Bilingualism: Language and Do native and non-native speakers make

Cognition
different judicial decisions?
cambridge.org/bil
Marie-Christine Rühle and Shiri Lev-Ari
Psychology Department, Royal Holloway, University of London, Egham, United Kingdom

Research Article Abstract

Cite this article: Rühle, M-C, & Lev-Ari, S. Bilinguals experience diminished emotion when using their foreign compared with
(2024). Do native and non-native speakers their native language. The diminished emotion has been shown to lead to more lenient
make different judicial decisions? Bilingualism:
moral evaluations in a foreign language. Here we show that non-native speakers of English
Language and Cognition, 1–8. https://doi.org/
10.1017/S136672892400018X are less sensitive to emotional mitigating circumstances of a crime than native speakers, pre-
sumably because of the diminished experience emotion. This can lead non-native speakers to
Received: 20 September 2023 provide harsher, rather than more lenient, evaluations. Native and non-native speakers of
Revised: 12 February 2024
English recommended sentence duration for crimes committed because of mitigating emo-
Accepted: 12 February 2024
tional circumstances (e.g., fraud to pay spouse’s medical treatment) or for selfish reasons
Keywords: (e.g., buying luxury goods). Native English speakers differentiated more between the two
foreign language effect; bilingualism; moral types of scenarios than non-native speakers did. The study thus provides preliminary evidence
evaluations that processing information in a foreign language can influence decisions, and that the direc-
Corresponding author: tionality of the effect depends on the role of emotion in the context.
Shiri Lev-Ari;
Email: [email protected]

Introduction
Many would agree that stealing money to pay for expensive medicine for a loved one is more
morally acceptable than stealing money to support an expensive life-style. Similarly, many
might find an assault less deplorable if the attacker assaulted an individual after witnessing
them taking advantage of a vulnerable close one. Indeed, such situations are often considered
as mitigating circumstances for a crime and can lead to reduced sentences. This paper tests
whether non-native speakers and native speakers judge crimes committed under such circum-
stances differently, such that native speakers are more swayed by the emotional mitigating cir-
cumstances than the non-native speakers.

Emotion in a foreign vs a native language

Bilinguals experience diminished emotional response in their second language. They rate emo-
tional phrases in their native language as more emotional than their translation equivalents
(Caldwell-Harris & Ayçiçeği-Dinn, 2009; Dewaele, 2004; Puntoni et al., 2009). They also
exhibit semantic priming in both their native and foreign language but they seem to exhibit
affective priming only in their native language, at least in cases where they do not use their
foreign language as often (Degner et al., 2011). Reduced responses to emotional words are
also evident in an attentional blink task: while Chinese–English bilinguals and native
English speakers exhibit similar attentional blink in English following neutral distractors,
the native English speakers exhibit a larger attentional blink than the Chinese–English bilin-
guals following English taboo/sexual distractors (Colbeck & Bowers, 2012). The reduced emo-
tional response in a foreign language can also be observed physiologically. Bilinguals exhibit
reduced galvanic skin response (Caldwell-Harris & Ayçiçeği-Dinn, 2009; Jankowiak &
Korpal, 2018) and smaller pupil dilation (Yao et al., 2023) in response to reprimands and
taboo words in their foreign vs their native language.
The reduced emotion that bilinguals experience in a foreign language leads bilinguals to
greater self-disclosure in a foreign language. For example, Bond and Lai (2001) had
© The Author(s), 2024. Published by Cantonese–English bilinguals interview each other in either Cantonese or English about
Cambridge University Press. This is an Open embarrassing topics, such as sexual attitudes of Chinese and Western individuals, and
Access article, distributed under the terms of
non-embarrassing topics, such as comparison of the educational systems in mainland China
the Creative Commons Attribution licence
(http://creativecommons.org/licenses/by/4.0/), and Hong-Kong. The researchers measured the time participants spent discussing each
which permits unrestricted re-use, distribution topic, excluding pauses. Notably, when the interviews were conducted in English, the partici-
and reproduction, provided the original article pants’ foreign language, they allocated a greater proportion of their time to discussing the
is properly cited. embarrassing topics (Bond & Lai, 2001). The reduced intensity of emotion also leads bilinguals
to report feeling less as though they are being untruthful when telling a lie in a foreign

This article has earned badges for transparent research practices: Open Data. For details see the Data Availability
Statement.

https://doi.org/10.1017/S136672892400018X Published online by Cambridge University Press

