Content uploaded by Tiago Jesse Souza Lima
Author content
All content in this area was uploaded by Tiago Jesse Souza Lima on Sep 25, 2019
Content may be subject to copyright.
RESEARCH ARTICLE
The legitimizing role of accent on discrimination against immigrants
Luana Elayne Cunha de Souza*, Cicero Roberto Pereira†,‡, Leoncio Camino‡,
Tiago Jessé Souza de Lima* & Ana Raquel Rosas Torres‡
* University of Fortaleza, Fortaleza, Brazil
†Institute of Social Sciences of the University of Lisbon, Lisbon, Portugal
‡Federal University of Paraíba, João Pessoa, Brazil
Correspondence
Luana de Souza, Postgraduate Program in
Psychology, University of Fortaleza, Brazil.
Postal Code: 60811-905.
E-mail: luana_elayne@hotmail.com
Received: 23 March 2015
Accepted: 14 April 2016
http://dx.doi.org/10.1002/ejsp.2216
Keywords: accent, prejudice,
discrimination, legitimization
Abstract
This article analyses the influence of accent on discrimination against immi-
grants by examining the hypothesis that accent leads to discrimination only
in more prejudiced individuals, merely because people speaking with a native
accent are perceived to be better qualified than those whose accent is non-
standard. In Study 1 (N= 71), we found that only prejudiced individuals
use accent to discriminate against immigrants. In Study 2 (N=124), we rep-
licated this effect and found that the influence of accent on discrimination is
mediated by the perceived quality of the accent. Study 3 (N= 105) r eplicated
the previous results even after controlling for the effect of stereotyping. These
results are the first experimental illustration of the hypothesis that accent
triggers intergroup discrimination only among prejudiced individuals be-
cause they evaluate native accents as being qualitatively better than accents
of immigrants, thereby legitimizing ingroup bias.
Non-standard accents are one of the most salient char-
acteristics of individuals from other countries who come
to live, work or study in a host country. An accent iden-
tifies and potentially stigmatizes people as not being na-
tive born (Derwing & Munro, 2009; Moyer, 2004) or
not being native speakers (Kinzler, Dupoux, & Spelke,
2007). In fact, more than 40years of research in this
field has established that a speaker’s accent constitutes
an important part of his/her social identity and conveys
a considerable amount of information that is useful in
evaluating a target (Edwards, 1999; Giles & Johnson,
1987; Gluszek & Dovidio, 2010; Lippi-Green, 1997).
This occurs because accent plays a central role in the
way individuals categorize speakers into social groups,
especially in relation to ethnic categorization (Rakić,
Steffens, & Mummendey, 2011). Indeed, previous re-
search suggests that non-standard accents are associated
with negative evaluations of the speaker (e.g., Fuertes,
Gottdiener, Martin, Gilbert, & Giles, 2012; Grondelaers,
van Hout, & Steegs, 2010; Hosoda, Stone-Romero, &
Walter, 2007; Ko, Judd, & Blair, 2006; Lev-Ari & Keysar,
2010; Lindemann, 2003, 2005), because the accent
elicits native individuals’negative attitudes towards the
social category to which the non-standard speaker be-
longs (Boyd, 2003; Bresnahan, Ohashi, Nebashi, Liu, &
Shearman, 2002; Fuertes et al., 2012; Lindemann,
2003). However, as far as we know, no research has
yet been published demonstrating the influence of ac-
cent on actual discrimination against non-standard
speakers, nor has there been any elucidation of the social
and psychological process underlying the accent effect.
Specifically, we do not know yet whether accent is
sufficient to lead to discrimination against non-standard
speakers or whether this discrimination is caused by a
combination of accent and a negative evaluation associ-
ated with the social categories that the accent reveals. In
this paper, we shed light on this question by proposing
that a target’s accent leads to discrimination against a
person only when listeners already have a prejudiced
attitude against the social groups to which the target be-
longs. This possibility suggests that the mere salience of
accent does not necessarily imply ingroup bias. Indeed,
Park and Judd (2005) reviewed the relationship be-
tween categorization and ingroup bias and argued that
although categorization under some circumstances
may lead to bias, an increase in the strength of category
boundaries is not necessarily associated with an increase
in ingroup bias (refer also to Costa-Lopes, Pereira, &
Judd, 2014). Although the question remains unre-
solved, we believe that consideration of the role played
by prejudice on the influence of accent on discrimina-
tion could help to resolve this issue. Specifically, we pro-
pose the hypothesis that accent can be innocuous in
motivating discriminatory behaviour in non-prejudiced
individuals but can have a pervasive influence on the
behaviour of highly prejudiced persons.
Nevertheless, because discrimination is discouraged
by the anti-prejudice norm (Dovidio & Gaertner,
2000), even more prejudiced individuals need to engage
in a legitimizing process in order to discriminate, and
they do so by using a seemingly unprejudiced justifica-
tion (e.g., Gaertner & Dovidio, 2005). Indeed, research
European Journal of Social Psychology 46 (2016) 609–620 Copyright © 2016 John Wiley & Sons, Ltd. 609
EJSP
on the legitimation of social inequality (refer to
Costa-Lopes, Dovidio, Pereira, & Jost, 2013 for a review)
suggests that individuals need to legitimize their dis-
criminatory behaviour because this resolves the psycho-
logical conflict between two contrasting motives: one
that promotes prejudiced attitudes and another that
promotes behaviours guided by egalitarian and fair
justice motives (e.g., Lima-Nunes, Pereira, & Correia,
2013). This possibility highlights the specific role played
by different aspects of accent in discrimination against
minority groups, which is an under-explored avenue
of research. What we propose is that more prejudiced
individuals discriminate against a target with a non-
standard accent because they use information about ac-
cents as a legitimizing mechanism for discriminating
against non-standard speakers.
The Legitimizing Role of Accent on
Discrimination
Prejudiced individuals can use information provided by
accents in many different ways to legitimize discriminat-
ing against non-standard speakers. For instance, be-
cause accent reveals target group membership,
stereotypes associated with the speaker’s group can jus-
tify ingroup bias (Allport, 1954; Tajfel, 1984; Yzerbyt,
Rocher, & Schadron, 1997). As stated by Operario and
Fiske (2001), stereotypes are consequences of the more
general process of categorization. The Stereotype Con-
tent Model (Fiske, Cuddy, Glick, & Xu, 2002) argues
that stereotypes are captured by two universal dimen-
sions (warmth and competence) and that these dimen-
sions are predicted by two variables long identified as
important in intergroup relations—status and competi-
tion. Warmth is the perception that members of the
group are tolerant, warm, good natured and sincere,
while competence is the perception that members of
the group are competent, confident, independent, com-
petitive and intelligent. The greater attribution of
warmth and the lesser attribution of competencestereo-
types to characterize immigrants can be used by individ-
uals when making decisions in critical social situations
(Fiske et al., 2002; Lee & Fiske, 2006). In fact,
warmth–competence stereotype content shows up in
situations involving language skills (e.g., Yzerbyt,
Provost, & Corneille, 2005): people with non-standard
accents (e.g., Scottish accents in UK and Chicano accents
in the USA) are perceived to be less competent but si-
multaneously friendly (Bradac, 1990; Ruscher, 2001).
This information can be pivotal when a manager needs
to make decisions about hiring applicants for a job.
