Växjö University

School of Humanities
English, EN3103
Supervisor: Hans Lindquist
Examiner: Ibolya Maricic
18 January 2008

Sexist language
Gender marking of occupational terms and the non-
parallel treatment of boy and girl

Sophia Magnusson
Abstract

In everyday life women are exposed to sexist language. Terms and usages that exclude or
discriminate women are referred to as sexist language. This takes into account that one
presumes that maleness is the standard, the norm, and that femaleness is the non-standard, or
the exception. The aim of this paper was to find whether gratuitous modifiers such as girl,
lady, female and woman are used more frequently than the male markings and whether girl is
used to a wider extent than boy to denote an adult. The aim includes two aspects of sexist
language. Firstly, the aspect of calling women girls and men men, called non-parallel
treatment. Secondly, the fact that it is more common for unmarked terms to refer to males
while when referring to females a marked term is needed. As primary source for the study the
Time Corpus was used, which is an online corpus containing over 100 million words and
ranges from 1923-2007. The conclusion of this essay was that the female sex is more
commonly marked and that woman/women are the most commonly used premodifiers. Gender
markings most likely apply to occupations and labels which are thought of as either typically
male or female. Furthermore, it was found that girl was used to a wider extent than boy to
denote an adult. In addition, the results presented a possible change of trends where girl
referred to a child to a larger extent in contemporary English.

Keywords: Sexist language, female marking, non-parallel treatment, boy, girl

Table of Contents

1. Introduction…………………………………………………………… 1
1.1 Aim and scope…………………………………………………… 1
2. Background…………………………………………………………… 2
2.1 Language and gender – Sexist language………………………… 2
2.2 Non-parallel treatment…………………………………………… 4
2.3 Linguistic markedness of women……………………………….. 4
3. Material……………………………………………………………… 5
4. Method………………………………………………………………… 5
4.1 Limitations………………………………………………………. 6
5. Results………………………………………………………………… 7
5.1 Marking of occupational terms………………………………….. 7
5.2 Non-parallel treatment of boy and girl………………………….. 10
5.3 Change in the use of: girl/woman boy/man and male/female…... 13
6. Discussion……………………………………………………………. 16
6.1 Gender premodifiers…………………………………………… 16
6.2 Girl vs. boy…………………………………………………….. 18
6.3 Resurrection of women………………………………………… 18
7. Conclusion…………………………………………………………… 20
References………………………………………………………………. 21
1. Introduction
Doyle (1998:149) states that terms and usages that exclude or discriminate women are
referred to as “sexist language”. This takes into account that one presumes that maleness is
the standard, the norm, and that femaleness is the non-standard, or the exception.
Holmes (2001) claims that in everyday life women are frequently victims of both
conscious and unconscious sexism and in language the prevalence of sexism is also present.
Through sexist language a culture or society can maintain sexist attitudes. Women are often
assigned subordinate status by virtue of their gender alone, and they are treated linguistically
as subordinate, regardless of their actual power or social status.
Litosseliti (2006:14) argues that since many occupational terms are traditionally male
the use of a female marker has been adopted for instance in lady or female doctor. This
diminishes a person’s prestige by drawing attention to his or her sex. Furthermore, Litosseliti
(2006:15) argues that in everyday talk girl is used in a much wider range than boy, for
example in weathergirl and weatherman, and salesgirl and salesman. Lei (2006) states that the
wider use of girl further enhances the trivialization of women. If a woman is referred to as
girl in the same context as a man is described as man, it implies, among other things, that
women are not considered to be grown ups.
The occurrence of sexist language is undoubtedly present. However, to what degree is
hard to determine. The discussion about sexist language has been prevalent for a long time.
Nevertheless, it is still a contemporary issue. There are many aspects and discussions about
sexism. However, in the following study I have chosen to focus on two aspects of sexist
language: the marking of women and the non-parallel treatment of boy and girl.

1.1 Aim and scope

This paper explores the issue of gender marking, by examining the premodifiers:
girl/lady/female/woman and its male counterparts: boy/gentleman/male/man in the company
of occupational terms. Furthermore, I will study whether there is a non-parallel treatment of
girl and boy.
The aim of the paper is to find out whether the use of gratuitous premodifiers; girl, lady,
female and woman are used more frequently than the male markings and whether girl is used
more frequently than boy to denote an adult. These two matters are both connected with sexist
language. The research questions investigated in the paper are:

