“Everybody does, don't they?”—A Corpus Study of a Nonsexist Personal Pronoun in Spoken British English

“Everybody does, don’t they?”—A Corpus Study of

a Nonsexist Personal Pronoun in Spoken British

English

Martin Persson

C Paper in Linguistics

Uppsala University, Department of English

Supervisor: Pia Norell

December 17, 2006

Abstract

This paper deals with the use of third person pronouns with generic refer-

ence to compound indefinite pronouns in spoken British English. Previous

research on this topic shows a decline in the use of the masculine generic

he, which some may consider sexist, and an increased use of the gender-

neutral they, which is one of the proposed nonsexist alternatives to generic

he. The paper contains two introductory sections which give appropriate

background information on language and gender, sexist and nonsexist lan-

guage and pronominal reference. In the third section, the findings of the

study are presented. The study was carried out by investigating the fre-

quency of the third person pronouns he, she and they with reference to the

indefinite compound pronouns (e.g. anybody) in spoken British English. The

data were collected from the British National Corpus, a corpus of contempo-

rary British English. The study shows that they is used in a clear majority

of the cases (about 95 percent), and that generic he is very infrequent, even

in formal registers. This paper also suggests that contemporary grammars

probably lag behind actual usage by stating that they used singularly and

generically is a feature of informal language only.

Contents

1 Introduction 2

1.1 Aim and Scope . . . . . . . . . . . . . . . . . . . . . . . . . . 3

1.2 Previous Research . . . . . . . . . . . . . . . . . . . . . . . . 3

1.3 Material and Method . . . . . . . . . . . . . . . . . . . . . . . 4

1.4 Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

2 Background 6

2.1 Language and Gender . . . . . . . . . . . . . . . . . . . . . . 6

2.2 Sexism in English and the Purpose of Nonsexist Language . . 8

2.3 Pronominal Reference . . . . . . . . . . . . . . . . . . . . . . 9

2.4 The Personal Pronouns of English . . . . . . . . . . . . . . . 10

2.5 Indefinite Compound Pronouns . . . . . . . . . . . . . . . . . 11

2.6 Generic He . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

2.7 Singular They . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

3 Presentation of Results 14

3.1 3PPs with antecedent Anybody . . . . . . . . . . . . . . . . . 14

3.2 3PPs with antecedent Anyone . . . . . . . . . . . . . . . . . . 15

3.3 3PPs with antecedent Everybody . . . . . . . . . . . . . . . . 16

3.4 3PPs with antecedent Everyone . . . . . . . . . . . . . . . . . 17

3.5 3PPs with antecedent Nobody . . . . . . . . . . . . . . . . . . 18

3.6 3PPs with antecedent Somebody . . . . . . . . . . . . . . . . 18

3.7 3PPs with antecedent Someone . . . . . . . . . . . . . . . . . 20

3.8 Distribution of 3PPs over Contextual Domains . . . . . . . . 21

4 Concluding Remarks 22

1

1 Introduction

The quest for a gender-neutral third person pronoun has been going on

for a long time. Since the 1880s, when a gender-neutral pronoun became a

popular discussion topic in literary academic journals, numerous suggestions

have been put forward to fill the gap in the English pronoun system (Baron

1985), whose forms are gender-neutral except for in the third person singular.

There we have to choose between she or he when referring to human beings.

Or are these really the only options we have? According to the standard, to

the authoritative grammars, they are, if we do not want to be considered

incorrect. But how do people in general, actual speakers of English, fill this

gap, the missing gender-neutral third person singular pronoun?

Studies show that the use of the so called singular they (henceforth re-

ferred to as sing. they), where the usually plural pronoun they have singular

reference, is becoming increasingly popular. Sing. they is used primarily with

generic reference (referring to people in general), as in (1), but it is also ob-

served with specific reference, as (2), a dialogue from an American TV series

(The L Word), is an example of.

(1) Everyone should mind their own business.1

(2) Tania I didn’t even ask you, why are you here? Are you OK?

Alice Yeah, uh, I’m just.. well.. visiting a friend.

Tania I hope they’re alright.

Alice Yeah, they’re.. I mean, they’re getting better.

This paper presents a study of the frequency of generic sing. they in

spoken British English (from now on referred to as BrE). This introduc-

tory section presents the aim of the study, previous research on the topic,

the primary material and methodology, and definitions of some important

terms. In Section 2, a general background is given, dealing with language

and gender, sexist and nonsexist language, personal and indefinite pronouns,

generic he and sing. they. The results of the study are presented in Section

3, and final concluding remarks are made in Section 4.

1Author’s own example.

