(Un)democratic Epicene Pronouns in Asian Englishes: A Register Approach

Abstract

The search for gender equality in language use is one of the most frequently cited cases of linguistic democratization (e.g., Farrelly & Seoane 2012:394). At the grammatical level, this process implies, for example, that pronouns such as generic he used with epicene antecedents are being replaced by singular they or by combined he or she, at least in inner-circle varieties of English. However, outer-circle varieties remain underexplored in this regard. For this reason, this paper analyzes three Asian English varieties, namely Hong Kong English (HKE), Indian English (IndE), and Singapore English (SgE), based on the relevant ICE corpora. More than 58,000 examples were retrieved from the corpora and manually filtered, resulting in 2120 tokens of epicene pronouns. The results show a very different picture for each variety. While overall HKE shows a high preference for the more democratic options they and he or she, IndE and SgE exhibit different patterns. IndE shows singular they in speech, but it is almost non-existent in writing, while in SgE there is a sharp contrast between the most spontaneous spoken register and all other registers. After testing different hypotheses, the findings are explained in socio-cultural terms, as a result of democratization possibly related to women’s movements in those territories.

Keywords

democratization epicene pronouns World Englishes register variation

1. Introduction

Pronouns used to refer to a gender-less antecedent (or to one whose gender is irrelevant) are usually called “epicene pronouns” (e.g., Baron 1981; Newman 1997). Because English does not have specific epicene pronouns (Baron 1981:85), speakers may resort to generic he and generic she (as in 1), he or she—and similar combinations—(as in 2),¹ and singular they (3).

(1) A child needs to be taught his heritage early, or it would be difficult to force it upon him when the influence of other cultures sets in. One effective way is by telling him stories. A busy parent who feels that she does not have enough time for her child can stretch her time by telling stories onto cassette tapes. (ICE-SIN:W2D-020)

(2) <[> They will not </[> </{> ordinarily occur <,> in a person who has got minor problems And then there are indicators that the doctor will ask you immedietly [sic] to stop.[. . .] No no further tests should be <{> <[> done </[> on him or her. (ICE-IND:S1A-068)

(3) You’ve told us the meaning of <.> secre </.> a secretor. That is a person <,> would be expected to posses [sic] appreciable quantity of their blood group substance in the other body fluids such as semen. (ICE-HK:S1B-069)

The choice of any of these pronouns is closely related to democratization processes in that generic he is considered sexist and the result of androcentric grammar (e.g., Bodine 1975), while he or she and singular they have long been proscribed in prescriptive grammar as either clumsy and pedantic, or just illogical, respectively (Bodine 1975:133). Despite prescriptivism, however, democratic (i.e., non-androcentric and egalitarian) options such as he or she and singular they have recently gained ground in inner-circle varieties of English (e.g., Pauwels 2001 for Australian English, AusE; Laitinen 2002 and Paterson 2014 for British English, BrE; Balhorn 2009 and LaScotte 2016 for American English, AmE).² Generic she (used to refer to the neutral term a busy parent in 1), on the other hand, is marginal (e.g., Laitinen 2002:160; LaScotte 2016:66, 67) and, for that reason, it will be omitted from this study.³

Epicene pronouns have tended to be studied from either of the two following perspectives: (i) a strictly linguistic approach that considers the role played by language-internal variables such as the collective or individualized meaning of the antecedent (e.g., Newman 1997; Laitinen 2002), or (ii) a social approach to pronominal variation based on the role played by prescriptive grammar and feminism in the use and expansion of the different forms (e.g., Pauwels 2001; Balhorn 2009; Paterson 2014). The view adopted here leans more clearly towards the second perspective, in that one of the aims is to determine the role played by democratization as a factor conditioning the use of non-sexist pronouns in three Asian varieties of English, although language internal variables will also be considered in the analysis. In addition, the adoption of a register approach (e.g., Biber 1988), based on the analysis of corpora (using the International Corpus of English , or ICE), will allow us to see how democratic options (he or she and they) diffuse through the different varieties.

The term and concept of democratization, described, for example, in Farrelly and Seoane (2012), will need to be used here with caution, as it is a western-centric concept applied to three eastern communities (see Loureiro-Porto & Hiltunen, this issue, for a discussion). In fact, when the term is found in the literature on World Englishes, it can refer to a wide variety of social and socio-political phenomena, such as the writing of a constitution (Schneider 2007); democratization is only occasionally used to refer to the kind of language change described in this issue (e.g., Wasserman & van Rooy 2014; Hackert & Deuber 2015). In order to properly interpret the (social and language-internal) factors conditioning the use of democratic epicene pronouns, the three Asian Englishes will be examined from different perspectives that allow us to capture their singularity.

The focus of this study, then, will be the variation between he, he or she (which includes all the variants in note 1), and they in Hong Kong, Indian, and Singapore English (HKE, IndE, and SgE) with the aim of answering the following questions:

RQ1: Is the distribution of epicene pronouns in these three Asian varieties similar to that of their matrilect, i.e., BrE?

RQ2: What does register variation tell us about the spread and diffusion of democratic epicene pronouns (he or she and singular they)?

RQ3: To what extent can the variation between epicene pronouns be explained in terms of socio-cultural phenomena related to each territory explored in this paper?

The paper is structured as follows. Section 2 describes the theoretical framework. Section 3 presents the methodology and decisions therein. Section 4 analyzes the data, taking into account both language internal and language external factors. Finally, section 5 discusses the main findings and draws some conclusions.

2. Theoretical Background

2.1. Epicene Pronouns: Diachrony and Synchrony

The use of epicene pronouns in English is closely related to the loss of grammatical gender. The traditional view states that, while in Old English times pronouns agreed with the grammatical gender of the antecedent, the leveling of inflections during the Middle English period ushered natural gender (sex) into the linguistic arena (Curzan 2003:42-46).⁴ Nonetheless, as early as Old English, examples of generic he (as in 4), coordinated he or she (5), and singular they (6) are all recorded (Curzan 2003:70-71), but singular they is found mainly with disjunctive coordinators where a plural reading is possible, as in (6).

(4) Ne fornime noðer oðer ofer will butan geð afunge, ðæm timum ðe he hine wille gebiddan, ac geæmtigeað ince to gebedum. (Alfred’s Cura pastoralis 399)

‘Do not, neither of you, deprive the other against [their (his?)] will without consent, at the times when he wants to pray, but have time to yourselves for prayers.’ (from Curzan 2003:65)

(5) weila he seið wa is me þt he oðer heo habbeð swuch word icaht . . . þt is muchel sorhe. for i feole oðer þing he oðer heo is swiðe to herien . . . (Ancrene wisse 47)

‘“Oh dear,” he [the backbiter] says, “Woe is to the one that he or she has caught such talk . . . that is a great sorrow, for in many other ways he or she is greatly to be praised”’ (from Curzan 2003:68; also quoted by Balhorn 2004:96)

(6) Gif oxa ofhnite wer oððe wif, þæt hie dead sien, sie he mid stanum ofworpod . . . (Alfred’s Introduction to laws 32)

‘If an ox gores a man or a woman, so that they be dead, may he [the ox] be killed with stones’ (from Curzan 2003:70-71)

Examples such as (6) have led some authors to claim that singular they has always been used (e.g., Bodine 1975:131), while others (e.g., Balhorn 2004) are more cautious, arguing that the first clear use of singular they is found in Chaucer, illustrated in (7).