1
 • Which gender premodifiers out of girl/lady/female/woman and boy/gentleman/
male/man do the following words chiefly have: student, doctor, secretary, nurse and
priest?
• Are boy and girl used to the same extent to refer to adults?
• Has there been an increase or decrease in the use of girl/woman boy/man and
male/female in the Time Corpus from the 1920s to the 2000s.
The first research question is included in the essay in order to see which premodifiers are the
most frequent and also to see which of the female or male markers is the most common one.
Doctor and priest are more often associated with male professions and would therefore
according to my hypothesis most likely have more female markers.
The second question is integrated in order to determine if girl is used in a wider sense
than boy. The final question is incorporated in the essay to find any possible patterns or
changes in the usage of girl/boy man/woman and female/male, and whether the words are
used more or less in contemporary English. My hypothesis about this is that there will be an
increase in the use of female words compared to the male ones in recent decades compared to
the beginning of the century. Furthermore I expect that the use of woman/women will have
increased and that girl(s) will have decreased.

2. Background

2.1 Language and gender – Sexist language

There is a difference between what is meant by sex and what is meant by gender. According
to Wareing (2004:76) sex refers to biological a category and gender refers to a social
category, which is associated with certain behaviour. She points out that sexist language also
uses stereotypes and most often to the disadvantage of women. Wareing further states that
“[s]exist language represents women and men unequally, as if members of one sex were
somehow less completely human and less complex and had fewer rights than members of the
other sex” (Wareing 2004:76) Moreover, Wareing discusses whether language can be sexist
against men or only against women. She claims that with the help of language one can present
men as less complex and with fewer rights than women. However, when one looks at power
distribution in society as a whole we see that men still hold most high-status posts and

2
occupations. For this reason, Wareing claims that one cannot truly claim that language can be
sexist against men.
According to Wareing (2004:77) there are two ways in which one can discuss sexist
language: firstly to what extent the English language is inherently sexist and secondly, the
extent to which some ways of using language are sexist. The first discussion of inherently
sexist language deals with symmetry or asymmetry in the vocabulary. One example of
asymmetry is in address forms where one only uses Mr for men, but Mrs, Miss and Ms for
women. (Wareing 2004:79)
According to Litosseliti (2006:9-10) one can view language both as a “fixed and closed
system”, or as system which is “dynamic, complex and subject to change”. The latter view of
a language presupposes that every time we use languages we make personal and significant
choices from the linguistic resources available. There is a view that language does not simply
reflect social reality, but is also constitutive of such reality, i.e. we are the way we are because
of the way we use language. Litosseliti (2006:19) further argues that if language use is
constitutive then it has the potential to help establish and maintain social and power relations,
values and identities and contribute to social change. Since language is always changing one
should actively influence such shifts and this awareness-raising may eventually lead to fairer
representation and empowerment of certain groups.
Pauwels (1998b:10) argues that the aim of sexist language reform is to rid sexist terms
from our language use, replacing them with non-sexist ones or with new expressions. Pauwels
also states that “[f]eminist language planning efforts have a predominantly social rather than a
linguistic motivation” (1998b:10).
Litosseliti (2006:13-4) claims that linguistic change concerning gender includes the
promotion of gender-neutral terms e.g. fire fighters. In addition, more inclusive language has
been encouraged such as words including people, persons, they and them. According to
Litosseliti raising of awareness about sexism and sexist language is still an important step
towards change. However, she believes that any effort of intervention must be accompanied
by broader changes in our discourse practice and in social and institutional structures.
Pauwels (1998a) has observed that there are many misuses of -person compounds since
it is primarily used to identify female incumbents while men continue to be referred to by
means of -man compounds. She further states that masculine compound nouns containing
-man no longer function generically. These terms are almost exclusively used about male
referents, while women tend to be referred to by female specific terms e.g. spokeswoman.

3
Pauwels argues that it is unclear whether the new -person compounds will be successful in
challenging the -man compounds in its generic function.

2.2 Non-parallel treatment

One kind of asymmetry is the way girl can be used to describe adult women while man would
almost certainly be used to refer to an adult man (Wareing 2004:78). Litosseliti (2006:15)
states that the occurrence of calling women girls and men men is an important aspect of sexist
language. This gives the impression that women are taken less seriously. This action can be
referred to as non-parallel treatment. Litosseliti further argues that the wider use of girl than
boy is one way of demeaning women in language e.g. weathergirl vs. weatherman. According
to Pauwels (1998b:52) it has been shown that in a variety of languages there is an
asymmetrical use in word pairs such as: girl-boy, lady-gentleman and man-woman where the
sense of triviality, sexuality and immaturity are present in the female word (Pauwels
1998b:52) whereas the connotation for the masculine term is either positive or non-
judgemental e.g. ladies’ man (Doyle 1998:153).