In addition to the role played by accent in
stereotyping, accent provides further information about
people, legitimatizing discrimination against them. Indi-
viduals evaluate the extent to which the quality of an
accent conforms to socially more desirable standards of
speaking. Accordingly, the way a person speaks can be
used as a basis for making arbitrary evaluations and, un-
like many other forms of discrimination, is commonly
accepted and perceived as legitimate by society (Gluszek
& Dovidio, 2010). For instance, Ng (2007) argues that
accent justifies discrimination because the perceived
quality of a group’s language provides a legal justifica-
tion for preferential treatment over other linguistic
groups, ‘making it seemingly rational on the grounds
that a particular language competence is necessary for
becoming a bona fide citizen, for performing well in
the job, or for benefiting from university education’
(p. 108). In addition, Lippi-Green (1997) states that in-
dividuals feel no compunctions about language-based
discrimination because they act as if accents were a lit-
mus test for exclusion, an excuse to turn away, to refuse
to recognize the other. In this sense, an individual’sper-
ception of the quality of a target’s accent might function
as a justifiable reason for discriminating against this
target in a way that allows the individual to see
his/her behaviour as correct and legitimate.
Examples of this kind of discrimination can be found
in recent research into how speakers with non-standard
accents are disadvantaged in the workplace (Hosoda &
Stone-Romero, 2010; Huang, Frideger, & Pearce,
2013). In a survey of 26 out of 27 European Union
countries, 34% of a representative sample of respon-
dents believed that a job applicant’s way of speaking
(principally accent) would put him or her at a disadvan-
tage compared with an equally qualified, non-accented
candidate. Among managers in a position to hire, this
figure rose to 45% (European Commission, 2008).
These results can be explained by the fact that
prejudiced attitudes affect behaviour and decisions in
both conscious and unconscious ways (Devine, 1989;
Dovidio, 2001), by allowing employers to decide
whether having an accent constitutes a communicative
impairment.
According to our rationale, prejudiced individuals are
more sensitive to social categories made salient by the
accent of a speaker. Subsequently, they are likely to per-
ceive the quality of a non-standard accent to be lower
than that of a standard one and will then consider it le-
gitimate to hire a native speaker rather than an immi-
grant for a particular job. We argue, thus, that only
more prejudiced individuals will discriminate against a
speaker on the basis of his/her immigrant accent (vs. a
native one) because they are more motivated to per-
ceive the quality of an accent as a justifiable reason for
not hiring a candidate who speaks with a non-standard
accent.
In summary, in this paper, we propose that individ-
uals speaking with non-standard accents will trigger dis-
crimination against themselves in highly prejudiced
people. This means that prejudice moderates the influ-
ence of speakers’accent on discrimination against them;
that is, the effect of accent is motivated by prejudice.
Moreover, we argue that the assessment of accents
works as a legitimizing mechanism for discrimination
in that the assessment of non-standard accents as being
worse than standard accents is perceived as a legitimate
reason for discrimination against immigrants. Accord-
ingly, this implies that assessment of accents should
function as a mediator of the influence that the
L. E. C. Souza, et al.Legitimization of accent on discrimination
European Journal of Social Psychology 46 (2016) 609–620 Copyright © 2016 John Wiley & Sons, Ltd.610
perception of immigrant versus native accents has in dis-
crimination against immigrants. So if the psychological
process occurs in thisway, it means that a mediated mod-
eration exists in the sense that accent discrimination oc-
curs only in prejudiced individuals, but even prejudiced
individuals need to legitimize discrimination, so the as-
sessment ofaccent functions as amediator of the process.
Overview of Studies
Through three studies, we examined how an immigrant
versus native accent influences discrimination against
immigrants and whether the perceived quality of an ac-
cent plays a legitimizing role in this discrimination in
Portugal. These studies were conducted with a
decision-making scenario used in recruiting candidates
for a job. We chose Brazilian immigrants as our target
group for two reasons. First, it is the largest immigrant
community that is legally resident in Portugal. Second,
although Brazilian immigrants speak the same language
as native Portuguese nationals, Brazilian and European
Portuguese accents are easily distinguishable by listeners.
First, we conducted a pilot study to test whether a
candidate’s accent is perceived as a legitimate reason
for not hiring an immigrant who applied for a job. Then,
in Study 1, we tested the hypothesis that accent influ-
ences discrimination against immigrants and that this
influence is moderated by prejudice. In Study 2, we rep-
licated and extended the previous study by testing the
hypothesis that the influence of an accent in activating
discrimination in more prejudiced individuals is medi-
ated by the perceived quality of the accent. In Study 3,
we replicated the mediated-moderation effect and went
further, by demonstrating the influence of the targets’
accents in discrimination even after taking into account
stereotypes about the targets.
Pilot Study
The aim of this study was to test the hypothesis that
using information about accents to decide in favour of
an ingroup target by discriminating against an immi-
grant is perceived as more legitimate than a decision
that does not take the target’s accent into account. The
subjects were 27 female and 10 male Portuguese uni-
versity students (M
age
= 20.4, SD = 3.75), who were ran-
domly allocated to one of two conditions (decision
based on candidate’s accent vs. not based on candidate’s
accent). The participants read a text in which a director
of human resources required help to choose between
one of two candidates for a job. The only difference be-
tween the two candidates was that one of them spoke
with a Brazilian accent. Thus, the director decided to
choose the candidate speaking with a native accent.
We defined a ‘Brazilian accent’as the accent of a Brazil-
ian man who lives and looks for a job in Portugal, not a
Brazilian tourist. It is important to distinguish these two
target groups. While Brazilian tourists typically are not
familiar with the way of speaking Portuguese in Portu-
gal, Brazilian immigrants already have some contact
with the manner of speaking in Portugal and therefore
try to use local colloquialisms. This characterize them
as specific targets, different from tourists who do not
represent a threat to the status of the ingroup. The goal
here is to analyse what people think of this particular
group (i.e., Brazilian immigrants) and not Brazilians in
general. Specifically, the text used was as follows:
‘A reputed company invited candidates to apply for a
series of jobs in business consulting and initiated a
process of recruitment and selection. The company’s
director of human resources received five applica-
tions. After a detailed analysis of the candidates’
CVs, the director selected the two best candidates
for a telephone interview. During the interview, the
director failed to notice any significant differences be-
tween the two candidates, except that one of them
spoke with a Brazilian accent.’
In both experimental conditions, we presented a justifi-
cation for the director’s behaviour, but in one of them,
we used the candidate’s accent for justification, and in
the other, condition we did not mention the accent. In
the condition of deciding mentioning the candidate’sac-
cent, the participants read that ‘the director decided to
hire the candidate speaking with a European Portu-
guese accent because he thought that he/she best fit
the role to be performed at the company, which is to in-
teract with customers who consider the accent of the
Portuguese spoken in Portugal more understandable
and correct’. In the condition of deciding without men-
tioning the candidate’s accent, the participants read that
the director decided to hire the Portuguese candidate
because after listening to the candidates he dispelled
any questions that he might have about what would
be the best choice. After reading the text, the partici-
pants answered a 7-point measure-of-legitimacy ques-
tionnaire evaluating whether the decision to select the
Portuguese candidate was correct, appropriate, neces-
sary and legitimate. A mean score of these items was
created (α=.92).
We hypothesized that if accent is perceived as a legit-
imate reason for discrimination, the participants should
evaluate the decision to hire the Portuguese candidate
as more legitimate, correct, appropriate and necessary
in the ‘mentioning accent condition’than in the ‘non-
mentioning accent condition’.At-test for independent
samples showed that the participants judged as more
legitimate the hiring decision in the mentioning accent
condition (M=4.30, SD = 1.31) than in the non-men-
tioning accent condition (M=3.25, SD =1.18), t(35) =
2.56, p=.015. In addition, we compared the means
obtained in each condition with the midpoint of the
scale and found that in the control condition (non-men-
tioning accent condition), discrimination was clearly not
legitimate because the mean was significantly below the
midpoint, t(16) = 2.62, p= .018. In contrast, in the ex-
perimental condition it was above the midpoint, but this
difference was not significant, t(19)= 1.02, p=.320.
