Content uploaded by Marie Gustafsson Sendén
Author content
All content in this area was uploaded by Marie Gustafsson Sendén on Nov 26, 2015
Content may be subject to copyright.
Hen can do it!
Effects of using a gender-neutral pronoun in recruitment
Emma A. Bäck, Anna Lindqvist, & Marie Gustafsson Sendén
During recent years there has been a heated debate in many media in
Sweden regarding a new gender-neutral pronoun; hen. It has been
suggested as an alternative to hon (she) and han (he). The debate has
revealed many strong opinions and feelings (Milles, 2011), where the use of
hen has been described as confusing and negative for children (Lagerwall,
2012). At the same time, hen is an important step toward gender equality
(Lagerwall, 2012; Milles, Salmson, & Tomicic, 2012). The attitudes sur-
rounding hen and its possible consequences are many and strong, but to
date there is no empirical research about the psychological consequences of
the addition of a gender-neutral pronoun in a natural gender language that
have gendered pronouns. For instance there is no research on how such an
addition could contribute to gender equality in society. The current situation
in Sweden is unique. No other country has successfully introduced a third
gender-neutral pronoun that has actually caught on in media and the popu-
lation. Hence, the main objective of the present research is to investigate
how such a pronoun may affect social cognitive thinking about gender.
Specifically we do this by investigating if gender bias in a recruitment sit-
uation may be reduced by the usage of hen, instead of using the gendered
pronouns hon or han when describing an applicant for a position.
Background
Gender is one of the primary categories in the human social world. Gender
is processed in a matter of milliseconds when meeting new people (Bennett,
Sani, Hopkins, Agostini, & Malucchi, 2000; Fiske & Taylor, 1991), and
this identification affects how subsequent information is processed, and the
conclusions drawn about the individual. The fact that gender is a primary
and active category is supposed to both contribute to a gender segregated
labor market (Eagly & Steffen, 1984) and explain why gender stereotypes
are difficult to change (Fiske & Taylor, 1991).
Language and communication have a large impact on the creation of a
common ground and reality, for instance concerning what is considered as
normal or desirable (Clark & Brennan, 1991; Hardin & Higgins, 1996).
Language is also seen as one of the most important tools for constructing
what is perceived as ’male’ or ’female’ (Stahlberg et al. 2007). For exam-
ple, a purely grammatical gender definition might affect judgments of an
object. In one study, German-speaking participants described a bridge (die
Brücke – feminine form) as ’beautiful’, ’slim’ and ’serene’, while Spanish-
speaking participants described a bridge (el Puente – masculine form) as
’great’, ’dangerous’ and ‘tall’ – that is, in agreement with stereotypes about
males and females. These effects were found even when the experiment
was conducted in English where nouns are genderless. Hence no cues of
gender, even grammatical, were present (Boroditsky, 2011; Boroditsky,
Schmidt, & Phillips, 2003; Phillips & Boroditsky, 2003). In addition, Gus-
tafsson Sendén and colleagues (in this volume) demonstrated that the
words ‘he’ and ‘she’ are used in different contexts in news media in such a
way that gender stereotypes are strengthened and maintained. For instance,
in news messages, ‘she’ is more often presented with gender labels and as
being passive, whereas ‘he’ is more active and heterogeneously described.
Thus, it is obvious that language and gender stereotypes are closely inter-
twined.
Three language groups
Most of the world’s languages can be categorized into three separate groups
depending on how much gender they contain (Prewitt-Freilino, Caswell, &
Laakso, 2012). While all languages have words for man and woman, there
are large differences between languages concerning to what extent gender
pervades words, grammar and syntax. The language group containing the
highest level of gender has gender specified nouns and gender specified
pronouns, such as French, Spanish or German. Then there is the group de-
fined as ’natural gender languages’ not containing gender specified nouns,
but gender specified pronouns, such as Swedish, English or Norwegian.
Finally, there are gender-neutral languages, which lack both gender speci-
fied nouns and gender specified pronouns, such as Indian, Turkish, or Finn-
ish (Stahlberg, Braun, Irmen, & Sczesny, 2007). Prewitt-Freilino and col-
leagues (2012) investigated if there was any relation between the level of
gender in a language and the national level of gender equality. They found
that the most gender equal countries were not the gender-neutral ones, but
rather the ‘natural gender’ countries with gender specified pronouns but not
gender specified nouns (Prewitt-Freilino et al. 2012). The Swedish lan-
guage belongs to this group. It may at first glance seem counterintuitive that
the countries with gender-neutral languages are not the most gender equal
countries, nevertheless it has been demonstrated that gender-neutral pro-
nouns, words or names often are subjected to what has been labeled a ‘male
bias’ (Lenton, Sedikides, & Bruder, 2009; Mckelvie & Waterhouse, 2005).
This means that ‘neutral’ words are in fact most often connected to mascu-
linity, and that the prototypical human being is male, unless there is explicit
evidence of the contrary (Stahlberg et al. 2007). Further, this bias has been
found in the natural gender languages as well, that is, the countries that are
most gender equal. Hence, by neutralizing a natural gender language by for
example replacing the feminine form with a generic (traditionally mascu-
line) form, one is left with a male bias (Gabriel, Gygax, Sarrasin, Garnham,
& Oakhill, 2008). To reduce the male bias, double forms are sometimes
used, i.e. she/he (Stahlberg et al. 2007). This change seems to have positive
effects on the individuals’ cognitive representations of gender; jobs that are
described using him/her seem to be more gender balanced (Romaine, 2001;
Stahlberg et al. 2007). Still, using double forms elicit representations of
mostly males (Wojahn, 2013).