 2 Marie‐Christine Rühle and Shiri Lev‐Ari

language compared to their native language (Caldwell-Harris & are more accepting of relatively harmless violations, such as
Ayçiçeği-Dinn, 2009), and to be more likely to free-ride when consensual incest between adult siblings that does not lead to
using a foreign rather than a native language (Urbig et al., 2016). pregnancy or eating the meat of your deceased dog, when
responding in their foreign language than their native one.
The robustness of the influence of language on moral decision
Decision making in a foreign vs native language
making has been verified in a couple of meta-analyses (Circi
The lower intensity of emotion that bilinguals experience in their et al., 2021; Del Maschio et al., 2022).
second language leads them to make different decisions in their
two languages. For example, people tend to avoid taking bets,
even when these are favourable, because of loss aversion. The
Mechanism underlying the foreign language effect
reduced emotionality in a foreign language reduces loss aversion,
increasing bilinguals’ willingness to take such favourable bets While there is robust evidence for the difference in decision
(Keysar et al., 2012). The reduced emotion in a foreign language making in bilinguals’ foreign vs native language, there is no con-
also increases bilinguals’ willingness to consume aversive but sus- sensus with regards to the underlying cause of it. The most com-
tainable products, such as recycled water and cookies made of mon account attributes the difference in decisions to
insects (Geipel et al., 2018), and it reduces framing effects that experiencing diminished emotion in a foreign language than a
are induced by loss aversion or risk seeking, as in the Asian native one. As mentioned earlier, bilinguals’ diminished emo-
Disease problem (Keysar et al., 2012). tion in their foreign languages has been documented in many
The influence of diminished emotion in a foreign language has studies spanning self reports, behavioral measures, and physio-
been most extensively studied with regards to moral dilemmas. logical measures (e.g., Caldwell-Harris & Ayçiçeği-Dinn, 2009;
Most studies presented participants with the footbridge dilemma. Colbeck & Bowers, 2012; Degner et al., 2011; Dewaele, 2004;
In that dilemma, participants are asked to imagine that they are Jankowiak & Korpal, 2018; Puntoni et al., 2009; Yao et al.,
standing on top of a bridge. They see a trolley coming. If nothing 2023). The account that attributes differences in decision mak-
is done, the trolley will hit five workers that work on the tracks. ing across languages to diminished emotion can also account
They are told that the only way to save these five workers is to for the effect of language on responses to the more direct/per-
push off the bridge a large man standing next to them. The trolley sonal version of moral dilemmas but not the indirect/impersonal
will then kill him but the five workers will be saved. Participants ones. The diminished emotion has also been linked to reduced
are then asked whether they would push the person. Multiple mental imagery in a foreign vs a native language, which has
studies found that bilinguals are more willing to push the person indeed been found to moderate the foreign language effect
when performing the task in a foreign vs a native language (Hayakawa & Keysar, 2018). But why do bilinguals experience
(Brouwer, 2019; Cipolletti et al., 2016; Costa et al., 2014; diminished emotion in their foreign language? Several non-
Dylman & Champoux-Larsson, 2020; Geipel et al., 2015b), at mutually exclusive and often correlated factors have been pro-
least when the languages are dissimilar enough and differ in posed, including differences in the context of language learning
amount of use. or language use – natural for the native language vs school-
Some of the studies specifically contrasted the footbridge setting for foreign languages (e.g., Bond & Lai, 2001; Dewaele,
dilemma with a less emotional version of the dilemma, in 2004; Harris et al., 2006; Sheikh & Titone, 2016), differences
which the trolley can be diverted with a switch. Pulling that in age of acquisition (Harris et al., 2006), differences in profi-
switch would lead the trolley to a different track where it ciency (e.g., Degner et al., 2011; Dewaele, 2004; Eilola &
would kill one worker rather than five. Even though in both ver- Havelka, 2011), differences in amount of use (Dewaele, 2004),
sions one innocent person is sacrificed to save five, people are and differences in accessibility of cultural norms (Geipel et al.,
more willing to pull the switch than to push a person as the 2015a). In line with these explanations, some studies found
less direct nature of the act of sacrifice renders it less emotional. that the foreign language effect is absent in bilinguals who are
Correspondingly, studies that contrasted the two dilemmas highly proficient and well acculturated in both their languages
show that responding to the dilemma in a foreign language (Degner et al., 2011; Dylman & Champoux-Larsson, 2020) and
influences responses in the emotional version of the dilemma a meta-analysis showed that language similarity moderates the
(pushing a man from a footbridge) but not in the less emotional foreign language effect (Circi et al., 2021).
version of the dilemma (pulling a switch; Cipolletti et al., 2016; While evidence exists for all these predictors, it is often
Costa et al., 2014; Geipel et al., 2015b). This pattern of an effect mixed. For example, while there’s evidence from individual stud-
of language in dilemmas that require direct personal acts and ies for the moderating effect of proficiency on the foreign lan-
lack of an influence in more indirect impersonal dilemmas guage effect (e.g., Degner et al., 2011; Dewaele, 2004; Eilola &
also extends to similar other moral dilemmas, such as suffocat- Havelka, 2011), both meta-analyses that were conducted on
ing your crying baby to avoid discovery of multiple people by the foreign language effect did not find proficiency to moderate
enemy soldiers vs pressing a switch to divert deadly fumes the effect of language on moral decision making (Circi et al.,
from a hospital room with three patients to a hospital room 2021; Del Maschio et al., 2022), though they examined profi-
with only one patient (Shin & Kim, 2017). One study also pro- ciency at the experiment level and not individual differences
vides suggestive evidence that bilinguals are not only more will- within a sample, potentially reducing the sensitivity of the ana-
ing to sacrifice others, but are also more willing to sacrifice lyses. It should be noted though that while differences in emo-
themselves in order to save others when responding in a foreign tionality and differences in moral judgments are both robust,
vs native language (Romero-Rivas et al., 2022). Bilinguals direct tests of the link between emotionality and moral judgment
exhibit greater acceptance of moral violations when using a for- yielded mixed results – some found evidence for such a link
eign language even when these are not conducted to avoid (e.g., Geipel et al., 2015a) but others (e.g., Romero-Rivas et al.,
greater harms. Geipel et al. (2015a) have shown that bilinguals 2022) did not.