(7) That euery wight hath deyntee to cheffare

With hem and eek to sellen hem hir ware. (Man of law’s tale 139-140)

‘That everyone wanted to buy from them and also to sell them their merchandise.’ (Balhorn’s [2004:93] paraphrase)

Though these first examples of singular they are justified in strictly linguistic terms (the possible plural or collective interpretation of the antecedent), social factors have also contributed to the use of (non-democratic) generic he, on the one hand, and of (democratic) he or she and singular they, on the other. The main social factors that have been seen as playing a significant role in the choice of epicene pronouns are prescriptive grammar and women’s movements (here understood as related to social democratization).

The end of the eighteenth century saw the publication of prescriptive grammars that clearly proscribed the use of singular they, which they claimed was illogical, proposing instead the use of generic he (such as Kirby 1746, cited in Bodine 1975). However, he or she and, particularly, singular they, survived in the spoken mode and second-wave feminism revived such forms under the argument that generic he is sexist (Paterson 2014:2-5). The impact of feminism as a democratizing force in language is nicely observed in grammars such as Quirk, Greenbaum, Leech, and Svartvik (1985), which accepts some uses of singular they (as do Biber, Johansson, Leech, Conrad & Finegan 1999:316-317; and Huddleston & Pullum et al. 2002:494); Quirk, Greenbaum, Leech, and Svartvik (1972), by contrast, notes its use as “frowned on in formal usage” (Meyers 1993:182). In fact, democratic epicene pronouns (especially singular they) are the preferred option in different inner-circle varieties of English, such as BrE, AmE, and AusE. As for he or she, an increase is found in BrE and AmE from the 1960s to the 1990s, and a considerable decline in the 2000s (Paterson, this issue).

Balhorn (2004) explores BrE searching pronouns referring to every-compound antecedents in the OED and finds a steady increase of singular they from an initial 9 percent in the sixteenth century to 45 percent in the twentieth century. Over the past century, singular they has continued to increase in frequency, particularly in certain registers, such as journalistic prose, for which singular they represented 11 percent of epicene pronouns in the 1960s and rose to 80 percent in 2007/08 (Paterson 2011:179). Register appears to be an important factor in the expansion of this democratic pronoun, as is also shown by Laitinen (2002:150), who found that singular they represents 84 percent of epicene pronouns with indefinite anaphora in imaginative texts (fiction) in the BNC, and “only” 68 percent in informative texts (including sciences, world affairs, arts, belief and thought, and leisure).

Similar results are found for AmE and AusE. Pauwels (2001) found that singular they increased in non-scripted Australian radio programs from a mere 0.4 percent to 74.9 percent (and he or she from 2.3 percent to 4.5 percent) between 1960 and the 1990s. As for AmE, singular they is particularly common in quoted speech in US newspapers in the early 2000s (circa 90 percent) and also in non-quoted text (60 percent) (Balhorn 2009:399). Adopting a different methodology (questionnaires and multiple participants), LaScotte (2016:67, 74) shows that in the same period singular they is overwhelmingly the preferred epicene pronoun when used with the antecedent the ideal student (75 percent). He also notes that singular they is preferred over he or she by some informants because the latter excludes those who do not identify with the gender binary (the same idea is found in Bjorkman 2017:2). The likelihood that singular they becomes the most popular gender-inclusive pronoun can be expected to increase as sensitivity towards people who do not identify in the binary gains ground in society. Jones and Mullany (2019), for example, defend its use to refer to transgender people.

2.2. Asian Englishes

Despite the attention that epicene pronouns have received in inner-circle varieties of English, their status in the outer-circle remains largely unexplored. For this reason, this paper focuses on three outer-circle (Asian) varieties: HKE, IndE, and SgE. The three share several characteristics, which makes them comparable for a study like this. To begin with, all three have BrE as their matrilect. Further, most of the native languages spoken in these territories exhibit gender-neutral third person singular pronouns, as seen in Table 1. Therefore, language contact effects should not justify differences among the three Englishes analyzed (as also mentioned in Loureiro-Porto 2020:195).

Table 1.

Third Person Singular Pronouns in Asian Substrates (adapted from Loureiro-Porto 2020:195)⁵

Substrate	Asian English Variety*	Nominative	Oblique
Cantonese	HKE, SgE	kéuih (héuih)	kéuih (héuih)
Hindi	IndE	Proximal: yəh Distal: vəh	Proximal: is Distal: us
Tamil	SgE, IndE	Masculine: avan Feminine: ava(ļ)	Masculine: avan Feminine: ava(ļ)
Mandarin	HKE, SgE	Tā	Tā
Malay	SgE	dia / ia	dia / ia
Hokkien	SgE	I	In

This column refers to the English varieties which are potentially affected by the substrates listed in the first column. By no means does it mean that the pronouns in the third and fourth columns are used in any Asian variety of English.

Lastly, the three varieties are documented in the International Corpus of English (ICE) project, and corpora for each of them were compiled in the 1990s and include texts that date from 1990 or later, which allows for synchronic comparisons of the data. The structure of ICE corpora, including twelve different registers (see section 3), also allows for the analysis of register variation, something that has proved to be very significant in the study of World Englishes. In fact, register has come to be considered as significant a variable as region (Kruger & van Rooy 2018:231) and, notably, the registers in the different ICE corpora have been found to be very similar to one another (Kruger & van Rooy 2018:237), which guarantees that the results of the present study will not be biased by corpus compilation decisions.

Notwithstanding these similarities, the three varieties differ in a number of ways, such as the date at which English was first spoken in each territory. It reached India in the 1600s when the country became a trade colony; it entered Hong Kong only in the 1840s, “in the wake of the first Opium War” (Schneider 2007:133); and it arrived in Singapore in 1819, although Singapore did not become a British colony until 1867 (Deterding 2007:2). Following the introduction of English in these territories, the sociolinguistic situation gave rise to two different types of varieties, according to eWAVE (Kortmann & Lunkenheimer 2013). Both HKE and IndE are indigenized L2 varieties, that is, English was introduced in the colonial era via the educational system and is still used in education and other official domains, there never having been a significant number of L1 speakers. As for SgE, eWAVE only records Colloquial Singapore English, classifying it as a high-contact L1 variety, spoken by an ethnically mixed population who consider themselves native speakers of a local variety. No classification is found for Standard Singapore English (SSE).

In addition, the three varieties differ regarding their evolutionary phase, according to Schneider’s (2007) Dynamic Model of postcolonial Englishes, which consists of five evolutionary phases: (1) Foundation (settlers establishment), (2) Exonormative stabilization (English is stabilized in the territories, according to British rules, although the lexicon incorporates localisms), (3) Nativization (mixed codes are used and grammar starts to diverge from BrE), (4) Endonormative stabilization (after political independence, speakers are aware of the idiosyncrasies of their own variety and, on occasions, dictionaries are published), and (5) Differentiation (new varieties emerge). According to Schneider (2007:135-139), HKE is the least advanced of the three varieties here, in that it entered phase 3 in the 1960s and shows no traces of having entered phase 4. Following HKE is IndE, which according to Schneider (2007:171) shows little evidence of having entered phase 4, although Mukherjee (2007:182) claims that it is endonormatively stabilized. Finally, SgE is the most advanced variety, having entered phase 4 in the 1970s (Schneider 2007:155-161).