Taken together, these studies clearly demonstrate that the use of only
gender-neutral pronouns (as in genderless languages) is not sufficient, nei-
ther is it sufficient to remove feminine forms and replace them with mascu-
line generic terms or use double forms. This means that we are left with a
dilemma of how to formulate texts about persons whose gender is not
known, or is not important, as when it is desired that gender does not influ-
ence the judgment of a person.
The Swedish case
The above literature review leaves a sour taste in the mouth. Even though
some reforms have indicated that it is possible to reduce gender bias, it still
seems that completely escaping may be impossible. Some scholars have
pointed to the need to be creative in this process and simply come up with
new words (see e.g. Wayne, 2004). As we began this article, this has al-
ready been made in Sweden, where a third gender-neutral pronoun has been
suggested as a complement to the previous two gendered pronouns han (he)
and hon (she). In 1994, the linguist Hans Karlgren suggested using the gen-
der-neutral pronoun hen in the language column of the Swedish newspaper
Svenska Dagbladet, but the word is actually much older, and the linguist
Rolf Dunås mentioned it in Upsala Nya Tidning already in 1966
(Björkman, 2012). However, it never caught on with the general population,
until very recently. The proposal to use hen is today a language amendment
with the explicit purpose to contribute to gender equality and to challenge
the heteronormativity trend (Milles, 2011).
To our knowledge, Swedish is the only language in the world that suc-
cessfully has introduced a complementary gender-neutral pronoun. Other
countries have tried, on several occasions, but the suggestions have never
really caught on in the general population (see e.g. Baron, 1986). Within
transgender English-speaking communities of today there is a trend to use
gender-neutral pronouns such as ze and hir (cf. Love, 2004; Schindel,
2008), yet these pronouns are not commonly known outside LGBT com-
munities although there has been explicit suggestions to use them more
broadly (Wayne, 2005).
Compared to the English language and the words ze and hir, the Swe-
dish word hen is well known among the Swedish-speaking population. The
usage of hen has expanded from transgender, feminist, linguistic, and gen-
der pedagogical communities to other parts of the society (see e.g. Jo-
sephsson, 2010), and during the last years there has been an ongoing cultur-
al debate about the existence and use of the word where a lot of people
express strong feelings in relation to hen (see e.g. Åsell, 2012; Dalén, 2012;
Lagerwall, 2012). In 2012, Språkrådet (The Language Council of Sweden,
Institute for Language and Folklore), which provides recommendation on
issues related to language, recommended that hen should not be used. How-
ever, in 2013, these guidelines were changed. The recommendation now is
that hen could be used as a gender-neutral option if it is used with care
since it may draw the recipients’ attention from the content (Språkrådet,
2013). Hen is thus breaking ground and will probably become a natural part
of the Swedish everyday vocabulary. This is also shown in how the word is
used in Swedish media. During 2012, the use of the word increased by over
1000%, compared to 2011 (Svensson, 2012). During the first 6 months of
2012 the word was mostly seen in the debate about the word itself, while
during the second half of 2012, the word was mostly used in texts, unrelat-
ed to the debate. In line with this, Ledin and Lyngfelt (2013) show that the
use of hen in blogs increased during 2012 where the most popular use was
to describe a person of unknown gender. However, the authors argue that
hen cannot be considered a natural part of the written language, since it has
not had a very large impact in newspapers. However, Milles (2013) recently
showed that the frequency of the word has increased in newspapers during
2013, but is still used relatively seldom. Ledin and Lyngfeldt conclude that
the use of hen is unevenly distributed among the general public, and is used
in different contexts.
A person described as hen could be a female, a male, or a transgendered
or inter-sexed person. However, since the phenomenon is so new there is a
lack of research investigating whether hen really is perceived as gender-
neutral or not. Our overall aim is to investigate psychological consequenc-
es, such as gender stereotyping, positive or negative attitudes, to the use of
the gender-neutral pronoun hen. We chose to do this in a recruitment situa-
tion for several reasons, which we will explain below.
The segregated labour market and its possible roots in language
The Swedish labor market is today heavily gender segregated, where only
13% of the adult population has occupations with an equal gender distribu-
tion (SCB, 2012). This is despite the laws against discrimination in Swe-
den; in relation to recruitment of new personnel, companies and organiza-
tions are obliged to actively work for achieve gender balance.
Since language is so closely connected to the transmission of stereo-
types, one way to work for a more balanced labor market could be to look
over the words used in job advertisements and descriptions of job candi-
dates. Language in recruitment advertisements can be more or less gender-
stereotypical, for instance in how a desired candidate is described. To ac-
tively work to change language in relation to recruitment could decrease
prejudiced and stereotyped (conscious or unconscious) influence of gender
in the recruitment process.
Advertisements associated to traditionally feminine jobs are more often
described with stereotypically feminine traits such as perceptiveness and
focus on relations, while jobs associated to masculinity are more often de-
scribed in more instrumental terms (Gaucher, Friesen, & Kay, 2011).
Choice of words in an ad, might thus prime (i.e. pre-activate) concepts re-
lated to male- and femaleness, which affects how men and women are
judged during a recruitment process, as well as affects to what extend men
and/or women are interested in the announced job. It has earlier been
demonstrated that both women and men are more interested in a job when
the language in the ad matches their own gender (Gaucher et al., 2011).