2

1.1 Aim and Scope

The aim of this study is to examine the frequency of sing. they in spo-

ken BrE. Apart from looking at the overall frequency of this grammatical

feature, a second purpose of this study is to find out if sing. they is more

frequent with some indefinite pronouns (i.e. someone, somebody, everyone,

everybody, anyone, anybody and nobody2) as antecedents than with others.

A third purpose is to investigate if sing. they is confined mostly to informal

language, an assertion made by some contemporary grammars (Greenbaum

1996, Quirk et al. 1985). Therefore, formal and informal speech will be com-

pared as regards the use of sing. they.

The study is limited to sing. they with reference to indefinite compound

pronouns only, as in (1) above. Although it occurs occasionally with reference

also to common-gender nouns (as in (2) above) as well, such occurrences are

not as easily identifiable in the data as where the indefinite pronouns can

be used as indicators, often triggering generic pronouns.

Furthermore, the study is limited to spoken BrE, and hence, neither

other varieties of English (e.g. American) nor written language will be looked

into.

My hypothesis is that sing. they is very frequent in spoken British En-

glish, and undoubtedly used more often than generic he. Regarding formal

and informal spoken language, I doubt there is much of a difference when it

comes to the use of sing. they, since previous research (which I will present

below) shows a general decline in generic he and an increase in the use of

sing. they. Regarding the question if sing. they is more frequent with some

indefinite pronouns than with others, my hypothesis is that they might be

more common in co-reference with everybody and everyone, since they are

often plural in meaning (see Section 2.5).

1.2 Previous Research

Previous research on the topic often shows a preferred use of sing. they over

generic he.

2No one gave too few hits in the corpus, and was therefore discarded.

3

Baranowski (2002) investigated the use of generic pronouns in American

and British newspapers. He concludes that “he is no longer the preferred

singular epicene3 pronoun in English”, and even that “singular they is now

the predominant form” (2002:295) in cases where before you would have used

the masculine generic he. Baranowski also observes that American writers

tend to be more conservative as regards generic pronouns, and less likely to

use sing. they than British writers.

Pauwels (2003) looked into generic pronominal use in spoken Australian

English. She compared radio programs and parliamentary debates from the

1960s, 1970s and 1990s. She observes a “steep decline in the use of generic

he” and a “significant turnaround for singular they” (2003:563), probably

encouraged by the second-wave feminist movement in the 70s and 80s, which

drew attention to and increased people’s awareness of sexist language.

In another study, Pauwels (2003) explored the use of generic pronouns

by (Australian) academics. Again researching spoken material, this time

focusing on university lectures, she confirms the marginalisation of generic

he4. Interestingly, Pauwels also found that the academics’ preferred choice

of generic pronoun was he or she, whereas other speakers would use sing.

they most frequently (2003:564).

Although a good deal of research has been carried out on generic pro-

nouns, it seems to cover mostly written language, and, in some cases, formal

speech. Spoken BrE has been sparsely investigated in this area. Therefore,

the focus of this study, spoken BrE, formal and informal, seems justified.

1.3 Material and Method

This study uses the British National Corpus (BNC) as its primary material.

This is a corpus of contemporary spoken and written British English. The

total size of the corpus is 100 million words, and the spoken part, which this

study examines, constitutes 10 million (i.e. 10%) of these5.

The spoken data of the BNC is divided into 4 registers, or “domains”

3Unisex4In Pauwels study, it constitutes only approx. 12% of the generic pronouns found.5Corpus information was retrieved from About the British National Corpus.

4

as they are labelled in the corpus: “Educational/Informative”, “Business”,

“Public/Institutional”, and “Leisure”. This categorisation makes it possi-

ble to separate informal (Educational/Informative and Leisure) and formal

(Business and Public/Institutional) speech in the collected data, which is

one of the aims of this study.

The corpus was searched for the third person pronouns he, she and they

(and their inflected forms6) following a compound indefinite pronoun (e.g.

anybody or someone) in a span of 7 words. This was done with one indefinite

pronoun at a time. The queries were randomly thinned down to give 150 hits

per indefinite pronoun, given the limited time for the study. Then followed

the time-consuming process of manually singling out the hits where the

indefinite pronoun was the actual antecedent of the personal pronoun, i.e.

identifying personal pronouns in co-reference with the indefinite pronoun.

Hits where the personal pronoun had another antecedent than the indefinite

pronoun were discarded. The results (now containing only personal pronouns

with reference to the indefinite pronoun in question) were then filtered on

different domains (i.e. formal/informal).

1.4 Terminology

The term gender-neutral is used for pronouns not denoting gender, for ex-

ample they. For pronouns denoting gender, the term gender-specific is used.

The terms gender-inclusive and gender-exclusive are also important when

discussing pronominal reference. They signal if the reference of the pronoun

in question includes both genders, or excludes one of them. The basic terms

gender, sex, and sexism are defined in Section 2.1.