https://doi.org/10.1017/S136672892400018X Published online by Cambridge University Press

 Bilingualism: Language and Cognition 3

Current study the duration of the audio file, indicating they did not listen to
the story in full. This left 655 observations from 190 participants2
Regardless of the precise mechanism underlying the foreign lan-
(F = 109, M = 73, Non-binary = 5, Other = 3). Participants’ age
guage effect, its implications are likely to be far reaching. One
ranged from 18 to 80 (M = 36, SD = 17.3). Sixty of the participants
domain where its influence might be particularly strong is judicial
were native speakers of British English and 130 were speakers of
decisions. As mentioned earlier, people judge relatively harmless
English as a foreign language, mostly native speakers of German
moral violations more leniently in their foreign than native lan-
(N = 123). Of the 130 speakers of English as a foreign language,
guage (Geipel et al., 2015a). Woumans et al. (2020) showed that
86 have never resided in an English-speaking country, 25 have
bilinguals also judge real homicide cases less severely in their for-
resided in an English-speaking country for under a year,
eign language than their native one. If the foreign language effect
and the rest resided in an English-speaking country for one
is a consequence of reduced emotionality in a foreign language,
year (N = 11), two years (N = 4), or three years (N = 2).
then it should not influence the perceived severity of all crimes
equally. Instead, it should influence the perceived severity of
Stimuli
crimes with emotional motivations and consequences more
Four crime scenarios were generated with two versions for each
than those where the offense was neither prompted by nor
scenario – one where the perpetrator acted out of strong emo-
induced great emotion.
tional distress and one where the circumstances were less emo-
Furthermore, diminished emotion might not always lead to
tional. The scenarios were written as statements from the
more lenient judgments in a foreign language. In some cases, it
perpetrator in which they narrated the sequence of events. In
could lead to reduced sensitivity to mitigating circumstances,
the emotional version of the scenario, the perpetrator acted out
thus resulting in harsher judgments. For example, people some-
of desperation or because they were faced with a difficult moral
times commit crimes when under great emotion that distorts
dilemma. In the non-emotional version of the scenarios, the
their judgment. Such emotional circumstances are sometimes
defendant acted out of selfish reasons and could have decided
considered as mitigating circumstances and lead to reduced sen-
against committing the crime at any point without facing any
tences. This study tests whether non-native speakers are less influ-
negative consequences. For example, in the fraud scenario, the
enced than native speakers by the emotional circumstances that
perpetrator committed the crime in order to afford medical treat-
led to the crime when assigning punishment. Specifically, the
ment for his wife that is not covered by insurance, and that they
study tests whether native speakers assign shorter sentences to
could not have afforded otherwise. In the non-emotional version
defendants who committed a crime under strong emotional cir-
of the scenario, the perpetrator committed the crime in order to
cumstances compared with defendants who committed the
afford luxury goods. The two versions of each scenario were as
same crime without such circumstances whereas non-native
similar as possible and only differed in the section that described
speakers assign more similar sentences to the two types of defen-
the motivation (See Appendix for all scenarios).
dants. That is, the emotional mitigating circumstances should
The scenarios were presented auditorily as a prior study sug-
influence the sentencing decisions of native speakers more than
gested that auditory presentation can enhance the foreign lan-
those of non-native speakers. If so, the findings will have both
guage effect, especially among speakers of typologically similar
practical implications and theoretical ones.
languages (Brouwer, 2021). Four amateur actors recorded the
scenarios. Recordings took place in a soundbooth at the
Study Psychology Department of the university. The actors were all
native speakers of English. All recordings were captioned to facili-
Method
tate comprehension. The scenarios were presented as videos with
Ethical considerations a black screen with captions appearing phrase by phrase aligned
The authors assert that all procedures contributing to this work with the speaker’s speech.
comply with the ethical standards of the relevant national and After each scenario participants were provided with informa-
institutional committees on human experimentation and with tion about the maximal sentence for such a crime in the UK
the Helsinki Declaration of 1975, as revised in 2008. (e.g., “The maximum punishment in the UK for financial fraud
is 10 years’ custody”). Then participants were asked: “If you
Participants were in a position to decide, what prison sentence would you
We recruited native speakers of English and speakers of English as give to the defendant?”. The scale ranged from 0 to the maximal
a foreign language. We aimed for a sample size of at least 146 par- sentence. In three of the scenarios, the maximal sentence was 10
ticipants, in line with Keysar et al.’s (2012) Experiment 3 which years. In one case, it was 5 years.
had a similar design of Language (native, non-native) manipu-
lated between participants, and emotionality of the question Procedure
manipulated within participants. Our recruitment relied on social The study took place online on the survey platform Qualtrics
media so we recruited until a set deadline. By that date two- (https://www.qualtrics.com). Participants first provided demo-
hundred and thirty-one participants completed the study. graphic information, including their native language. If they indi-
Twenty-one participants were excluded for being native speakers cated that their native language was English, they were further
of a variety of English other than British English1. Eight partici- asked to indicate the precise variety of English they speak. If
pants in the non-native speaking group were excluded for residing they indicated a native language other than English, they were fur-
in an English-speaking country for longer than 3 years, as differ- ther asked how long they have resided in an English-speaking
ences in context and amount of use are argued to underlie the for- country, with the option ‘Never/0’ included. Participants next per-
eign language effect (Bond & Lai, 2001; Dewaele, 2004; Harris formed a short audio test.
et al., 2006; Sheikh & Titone, 2016). Lastly, 143 scenarios were Once participants completed the audio test, they were asked to
excluded because participants’ response time was shorter than imagine they were a jury member. They were informed that they

https://doi.org/10.1017/S136672892400018X Published online by Cambridge University Press