L. E. C. Souza, et al. Legitimization of accent on discrimination
European Journal of Social Psychology 46 (2016) 609–620 Copyright © 2016 John Wiley & Sons, Ltd. 611
Thus, the data gathered in this pilot study support our
hypothesis that accent is perceived as a legitimate rea-
son for favouring ingroup rather than outgroup
speakers.
Study 1
The pilot study showed that information about the tar-
get’s accent is perceived as a fair and legitimate reason
to discriminate. This study aimed to go further, by test-
ing the hypothesis that discrimination against non-
native speakers is influenced by the target’s accent only
in individuals who hold prejudiced attitudes against
that social category activated by the speaker’s accent.
In this sense, we proposed that the target’s accent
(native vs. non-native) will influence discrimination
in more-prejudiced individuals but not in less-
prejudiced ones. The participants were invited to en-
gage in an online study conducted using a decision-
making scenario used in recruiting candidates for a
job. We presented five CVs and then stated that only
two candidates had been selected for interview. In the
next section, we said that the participants would listen
to an excerpt from an interview with one of the candi-
dates. We manipulated the candidates’accents, which
were either European or Brazilian Portuguese. The
participants then indicated the extent to which each
applicant should be hired (discrimination measure).
According to our rationale, only individuals highly
prejudiced against Brazilian immigrants would use the
target’s accent in their decision about hiring; that is,
the target speaking with a Brazilian immigrant accent
will be hired less often than the target speaking with a
Portuguese accent, indicating discrimination against
the Brazilian immigrant based on his accent. Otherwise,
for low-prejudiced individuals, there will be no differ-
ence in hiring the Portuguese or the Brazilian candi-
date, indicating absence of discrimination.
Method
Participants and design. Seventy-one Portuguese
university students participated in this study
(M
age
=21.8, SD = 4.27; 34 male and 37 female). The
participants were randomly allocated to one of two con-
ditions (Portuguese accent vs. Brazilian accent) in a
between-subject single-factor design; 35 participants
evaluated the Brazilian candidate, and 36 participants
evaluated the Portuguese candidate.
Accent manipulation. All participants heard a
23-second excerpt from a job interview in which
the candidate introduced himself. Both conditions
had the same text and duration; the only difference
being the target’s accent, which was recorded either
by a native Portuguese man or a native Brazilian
man. The excerpt spoken by the targets was as
follows:
‘I am applying for the position of business consultant
in the area of telecommunications and information
technology. I have a degree in management. I have
three years’experience in this area and recently com-
pleted a postgraduate degree in business administra-
tion. I am also available to work full time in the
offices of Amoreiras, Saldanha, and Parque Expo.’
Discrimination measure. We operationalized dis-
crimination by asking the participants to ‘consider the
information that you have about the candidate. If
you were the manager of the company, what is the
likelihood of your hiring the candidate?’The partici-
pants chose answers from a 4-point scale varying from
1(not at all likely)to4(very likely).
Prejudice measure. We measured prejudice against
Brazilian immigrants by using six items from the
Portuguese version of the blatant and subtle prejudice
scale (Lima-Nunes et al., 2013; Pettigrew & Meertens,
1995; Vala, Lopes, & Lima, 2008). We selected the best
six items having as a criterion those with the highest
factor loadings in the study by Lima-Nunes et al.
(2013) of prejudice and discrimination against
Brazilian immigrants (e.g., ‘Brazilian immigrants come
from less able races and this explains why they are not
as well off as most Portuguese people’and ‘Compared
with the Portuguese people, Brazilian immigrants are
very different in the language that they speak’). The
participants indicated the extent to which they agreed
with each statement on a 7-point scale varying from 1
(totally disagree)to7(totally agree). The scores were sub-
mitted to an exploratory factor analysis (using the princi-
pal axis factoring method of extraction), which revealed
only one factor that explained 34.52% of the variance
(eigenvalue = 2.07; factor loadings from .38 to .69;
α= .72). We compared the mean prejudice scores
between the two experimental conditions. A t-test for in-
dependent samples showed that there was no significant
difference between participants who heard the Brazilian
accent (M=2.20; SD =0.80) and those who heard the
Portuguese accent (M=2.18; SD =0.99), t(67) = 0.09,
p= .928.
Manipulation check. At the end of the procedure,
we asked the participants to indicate the candidate’sac-
cent. Only one participant answered this question incor-
rectly. As a result, he was excluded from the analysis
and the final sample was therefore 71 students.
Results
We regressed discrimination on the accent (Portuguese
vs. Brazilian), prejudice and the interaction term. The
results showed a reliable main effect of accent, b=.37,
t(67) = 2.21, p=.031, η
2
p
= .26. Importantly, this effect
was qualified by a reliable interaction between preju-
dice and accent, b=.51, t(67) = 2.67, p=.010, η
2
p
=.31.
This means that prejudice moderates the influence of
L. E. C. Souza, et al.Legitimization of accent on discrimination
European Journal of Social Psychology 46 (2016) 609–620 Copyright © 2016 John Wiley & Sons, Ltd.612
accent on discrimination. As can be seen in Figure 1a,
low-prejudice participants (1.0 SD below the prejudice
mean) were equally likely to hire the Portuguese
(M= 2.97, SD = 0.16) and the Brazilian (M=3.06,
SD = 0.18) candidate, t(67) = 0.37, p= .714. However,
when prejudice was high (+1 SD above the prejudice
mean), the probability of hiring the Portuguese candi-
date (M=3.25,SD = 0.16) was greater than the probabil-
ity of hiring the Brazilian candidate (M=2.43,
SD = 0.18), t(67) = 3.46, p= .001. This means that listen-
ing to the target’s accent had an influence on discrimi-
nation only in participants who were more prejudiced
against Brazilian immigrants.
Discussion
This study provides the first experimental evidence that
listening to a target’s accent (native vs. immigrant) in-
fluences discrimination against this target, with this in-
fluence being moderated by prejudice. This result
supports our first hypothesis, according to which only
prejudiced individuals should be more likely to use in-
formation about targets’accents when making a deci-
sion about hiring candidates for a job, even when
these candidates have the same professional experience.
This means that accent leads to discrimination only in
individuals who already have negative attitudes about
atarget’s membership group. According to our ratio-
nale, this occurs because individuals engage in a legiti-
mizing process in which the perception of the quality
of an accent plays a central role. The next study ad-
dresses this hypothesis.
Study 2
This study intended to replicate Study 1 and go further
by analysing how information about accents leads to
discrimination in highly prejudiced individuals. As in
the previous study, the participants were invited to take
part in an online study conducted using a decision-
making scenario used in recruiting candidates for a
job. Once again, we manipulated the candidates’
accents, which were either European or Brazilian
Portuguese. The participants then answered a
measure-of-accent assessment and indicated the proba-
bility of their hiring the candidate (discrimination mea-
sure). We predicted that hearing a target accent (native
vs. immigrant) influences discrimination against this
target only in highly prejudiced individuals. Considering
that the perception of accent quality is a legitimizing
factor of discrimination, then this perception should
work as a mediator of the accent influence in the dis-
crimination, process similar to that seen in previous
studies (e.g., Lima-Nunes et al., 2013, Study 1). Accord-
ing to our rationale, and based on the results of the pilot
study showing that discrimination based on accent is
perceived as legitimate, the influence of accent on the
hiring decision should be mediated by an assessment
of the candidate’s accent, because accents trigger a pro-
cess of legitimation that is needed to justify the decision
to hire a member of the ingroup.
