Sex Roles (2012) 66:268–281

DOI 10.1007/s11199-011-0083-5

ORIGINAL ARTICLE

The Gendering of Language: A Comparison of Gender

Equality in Countries with Gendered, Natural Gender,
and Genderless Languages
Jennifer L. Prewitt-Freilino & T. Andrew Caswell &
Emmi K. Laakso

Published online: 18 October 2011

# Springer Science+Business Media, LLC 2011

Abstract Feminists have long argued that sexist language within it. Today, there are no languages, which do not
can have real world consequences for gender relations and distinguish between the genders at all, leading linguists and
the relative status of men and women, and recent research psychologists to believe that gender may be “so fundamental
suggests that grammatical gender can shape how people to social organization and social structure that linguistic means
interpret the world around them along gender lines to refer to this category are indispensable for speech
(Boroditsky 2009). Although others have theorized about communities” (Stahlberg et al. 2007, p. 163). However,
the connection between grammatical gender in language references to grammatical conventions of gender in language
and societal gender equality (Stahlberg et al. 2007), the have prompted contemporary concerns over the power of
current work tests this link empirically by examining language to shape social stereotypes about gender, and
differences in gender equality between countries with perhaps ultimately shape status distinctions between men and
gendered, natural gender, and genderless language systems. women. The feminist language critique, in particular, deems
Of the 111 countries investigated, our findings suggest that language to be overwhelmingly androcentric, putting girls
countries where gendered languages are spoken evidence and women at a disadvantage in personal and professional
less gender equality compared to countries with other relationships (Stahlberg et al. 2007), and some countries,
grammatical gender systems. Furthermore, countries where such as Norway, have actively reformed their languages to
natural gender languages are spoken demonstrate greater reflect a more genderless outlook (Gabriel and Gygax 2008).
gender equality, which may be due to the ease of creating In spite of attempts at language reform already underway,
gender symmetric revisions to instances of sexist language. numerous questions remain regarding the relationship between
the social aspects of gender and language and the potential
Keywords Grammatical gender . Gender equality . benefits of modifying languages to be more gender-neutral.
Gendering Language Although Stahlberg and colleagues (2007) have theorized
about the link between grammatical gender in language and
the relative social standing of men and women in society and
Introduction recent work highlights a link between the grammatical gender
of language and sexist attitudes (Wasserman and Weseley
It is quite likely, that as long as language has existed, the 2009), we believe more work is needed to determine the
distinction between male and female has also been present precise relationship between language conventions and
gender equality. Thus, in the current work we investigate
this relationship directly in an attempt to determine whether
J. L. Prewitt-Freilino (*) : E. K. Laakso
the grammatical gender of a language can predict societal
Rhode Island School of Design, markers of gender equality.
Providence, RI 02903, USA Recent theorizing suggests that language not only
e-mail: jprewitt@risd.edu reflects the conventions of culture and particular patterns
T. A. Caswell
of thought, but systems of language can actually shape our
University of South Florida, cognitive understanding of the world around us (Boroditsky
Tampa, FL, USA 2009; Deutscher 2010). Specifically, the gendering of
Sex Roles (2012) 66:268–281 269

language (even that which appears mundane and purely complete lack of grammatical gender distinction in the noun
grammatical, such as the use of la versus le in French) can system. In Finnish, for example, hän refers to both he and
actually impact our perceptions. For example, researchers she, and so has no gender. Genderless languages generally
have discovered that the grammatical gender of a term for an belong to the Uralic (Finnish), Turkic (Turkish), Iranian
inanimate object can influence people’s perceptions of the (Persian), Sinitic (Chinese), and Bantu (Swahili) language
masculine or feminine characteristics of that object, and this families, along with some others.
cannot be due merely to the properties of the object as the Given recent research tying gender in language to gendered
researchers used terms that were grammatically masculine in perceptions of the world (e.g., Boroditsky et al. 2003), one
one language and feminine in another (see Boroditsky et al. could infer that when language constantly calls attention to
2003; Konishi 1993). The same findings are true even when gender distinctions by discriminating between masculine and
pictures are used instead of text (Sera et al. 1994). feminine nouns and pronouns—as is the case in gendered
Furthermore, when Jakobson (1966) had participants choose languages—that individuals may be more apt to draw
voices to personify the days of the week, Russian speakers distinctions between men and women. If, in fact, language
consistently selected male or female voices to match the plays a role in how people organize their beliefs about
grammatical gender of that particular day. gender, then it stands to reason that differences in the gendered
If conventions of grammatical gender can affect people’s language systems across different cultures could play a role in
perceptions of gender in non-human objects, could it societal differences in beliefs, attitudes, and behavioral
similarly affect the real world social relations of men and practices about the role and status of men and women.
women? If so, then the extent to which a language In an empirical test of this assumption, Wasserman and
distinguishes grammatically between the masculine and Weseley (2009) assigned suburban, New York high school
feminine could have serious consequences for the social, students to read a passage in either English (a natural
economic, and political standing of women relative to men. gender language) or a gendered language (either Spanish or
Recent work highlights how grammatical gender can French) and then complete a measure of sexism. Students
increase sensitivity to the gender of a person, as relative who read the passage and completed the sexism measure in
to non-gender related questions, English speakers were English expressed less sexist attitudes compared to students
faster and more accurate than Chinese speakers in responding who read the passage and completed the sexism measure in
to gender relevant questions (Chen and Su 2011), suggesting a gendered language, lending support to the notion that the
that grammatical gender aids gender-relevant processing of grammatical gender of language can impact sexist
social information. attitudes.
Although all languages distinguish between genders, the Although it is impossible to isolate whether the grammatical
degree to which they do so varies. Grammatically, almost gender of a country’s language system serves as a causal factor
all languages can be divided into three gender-related in indicators of gender equality, we can determine whether
groups: grammatical gender languages, natural gender countries with gendered language systems are generally the
languages, and genderless languages (for overview of same countries in which men and women remain unequal in
definitions and classification of grammatical gender across their access to political and economic power. In the current
language families, see Stahlberg et al. 2007). Grammatical work, we explore the possible links between gendered
gender languages (or gendered languages) are characterized language systems and gender inequality by first reviewing
by their nouns, which are always assigned a feminine or the literature on the many ways gendering occurs in language
masculine (or sometimes neuter) gender. When said nouns (beyond conventions of grammatical gender) and how such
refer to people, they generally reflect the gender of the gendering in language has been shown to impact status
individual in question, and other dependant forms, such as relevant social decision-making and behavior. We then address
adjectives and pronouns carry the same gender markers whether the grammatical gender of a language system can help
as the nouns to which they refer. Generally, gendered to predict cross-cultural variations in gender equality over and
languages belong to the following linguistic families: Slavic above the predictive power of other cross-cultural factors by
(Russian), Germanic (German), Romance (Spanish), Indo- comparing countries with gendered, natural gender, and
Aryan (Hindi), or Semitic (Hebrew), with some exceptions. genderless language systems on archival indices of gender
English (a West Germanic language), and Northern Germanic equality.
(or Scandinavian) languages, belong to what are called
natural gender languages. While these natural gender Gender in Language
languages distinguish gender through pronouns (such as he
or she), most nouns have no grammatical marking of gender, As noted above, the use of gender specific nouns and
unlike the gendered languages. Finally, some languages, pronouns is one way of classifying gender in language. A
called genderless languages, are characterized by their lack of grammatical gender, however, does not necessarily
270 Sex Roles (2012) 66:268–281