2.3 Linguistic markedness of women

Wareing (2004:80) states that the concept of semantically marked and unmarked terms is one
type of asymmetry, but a specific kind. It is quite common for unmarked terms to refer to
males while a marked term is needed when referring to females. Consequently, this can make
terms used for females appear to differ from the standard.
According to Romaine (1999:131) women are more generally marked e.g. in
lady/woman/female doctor. It is assumed that a doctor is a man; therefore a woman who is a
doctor needs to be marked in some way. Conversely, one would not normally call a doctor
who is a man a gentleman/man/male doctor. Examples of where male occupations have been
marked are male nurse or male midwife. Furthermore, we often hear the term career woman,
and even career girl, but never or rarely career man.
Romaine (1999:125-6) claims that men have been reluctant to use the term woman.
Consequently, the most common marked terms for women are lady and girl. Lady is not
simply the polite equivalent of gentleman. Lady is for instance used in circumstances where
gentleman would not be, for example we say cleaning-lady but not garbage-gentleman.

4
 Cruse (2004:254) argues that in a binary system the features which are associated
together in pairs, for instance [MALE] and [FEMALE] form an obvious binary pair.
However, one would only need one of the two components to have one of two values + or -.
Consequently, either [+/- MALE] or [+/- FEMALE] can be used. Nevertheless, the former
component is most commonly used. If one were to use [+/- MALE] female would be the
marked term. Cruse believes that one reason why this is the case is that a great many words
referring to females are formed by referring to the male counterpart and adding a
morphological mark in the form of an affix. For example prince/princess; steward/stewardess
and usher/usherette. Words which refer to male and are derived from the female word are
very rare in English, although widow/widower is one example of this practise. Another
feature of the marked nature of female is that most often the male term can also have a
generic use. For instance actors can designate a group of males and females, while actress
only can be used for female actors (Cruse 2004:254).

3. Material
The primary source of this study is the BYU Time Corpus (Time corpus). This is an online
corpus containing over 100 million words. It is based on American English and ranges from
1923 to 2007. All the figures and numbers presented in this study are drawn from the Time
Corpus only. The Time Corpus allows you to quickly and easily search a huge corpus as
found in TIME magazine. For instance, one can see how words and phrases have increased or
decreased in usage and see how words have changed meaning over time. The corpus also
allows searches for only nouns or adjectives etc. which is good in case the word one is
searching for has different meanings. Furthermore, the results in the database are presented
either by all the number of tokens for each decade or as words per one million, which is
convenient if one wants to compare a word between the decades.

4. Method
The first research question: Which gender premodifiers out of girl/lady/female/woman and
boy/gentleman/male/man do the following words chiefly have: student, doctor, secretary,
nurse and priest? was answered by investigating the premodifiers using the Time Corpus. The
words chosen for this study; student, doctor, secretary, nurse and priest, were all chosen for a
reason. The first word student was chosen because it is a rather neutral word. The other words

5
were chosen since they are more often associated with either males or females. Doctor and
priest are more often associated with male professions and would therefore according to my
hypothesis probably have more female markings. Secretary and nurse, on the other hand, are
more associated with female occupations and would consequently have more male markings.
In order to answer the second research question: Are boy and girl used to the same
extent to refer to adults? 100 concordance lines from each decade were drawn from the Time
Corpus. The concordance lines were drawn from the 1920s, 40s, 60s, 80s and the 2000s. The
100 lines of boy and the 100 lines of girl were then evaluated to find whether they referred to
adults or not. This study is a diachronic study, in that it compares any possible differences
between these time periods. The three categories used in this study are; “adult”, “child” and
“other”. In the category “adult” I included those who were over 18 (where their age could be
found). In “child” I included those under 18 and those who referred to children. The final
category, “other”, was dedicated to expressions and other unrelated subjects such as movie
titles e.g. good girl and boy scouts.
The final research question: Has there been an increase or decrease in the use of
girl/woman boy/man and male/female in the Time Corpus from the 1920s to the 2000s? was
investigated by examining the number of tokens of each word and also in its plural form. The
numbers were then presented per one million words in order to get an accurate reading of any
possible changes. My hypothesis about this was that there would be changes in the
contemporary use of the words compared to the beginning of the century. Firstly, that the use
of woman/women would have increased and that of girl(s) would have decreased. This could
then be connected with the second research question and one could draw conclusions about a
possible change of the use of girl(s). In addition, it was presupposed that the use of the female
words would have increased compared to the male counterparts.