Method
Participants and design. One hundred and
twenty-nine Portuguese university students partici-
pated in this study. However, five participants missed
Fig. 1: Probability of hiring the Portuguese and Brazilian candidates as a function of candidates’accent and prejudice in Study 1 (a), Study 2 (b) and
Study 3 (c)
L. E. C. Souza, et al. Legitimization of accent on discrimination
European Journal of Social Psychology 46 (2016) 609–620 Copyright © 2016 John Wiley & Sons, Ltd. 613
the manipulation check, so the final sample was 124
students (M
age
=23.7, SD = 4.83; 60 male and 64 fe-
male). The participants were randomly allocated to
one of two conditions (Portuguese accent vs. Brazilian
accent) of a between-subject single-factor design.
Accent manipulation. The content of the manip-
ulation was the same as that used in Study 1. Thus, the
participants were exposed to and evaluated one of
the two candidates, this presentation being fully
randomized.
Assessment of accent measure. We developed a
12-item measure to assess the candidate’s accent (e.g.,
‘understandable’,‘sounds good’,‘correct’,‘ugly’and
‘nice’). The participants indicated the extent to which
each of the adjectives described the speech that they
had just heard using a 7-point scale (1=not at all;7=a
lot). The scores were submitted to exploratory factor
analysis (using the principal axis factoring method of
extraction), which revealed only one factor that ex-
plained 32.39% of the variance (eigenvalue = 3.89; fac-
tor loadings from .32 to .93). Three items were excluded
because they had low factor loadings (under .30). Reli-
ability analysis indicated that this scale had good inter-
nal consistency (α=.84).
Discrimination measure. The measure was the
same used in the previous study, in which participants
indicated the probability of their hiring each candidate
on a 4-point scale varying from 1 (not at all likely)to4
(very likely).
Manipulation checks. We used the same manip-
ulation check as in the previous study. Five participants
did not answer this question correctly and were ex-
cluded from the analysis.
Prejudice measure. The measure consisted of the
same six items from the Portuguese version of the bla-
tant and subtle prejudice against Brazilian immigrants
scale used in a previous study (α= .72). We compared
the mean scores of prejudice between the two condi-
tions of the experiment. A t-test for independent sam-
ples showed that there was no significant difference
between participants who heard the Brazilian accent
(M=2.26; SD = 0.85) and those who heard the Portu-
guese accent (M=2.22; SD =1.05),t(72) = 0.16, p= .874.
Results
We used a multiple-regression approach (Baron &
Kenny, 1986) to test whether the influence of accent
on discrimination is moderated by prejudice and medi-
ated by the assessment of the accent in question. For this
purpose, we estimated three regression models. The
parameters estimated for the three steps are shown in
Table 1.
In the first model, we regressed discrimination on ac-
cent, prejudice, and the interaction term. The results
showed a reliable main effect of accent. Importantly,
the interaction between prejudice and accent predicted
discrimination. This means that our first hypothesis that
prejudice moderates the relationship between accent
and discrimination was corroborated. As can be seen in
Figure 1b, low-prejudice participants (1.0 SD below
the prejudice mean) were equally likely to hire the Por-
tuguese (M=2.86, SD = 0.11) and the Brazilian
(M=2.80, SD = 0.11) candid ate, t(120) = 0.37, p=.714.
However, in participants with high prejudice (+1 SD
above the prejudice mean), the Portuguese candidate
(M=2.98, SD =0.11)wasmorelikelytobehiredthan
the Brazilian one (M=2.35, SD =0.10), t(120) = 4.20,
p<.001. This means that the accent had an influence
on discrimination only in participants who were more
prejudiced against Brazilian immigrants.
In the second model, we regressed the assessment of
accent on prejudice, accent and the interaction term.
The results indicated that accent has a reliable effect
on the assessment of accent, demonstrating that the
Portuguese accent (M=4.80, SD = 0.11) is more posi-
tively evaluated than the Brazilian accent (M= 4.39,
SD =0.11), t(120) = 2.63, p= .010. Importantly, this ef-
fect is moderated by prejudice in such a way that low-
prejudice participants (1.0 SD below the prejudice
mean) showed no difference in the way they assessed
the Portuguese accent (M=4.67,SD = 0.16) and the Bra-
zilian accent (M=4.60, SD =0.16), t(120) = 0.32,
p= .745. However, high-prejudice participants (+1 SD
above the prejudice mean) assessed the Portuguese
accent more positively (M=4.94, SD = 0.16) than the
Brazilian accent (M=4.19, SD =0.15), t(120) = 3.37,
p= .001. This means that only individuals with more
negative attitudes against Brazilian immigrants perceive
Table 1. Parameters estimated according to regression models used in
analysis of the role of assessment of accent in the influence of accent on
discrimination
Criterio n variables
Step 1: D Step 2: AA Step 3: D
Predictors bb b
Intercept 2.75** 4.60** 2.75**
Accent (A) 0.35** 0.41** 0.19*
Prejudice (P) 0.09 0.03 0.07
A × P 0.31** 0.36* 0.17
Assessment of
accent (AA) 0.38**
AA × P 0.00
Model
information R=.39 R=.30 R=.65
R
2
Adjusted
=.13 R
2
Adjusted
=.07 R
2
Adjusted
=.40
F(3,120) = 7.22 F(3,120) = 4.01 F(5,118) = 17.10
p<.001 p<.01 p<.001
Note:b= unstandardized coefficients; D, discrimination.
*p<.05.
**p<.01.
L. E. C. Souza, et al.Legitimization of accent on discrimination
European Journal of Social Psychology 46 (2016) 609–620 Copyright © 2016 John Wiley & Sons, Ltd.614
the quality of a Portuguese accent as better than a
Brazilian one.
In the third model, we added the assessment of accent
and its interaction with prejudice to predict discrimina-
tion. The results indicated that the effect of the assess-
ment of accent on discrimination was significant. The
interaction effect between accent and prejudice de-
clined substantially and was no longer significant. This
result indicates that the relationship between accent
and discrimination, besides being moderated by preju-
dice, is mediated by the assessment of the target’sac-
cent. The mediation effect was reliable with 5000 bias
corrected resamples from bootstrapping (indirect
effect= 0.16; 95%CI: 0.04 to 0.31). To estimate the con-
fidence intervals through bootstrapping, we used struc-
tural equation modelling with AMOS for Windows. In
the Hayes PROCESS, this procedure is equivalent to
model 59 (Hayes, 2013).
In order to interpret this mediated-moderation, we
analysed the effect of accent on discrimination against
the Brazilian candidate by taking into account different
levels of prejudice. Specifically, we broke down the me-
diating effect of assessment of accent into participants
with a low prejudice (i.e., those with 1.0 SD below
the prejudice mean) and those with a high prejudice
(i.e., those with +1.0 SD above the prejudice mean). As
Figure 2a shows, for participants with higher levels of
prejudice, the effect of accent on discrimination was
mediated by the assessment of the accent (indirect ef-
fect = 0.29; 95%CI: 0.08 to 0.54 with 5000 bias
corrected resamples through bootstrapping). In other
words, for highly prejudiced individuals, the accent of
a candidate speaking European Portuguese is better
evaluated than that of a candidate speaking Brazilian
Portuguese, which in turn leads to the Portuguese can-
didate being hired instead of the Brazilian candidate.
For participants with lower levels of prejudice, accent
predicted neither discrimination nor the assessment of
accent (indirect effect = 0.10; 95%CI: 0.29 to 0.05).