At the social level, as Brooks (2007:3) has noted, there are regional disparities concerning gender equality in the Asian region, with Hong Kong and Singapore being said to have “a high level of social development.” In fact, Hong Kong has undergone the consequences of intensive industrialization since the 1970s, when “increasing numbers of women entered the labour force” (Göransson 2010:199-200), while Indian women and girls are considered “preservers of the ethnic and religious authenticity [. . .] the bearers of traditional values that are to be inculcated into future generations” (Rydstrøm 2010:11). In such a scenario it is not surprising that the presence of a women’s movement in these territories differs considerably. Thus, because Hong Kong remained a British colony until 1997, the women’s movement in Hong Kong broadly mirrors that of western feminism (Lim 2010:144). Singapore, meanwhile, has enjoyed its own strong feminist tradition since the late 1800s, which made possible the emergence of several feminist societies, such as the Singapore Council of Women (after WWII), the National Council of Women (1975), and the Association of Women for Action and Research (1985) (Lyons 2010:76-80). The autonomous feminist movement in India, however, is said to have developed in the 1970s (Madhok 2010:225), yet by the 1990s the situation of women in public was the following:⁶

1990s woman can step out, but she must demonstrate moral purpose—she is off to work, to buy food, to pick up the kids from school and so on. She must also perform respectability on the street, by avoiding eye contact, looking busy, moving with purpose. We have many reports of sexual harassment, of the difficulties for women of renting property and living without a man or family. (Osella 2017:230)

Nonetheless, it should not be forgotten that all these approximations to gender equality in Asia are made from a western perspective, while the true gender picture in Asia is in fact far more complex, as seen, for example, in the challenge to binary gender represented by Indian hijras, a religious community of individuals who embody male and female characteristics, who prefer to be referred to in the feminine gender, and who belong to an institutionalized third gender (Nanda 1999:ix-xiv). In addition, in 2009 a third gender (beyond the male/female distinction) was made available on ballot forms by the Indian election commission (Osella 2017:241). Thus, any working sociological description of gender equality in these Asian territories can only be taken to be superficial, as a means of contextualizing the possible role of social forces in the use of democratic epicene pronouns.

3. Methodology

The methodology adopted for this study is corpus-based, and the corpora chosen for the analysis of the three Asian varieties are the HKE, IndE, and SgE components of ICE (ICE-HK, ICE-IND, and ICE-SIN, henceforth). As stated, the three corpora were released in the 1990s and the texts they include were produced in that decade. Although this material is more than twenty years old, they are the most recent carefully curated corpora of the varieties under analysis, and their advantages outweigh their limitations. For example, they enable synchronic cross-linguistic comparisons, and they have proven to be tidier than larger, more recent corpora, such as GloWbE (Davies 2013), for example (Loureiro-Porto 2017).

The ICE project aims to provide comparable representative corpora of varieties of English throughout the world (Greenbaum 1996). Each ICE corpus consists of one million words (60 percent of spoken material, 40 percent of written material) in twelve broad registers, as shown in Table 2.

Table 2.

Registers Included in ICE

Spoken	Dialogues	Private	S1A	200,000w
	Dialogues	Public	S1B	160,000w
	Monologues	Unscripted	S2A	140,000w
	Monologues	Scripted	S2B	100,000w
Written	Non-printed	Student writing	W1A	40,000w
	Non-printed	Letters	W1B	60,000w
	Printed	Academic writing	W2A	80,000w
		Popular writing	W2B	80,000w
		Reportage	W2C	40,000w
		Instructional writing	W2D	40,000w
		Persuasive writing	W2E	20,000w
		Creative writing	W2F	40,000w

None of the three corpora is tagged or parsed. Therefore, examples were retrieved using AntConc (Anthony 2014) and then manually pruned. The word forms searched for include all word forms of the lexemes they and he (which rendered examples of both generic he and he or she), namely: they, them, their, themselves, he, him, his, himself, and also non-standard forms, such as themself, theirselves, theirself, and hisself.⁷ The number of scrutinized tokens (58,112) is shown in Table 3.

Table 3.

Scrutinized Forms per Variety

Forms	HKE	IndE	SgE	Total
They	6472	5724	5716	17912
Them	1662	1627	1608	4897
Their	2386	2036	2403	6825
Theirs	9	6	13	28
Themselves	149	178	194	521
Themself	3	1	0	4
Theirselves	1	0	0	1
Theirself	0	0	0	0
He	5220	5913	5550	16683
Him	1075	1231	1183	3489
His	2039	2744	2503	7286
Himself	108	176	182	466
Hisself	0	0	0	0
Total	19124	19636	19352	58112

Pruning these examples implied much decision-making having to do with (i) deciding which antecedents allow for variation between generic he, singular they, and he or she (following Labov’s 2010:550 principle of accountability); and (ii) accounting for morphological variability in the three outer-circle varieties.

Regarding the first of these issues, it was decided that there are probably cultural reasons why generic he will be the default pronoun (i.e., not allowing for variation) in the following cases found in the corpora and, therefore, they were not included in the database. First, whenever the antecedent was narrator, particularly when it was an omniscient one, because in that case narrator does not refer to a gendered individual but to a narrative voice. Second, with culture-bound male referents, such as the runner (in Zeno’s paradox, for example) or God, the orator (in Cicero’s times), because the existence of these antecedents since ancient times in which women did not have a social voice makes them likely candidates to be invariably associated to the pronoun he, maybe not as a generic pronoun, but as an exclusively masculine one. Third, with animals such as dog and grasshopper (where, if anything, he can be in variation with she or with it, but not with singular they). As for singular they, I excluded cases of impersonal they (where no clear antecedent is identified), and antecedents having collective meanings which do not allow an anaphoric he were also excluded, such as countries (e.g., China), companies (e.g., my company), and common collective nouns (e.g., group). In all these cases, they, if anything, stands in variation with it, but not with he or he or she, and, for that reason, these cases were not included in the data.

Identifying the antecedent of a pronoun often requires the analysis of an extensive context, because a short one may lead to misinterpretations. On occasions, however, an extended context is not available. Where a pronoun appeared in a context in which the antecedent could not be retrieved, it was not included in the data.

The second important issue to take into account when pruning these data concerns the variable morphology in outer-circle varieties of English, where plural -s is not as systematically used as in Standard English (Mesthrie & Bhatt 2008:52). This makes it difficult to establish whether an antecedent is singular or plural:

(8) Now if we observe L two learner second language learner <,,> they are mostly adult <,,> because <,,> if they are not adults <,,> and if first language and second language <,,> uhm if a child acquires <,> then even that second language is like first language <,> uhm if the child acquire simultaneously <,,> [. . .] Second language learners are more cognitively mature <,,> They <-> are </-> <,,> have <,> more metalinguistic awareness. (ICE-IND:S2A-041)

(9) And the other composer also affected by him and go to use parody in their drama. (ICE-HK:W1A-014)

(10) When a student go over to France to do a degree perhaps in business or even in engineering how marketable are their degrees. (ICE-SIN:S1B-049)

In sentence (8) the first italicized they is not considered to be an epicene pronoun, despite its antecedent being second language learner, in that the following context suggests that learner is actually an unmarked plural form. Likewise, in (9) the other composer cannot be taken to be a clear singular antecedent, because the verb go does not agree with a singular subject. Something similar happens in (10), where a student, which seems to be undoubtedly singular because of the presence of the determiner a, is followed by the base form go, leaving room for an ambiguous interpretation. For the sake of rigor, and with the aim of minimizing errors, cases such as these were omitted from the analysis, and only very clear instances of epicene antecedents (such as those in 1-3 above) were included in the data.

A total of 2120 tokens of epicenes were found (see Table 4). All of these were then entered into a database including the variables below, which are analyzed through cross-tabulations and descriptive statistics:

Table 4.