4.1 Limitations
One limitation of the study was that when the premodifiers were investigated in the Time
Corpus the study did not investigate any changes of the premodifiers over time i.e. no changes
in the usage of the premodifiers between the decades were recorded. However, it could be
included in another study since it would have been too extensive in this paper.
One limitation of working with the Time Corpus was that occasionally the words and
concordance lines were presented twice, hence giving a higher result than it should be.
Nevertheless, this only occurred a few times and it did not have any influence on the study of

6
 concordance lines for boy and girl, since I deleted all the doublets in this investigation. Still,
this could have a slight impact on the numbers presented for the number of tokens for
girl/woman boy/man and male/female; however this will not have any major impact on the
results

5. Results

5.1 Marking of occupational terms

The results for the first research question are presented in Table 5.1. The data were collected
from the Time Corpus and all the data from the 1920s to 2000s were used for each word.
From Table 5.1, one can see how many times the premodifiers girl/lady/female
/woman and boy/gentleman/male/man occur together with the words chosen in this study:
student, doctor, secretary, nurse and priest.

Table 5.1 Number of tokens for each word together with the premodifiers girl/lady/female/woman and
boy/gentleman/male/man
Student(s) Doctor(s) Priest(s) Nurse(s) Secretary(ies) Total
Girl 52 1 0 1 7 61
Boy 7 0 0 0 0 7
Woman/women 76 72 77 6 9 240
Man/men 23 3 0 2 3 31
Female 47 8 18 12 15 100
Male 48 14 3 48 25 138
Lady 0 4 2 1 0 7
Gentleman/men 1 0 0 0 0 1
Total 254 87 100 70 59

Firstly, if one starts with the word student one can see that it is the most common word, with a
total of 254 tokens. Student was thought to be a neutral word and would therefore not have
many markings. If one looks at male (48) and female (47) the markings are equally
distributed; it is equally common to say male student(s) or female student(s). However, the
difference between girl student(s) (52) and boy student(s) (7) is greater and the same applies
for woman/women student(s) (76) and man student(s) (23). Therefore, one cannot conclude

7
that student is a gender neutral word; on the contrary it is fairly often gender marked and most
often it is a female marking.
Some examples from the concordance lines with a gender marking and student are
shown in (1) and (2).

(1) …a 16-year-old schoolboy who falls in love with a girl student. (1952)
(2) They feel that the proportion of female students will eventually rise, resulting
in a smaller number of male students and thus fewer sons of Harvard fathers.
(1975)

For the second word, doctor, the hypothesis was that the word would have more female
markings since it is more perceived as a male occupation. This hypothesis was proved right,
since if one looks at Table 5.1 one can see that between woman/women (72) and man/men (3)
there is a great difference. However, the differences between female (8) and male (14), lady
(4) and gentleman(men) (0) are not equally evident. On the contrary, these results suggest that
male doctor (14) is more often used than female doctor (8). Nevertheless the figures are fairly
weak so one cannot draw this conclusion with great reliability. In conclusion we see that
doctor more commonly has female markings. Some examples of where doctor is used with a
female premodifier are shown below.

(3) No, I've never seen a woman doctor, but I resent them anyway.(1972)
(4) She bristles whenever she hears the term lady editor, lady painter, lady doctor,
lady lawyer; they are, she will insist with quiet fury, editors or painters or
doctors or lawyers who also happen to be women. (1972)
(5) …but when the girl doctor of his dreams rushes to his bedside. (1955)

Priest is also more often perceived as a male profession and because of this it would normally
have more female markings. In Table 5.1 there are 77 instances of woman/women priest(s)
and zero of man/men priest(s). There are also differences between female (18) and male (3),
lady (2) and gentleman/men priest(s) (0). However, the difference between lady and
gentleman/men priest(s) is quite small hence one can draw the conclusion that neither of the
premodifiers are normally used. As a conclusion one can see that it is common to mark priest
if the sex of the priest is female. The most common example in the Time Corpus was

8
women/women priests, with 77 tokens in the corpus database. Most examples originated from
the 1970s (34) and most concordance lines dealt with the idea of ordaining women priests as
seen in example (6) and (7).

(6) The idea of women priests is gaining force among U.S. Episcopalians. (1974)
(7) Some priests threaten to bolt the church if women priests are permitted, while
one bishop has vowed to resign his office if they are not. (1975)

Nurse on the other hand is thought to be a female occupation therefore it would normally have
more male premodifiers. From Table 5.1 one can see that the main difference is between
female (12) and male (48) nurse, where it is four times more common to say male nurse than
female nurse. Nevertheless, according to Table 5.1 it is more common to say woman/women
nurse(s) (6) than it is to say man/men nurse(s) (2). However, since male nurse was more
frequent than female to such a large extent, one can draw the conclusion that nurse more often
has male premodifiers. From the concordance lines drawn from the corpus I draw the
conclusion that male nurse is not a derogatory expression in itself. However, it is sometimes
hinted that it might not be a suitable occupation for a male. Examples of male nurse can be
seen in (8) and (9), in both examples a male nurse is compared to being a doctor.

