Syntactic variation in Hong Kong English

1. Introduction

Kirkpatrick and Lixun (2021) maintain that two significant morphosyntactic processes have been at play in early Englishes. These are simplification and regularization. Simplification refers to the relatively simplified inflectional morphology in English today. Kirkpatrick and Lixun (2021) provide an example for the word stan (i.e., stone in Old English) that showed great differences in the singular and plural form in nominative, accusative, genitive and dative case in Old English. Another process is regularization, through which some of the strong verb forms for past tense in English have changed to take the weak or the regular form. To illustrate, the past tense of work was wrought but over time, it has changed to worked.

In her study on the syntactic variation characterizing English as a Lingua Franca (ELF) in informal conversations, Meierkord (2004) states that new English varieties are also usually characterized by levelling and simplification. While levelling involves ‘reduction or attrition of marked variants’ (Trudgill, 1986: 98), simplification refers to an increase in regularity, particularly pertaining to morphology. These concepts of simplification and levelling (or regularization) are significant as they may also be at play in Hong Kong English today.

Platt, Weber and Ho (1984) outline a number of features found across different English varieties. Among these non-standard features within the noun phrase (NP) are changing the word order of an NP, lack of plural marking, and lack of a distinction between the third person pronouns he and she. In terms of the verb phrase (VP), speakers have an inclination to extend the progressive aspect to stative verbs (e.g., *I am loving), lack of verbal marking for third person singular in the present tense or for past tense, and lack of a distinction between tense and aspect systems. Finally, at the clause level, we see omission of subject and object pronouns, copying of pronouns, uninverted syntactic units in interrogative sentences, and the use of an invariant question tag form are.

Despite a few monographs (see among others Bolton, 2002; Setter, Wong & Chan, 2010; Cummings & Wolf, 2011; Evans, 2016; Wong, 2017) in the last two decades on the general properties of Hong Kong English (HKE), the broad range of syntactic variation observed in HKE has rarely been fully explored in a single empirical study. Bolton's (2002) first edited monograph on HKE explored issues ranging from HKE in context, to the various forms in HKE as well as dwelling on the creativity of Hong Kong speakers particularly in writing. The only syntactic property addressed was Gisborne's (2002) chapter on relative clauses in English, which provides a comprehensive range of syntactic features Hong Kong speakers exhibit in their use of relative clauses. These range from zero-subject relatives, in which an obligatory subject relative pronoun is omitted as in This is the student did it (p. 144), to the omission of prepositions, the use of resumptive pronouns, and the absence of the restrictive/non-restrictive relative clause contrast.

Setter et al.'s (2010) monograph on HKE surveyed more syntactic properties such as what the authors refer to as random morphological markings, which they categorize into three: (a) the use of a singular count noun in bare form, (b) the omission of the plural suffix in plural nouns, and finally (c) the use of the plural suffix to mark singular count nouns. Setter et al. (2010) also discuss double morphological markings including forms such as more better. Another property the authors discuss is what they call tense switching, which demonstrates Hong Kong speakers’ tendency to use the present tense for past and future and using a lexical indicator of time reference (e.g., yesterday, tomorrow) rather than signaling tense through morphological markings. Among other phenomena scrutinized in Setter et al. (2010) are issues relating to subject-verb agreement, particularly the omission of the -es suffix for third person singular subjects used with present tense, zero subjects as discussed in Bolton (2002), as well as double subjects indicated by the italic noun phrases in (1).

1. 1) Passengers who take the ferry service from Ma Liu Shui, they can enjoy a free ride from Tap Mun to Wong Shek (p. 57).

Setter et al. (2010) also highlight Hong Kong speakers’ not being able to distinguish count nouns from mass nouns. That includes omitting the indefinite article from singular count nouns (e.g., there will be giraffe), and pluralizing canonically mass nouns (e.g., equipments) (p. 60). The authors also bring attention to an interesting phenomenon whereby Hong Kong speakers use a non-standard word order in which modifiers immediately precede heads. An example Setter et al. (2010) provide is He very like dancing, a standard version of which would be He likes dancing very much. The authors also underline the use of redundant prepositions (e.g., you have to face to the south), as well as conversion of grammatical categories (e.g., it's less physically demand) (pp. 64–65). A major limitation of Setter et al. (2010) was the small number of participants. Setter et al. (2010) provided data from only five participants studying at the University of Reading or the University of Oxford. The researchers collected data through two speaking tasks. Subjects were asked to recount a happy event from their childhood, and also to perform a map task that required the participants to guide the interviewer from a starting point to the meeting point on a map.