A general distinction is made between generic and specific reference of

pronouns. Pronouns with specific reference point to a specific person or thing

in the text or in the context, whereas pronouns with generic reference point

more generally to people (or things). Most of the pronouns discussed in

this paper have generic reference. Further definitions of terms dealing with

pronominal reference are presented in Section 2.3.

6i.e. the subjective, objective, possessive, and reflexive forms were included for each3PP.

5

Abbreviations used in this paper include 3PP for third person pronoun

(i.e. he, she, it and they), the BNC for the British National Corpus, and

BrE for British English. Singular they is shortened to sing. they.

2 Background

2.1 Language and Gender

When discussing language and gender, it is first necessary to explain one’s

definition of the term gender, in order to avoid ambiguity and confusion.

In society, gender is the broad term for the different social roles of women

and men. In contrast to sex, which is a solely biological feature, gender

is something that is learnt—it is socially constructed. As Talbot (1998:7)

puts it, “people acquire characteristics which are perceived as masculine

and feminine”. The words feminine and masculine can be said to denote

gender, whereas the corresponding words denoting sex are female and male.

Since we live in a society in which sex difference imply gender difference

(e.g. that a female should be feminine), the distinction between gender and

sex (i.e. the distinction between what is social and what is biological) is quite

often forgotten, or ignored. The fact that people confuse biological differ-

ences between women and men with socially constructed gender differences

is unfortunate, according to Talbot (1998:9), since it prevents women and

men from breaking free from their traditional gender roles. Cameron (1993)

points out that it should not be forgotten, when discussing gender roles,

that the feminine and the masculine roles are not only different—but also

unequal. She claims that there is inferiority built into the feminine role, and

superiority built into the masculine (Cameron 1993:100). This fundamental

unbalance between the roles of women and men in society is generally re-

ferred to as sexism. Sexist language will be discussed further on in the paper,

but let us first take a look at the different roles gender plays in linguistics.

In language, we find grammatical gender and natural gender. Latin,

French, Italian and German are examples of languages with grammatical

gender. In these languages, all nouns are placed into gender categories, and

6

this classification then governs the words’ grammatical behaviour, such as

inflectional endings, adjective agreements, article use and pronoun choice

(Cameron 1992). In German and Latin7, there are 3 categories of grammat-

ical gender: feminine, masculine, and neuter. French and Italian have only

2 categories, feminine and masculine. But these are nothing but abstract

grammatical categories, which have little to do with biological sex8, since

apparently sexless words can be masculine or feminine, as is the case with

the Italian mela (apple), which is feminine, and gelato (ice cream), which is

masculine.

In English, however, we find only natural gender. Natural gender is when

the biological sex of a person determines certain grammatical features, as is

the case with the English third person singular pronouns. If you are speaking

about a girl, you will use the pronoun she and if you are speaking about a

boy, the appropriate pronoun is he. Non-human words, such as book, knitting

and rice, will go with the pronoun it. Animals sometimes go with it and

sometimes with he or she, depending on their relationship to the speaker9.

Natural gender also sometimes determines the ending of words, for example

in widower/widow, host/hostess and waiter/waitress. There are also words of

common gender, which refer to a human (animate) being of either sex, such

as student, cousin and friend. Gender with these words is revealed by choice

of pronoun (She is my friend), or by adding a gender-specific adjective, as

in male cousin (Crystal 1984).

Why is language and gender studies important? The answer to this ques-

tion depends on what role we believe language to have in society. The Sapir-

Worf hypothesis suggests that language is a mirror of society, a mirror which

not merely reflects society, but also influences it (Chalker and Weiner 1998).

The belief that structures we find in society, we will also find in language, is

widely accepted among linguists (feminist or not), and it initiated the study

of sociolinguistics. Applying the theory of language as a mirror to gender in-

7Latin is the language from which grammatical gender originally derives.8It is not true, really, to say that grammatical gender has nothing at all to do with

social gender, Cameron (1993:91 ff.), for instance, problematises this assertion.9i.e. pets are usually called he or she by their owners.

7

equalities, Talbot (1998) recognises the importance of language and gender

studies (from a feminist perspective) in two ways. First, by looking upon

language as a mirror of society, studies on language and gender can help us

identify and understand gender relations and social power structures. Sec-

ond, if we believe that language not only reflects society, but also has an

influence on it, studies on language and gender may help us to “raise people’s

awareness [of sexism] and stimulate social change” (Talbot 1998:15).

2.2 Sexism in English and the Purpose of Nonsexist Lan-

guage

Before I move on to describing generic he and sing. they in detail, let us

first take a brief look at some linguistic features of English which may be

considered sexist (i.e. indicators of male dominance), and some suggested

nonsexist solutions for these.