Thus, women are more likely to apply for a job that is described in feminine
terms, whereas men are more likely to apply for a job described in mascu-
line terms. This functions to uphold the segregated market, when ads are
termed in relation to the stereotypicality of the job. In Sweden, gender-
discrimination in recruitment is prohibited. In this, advertisements for jobs
should be presented as to decrease gender inequality within the field. It is
therefore likely that the effects found in Gaucher and colleagues’ studies
may be less pronounced in a Swedish context. Still, studies show that de-
scriptions of what competence is needed are strongly in line with what gen-
der dominates the field (Westberg, 1996).
Regarding the use of pronouns, matches between an applicant and an ad
affects judgments such that a ‘she’ (compared to a ‘he’) who applies for a
typically feminine job constitutes a better match, which in turn leads to
more positive judgments and higher probability of employment (Heilman,
1981). The overall purpose of the present research is to compare which kind
of mental representations are evoked from the use of an old and a new form
of gender-neutral person descriptions. The hypothesis is that an old form
evokes a stronger male bias than a new one, since the new is consciously
implemented to reduce the effects of gender (Milles, 2011). Thus, we aim
to establish the gender neutrality of hen as compared to a traditional gender-
neutral description.
Bases for positive or negative evaluations
There are several reasons for expecting that a person being described with
the gender-neutral pronoun hen, or a person who use the word, would be
negatively evaluated. First of all, most people seem to oppose and dislike
the word itself (Wojahn, 2013). In addition, there are factors that work
against the implementation of anything that is new. For instance, in political
psychology, a large body of research indicates that people strongly prefer
the system that they currently live in, even though it may not actually be
beneficial to themselves (Jost, Banaji, & Nosek, 2004). In addition, people
are reluctant to risk what is known and safe (Bäck, 2013; Bäck &
Lindholm, 2014; Eidelman, Pattershall, & Crandall, 2010; Kahneman &
Tversky, 1979). People prefer to keep things stable and predictable. A new
word, any word, would thus elicit some resistance. However, there is reason
to believe that a word that explicitly challenges gender roles should elicit
even more resistance. Not knowing the gender of a person leads to uncer-
tain predictions, since no prior knowledge is available to apply, which
should be experienced as discomfort. However, there are still some factors
that may lead to positive evaluations of those described with, or who use,
hen. For instance, it signals progression, compliance to norms of being
politically correct, and the taking of gender issues seriously – matters that
may be especially important in a recruitment situation.
Overview of present research
Gender bias may enter at different stages of the recruitment process. We are
interested in the stage where a professional recruiter has made written eval-
uations of the applicants. Judgments of these evaluations may be affected
by what pronoun is used. Our main goal is to investigate if hen is perceived
as gender-neutral in comparison to another gender-neutral description. The
reason for this is that many of the opponents of the word hen claim that it
does not add anything to the language, since it is already possible to use
gender-neutral descriptions if one wants to express gender in a neutral way.
By varying fictional written evaluations of candidates described as either
hen or ‘the applicant’ (den sökande in Swedish), we investigate 1) if there is
a male bias connected to ‘the applicant’, as would be expected (Stahlberg et
al. 2007), but less so to hen, and 2) if negative evaluations of the word hen
spill over to negative evaluations of a) a person being described as hen, or
b) a person using the word hen. Or, put differently: how is a person de-
scribed as hen perceived and how is someone using hen perceived? We aim
to answer these questions with two separate experimental studies, where the
second experiment was designed to cover possible caveats in the first.
Study 1
In our first experiment we chose a gender-neutral position as team leader
for an insurance sales team. This position is according to Official Swedish
statistics (SCB, 2012) gender-neutral in the sense that there is an equal
distribution of men and women employed within the branch. The reason for
choosing a gender-neutral position was to avoid possible stereotypes about
the profession affecting later judgments, that is, to isolate our independent
variable.
In order to evaluate our overall goal we formulated the following hy-
potheses:
H1: A description of a person described as hen evokes more gender-neutral
mental representations than other gender-neutral descriptions (i.e. ‘the
applicant’), such that:
a) hen will be less subjected to a male-bias than will the other gender-
neutral description, and:
b) hen will be percieved less agentic than the other gender-neutral
description (which would be subjected to a male-bias).
H2: Individuals described as hen are perceived more negatively, and are
seen as less hireable than individuals described as he/she/‘the applicant’.
H3: Individuals who use the pronoun hen are perceived more negatively,
than individuals who use he/she/‘the applicant’.
Participants and design
To get a fairly representative sample, we conducted the study at Stockholm
Central station. Participants were 80 travellers (54 women, 26 men, mean
age = 31.6, SD = 13.7), who were approached and asked to take part in a
short survey about recruitment, where their task was to judge a candidate
for a position, as well as to evaluate the recruiter who had ostensibly written
the evaluation. They were informed about anonymity and after participation
they were reimbursed with a lottery ticket.
The experiment had four conditions: the described applicant for a fic-
tional job was presented as a hen/‘the applicant’/he/she. The last two condi-
tions were mainly used as controls.
Procedure and material
Participants were first given a job description of a neutral position as a sales
manager for insurances (SCB, 2012). The description was carefully formu-
lated to avoid interpretations as directed to either a male or female.
The participants were then asked to read a description of a candidate,
supposedly written by a professional recruiter. This description was also
carefully formulated to avoid any gender references. For instance, the de-
scription stated that the applicant was 32 years old, had a bachelor’s degree
in business, mentioned previous jobs and stated that the applicant’s interests
were travelling, training and socializing. The description started with a
paragraph referring to ‘the applicant’, after this first paragraph and
throughout the rest of the text the applicant was then either referred to as
‘the applicant’, he, she, or hen. Conditions were randomly assigned to the
participants.