Sung (2015) discusses HKE from linguistic and sociolinguistic perspectives by highlighting the variation in noun phrases, verb phrases, and clause and sentence structures. Regarding noun phrases, two of the properties Sung (2015) underlines are the non-distinction between count and mass nouns in HKE (Kirkpatrick, 2007; Setter et al., 2010), and zero morphological markings (Setter et al., 2010). With respect to verb phrases, a feature Sung (2015) emphasizes is that HKE speakers sometimes do not make a distinction between active and passive verbs, employ double morphological markings (Setter et al., 2010), or display non-tense marking (Li, 2000). In terms of clause and sentence structures, Sung (2015) illustrates the use of periphrastic topic construction in HKE as in In the above examples, it shows that learners (Li, 2000), as well as the non-use of the ‘Adj for NP to V’ structure (Li, 2000). Finally, Sung (2015) discusses unique features of HKE within relative clauses. One such feature is zero subject relatives (Newbrook, 1988; Gisborne, 2000). One must exercise caution, however, in labelling those as features of HKE as such variation may also stem from the L2 acquisition process (Hung, 2012). Moreover, as Sung (2015) reiterates, the features he outlines are not unique to HKE only but some of them are also found in Philippine English, Indian English and Singapore English. Kortmann (2010) refers to such features that can be found in multiple English varieties as a morphosyntactic angloversal. According to Sung (2015), HKE exhibits some phonological, grammatical and lexical features that make it a distinct variety. It should be noted, however, that Sung (2015) was only a survey and lacked empirical data, which still remains a research gap in investigations regarding the linguistic features of HKE.

Empirical investigations on HKE has been limited. The current research aims to fill this gap by scrutinizing the syntactic variation in HKE observed in college student essays through a careful linguistic analysis between the participants’ first language, Cantonese, and English.

2. The current study

The current study was motivated by the research question: what is the syntactic variation in HKE observed in college student essays and what kind of syntactic variation can be categorized as a distinct feature of HKE?

2.1 The participants

The participants were 79 first-year undergraduate students studying mainly engineering and business administration at a major research university in Hong Kong. All of them were native speakers of Cantonese, a language with very few inflectional and derivational markings (Matthews, & Yip 2013), and they started learning English as a second language (L2) in both formal and informal settings as early as 4.3 years of age. All the participants signed informed consents for their data to be used for research. Table 1 illustrates the descriptive statistics.

Table 1. Descriptive statistics from the participants

2.2 Data collection instruments

The data came from two 300-word graded reflection essays that the students were asked to write before they finished a competency-based upper-intermediate level English language course in the Spring semester in 2022. The first essay, called the writing task, asked students to write about how multimodality was used to enhance meaning in the students’ multimodal group project. The second essay, called the looking forward task, asked students whether they learnt something that they did not expect to learn, what they did exceptionally well in the course, what competency was the most challenging in the course, etc.

2.3 Data analysis

Out of 158 300-word reflection essays, 50 essays were randomly selected and analyzed. Any non-standard form that appeared at least once in at least 20% of the essays (i.e., ten essays) was selected for inclusion in the study. The rationale behind this was to make sure to exclude instances of non-standard forms that might have occurred purely due to occasional performance errors (Van Rooy, 2013). The analysis unearthed two general patterns. These were omission, and using certain grammatical and lexical forms with meanings not found in standard English. For the omission cases, I determined the grammatical construction, the number of obligatory contexts for that grammatical construction in all the 50 essays, the number of tokens in obligatory contexts, the percentage of tokens in obligatory contexts, the number of omission in obligatory contexts, and the percentage of omission in obligatory contexts. For cases in which the students used certain grammatical and lexical forms with meanings not found in standard English, I decided on the grammatical construction, the total number of tokens, the number of standard usage, the percentage of standard usage, the number of non-standard usage, and the percentage of non-standard usage.

3. Results

The analysis revealed two general patterns: (1) the omission of various grammatical forms in obligatory contexts: (2) using certain grammatical and lexical forms with intended meanings not found in standard English.