 4 Marie‐Christine Rühle and Shiri Lev‐Ari

would be presented with four court trials. In all cases, the defend- (β = 1.42, SE = 0.41, t = 3.50). As can be seen in Figure 1, the
ant was pleading guilty to the charges and their task was to decide interaction indicates that the effect of Emotion is larger for native
the length of the defendant’s sentence. They then heard four speakers than non-native speakers.
recordings, one of each scenario, in random order. Each partici-
pant listened to two emotional and two non-emotional scenarios.
Participants could take as long as they needed to listen to the General discussion
recordings, including pausing and replaying them. During play-
back, captions of the recordings appeared, one line at a time to We tend to treat language as a vehicle to communicate a message.
facilitate comprehension. We assume that as long as the addressee understood the message,
For each scenario, participants indicated their recommended sen- it does not make a difference whether the message was commu-
tence for the defendant (in years) by moving a slider on a scale from nicated in the addressee’s native or foreign language. A growing
0 to the maximum penalty for this kind of crime in the UK. body of literature, however, indicates that this is not the case.
That literature shows that bilinguals experience greater intensity
of emotion when processing content in their native language vs
Results
their foreign language, at least in cases where the two languages
All data and the analysis script are available at https://osf.io/8xtu6/ differ in age of acquisition and context of learning and use.
?view_only=b63f710b742e4354b87509ec1865d093. Furthermore, the diminished emotion also leads bilinguals to
Before analysing the data, we transformed the dependent vari- make different decisions and draw different judgments in their
able to make it comparable across scenarios. As mentioned earl- two languages. This study provides further support for the influ-
ier, the maximal sentence for one of the scenarios was 5 years ence of diminished emotion on moral judgment, while focusing
whereas the maximal sentence for the other scenarios was 10 on a situation that could have real world implications. The
years. Therefore, before running the analysis, responses to the study shows that non-native speakers are less influenced by emo-
scenario with a maximal sentence of 5 years were doubled to tional mitigating circumstances when allocating punishments for
make them comparable to the responses to the other scenarios3. a crime. These results could suggest that bilinguals’ evaluations of
We ran a mixed effects regression analysis with Language crime could be influenced by the language in which these crimes
(Native, Non-native; reference level=Non-native), Emotion are described. While prior studies found that bilinguals are more
(Emotional, Non-emotional; reference level = Emotional) and lenient when judging a crime in a foreign language (Woumans
their interaction as fixed effects and Participants and Scenarios et al., 2020), this study shows that the effect of language might
as random effects. The random structure also included influence judgments in both directions, depending on whether
by-participant and by-scenario slopes for Emotion4. We did not and how emotion is involved in the motivations and conse-
include a by-scenario slope for Language as that slope led the quences of the crime.
model to fail to converge. Results revealed an effect of Emotion That said, an examination of the pattern of results revealed
(β = 1.20, SE = 0.53, t = 2.26) at the reference level (Non-native), that native and non-native participants did not differ in the pun-
and crucially, an interaction between Language and Emotion ishment they assigned to emotional cases (there was no effect of

Figure 1. Assigned sentence duration as predicted by whether participants responded in their native or non-native language and scenario’s emotionality. Black
dots indicate condition means.

https://doi.org/10.1017/S136672892400018X Published online by Cambridge University Press