Following this, the participants were given a questionnaire with ques-
tions regarding their judgments. Our dependent measures were ascribed
agentic and communal traits to the applicant, hireability (Rudman & Glick,
2001), feelings and judgements of the recruiter, and remembered gender of
the applicant. The agentic traits were individualistic, self-sufficient, com-
petitive, independent, hierarchical and autonomous (Cronbach’s α = .82),
and the communal traits were cooperative, supportive, kinship-oriented and
connected (α = .73). All traits were rated on 7-point Likert scales indicating
to what extent each trait was characterizing the applicant (1 = not at all, to
7 = very much), and combined to create mean indices of agentic and com-
munal traits.
Hireability was assessed, using a scale by Rudman & Glick (2001)
where the participants were asked to indicate to what extent they:
(a) would be willing to call the applicant to an interview if they had
been recruiters;
(b) believed the applicant was appropriate for the job;
(c) would be willing to hire the applicant;
(d) considered the applicant’s qualifications as sufficient; and finally:
(e) evaluated the overall qualifications of the candidate.
All items were assessed on 7-point Likert scales where 1 = ‘not at all will-
ing’ / ‘not at all appropriate’, etc. and 7 = ‘very willing to call to inter-
view’, ‘very appropriate’, etc. Items were combined to create a mean index
of hireability (α = 0.63).
We also included a feeling thermometer, where the participants indicat-
ed their own feelings towards the applicant ranging from 0 = ‘very cold
feelings’ to 100 = ‘very warm feelings’ on a continuous scale. After this, we
assessed what the participants thought about the recruiter who had suppos-
edly written the description, with our intention to measure how people per-
ceive a person using the word hen. Participants judged the recruiter on how
professional, objective, and trustworthy they believed this person had been,
and if they had confidence in the recruiter. Again, answers were assessed on
7-point Likert scales, ranging from 1= ‘not at all professional/objective’,
etc. to 7 = ‘very professional/objective’, etc. The items were combined to a
mean index (α = 0.87) of recruiter judgments. Finally, we also had a feeling
thermometer assessing the participants feelings for the recruiter, ranging
from 0 = ‘very cold feelings’ to 100 = ‘very warm feelings’. Lastly, the
participant was asked to indicate what gender they remembered that the
applicant in the description had. They chose from he, she or hen.
Results
To test our first hypothesis, that a description of a person described as hen
would evoke more gender-neutral mental representations than other gender-
neutral descriptions (i.e. ‘the applicant’), we tested first, if participants
remembered ‘the applicant’ as being male to a larger extent than when
using hen, and second if hen was rated as less agentic than ‘the applicant’.
In order to test the first part, a chi-square test between condition
(hen/‘the applicant’) and the remembered gender (he, she, hen) was con-
ducted. The relation between these variables were significant, χ
2
(3, N = 40)
= 9.14, p = .03). Results are presented in Table 1, where we have also in-
cluded the conditions he/she, for an overview (we have not hypothesis-
tested these since it is not relevant here). As can be seen in Table 1, when
he or she was used in the description the participants correctly remembered
the applicant as a he or as a she respectively. When ‘the applicant’ was
used, most participants thought they had read about a man, even though a
few of them did not state an answer. When hen was used, most people re-
membered this person as a hen, and thus using hen evokes a more gender-
neutral representation of an individual.
Table 1. Cross-table of remembered gender in description of candidate com-
pared to condition
Perceived gender
he
she
hen
no answer
Gender in
he
17
0
0
3
description
she
0
17
0
2
hen
6
3
9
2
‘the applicant’
9
4
1
6
The second part of hypothesis 1 was that individuals described as hen
would be perceived as more gender-neutral than ‘the applicant’. Here, we
expected that ‘the applicant’ would be subjected to a male bias and thus
percieved as more agentic than hen. In order to test this, we performed an
independent samples t-test, using condition (hen/‘the applicant’) as
independent variables, and the agentic index as dependent variable, but no
significant difference was found, t = 0.02, p>.05. Thus, it could not be
concluded that hen is percieved as less agentic and more gender-neutral
than ‘the applicant’.
Our second hypothesis was that individuals described as hen would be
perceived more negatively, and as less hireable than individuals described
as he/she/‘the applicant’. In order to test this, we conducted a one-way
ANOVA, using condition (he/she/hen/‘the applicant’) as independent
variable and hireability index as dependent variable. No differences in
hireability were found between the conditions, F(3, 78) = 0.20, p > .05.
Thus, when hen was used to describe an applicant for a position, the person
was not percieved to be less hireable than when using he, she or ‘the
applicant’. We also conducted another one-way ANOVA with the feeling
thermometer as dependent variable. This analysis did not reveal any
significant differences, F(3,78) = 0.07, p > .05. Again, a person described
as hen was not percieved as less warm than a person described as a he, she
or ‘the applicant’.
Our third hypothesis was that individuals who use the pronoun hen
would be perceived more negatively than individuals who use he/she/‘the
applicant’. Thus, we expected that the recruiter, who in this particular case
was using hen, would be percieved more negatively and receive lower
ratings both on the recruiter index, and on the feeling thermometer. To test
this, we conducted two one-way ANOVAs with condition (he, she, hen,
‘the applicant’) as independent variable and the recruiter index and feeling
thermometer as the dependent variables.