A note of caution is needed here. Trying to identify the source of non-standard English forms in a new variety of English, one may easily fall into the pitfall of attributing all the non-standard syntactic or morphosyntactic variation to the substrate or the first language (L1) of the speakers using the variety. Regarding this danger, Kirkpatrick and Lixun (2021) write ‘ . . . explaining the presence of non-standard morphosyntactic forms solely or mainly by reference to the substrate seems ill-advised, as so many of the same non-standard morphosyntactic forms occur in so many different varieties of vernacular English’ (p. 139). Meierkord (2004), for example, found that in interactions across international Englishes, there was an ‘overwhelming correspondence to the rules of L1 Englishes, transfer phenomena, developmental patterns and nativised forms¹, simplification, regularisation and levelling processes’ (p. 128). Therefore, in the following sections, I am going to try to pinpoint the locus of the non-standard syntactic variation by appealing to L1 transfer (or the use of a nativized form), as well as referring to concepts of simplification and levelling (or regularization) discussed in detail in Trudgill (1986), Meierkord (2004), and Kirkpatrick and Lixun (2021).

3.1 Omission of various grammatical forms in obligatory contexts

3.1.1 Omission of the indefinite article

The first omission category was the omission of the indefinite article a/an. This has been documented in previous research as zero morphological marking (Setter et al., 2010). It should be noted that Cantonese lacks articles altogether. A bare singular count noun in the subject (2) or the object position (3) can refer to singular or plural entities (Cheng & Sybesma, 1999).

Out of 313 contexts that required the use an indefinite article, the participants supplied the indefinite article in 232 contexts. Example (4) demonstrates such omission, where the blank (here and henceforth) indicates the omitted grammatical form.

1. 4) We also used images and stock footage to let the audience have ___ better idea of what we are talking about.

3.1.2 Omission of the definite article

The participants also omitted the definite article the, also referred to as zero morphological marking in prior literature (Setter et al., 2010). In Cantonese, there is no definite article. Definiteness can be conveyed through context and discourse via bare singulars, shown in (5).

Lack of a definite article in Cantonese may have led to negative L1 transfer. Out of 364 obligatory contexts that necessitated the use of a definite article, the participants provided them only in 258 contexts. Example (6) indicates this.

1. 6) ___ Final solution is to use the white noise of the sea waves to create the atmosphere on the beach together with the softened, relaxing tone in the voiceover.

3.1.3 Omission of the subject relative pronoun

Participants also omitted the subject relative pronouns in a relative clause. This omission case is referred to as zero subject relatives in previous research (Newbrook, 1988; Gisborne, 2000). Note that English (7) and Cantonese (8) employ different syntactic means for relative clauses.

The relative clause in the English example in (7) is constructed by a head noun, people, that is followed by a relative pronoun, who, which modifies the head noun. Such a construction can be reduced by deleting the relative pronoun, who, and at the same time using a gerund form of the verb. The relative clause construction in Cantonese in (8) is radically different. If the relativizer ge is considered to have the same function as who in English, the relative clause construction in Cantonese seems to be almost the reverse of such a construction in English. I believe that the reason why Cantonese students omit the subject relative pronoun in a relative clause in English is because in their L1, relative clauses are not introduced by a head noun followed by a relativizer. In English, a relative clause follows a head noun. However, in Cantonese, the head noun follows the relative clause. Thus, Cantonese students may not realize that relative clauses always need to be introduced by a head noun followed by a relativizer unless the relative clause is reduced. Out of 93 contexts that required the use of subject relative pronouns, the participants provided them in 66 contexts. Example (9) demonstrates this phenomenon.

1. 9) However, sometimes our group members ___ worked on video editing couldn't reply in time or miss the messages sent by others.

3.1.4 Omission of the to be verb in a relative clause construction

The participants also omitted the to be verb in a relative clause construction particularly in cases where the subject of the relative clause was in passive voice.² This is reported in the literature as non-distinction between active/passive verbs (Li, 2000). Out of 44 obligatory contexts that required a to be verb, the participants supplied it 34 times. Example (10) shows this omission.

1. 10) The second video was better than the first one which ___ recorded 2 weeks ago.

3.1.5 Omission of the past tense marker

Another omission pattern was the omission of the past tense marker and using lexical past instead, which has also been observed in Li (2000) as non-tense marking. It should be noted that in Cantonese, there is no suffix that could be equated to the English past tense morpheme -ed. Lexical past is quite frequent in Cantonese, illustrated in (11).