Discussion
This study reinforces evidence that the target accent
triggers discrimination against this target, which sup-
ports our first hypothesis. In fact, the results replicate
the previous study, showing that a candidate with a
Portuguese accent is significantly more likely to be hired
than one having a Brazilian accent, by more prejudiced
participants but not by less prejudiced ones.
Importantly, confirming our second hypothesis, the
results showed that the influence of accent on discrimi-
nation is mediated by individual perception of the qual-
ity of the accent concerned. This means that the
assessment of an accent’s quality can be the psycholog-
ical mechanism that underlies the influence of accent
on discrimination. In other words, in assessing a
speaker’s accent, prejudiced individuals tend to evaluate
the ingroup accent more positively in a manner that le-
gitimates their decision, which was made in a context of
intergroup discrimination.
Despite providing evidence for our hypotheses, the
findings would be stronger if we had given the partici-
pants the opportunity to hear both candidates and
then decide which one to hire. Moreover, stronger
evidence for the role played by accent quality would
have been obtained if we had controlled for the effect
of stereotyping, because the accent effect can be
confounded with the way individuals use warmth and
competence stereotypes when evaluating standard and
non-standard accent speakers (e.g., Yzerbyt et al.,
2005). To overcome these limitations, we carried out a
third study with a within-subject design.
Study 3
This study intended to replicate the previous ones by
using a within-subject design where accent leads to dis-
crimination, with this influence being moderated by
prejudice and mediated by assessment of the candidate’s
Fig. 2: The effect of accent on discrimination moderated by prejudice (coefficients in bold refer to participants with higher levels of prejudice) and
mediated by assessment of accent in Study 2 (a) and Study 3 (b)
L. E. C. Souza, et al. Legitimization of accent on discrimination
European Journal of Social Psychology 46 (2016) 609–620 Copyright © 2016 John Wiley & Sons, Ltd. 615
accent. Moreover, we measured warmth and compe-
tence stereotypes to test the mediating role of accent
quality even after controlling for the effect of
stereotyping. As in the previous study, the participants
were invited to take part in an online study conducted
using a decision-making scenario used in recruiting can-
didates for a job. They first heard one of two recorded
excerpts of a job interview. Next, they answered ques-
tions designed to assess the accent, attributed scores to
the candidate’s warmth and competence, and indicated
the probability of their hiring the candidate (discrimina-
tion measure). Then, the participants heard the other
recording and answered the questions again for the
other candidate. At the end, the participants answered
the prejudice questionnaire. We hypothesized that ac-
cent influences discrimination in highly prejudiced par-
ticipants. Importantly, if the influence of accent is
because of participants’perception of the features of
the target’s accent, this influence should be mediated
by the perception of the accent’s quality. Instead, if the
influence of accent is motivated by stereotyping, this in-
fluence should be mediated by the greater attribution of
warmth stereotypes to the Brazilian than to the Portu-
guese target and lesser attribution of competence ste-
reotypes to the Brazilian than to the Portuguese target,
because Brazilians are perceived having low status and
forming an uncompetitive outgroup (Lima-Nunes,
2013).
Method
Participants and design. One hundred and five
Portuguese university students participated in this study
(M
age
=24.1, SD = 4.37; 58 male and 47 female). This
experiment consisted of a single factor (Portuguese
accent vs. Brazilian accent) within-subject design.
Accent manipulation. The content of the manip-
ulation was the same as that used in Studies 1 and 2.
However, because we used a within-subject design,
the participants were exposed to and evaluated one
candidate and then were exposed to and evaluated
the other candidate, this presentation being fully
randomized.
Assessment of accent measure. The measure was
the assessment of the candidates’accents that was used
in a previous study (α= .85 for Portuguese accent and
α= .89 for Brazilian accent).
Stereotype measure. We used the competence
and warmth stereotypes (Fiske et al., 2002) to measure
Brazilian and Portuguese stereotypes. We asked the par-
ticipants to indicate the extent to which the candidates
presented each of seven traits, using a 7-point scale
(1 = not at all;7=alot). We submitted the scores to two
exploratory factor analyses (using the principal axis
factoring method of extraction) with varimax rotation.
For the Portuguese stereotypes, the results revealed
two factors that explained 49.98% of the variance
(eigenvalues = 2.09 and 1.90) for competence stereo-
type factor loadings from .58 (competitive) to .68 (inde-
pendent), α=.73 and warmth stereotypes from .55
(trustworthy) to .80 (warm), α= .77. For the Brazilian
stereotypes, the results also revealed two factors that ex-
plained 50.70% of the variance (eigenvalues=2.52 and
1.53) for competence stereotype factor loadings from
.34 (competent) to .72 (competitive), α= .76 and
warmth stereotypes from .45 (warm) to .82 (trustwor-
thy), α= .78.
Discrimination measure. We used the same mea-
sure employed in the previous studies; that is, the partic-
ipants indicated the probability of their hiring each
candidate on a 4-point scale varying from 1 (notatall
likely)to4(very likely).
Prejudice measure. The measure consisted of the
same six items from the Portuguese version of the bla-
tant and subtle prejudice scale used in the previous
studies (α= .79).
Results
We expected that the influence of accent on the proba-
bility of the candidates being hired would be moderated
by prejudice. That is, the Portuguese candidate would be
more likely to be hired than the Brazilian candidate, but
only by more prejudiced participants. Moreover, we
predicted that this moderation would be mediated by
the assessment of the candidate’s accent, even when
taking into account the role played by the warmth and
competence stereotypes attributed to the targets. To test
these hypotheses, we estimated a mediated-moderation
model in a within-subject design. In line with the sug-
gestions of Judd, Kenny and McClelland (2001), we
used difference scores to represent the within subject-
effect of the target on discrimination (i.e., the dependent
variable), the assessment of the candidate’s accent and
the warmth and competence stereotypes (i.e., the medi-
ators) in multiple regression analyses. In this sense, we
estimated four regression models. The parameters esti-
mated for the three steps are shown in Table 2.
In the first model, we regressed the difference scores
for discrimination (D) on prejudice. In this model, the
intercept represents the main effect of the targets’
accents on discrimination, while the prejudice effect
represents the two-way interaction between accent
and prejudice. As predicted, the results showed that
the accent per se was not sufficient for discrimination
to take place. However, the effect of prejudice was reli-
able and indicated that prejudice moderates the influ-
ence of accent on discrimination. As can be seen in
Figure 1c, low-prejudice participants (1SD from prej-
udice mean) were just as likely to hire the Portuguese
candidate (M=2.72, SD = 0.09) as the Brazilian one
L. E. C. Souza, et al.Legitimization of accent on discrimination
European Journal of Social Psychology 46 (2016) 609–620 Copyright © 2016 John Wiley & Sons, Ltd.616
(M= 2.83, SD =0.09),t(102) = 0.98, p= .331. However,
when prejudice was high (+1 SD from prejudice mean),
the probability of hiring the Portuguese candidate
(M= 2.72, SD = 0.09) was greater than the probability
of hiring the Brazilian candidate (M=2.44, SD =0.09),
t(102) = 2.50 p= .014. As in Studies 1 and 2, this means
that the targets’accents influenced discrimination only
in participants who were more prejudiced against
Brazilian immigrants.
In the second model, we regressed the difference
scores of the mediators (AA, CS and WS) on prejudice.
In this model, the intercept represents the mean differ-
ence between the experimental conditions used in
assessing accent and competence and warmth stereo-
types. The results indicated that accent has a reliable ef-
fect only in assessing accent and warmth stereotypes.