(8) But because he was only a male nurse and not a licensed doctor (1946)
(9) He's the kind of guy who would rather be a male nurse than a doctor. (2000)

The last word which was analyzed was secretary. This word is also more associated with
females and would therefore, naturally, have more male markings. From Table 5.1 one can
see that there is a difference between male (25) and female (15) secretary(ies). Hence, it is
more common to say a male secretary. Nonetheless, there are more tokens for both
woman/women (9) and girl (7) compared to man/men (3) and boy (0) secretary(ies). When
reading the concordance lines of male secretary the examples shown are neutral and male is
only added to show the sex of the secretary, there is no implication that a man is doing a
woman’s job.

(10) His female secretary tells his male secretary. (1933)

9
5.2 Non-parallel treatment of boy and girl
The results for the second research question are presented in Tables 5.2 - 5.3 and Figures 5.1 -
5.2. 100 concordance lines were drawn for boy and girl respectively, from the 1920s, 40s, 60s,
80s and the 2000s. This was done in order to find a difference between the use of boy and girl,
and in addition see if there is a difference between the decades. Table 5.2 and Figure 5.1
present the results for the use of girl in a diachronic study.

Table 5.2 No. of instances of girl denoting a child or an adult from a diachronic perspective.

1920s 1940s 1960s 1980s 2000s

Adult +18 69 69 71 58 33
Child 23 22 21 21 38
Other 8 9 8 21 29
Total 100 100 100 100 100

80

70

60

50
Percentage

Adult
40 Child
Other
30

20

10

0
‘ 1920s 1940s 1960s 1980s 2000s

Figure 5.1 Diachronic study of girl in the Time Corpus

10
 From Table 5.2 one can see that the first three decades, the 1920s, 40s and 60s are very
similar in the way girl denotes an adult or child. This can also be seen in Figure 5.1 where the
lines are almost straight for the first decades. In the first three decades girl denotes an adult in
about 70 per cent of the times in the Time Corpus. The final decade, the 2000s, presents the
most interesting result; here the percentage of where girl denotes a child is even higher (38
per cent) than when it is denoting an adult (33 per cent). These results would then suggest that
there has been a change in the use of girl, that girl in the contemporary decades more often
denotes a child rather than an adult. Some examples of where girl is used to denote an adult
are shown in example (11) and (12).

(11) …a tale of a man and his work, a man and his girl. (1940)
(12) " She is a great girl. " Well, had he popped the question? (1980)

In (11) one can see that girl refers to an adult since it is used together with man, and in
(12) the person is talking about if the man has asked the girl to marry him, this would indicate
that the girl is an adult.

Sentence (13) and (14) are examples of where girl indicates that it is a child, which is
evident in the words used before girl, namely baby and little.

(13) Morgan finds himself delivering a baby girl in the back seat of a car. (1980)
(14) …a plain little girl, insists on playing with Laurence (1923)

The following sentences belong to the category of “other”. In (15) the girl scouts is an
association and (16) boy meets girl is an expression which one would use no matter the age of
the people referred to.

(15) As a starter they picked the Boy and Girl Scouts (1940)
(16) Boy meets girl, girl marries boy and promises to love, honor and cherish
(2000)

11
Table 5.3 and Figure 5.2 present the results of the use of boy in the Time Corpus and whether
the word denotes a child or adult.

Table 5.3 No. of instances of boy denoting a child or an adult from a diachronic perspective.

1920s 1940s 1960s 1980s 2000s

Adult +18 18 28 13 30 7
Child 60 50 74 53 82
Other 22 22 13 17 11
Total 100 100 100 100 100

90

80

70

60
Percentage

50 Adult
Child
40
Other

30

20

10

0
1920s 1940s 1960s 1980s 2000s

Figure 5.2 Diachronic study of boy in the Time Corpus

In Table 5.3 and Figure 5.2 one can see that it has always been most common that boy
denotes a child. However, the figures for both “adult” and “other” fluctuate but there is no big
difference between the decades in whether boy denotes an “adult” or “other”. In Figure 5.2
one can see that the lines do not change drastically, but remain at practically the same level in
all the decades. Hence from this study one can draw the conclusion that what boy denotes has
not changed particularly over time and that in most cases it denotes a child. However, there is
a weak tendency that boy also is used to a lesser extent to denote an adult, like girl, in the
contemporary decades.
Two examples where boy denotes an adult are shown in (17) and (18):

12
 (17) 43-year-old country boy (1980)
(18) The other was 31-year-old James Regis Fitzgibbon, a Pittsburgh boy (1940)

In both sentences boy is an indication of where the person comes from and the sex of the
person, rather than conveying that he is a child.
One example of where boy denotes a child is seen in (19) where boy refers to the fact
that he lived in America as young, hence as a child.