The lexical past kàhmyaht in (11) is sufficient for the utterance to have a past reference. The perfective (PFV) aspectual suffix -jó, which indicates the completion of an action, is optional in the sentence. Such morphosyntactic differences between Cantonese and English in the realization of the past tense reference may have led the Cantonese speakers to omit the past tense marker in English. Out of 677 obligatory contexts that required the use of a past tense morpheme, the Cantonese speakers used the past tense only in 487 of them. (12) indicates an example.

1. 12) During our meetings, my group mates have a great output of ideas but our discussions go all over the place.

3.1.6 Omission of the preposition to

Cantonese has two elements that can be equated to the English preposition to. The first one is the verb bei, which literally means give. In addition to its verbal meaning, bei may sometimes be translated as to or for. This is illustrated in examples (13) and (14).

The second element that can be translated as to is heui, which primarily means go, shown in (15).

However, when there is a dative argument indicating a location, heui takes on a new prepositional function meaning to, illustrated in (16).

As the examples (13) to (16) indicate, although Cantonese speakers have at their disposal multiple morphemes to communicate the meaning of the preposition to in English, the syntactic and semantic properties of bei and heui are more complicated than the English to. Therefore, most likely due to interference from their L1, the Cantonese speakers omitted the preposition to in 25 contexts out of 113 obligatory contexts. Example (17) indicates such omission.

1. 17) When I was 14, I applied ___ a piano competition.

3.1.7 Omission of the preposition for

As discussed previously, the verbal and prepositional element bei in Cantonese is sometimes translated as the benefactive for. This is illustrated in (18).

However, such usage of bei is only limited to cases with a benefactive meaning. In cases where one refers to duration, bei is not licensed. This is indicated in (19).

The inclusion of bei in (19) renders the sentence ungrammatical. Due to such properties of Cantonese, participants may have omitted for. Out of 155 obligatory cases, the participants provided for only in 131 contexts. Example (20) illustrates an omission case.

1. 20) One of the group member was always late ___ 20 minutes.

Similar omission of prepositions discussed in this and the previous subsection has been documented in Gisborne (2000), although those are omission cases within relative clauses.

3.1.8 Omission of pleonastic subjects

The English syntax requires that every sentence have a subject, whether explicit or implicit. To satisfy this syntactic requirement, sentences may appear with a pleonastic subject it, which is also referred to as an expletive or a dummy pronoun. Example (21) illustrates this point.

1. 21) It is nice to meet you.

The pronoun it in (21) has no semantic meaning. It is there purely for syntactic reasons. The lack of a pleonastic subject in Cantonese may have led the participants to omit it, and alter the syntactic structure of sentences that would otherwise require a pleonastic subject. Out of 11 obligatory contexts, the participants provided pleonastic subjects in only 3 instances. The italicized part in example (22) illustrates this.

1. 22) In our group meetings, I am difficult to concentrate.

A grammatical version of (22) would be (23).

1. 23) In our group meetings, it is difficult for me to concentrate.

This omission case has been documented in previous literature as non-use of the ‘Adj for NP to V’ structure in Li (2000). Table 2 illustrates all the eight omission cases.

Table 2. L1 Cantonese L2 English speakers’ omission patterns in writing in L2 English

Figure 1 demonstrates the number of obligatory contexts, the number of tokens, and the number of omission in those contexts.

Figure 1. L1 Cantonese L2 English speakers’ omission patterns in writing in L2 English

In all of the omission cases, a crosslinguistic comparison of the participants’ L1 Cantonese and English indicates that the non-standard English forms could be due to the L1 or they could stem from a simplification process attested in the literature (Trudgill, 1986; Meierkord, 2004).

A significant question that arises is whether to accept all those eight omission patterns as features of HKE. To do so, for each case discussed above, we need to determine if that is an occasional non-standard feature, or whether the omission can be classified as a regular feature of the variety under investigation. According to Kirpatrick and Lixun (2021), when the standard usage of a feature exceeds that of non-standard usage, such non-standard usage cannot be considered as a distinct feature of the variety under investigation. Scrutinizing the eight cases, we see that the learners were more likely to use a standard form compared to a non-standard one in all of the cases except for the last one, pleonastic subjects. Since the non-standard usage in the domain of pleonastic subjects far exceeds the number of standard usage, we can maintain that this could be a distinctive feature of HKE.