reflect gender neutrality (Braun 2001; Engelberg 2002), and so terms and idiomatic expressions (see Hellinger and
it would be mistaken to believe that the grammatically Bußmann 2001). Address terms refer to the use of formal
genderless languages automatically lead to a more gender- structure as compared to informal structure within a
neutral society. Linguistic asymmetries and false generics in language depending on who one is speaking to, or the use
discourse cut across grammatical gender language systems, of honorary titles, such as Mr., Mrs., Miss, and Ms. in
and likely contribute to how individuals organize and interpret English. In the Javanese language, for example, a husband
gender relevant information in everyday communication addresses his wife as dik, or “younger sister,” where as he is
(Stahlberg et al. 2007). Below, we give a brief overview of called mas, or “older brother” regardless of his actual age.
several ways in which language becomes gendered and how In addition to asymmetry, the use of false generics can
such linguistic patterns convey status differentially to men and affect how gender intersects with language. A solely
women. masculine or feminine term, used generically to represent
First, lexical gender refers to whether a word is gender- both men and women, is called a false generic (Hellinger
specific (for example father, sister, grandmother), or gender and Bußmann 2001). Most false generics are masculine and
neutral (citizen, patient, individual), and gender asymmetry are used to refer to males as well as females, such as the
is created when gender is lexically marked when it does not word lakimies (literally, lawman, or lawyer) in Finnish.
need to be (Hellinger and Bußmann 2001). For example, in Another example is the general he in English, as in “When
the case of the English words steward and stewardess, a student drops a pencil, he should also pick it up.” The
stewardess becomes a specific marked term, or separate only known languages in which the generic is female, are in
from the term of steward, and also gains a secondary some Iroquois languages (Seneca and Oneida), as well as
meaning entirely. According to the Oxford English Dictionary some Australian aboriginal languages (Hellinger and
(1989), steward is taken to primarily refer to “an official who Bußmann 2001).
controls the domestic affairs of a household, supervising the
service of his master's table, directing the domestics, and Language and Gender Equality
regulating household expenditure…”; however, a stewardess
primarily refers to either “a female steward” or “a female Increasingly scholars and researchers recognize the power
attendant on a passenger aircraft who attends to the needs and that asymmetries in lexical gender, male false generics, and
comfort of the passengers.” The female counterpart of the systematic way language becomes gendered can have
steward references the male term, which becomes the neutral on social gender stereotypes and inequities in status
form, and while the secondary profession of stewardess between men and women (Schneider 2004). For example,
possesses some similarities to the roles of the steward, the feminist scholars have long decried that masculine generics
masculine counterpart implies more authority, through the use are androcentric, and make women seem invisible in
of words such as “control,” and “supervise.” Furthermore, historical and contemporary discourse (see Cameron
female counterparts for male words are often derived from 1998). With empirical research highlighting the real world
the masculine term, and are more complex, demonstrating impact that gendered language can have on people’s social
that the masculine is the generic form, as the feminine form is judgments, decisions, and behavior, many have begun to
generally only used when females are specifically involved. rally behind the idea that change in language is needed to
For example, in Norwegian, forfatter (meaning writer) is the curb social inequalities in society (Martyna 1980).
generic and masculine term, where as forfatterinne, the Although opponents of language reform argue that male
feminine form, is derived from forfatter, is considerably false generics remain mere grammatical convention, too
longer in length, and would never be used as a generic term widespread to expect change, and irrelevant to gender
for a writer, unless a female writer were specifically being inequality, empirical evidence supports what feminists have
referenced (Bull and Swan 2002). Moreover, some languages long known (see review by Martyna 1980). For example,
use compounding in nouns to create gender specific studies have shown that the male generic is in fact not
structures of non-traditional professions (Hellinger and simply a grammatical convention but that speakers actually
Bußmann 2001). A male-nurse or a female-surgeon points visualize males when the word “he” or “his” is used in its
to the need to categorize, in language, which gender generally generic form (Gastil 1990; Hyde 1984; Moulton et al.
pursues which occupation, and to mark the exceptions with 1978). Moreover, a “chairman” primes male pronouns and
marked names. The marked nature of such exceptions results is rated as more masculine than a “chairperson” (Banaji and
in specific patterns of the perceptions of social gender, or the Hardin 1996; McConnell and Fazio 1996). If men and
use of stereotypes in deciding generic pronouns for specific women interpret male generics in a gendered way, then it
occupations and roles. stands to reason that these gendered impressions could have
Gender asymmetry may also be manifested through a lasting impact on real world gender stereotyping and role
gender-related messages within a language, such as address behavior. For example, consider the implications for career
Sex Roles (2012) 66:268–281 271

choice. Masculine forms of nouns, such as are found in lexical relationship between all three grammatical gender groups
gender and gender related structures, are problematic, and gender equality. Wasserman and Weseley (2009)
specifically when it comes to occupational titles and positions demonstrate that gendered languages promote an increase
because women may have trouble identifying with the in expressed sexist attitudes. Moreover, Stahlberg and
masculine forms, and so may choose to not pursue a career colleagues (2007) have noted that languages that create
which implicitly excludes women. For example, Bem and significant gender distinctions (i.e., grammatical gender
Bem (1973) found that only 5% of female participants languages) are often thought to lead to greater sexism,
applied for a traditionally male job which used male generics while languages that do not distinguish grammatical gender
in its description, whereas 25% of women applied when it (i.e., genderless languages) may on the surface appear less
was described in a gender neutral way. Moreover, Briere and sexist. However, as they note:
Lanktree (1983) found that people rated women’s attraction
All language types… could in principle be used in
to a future career in psychology as lesser when they had read
a symmetrical and gender-fair way: In grammatical
an excerpt about ethical standards for psychologists worded
gender languages the feminine could be used
using male generics, as opposed to versions that were
consistently in referring to female persons and
worded in a gender neutral way. Thus, over and above the
the masculine in reference to males. In natural
influence of stereotypes about traditional gender roles
gender languages symmetry could be achieved,
associated with a particular occupation, the gendering of
above all, by the consistent use of sex-marking
language may influence women’s desire to seek certain
pronouns. And in genderless languages sex can be
employment opportunities.
disregarded symmetrically (Stahlberg et al. 2007;
Generics may also pose legal issues for women, when legal
p. 167).
documents (especially from the past) do not clarify whether
they pertain to all people, or explicitly to men. Such linguistic However, as we have seen from the above examples, this
conventions can even have legal consequences, as Hamilton et is rarely the case, and all grammatical groups display
al. (1992) have demonstrated the dangers of gendered gender asymmetry, as it is expressed in language through
pronouns in a legal context. In their study, students enrolled lexical structures, generic terms, social use of language, and
in introductory psychology courses at the University of gender related word structures.
Kentucky played the part of the jury in a mock murder case, Despite the fact that gender neutral conventions can be
and were asked to determine whether a woman had acted in developed for languages within all three grammatical
self-defense. As a part of the ‘case,’ 72 participants (24 in groups, this does not imply it is equally easy to address
each group) were given a definition of ‘self-defense’ with gendered grammar conventions across these groups. In fact,
either the use of he, he or she, or she, in the description. Only Stahlberg and colleagues (2007) note that grammatical
five of the participants reading the description with he were gender languages (like German) involve much more effort
willing to acknowledge self-defense, whereas sixteen did with to create a gender neutral configuration—compared to
he or she, and eleven with she. This suggests that male natural gender languages like English—because such
generic wording in legal proceedings can dramatically affect reconfigurations require changing a large number of
people’s perceptions of an individual’s guilt or innocence. personal nouns in addition to pronouns.
Given the above research demonstrating how small Furthermore, although it might appear that genderless
changes in the use of gendered language can have a languages already exhibit a gender fair grammatical style,
dramatic impact on people’s judgments, decisions, and there is evidence that gender neutral nouns and pronouns
behaviors, it is perhaps unsurprising that there have been can be interpreted with an implicit male bias (Stahlberg et
efforts on the part of “government agencies, feminist groups, al. 2007). Take, for example, research showing that
professional associations, religious organizations, educational different solutions to the use of masculine generics are not
institutions, publishing firms, and media institutions” to equally effective in natural gender languages like English.
reduce the use of male generics (Martyna 1980, p. 491). In Several studies have shown that replacing masculine
many languages, people have begun to shy away from the generics with gender-symmetrical terms, like he/she, led
use of masculine generics. to greater visualization of female actors compared to gender
neutral terms, like the singular they (Hyde 1984; Switzer
Grammatical Gender in Language and Gender Equality 1990). Hyde (1984) found that when children were asked to
write a story in response to the prompt “When a kid goes to
Although ample empirical research demonstrates that school, [he/they/he or she] often feels excited on the first
reducing the use of masculine generic pronouns promotes day,” (p. 699) only 12% and 18% wrote about female
the inclusion of women, there are no empirical studies that characters when he and they were used, respectively,
we are aware of which attempt to investigate the overall whereas 42% wrote about a female character when he or
272 Sex Roles (2012) 66:268–281