 Bilingualism: Language and Cognition 5

Language at the reference level, which is emotional scenarios), task in German to reduce potential effects of cultural differences.
whereas native and non-native speakers differed in the sentences Importantly, the study tested for the existence of an interaction.
they assigned in the non-emotional cases. It is important to note That is, rather than testing whether one group gives more lenient
that the pattern of results does not seem to be driven by floor or harsher punishment, it compared the sensitivity of each group
effects in the sentencing decisions in the emotional scenarios. to different types of scenarios. The presence of emotional mitigat-
While they were lower than the sentences assigned in the ing circumstances was manipulated within participants, thus
non-emotional cases, they averaged over 2.5 years, and so could allowing for the test of the effect of language while controlling
have allowed for a further drop. The pattern of results, then, for baseline group differences. Nevertheless, there remains the
seems to reflect baseline differences between the groups. What possibility that members of different cultures differ in their sensi-
could explain the baseline differences? Several options present tivity to emotional mitigating circumstances. Future research
themselves. One might be due to cultural differences. German should control for this possibility.
and British participants might view the appropriateness of prison It is also worth noting that the native and non-native partici-
sentences differently. Alternatively, the results might reflect two pants in our study responded differently even though the native
distinct foreign language effects working in tandem: general per- language of most of our non-native speaking participants
ception of crimes less severely in a foreign language, as was also – German, is highly similar to English (both are Germanic lan-
reported in Woumans et al. (2020), and reduced sensitivity to guages), and foreign language effects are often reduced with
emotional circumstances in a foreign language. Together, these such pairs of languages (e.g., Circi et al., 2021). Our study thus
effects would yield the observed pattern of results: In the might have under-estimated the size of the foreign language effect.
non-emotional scenarios, the diminished emotional response to Another limitation of the study is the fact that participants’
the act and its outcome led non-native speakers to judge the comprehension was not tested. While we cannot rule out the pos-
crimes more leniently (note that in these scenarios the act and sibility that the reduced sensitivity to emotional mitigating cir-
its outcome can induce emotion; these scenarios were labelled cumstances is due to lack of comprehension of these
non-emotional because they were not prompted by emotional circumstances, participants’ general behavior does not suggest
mitigating circumstances). In the case of the emotional scenarios, that. The non-native speakers in our study were not more likely
non-native speakers experienced reduced emotional response to to provide responses in the middle of the scale. Furthermore,
the crime and its outcome, similarly to the case of non-emotional the native and non-native speaking participants were also similar
scenarios, which led to more lenient judgments. At the same time, in their ranking of the severity of the scenarios. For example,
their diminished emotion also led them to be less influenced by when considering the non-emotional versions of the scenarios
the lenient mitigating circumstances, which led them to assign (where emotionality should not lead to group differences), both
harsher sentences than they would have had they experienced groups assigned the longest sentence to the darknet scenario
an emotional reaction to the mitigating circumstances. and the shortest sentence to the grandparent scenario (see
Together, the two effects of diminished emotion cancelled each Appendix for the scenarios) suggesting comprehension of the
other leading to similar judgments to native speakers. depicted crimes. The similar ranking of the crimes also suggests
One interesting alternative interpretation of the results of the that participants in the two groups possess similar values when
study is that the non-native speakers were less sensitive to the evaluating such crimes.
mitigating circumstances not because of diminished emotion in Together, the study provides further support for the foreign
a foreign language, but because of reduced attention to intentions language effect. Furthermore, it suggests that differences in
compared with consequences when processing a foreign language. experienced emotion in a native vs a foreign language can influ-
Geipel et al.’s (2016) presented participants with acts prompted ence judicial decisions. Unlike prior literature, the study shows
by good intentions that had negative consequences (e.g., a person that diminished emotion might not influence judgment of all
giving a poor boy money, but then the boy uses it to buy drugs crimes similarly and its effect might be stronger for some scen-
and dies of an overdose) and with acts prompted by dubious arios than others. It also shows that diminished emotion does
intentions that had positive outcomes (e.g., a company donating not always lead to more lenient sentences, as it can lead to
money to increase profits). Bilinguals in that study rated the for- reduced sensitivity to mitigating circumstances, thus resulting in
mer acts more negatively and the latter more positively when harsher judgment than had they been taken into account.
evaluating them in their foreign vs their native language (Geipel While the study presented participants with crime scenarios,
et al., 2016). The non-native participants in our study, therefore, the results are likely to apply to judgments of behavior of friends
might have similarly focused on the outcome more than the and colleagues as well. Native and non-native speakers might
intention, reducing their sensitivity to the mitigating circum- judge impolite or transgressive acts such as hurtful comments
stances. Such an account, thus, similarly predicts that individuals or ungenerous acts differently, and might differ in how under-
would be less sensitive to emotional mitigating circumstances in standing they would be to the emotional turmoil that prompted
their foreign language, but it also suggests that non-native speak- the behavior. The influence of the foreign language effect on
ers might be less sensitive to other, non-emotional, mitigating cir- social behavior might therefore be pervasive and complex.
cumstances as well. Future research should compare scenarios
with different types of mitigating circumstances to better under- Acknowledgement. The research was funded by a Leverhulme grant
awarded to S.L. (RPG-2022-270). We would like to thank Niki Agrios,
stand the extent of the effect and the underlying mechanism.
Samuel Clark, Matthew Davies, and Elyas Yusuf for recording the stimuli.
Several limitations of the study should be taken into account.
The study compares two different groups of participants – native
speakers of British English and (mostly) German–English bilin-
guals. That is, participants were not randomly assigned to a lan- Notes
guage condition. It would be good to repeat the study with native 1
As the stimuli were presented auditorily, we worried that processing the
German speakers and English–German bilinguals performing the scenarios in a non-native accent would create distance and reduce emotion.

https://doi.org/10.1017/S136672892400018X Published online by Cambridge University Press