That is, the Portuguese candidate was judged to have a
better accent (M=4.59, SD = 0.83) than the Brazilian
candidate (M=4.36,SD = 0.97), but the Brazilian immi-
grant was evaluated as being warmer (M=4.37,
SD = 0.77) than the Portuguese candidate (M=4.12,
SD = 0.77).
In the last model, we regressed discrimination on
prejudice, on the difference scores of the mediators,
and on the interaction terms. The results indicated that
only the assessment of accent predicted discrimination
as this was the only mediating variable in the process ac-
cording to 5000 bias corrected resamples obtained by
bootstrapping (indirect effect=0.04; 95%CI: 0.01 to
0.10), even when controlling for warmth and compe-
tence. To estimate the confidence intervals through
bootstrapping, we used structural equation modelling
with AMOS for Windows. There is no equivalent model
for this mediated moderation in the within-subject de-
sign in the Hayes PROCESS (Hayes, 2013). Moreover,
the effect of prejudice decreased substantially and was
no longer significant, which indicates that assessment
of accent mediates the role played by prejudice in the re-
lation between accent and discrimination.
In order to interpret this mediated moderation, as in
the previous study, we analysed the effect of accent on
discrimination against immigrants by taking into ac-
count different levels of prejudice. Specifically, we
broke down the mediating effect of assessmentof accent
into low-prejudice participants (i.e., those with 1.0 SD
below the prejudice mean) and high-prejudice partici-
pants (i.e., those with +1.0 SD above the prejudice
mean). As Figure 2b shows, for participants with higher
levels of prejudice, the effect of accent on discrimination
was mediated by the assessment of accent (indirect ef-
fect = 0.08; 95%CI: 0.01 to 0.25). In other words, being
highly prejudiced significantly predicted a better evalu-
ation of the Portuguese accent than the Brazilian accent,
which in turn led to the Portuguese candidate to be
hired more thanthe Brazilian one. For participants with
lower levels of prejudice, the candidate’s accent pre-
dicted neither discrimination nor the assessment of ac-
cent (indirect effect = 0.00; 95%CI: 0.07 to 0.06).
Discussion
The main effect of accent we found in the two previous
studies did not arise in this one. We think this is because
of the within-subject experimental design used here. In-
deed, participants had the opportunity to compare the
two candidates for the job, and thus, it is very likely that
the comparison aroused anti-prejudiced concerns that
motivated the reduction in accent strength. However,
the pattern is a strong test for our prediction according
to which the influence of accent on discrimination is
moderated by prejudice and mediated by assessment
of accent. Using a different design, results demonstrated
that targets’accents influenced discrimination only
when prejudice was high, and this influence was
Tab le 2 . Parameters estimated for mediated moderation
Criterio n variables
Step 1: D Step 2: Step 3: D
AA CS WS
Predictors bb b bb
Accent (I) 0.08 0.23* 0.10 0.25* 0.02
Prejudice (P) 0.18* 0.18
+
0.01 0.13 0.09
AA 0.39**
CS 0.14
WS 0.04
P*AA 0.06
P*CS 0.02
P*WS 0.04
Model Information R=.26 R=.27 R=.12 R=.28 R=.65
R
2
Adjusted
=.05 R
2
Adjusted
=.06 R
2
Adjusted
=.005 R
2
Adjusted
=.06 R
2
Adjusted
=.38
F(2,102)= 3.60 F(2,103)= 4.10 F(2,101) = 0.74 F(2,101) = 4.39 F(8,95) = 8.80
p=.031 p=.019 p=.478 p=.015 p<.001
Note:b= unstandardized coefficients; D, discrimination; I, intercept; AA, assessment of accent; CS, competence stereotype; WS, warmth stereotype.
*p<.05.
**p<.01.
+
p=.065.
L. E. C. Souza, et al. Legitimization of accent on discrimination
European Journal of Social Psychology 46 (2016) 609–620 Copyright © 2016 John Wiley & Sons, Ltd. 617
mediated by the assessment of the target’s accent. The
effect occurred independently of stereotyping about
the target. Although accent predicted warmth stereo-
types, in the sense that the participants attributed more
warmth stereotypes to the Brazilian candidate than to
the Portuguese one, which corroborates other research
(Edwards, 1982; Fuertes et al., 2012), these stereotypes
did not mediate the influence of accent on discrimina-
tion against immigrants.
General Discussion
We analysed the influence of accent on discrimination
against immigrants. Study 1 explored the role of preju-
dice in this influence and revealed that listening to an
immigrant’s accent led to discrimination against this im-
migrant only in the case of the more prejudiced partici-
pants. The question that arises here is whether the mere
fact of an accent being perceived as native versus non-
native is sufficient to lead to discrimination against the
non-native speaker. Indeed, in people with low levels
of prejudice, accent does not lead to discrimination
against the immigrant. Importantly, this pattern of re-
sults was replicated in Studies 2 and 3.
The mediating role played by the assessment of accent,
as demonstrated in Study 2 and replicated in Study 3,
follows and extends the results of previous studies, dem-
onstrating the importance of legitimizing factors to un-
derstand discrimination (Pereira, Vala, & Costa-Lopes,
2010; Pereira, Vala, & Leyens, 2009). One of the innova-
tive aspects of this research programme is demonstrating
that the evaluation that people make about the accent of
a target group assumes this role in legitimizing
prejudiced individuals. In accordance with our predic-
tions, for more prejudiced individuals, information
about accent is an important criterion when they make
hiring decisions for job candidates; insofar, this decision
is mediated by assessment of the quality of the target’s
accent. Our rationale is that the assessment of a speaker’s
accent is perceived as a non-prejudiced way of justifying
discriminatory behaviour, as we found in the pilot study.
While it is morally and ethically unacceptable to use in-
formation about race or ethnicity to discriminate
(Lippi-Green, 1997), evaluating the way that a person
speaks is commonly accepted and perceived as legitimate
in society (Gluszek & Dovidio, 2010; Ng, 2007).
Theoretical Implications
In current day globalized societies where immigrants
are competing with nationals for jobs, the study of hir-
ing discrimination has high social relevance. Indeed,
accent is one of the key markers differentiating
immigrants from the national majority. Studying the ac-
cent effect has practical implications for social psycholo-
gists not only because it provides insights about the
dynamics of intergroup relations involving linguistic
skills but also because it is useful for planning interven-
tion in several social situations, especially in the labour-
market context. The current research programme goes
beyond the implications of the applied effect of
speakers’accent, because it also contributes to theoriz-
ing and research on intergroup discrimination in at least
three ways: (i) it sheds light on literature about the ac-
cent effect because it shows experimental evidence for
the influence of accent on actual discrimination against
immigrants; (ii) it suggests that accent per se is not
enough to trigger discrimination because the accent ef-
fect would depend on negative attitudes that individuals
have about the target groups; and (iii) it highlights the
legitimizing role played by accent in social inequality.
In fact, this paper contributes to a better understand-
ing of the role played by categorization in intergroup
discrimination (Park & Judd, 2005) because it shows ex-
perimentally that the accessibility of social categories per
se is not sufficient to lead to discrimination and that this
pervasive effect occurs only in individuals who have
negative attitudes towards the target group. We show
that accents are really linked to categorization and that
this categorization then triggers negative evaluations
about targets (Fuertes et al., 2012; Rakićet al., 2011)
and leads to discrimination against them, specifically
immigrants with different accents. Our research goes
further and shows in Study 3 that the psychological
mechanism that explains the influence of accent on dis-
crimination is, in fact, the assessment of the quality of
the target’s accent even after controlling for the effect
of stereotyping.