Frequency according to Variety

Form	HKE	IndE	SgE	Total
Singular they	146	78	97	321
he or she	116	55	33	204
Generic he	341	603	651	1595
Total	603	736	781	2120

epicene pronoun: he, they, he or she

variety: HKE, IndE, or SgE

antecedent: indefinite pronoun, noun phrase (NP) with quantifier, indefinite NP, definite NP

speaker gender (only for HKE and IndE): woman or man, as found in the ICE metadata⁸

speaker age (only for HKE and IndE): age range assigned to speakers in the ICE metadata

4. Data Analysis

4.1. General Results

The 2120 tokens of epicene pronouns in the three Asian varieties under investigation are distributed as shown in Figure 1 (which is a graphic representation of Table 4). Figure 1 shows a marked and significant difference between HKE,⁹ on the one hand, and IndE and SgE, on the other, the former exhibiting an overall more democratic picture than the latter, with 43.5 percent of democratic epicenes, i.e., singular they and he or she (versus 18.1 percent in IndE and 16.7 percent in SgE). In fact, despite the fact that the three ICE corpora studied here were compiled with texts from the 1990s, the rate of singular they in IndE and SgE is similar to that obtained for BrE in LOB from the 1960s: Paterson (2011:179) finds that singular they represents 11 percent of all epicenes. In addition, it is worth noting that, among democratic options, the three varieties prefer singular they over he or she. Particularly surprising is the frequency of he or she in HKE (116 in the one-million-word corpus), much higher than the frequencies of this combined pronoun in BrE and AmE in the same period (see Paterson, this issue).

Figure 1

Frequency of Epicene Pronouns per Variety

4.2. Type of Antecedent

This section explores the type of antecedents of the three epicene pronouns analyzed, following the classification used by Paterson (2014:50; cf. Paterson, this issue), which presents antecedents on a scale from most indefinite to most definite:

(i) indefinite pronoun (e.g., somebody, everyone), as in (11).

(11) Now when somebody <,> is treated in hospital they get rehabilitation there but they want they will have to go out of hospital. (ICE-HK:S1B-025)

(ii) NP with quantifier (e.g., any person, each child, every teacher, no student, some assistant), as in (12).

(12) So surely there’s no reason to assume that every external law degree holder is inferior. And what’s more to carry that assumption so far as to deny them the chance to prove themselves. (ICE-SIN:S2B-024)

(iii) indefinite NP (e.g., a person, another child), as in (13).

(13) I also felt always <,> that a teacher needs to be free from any obsession <,> with one particular theory of literature <,> or <,> or criticism. If <}> <-> you </-> <+> they </+> </}> want to save themselves from the dangers of distortion. (ICE-IND:S2A-025)

(iv) definite NP (e.g., the person, her friend, the professor’s student), as in (14).

(14) If the child asks money to their parents they give money without checking for what purpose he wants money. (ICE-IND:W1A-004)

Type of antecedent is a key factor in the analysis of epicene pronouns from the perspective of democratization, because there are historical reasons why singular antecedents with a potential plural referent may foster the use of they or the coordinate he or she, as shown in section 2.1. Thus, indefinite pronouns such as everyone, anyone, someone, and NPs with quantifiers such as every student, each neighbor, allow for a collective reading and in fact are recorded historically in both Chaucer (see 7 above) and in Jane Austen, as in (15).

(15) Every body was punctual, every body in their best looks: not a tear, and hardly a long face to be seen. (Emma, 1815)

Therefore, clear evidence for democratization must be sought in antecedent types (iii) and (iv), namely indefinite and definite NPs.¹⁰

Figure 2 sets out the distribution of epicene pronouns per variety, taking into account the four types of antecedents. It shows that the frequency of democratic pronouns is consistently lower with antecedent types (iii) and (iv) than with antecedents (i) and (ii) in all three varieties (35.4 percent in HKE, 9.6 percent in IndE, and 9.9 percent in SgE, which comes ahead of IndE when purely democratic antecedents are considered). This finding is not surprising and goes hand in hand with data for BrE for the twenty-first century, as shown by Paterson (2014:60). She finds that while indefinite pronouns select singular they 95 percent of the time, the percentage decreases to 71 percent with indefinite NPs, and 52 percent with definite NPs. Despite the declining frequency of democratic epicene pronouns in all varieties, the overall picture of the three Asian Englishes remains constant: a clear boundary appears once more between the more democratic HKE (at least as far as epicene pronouns are concerned), with 27 percent and 41 percent of democratic options with indefinite and definite NPs, respectively, and the varieties IndE and SgE, which again exhibit fairly similar percentages, ranging from 8 percent to 13 percent.

Figure 2.

Type of Antecedent per Variety and Pronoun (Raw Frequencies)

4.3. Register Analysis

This section presents the distribution of the three epicene pronouns in the twelve registers included in ICE, reproducing the text codes used in the project (briefly, codes beginning with S indicate spoken registers, and those beginning with W indicate written registers; see Table 2 above for further information). As seen in Figures 3 -7, the frequencies of epicene pronouns in some registers are very low (because the size of each corpus section is small). For that reason, the subsequent analysis will mainly focus on general tendencies, rather than on specific fine-grained differences between registers. Figure 3 shows the first overview of the data. Consistent with the results seen so far, a clear difference is observed between HKE, on the one hand, and IndE and SgE, on the other: HKE exhibits democratic epicene pronouns across all registers,¹¹ and the presence of democratic options in IndE and SgE is restricted to certain registers. Particularly striking in IndE and SgE is the frequency of democratic epicene pronouns in W2F (fiction) and, in the case of SgE, in S1A (private dialogues). The fact that fiction often exhibits features of spoken registers has been explained as a result of the tendency for it to contain (informal) dialogue (e.g., Kruger & van Rooy 2018:231). Given that in section 4.2 we have seen that the type of antecedent is an important factor in measuring the democratic value of singular they, the following figures compare the distribution of epicene pronouns across registers per variety, both when all antecedent types are considered and when only antecedents type (iii), indefinite NPs, and (iv), definite NPs, are taken into account.

Figure 3.

Distribution of Epicene Pronouns across ICE Registers per Variety (Raw Frequencies)

Figure 4.

Distribution of Epicene Pronouns in ICE-HK Registers: Comparison between All Types of Antecedents and Only Antecedents (iii) and (iv) (Raw Frequencies)

Figure 5.

Distribution of Epicene Pronouns in ICE-IND Registers: Comparison between All Types of Antecedents and Only Antecedents (iii) and (iv) (Raw Frequencies)

Figure 6.

Distribution of Epicene Pronouns in ICE-SIN Registers: Comparison between All Types of Antecedents and Only Antecedents (iii) and (iv) (Raw Frequencies)

Figure 7a

Epicene Pronouns in HK: age and gender Combined: All Antecedent Types (Raw Frequencies in Table A1 in the Appendix)

Figure 7b.

Epicene Pronouns in HK: age and gender Combined: Antecedent Types (iii) and (iv) (Raw Frequencies in Table A1 in the Appendix)

Figure 4 shows the results for HKE, where the picture does not differ much if we compare all antecedents with antecedents (iii) and (iv). Beginning with spoken registers (S1A to S2B), we see a clear constant decrease in the percentage of singular they. Since the four spoken registers in ICE correspond to private dialogues (S1A), public dialogues (S1B), non-scripted monologues (S2A), and scripted monologues (S2B), it may well be concluded that the proportion of singular they decreases as does the degree of spontaneity of the text. Likewise, singular they is clearly the preferred democratic pronoun in all registers, with the exception of S2B (scripted monologues), W1A (student writing), W2A (academic writing), W2D (instructional writing), and W2E (persuasive writing). The only difference between the top and the bottom part of Figure 4 concerns the frequency of democratic pronouns in W2C (reportage) and W2F (fiction), which disappears if we exclude antecedents which admit a collective reading (i.e., types i and ii).