The first example of sexist English that the authors of The Handbook of

Nonsexist Writing (Miller and Swift 2001) take up is the generic use of man

to refer to people of both sexes, as is the case with the sentence Man is a

dreamer. Here, man refers to the human species in general, and not only to

males. This leaves women linguistically in the shade of men, being reduced

only to “variants of males”. A nonsexist rephrasing of this sentence could

be Human beings are dreamers.

Another area in which linguistic sexism appears is occupational titles.

Many traditionally male dominated occupations have titles ending in the

suffix -man, as in fireman, fisherman and policeman. Here, Miller and Swift

(2001) propose using the gender-neutral fire fighter, fisher and police officer

instead.

There are also occupations where, even if the occupational title in itself

does not contain a masculine affix (e.g. -man), the title is traditionally un-

derstood as masculine. Sexist norms have created a “need” for indicators

of women in these occupations, such as lady doctor and woman writer or

authoress (Talbot 1998).

If we turn to social titles, we see another sexist pattern. Women are

8

addressed according their marital status (Miss or Mrs), whereas men are

not (they have only Mr). In this field, however, the nonsexist alternative Ms,

which is used by women (whether married or not) who refuse to be defined

in terms of their marital status, has been very successful, and is now the

unmarked female title in the United States (Cameron 1993).

Cameron (1993) identifies a general sexist pattern in men’s addressing of

women. Men commonly address women as subordinates, by calling them by

their first names or using so called endearment terms, e.g. dear or honey. “A

man can choose to call any woman dear unless she is directly in authority

over him”, but “if a woman calls a man dear she is normally either older

than him or intimate with him” (1993:105).

Lexicographically, there seem to be more insulting words for women than

for men. Spender (1985:15) refers to a study which found 220 words for a

sexually promiscuous female, but only 20 for a sexually promiscuous male.

This, according to Cameron, reflects the sexist idea that “women should

have no sexual desires whereas men should be insatiable” (1993:108).

Few people believe nonsexist language to be the one method with which

to fight gender inequalities, but some acknowledge its influence on society,

and decide to fight sexism linguistically as well as in other fields. “We can de-

mand equal pay and nonsexist language too”, Cameron (1993:2) concludes.

2.3 Pronominal Reference

One way of defining pronouns is to say that they “replace nouns”, or “stand

instead of nouns”. When taking a closer look at pronouns, however, we find

that it is more a question of reference than of substitution (Wales 1996). Pro-

nouns can be used to “substitute” a noun, by means of avoiding repetition,

for example, as in (3) below, but they need not always do that.

(3) Sarah sat down on the sofa. She was exhausted.10

(4) I wouldn’t do that if I were you.

(5) It is nice to have a cup of coffee in the morning.

10Author’s own examples.

9

Consider example (4) above, where neither I nor you can be said to

substitute something. Instead, we can talk about reference. In (4), I refers

to the person speaking, and you refers to the addressee. They have exophoric

(or contextual) reference, referring to something outside the text. In (3)

above, we have an example of endophoric (or textual) reference, reference

within the text. Here, she refers back to Sarah. Backwards reference is called

anaphoric reference. The pronoun she refers to its antecedent, Sarah, which

refers to the actual person (exophorically) Sarah. But we can also consider

both the pronoun she and the (proper) noun Sarah to refer to the person

Sarah. They are then said to be co-referential with each other. Pronouns

referring forwards in the text have cataphoric reference, which (5) above is

an example of.

2.4 The Personal Pronouns of English

Table 1: The Personal Pronouns of English (subjective case)

Person Singular Plural

1st I We2nd You You3rd He, She, It They

Table 1 lists the English personal pronouns. We can see that it is only

in the third person singular that the standard system lacks a gender-neutral

form for human reference. But people do not only use the strict systems of

grammarians when they produce language. Even in the category of pronouns,

which is considered to be a closed word class (where we do not normally take

in new words), there is a considerable amount of variation and change. Wales

(1996:xii) claims that:

Users of English have always created their own “systems” ofpronouns or “rules” of use for their own needs and strategies. . . often in complete disregard of grammarians’ notions of logic.

Therefore, instead of treating pronouns prescriptively, Wales argues, it is

10

more rewarding to look at them from a descriptive perspective.

If we take the case of sing. they, for example, it is only recently that

contemporary grammars have accepted it, even though it has been used by

a great many English speakers for a long time (see Section 2.7). This reflects,

according to Wales (1996:126), “the inheritance of prescriptivism on even

contemporary grammars”.

There is another pronoun in English which, similarly to they, started

by having plural-only reference, but expanded to cover singular reference as

well, namely you. Nowadays, we do not find this form strange or confusing,

and this shows that pronouns are not a fixed grammatical category at all,

but a living word class which constantly changes.