she was used. Thus, even the use of gender neutral evolution of not only language systems (which map onto
terminology (such as member of congress) may implicitly geographical shifts in population), but the development of
convey a gendered interpretation more often than gender- regionally specific cultural mores and practices that might
symmetrical terminology (such as congressman/woman), impact everyday gender relations, as well. In line with our
which would make it more difficult to convey a gender assumption, the global gender gap report confirms that
neutral interpretation in genderless languages that lack widespread variations in gender equality by geographic region
gendered pronouns. exist (Hausmann et al. 2009).
As cornerstones of culture, we believe dominant religious
Overview of Current Investigation tradition and political systems have the power to influence
gender equality at the societal level, and recent theorizing
In the current investigation, we explored the relationship supports this view (Razavi and Jenichen 2010). Razavi and
between countries’ grammatical gender language systems Jenichen (2010) argue that especially in the last 30 years
and indices of gender equality. To do this, we categorized religious fundamentalism has shaped political governance in
an extensive list of countries as having either a gendered, a way that impacts women’s rights. Given the link between
natural gender, or genderless orientation, and then used both religious fundamentalism (Hill et al. 2010) and right
grammatical gender system as a fixed grouping variable to wing authoritarianism (Sibley et al. 2007) with benevolent
compare the average level of gender equality. sexism, we expect that both religious and governance
Because the gendering of language has the power to systems should predict variation in gender equality. Moreover,
impact how we think at an individual level, systems of variation in human development, which comprises life
language have the potential to shape entire social structures expectancy, literacy, education, and standard of living
(Boroditsky 2009; Deutscher 2010). Gender equality is not (Klugman et al. 2010), may also help to explain why
a unitary construct, and as such, we were interested in countries differ with regard to gender equality, as the global
exploring the specific domains in which grammatical gender gap report also notes the importance of development
gendering of the language system predicted gender inequality. of human talent as a key factor in cross-cultural variations in
Thus, we wanted a measure of gender equality that allowed gender equality (Hausmann et al. 2009).
for exploring how men and women fair across a wide array of In the current analyses, we included these factors in our
societal indicators. The World Economic Forum’s Global model because we predict that even when controlling for other
Gender Gap Index and sub-indices (Hausmann et al. 2009) characteristics of countries that could be related to gender
highlight the gap between men and women in economic, equality (e.g., religious tradition, system of governance), we
educational, political, and health spheres in countries around believe differences in gender equality will emerge between
the world. countries with divergent grammatical gender language systems.
By its very nature, differential research (i.e., research that
tests for differences in pre-existing groups) does not allow for Hypotheses
the testing of causal relationships; therefore, we wanted to
ensure that any differences between the grammatical groups Given the previous research outline above demonstrating the
could not be completely explained by other factors that may difficulty involved in crafting gender fair communications in
relate to societal indicators of gender equality. Thus, to gendered languages, we anticipated the following:
explore whether the impact of grammatical gender remained
Hypothesis 1. Countries predominated by a gendered
even when other potential influences on gender equality were
language system should consistently evidence less
accounted for, we included several possible covariates of
gender equality across the various indexes than
gender equality in our model.
countries where natural gender or genderless languages
In brainstorming for characteristics of countries that
are spoken, even when controlling for geographic,
could be related to gender equality that were either easily
religious, political, and developmental variations that
categorizable or already reliably quantified, we arrived at
could also explain differences in gender equality
the following list of covariates: geographic location,
among countries.
religious tradition, system of government, and relative
human development. This list is not considered exhaustive Although previous work has demonstrated a link between
of the possible influences on variation in gender equality in gendered languages and sexism (Wasserman and Weseley
different countries, but was a starting place for trying to 2009), we are unaware of prior research on the impact of
account for such variance. We believe that over time, genderless versus natural gender language usage on sexism.
variations in geographic location could contribute to cultural Despite the fact that genderless languages generally avoid
shifts that impact the relative treatment of men versus women. the problem of the masculine generic, even gender neutral
For example, such variation could be an impetus for divergent terminology (e.g., the singular “they”) can still convey a
Sex Roles (2012) 66:268–281 273

male bias in the mind of the audience, and unlike natural matical gender grouping. If only official languages were
gender languages, genderless languages do not always allow listed, those languages were used to determine the
for adopting gender symmetric terminology (e.g., “he/she”). grammatical gender.
Thus, because natural gender languages allow for greater Using the dominate language or languages for each
flexibility compared to genderless languages to address the country, grammatical gender group was specified using the
problem of the male generic, we predicted: criteria put forth in Stahlberg et al. (2007) for delineating
the difference between gendered, genderless, and natural
Hypothesis 2. Countries predominated by a natural
gender languages. Stahlberg and colleagues categorize
gender language system should evidence greater
some languages and language families directly, and for
gender equality than countries with other grammatical
those languages, we used their classification. For languages
gender systems.
not classified directly by Stahlberg and colleagues (2007),
Unlike our prediction about gendered languages, which we consulted information from the series Gender across
we believe to be robust even when accounting for other Languages (Hellinger & Bußmann 2001–2003) and work
societal influences on gender equality, our predictions about by Ibrahim (1973) to determine the grammatical gender
the relative differences in gender equality between countries structure of specific languages. For the languages not
with natural gender and genderless languages are based on covered in either of these sources we conducted online
assumptions about the ease of language reform. Because searches to ascertain the grammatical gender structure of
social factors (i.e., religious tradition, political system) can the language in order to classify it.
influence the desire and ability to reform language, our If no single grammatical gender was predominant, the two
predictions about differences are much more circumspect, most predominant grammatical genders were listed (e.g.,
as we do not necessarily believe these differences could “gendered and natural gender”). A very small percentage of
emerge when controlling for social factors that influence countries could not be categorized into a specific grammatical
efforts at language reform. If indeed, natural gender gender category and comprise the “other”, either because they
languages demonstrate greater gender equality, we believe spoke a Creole or hybrid language without a determinable
such a difference would shed light on the need as well as grammatical gender, or if the language did not fit into any of
potential for effective language reform across different the groupings.
language systems. In order to validate our coding of grammatical gender
group, the second author classified a random sub-set of the
countries by grammatical gender group using the same
Method process. The second author’s judgment matched that of the
third author on sixty of the sixty-one randomly chosen
Sample countries, yielding a kappa of .96. Thus, we felt comfort-
able with the reliability of the coding procedure and used
In the current investigation, the third author categorized 134 the entirety of the third authors’ categorizations in our
countries (those represented in the 2009 Global Gender Gap analyses.
Report; Hausmann et al. 2009) based on the predominant In the current investigation, we only analyzed the
language(s) spoken as either gendered (54.5%), natural differences between countries that fell neatly into one
gender (9%), genderless (19.4%), gendered and natural gender language category (i.e., gendered, natural gender, or
gender (4.5%), gendered and genderless (4.5%), natural genderless), which left 111 countries in our sample (73
gender and genderless (5.2%), or other (3%). Each gendered, 12 natural gender, and 26 genderless; see Tables 1
country’s grammatical gender group or groups were and 2).
determined by first identifying the primary language for
each country. The languages spoken in each country were Indicators of Gender Equality
found through the Central Intelligence Agency’s (CIA)
Factbook (2010), which lists each country, the languages In order to compare countries’ level of gender equality, we
spoken, as well as the percentage of population which utilized data from the 2009 World Economic Forum’s
speaks each language, or, if percentages are not available, Global Gender Gap Report (Hausmann et al. 2009).
the official language(s). Given that many of the countries According to its authors, the Global Gender Gap Index
listed speak multiple languages, if over 70% spoke a single “benchmarks national gender gaps on economic, political,
language or multiple languages with a single type of education- and health-based criteria, and provides country
grammatical gender, those languages’ linguistic families— rankings that allow for effective comparisons across regions
determined using Lewis’s (2009) Ethnologue: Languages of and income groups” (p. 3). Rather than focusing on policy,
the World—were used to determine the appropriate gram- culture, or norms, the report attempts “to provide a snapshot
274 Sex Roles (2012) 66:268–281