Figure 5 shows that IndE spoken registers also exhibit a decreasing percentage of singular they from private dialogues (S1A) to scripted monologues (S2B). In addition, if we exclude antecedents (i) and (ii), Figure 5 shows a notable result: democratic pronouns practically vanish from the picture. In particular, several of those (few) columns in which democratic pronouns are more frequent when all antecedent types are considered in fact disappear when the focus falls only on antecedent types (iii) and (iv), namely: S1B (public dialogue), S2B (scripted monologue), W1B (letters), W2F (fiction). The presence of democratic pronouns is notable only in S1A (private dialogue), W2B (popular writing), and W2D (instructional writing). In addition, when only antecedents (iii) and (iv) are considered, the most common democratic pronoun is he or she, rather than singular they, with the exception of S1A (private dialogue). This important difference between IndE and other World Englishes is also noted by Xiao (2009), whose multi-dimensional analysis places this variety in the most distant position as regards BrE, while other Asian Englishes, such as HKE and SgE, are similar and somewhere in between, especially regarding Dimension 1 in Biber (1988), “Involved vs. Informational Production.” According to this dimension, which in some ways measures the degree of formality, IndE is found to be the most elaborate, BrE is the most interactive, while HKE and SgE are very similar (Xiao 2009:447). As far as epicene pronouns are concerned, IndE is also the furthest from BrE. HKE and SgE, however, do show some important differences.

As shown in Figure 6, the situation in SgE does not change much when all types of antecedents are compared to antecedent types (iii) and (iv). If anything, the elimination of antecedents (i) and (ii) from the analysis makes even more evident the sharp division between S1A (private dialogues) and all other registers, where democratic pronouns are seldom used. In addition to the decreasing proportion of singular they from S1A to S2B (as was the case with HKE and IndE), particularly interesting is the case of W2F, fiction, where all six cases of singular they correspond to indefinite pronouns or NPs with quantifiers. SgE exhibits features that set it apart from IndE, such as the fact that singular they is the preferred democratic epicene pronoun, and that in private conversations (S1A) its frequency ranges from 80 percent (only antecedents iii and iv) to 90 percent (all sorts of antecedents). This variety also differs from HKE in the uneven distribution of democratic epicene pronouns: while in HKE these pronouns are found across registers (see Figure 4), in SgE they are almost wholly restricted to private dialogues. This clear divide between private sections (S1A) and the other eleven registers in ICE-SIN has also been identified in other studies, such as Bao and Hong (2006), who test whether the diglossia between Colloquial Singapore English (CSE) and Standard Singapore English (SSE) is visible in ICE-SIN by characterizing the SSE-CSE diglossia as a case of register variation (2006:106). To this end, two substrate-derived additional meanings of already and also in the four spoken registers of ICE-SIN are studied and the results “point to the sharp division of labor in Singapore English between private dialogue and the other registers” (Bao & Hong 2006:110-111). Their quantitative analysis shows that the CSE-SSE diglossia is supported by corpus data, as “CSE as the vernacular L variant associated with informal occasions and SSE as the local standard H” (Bao & Hong 2006:112). Therefore, the overwhelming preference of singular they in private dialogues in ICE-SIN, which contrasts sharply with all other registers, can also be interpreted in this way: singular they is pervasive in CSE, while it is only marginal in SSE, which, according to Gupta (1994:7), does not differ much from the standard variety of English elsewhere. However, the analysis of epicene pronouns in ICE-SIN provides significant evidence against this claim, as registers other than private dialogues exhibit a far lower frequency of democratic epicene pronouns than BrE in the 1990s, as seen in section 2.1.

Summing up, despite the fact that the frequencies of epicene pronouns in some registers are too low to draw some fine-grained conclusions, the register analysis of these three varieties yields interesting results that point towards general tendencies. Firstly, the three varieties exhibit a higher incidence of democratic epicene pronouns in spoken registers, particularly in the most spontaneous ones, and in fiction, which is expected to include much conversation (e.g., Kruger & van Rooy 2018:231). Secondly, in HKE the use of democratic epicene pronouns is spread across all registers, and singular they is clearly preferred over he or she in the majority of registers. Thirdly, in IndE the use of democratic epicene pronouns is mainly related to antecedent types (i) and (ii), i.e., indefinite pronouns and NPs with quantifiers, because, when these antecedents are not considered, singular they practically vanishes from the picture, and he or she becomes the main, though not very strong, alternative to generic he. Finally, SgE differs from both of its neighbor varieties: unlike in IndE, singular they is preferred over he or she, and, unlike in HKE, democratic epicene pronouns are not spread across all registers, but are almost wholly restricted to private dialogues.

4.4. Speaker Age and Gender

This section considers two external variables that may help explain the variation in epicene pronouns in HKE and IndE, namely the age and gender of the speakers. This will allow us to focus on potential genderlectal differences in the two varieties (Labov 1990), as well as to describe a possible change in progress by adopting an apparent-time approach (Labov 1963). As mentioned above (section 3), the metadata for ICE-SIN are not available, and hence this variety cannot be examined on these lines. Before looking at the results, several clarifications are in order:

The registers for which metadata is available on the speaker’s gender and age are the spoken ones, namely S1A (private dialogue), S1B (public dialogue), S2A (non-scripted monologue), and S2B (scripted monologue).

The dichotomous approach to gender, differentiating only male and female speakers, represents a simplification of reality and excludes speakers who associate with a third gender, such as the hijras in India. Nonetheless, because this religious group has only gained democratic rights recently (2009, according to Osella 2017:241, as seen in 2.2), they are highly unlikely to have been informants in the compilation of ICE-IND, which took place in the 1990s and which stipulated that authors and speakers had to have been educated through the medium of English (according to the ICE compilation criteria). Other gender identities not culturally recognized in India and Hong Kong might, however, have been included in ICE corpora without proper labeling. Unfortunately, that is something that cannot be known, and the analysis of the data must be conducted without forgetting that such simplification of reality is unavoidable in corpora compiled in the 1990s.

ICE-HK and ICE-IND classify age groups differently. Since the original classification is followed here, age groups will vary from one variety to another.

The number of speakers per age group is very unbalanced. However, the rate of epicene pronouns will be presented in percentage form, which normalizes the frequency of each of the three available forms.

Figures 7a and 7b show the percent distribution of epicene pronouns in HKE according to gender and age groups. Figure 7a includes all types of antecedents, while Figure 7b focuses exclusively on antecedents (iii) and (iv). Overall, the HKE picture does not point to any clear tendency, although the youngest groups (less than thirty years old) use democratic pronouns more often than generic he in general. No clear difference is observed between men and women, other than that the former exhibit higher values for the coordinate form he or she (in all its possible variants) than women, who almost exclusively prefer singular they.

Figures 8a and 8b show the picture for IndE. These figures reveal that, contrary to what happened in IndE written registers (see section 4.3), coordinate he or she is disappearing from the spoken mode of this variety, as its use is concentrated mainly in the oldest groups of speakers. In addition, Figures 8a and 8b provide some very relevant new information about IndE: although the overall description showed a much lower proportion of democratic pronouns in this variety than in HKE, if we focus on genderlectal differences, and combine this focus with an apparent-time approach, an altogether new picture emerges, with two important readings. First, IndE women are responsible for the vast majority of democratic epicene pronouns (both when all antecedents are considered, as shown in Figure 8a, and when focusing only on antecedents iii and iv, as shown in Figure 8b). Second, a clear diachronic tendency is observed among IndE women, in that the percentage of singular they increases progressively from the oldest group for which data are available (50-59) to the youngest group (18-25). That women are leaders of linguistic innovation has been noted repeatedly in the literature for inner-circle varieties of English from Labov (1990) onwards (see, e.g., D’Arcy 2004:324; Tagliamonte & D’Arcy 2009:91; Nevalainen, Ramoulin-Brunberg & Mannila 2011). Outer-circle varieties are no exception here, and they confirm that women tend to lead changes (particularly IndE) such as the colloquial discourse particle yaar ‘pal, mate, guy’ (Lange 2009) and the colloquial quotative be like (Davydova 2015). It seems as if Indian women are also leading the democratization of the set of epicene pronouns, with the spread of singular they as a replacement of non-democratic generic he.