 6 Marie‐Christine Rühle and Shiri Lev‐Ari

We therefore restricted native participants to native speakers of the same var- Geipel, J., Hadjichristidis, C., & Surian, L. (2016). Foreign language affects the
iety as the recorded actors, namely, British English. contribution of intentions and outcomes to moral judgment. Cognition, 154,
2
We conducted a sensitivity power analysis after completing the study using 34–39.
the simr package (Green & McLeod, 2016) in R. Results indicated that a sam- Geipel, J., Hadjichristidis, C., & Klesse, A. K. (2018). Barriers to sustainable
ple of 190 participants provides >80% power to detect an effect as small as consumption attenuated by foreign language use. Nature Sustainability, 1,
β = 1.2. Our results yielded an effect with β=1.42, around 5 SEs larger than 1, 31–33.
this. It seems, then, that our study was sufficiently powered to detect an effect Green, P., & MacLeod, C. J. (2016). simr: an R package for power analysis of
of the size we found. generalised linear mixed models by simulation. Methods in Ecology and
3
To make sure that the transformation of the responses to one scenario did Evolution, 7, 4, 493–498. doi:10.1111/2041-210X.12504, https://CRAN.R-
not distort responses, we also ran an alternative analysis in which responses project.org/package=simr.
were normalized by scenario. In this analysis, the random variable Scenario Harris, C., Gleason, J., & Ayçiçeği, A. (2006). When is a First Language More
did not account for any variance and led to singular fit. It was therefore Emotional? Psycholinguistic Evidence from Bilinguals. In Bilingual Minds:
removed. The analysis therefore included Language (Native, Non-native), Emotional Experience, Expression, and Representation, edited by
Emotion (Emotional, Non-emotional) and their interaction as fixed effects, A. Pavlenko, 257–284. Clevedon: Multilingual Matters.
Participants as a random effect, and a by-participant slope for Emotion. Hayakawa, S., & Keysar, B. (2018). Using a foreign language reduces mental
Results were comparable to those of the main analysis: An effect of Emotion imagery. Cognition, 173, 8–15.
(β = 0.44, SE = 0.08, t = 5.52) at the reference level (Non-native) and an inter- Jankowiak, K., & Korpal, P. (2018). On modality effects in bilingual emotional
action between Language and Emotion (β = 0.52, SE = 0.15, t = 3.49) indicating language processing: Evidence from galvanic skin response. Journal of
that the effect of Emotion is larger for native speakers than non-native Psycholinguistic Research, 47, 663–677.
speakers. Keysar, B., Hayakawa, S. L., & An, S. G. (2012). The foreign-language effect:
4
The model was: Thinking in a foreign tongue reduces decision biases. Psychological science,
Punishment ∼ Language*Emotion + (1+Emotion|Participant) 23, 6, 661–668.
+ (1+Emotion|Scenario) Puntoni, S., De Langhe, B., & Van Osselaer, S. M. J. (2009). Bilingualism and
the Emotional Intensity of Advertising Language. Journal of Consumer
Competing interest. The authors declare none Research, 35, 6, 1012–1025.
Romero-Rivas, C., López-Benítez, R., & Rodríguez-Cuadrado, S. (2022).
Would you sacrifice yourself to save five lives? Processing a foreign language
References increases the odds of self-sacrifice in moral dilemmas. Psychological
Bond, M. H., & Lai, T. (2001). Embarrassment and code switching into a Reports, 125, 1, 498–516.
second language. The Journal of Social Psychology, 126, 2, 179–186. Sheikh, N. A., & Titone, D. (2016). The embodiment of emotional words in a
Brouwer, S. (2019). The auditory foreign-language effect of moral decision second language: An eye-movement study. Cognition & Emotion, 30, 3,
making in highly proficient bilinguals, Journal of Multilingual and 488–500. https://doi.org/10.1080/02699931.2015.1018144
Multicultural Development, 40, 10, 865–878, https://doi.org/10.1080/ Shin, H. I., & Kim, J. (2017). Foreign language effect and psychological dis-
01434632.2019.1585863 tance. Journal of Psycholinguistic Research, 46, 1339–1352.
Brouwer, S. (2021). The interplay between emotion and modality in the Urbig, D., Terjesen, S., Procher, V., Muehlfeld, K., & Van Witteloostuijn, A.
Foreign-Language effect on moral decision making. Bilingualism: (2016). Come on and Take a Free Ride: Contributing to Public Goods in
Language and Cognition, 24, 2, 223–230. Native and Foreign Language Settings. Academy of Management Learning
Caldwell-Harris, C. L., & Ayçiçeği-Dinn, A. (2009). Emotion and lying in a & Education 15, 2, 268–286.
non-native language. International Journal of Psychophysiology, 71, 3, Woumans, E., Van der Cruyssen, I., & Duyck, W. (2020). Crime and punish-
193–204. ment: Morality judgment in a foreign language. Journal of Experimental
Cipolletti, H., McFarlane, S., & Weissglass, C. (2016). The Moral Psychology: General, 149, 8, 1597.
Foreign-Language Effect. Philosophical Psychology, 29, 1, 23–40. Yao, Y., Connell, K., & Politzer-Ahles, S. (2023). Hearing emotion in two lan-
Circi, R., Gatti, D., Russo, V., & Vecchi, T. (2021). The foreign language effect guages: A pupillometry study of Cantonese–Mandarin bilinguals’ percep-
on decision-making: A meta-analysis. Psychonomic Bulletin & Review, 1–11. tion of affective cognates in L1 and L2. Bilingualism: Language and
Colbeck, K. L., & Bowers, J. S. (2012). Blinded by taboo words in L1 but not Cognition, 1–14.
L2. Emotion, 12, 2, 217.
Costa, A., Foucart, A., Hayakawa, S., Aparici, M., Apesteguia, J., Heafner, J., &
Keysar, B. (2014). Your morals depend on language. PloS one, 9, 4, e94842. Appendix – Scenarios
Degner, J., Doycheva, C., & Wentura, D. (2011). It matters how much you talk:
Grandson scenario, emotional version
On the automaticity of affective connotations of first and second language
words. Bilingualism: Language and Cognition, 15, 181–189. You see, my grandpa hasn’t been well for quite some time. He’s in his eighties
Del Maschio, N., Crespi, F., Peressotti, F., Abutalebi, J., & Sulpizio, S. (2022). now and he’s had Alzheimer’s for many years. He is not aggressive or any-
Decision-making depends on language: A meta-analysis of the Foreign thing, just very confused, but still very lovely and trusting. I still visit him
Language Effect. Bilingualism: Language and Cognition, 25, 4, 617–630. every week, even though I live ages away now. I’m still responsible for his
Dewaele, J. M. (2004). The Emotional Force of Swearwords and Taboo Words insurance, finances and everything, I also have access to his bank account,
in the Speech of Multilinguals. Journal of Multilingual and Multicultural but I don’t touch his money.
Development, 25, 204–222. When he was younger, he always told us he didn’t want to end up in an old
Dylman, A. S., & Champoux-Larsson, M. F. (2020). It’s (not) all Greek to me: people’s home, so when grandma passed away, my parents took him to their
Boundaries of the foreign language effect. Cognition, 196, 104148. place. He’s living in my old room now. But caring for him is very time-
Eilola, T. M., & Havelka, J. (2011). Behavioural and physiological responses to consuming and exhausting, I guess – on the bad days, he can’t even go to
the emotional and taboo Stroop tasks in native and non-native speakers of the bathroom by himself. Since I’m not living at home anymore and both of
English. International Journal of Bilingualism, 15, 3, 353–369. my parents still work full-time, they hired some guy who was a trained health-
Geipel, J., Hadjichristidis, C., & Surian, L. (2015a). How Foreign Language care worker to be at our house during the weekdays and care for him and keep
Shapes Moral Judgment. Journal of Experimental Social Psychology, 59, 8– him company. My parents did background checks and talked to the facility he
17. had worked at before, everything seemed cool. It went well for a couple of
Geipel, J., Hadjichristidis, C., & Surian, L. (2015b). The foreign language effect months, but one day, grandpa’s bank called me and informed me that there
on moral judgment: The role of emotions and norms. PloS one, 10, 7, was suspicious activity going on in his bank account. Over the past few
e0131529. months, someone had regularly withdrawn large sums of money from it,