Table 1 Summary of data by country

Country Lang. Group GGG Index Econ. Part. Educ. Attain. Health/Surv Pol. Emp. Continent Religion Govt HDI

Overall Mean – .68 .63 .95 .97 .16 – – – .77

Algeria Gendered .61 .47 .95 .97 .06 Africa Islam AR .73
Argentina Gendered .72 .60 .99 .98 .31 SA Christianity DR .87
Armenia Gendered .66 .67 1.00 .93 .04 Europe Christianity DR .78
Australia Natural .73 .75 1.00 .97 .19 Oceania Christianity DCM .96
Austria Gendered .70 .57 .99 .98 .27 Europe Christianity DR .95
Azerbaijan Genderless .66 .69 .97 .94 .06 Europe Islam AR .75
Bahamas Natural .72 .83 1.00 .98 .07 NA Christianity DCM .85
Bahrain Gendered .61 .48 .99 .96 .02 Asia Islam AM .87
Bangladesh Gendered .65 .46 .91 .95 .29 Asia Islam DR .55
Barbados Natural .72 .79 1.00 .98 .13 NA Christianity DCM .89
Belarus Gendered .71 .73 .99 .98 .16 Europe Christianity AR .8
Belgium Gendered .72 .65 .99 .98 .24 Europe Christianity DCM .95
Belize Gendered .66 .62 1.00 .98 .05 NA Christianity DCM .78
Bolivia Gendered .67 .59 .97 .97 .15 SA Christianity DR .7
Botswana Genderless .71 .74 1.00 .95 .13 Africa Christianity DR .65
Brazil Gendered .67 .64 1.00 .98 .06 SA Christianity DR .8
Brunei Darus. Genderless .65 .62 .99 .97 .03 Asia Islam AM .89
Bulgaria Gendered .71 .69 .99 .98 .16 Europe Christianity DR .82
Burkina Faso Gendered .61 .64 .73 .97 .10 Africa Islam DR .37
Cambodia Genderless .64 .65 .86 .98 .08 Asia Buddhism DCM .6
Cameroon Genderless .61 .55 .84 .97 .08 Africa Other AR .53
Chad Gendered .54 .65 .47 .98 .07 Africa Islam AR .39
Chile Gendered .69 .52 1.00 .98 .26 SA Christianity DR .87
China Genderless .69 .70 .98 .95 .14 Asia Other CS .78
Colombia Gendered .69 .69 1.00 .98 .10 SA Christianity DR .79
Croatia Gendered .69 .65 .99 .98 .16 Europe Christianity DR .85
Cuba Gendered .72 .60 1.00 .97 .29 NA Christianity CS .84
Czech Republic Gendered .68 .64 1.00 .98 .09 Europe Irreligion DR .89
Denmark Natural .76 .75 1.00 .97 .33 Europe Christianity DCM .95
Dominican Rep. Gendered .69 .65 1.00 .98 .12 SA Christianity DR .78
Ecuador Gendered .72 .63 1.00 .98 .28 SA Christianity DR .77
Egypt Gendered .59 .45 .90 .97 .02 Africa Islam AR .71
El Salvador Gendered .69 .58 .99 .98 .23 NA Christianity DR .74
Estonia Genderless .71 .71 1.00 .98 .16 Europe Other DR .86
Ethiopia Gendered .59 .60 .70 .97 .11 Africa Christianity DR .41
Finland Genderless .83 .75 1.00 .98 .57 Europe Christianity DR .95
France Gendered .73 .66 1.00 .98 .29 Europe Christianity DR .95
Gambia Natural .68 .74 .85 .98 .13 Africa Islam AR .5
Germany Gendered .74 .70 1.00 .98 .31 Europe Christianity DR .94
Ghana Genderless .67 .75 .89 .97 .07 Africa Christianity DR .55
Greece Gendered .67 .61 .99 .98 .09 Europe Christianity DR .93
Guatemala Gendered .62 .51 .94 .98 .06 NA Christianity DR .69
Guyana Natural .71 .61 1.00 .98 .25 SA Other DR .75
Honduras Gendered .69 .60 1.00 .98 .17 NA Christianity DR .7
Hungary Genderless .69 .67 .99 .98 .11 Europe Christianity DR .87
Iceland Natural .83 .75 1.00 .97 .59 Europe Christianity DR .97
India Gendered .62 .41 .84 .93 .27 Asia Hinduism DR .62
Indonesia Genderless .66 .57 .97 .97 .12 Asia Islam DR .73
Sex Roles (2012) 66:268–281 275

Table 1 (continued)

Country Lang. Group GGG Index Econ. Part. Educ. Attain. Health/Surv Pol. Emp. Continent Religion Govt HDI