2.5 Indefinite Compound Pronouns

The indefinite compound pronouns are any-/every-/no-/some- attached to

-body or -one. They are indefinite in that they “[lack] the definiteness of

reference inherent in personal, reflexive, possessive, and demonstrative pro-

nouns” (Chalker and Weiner 1998:200).

The indefinite compound pronouns are “singular in form, but often plural

in meaning”, as Baron (1985:193) puts it. This is especially true for everybody

and everyone, which “notionally . . . clearly denotes more than one person”

(Wales 1996:128).

The indefinite compounds can also vary in reference. Sometimes they

point to a specific person (or specific persons) in the context, but more

often they have generic reference, pointing to people in general. Examples of

indefinite pronouns with generic as well as specific reference will be presented

in Section 3.

2.6 Generic He

One grammatical feature in English which has acquired much attention in

the debate on sexist language is the so called generic he. Generic he is the

traditional use of a masculine 3PP to refer to persons of both sexes (or of

either sex), as in the following sentence:

11

(6) Everyone holds his breath for half a minute.11

Historically, this use has been prescribed by grammarians since the 18th

century, when the authoritative grammarians decided that only one pro-

noun, the masculine, was the ”correct” form to use for generic reference

(referring to people in general) (Wales 1996). At that time, several options

for making generic reference (including sing. they, as we shall see further

on) were available. But the grammarians, who were “preoccupied with rules

of ’concord’ or ’agreement’ on the basis of Latin models of grammatical

description” (Wales 1996:112), decided in favour of generic he, which, it

was argued, agreed with a singular gender-neutral antecedent (In (6), the

antecedent is everyone) both in person (third) and in number (singular).

Critics of generic he, however, have argued that even though it agreed

in person and number, it disagreed in gender (Baranowski 2002:378). The

grammarians of the 18th century maintained that, as described by Wales

(1996:113):

generic he was semantically justifiable on the grounds that thelanguage (like the culture) was indeed biased in favour of males,who were the “superior” . . . or the “worthier” sex, and that fe-male reference could be ”subordinated” to male reference of thiskind.

But many women (some of them feminists) have refused to accept this

sexist convention (c.f. also the case of generic man discussed above), feeling

excluded in reference. Wales (1996:114) claims that “the pronoun he is so

crucially important for male reference in modern English”, and can therefore

not be a “true generic”, without “male connotations whatsoever”. She calls

it “pseudo-generic he”, emphasising the weak semantic gender-inclusiveness

of generic he.

There are several alternatives, more or less convenient (and more or less

accepted), if you want to avoid using the masculine generic. These include

sing. they, he or she, (s)he or s/he, generic she, pluralising the sentence, and

11Author’s own example.

12

neologisms such as thon or hesh. Let us now take a closer look at one of

these alternatives, namely sing. they.

2.7 Singular They

The use of they with singular reference can be traced back to the 16th

century. It has been “well established in informal usage for centuries” (Wales

1996:126). Baron (1985) lists some “major English writers” who have used

sing. they extensively, among them Jane Austen12.

But what are the advantages of sing. they over the other alternatives to

generic he? Wales (1996:126) sums it up as follows:

It is certainly pronounceable [c.f. writing-only variants such ass/he], well established and non-“alien” [c.f. hesh and thon], nor-mally unstressed and unfocussed, and brief . . . it “avoids” cum-bersome alternatives [c.f. he or she], and it is itself also “difficultto avoid” across clause boundaries and in tag questions.

The main advantage of sing. they over generic he is, according to Wales

(1996), its unmarkedness for gender. It fills the gap of a gender-neutral

singular 3PP conveniently.

In the previous section, I discussed the opinions on generic pronouns of

the grammarians of the 18th and 19th century. But what about contempo-

rary grammars, what do they say about sing. they? Greenbaum (1996:20)

sees sing. they as a convenient way of avoiding a (possibly sexist) generic

he. Moreover, he considers sing. they to be mostly a feature of (informal)

speech, but predicts it becoming “[increasingly acceptable] as the generic

singular even in formal style”, in the future. Quirk et al. (1985:342) recog-

nise the use of sing. they with generic reference, but point out its informality

and “defiance of strict number concord”. Echoing Wales (1996), they also

note that they is practically unavoidable in tag questions and across clause-

boundaries.

12See http://www.crossmyt.com/hc/linghebr/austheir.html for further informationabout sing. they in Jane Austen’s works.

