing their proficiency in English even though these respondents could possibly be proficient speakers of SLE.The remaining 80 responses were used in the analysis.

Social Variables in the Sample
The respondents were from the age range 20-50 years except one respondent who was over 50 years old.53 respondents were English teachers in tertiary educational institutes or leading international schools in Colombo and 61 possessed a degree or diploma related to English or were reading for a degree.It was attempted to include equal numbers of males and females in the sample.However, the sample was predominantly female as there were 64 females and only 16 males.Even though this could be seen as a limitation of the sample, this gender bias could also be a reflection of the male and female composition in occupations related to language, communication and teaching in Sri Lanka.

Ethical considerations, informed consent and confidentiality
The study received approval from the Ethics Review Committee for Social Sciences and Humanities (ERCSSH) of the University of Colombo.A consent form was used to obtain the informed consent of the respondents to participate in the study.The consent form contained basic information about the study and the fact that participation in the study is entirely voluntary and optional.The respondents were assured of the confidentiality of the data they provide and that their anonymity would be strictly maintained.

Limitations of AJT as a Research Tool and Steps Taken to Minimize their Effects
A limitation of AJTs is that acceptability can only be measured indirectly through reported perception.It requires the respondent to objectify grammatical knowledge and make a metalinguistic judgment (Tremblay, 2005;Schutze & Sprouse, 2013).Some scholars have questioned the validity of relying on an assumed relationship between objectification skills and actual grammatical knowledge (Bresnan, 2007, as cited in Schütze & Sprouse, 2013).However, Schütze and Sprouse (2013) state that reported perceptions have been considered and widely utilized as a reliable tool not only in linguistic research but also in other cognitive research and that they "tend to be systematic in ways that can lead to the construction of falsifiable theories" (p.3).

Grammaticality as a gradient concept
There seems to be considerable debate over whether grammaticality is a dichotomous or gradient concept as there is no clear criteria for determining the nature of grammaticality (Tremblay, 2005).Most of the formal models assume grammaticality to be dichotomous and expect respondents to make binary judgements of sentences as grammatical or ungrammatical.However, Tremblay (2005) shows that Chomsky had recognized the gradient nature of grammaticality as early as in 1975 because he had used multiple annotations to denote varying degrees of ungrammaticality showing that one sentence could be more grammatical or ungrammatical than another.
The gradient nature of grammaticality is created by both intraspeaker and interspeaker variation.Intraspeaker variation refers to the fact that the same person's notions of grammaticality are not static; i.e., they may judge the same sentence in two different ways in two different datasets or at two different times (Tremblay, 2005).Thus their perceptions of grammaticality could be in a process of evolution which may make making absolute judgments difficult.Interspeaker variation refers to the fact that people speaking different versions of the same language may differ in their judgments of acceptability (Tremblay, 2005).These people may recognize that a feature is different from their version of the language but may feel reluctant to judge it as completely unacceptable.Thus asking respondents to make binary judgments based on their perceptions of grammaticality could lead to misleading data.
Further, a person's internal grammar consists of a core and a periphery which enable the recognition of intermediate levels of grammaticalness or grammatical indeterminacy (Tremblay, 2005).Core grammar consists of regular or unmarked features whereas periphery grammar includes marked features which are irregular and less rule-governed.Therefore, there is a possibility that grammaticality judgments related to periphery grammar may not always be as firm as those related to the core and may probably be rated with intermediate levels of acceptability.The worksheet used in the present article did not expect the respondents to make dichotomous decisions regarding the acceptability of the sentences because of the above arguments.