Iran Genderless .58 .38 .96 .98 .02 Asia Islam AR .76
Ireland Gendered .76 .69 1.00 .97 .37 Europe Christianity DR .96
Israel Gendered .70 .69 1.00 .97 .15 Asia Judaism DR .93
Italy Gendered .68 .59 1.00 .97 .16 Europe Christianity DR .94
Jamaica Natural .70 .74 1.00 .97 .09 NA Christianity DCM .74
Jordan Gendered .62 .45 .99 .97 .06 Asia Islam DCM .77
Kazakhstan Gendered .70 .76 1.00 .98 .07 Asia Other AR .79
Korea Rep. Gendered .61 .52 .89 .97 .07 Asia Irreligion DR .92
Kuwait Gendered .64 .56 .98 .96 .04 Asia Islam AM .89
Kyrgyz Rep. Genderless .71 .69 .99 .98 .16 Asia Islam DR .7
Latvia Gendered .74 .75 1.00 .98 .23 Europe Irreligion DR .86
Lithuania Genderless .72 .75 .99 .98 .15 Europe Christianity DR .86
Luxembourg Gendered .69 .64 1.00 .97 .14 Europe Christianity DCM .94
Macedonia Gendered .70 .67 .99 .96 .16 Europe Christianity DR .8
Malawi Genderless .67 .69 .88 .96 .16 Africa Christianity DR .44
Malaysia Genderless .65 .57 .99 .97 .06 Asia Islam DCM .81
Maldives Gendered .65 .58 1.00 .95 .06 Asia Islam AR .74
Malta Gendered .66 .56 1.00 .97 .12 Europe Christianity DR .88
Mauritania Gendered .61 .49 .85 .98 .12 Africa Islam AR .55
Mexico Gendered .65 .51 .98 .98 .13 NA Christianity DR .83
Moldova Gendered .71 .73 .99 .98 .14 Europe Christianity AR .71
Mongolia Genderless .72 .83 1.00 .98 .08 Asia Buddhism DR .7
Morocco Gendered .59 .45 .86 .97 .10 Africa Islam DCM .65
Mozambique Genderless .72 .81 .78 .98 .30 Africa Christianity DR .38
Namibia Genderless .72 .72 .98 .97 .20 Africa Christianity DR .65
Nepal Gendered .62 .50 .82 .96 .22 Asia Hinduism DCM .53
Netherlands Gendered .75 .69 1.00 .97 .34 Europe Christianity DCM .95
Nicaragua Gendered .70 .56 1.00 .98 .26 NA Christianity DR .71
Norway Natural .82 .78 1.00 .98 .53 Europe Christianity DCM .97
Oman Gendered .59 .41 .97 .97 .02 Asia Islam AM .81
Pakistan Gendered .55 .34 .75 .95 .15 Asia Islam AR .55
Panama Gendered .70 .68 .99 .98 .15 NA Christianity DR .81
Paraguay Gendered .69 .67 1.00 .98 .10 SA Christianity DR .76
Peru Gendered .70 .64 .98 .97 .22 SA Christianity DR .77
Poland Gendered .70 .64 1.00 .98 .18 Europe Christianity DR .87
Portugal Gendered .70 .68 .99 .97 .16 Europe Christianity DR .9
Qatar Gendered .59 .40 .99 .95 .02 Asia Islam AM .88
Romania Gendered .68 .71 .99 .98 .04 Europe Christianity DR .81
Russian Fed. Gendered .70 .74 1.00 .98 .08 Asia Other DR .8
Saudi Arabia Gendered .57 .31 .97 .98 .00 Asia Islam AM .81
Senegal Gendered .64 .64 .82 .97 .14 Africa Islam DR .5
Slovakia Gendered .68 .65 1.00 .98 .11 Europe Christianity DR .86
Slovenia Gendered .70 .72 1.00 .97 .10 Europe Christianity DR .92
South Africa Genderless .77 .66 1.00 .98 .45 Africa Christianity DR .67
Spain Gendered .73 .60 .99 .97 .37 Europe Christianity DCM .95
Sri Lanka Gendered .74 .57 .99 .98 .42 Asia Buddhism DR .74
Sweden Natural .81 .79 1.00 .97 .50 Europe Christianity DCM .96
Switzerland Gendered .74 .69 .98 .98 .33 Europe Christianity DR .96
Syria Gendered .61 .46 .93 .98 .06 Asia Islam AR .59
276 Sex Roles (2012) 66:268–281

Table 1 (continued)

Country Lang. Group GGG Index Econ. Part. Educ. Attain. Health/Surv Pol. Emp. Continent Religion Govt HDI

Tanzania Genderless .68 .68 .87 .97 .20 Africa Other DR .47
Thailand Genderless .69 .72 .99 .98 .07 Asia Buddhism DCM .78
Tunisia Gendered .62 .45 .96 .97 .11 Africa Islam AR .77
Turkey Genderless .58 .40 .89 .97 .07 Asia Islam DR .78
Ukraine Gendered .69 .72 1.00 .98 .06 Europe Christianity DR .79
United Arab Em. Gendered .62 .41 .99 .96 .11 Asia Islam AM .87
United States Natural .72 .75 1.00 .98 .14 NA Christianity DR .95
Uruguay Gendered .69 .65 1.00 .98 .14 SA Christianity DR .85
Uzbekistan Genderless .69 .77 .94 .98 .08 Asia Islam AR .7
Venezuela Gendered .68 .62 1.00 .98 .14 SA Christianity DR .79
Vietnam Genderless .68 .73 .90 .97 .12 Asia Irreligion CS .73
Yemen Gendered .46 .23 .61 .98 .02 Asia Islam AR .51
Zambia Genderless .63 .59 .87 .96 .11 Africa Christianity DR .43
Zimbabwe Natural .65 .62 .93 .95 .10 Africa Other AR .51

Note: GGG stands for Global Gender Gap and HDI stands for Human Development Index. The full names of the sub-scores for the Global Gender
Gap Index are "Economic Participation and Opportunity", "Educational Attainment", "Health and Survival", and "Political Empowerment." All
quantitative variables range from 0 to 1. Under the Continent column, NA stands for "North America" and SA stands for "South America". Under
the Government column, AR stands for "Authoritarian Republic," AM for "Absolute Monarchy," DR Stands for "Democratic Republic," DCM for
"Democratic Constitutional Monarchy," and CS for "Communist State"

of where men and women stand with regard to some from 0 to 1 (higher scores representing greater gender
fundamental outcome variables related to basic rights” (p. 3). equality). The sub-indices themselves are calculated using
The report provides scores for each country in the form of a weighted averages of individual indicators (see Hausmann et
composite rating to represent the country’s overall gender gap, al. 2009 for further details). It is important to highlight that
as well as individual scores for each of the four sub-indices these ratings represent the relative gap between men and
that are used to calculate the un-weighted average composite women in a given country, rather than the overall level for
rating (Economic participation and opportunity, Educational each indicator. Thus, countries in which the average
attainment, Political empowerment, and Health and survival), educational attainment is relatively high might still rate as
with scores for both the overall index and sub-indices ranging relatively low in gender equality if the gap between men and

Table 2 Descriptive statistics for indicators of gender equality by grammatical language group

Dependent measure Grammatical language group

Gendered M (SD) Natural M (SD) Genderless M (SD) One-way p-value df

With covariates
Overall GGG rating .67 (.004)a .72 (.011)b .70 (.008)b < 0.001 2,103
a
Economic participation .59 (.010) .71 (.025)b .68 (.019)b < 0.001 2,103
Educational attainment .95 (.007) .95 (.016) .98 (.012) .13 2,103
Health and survival .97 (.001) .97 (.003) .97 (.002) .92 2,103
Political empowerment .15 (.013) .23 (.032) .16 (.024) .06 2,103
Without covariates
Overall GGG rating .67 (.007)a .74 (.016)b .68 (.011)a < 0.001 2,108
a b
Economic participation .59 (.013) .74 (.031) .67 (.021)b < 0.001 2,108
Educational attainment .95 (.010) .98 (.025) .94 (.017) .44 2,108
Health and survival .97 (.001) .97 (.003) .97 (.002) .48 2,108
Political empowerment .15 (.014)a .26 (.034)b .15 (.02)a .02 2,108