13

3 Presentation of Results

Out of a total of 1050 hits (150 for each indefinite pronoun investigated) in

the BNC, 382 hits where the 3PP referred to the indefinite pronoun were

found13. Table 2 below shows the distribution of 3 different 3PPs (they, he,

and she) with the 7 indefinite pronouns investigated14 as antecedents. As can

be seen, they is the predominant 3PP to refer to indefinite pronouns, being

used in nearly 95 percent of the cases. He is only used in about 4.5 percent

of the cases, and she occurs only three times, which makes a percentage of

about 0.8.

Table 2: 3PPs with an indefinite pronoun as antecedent

Pronoun Number (382) %

They 362 94.76He 17 4.45She 3 0.79

3.1 3PPs with antecedent Anybody

In Table 3 below, the distribution of 3PPs with anteceding anybody is pre-

sented. 54 of 56 3PPs were they, constituting about 96 percent of the total

3PPs. In the remaining 2 cases, we find he, which was used in just under 4

percent of the cases.

Table 3: 3PPs with antecedent anybody

Pronoun Number (56) %

They 54 96.43He 2 3.57She 0 0

13In the other cases, the 3PP had further anaphoric reference to a specific person in thecontext.

14These were anybody, anyone, everybody, everyone, somebody, someone, and nobody.

14

Some examples from the data are:

(7) Jean Oh I wouldn’t mind but I wouldn’t marry anybody

Brenda You might live with him.

(KBF 8722–23)

(8) No sympathy for anybody like that, they know what they ’redoing. They ’re not much cop when they can open their legs likethat.

(KSS 5021–22)

In (7), he is used to refer to anybody. From the context, however, we gather

that Jean and Brenda are discussing future male partners, and thus the

reference of he (and, accordingly, the reference of anybody) is limited to

men only. The pronouns are generic in the sense that it can refer to any

man, but they are specific in the sense that it refers to men only, being

gender-exclusive rather than gender-inclusive.

(8) is an example of sing. they, where anybody like that and they have

co-reference to the same exophoric person or persons.

3.2 3PPs with antecedent Anyone

Table 4 shows the distribution of 3PPs with anteceding anyone. Again, about

96 percent of the 3PPs are they. Also similarly to anybody, about 4 percent

of the 3PPs are he.

Table 4: 3PPs with antecedent anyone

Pronoun Number (52) %

They 50 96.15He 2 3.85She 0 0

Let us have a look at these two cases of generic he:

15

(9) Anyone can set his newspaper in the New York Times letteringstyle without paying the Times a nickle <laugh>.

(GYU 153)

(10) You could not . . . go down er a pit and say to anyone down a pit,and say to anyone down a pit, that he’s not a miner.

(KS2 508)

Both (9) and (10) are examples of a generic use of he. Here, the reference

to traditionally male-dominated occupations (typesetters (9) and miners

(10)) is probably what triggers the use of a masculine generic pronoun15.

3.3 3PPs with antecedent Everybody

Table 5 shows the distribution of 3PPs with everybody as antecedent. Of the

70 3PPs found, all of them were they. Neither he nor she with reference to

everybody was found.

Table 5: 3PPs with antecedenteverybody

Pronoun Number (70) %

They 70 100.00He 0 0She 0 0

Below are some examples from this part of the data:

(11) Everybody does, don’t they?

(KCL 1514)

15c.f the use of occupational titles discussed in Section 2.2.

16

(12) And nobody knew anything about spores or farmer’s lung or an–Everybody are very conscious and they wear masks when they ’refeeding hay now.

(HEM 389)

The plural verb forms of (11) and (12) confirms a strong plural mean-

ing within everybody (as discussed in Section 2.5), and this also makes the

choice of a plural 3PP more logical. Also, both (11) and (12) exemplify the

assumption made by contemporary grammarians (see Section 2.7) that it is

hard to avoid sing. they in tag questions (11) and across clause boundaries

(12). Everybody does, doesn’t he, or Everybody are very conscious and he

wears masks would not seem as natural as (11) and (12) do.

3.4 3PPs with antecedent Everyone

Table 6 shows the distribution of 3PPs with reference to everyone. 60 such

3PPs were found, 59 they (about 98 percent), and 1 he (just under 2 percent).

Table 6: 3PPs with antecedent everyone

Pronoun Number (60) %

They 59 98.33He 1 1.67She 0 0

Let us have a look at some examples of they :

(13) Paul Last time I went in everyone got called!

Terry Yep, they all that <unclear>

(F7G 685–86)

(14) So everyone, everyone has his own

(KCV 1370)

17

(15) What I’m trying to say is, everyone joins for their own differentreasons, you can’t generalise that.

(K74 264)

Again, the plural meaning of everyone, similarly to everybody, seems

strong. An indicator of this is (13), in which everyone is referred to as they

all. It can be argued that when he is used, as in (14), it stresses the singular-

ity of the separate units of the indefinite pronoun. This is not very probable,

however, since we find more cases with they expressing this singularity of

meaning, as in (15).