Findings of the corpus analysis
The following table presents the number of occurrences of each pv with unique SLE meaning and their equivalent Standard English meaning as a proportion of the total number of occurrences of the pv in each corpus.These numbers are then presented as a percentage.The percentages are rounded to the first decimal number.
3 There is no equivalent BE phrasal verb 4 There is no equivalent BE phrasal verb The use of the pv pass out with the localized meanings 'to graduate' / 'to complete a course' in the corpora is significant as it has been used with the unique meaning in almost 75% of the occurrences making it the rule rather than the exception.For instance, out of the 37 occurrences of the different conjugations of the pv 'pass out' in the SAVE-SL corpus, 29 occurrences carry the SLE meaning 'graduate' or 'complete a course' while only 08 occurrences carry the BE meanings 'finish a military course/ become unconscious'.Therefore, 29 occurrences out of the 37 hits carry a localized meaning which comprises 78% of the total occurrences.In SLENC-2015, 29 out of the total 39 occurrences carry a variant localized meaning which comprises 74% of the total occurrences.
Using the unique local pv bring down to mean 'to import', come down to mean 'come from abroad' and go behind in place of 'go after', are also observable in both corpora, but to a comparatively less significant extent.For example, occurrences of bring down with the unique SLE meaning make less than 30% of the total occurrences.Also, the occurrence of some of these features are limited to the same newspaper in both corpora.For instance, go behind has occurred only in The Daily News in both corpora, and come down with the localized meaning has occurred only in The Daily Mirror in both corpora.
Similarly, go as, and make out with the localised meanings have occurred only in the SAVE-SL corpus.Taken up with the SLE meaning has not occurred in SAVE-SL but has appeared in SLENC-2015.

Findings of the AJT
Considering the gradient nature of grammatical acceptability, the respondents were instructed to rate the sentences as 'Completely unacceptable', 'Sounds awkward', 'Slightly unnatural but okay', 'Sounds okay' or 'Completely acceptable'.For purposes of analysis, these options are assigned the following numerical values in this data analysis;

Rating
Completely unacceptable

Sounds awkward
Slightly unnatural but okay

Sounds okay
Completely acceptable The respondents could choose the "not sure" option if they felt uncertain about the grammatical un/acceptability of a sentence and therefore did not want to rate it.The following table presents the data collected through the acceptability judgment task.The respondents were also requested to write down any errors/ awkward expressions that they see in each sentence.The responses received are incorporated into the discussion of the findings of the acceptability study.The use of come down with the unique localized meaning 'to come from abroad' has been rated as 'sounds okay' or 'completely acceptable' by 55 respondents out of 80 respondents (68.75%).Another 07 (8.75%) have rated it as 'sounds unnatural but okay'.In addition, only 28 respondents (35%) respondents had identified come down as an error or an awkward expression.This finding suggests that a significant majority of the respondents (65%) considered that come down is not a language error or an awkward expression produced due to lack of language proficiency.It is also interesting to note that the number of hits of come down in the corpora are relatively low, even though it has received a significant level of acceptance among the respondents.This may suggest that come down is a pv familiar to the respondents through recurrent exposure, possibly in a different genre.
Another pv that seems to have a comparatively higher level of acceptance among the respondents in the sample is pass out with the meaning 'graduate' as used in the following sentence: He passed out from the University of Bolton with a first class degree in civil engineering.44 respondents (55%) have rated this as either "sounds okay" or "completely acceptable" while 12 more respondents (15%) have rated it as "sounds unnatural but okay".Only 24 respondents (30%) have rated it as 'completely unacceptable' or 'sounds awkward'.However, 35 respondents (43.75%) respondents had identified pass out as an error or an awkward expression in the sentence even if some of these respondents had rated the sentence as 'sounds okay' or 'sounds unnatural but okay'.However, it is important to note that 45 respondents (56.25%) considered that the use of pass out in this sentence is not a language error or an awkward expression produced due to lack of language proficiency.
The following are 04 comments given in response to the statement; "Passed out?But, I wouldn't say it's awkward".(ID 50) "Passed out is a purely Sri Lankan expression to mean 'graduate'.(ID 53) "Passed out is widely used to refer to graduation, but I'm not too convinced, it sounds a bit awkward".(ID 29) "Passed out usually indicates someone falling or going into deep sleep.A more correct word choice would have been graduated".(ID 110) The responses given to bring down also shows lack of consensus regarding its acceptability or non-acceptability.This pv has been rated as either 'sounds okay' or 'completely acceptable' by 36 respondents (45%) while it has been rated it as "sounds unnatural but okay" by 15 more respondents (18.75%).However, 25 respondents (31.25%) have rated it as 'completely unacceptable' or 'sounds awkward'.43 respondents (53.75%) had identified bring down as an error or an awkward expression in the sentence even if some of them rated the sentence as 'sounds okay' or 'sounds unnatural but okay'.Therefore, it seems to be difficult to come into a conclusion regarding the acceptability of this pv.
The reception of go behind is interesting.Even though a majority of the respondents have rated it as 'completely unacceptable' or 'sounds awkward', many respondents have identified it as a non-standard usage, or a colloquial expression rather than a language error.For instance, 45 respondents (56.25%) have rated the sentence as 'completely unacceptable' or 'sounds awkward'.Only 19 respondents (23.75%) have rated this as either 'sounds okay' or 'completely acceptable' and 14 more respondents (17.5%) have rated it as 'sounds unnatural but okay'.In addition to this, only 31 respondents (39%) respondents had identified go behind as an error or an awkward expression in the sentence.Thus, a majority of the respondents seem to consider it a non-standard usage rather than an error and only a minority considered the use of go behind in this sentence as a language error or an awkward expression produced due to lack of language proficiency.The comments given by the respondents further show that go behind is viewed more as a colloquial expression rather than an error that should be avoided in all contexts; "I am aware of this expression being used, so I can't put it down as an error".(ID 12) "The expression 'going behind' is more colloquial than formal".(ID 28) "Going behind' is a direct translation from Sinhala to English".(ID 49) "Going behind someone is colloquial, but I wouldn't say it's an error".(ID 50)