Note. Subscript letters indicate group differences obtained by pairwise comparisons significant at the 0.05 level. GGG stands for Global Gender
Gap. Covariates included contrasts for geographic location, religious tradition, system of government, and the continuous measure of the human
development index (HDI)
Sex Roles (2012) 66:268–281 277

women’s educational attainment is large, whereas countries countries. Thus, in our final analyses our indicator of
with low average educational attainment may be represented religious tradition contrasted countries in which Islam and
as more equal, as both men and women fail to achieve Hinduism were the primary religious traditions versus
educational attainment. For the current analyses, we utilized countries where Christianity, Judaism, Buddhism, irreli-
the overall rating, as well as the individual sub-indices as gion, or other religions predominate.
outcome variables, to determine whether differences in gender To assess differences in gender equality in countries with
equality exist among countries with different grammatical varied political systems, we compared gender equality among
gender language systems. countries with various government structures (democratic
republics, democratic constitutional monarchy, absolute mon-
Covariates archy, authoritarian republic, military government, communist
state, and transitional government). We found persistent
In order to determine if any existing differences in gender differences in gender equality between countries with either
equality between countries with different grammatical democratic republics or democratic constitutional monarchies
gender language systems would persist even when other compared to other forms of government. Thus, in our final
factors that could impact gender equality were accounted model we entered a contrast comparing democratic republics
for, we entered several other factors into our final model as and constitutional monarchies versus other forms of govern-
potential covariates. Because most of the potential covariates ment as a covariate.
were categorical variables with many levels, we first conducted Finally, we utilized the 2010 Human Development Index
a series of one-way analyses of variance (ANOVAs) on the (HDI) ratings published by the United Nations as another
Global Gender Gap Index with each potential categorical factor which might reliably predict differences in gender
covariate as the grouping variable to determine if there were equality as a covariate in our final model. The United Nations
any consistent patterns in the group means that could be conceptualizes human development as “the expansion of
represented by a planned contrast in the final model. Without a people’s freedoms to live long, healthy and creative lives; to
planned contrast, representing each level of the variable would advance other goals they have reason to value; and to engage
have required an extensive number of dummy coded variables actively in shaping development equitably and sustainably on
which would have required a larger sample size to ensure a shared planet” and as an indicator of human development,
enough power in the model for our primary analyses. the HDI “combines information on life expectancy, schooling
Furthermore, some of the levels had a very small sample size and income in a simple composite measure” (Klugman et al.
(e.g., countries with a communist state as the system of 2010; p. 2–3). The HDI is a continuous measure in which
government) and thus would not have allowed for comparison. higher scores indicate greater overall development, and the
Creating the planned contrast to represent the categorical HDI demonstrated a positive relationship with all five
covariates allowed for a reduction in the number of dummy indicators of gender equality (all ps<.01 except the
coded variables in the final model, while still representing the relationship with Economic participation, which merely
variability explained by that particular factor. The contrasts approached significance, p=.06).
chosen as covariates were those which maximized the variance In order to ensure that none of our covariates interacted
in GGG ratings explained. with our language variable, we submitted the dependent
First, we explored whether geographic region predicted variables to a series of Multivariate Analyses of Covariance
gender equality by comparing differences in gender equality (MANCOVAs) with the dichotomous covariates (geographic
by continent (Asia, Africa, North America, South America, location, religious tradition, and system of government), as
Europe, and Australia). A series of one-way ANOVAs well as a regression on the HDI scores to test for
comparing the differences in means by continent suggested moderation with our primary predictor (grammatical gender
that a contrast comparing Eastern (Africa and Asia) versus language group). None of the interactions were significant,
Western cultures (North and South America, Europe, and and thus, we felt comfortable entering our covariates into
Australia) would best represent the variance based on the final model.
geographic location. Thus, in our final analyses, our indicator
of geographic location contrasted countries in Eastern versus
Western cultures. Results
Next, we once again used a series of one-way ANOVAs
to assess the differences in gender equality among countries To test the effects of language classification on gender
with various predominant religious traditions (Christianity, equality, we submitted the overall Global Gender Gap
Islam, Judaism, Buddhism, Hinduism, irreligion, or other) (GGG) ratings and four individual indices to a MANCOVA
and found that countries which primarily practiced Islam with the government contrast, religion contrast, geography
and Hinduism demonstrated less gender equality than other constrast, and HDI scores as covariance. This yielded a
278 Sex Roles (2012) 66:268–281

significant effect for language, Λ=.69, F(10,198)=4.11, language (M=.15, SE=.014), with no significant difference
p<.001, with religion, Λ=.79, F(5,99)=5.12, p<.001, and between the latter two groups. For the economic participation
HDI, Λ = .49, F(5,99) = 20.46, p < .001, emerging as rating, countries with a gendered language scored lower in
significant covariates. The political system covariate equality (M=.59, SE=.013) than countries with a natural
approached significance, Λ= .90, F(5,99) =2.16, p <.07, gender (M=.74, SE=.031) or a genderless language (M=.67,
and geography was not statistically significant, Λ=.98, F SE=.021). The difference between the latter two groups
(5,99)=.34, p>.89. approached significance, p=.06.
We next submitted each of the five dependent variables to a
separate Analysis of Covariance (ANCOVA) with the same
covariates described above. To control the Type I error rate, Discussion
the Bonferroni approach was employed. With five univariate
follow-up tests, the alpha was set at α=.01 (.05/5). These Previous research highlights how grammatical gender in
analyses yielded a significant effect of language classification language can shape people’s perceptions (e.g., Boroditsky
for the overall GGG ratings, F(2,103)=12.68, p<.001, and et al. 2003) as well as how minor changes in the gendering
the economic participation rate, F(2,103)=15.85, p<.001. of everyday language can impact an individual’s judgments,
The analyses on the remaining dependent variables did not decisions, and behavior in ways that could influence the
yield significant results, Fs<2.9, ps>.06. relative status and treatment of men and women at an
As predicted in Hypothesis 1, countries with gen- interpersonal level (e.g., Bem and Bem 1973; Hamilton et
dered language scored lower in the GGG overall rating al. 1992). The current work attempts to merge these two
(M=.67, SE=.004) than countries with a natural gender areas by exploring how grammatical gender in language
(M=.72, SE=.011) or a genderless language (M=.70, relates to gender equality at the societal level. Taken
SE=.008). The difference between the latter two groups together, the current findings suggest a relationship between
was not significant. An identical pattern was found among the the gendering of language at a macro level and society-wide
economic participation rating with countries with a gendered indicators of gender equality.
language scoring lower in equality (M=.59, SE=.01) than In the current work, we anticipated that the grammatical
countries with a natural gender (M=.71, SE=.025) or a gender language classification of the primary language spoken
genderless language (M=.68, SE=.012). within a given country had the potential to predict overall
To ensure that the covariates were not masking any levels of gender equality in that country. As predicted, it
significant effects, we ran the analyses with the covariates appears that countries that speak gendered languages evidence
removed. We submitted the five ratings of gender equality to a less gender equality than countries that speak natural gender or
one-way Multivariate Analysis of Variance (MANOVA) with genderless languages–especially in terms of gender differences
the language classification as the independent variable. This in economic participation—even when other factors that could
yielded a significant effect for language classification, Λ=.69, influence variations in gender equality (e.g., religious tradition,
F(10,198)=4.11, p<.001. system of government) are taken into account (Hypothesis 1).
We next submitted each of the five dependent variables to a Moreover, the current findings suggest that countries that
separate Analysis of Variance (ANOVA) with the more speak natural gender languages may be even more apt to
conservative alpha level (α=.01) described above. This exhibit gender equality—especially in the form of women’s
yielded a significant effect of language classification for the greater access to political empowerment—than in countries
overall GGG ratings, F(2,108)=8.58, p<.001, the economic where gendered or genderless languages are spoken (support
participation rate, F(2,108)=13.34, p<.001, and the effect for Hypothesis 2), as countries that speak genderless languages
for political empowerment approached significance, F appear to fall in between countries that speak gendered and
(2,108)=4.31, p=.016. The analyses on the remaining natural gender languages in terms of the various indicators of
dependent variables did not yield significant results, Fs<1, gender equality.
ps>.43. As Stahlberg and colleagues (2007) have noted, despite
In support of Hypothesis 2, countries with natural gender the assumption that genderless languages are gender fair or
language scored higher in the GGG overall rating (M=.74, neutral, research has shown that a seemingly gender neutral
SE = .016) than countries with a genderless (M = .68, term (e.g., “they”) can be interpreted in a gender biased way.
SE=.011) or a gendered language (M=.67, SE=.007). The For example, in their review, Stahlberg and colleagues
difference between the latter two groups was not significant. recount research investigating possible corrections for the
An identical pattern was found for the political empowerment masculine generic in Spanish, English, German, and Turkish
ratings. Countries with natural gender language scored higher that have all demonstrated that gender neutral terms can
in the political empowerment rating (M=.26, SE=.034) than continue to connote a male bias in the mind of the audience,
countries with a genderless (M=.15, SE=.023) or a gendered and in cases where a gender symmetrical version is
Sex Roles (2012) 66:268–281 279