(19) As a boy he lived in America… (1923)’

The following are examples of where boy is used for something else than a child or adult,
used in the “other” group:

(20) …hero of school boy fiction (1923)

(21) …an Airedale, half brother of Laddie Boy.

(20) and (21) belonged to the “other” category where one can see that the boy in (20) refers to
a genre and (21) refers to a dog’s name.

5.3 Change in the use of: girl/woman boy/man and male/female

The results of the third research question are presented in Figures 5.3 - 5.5. One should take
into consideration that the figures are not to scale, and can as a result be somewhat
misleading. For instance in Figure 5.3 the scale’s highest point is 1000 while in Figure 5.4 the
scale goes to 2500. However, if one looks at the scales thoroughly the results can still be
discerned.
The hypothesis for the third research question was that the use of girl(s) had decreased
and woman/women had increased in the most recent decades as a result of a more feministic
approach where women are considered more equal to men. Furthermore, it was believed that
there would have been an increase in the use of the female words compared to the male
words.

13
 Figure 5.3 shows the usage of girl(s) and woman/women in the Time corpus in a
diachronic perspective.

1000

900

800

700
per million words

600
Girls(s)
500
Woman/women
400

300

200

100

0
1920s 1930s 1940s 1950s 1960s 1970s 1980s 1990s 2000s

Figure 5.3 No. of tokens of girl(s) and woman/women per million words in the Time Corpus

Firstly, if one looks at Figure 5.3 of girl(s) and woman/women one can see that
woman/women have always been used more than girl(s). Nevertheless, if we concentrate on
the curve for woman/women one can see that in the first decade, the 1920s, there was a high
number of tokens, while during the 1930s to the 1960s the curve is steady with approximately
the same number of tokens per million words. Consequently, one can ask the question why
there is such a high number in the 1920s. I tried to find an answer to this question but I found
no certain answers. One reason could be that this was the beginning of women’s liberation
and therefore the Time magazine wrote more about women at this time period. From the
1970s and onwards the number for woman/women increases.
Secondly, the curve for girl(s) shows that the number of tokens was fairly equally
distributed in the first decades but from the 1960s there is a decrease to the 1980s when there
is a slight increase. However, the number of tokens per million words has still decreased in
the contemporary decade compared to the 1920s.

14
 Figure 5.4 presents the results for boy(s) and man/men in the Time Corpus. The figures
range from the 1920s to the 2000s.

2500

2000
per million words

1500
Boy(s)
Man/men
1000

500

0
1920s 1930s 1940s 1950s 1960s 1970s 1980s 1990s 2000s

Figure 5.4 No. of tokens of boy(s) and man/men per million words in the Time Corpus

In Figure 5.4 one can see that the difference between the number of tokens per million
words in the Time Corpus of boy(s) and man/men is great. The main reason for this is in all
probability that man can also be used to denote mankind and people in general and not
necessarily just the male sex. The line for boy(s) is steady throughout the decades, with
approximately 200 to 300 tokens per million words, which would indicate that the usage and
meaning of boy has been the same. The line for man/men, on the other hand, indicates that
there has been a decrease in the use of the words. In the beginning of the century there is a
sudden drop in the number of tokens in the 1930s. These numbers seems to be deviant
compared to the decades before and after which both show a number above 2000 tokens per
million words. The numbers decreased drastically from the 1960s to the 1980s, from this
point the line has been steady in the contemporary decades. The number of tokens per million
words has gone from about 2000 in the 1920s to about half the numbers, 1000 tokens per
million words, in the 2000s. However, whether man/men are used less now to the benefit of
woman/women cannot be concluded with certainty.

15
 Figure 5.5 shows, similarly to Figure 5.3 and 5.4, the results for female/male used in the
Time Corpus from the 1920s to the 2000s.

120

100
per million words

80

Female(s)
60
Male(s)

40

20

0
1920s 1930s 1940s 1950s 1960s 1970s 1980s 1990s 2000s

Figure 5.5 No. of tokens of female(s) and male(s) per million words in the Time Corpus

Figure 5.5 presents the results for the use of male(s) and female(s) in the Time Corpus.
Both lines show similar trends, where the number of tokens for both male and female were
quite low in the beginning of the century, approximately 55. The line then rose steadily from
the 1960s and in the 1990s the number of tokens per million words for both words was about
110. Female has increased somewhat more than male in the most recent decades. This could
indicate that female has been more used either because more females were written about or
because one wanted to find a better word, compared to woman or girl. The steady increase of
both male and female also indicates that the words are used more commonly in contemporary
English language.