We also believe that this paper contributes to the study
of processes by which social inequalities are legitimated
(Costa-Lopes et al., 2013), as it reveals a new, seemingly
unprejudiced argument that is capable of legitimizing
discrimination (Pereira et al., 2010). Indeed, perceiving
the quality of an accent to be legitimate information in
deciding in favour of a native candidate is in line with so-
cial psychology research and theory, which reveals that
prejudice and discrimination persist because individuals
have developed indirect ways and legitimated forms of
discrimination (e.g., Crandall & Eshleman, 2003;
Dovidio & Gaertner, 2004; Sidanius & Pratto, 1999).
Moreover, as pointed out by Pereira et al. (2009), the re-
lationship between prejudice and discrimination needs
to be legitimized because individuals’self-concepts hold
internalized egalitarian justice principles that require
them to have good reasons to discriminate. This paper
shows that the assessment of the way a person speaks
can work as a strong argument to justify discrimination.
Limitations and Further Directions
Although the results reported here support our hypoth-
eses, these studies have some important limitations.
Although contributing to the field of accent research
by experimentally demonstrating the prejudiced-based
accent effect on actual discrimination, we did not ma-
nipulate prejudice. The observational nature of the role
played by prejudice on accent effect weakens the evi-
dence in support the idea that discrimination against
non-standard speakers is motivated by prejudice. We
think that further research can improve the evidence
L. E. C. Souza, et al.Legitimization of accent on discrimination
European Journal of Social Psychology 46 (2016) 609–620 Copyright © 2016 John Wiley & Sons, Ltd.618
for this hypothesis by manipulating not only the accent
but also prejudice. In addition, our predictions clearly
posit that accent influences only individuals who
already have prejudiced attitudes against targets’social
category, but does not provide insight about the accent
effect on individuals’prejudiced attitudes. Future
research can better investigate this issue by testing the
role played by accent on factors known to directly
influence prejudice, such as intergroup contact and the
quality of this contact.
Despite these limitations, the research presented here
shows strong evidence that accent influences discrimi-
nation only in more prejudiced people. Furthermore,
this effect is mediated by assessment of the target’sac-
cent in the sense that the participants tend to evaluate
the ingroup accent more positively as a way of justifying
their discriminatory behaviour as legitimate.
Acknowledgements
This work was partially supported by a grant from CNPq,
Ministério da Ciência e Tecnologia, Brazil, awarded to
Luana de Souza (SWE-245390/2012-0), by a grant from
the Institute of Social Sciences of the University of Lisbon
(PEst-OE/SADG/LA0013/2013), and by a grant from the
Fundação para a Ciência e a Tecnologia, awarded to
Cicero Roberto Pereira (PTDC/PSI-PSO/114159/2009).
We thank the Lisbon SPARC Research Group of the In-
stitute of Social Sciences of the University of Lisbon for
their helpful comments on previous versions of these
studies. We would also like to thank Eva G. T. Green
and two reviewers for their insightful comments that
greatly contributed to improving this paper.
References
Allport, G. W. (1954). The nature of prejudice. Wokingham:
Addison-Wesley.
Baron, R. M., & Kenny, D. A. (1986). The moderator-
mediator variable distinction in social psychological re-
search: Conceptual, strategic, and statistical considerations.
Journal of Personality and Social Psychology,51,1173–1182.
http://dx.doi.org/10.1037/0022-3514.51.6.1173
Boyd, S. (2003). Foreign-born teachers in the multilin-
gual classroom in Sweden: The role of attitudes to
foreign accent. International Journal of Bilingual Education
and Bilingualism,6,283–295. http://dx.doi.org/10.1080/
13670050308667786
Bradac, J. J. (1990). Language attitudes and impression for-
mation. In H. Giles & W. P. Robinson (Eds.), Handbook of
language and social psychology (pp. 387–412). New York:
Wiley.
Bresnahan, M. J., Ohashi, R., Nebashi, R., Liu, W. Y., &
Shearman, S. M. (2002). Attitudinal and affective
response toward accented English. Language and Com-
munication,22, 171–185. http://dx.doi.org/10.1016/
S0271-5309(01)00025-8
Costa-Lopes, R., Dovidio, J. F., Pereira, C. R., & Jost, J. T.
(2013). Social psychological perspectives on the legitima-
tion of social inequality: Past, present and future. European
Journal of Social Psychology,43,229–237. http://dx.doi.org/
10.1002/ejsp.1966
Costa-Lopes, R., Pereira, C. R., & Judd, C. M. (2014).
Categorisation salience and ingroup bias: The buffering
role of a multicultural ideology. International Journal
of Psychology,49 508–512.http://dx.doi.org/10.1002/ijop.
12044
Crandall, C. S., & Eshleman, A. (2003). A justification-
suppression model of the expression and experience of
prejudice. Psychological Bulletin,129,414–446. http://dx.
doi.org/10.1037/0033-2909.129.3.414
Derwing, T. M., & Munro, M. J. (2009). Putting accent
in its place: Rethinking obstacles to communication.
Language Teaching,42,476–490. http://dx.doi.org/10.1017/
S026144480800551X
Devine, P. G. (1989). Stereotypes and prejudice: Their auto-
matic and controlled components. Journal of Personality
and Social Psychology,56,5–18. http://dx.doi.org/10.1037/
0022-3514.56.1.5
Dovidio, J. F. (2001). On the nature of contemporary preju-
dice: The third wave. Journal of Social Issues,57,829–849.
http://dx.doi.org/10.1111/0022-4537.00244
Dovidio, J. F., & Gaertner, S. L. (2000). Aversive racism and se-
lection decisions: 1989 and 1999. Psychological Science,11,
315–319. http://dx.doi.org/10.1111/1467-9280.00262
Dovidio, J. F., & Gaertner, S. L. (2004). Aversive racism.In M.
P. Zanna (Ed.), Advances in experimental social psychology
(Vol. 36, pp. 1–52). Thousand Oaks, CA: Sage.
Edwards, J. (1982). Languageattitudes and their implications
among English speakers. In E. B. Ryan, & H. Giles (Eds.),
Attitudes toward language variation: Social and applied contexts
(pp. 20–33). London: Edward Arnold.
Edwards, J. (1999). Refining our understanding of language
attitudes. Journal of Language and Social Psychology,18,
101–110. http://dx.doi.org/10.1177/0261927X99018001007
European Commission. (2008). Discrimination in the European
Union: Perceptions, experiences and attitudes (Special
Eurobarometer 296, Wave 69.1). Retrieved 7 January,
2013, from http://ec.europa.eu/public_opinion/archives/
ebs/ebs_296_en.pdf
Fiske, S. T., Cuddy, A. J. C., Glick, P., & Xu, J. (2002). A
model of (often mixed) stereotype content: Competence
and warmth respectively follow from perceived status
and competition. Journal of Personality and Social Psychol-
ogy,82,878–902. http://dx.doi.org/10.1037/0022-3514.82.
6.878
Fuertes, J. N., Gottdiener, W., Martin, H., Gilbert, T. C., &
Giles, H. (2012). A meta-analysis of the effects of speakers’
accents on interpersonal evaluations. European Journal of
Social Psychology,42,120–133. http://dx.doi.org/10.1002/
ejsp.862
Gaertner, S. L., & Dovidio, J. F. (2005). Understanding and
addressing contemporary racism: From aversive racism to
the common ingroup identity model. JournalofSocial
Issues,61, 615–639. http://dx.doi.org/10.1111/j.1540-4560.