In the first ANOVA, we found a significant difference in the recruiter
judgments, F(3,78) = 2.31, p = 0.08. We performed a Tukey’s post-hoc test
in order to investigate what conditions differed. The post-hoc test revealed
that when the participant had read about hen, they rated the recruiter
significantly more positive (M = 5.10, SD = 0.61) than when they had read
about a he (M = 4.20, SD = 1.25). Thus, this result contradicted our
hypothesis that a person who uses the word hen would be percieved more
negatively than a person who use he, she or ‘the applicant’.
The second ANOVA, assessing whether there were any differences in
feelings towards the recruiter did now show any significant effects, F(3,78)
= 1.79, p > .05. Thus, we did not receive support for the notion that a
person who use the word hen would be percieved more negatively than a
person who use another pronoun or a neutral word.
Discussion
Our first hypothesis was partly supported in that most participants who read
about ‘the applicant’ tended to remember this person as being a male, while
those who read about hen had more difficulties to assign hen a gender label.
However, we did not find any effects on the agentic dimension, indicating
that even though our participants percieved the applicant as a male, they did
not find this person to be more characterized by male traits, than when hen
was used. Nor did we find any differences in feelings towards the
candidate, indicating that people do not dislike a person who is described as
a hen.
Even though a significant effect was found when evaluating the user of
hen, this effect was in the opposite direction to what we had hypothesized –
a person who use hen was percieved more positively than a person who use
gendered pronouns or another neutral words to describe a person. One
possible explanation to this finding is that the usage of hen might be
percieved as progressive, and that the recruitment situation employed in the
current study might have elicited thoughts about equality and fairness. In
this particular kind of situation, hen may be percieved as a probable way to
eliminate gender biases.
One possible reason for the lack of results regarding the agentic
dimension may be that the descriptions used in the current study were very
neutral, and thus it may have been difficult for participants to judge the
candidate on gender-specific traits, since information to form such
judgment was lacking. Consequently, in order to research if this was the
case, a second study was performed.
Study 2
In Study 1, both the job description and the description of the candidate had
a neutral language and a neutral description of the announced position. This
might have accounted for the lack of effects, since it would leave partici-
pants with too little information to form their judgements on. Thus, in Study
2, we aimed at providing participants with more information in the descrip-
tion of the candidate. Study 2 was similar to Study 1, with the exception
that we “gendered” the job advertisement to match either a stereotypical
man or a stereotypical woman. Another reason for this extension was to
investigate how the context affects perceptions of a candidate, depending
on the used pronoun or neutral word defining the candidate. Hence, the ad
describing a person being suitable for the position contained either typical
feminine or masculine traits, such as a person with a passion for people
(typical ‘feminine’ trait) or a seller with a winner instinct (typical ‘mascu-
line’ trait).
We formulated the following hypothesis:
H1: An individual described as hen will evoke more gender-neutral mental
representations, compared to another gender-neutral description, which will
be influenced by the context. That is, represented as more female in a femi-
nine context and more male in a masculine context, such that:
a) ‘the applicant’ will be remembered as male to a larger extent than
hen, and:
b) hen will be rated as less communal than ’the applicant’ in the femi-
nine context and less agentic than ‘the applicant’ in the masculine con-
text.
Participants and design
The participants were again recruited from Stockholm Central train station.
This time 194 persons participated in the experiment (109 women, 76 men,
mean age: = 37.0, SD =18.8). Again, we varied the pronoun in the descrip-
tion of the candidate, such that the candidate was described as a he, she, hen
or as ‘the applicant’. In addition we varied the context of the job ad so that
it was presented in either typical masculine or in typical feminine terms.
Thus, the design was a 4 (gender in candidate description: he/she/hen/‘the
applicant’) X 2 (context gender: male/female).
Procedure and material
The procedure was similar to the procedure in Study 1, where the partici-
pants first read the job description, then the description of the candidate,
and finally were given the questionnaire assessing communal and agentic
traits. The description of the job was now phrased in either masculine or
feminine terms. We varied the following attributes of the description, where
the underlined parts refer to the female context: Applicant trait: Seller with
winner instinct/ passion for people, Abilities: coaching, relation oriented,
empathic, supportive, feed-back/ competitive climate, self-confident, au-
tonomous, tough negotiator, Market: highly competitive/ good relations,
and Goals: collective/ individual. In all else, the two texts were identical.
Since we argue that most people oppose the use of hen, which may be
one reason a person described as hen, or a person using the pronoun hen,
could be perceived negatively, we also wanted to be certain that people in
fact do oppose the use of hen. Hence, we added an item asking participants
what they thought about using hen as a gender-neutral pronoun. Answers
were made on a 7 point Likert scale from 1 = ‘not good at all’ to 7 = ‘very
good’. We have also argued that people have strong feelings about hen, and
thus we added an item asking the participants how strong their opinion
regarding the use of hen as gender-neutral pronoun was. Again answers
ranged from 1 to 7, where 1 = ’not at all strong’ to 7 = ’very strong’.
Results
First we wanted to test if people actually dislike the use of hen as a gender-
neutral pronoun. The mean for this item was 2.87 (SD = 2.17), indicating
that people in fact do not like the word hen, but the variation is quite large,
nevertheless, most people (42%) stated that they thought that the use of hen
was not good at all (=1). The participants also held fairly strong opinions
about hen, with an average rate of 4.08 (SD = 2.16). A correlation analysis
revealed no relation between attitudes to hen and how strong those attitudes
were, r = -.03, p > .05. Thus, strong feelings may be present both with the
proponents and the opponents.