Figure 8a.

Epicene Pronouns in IND: age and gender Combined: All Antecedent Types (Raw Frequencies in Table A2 in the Appendix)

Figure 8b

Epicene Pronouns in IND: age and gender Combined: Antecedent (iii) and (iv) (Raw Frequencies in Table A2 in the Appendix)

Summing up, the HKE data do not point towards any clear tendency, other than younger speakers use democratic pronouns more often than generic he, and no difference is observed between men and women. The IndE data, however, reveal several interesting results. Firstly, despite the large proportion of he or she as a democratic pronoun in this variety, this option is disappearing from the spoken mode (it is only used by the oldest group of speakers). Secondly, women produce the vast majority of democratic pronouns. Thirdly, the apparent-time analysis shows that the rate of singular they clearly increases from the oldest to the youngest group, which suggests that Indian women are leading the change towards this specific aspect of the democratization of the language.

5. Discussion and Conclusions

Based on the HKE, IndE, and SgE ICE corpora, this paper has explored the variation between three possible epicene pronouns in 1990s data, one considered sexist, namely generic he, and two considered democratic options, namely he or she and singular they. The results were analyzed from different perspectives, including language external and language internal factors, with the aim of profiling the process of democratization in the three varieties. After noting that the frequency of democratic epicene pronouns is far lower in Asian Englishes than in inner-circle varieties (see, e.g., Balhorn 2009; Paterson 2014), but also far higher in HKE (43.4 percent) than in IndE (18.1 percent) and SgE (16.6 percent),¹² it was shown that the “historical” antecedents of he or she and singular they (namely indefinite pronouns and NPs with quantifiers) do not appear to justify a democratization reading. Meanwhile, antecedents represented by a definite or an indefinite NP (labeled antecedent types iii and iv, respectively) are far more telling of the potential democratization of these varieties, in that no plural reading is grammatically justified in these kinds of antecedents. For this reason, a distinction is established between the whole dataset and the section accounting for antecedent types (iii) and (iv) in the other variables examined. The first consequence of this is that the ranking of varieties in terms of their degree of democratization regarding epicene pronouns is as follows: HKE (35.4 percent) > SgE (9.9 percent) > IndE (9.6 percent).

Several hypotheses can be explored to account for this ranking, the first of which involves the specific phase of development of each variety according to Schneider’s (2007) Dynamic Model. As seen in section 2.2, the most advanced variety is SgE, the least advanced one is HKE, and IndE exists somewhere in between. There is, then, no correlation between degree of progress in postcolonial terms and a preference for democratic epicene pronouns. This seems to be an important finding, because it has been shown that the developmental phase does indeed correlate with degree of informality. As Kruger and van Rooy (2018:237) observe, “the more advanced non-native varieties become along the stages of the Dynamic Model (Schneider 2007), the more they resemble native varieties in their use of informal features in written registers.” However, HKE has been shown to use democratic pronouns in written registers to a far greater extent than the other Asian varieties, although they are more advanced in Schneider’s (2007) model. The same applies to the classification of World Englishes as indigenized L2 varieties (HKE and IndE, according to Kortmann and Lunkenheimer’s 2013 eWAVE) and high-contact L1 varieties (which would be the case of CSE). According to Kruger and van Rooy (2018:231), the lowest degree of informality is to be found in high-contact varieties, which does not help to explain the lowest position of IndE in the ranking here. Therefore, perhaps democratic epicene pronouns cannot be interpreted as an informal feature, an assumption which would help disentangle democratization from phenomena such as informalization, although in fact these have been claimed to be closely related (Farrelly & Seoane 2012:393).

Another hypothesis that might explain the differences between the varieties is substrate influence. However, as shown in section 2.2, the languages that English enters into contact with in these three territories are very similar, in the sense that they all have a gender-neutral third person singular pronoun in their inventory, which would foster the use of democratic options rather than generic he. With different hypotheses thus having been proved to be wrong, an explanation involving different degrees of linguistic democratization in the three Asian varieties of English gains ground.

In fact, converging evidence is found in Loureiro-Porto (2019), a study on the variation regarding another democratization marker (Farrelly & Seoane 2012:393), namely the replacement of modal must with semi-modals have (got) to, need to, want to. Based on a series of morphosyntactic and semantic features related to grammaticalization, the study shows a ranking of Asian varieties of English according to which HKE is more advanced than SgE, which in turn is ahead of IndE (Loureiro-Porto 2019:134). The correlation between that ranking and the one presented here for epicene pronouns seems to suggest that democratization is at work at different levels in Asian Englishes. In the specific case of epicene pronouns, in addition, there are social reasons, these from a feminist perspective, that explain why HKE exhibits the highest rate across registers of democratic epicene pronouns. In fact, different sources point to a higher level of social development in Hong Kong than in other Asian territories (Brooks 2007:3; Lim 2010:144), while the feminist movement in India only emerged in the 1970s, and by 1990 inequality was still very much in evidence in the public role of women (Madhok 2010:225). The differences between HKE and IndE regarding gender equality in language, then, seems to correlate with gender equality in society.

In addition to these general results, this paper also adopted a register approach so as to determine the paths followed in the diffusion of democratic epicene pronouns in World Englishes. The three main findings that emerge are consistent in the analysis of the three varieties. First, singular they is the clearly preferred democratic option in spoken registers (70 percent of all democratic pronouns in HKE, 78 percent in IndE, and 89 percent in SgE), while he or she is essentially restricted to written registers, where it clearly is the preferred democratic pronoun in the three varieties (67 percent in HKE, 70 percent in IndE, and 55 percent in SgE). The contrast between spoken and written registers is too strong to be due to mere chance. A possible explanation is the following. Speakers may be driven by the more economical, “dense,” nature of singular they, as opposed to he or she, in accordance with other densification processes described in the literature, which involve “compacting meaning into a smaller number of words” (Leech, Hundt, Mair & Smith 2009:249-252; see also Baker 2017:238-240); writers, meanwhile, may still feel the prescriptivist load described in section 2.1. This difference between the spoken and the written mode is expected to be eroded once two social events are well established, namely the increasing prescriptive acceptance of singular they, and also an increasing gain in the rights and visibility of non-binary people, who may rather be represented by singular they than by combined he or she (Jones & Mullany 2019; see also Paterson, this issue). A second important finding in the register analysis conducted in this paper is that democratic pronouns are more frequent, in general, in spoken registers than in written ones. Third, when considering the four spoken registers in ICE corpora, we observe that the percentage of singular they declines as the spontaneity of the register decreases. Therefore, it is evident that singular they diffuses from spontaneous, spoken registers to other registers, at least in the cases of HKE and IndE. SgE, though, represents a special case, in that the contrast between private conversations and all other registers is very sharp. Since this finding correlates with other register-based studies of SgE, such as Bao and Hong (2006), the explanation given by authors to account for the sharp divide is supported here, in the sense that the diglossia noted between CSE and SSE is reflected in ICE-SIN: the private conversation section (S1A in ICE terms) is closer to CSE, while all other registers represent SSE. The SgE data, then, confirm Mollin’s (2007) view that register “plays an important role in establishing a version of English as a variety in its own right” (Neumann 2012:75).