2005.00424.x
Giles, H., & Johnson, P. (1987). Ethnolinguistic identity the-
ory: A social psychological approach to language mainte-
nance. International Journal of the Sociology of Language,68,
69–99. http://dx.doi.org/10.1515/ijsl.1987.68.69
Gluszek, A., & Dovidio, J. F. (2010). The way they speak: A
social psychological perspective on the stigma of non-
L. E. C. Souza, et al. Legitimization of accent on discrimination
European Journal of Social Psychology 46 (2016) 609–620 Copyright © 2016 John Wiley & Sons, Ltd. 619
native accents in communication. Personality and Social Psy-
chology Review,14,214–237. http://dx.doi.org/10.1177/
1088868309359288
Grondelaers, S., van Hout, R., & Steegs, M. (2010). Evaluat-
ing regional accent variation in standard Dutch. Journal of
Language and Social Psychology,29,101–116. http://dx.doi.
org/10.1177/0261927X09351681
Hayes, A. F. (2013). Introduction to mediation, moderation, and
conditional process analysis: A regression-based approach.New
York: Guilford Publications.
Hosoda, M., & Stone-Romero, E. (2010). The effects of for-
eign accents on employment-related decisions. Journal of
Managerial Psychology,25,113–132. http://dx.doi.org/
10.1108/02683941011019339
Hosoda, M., Stone-Romero, E. F., & Walter, J. N. (2007). Lis-
teners’cognitive and affective reactions to English speakers
with standard American English and Asian accents. Percep-
tual and Motor Skills,104,307–326. http://dx.doi.org/
10.2466/pms.104.1.307-326
Huang, L., Frideger, M., & Pearce, J. L. (2013). Political skill:
Explaining the effects of nonnative accent on managerial
hiring and entrepreneurial investment decisions. Journal
of Applied Psychology,98,1005–1017. http://dx.doi.org/
10.1037/a0034125
Judd, C. M., Kenny, D. A., & McClelland, G. H. (2001). Esti-
mating and testing mediation and moderation in within-
subject designs. Psychological Methods,6,115–134. http://
dx.doi.org/10.1037/1082-989X.6.2.115
Kinzler, K. D., Dupoux, E., & Spelke, E. S. (2007). The native
language of social cognition. Proceedings of the National
Academy of Science,104,12,577–12,580. http://dx.doi.org/
10.1073/pnas.0705345104
Ko, S. J., Judd, C. M., & Blair, I. V. (2006). What the voice re-
veals: Within- and between-category stereotyping on the
basis of voice. Personality and Social Psychology Bulletin,32,
806–819. http://dx.doi.org/10.1177/0146167206286627
Lee, T. L., & Fiske, S. T. (2006). Not an outgroup, not yet an
ingroup: Immigrants in the Stereotype Content Model. In-
ternational Journal of Intercultural Relations,30,751–768.
http://dx.doi.org/10.1016/j.ijintrel.2006.06.005
Lev-Ari, S., & Keysar, B. (2010). Why don’t we believe non-
native speakers? The influence of accent on credibility.
Journal of Experimental Social Psychology,46,1093–1096.
http://dx.doi.org/10.1016/j.jesp.2010.05.025
Lima-Nunes, A., Pereira, C. R., & Correia, I. (2013).
Restricting the scope of justice to justify discrimination:
The role played by justice perceptions in discrimination
against immigrants. European Journal of Social Psychology,
43,627–636. http://dx.doi.org/10.1002/ejsp.1981
Lima-Nunes, A. V. (2013). Justice seems not to be for all: the
role played by justice perceptions in discrimination against
immigrants. (Doctoral dissertation). ISCTE-IUL, Lisbon,
Portugal. Retrieved from http://hdl.handle.net/10071/
6925
Lindemann, S. (2003). Koreans, Chinese or Indians? Atti-
tudes and ideologies about nonnative English speakers in
the United States. Journal of Sociolinguistics,7,348–364.
http://dx.doi.org/10.1111/1467-9481.00228
Lindemann, S. (2005). Who speaks “broken English”?
US undergraduates’perception of nonnative English.
International Journal of Applied Linguistics,15,187–212.
http://dx.doi.org/10.1111/j.1473-4192.2005.00087.x
Lippi-Green, R. (1997). English with an accent. London:
Routledge.
Moyer, A. (2004). Age, accents, and experience in second language
acquisition: An integrated approach to critical period inquiry.
Clevedon, UK: Multilingual Matters.
Ng, S. H. (2007). Language-based discrimination:Blatantand
subtle forms. Journal of Language and Social Psychology,26,
106–122. http://dx.doi.org/10.1177/0261927X07300074
Operario, D., & Fiske, S. T. (2001). Stereotypes: Content,
structures, process, and context. In R. Brown & S.
Gaertner (Eds.), Blackwell handbook of social psychology:
Intergroup process (pp. 22–44). Malden, MA: Blackwell
Publishers Ltd.
Park, B., & Judd, C. M. (2005). Rethinking the link
between categorization and prejudice within the so-
cial cognition perspective. Personality and Social Psy-
chology Review,9,108–130. http://dx.doi.org/10.1207/
s15327957pspr0902_2
Pereira, C., Vala, J., & Costa-Lopes, R. (2010). From prejudice
to discrimination: The legitimizing role of perceived threat
in discrimination against immigrants. European Journal of
Social Psychology,40, 1231–1250. http://dx.doi.org/10.1002/
ejsp.718
Pereira, C. R., Vala, J., & Leyens, J. P. (2009). From infra-
humanization to discrimination: The mediation of sym-
bolic threat needs egalitarian norms. Journal of Experimental
Social Psychology,45,336–344. http://dx.doi.org/10.1016/j.
jesp.2008.10.010
Pettigrew, T. F., & Meertens, R. W. (1995). Subtle and bla-
tant prejudice in Western Europe. European Journal of
Social Psychology,25,57–75. http://dx.doi.org/10.1002/
ejsp.2420250106
Rakić, T., Steffens, M. C., & Mummendey, A. (2011). Blinded
by the accent! The minor role of looksin ethniccategoriza-
tion. Journal of Personality and Social Psychology,100,16–29.
http://dx.doi.org/10.1037/a0021522
Ruscher, J. B. (2001). The social psychology of prejudiced commu-
nication. New York: Guilford Press.
Sidanius, J., & Pratto, F. (1999). Social dominance. New York:
Cambridge University Press.
Tajfel, H. (1984). Intergroup relations, social mythsand social
justice in social psychology. In H. Tajfel (Ed.), The social di-
mension (Vol. 2, pp. 695–715). Cambridge: Cambridge Uni-
versity Press.
Vala, J., Lopes, D., & Lima, M. (2008). Black immigrants in
Portugal: Luso-tropicalism and prejudice. Journal of Social
Issues,64,287–302. http://dx.doi.org/10.1111/j.1540-
4560.2008.00562.x
Yzerbyt, V., Provost, V., & Corneille, O. (2005). Not competent
but warm…Really? Compensatory stereotypes in the
French-speaking World. Group Process Intergroup Relations,
8, 291–308. http://dx.doi.org/10.1177/1368430205053944
Yzerbyt,V.Y.,Rocher,S.J.,&Schadron,G.(1997).Stereo-
types as explanations: A subjective essentialistic view of
group perception. In R. Spears, P. J. Oakes, N. Ellemers, &
S. A. Haslam (Eds.), The social psychology of stereotyping and
group life, (Vol. 8, pp. 20–50). Malden: Blackwell Publishing.
L. E. C. Souza, et al.Legitimization of accent on discrimination
European Journal of Social Psychology 46 (2016) 609–620 Copyright © 2016 John Wiley & Sons, Ltd.620