The first part of our hypothesis was that hen, regardless of context,
would not be more often remembered as neither a female nor a male in
comparison to when another gender-neutral term was used. Especially, such
a term would be more influenced by the gendered context. Thus, we first
expected the applicant to be remembered as a female in the feminine con-
text and as male in the masculine context, but that hen, regardless of con-
text, would be remembered as neither a female nor a male. Second, we
expected that hen would be perceived more gender-neutral than ‘the appli-
cant’, such that hen would be rated less communal than ‘the applicant’ in
the feminine context and less agentic than ‘the applicant’ in the masculine
context. To investigate the first part of this hypothesis, we conducted two
chi-square analyses, one for the typically masculine ad and one for the typi-
cally feminine ad. In both conditions, the applicant was most often remem-
bered as being male; while hen was most often remembered as a hen in both
conditions, see Figure 1 (χ
2
(3, N = 47) = 12.28, p = .006, and χ
2
(3, N = 48)
= 20.30, p < .001, for the male and female ad, respectively). There was a
small tendency for hen to be remembered as a male in the masculine con-
text, suggesting that under certain conditions, hen may also be subjected to
a male bias. Interestingly, ‘the applicant’ in the feminine context was not
affected by the context. That is, ‘the applicant’ was still perceived as a
male, even when the context suggested that the applicant would be a fe-
male, indicating that what is considered as a gender-neutral description in
fact is subject to a strong male bias.
The second part of the hypothesis was that individuals being described
as hen would be perceived more gender-neutral than other gender-neutral
descriptions, which would be influenced by context. Thus, we expected that
‘the applicant’ in a feminine context would be perceived as more communal
than hen, and that ‘the applicant’ in a masculine context would be perceived
as more agentic than hen. To test this two t-tests were performed, one for
communal traits and one for agentic traits, using gender in the description
(hen/‘the applicant’) as grouping factor. None of these demonstrated a sig-
nificant difference, t’s < 1.23, p’s > .05. We did not receive support for the
idea that hen would be perceived more gender-neutral than ‘the applicant’,
not even when using more ‘gendered’ language in the descriptions.
Figure 1. How hen and ‘the applicant’ were remembered in the masculine con-
text (left diagram) and in the feminine context (right diagram).
To sum up, our hypothesis received partial support in that hen was remem-
bered as a hen, while ‘the applicant’ was remembered as a he in both the
masculine and the feminine context. We had hypothesized that ‘the appli-
cant’ would be remembered as female in the feminine context. This result
can be interpreted in two ways, either that the male bias is so strong that a
supposedly gender-neutral formulation is perceived as a masculine even
0
2
4
6
8
10
12
14
16
18
he she hen no answer
Hen The applicant
0
2
4
6
8
10
12
14
16
18
he she hen no answer
Hen The applicant
though contextual cues might indicate the opposite. Another interpretation
is that our manipulation was too weak. However, there is some reason to
believe that ‘the applicant’ (den sökande in Swedish) due to its historic
generic masculine form might have influenced the participants to perceive
‘the applicant’ as a man. Most Swedish nouns ending with –e and used as
generic forms are the historic masculine forms such as lärare, which is the
masculine form of ‘teacher’ that has replaced the feminine form lärarinna
as generic for all teachers despite of gender (Gabriel & Gygax, 2008). Even
if there has not been a feminine form of the Swedish word for ‘applicant’, it
is possible that the general –e ending might have activated associations to
masculinity.
Concluding discussion
Recently, there has been much debate about a third, gender-neutral Swedish
pronoun– hen. Hen could be seen as either a way to create efficiency in
language, as a complement to the gendered pronouns he and she, to be used
when gender is not known or is considered as being irrelevant. However,
people who do not feel that they fit in the traditional gender dichotomy also
use it. Hence, the word hen is not only for practical reasons, but it also ex-
plicitly challenges the existing gender dichotomy (Milles, 2011). This has
led to heated debates about its existence. One argument against the usage of
hen is that the word is not adding anything to the language, since there are
already possibilities to express oneself gender neutrally using already exist-
ing words. In two experiments, we tested if the newly introduced and heavi-
ly debated word hen truly is gender-neutral, especially in comparison to
other gender-neutral descriptions.
In general, we received partial support for this notion. In both studies,
the use of a supposedly gender-neutral word, in this case ‘the applicant’,
made the participants believe that they had read about a man. However,
when using hen, the participants could not state if they had read about a
male or female. This makes us believe that when hen is used, people get
motivated to think “outside the box”. That is, gender becomes more im-
portant, than when using other neutral words, which are subject to a tradi-
tional male-bias (Stahlberg et al. 2007). However, we did not find any re-
sults supporting that hen would be perceived as less agentic or communal,
which also indicates more gender-neutrality than when using ‘the appli-
cant.’ No evaluative differences were found, neither in the feeling ther-
mometer nor in hireability. In study 1, we found that a person who use hen
actually was perceived more positively than a person who use he. One posi-
tive interpretation of our results, or lack thereof, is that even though people
dislike the use of hen, as our results from Study 2 demonstrated, this might
not rub off on judgments about people being described as a hen. On the
contrary, in Study 1 we actually found that participants rated the user of the
word hen more positively.
Nonetheless, people do not seem to like hen, as revealed in Study 2.
Then, why did we not find any other results? At first sight, it might be ar-
gued that Sweden is a very egalitarian country. However, there is still a
strongly segregated labor market, where women hold the lower positions,
with lower wages (SCB, 2012). Thus, Sweden could not be considered as
gender equal.