3.5 3PPs with antecedent Nobody

Table 7 shows the distribution of 3PPs with nobody as antecedent. All of

the 38 3PPs found were they. Neither he nor she with reference to nobody

was found.

Table 7: 3PPs with antecedent nobody

Pronoun Number (38) %

They 38 100.00He 0 0She 0 0

(16) below shows the use of the plural reflexive with reference to nobody.

(16) . . . nobody regards themselves as aged these days do they?

(G4F 65)

3.6 3PPs with antecedent Somebody

In Table 8, the distribution of 3PPs with anteceding somebody is presented.

51 such 3PPs were found. 47 (about 92 percent) of these were they, and

3 (about 6 percent) of them he. Interestingly, we also find one she with

reference to the indefinite pronoun in question (see (17) below).

18

Table 8: 3PPs with antecedent somebody

Pronoun Number (51) %

They 47 92.16He 3 5.88She 1 1.96

Let us again consider some examples from the data:

(17) I asked somebody the way out and she offered to show me . . .

(KBP 1252)

(18) I mean I actually had somebody last Friday, they told me whatto do, a four letter word, they told me what to do, and put thephone down.

(JSN 745)

(19) . . . to deprive somebody of the right to express their opinions inpublic would be regarded as an infringement of his civil liberties,which would be unthinkable in a republican erm situation.

(KRH 2403)

In (17), somebody is already specific, referring anaphorically to a specific

person who was asked for the way out by the speaker. The fact that the

person asked was a female triggers the feminine 3PP here. However, in a

similar case of somebody with specific reference, (18), they is the 3PP used.

Here it is “the somebody from last Friday” who is being discussed.

Moving on to a more generic usage of somebody, (19) is an interesting

example of a speaker first using they and then switching to he. The speaker’s

switching to the masculine 3PP may perhaps be due to the strong formal

fixedness of the expression civil liberties.

19

3.7 3PPs with antecedent Someone

The distribution of 3PPs with anteceding someone is presented in Table 9.

Out of 55 3PPs found, 44 (80 percent) were they, 9 (about 16 percent) were

he, and 2 (almost 4 percent) were she.

Table 9: 3PPs with antecedent someone

Pronoun Number (55) %

They 44 80.00He 9 16.36She 2 3.64

Let us look at some of these cases in detail:

(20) Erm, and it can, it can happen overnight, can’t it, from some–someone is er, he’s he’s he’s a bit of the one with the ladies.

(KGP 1071)

(21) no, when you get someone’s relative phoning up because they ’vejust been crushed under a lump of machinery, that’s is veryfraught.

(KE3 3814)

(22) eventually after some searching they nominated someone whothen resigned from her post

(F7C 1259)

(23) Someone put her poo in my locker16

(KPA 1844)

(20) is similar to (7) in that it is gender-exclusive. Its reference is meant to

include males only, “someone who is a bit of the one with the ladies”.

16The author is aware of the obscene content of this utterance, but it had to be includedin the discussion, since it is one out of two cases of she in this section.

20

In (21), a certain ambiguity evolves. Is they referring to someone, some-

one’s relative, or both? Who was crushed under a lump of machinery in

this sentence? This kind of ambiguity would have been the same with a

masculine 3PP, however, and happens from time to time with anaphoric

pronominal reference, especially in speech (Wales 1996). Here we have to

use “our knowledge of the world” (Wales 1996:22) to make the assumption

that it is probably the someone who was crushed in (21), since the relative

would probably not be on the phone if she herself had been the one crushed.

(22) and (23) are the two cases of she found in this part of the data.

In (22), the reference of the 3PP is clearly specific, pointing backwards to

“the person who was nominated and then resigned”. The specific reference

triggers a gender-specific 3PP here, and is similar to (17) in Section 3.6.

In (23), the indefinite compound someone is the antecedent of the 3PP

her (the 3PP has no further anaphoric reference). But here, the reference

of the 3PP is probably meant to include females only, since she does not

have the same strong tradition of generic reference as he has. However, the

possibility that this example of she is generic and gender-inclusive cannot be

completely eliminated. It could also be argued that the 3PP here has specific

reference, that the speaker knows who put the “thing” in the locker, and

that it is a female, which triggers the feminine 3PP (similarly to (22) and

(17)). Unfortunately, the conversation from which this utterance is taken is

very fragmentary, and the context does not provide any extra information

as regards the reference of this 3PP.