https://doi.org/10.1017/S136672892400018X Published online by Cambridge University Press

 Bilingualism: Language and Cognition 7

over 3,000 pounds in total. At first, I thought someone had stolen his credit new treatment that gave us some hope. The treatment, however, was very
card, but when I asked grandpa about it, he showed me his card. He was expensive, and our insurance wouldn’t cover it. We’re not poor, I work as
too far gone to understand what I was making such a fuss about, but he men- an accountant after all, but also not super-rich. There was just no way to
tioned that “nice young man” recently taking him to the ATM to pay for pay for the treatment. But time was running away, and we were getting
groceries. more and more desperate. It was clear that the basic medical treatment that
I know that the responsible thing to do would have been to call the police was covered by the insurance would not save her life and only prolong her suf-
immediately. I know I wouldn’t have to make this statement otherwise. But I fering, while there was this thing right there that could actually save her if we
was just too upset at that moment. I left the house, but instead of driving could just afford it. I just could not sit still and do nothing while my wife was
home, I lingered around the street and waited for the care worker. He arrived wasting away, I just couldn’t.
shortly after. I wanted to go after him there and then, but there was still some One day at work, I discovered this loophole in our IT system almost by
doubt in my mind about whether this was all just a misunderstanding, so I accident. I transferred a two-digit sum of money into my own bank account,
waited for a while. It didn’t even take half an hour before I saw both of just to see whether it was possible, and I was baffled by the fact that it didn’t
them leave the house together, my grandpa holding onto the man’s arm, have any consequences, and then, something in me started to wonder whether
who took him to the ATM under some made-up story to withdraw another this could actually be a sign from above. Instead of reporting it, I did it again,
three-digit sum of my grandpa’s money. I just lost it then. I walked towards and then, gradually, I began to do what I was eventually arrested for: It started
them – the care worker recognized me, let go of my grandpa’s arm, and turned small at first, just changing a few numbers here and there. But it soon turned
around – probably to run away – and then I just went at him. I think I was into something bigger. I soon realized that it was possible to funnel large sums
scared that he would just disappear with the 3,000 pounds he had already sto- of my company’s profits into my bank account without anyone noticing, and
len from grandpa, and I was also really, really angry. I must have hit him three so I did exactly this when the next medical bill arrived for Helen, and the next
or four times, I don’t remember. I do remember him lying on the ground at one after that. I justified it with the thought that paying for Helen’s treatment
some point, with his face bleeding. I’m not sure I would do it again, but I was, from an ethical point of view, more important than increasing the com-
also don’t regret it. pany’s worth. When I was caught after a year or so, I had funneled several
thousand pounds into my own bank account. It has caused damage to the
company and my co-workers, I know that. I regret it very much. I didn’t
Grandson scenario, non-emotional version know what else to do at the time. I just couldn’t bear the thought of watching
her suffer without being able to do a single thing about it. All the money
You see, my grandpa hasn’t been well for quite some time. He’s in his eighties only ever went into paying for the treatment, I did not keep a single pound
now and he’s had Alzheimer’s for many years. He is not aggressive or any- to myself.
thing, just very confused, but still very lovely and trusting.
When he was younger, he always told us he didn’t want to end up in an old
people’s home, so when grandma passed away, my parents took him to their
Financial fraud scenario, non-emotional version
place. He’s living in my old room now. But caring for him is very time-
consuming and exhausting, I guess – on the bad days, he can’t even go to I work as an accountant. My wife and I are not poor, but it would be a huge
the bathroom by himself. Since I’m not living at home anymore and both of overstatement to say we’re rich. By the end of the month, we can pay all the
my parents still work full-time, they hired some guy who was a trained health- bills and treat ourselves occasionally, but it’s never enough for a big holiday,
care worker to be at their house during the weekdays and care for him and a new car, or retirement savings.
keep him company. My parents did background checks and talked to the facil- One day at work, I discovered this loophole in our IT system almost by
ity he had worked at before, everything seemed cool. I never liked him much, accident. I transferred a two-digit sum of money into my own bank account,
though. In my opinion, that care guy was obnoxious and resentful, and, to be just to see whether it was possible, and I was baffled by the fact that it didn’t
frank, just not a very nice person. He would constantly get into arguments with have any consequences. Instead of reporting it, I did it again, and then, grad-
my parents over money or his work schedule, get angry with my grandpa for ually, I began to do what I was eventually arrested for: It started small at first,
little things and was generally very unreliable. just changing a few numbers here and there. But it soon turned into something