Another possibility may be that the recruitment situation used in the pre-
sent studies is not the best way to test the gender-neutrality of hen. Such a
situation may activate neutrality schemas, since most people are aware that
a recruitment process should be free of sexism and gender bias. Thus, peo-
ple may have been reluctant to negative ratings in the hen-condition, and
also to ascribe gender stereotypical traits to either ‘the applicant’ or hen. It
is also important to note that even though the contexts in Study 2 were
“gendered”, the job was still for a non-gender stereotypical job. The results
might have been different if we had used, for example midwifes or engi-
neers (Gabriel & Gygax, 2008). Another problem with the present research
is that we used explicit judgments, and given that the recruitment situation
may have activated schemas related to neutrality, it could have been better
to use implicit measures. These drawbacks imply that further research is
strongly needed to investigate the psychological consequences of using hen.
Considering the currently occurring and rapid changes in the Swedish lan-
guage, there is an urgent need for more empirical research on the subject.
References
Baron, D. E. (1986) Grammar and Gender. New Haven, CT: Yale University
Press.
Bennett, M., Sani, F., Hopkins, N., Agostini, L., & Malucchi, L. (2000)
Children's Gender Categorization: An Investigation of Automatic
Processing. British Journal of Developmental Psychology, 18(1), 97-
102.
Björkman, A. Q. (2012, 03-08). ”Hen” föreslogs av språkforskare redan 1994
[”Hen” was proposed by linguist already 1994, in Swedish]. Svenska
Dagbladet. Retrieved from http://blog.svd.se/kultur/2012/03/08/”hen”-
foreslogs-av-sprakforskare-redan-1994-–-i-svd/
Boroditsky, L. (2011) How Language Shapes Thought. Scientific American,
304(2), 62-65.
Boroditsky, L., Schmidt, L. A., & Phillips, W. (2003) Sex, Syntax, and
Semantics. Language in Mind: Advances in the Study of Language and
Thought, 61-79.
Bäck, E. A. (2013) Position toward the Status Quo: Explaining Differences in
Intergroup Perceptions between Left- and Right-Wing Affiliates.
Journal of Applied Social Psychology, 43(10), 2073-2082.
Bäck, E. A., & Lindholm, T. (2014). Defending or Challenging the Status Quo:
Position Effects on Biased Intergroup Perceptions. Journal of Social
and Political Psychology, 2, 77-97.
Clark, H. H., & Brennan, S. E. (1991) Grounding in Communication.
Washington, DC, US: American Psychological Association.
Dalén, K. (2012, 02-24). Starka reaktioner efter intervjuer om ”Hen” [Strong
reactions after interviews about 'Hen', in Swedish]. Dagens Nyheter.
Retrieved from http://www.dn.se/kultur-noje/starka-reaktioner-efter-
intervjuer-om-hen/
Eagly, A. H., & Steffen, V. J. (1984) Gender Stereotypes Stem from the
Distribution of Women and Men into Social Roles. Journal of
Personality and Social Psychology, 46(4), 735-754.
Eidelman, S., Pattershall, J., & Crandall, C. S. (2010) Longer Is Better. Journal
of Experimental Social Psychology, 46, 993-998.
Fiske, S. T., & Taylor, S. E. (1991) Social Cognition. New York [etc.]:
McGraw-Hill.
Gabriel, U., Gygax, P., Sarrasin, O., Garnham, A., & Oakhill, J. (2008) Au
Pairs Are Rarely Male: Norms on the Gender Perception of Role
Names across English, French, and German. Behavior research
methods, 40(1), 206-212.
Gaucher, D., Friesen, J., & Kay, A. C. (2011) Evidence That Gendered
Wording in Job Advertisements Exists and Sustains Gender
Inequality. Journal of Personality and Social Psychology, 101(1), 109-
128.
Gustafsson Sendén, M. (2014). “He” and “She” in News Media Messages:
Pronoun Use Reflects Gender Stereotypes. Paper presented at the The
eighth Nordic conference on language and gender, Södertörn,
Stockholm, Sweden.
Gustafsson Sendén, M. (2014). Personal Pronouns in Evalutive
Communication. Stockholm University, Stockholm.
Hardin, C. D., & Higgins, T. E. (1996) Shared Reality: How Social Verification
Makes the Subjective Objective. Handbook of Motivation and
Cognition (pp. 28-84). New York, NY, US: Guilford Press.
Heilman, M. E. (1981) Sex Bias in Work Settings: The Lack of Fit Model.
Research in Organizational Behavior, 5, 269-298.
Josephsson, O. (2010, 02-15). Sensus Inför Ett Tredje Kön [the Adult
Educational Association 'Sensus' Introduces a Third Gender, in
Swedish]. Svenska Dagbladet. Retrieved from
http://www.svd.se/kultur/sprakspalt/sensus-infor-ett-tredje-
kon_4264467.svd
Jost, J. T., Banaji, M. R., & Nosek, B. A. (2004). A Decade of System
Justification Theory: Accumulated Evidence of Conscious and
Unconscious Bolstering of the Status Quo. Policital Psychology, 6,
881-919.
Kahneman, D., & Tversky, A. (1979) Prospect Theory: An Analysis of Deci-
Sion under Risk. Econometrica, 47, 263-292.
Lagerwall, K. (2012, 02-14). "Hen" gör barn förvirrade ['Hen' makes the
children confused, in Swedish]. Dagens Nyheter. Retrieved from
http://www.dn.se/nyheter/sverige/kritiker-hen-gor-barn-forvirrade/
Lagerwall, K. (2012, 02-14). På Egalia tas fokus bort från könet [at Egalia,
focus ss removed from the gender, in Swedish]. Dagens Nyheter.