Two language external factors were taken into account for HKE and IndE, based on the metadata that accompany ICE corpora. The variables analyzed were gender and age of the speakers (on the four spoken registers in ICE). The most interesting results are obtained for IndE, where, despite the general low frequency of democratic pronouns, genderlectal differences clearly show that women are leaders in the increasing rate of singular they in this variety (he or she being on the verge of extinction in the spoken component of ICE-IND, although it is the preferred choice in several written registers). Data on Indian women, then, confirm Labov’s (1990) idea that women lead language change, as shown in Figures 8a and 8b, which reinforces results from previous studies by Lange (2009) and Davydova (2015) regarding this variety. As for HKE, it was shown (Figure 7a) that the youngest groups of speakers (less than thirty-years-old) use democratic pronouns far more often than generic he, something which is not true for all older groups. The fact that no differences between men and women speakers are observed in HKE (as opposed to IndE) probably means that the use of democratic pronouns in HKE is in an advanced stage, since genderlectal differences tend to neutralize after the initial stage (Ferrara & Bell 1995).

Summing up, the analysis presented here allows us to answer the three questions posed in the introduction. Thus, the presence of democratic epicene pronouns in Asian Englishes does not mirror that of BrE in the 1990s, although HKE is far closer to its matrilect than IndE and SgE (RQ1). The register approach clearly shows that singular they spreads from spoken, spontaneous registers to other registers and also shows that the linguistic situation of SgE is very heterogenous (RQ2). Finally, the answer to RQ3 is affirmative: democratization seems to be the factor that accounts for regional variation regarding gender-equality at the linguistic level, since (i) no other linguistic hypothesis explains the variation, and (ii) social factors such as the advancement of women’s movements in the three territories studied correlate with the relative frequency of singular they. In addition, although IndE exhibits the lowest rate of democratic epicene pronouns, it is also shown than young women are leading the change towards democratization, probably as a parallel phenomenon to the establishment of an autonomous Indian feminist movement, which took place in the 1970s, just twenty years before ICE-IND was compiled. These observations, however, cannot be extrapolated to other varieties of English, where other language internal and language external factors may play different roles. Therefore, future work must explore other democratization markers, from perspectives such as gender equality, among others, so as to continue profiling the paths that this recent process has taken, and still takes, in different varieties of English.

Footnotes

Appendix

This appendix includes the tables corresponding to Figures 7a and 7b (Table A1) and 8a and 8b (Table A2). Because the raw frequencies of some pronouns in each age group were on occasions very low, no statistical test could be used.

Acknowledgements

I am grateful to the reviewers and Turo Hiltunen for their very insightful comments. The usual disclaimers apply.

Funding

The author received financial support for the research, authorship, and/or publication of this article: For financial support, thanks are due to the Spanish Ministry of Economy and Competitiveness (grant FFI2017-82162-P) and the Academy of Finland (decision 258434).

Notes

Author Biography

Lucía Loureiro-Porto is associate professor in the Department of Spanish, Classical, and Modern Languages at the University of the Balearic Islands, Spain. Her research interests involve grammatical variation in World Englishes using corpus-based approaches. She has paid special attention to grammaticalization of modal verbs, variation in epicene pronouns, and general phenomena such as democratization, colloquialization, and Americanization in different varieties of English.

References

Davies

Mark

. 2013. Corpus of Global Web-Based English: 1.9 billion words from speakers in 20 countries (GloWbE). Available online at https://www.english-corpora.org/glowbe/.

International Corpus of English. 1990-. Available through the ICE Project. https://www.ice-corpora.uzh.ch/en.html.

Anthony

Laurence

. 2014. AntConc. Version 3.2.4, retrieved May 2015 from http://www.laurenceanthony.net/software.

Baker

Paul

. 2017. American and British English. Divided by a common language? Cambridge: Cambridge University Press.

Balhorn

Mark

. 2004. The rise of epicene they. Journal of English Linguistics 32(2). 79-104.

Balhorn

Mark

. 2009. The epicene pronoun in contemporary newspaper prose. American Speech 84(4). 391-413.

Bao

Zhiming

Hong

Huaqing

. 2006. Diglossia and register variation in Singapore English. World Englishes 25(1). 105-114.

Baron

Dennis E.

1981. The epicene pronoun: The word that failed. American Speech 56(2). 83-97.

Biber

Douglas

. 1988. Variation across speech and writing. Cambridge: Cambridge University Press.

10.

Biber

Douglas

Johansson

Stig

Leech

Geoffrey

Conrad

Susan

Finegan

Edward

. 1999. Longman grammar of spoken and written English. London: Longman.

11.

Bjorkman

Bronwyn M.

2017. Singular they and the syntactic representation of gender in English. Glossa: A Journal of General Linguistics 2(1), 80. 1-13.

12.

Bodine

Ann

. 1975. Androcentrism in prescriptive grammar: Singular “they,” sex-indefinite “he,” and “he or she.” Language in Society 4(2). 129-146.

13.

Brooks

Ann

. 2007. Gendered work in Asian cities. The new economy and changing labour markets. Aldershot: Ashgate.

14.

Chappell

Hilary

. 2019. Southern Min. In Vittrant

Alice

Watkins

Justin

(eds.), The Mainland Southeast Asia linguistic area, 176-233. Berlin: Mouton de Gruyter.

15.

Curzan

Anne

. 2003. Gender shifts in the history of English. Cambridge: Cambridge University Press.

16.

DailyO. 2019. India has the lowest divorce rate in the world: 5 reasons why it’s not a good thing. https://www.dailyo.in/variety/india-has-the-lowest-divorce-rate-in-the-world-5-reasons-why-its-not-a-good-thing/story/1/29213.html (15 February, 2020).

17.

D’Arcy

Alexandra

. 2004. Contextualizing St. John’s youth English within the Canadian quotative system. Journal of English Linguistics 32(4). 323-345.

18.

Davydova

Julia

. 2015. Linguistic change in a multilingual setting. A case study of quotatives in Indian English. In Collins

Peter

(ed.), Grammatical change in English world-wide, 297-334. Amsterdam: John Benjamins.

19.

Deterding

David

. 2007. Singapore English. Edinburgh: Edinburgh University Press.

20.

Farrelly

Michael

Seoane

Elena

. 2012. Democratisation. In Nevalainen

Terttu

Traugott

Elizabeth Closs

(eds.), The Oxford handbook of the history of English, 392-401. Oxford: Oxford University Press.

21.

Ferrara

Kathleen

Bell

Barbara

. 1995. Sociolinguistic variation and discourse function of constructed dialogue introducers: The case of be + like. American Speech 70(3). 265-290.

22.

Göransson

Kristina

. 2010. Gendered expectations and intergenerational support among Chinese Singaporeans. In Rydstrøm

Helle

(ed.), Gendered inequalities in Asia. Configuring, contesting and recognizing women and men, 191-220. Copenhagen: Nordic Institute of Advanced Studies Press.

23.

Greenbaum

Sidney

(ed.). 1996. Comparing English worldwide: ICE. Oxford: Clarendon Press.

24.

Gupta

Anthea F.

1994. The step-tongue: Children’s English in Singapore. Clevedon: Multilingual Matters.

25.

Hackert

Stephanie

Deuber

Dagmar

. 2015. American influence on written Caribbean English: A diachronic analysis of newspaper reportage in the Bahamas and in Trinidad and Tobago. In Collins

Peter

(ed.), Grammatical change in English world-wide, 389-410. Amsterdam: John Benjamins.

26.

Huddleston

Rodney

Pullum

Geoffrey K.