6. Discussion

6.1 Gender premodifiers

In the results section and in Table 5.1 one can see that one seldom uses lady and
gentlemen/man as predeterminers. It was only used once in the male sense, gentleman student,
and 7 times with lady as predeterminer. Furthermore, it can be seen that both girl and boy are

16
infrequently used as predeterminers, except for girl (52) and boy (7) student(s) which was
used more often. The reason why student often takes on the predeterminers boy or girl can be
that generally students are rather young.
The biggest difference can be seen in the use between woman/women and man/men. If
one looks at the total number of tokens by each predeterminer one can see that woman/women
have 240 tokens while man/men have 31. Conversely, female has 100 tokens while male has
124 tokens. However, the reason why male has more tokens than female can be that in the
case of male markings it sounds better to say male rather than man/men, while both
woman/women and female can be used to an almost equal extent. If one were to add up
woman/women/female it would have 340 tokens while man/men/male would have 169. Hence
one can see that female markings still are twice as common as male markings.
Furthermore, one can look at which premodifiers are used more often together with the
words chosen in the study. To begin with doctor, one can see that the most common marking
was woman/women doctor(s). Similarly, for priest the most common marking was
woman/women. The two words which in the beginning were expected to be male-marked,
nurse and secretary, both have male as the most common marking. Student’s most common
marking is woman/women however the premodifiers girl and female are also used to a large
extent, as shown in Table 5.1. On the other hand, the male counterparts are used together with
student to a greater extent than in doctor and priest. Student was included in the study because
it was thought to be neutral, nevertheless it appeared to have more female marking. However,
the male marking was more evident than in any of the other words so in that sense student
was more neutral than doctor, priest, nurse or secretary.
To sum up, we see that woman/woman and male are the most common gender
premodifiers. However, both girl and female are also fairly often used for female marking,
while male was the only common male counterpart. Nonetheless, these results are based
solely on the five words used in this study and the words were primarily connected with either
male or female occupations. Therefore, one cannot conclude with certainty that these are the
most common markings for all words. Nevertheless, the gender markings probably apply to
occupations and labels which are either thought of as typically female or male.
Romaine (1999:125-6) claimed that men have been reluctant to use woman as a marked
term and consequently lady and girl have been used instead. This has nonetheless been proved
wrong if one looks at the results from Table 5.1 and the previous discussion, which showed
that woman/women is the most common marking used for women.

17
 The first research question for this study has been answered in the above discussion:
which gender premodifiers do the following words have: student, doctor, secretary, nurse and
priest? One can clearly see that the female marking is more common and that mostly
woman/women are used. However, in the male marking male is more commonly used than
man/men.

6.2 Girl vs. boy

From the results in Section 5.2 one can see that there is an apparent difference in what girl
and boy denote. During the first three decades investigated, the 1920s, 40s and 60s, one can
see in Table 5.2 that in the case of girl, about 70 per cent of the times girl denoted an adult
and around 20 per cent a child. The following decades show a clear trend of change with an
increase of when a girl refers to a child, which even surpasses the numbers for when girl
denotes an adult. Boy on the other hand, shows no apparent trends, but fluctuates between the
decades. However, what was consistent throughout the decades was that boy always denoted a
child more often than an adult. Nevertheless, there is also a slight trend that boy, as girl,
denotes an adult to a lesser extent compared to a child in the contemporary decade.
The second research question was: Are boy and girl used to the same extent to refer to
adults? From the previous discussion and results one can clearly see that girl is used to a
wider extent than boy to denote an adult. In addition, there is an indication of a change of
trend, where girl is on the way of denoting a child and adult to either the same degree or more
often a child. Consequently, the second research question has been answered.

6.3 Resurrection of women

From the results in section 5.3 and Figure 5.3 one can see that the first word pair, girl/woman
had changed in usage in the Time Corpus. Except for the deviation in the first decade, there
has been an increase in the use of woman since the 1960s. Conversely, girl has decreased to
some extent since the 1920s. Hence, my hypothesis was confirmed that the use of
woman/women would increase and that girl(s) would decrease. If one were to connect this
with the second research question one could also here draw the conclusion that there has been
a change in the meaning of girl i.e. that girl nowadays more often denotes a child rather than
an adult. However, the decrease is not as evident as expected and therefore the meaning of
girl cannot be said to have changed altogether, with certainty.