Retrieved from http://www.dn.se/nyheter/sverige/pa-egalia-tas-fokus-
bort-fran-konet/
Ledin, P., & Lyngfelt, B. (2013). Olika Hen-syn. Om bruket av hen i bloggar,
tidningar, och studentuppsatser. Språk och Stil, 23, 141-174.
Lenton, A., Sedikides, C., & Bruder, M. (2009) A Latent Semantic Analysis of
Gender Stereotype-Consistency and Narrowness in American English.
Sex Roles, 60(3-4), 269-278.
Love, H. (2004) ‘The Right to Change My Mind’: New Work in Trans Studies.
Feminist Theory, 5(1), 91-100.
McKelvie, S. J., & Waterhouse, K. (2005) Impressions of People with Gender-
Ambiguous Male or Female First Names. Perceptual and Motor Skills,
101(2), 339-344.
Milles, K. (2011) Feminist Language Planning in Sweden. Current Issues in
Language Planning, 12, 21-33.
Milles, K. (2013). En öppning i en sluten ordklass? Den nya användningen av
pronomenet hen. Språk och Stil, 23, 107-140.
Milles, K., Salmson, K., & Tomicic, M. (2012, 01-20). Det behövs ett nytt ord i
svenska språket [the Swedish language needs a new word, in
Swedish]. Svenska Dagbladet. Retrieved from
http://www.svd.se/opinion/brannpunkt/det-behovs-ett-nytt-ord-i-
svenska-spraket_6784859.svd
Phillips, W., & Boroditsky, L. (2003) Can Quirks of Grammar Affect the Way
You Think? Grammatical Gender and Object Concepts. Proceedings
of the Twenty-Fifth Annual Conference of the Cognitive Science
Society, Pts 1 and 2, 928-933.
Prewitt-Freilino, J. L., Caswell, T. A., & Laakso, E. K. (2012) The Gendering
of Language: A Comparison of Gender Equality in Countries with
Gendered, Natural Gender, and Genderless Languages. Sex Roles,
66(3-4), 268-281.
Romaine, S. (2001) A Corpus-Based View of Gender in British and American
English. Gender across Languages: The Linguistic Representation of
Women and Men, Vol. 1. (pp. 153-175). Amsterdam, Netherlands:
John Benjamins Publishing Company, Amsterdam.
Rudman, L. A., & Glick, P. (2001) Prescriptive Gender Stereotypes and
Backlash toward Agentic Women. Journal of Social Issues, 57(4),
743-762.
SCB. (2012) På tal om kvinnor och män: Lathund om jämställdhet. Stockholm:
Statistiska centralbyrån, Funktionen för jämställdhets-statistik.
Schindel, J. E. (2008) Gender 101 Beyond the Binary: Gay-Straight Alliances
and Gender Activism. Sexuality Research and Social Policy, 5(2), 56-
70.
Språkrådet. (2013, 05-27). Kan man använda pronomenet Hen? Veckans
språkråd. Retrieved from http://www.sprakradet.se/16474
Stahlberg, D., Braun, F., Irmen, L., & Sczesny, S. (2007) Representation of the
Sexes in Language. In K. Fiedler (Ed.), Social Communication. (pp.
163-187). New York: Psychology Press.
Svensson, A. (2012, 02-14). Hens uppgång har nått Hovrätten [The Hen boom
has reached the Court of Appeal, in Swedish]. Språktidningen.
Retrieved from http://spraktidningen.se/blogg/hens-uppgang-har-natt-
hovratten
Wayne, L. D. (2004) Neutral Pronouns: A Modest Proposal Whose Time Has
Come. Canadian Women Studies, 24(2/3), 85-92.
Westberg, H. (1996). Kvinnor och män märks: Könsmärkning av arbete –
en dold lärandeprocess. Doktorsavhandling. Stockholm: Arbetslivsin-
stitutet.
Wojahn, D. (2013) De personliga pronomenens makt: En studie av hur prono-
men styr våra föreställningar om personer. In B. Bihl, P. Andersson &
L. Lötmarker (Eds.), Svenskans Beskrivning 32: Förhandlingar Vid
Trettioandra Sammankomsten För Svenskans Beskrivning, Karlstad
Den 13-14 Oktober 2011, S. 356-367. Karlstad, Sweden.
Åsell, E. M. (2012, 05-07). Argumenten mot ”Hen” påminner om rysk
homofobi [the arguments against 'Hen' remind of Russian
homophobia, in Swedish]. SVT Debatt. Retrieved from
http://debatt.svt.se/2012/05/05/argumenten-mot-hen-paminner-om-
rysk-homofobi/
Emma A. Bäck, Fil. Dr., är lektor i psykologi vid Lunds Universitet, där hon
främst arbetar med politisk psykologi, samt genus- och språk. I relation till
det senare är hon speciellt intresserad av det könsneutrala pronomenet hen
och vilka mentala representationer av kön som detta utlöser hos individer.
Marie Gustafsson Sendén, Fil. Dr. i psykologi, arbetar vid Stockholms Uni-
versitet, psykologiska institutionen, med inriktning på socialpsykologi och
hur attityder och värderingar sprids och vidmakthålls genom språket. Ett
särskilt intresse är hur pronomen kan användas för att mäta värderingar.
Anna Lindqvist, Fil. Dr. i psykologi, har ett stort intresse för genuspsykologi,
särskilt när det gäller att integrera experimentell psykologi med genusper-
spektiv. Hon skrev sin avhandling om hur parfymers könsklassificering
märks på flaskans etikett – inte på dess doft.