, et al. 2002. The Cambridge grammar of the English language. Cambridge: Cambridge University Press.

27.

Jones

Lucy

Mullany

Louise

. 2019. The problematic case of gender-neutral pronouns: A response to “A Modest Proposal.” International Journal of Transgenderism 20(2-3). 337-340.

28.

Kachru

Braj B.

1985. Standards, codification and sociolinguistic realism: The English language in the outer circle. In Quirk

Randolph

Widdowson

Henry

(eds.), English in the world: Teaching and learning the language and literatures, 11-36. Cambridge: Cambridge University Press.

29.

Kachru

Yamuha

. 2006. Hindi. Amsterdam: John Benjamins.

30.

Kortmann

Bernd

Lunkenheimer

Kerstin

(eds.). 2013. The electronic world atlas of varieties of English (eWAVE). Leipzig: Max Planck Institute for Evolutionary Anthropology. https://ewave-atlas.org/ (23 April, 2019).

31.

Kruger

Haidee

van Rooy

Bertus

. 2018. Register variation in written contact varieties of English. A multidimensional analysis. English World-Wide 39(2). 214-242.

32.

Labov

William

. 1963. The social motivation of a sound change. Word 19(3). 273-309.

33.

Labov

William

. 1990. The intersection of sex and social class in the course of linguistic change. Language Variation and Change 2(2). 205-251.

34.

Labov

William

. 2010. Principles of linguistic change. Internal factors. Oxford: Blackwell.

35.

Laitinen

Mikko

. 2002. HE and THEY in indefinite anaphora in written present-day English. International Journal of Corpus Linguistics 7(2). 137-164.

36.

Lange

Claudia

. 2009. “Where’s the party, yaar!”: Discourse particles in Indian English. In Hoffmann

Thomas

Siebers

Lucia

(eds.), World Englishes—Problems, properties and prospects, 207-225. Amsterdam: John Benjamins.

37.

LaScotte

Darren K.

2016. Singular they: An empirical study of generic pronoun use. American Speech 91(1). 62-80.

38.

Leech

Geoffrey

Hundt

Marianne

Mair

Christian

Smith

Nicholas

. 2009. Change in contemporary English: A grammatical study. Cambridge: Cambridge University Press.

39.

Lim

Adelyn

. 2010. The Hong Kong women’s movement. Towards a politics of difference and diversity. In Roces

Mina

Edwards

Louise

(eds.), Women’s movements in Asia. Feminisms and transnational activism, 144-165. London: Routledge.

40.

Loureiro-Porto

Lucía

. 2017. ICE vs GloWbE: Big data and corpus compilation. World Englishes 36(3). 448-470.

41.

Loureiro-Porto

Lucía

. 2019. Grammaticalization of semi-modals of necessity in Asian Englishes. English World-Wide 40(2). 115-142.

42.

Loureiro-Porto

Lucía

. 2020. Singular they in Asian Englishes: A case of linguistic democratization? In Núñez-Pertejo

Paloma

José López-Couso

María

Méndez-Naya

Belén

Pérez-Guerra

Javier

(eds.), Crossing linguistic boundaries: Systemic, synchronic and diachronic variation in English, 187-209. London: Bloomsbury.

43.

Lyons

Lenore

. 2010. Transnational networks and localized campaigns: The women’s movement in Singapore. In Roces

Mina

Edwards

Louise

(eds.), Women’s movements in Asia. Feminisms and transnational activism, 75-89. London: Routledge.

44.

Madhok

Sumi

. 2010. Rights talk and the feminist movement in India. In Roces

Mina

Edwards

Louise

(eds.), Women’s movements in Asia. Feminisms and transnational activism, 224-241. London: Routledge.

45.

Mesthrie

Rajend

Bhatt

Rakesh M.

2008. World Englishes. The study of new linguistic varieties. Cambridge: Cambridge University Press.

46.

Meyers

Miriam W.

1993. Forms of they with singular noun phrase antecedents: Evidence from current educated English usage. Word 44(2). 181-192.

47.

Mollin

Sandra

. 2007. New variety or learner English? Criteria for variety status and the case of Euro-English. English World-Wide 28(2). 167-185.

48.

Mukherjee

Joybrato

. 2007. Steady states in the evolution of new Englishes: Present-Day Indian English as an equilibrium. Journal of English Linguistics 35(2). 157-187.

49.

Nanda

Serena

. 1999. Neither man nor woman: The Hijras in India. 2nd edn. Belmont, CA: Wadsworth.

50.

Neumann

Stella

. 2012. Applying register analysis to varieties of English. In Fludernik

Monika

Kohlmann

Benjamin

(eds.), Proceedings of Anglistentag 2011, 75-94. Trier: Wissenschaftlicher Verlag.

51.

Nevalainen

Terttu

Ramoulin-Brunberg

Helena

Mannila

Heikki

. 2011. The diffusion of language change in real time: Progressive and conservative individuals and the time depth of change. Language Variation and Change 23(1). 1-43.

52.

Newman

Michael

. 1997. Epicene pronouns. The linguistics of a prescriptive problem. New York: Garland.

53.

Osella

Caroline

. 2017. Utopia interrupted: Indian sex/gender dissident activism and the everyday search for a life worth living. In Kuldova

Tereza

Varghese

Matthew A.

(eds.), Urban utopias. Excess and expulsion in neoliberal South Asia, 227-246. Cham: Palgrave Macmillan.

54.

Paterson

Laura Louise

. 2011. Epicene pronouns in UK national newspapers: A diachronic study. ICAME Journal 35. 171-184.

55.

Paterson

Laura Louise

. 2014. British pronoun use, prescription, and processing. Linguistic and social influences affecting “they” and “he.” Basingstoke: Palgrave.

56.

Pauwels

Anne

. 2001. Non-sexist language reform and generic pronouns in Australian English. English World-Wide 22(1). 105-119.

57.

Quirk

Randolph

Greenbaum

Sidney

Leech

Geoffrey

Svartvik

Jan

. 1972. A grammar of contemporary English. London: Longman.

58.

Quirk

Randolph

Greenbaum

Sidney

Leech

Geoffrey

Svartvik

Jan

. 1985. A comprehensive grammar of the English language. London: Longman.

59.

Ross

Claudia

Sheng Ma

Jing-heng

. 2006. Modern mandarin Chinese grammar. A practical guide. London: Routledge.

60.

Rydstrøm

Helle

. 2010. Introduction. In Rydstrøm

Helle

(ed.), Gendered inequalities in Asia. Configuring, contesting and recognizing women and men, 1-18. Copenhagen: Nordic Institute of Advanced Studies Press.

61.

Schiffman

Harold F.

1999. A reference grammar of spoken Tamil. Cambridge: Cambridge University Press.

62.

Schneider

Edgar W.

2007. Postcolonial Englishes: Varieties of English around the world. Cambridge: Cambridge University Press.

63.

Tagliamonte

Sali A.

D’Arcy

Alexandra

. 2009. Peaks beyond phonology: Adolescence, incrementation, and language change. Language 85(1). 58-108.

64.

Wasserman

Ronel

van Rooy

Bertus

. 2014. The development of modals of obligation and necessity in White South African English through contact with Afrikaans. Journal of English Linguistics 42(1). 31-50.

65.

Windstedt

R. O.

1913. Malay grammar. Oxford: Oxford University Press.

66.

Xiao

Richard

. 2009. Multidimensional analysis and the study of World Englishes. World Englishes 28(4). 421-450.

67.

Yip

Virginia

Matthews

Stephen

. 2000. Basic Cantonese: A grammar and workbook. London: Routledge.