investigated (e.g., “he or she”), the gender symmetrical Although in the current work, we have classified
version promotes greater inclusion of women (see, for languages into three distinct categories on the basis of the
example, Braun 2001; Hyde 1984; Nissen 2002; Scheele extent to which grammatical gender is evident, it is
and Gauler 1993). Therefore, genderless languages—such as important to note that there are vast differences in the
Finnish—can include seemingly gender neutral terms that in way grammatical gender manifests within each individual
fact connote a male bias (just as natural and gendered language. For instance, gendered languages differ greatly in
languages), but because they do not possess grammatical the extent to which lexical gender is tied to the semantic
gender, it is not possible to use female pronouns or nouns to categories of sex and gender (i.e., distinctions between male
“emphasize women’s presence in the world,” which could and female or masculine and feminine; Comrie 1999), and
mean “androcentricity in a genderless language may even these differences can impact the extent to which grammatical
increase the lexical, semantic and conceptual invisibility of gender influences gendered perceptions. For example,
women” (Engelberg 2002, p.128). Vigliocco et al. (2005) found that grammatical gender affects
In contrast, gendered languages are so fundamentally based meaning more when the mapping between grammatical
in gender that it is complicated to attempt to reform the gender gender and the semantic categories for sex and gender are
asymmetry that is present in pronouns, nouns, dependant more direct (e.g., in Italian) compared to when the
forms, and grammatical rules of agreement (Stahlberg et al. relationship between grammatical gender and semantic
2007), and including gender-symmetrical forms of nouns and gender are largely arbitrary (e.g., German). However, despite
their dependant forms within sentences would significantly the variations in the way grammatical gender manifests
disturb one’s ability to read them (Koniuszaniec and within gendered, natural gender, and genderless languages,
Blaszkowska 2003). Therefore, natural gender languages the fact that consistent differences between these broad
may be the most successful at promoting gender-inclusive categories emerged is perhaps even more impressive. Future
language, because unlike genderless languages they are able research should investigate the extent to which semantic
to include gender-symmetrical forms in pronouns and nouns mapping plays a role in cross-cultural language differences
(thus increasing the visibility of women), but compared to and how this could impact both efforts at reforming gendered
gendered languages they do not depend upon gendered language and the extent to which grammatical gender
structures that would limit the legibility or intelligibility of impacts everyday gender relations. Is it the case that
symmetrical revisions. countries with gendered languages where the mapping
Although the current findings signal the existence of is largely arbitrary demonstrate less gender inequality
differences in gender equality between countries with compared to countries where the mapping is consistent? The
various grammatical gender language systems, they cannot current findings suggest that perhaps it is less about whether
address questions about the process through which these the language contains grammatical gender per se, but rather the
differences emerge or if language systems play a causal ability to reducing sexist language that predicts societal
role. Fortunately, by including other possible causal factors indicators of gender equality. Given this, we would expect
in our model, we know that the pattern of variation in that whichever language form allows for creating gender
gender equality cannot be completely explained by divergent symmetry should be associated with greater gender equality.
geographic locations, religious traditions, government systems, It is also important to note that even in languages that are
or level of development alone, as differences by grammatical deemed “genderless” in our grammatical gender classifica-
gender language remained even after controlling for other tion, language conventions exist that may allow for gender
factors. Future research could explore other possible factors that inequalities. For example, in Chinese, characters are often
predict gender equality cross-culturally or other ways of comprised of a collection of radicals (which themselves can
operationally defining the variables to be included in the be stand alone characters). Research suggests that terms
model. For example, in the current study, we used continent as comprising a radical that as a standalone character
basis for defining the various geographic locations. However, represents the word “son” are generally viewed more
smaller regional differences might be more appropriate for positively than terms comprising the character for “woman”
teasing apart geographic influences on gender equality. (Cherng et al. 2009), despite the fact that both sets of terms
Whereas Europe did not stand out from all other continents as were generally judged to be positive. Thus, grammatical
having greater gender equality, five of the top ten countries gender is only one facet of how language becomes
for overall gender equality are Nordic countries (Denmark, gendered which may have wider social implications.
Finland, Iceland, Norway, and Sweden). However, in addition For many years, feminist critics have been calling for
to a shared geographic proximity, these countries (with the language reform to reduce or eliminate the use of gender
exception of Finland) also share a history of Scandinavian asymmetries, masculine generics, and other biases in
cultural and linguistic tradition. Thus, it is difficult to tease language and make language more gender fair (Martyna
apart these influences. 1980). There have been clear efforts on the part of
280 Sex Roles (2012) 66:268–281