18
 The result for boy/man shows a higher number of tokens for man/men than for boy(s)
(Figure 5.4.) The reason for the higher number is most likely that man in addition can denote
mankind and people as such, rather than just the male sex. The usage of boy(s) has remained
practically the same throughout the decades while the usage of man/men in the Time Corpus
has decreased from the 1960s and to the 2000s. This trend suggests that man/men are used
less to the benefit of other words, but whether it is to the benefit of female words cannot be
claimed with certainty.
The words female/male, in Figure 5.5, show a similar trend from the 1920s and onwards.
Both words have a low usage in the 1920s, but from the 1960s they have increased steadily.
Female(s) has a slightly higher increase in the last decade while male(s) shows the opposite
trend. The trends presented in the figure suggest that the words have become more popular to
use and this might be because the words are thought of as unbiased.
In addition, one can compare the results for woman/women and man/men. My
hypothesis was that there would be an increase in the use of female words compared to the
male ones during the decades. From Figure 5.3 one can see that the line for woman/women
starts, if one disregards the first decade, at approximately 500 tokens per million words and
ends at about 900 tokens per million words. From Figure 5.4 we see that the starting point of
the line for man/men is at about 2000 tokens per million words and ends at just above 1000
tokens per million words. Consequently, the number of tokens for both woman/women and
man/men is nearly at the same level in the contemporary use, which would suggest that they
are used to an almost equal extent.
Moreover, one can also compare the number of tokens per million words for girl(s) and
boy(s). Figure 5.3 illustrates that girl(s) started at approximately 350 and ended at about 280.
In Figure 5.4 one can see that boy(s) is at approximately the same level for the entire study, at
about 250 per million words. Hence the difference in the use of girl and boy is not so big in
the Time Corpus in the contemporary decades. If girl was used to a wider extent than boy then
one would expect a higher number of tokens for girl in the corpus throughout the decades.
One can clearly see that all the words investigated have gone through changes during the
century and the above discussion has answered the final research question: Has there been an
increase or decrease in the use of girl/woman boy/man and male/female in the Time Corpus
from the 1920s to the 2000s.

19
7 Conclusion
The aim of this paper was to find out whether the gratuitous modifiers girl, lady, female and
woman are used more frequently than the male markings and whether girl is used to a wider
extent than boy to denote an adult. The aim included two aspects of sexist language and by
using the Time Corpus the research questions stated in the introduction have been examined
and answered.
One can clearly see that the female sex is more commonly marked and that mostly
woman/women are used as a premodifier. However female and girl are also used, but to a
lesser extent. Male markings are also present but to a lesser degree. Normally, male is used,
while boy, man/men and gentlemen are rarely used. These gender markings most likely apply
to occupations and labels which are either thought of as typically female or male.
Additionally, it was clearly found that girl was used to a wider extent than boy to denote
an adult. Nevertheless, the results showed a possible change of trends where girl denoted a
child to a larger extent. The results showed that girl denotes a child and adult to either the
same degree or more often a child in the more contemporary decades, while what boy denotes
remains practically the same.
Further studies on the same subject could look into the marking of more occupational
terms in order to get more valid and better results. In addition, one could study possible
changes in the use of the different premodifiers, that is for example, if woman + profession
was used more in the previous decades and whether male + profession and female +
profession are used more in the contemporary decades.
In the end, one can say that sexist language still prevails in English, however, it is
slowly changing with an increased use of gender neutral words and generic compounds.

20
References

Primary source
Time Magazine Corpus. [online]
available at: http://corpus.byu.edu.time/
[accessed: November 15, 2007]

Secondary sources
Cruse, Alan. 2004. Meaning in language an introduction to semantics and pragmatics. New
York:Oxford University Press.
Doyle, Margaret. 1998. Introduction to the A-Z of non-sexist language. In Cameron, Deborah.
The feminist critique of language: A reader. 2nd ed. London:Routledge. 149-154.
Holmes, Janet and Sigley, Robert. 2001. What’s a word like girl doing in a place like this? In
Smith, Adam and Peters, Pam (eds.). New frontiers of corpus linguistics. Amsterdam:
Rodopi, 247-263.
Lei, Xiaolan. 2006. Sexism in language. Journal of Language and Linguistics. 5 (1): 87-95
ISSN 1475-8989.
Litosseliti, Lia. 2006. Gender and language: theory and practice. New York: Oxford
University Press.
Pauwels, Anne. 1998a. Feminist language planning: has it been worthwhile? Linguistik online
1/1998. [online].
Available at: http://www.linguistik-online.de/heft1_99/pauwels.htm
[accessed: November 20, 2007]
Pauwels, Anne. 1998b. Women changing language. New York: Addison Wesley Longman.
Wareing, Shân 2004. Language and Gender. In Sing, Ishtla and Stilwell Peccei, Jean (eds).
Language, society and power. London: Routledge. 75-92.

21