Conclusions and avenues for further research
The findings of the corpus analysis and the acceptability study show that the pvs with unique meanings focused on in this study continue to occur in Sri Lankan English to varying degrees.All 07 pvs that were analysed in the corpus analysis appeared at least in one corpus and 04 of them occurred in both corpora.Thus, it could be observed that a significant number of these localised pvs have continued to occur in authentic SLE writing over a period of 15 years as the reference corpus has been compiled from 2001 -2007 and the monitor corpus was compiled in 2015.Especially, the use of pass out for the two localised meanings, 'graduate/complete a course of study', occur as the rule rather than the exception in the corpora indicating the localised meanings of pass out has a very strong and long-standing presence in written SLE.
As regards acceptance, it was observed that there is no strong consensus regarding the acceptability of these pvs except come down with its unique SLE meaning.Although the localised meanings of pass out have a strong presence in the corpora, a strong consensus cannot be observed regarding its acceptability among the respondents.This contradiction may indicate an unwillingness among the speech community to overtly acknowledge the legitimacy of a fairly widespread localized feature which could be a sign of linguistic schizophrenia in the Sri Lankan context in relation to this feature.
Further, the mismatch between the positive ratings and the responses given to the statement "Please write down errors / awkward expressions that you see in the sentence, if any", as in the cases of pass out and bring down, exemplify the ambivalence existing in the SLE speech community regarding the legitimacy of using SLE for writing purposes.This is further elaborated in comments such as "Passed out is widely used to refer to graduation, but I'm not too convinced, it sounds a bit awkward" (ID 29), and "The expression 'going behind' is more colloquial than formal".(ID 28).These comments show that SLE pvs are not completely unaccepted, but they are considered more suitable in less formal domains.This may possibly indicate that English in Sri Lanka seems to be becoming diglossic in the same way as the main two local languages of Sri Lanka, Sinhala and Tamil.
It is also interesting to note that none of these pvs have been rated as unacceptable or awkward by all respondents which shows that all these features have gained some level of acceptance among proficient users of English in Sri Lanka which may possibly grow further in the future.These findings of the corpus analysis and the acceptability study affirm Gunasekera's (2005) statement that SLE writing seems to be in a process of "slowly shaking itself free" and being used in writing more frequently even though currently there does not seem to be much consensus among the sample regarding the acceptability of a majority of the variant features.(pp. 130-132).
As regards avenues for further research, the present study revealed that the localised meanings of pass out occur in Sri Lankan writing as the rule rather than the exception even though acceptance lags behind.It would be interesting to further investigate the concerns of the Sri Lankan speakers of English when they choose not to accept it as appropriate to be used in writing which may reveal the linguistic power relations that come into play when making a conscious decision regarding the acceptability of a new variety of English.
In future research in this area of study, it would also be interesting to see whether the pv 'come down' continues to occur in different genres of Sri Lankan writing and whether it continues to maintain the acceptance of the Sri Lankan speakers of English as a feature of the local standard appropriate to be used in writing.

Table 2 : Findings of the corpus analysis
1 DN refers to the newspaper Daily News 2 DM refers to the newspaper Daily Mirror