3.2 Using certain grammatical and lexical forms with intended meanings not found in English

The participants also used grammatical and lexical forms with intended meanings not found in standard English.

3.2.1 Present will

Out of 187 uses, the participants used will correctly with a future reference in 112 contexts. However, in 31 contexts, they used will with a present tense interpretation. The italicized will in example (24) indicates this usage.

1. 24) At the end of the LANG1003 course, all students in a class will be broken down into groups of 4 to 5 and each group will have to present upon a multimodal project by making an 8-minute-long video.

3.2.2 Present habitual will

Out of 187 uses, the participants used will correctly with a future reference in 112 contexts. In 18 cases, they used will with a present tense habitual interpretation, indicated in (25).

1. 25) I have a pretty terrible habit in which I will compare the quality of my work to others.

3.2.3 Past will

The participants used will with a past tense interpretation in 20 out of 187 contexts. Example (26) shows such a usage.

1. 26) All of us paid huge effort into the project, and we do want our outcome to be the best. That is why some of us will stand very strong in our opinions.

3.2.4 Past habitual will

The Cantonese speakers used will with a past tense habitual interpretation in 6 out of 187 contexts, illustrated in (27).

1. 27) The success of my group's presentation is attributed to my groupmates' mutual effort and understanding. At the beginning of the project, we will meet 1 hour every Tuesday.

The four cases of will discussed above can be an example of what Setter et al. (2010) refer to as tense switching, although Setter et al. (2010) mainly discuss the switching between present and past tense forms. However, given the use of will with an intended habitual meaning in 3.2.2. and 3.2.4., we can argue that the speakers may also be mixing the tense and aspect systems as discussed in Meierkord (2004).

3.2.5 Pluralizing mass nouns (i.e. equipments, staffs, stuffs)

Unlike in English, mass-count distinction is not fully grammaticalized in Cantonese. It may have been due to the mismatch between English and Cantonese that led the Cantonese speakers to consistently pluralize mass nouns equipment, staff, and stuff. This has been referred to as non-distinction between count and mass nouns in prior literature (Kirkpatrick, 2007; Setter et al., 2010). The Cantonese speakers may have only considered whether a noun is atomic or non-atomic (see Choi et al., 2018). Nouns like equipment, staff, and stuff are atomic and they denote bounded entities (unlike a mass noun like water), but they are syntactically mass, and they cannot be pluralized in English. I believe that the participants were influenced by the atomicity of those nouns, thereby conceptualizing them as count nouns which can be pluralized. The participants correctly used those three nouns in 13 contexts, but pluralized them in 10 contexts. Examples (28) and (29) indicate this phenomenon.

1. 28) For the video recording, we didn't use many equipments.

2. 29) We had to visit the library and staffs were very helpful.

Mass and count nouns have been examined in previous literature. To illustrate, in their study on Outer Circle, Expanding Circle and lingua franca Englishes, Hall, Schmidtke and Vickers (2013) investigated 19 mass nouns such as stuff in the VOICE corpus and 25 mass nouns such as equipment in various websites. Although Hall et al. (2013) found considerable tokens of mass nouns used countably, such usage was extremely infrequent compared to the usage of mass nouns with mass syntax. To illustrate, Hall et al. (2013) identified 152 tokens of the word stuff, of which only one was used with count syntax, while 151 was used in line with standard English norms. That is, the countable use of the word stuff only accounted for 0.7 per cent. Another significant finding of Hall et al. (2013) is that overall, the authors found little evidence indicating a role for substrate (L1) influence.

In the current study, a reanalysis of all the mass nouns in the 50 papers indicated that there was a total of 27 different mass nouns used with a total token number of 217. The only mass nouns pluralized were equipments, staffs, and stuffs. That is, out of 217 tokens, mass nouns were pluralized in a non-standard manner ten times, which amounts to a non-standard usage of about 4.6 per cent of the time. Even though pluralization of mass nouns is higher compared to previous research (note that such countable usage of mass nouns is 2.9 percent in Hall et al. 2013 and 2.5 percent in Mollin 2007), it still shows that the L2 learners were much more likely to use a mass noun in a canonical mass context.