professional organizations, publishing companies, and even Braun, F. (2001). The communication of gender in Turkish. In M.
governmental organizations in different countries and across Hellinger & H. Bußmann (Eds.), Gender across languages (Vol. 1,
pp. 283–310). Philadelphia: John Benjamins Company.
many different languages to address these concerns (Gabriel Bull, T., & Swan, T. (2002). The representation of gender in
and Gygax 2008; Martyna 1980; Stahlberg et al. 2007). Norwegian. In M. Hellinger & H. Bußmann (Eds.), Gender
Despite the notion that grammatical gender is harmless (or across languages (Vol. 2, pp. 219–249). Philadelphia: John
even poetic; see Deutscher 2010), given the differences in Benjamins Company.
Cameron, D. (1998). The feminist critique of language: A reader (2nd
gender equality that emerged in the current work, it might be ed.). New York: Routledge.
worth considering how grammatical gender impacts attempts Central Intelligence Agency. (2010). The world factbook: Languages.
at language reform, and whether the limitations of gendered Retrieved from https://www.cia.gov/library/publications/the-
and genderless languages to adequately reform sexist world-factbook/fields/2098.html.
Chen, J.-Y., & Su, J.-J. (2011). Differential sensitivity to the gender of a
language could impact real world perceptions of gender, person by English and Chinese speakers. Journal of Psycholinguistic
and ultimately the everyday lives of women and men. Research, 40, 195–203. doi:10.1007/s10936-010-9164-9.
The current investigation explored cross-cultural variations Cherng, R.-J., Chang, C.-L., & Chen, J.-Y. (2009). A new look at
in grammatical gender in language and gender equality. gender inequality in Chinese: A study of Chinese speakers’
perception of gender-based characters. Sex Roles, 61, 427–433.
Taking such a macro approach allows for the discovery doi:10.1007/s11199-009-9639-z.
of sweeping trends, but does not allow for the discovery Comrie, B. (1999). Grammatical gender systems: A linguist’s
of nuanced differences between individual countries and assessment. Journal of Psycholinguistic Research, 28, 457–466.
languages or the understanding of how languages are doi:10.1023/A:1023212225540.
Deutscher, G. (2010). Through the language glass: Why the world
communicated and social aspects of gender status are looks different in other languages. New York: Metropolitan
negotiated in everyday situations. Thus, we hope the current Books.
work will be a catalyst for future work to further explore how Engelberg, M. (2002). The communication of gender in Finnish. In M.
grammatical gender operates in everyday communication and Hellinger & H. Bußmann (Eds.), Gender across languages (Vol. 2,
pp. 109–132). Philadelphia: John Benjamins Company.
how differences in the use of grammatical gender across Gabriel, U., & Gygax, P. (2008). Can societal language amendments
languages may not only impact cognitive interpretations (as change gender representation? The case of Norway. Scandinavian
others have already noted; e.g., Deutscher 2010; Boroditsky et Journal of Psychology, 49, 451–457. doi:10.1111/j.1467-
al. 2003), but possibly the gender relations of men and 9450.2008.00650.x.
Gastil, J. (1990). Generic pronouns and sexist language: The
women. To truly understand the intersection of gender in oxymoronic character of masculine generics. Sex Roles, 23,
language, cognition, and culture, researchers will ultimately 629–643. doi:10.1007/BF00289252.
need to draw connections between large-scale cross-cultural Hamilton, M. C., Hunter, B., & Stuart-Smith, S. (1992). Jury instructions
trends, cognitive process models, and experimental research worded in the masculine generic: Can a woman claim self-defense
when “he” is threatened? In J. C. Chrisler & D. Howard (Eds.), New
on interpersonal behavior. Moreover, although language directions in feminist psychology: Practice, theory, and research
may very well play a role in gender equality and (pp. 340–347). New York: Springer.
language reform could be a fruitful avenue for improving Hausmann, R., Tyson, L.D., & Zahidi, S. (2009). The global gender gap
the status of women, it is important to remember that 2009. Geneva, Switzerland: World Economic Forum. Retrieved
from http://www.weforum.org/pdf/gendergap/report2009.pdf
linguistic modification must be accompanied by social Hellinger, M., & Bußmann, H. (2001). The linguistic representation of
and political adjustments in order to truly change women and men. In M. Hellinger & H. Bußmann (Eds.), Gender
existing asymmetries in gender. across languages (Vol. 1, pp. 1–25). Philadelphia: John Benjamins
Company.
Hellinger. M., & Bußmann, H. (Eds.). (2001–2003). Gender across
languages (Vols. 1–3) Philadelphia: John Benjamins Company.
References Hill, E. D., Terrell, H. K., Cohen, A. B., & Nagoshi, C. T. (2010). The
role of social cognition in the religious fundamentalism-prejudice
relationship. Journal for the Scientific Study of Religion, 49,
Banaji, M., & Hardin, C. (1996). Automatic stereotyping. Psychological 724–739. doi:10.1111/j.1468-5906.2010.01542.x.
Science, 7, 136–141. doi:10.1111/j.1467-9280.1996.tb00346.x. Hyde, J. S. (1984). Children's understanding of sexist language.
Bem, S. L., & Bem, D. J. (1973). Does sex-biased job advertising “aid Developmental Psychology, 20, 697–706. doi:10.1037/0012-
and abet” sex discrimination? Journal of Applied Social 1649.20.4.697.
Psychology, 3, 6–18. doi:10.1111/j.1559-1816.1973.tb01290.x. Ibrahim, M. H. (1973). Grammatical gender: Its origin and
Boroditsky, L. (2009). How does our language shape the way we development. The Hague: Mouton & Co.
think? In M. Brockman (Ed.), What's next? Dispatches on the Jakobson, R. (1966). On linguistic aspects of translation. In R. A.
future of science (pp. 116–129). New York: Vintage. Brower (Ed.), On translation (pp. 232–239). New York: Oxford
Boroditsky, L., Schmidt, L. A., & Phillips, W. (2003). Sex, syntax, University Press.
and semantics. In D. Gentner & S. Goldin-Meadow (Eds.), Klugman, J., et al. (2010). Human development report 2010 (20th
Language in mind: Advances in the Study of Language and anniversary edition): The real wealth of nations: Pathways to
Cognition (pp. 61–79). Cambridge: MIT. human development. New York: United Nations Development
Briere, J., & Lanktree, C. (1983). Sex-role related effects of sex bias in Programme. Retrieved from http://hdr.undp.org/en/media/
language. Sex Roles, 9, 625–632. doi:10.1007/BF00290069. HDR_2010_EN_Complete_reprint.pdf
Sex Roles (2012) 66:268–281 281

Konishi, T. (1993). The semantics of grammatical gender: A cross- and female scientists differ in their selection of research problems?
cultural study. Journal of Psycholinguistic Research, 22, 519–534. The problem of gender-sex bias as a paradigmatic example of
doi:10.1007/BF01068252. linguistic relativity]. Sprache & Kognition, 12, 59–72.
Koniuszaniec, G., & Blaszkowska, H. (2003). Language and gender in Schneider, D. J. (2004). The psychology of stereotyping. New York:
Polish. In M. Hellinger & H. Bußmann (Eds.), Gender across Guilford.
languages (Vol. 3, pp. 259–285). Philadelphia: John Benjamins Sera, M., Berge, C., & Del Castillo, J. (1994). Grammatical and
Company. conceptual forces in the attribution of gender by English and
Lewis, M. P. (2009). Ethnologue: Languages of the World (16th ed). Spanish speakers. Cognitive Development, 9, 261–292.
Dallas, TX.: SIL International. Retrieved from: http://www. Sibley, C. G., Wilson, M. S., & Duckitt, J. (2007). Antecedents of men's
ethnologue.com/family_index.asp. hostile and benevolent sexism: The dual roles of Social Dominance
Martyna, W. (1980). Beyond the “he/man” approach: The case for Orientation and Right-Wing Authoritarianism. Personality and
nonsexist language. Signs, 5, 482–493. Retrieved from http:// Social Psychology Bulletin, 33, 160–172. doi:10.1177/
www.jstor.org/stable/3173588. 0146167206294745.
McConnell, A. R., & Fazio, R. H. (1996). Women as men and people: Stahlberg, D., Braun, F., Irmen, L., & Sczesny, S. (2007). Representation
Effects of gender-marked language. Personality and Social Psychol- of the sexes in language. In K. Fiedler (Ed.), Social communication
ogy Bulletin, 22, 1004–1013. doi:10.1177/01461672962210003. (pp. 163–187). New York: Psychology.
Moulton, J. G., Robinson, M., & Elias, C. (1978). Sex bias in language Oxford English dictionary (2nd ed.). (1989). Oxford: Clarendon Press
use: “Neutral” pronouns that aren't. American Psychologist, 33, (Vol. 16, pp. 666).
1032–1036. doi:10.1037/0003-066X.33.11.1032 Switzer, J. Y. (1990). The impact of generic word choices: An
Nissen, U. K. (2002). Gender in Spanish: Tradition and innovation. In empirical investigation of age and sex-related differences. Sex
M. Hellinger & H. Bußmann (Eds.), Gender across languages Roles, 22, 69–82. doi:10.1007/BF00288155.
(Vol. 2, pp. 251–277). Philadelphia: John Benjamins Company. Vigliocco, G., Vinson, D. P., Paganelli, F., & Dworzynski, K. (2005).
Razavi, S., & Jenichen, A. (2010). The unhappy marriage of religion and Grammatical gender effects on cognition: Implications for
politics: Problems and pitfalls for gender equality. Third World language learning and language use. Journal of Experimental
Quarterly, 31, 833–850. doi:10.1080/01436597.2010.502700. Psychology, 104, 501–520. doi:10.1037/0096-3445.134.4.501.
Scheele, B., & Gauler, E. (1993). Wählen wissenschaftler ihre probleme Wasserman, B. D., & Weseley, A. J. (2009). ¿Qué? Quoi? Do
anders aus als wissenschaftler innen? Das genus-sexus-problem als languages with grammatical gender promote sexist attitudes? Sex
paradigmatischer fall der linguistischen relativitätsthese [Do male Roles, 61, 634–643. doi:10.1007/s11199-009-9696-3.