3.8 Distribution of 3PPs over Contextual Domains

Table 10: 3PPs with an indefinite pronoun as antecedent, distribution overdomains

Pronoun Business (67) Edu./Informative (60) Leisure (54) Publ./Inst. (46)

They 65 53 52 45He 1 7 2 1She 1 0 0 0

21

The BNC has 4 categories of contextual domains: Business, Educa-

tional/Informative, Leisure and Public/Institutional. Of these, Business and

Public/Institutional can be considered formal, whereas the other two are

informal. Table 10 shows how the co-referential 3PPs of this study were dis-

tributed over contextual domains. Not all hits were tagged for domain, but

227 out of 382 hits carried this tag.

The table shows that they is the dominating 3PP in all four categories,

and is not at all confined only to informal usage, as some contemporary

grammars claim (see Section 2.7). He occurs most frequently in the Educa-

tional/Informative domain, which suggests that he in this study was used

more frequently in informal contexts than in formal ones.

4 Concluding Remarks

This study had three aims. The first aim was to investigate the frequency of

sing. they in spoken BrE. The study shows that sing. they is undoubtedly the

most common 3PP used in co-reference with indefinite compound pronouns.

They was used in 94.76 percent of the cases, which clearly confirms (and

possibly even goes further than) my hypothesis that it should be common.

Another aim of this study was to find out if there is a difference in

frequency of sing. they over different indefinite compound pronouns as an-

tecedents. Here, my hypothesis was that it would be more frequent with

everybody or everyone as antecedent, because of their strong plural charac-

ter. The study proves this to be true, with they being the only 3PP used

with everybody, and in 98.33 percent of the cases with everyone. Interest-

ingly, it is also the only 3PP used with nobody, which might suggest that

nobody also has strong semantic plurality. The figures are slightly lower for

sing. they with anybody/anyone (around 96 percent) and somebody/someone

(between 80 and 92 percent), from which the conclusion can be drawn that

these indefinite compound pronouns are used more often with specific refer-

ence, thus triggering gender-exclusive instead of gender-inclusive 3PPs.

The third aim of this study was to investigate if sing. they mostly is

used in informal contexts, as our contemporary grammars claim. This study

22

shows that this is not the case. Sing. they is the dominating generic 3PP in

both informal and formal contexts. The domain in which this study found

most generic he (which is claimed to be the formal alternative), is actually

an informal one, namely Educational/Informative. Even though this study

might not be reliable as regards contextual domains (too few hits on generic

he), it raises doubts about the contemporary grammars’ claim that sing.

they is only an informal generic pronoun.

This study has shown that the nonsexist 3PP they, used singularly, is

the pronoun which BrE speakers prefer in co-reference with an indefinite

compound pronoun. It serves the purpose of a “gap-filler” for a singular

gender-neutral 3PP, and is on its way to rule out the earlier prescribed

pseudo-generic he.

For future research, it would be worthwhile investigating sing. they with

specific singular reference to common-gender nouns, as in the example from

The L Word, (2) in Section 1. It may also be interesting to carry out stud-

ies comparing British and American English on the use of sing. they. How

does, compared to the British, everybody in America choose their generic

pronouns?

23

References

About the British National Corpus. 2006. URL:

http://www.natcorp.ox.ac.uk/corpus/index.xml. Accessed on November

10.

Baranowski, Maciej. 2002. Current usage of the epicene pronoun in written

English. Journal of Sociolinguistics, pages 378–397.

Baron, Dennis. 1985. Grammar and Gender. Yale University Press, London.

Cameron, Deborah. 1993. Feminism and Linguistic Theory. Macmillan

Press LTD, London.

Chalker, Sylvia and Edmund Weiner. 1998. The Oxford Dictionary of En-

glish Grammar. Oxford University Press, Oxford, second edition.

Crystal, David. 1984. Who Cares About English Usage? Penguin Books,

Harmondsworth.

Greenbaum, Sidney. 1996. The Oxford English Grammar. Oxford University

Press, Oxford.

Miller, Casey and Kate Swift. 2001. The Handbook of Nonsexist Writing.

iUniverse.com, Inc, Lincoln.

Pauwels, Anne. 2003. Linguistic Sexism and Feminist Linguistic Activism. In

Janet Holmes and Miriam Meyerhoff, editors, The Handbook of Language

and Gender, chapter 24. Blackwell Publishing LTD, Oxford.

Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech, and Jan Svartvik.

1985. A comprehensive grammar of the English language. Longman, Lon-

don.

Spender, Dale. 1985. Man Made Language. Pandora Press, London, second

edition.

Talbot, Mary M. 1998. Language and Gender. Polity Press, Cambridge.

24

The British National Corpus. Further information can be retrieved at URL:

http://www.natcorp.ox.ac.uk/.

The L Word. 2006. American TV Series. Episode 10, Season 3. Air date:

March 12.

Wales, Katie. 1996. Personal pronouns in present-day English. Cambridge

University Press, Cambridge.

25