3.2.6 Conditionals without a conditional marker³

The participants treated even as a conditional marker equating it with even if / even though. They correctly used even if / even though 19 times. However, they omitted if in 9 contexts. Example (30) shows such usage.

1. 30) Sometimes, even ___ I don't agree with my group mates, I don't say anything.

3.2.7 Counterfactual conditionals instead of factual conditionals

The participants also used counterfactual conditionals when they needed to use a factual conditional. The reason for this may have been due to their confusion of tense and aspect as reported in Meierkord (2004). While tense refers to the time when an event or an action takes place or a state holds, aspect is about the internal constituency of events, indicating whether an event is completed, ongoing, repeated. The Cantonese speakers tended to use a past tense form of the verb, and creating a counterfactual conditional in cases where they would need a factual conditional. Out of 17 contexts, they only used counterfactual conditionals correctly in two contexts. Example (31) shows a non-standard usage of counterfactual conditionals.

1. 31) If we found out it is impossible to talk about all aspects of the issue (which is our case), we need to address the audience first by saying we will focus on certain part of it.

In English, the use of such a conditional would be interpreted as referring to a hypothetical situation in the present. However, what the Cantonese speaker means is an actual possibility in the present or the future rather than a hypothetical situation.

3.2.8 Conjunction doubling in subordinating conjunctions with although

Another non-standard English construction that the Cantonese participants used was conjunction doubling in subordinating conjunctions with although. This is referred to as double morphological marking by Setter et al. (2010). Note that such structures including although and but are perfectly grammatical in Cantonese, indicated in (32).

Due to negative L1 transfer, Cantonese speakers may have made use of conjunction doubling. Out of 45 contexts in which they used an although, they went on to use but in 12 of them. Example (33) indicates such usage.

1. 33) Although our video is neither fancy nor full of effects, but I believe we present our ideas quite clearly by the use of multimodality.

Table 3 indicates the various non-standard grammatical constructions the participants used.

Table 3. L1 Cantonese L2 English speakers’ non-standard grammatical forms in writing in L2 English

Figure 2 shows the total number of tokens, the number of standard and non-standard usage.

Figure 2. L1 Cantonese L2 English speakers’ non-standard grammatical forms in writing in L2 English

In all of the non-standard uses of the syntactic forms in this section, a comparison of the participants’ L1 Cantonese and English shows that the non-standard English forms could stem from the participants’ L1 or they could be due to a simplification process (Trudgill, 1986; Meierkord, 2004). However, participants’ pluralization of mass nouns (i.e. equipments, staffs, stuffs) is an exception since it cannot be due to the L1 neither could it be a simplification process.

Similar to the analysis in the omission cases, we need to establish which cases are a good candidate to be categorized as a distinct feature of HKE rather than an occasional property. Using the same logic utilized before and following Kirpatrick and Lixun (2021), we can maintain that those cases where the usage of the non-standard forms far exceeds the number of standard forms are potential candidates that can characterize the distinct English variety under investigation. Thus, the only such candidate in this section is counterfactual conditionals instead of factual conditionals, which is why that construction is the only one that could be conceptualized as a systematic feature of Hong Kong English.

4. Discussion and conclusion

This research investigated the nature of syntactic variation in HKE observed in college student essays. The results indicated that Hong Kong speakers are quite creative in using non-standard syntactic and lexical forms in HKE. The two broad patterns the participants exhibited in HKE, namely the omission of various grammatical forms in obligatory contexts, and using certain grammatical and lexical forms with intended meanings not found in standard English, demonstrate an effect of participants’ first language, Cantonese, on HKE. A comparative analysis between Cantonese and English for most of the properties observed provided evidence that the participants relied on using the structures of their first language in HKE. However, it has been shown that most of the non-standards forms could also be due to a simplification process.

Another goal of the research was to determine whether the syntactic variation discussed was simply an occasional presence of a non-standard feature, or whether such forms could be established as a regular feature of HKE. An analysis revealed that in the domain of pleonastic subjects and counterfactual conditionals instead of factual conditionals, the non-standard usage far exceeded the number of standard usage, which is why we can maintain that these two features could be distinctive features of HKE rather than indicating an occasional presence. Future studies should look into more contexts where English is used including oral discourses to pinpoint what kind of syntactic variation can indeed be categorized as a distinct feature of HKE.