So, two months ago, I was at my parents’ house, I was having a really bad bigger. I soon realized that it was possible to funnel large sums of my com-
day and I was super fed up with several things. My girlfriend had recently bro- pany’s profits into my bank account without anyone noticing. The first
ken up with me and things at my job weren’t going smoothly. When this guy thing I bought from that money was a necklace for my wife, then a new TV
came out of grandpa’s room, he didn’t even say Hi to me, he immediately for myself, then a leather couch. When my wife or my neighbours asked me
started ranting about my grandpa wetting himself again and about the extra how I got afford all this stuff, I told them I had been getting bonuses at
hours he had had to put in recently. He said that my grandpa was too work for my achievements. Their admiration and the social status those
much work and that if we wouldn’t pay him more, he would quit immediately. items bought me was even better than the items themselves. I justified it
I told him to talk to my parents about it but he said he wanted things to be with the thought that my wife and I were more deserving of that money
sorted right now. He was right at my face and I could smell his breath. I turned than the gigantic company. When I was caught after a year or so, I had fun-
away and started to walk away in order to not escalate the situation, but the neled thousands of pounds into my own bank account. It has caused damage
other guy just grabbed my arm to keep me listening and kept shouting at to the company and my coworkers, I know that. I regret it very much.
me. Tthat was when I just lost it – I turned around, and while turning around
I drew back my fist and landed a blow directly on his nose. He stumbled back-
ward, I hit him a second and a third time before I came to my senses. He was Policeman scenario, emotional version
bleeding from his face and I immediately felt bad about it. I’m not sure I would
do it again, but I also don’t regret it. I used to be a police officer in London and I really enjoyed my job. I know a lot
of people don’t like cops, but I swear I worked in this job to do something
good for the community, shield people from harm, and make the world a little
bit more just.
Financial fraud scenario, emotional version
About a year ago, we investigated the kidnapping of a young woman. It was
Helen and I have been together for over 30 years, I can’t even begin to tell you all over the news, she was the youngest member of a rich family, and the kid-
how much she means to me. She’s a sweet woman, very, very kind to everyone. napper demanded a ransom of 5 million pounds in order to let her go. Along
Two years ago, she came home and asked me whether I know how to use the with the family, we staged a ransom drop and succeeded in arresting the kid-
washing machine on my own. Turned out she had been diagnosed with napper. However, we still couldn’t find the kidnapped woman and he wouldn’t
Amyotrophic Lateral Sclerosis, a very nasty muscle disease that gets progres- tell us her whereabouts or anything else. This was the most difficult situation
sively worse until you die. There is no cure for that yet, but after visiting doz- of my entire career – assuming that she was probably still alive, but most likely
ens of doctors who said they couldn’t do much, we read about a promising locked away somewhere and possibly without access to water and food, it was a

https://doi.org/10.1017/S136672892400018X Published online by Cambridge University Press

 8 Marie‐Christine Rühle and Shiri Lev‐Ari

race against time. We questioned the suspect for two days, but he wouldn’t few times ever since, but from day one, he seemed to develop some sort of
talk. Pressure from the media and the victim’s family increased, and we obsession with me. It started with texts and little letters. I told him very
became desperate. I didn’t sleep at all during the two days, I was constantly early on that I wasn’t interested in him and I was nice about it, but that
thinking about that girl. On the third day, both my colleague and the suspect’s only made things worse. He started calling me at random hours, spamming
attorney left the room for a minute or two. I was not supposed to talk to him me on all social networks, and sometimes randomly showing up at my cam-
during this time, but I did. pus, even though he doesn’t study there. I didn’t think much about it at first,
I openly confess to what I did when I was alone with him: I told him that if but it became creepy very soon, and I started blocking him everywhere.
he didn’t tell us where the woman was, I would shoot him in the knee. When However, he would just make a new fake account or change his number to
he didn’t respond, I took my gun out of the holster and acted like I was loading keep messaging me. At first, he just wanted to go on a date with me, but
it, while verbally threatening him with the most intense pain he had ever felt in after I told him numerous times that I wanted him to leave me alone, his
his life, the possibility of never walking again, and how much worse life in texts became more and more threatening and disturbing.
prison would be for someone who couldn’t walk. He started crying, and After a few months, I reported him to the police, but they didn’t take me
under his breath, he finally told me where the girl was. We found her alive seriously. They said that as long as he didn’t send murder threats or physically
and physically well – thankfully, she had had access to food and water during assault me, he was no real danger to me and that they had more important
all this time. I was suspended from my job the next day, as expected, for threa- cases to take care of. This really upset me – it’s true, he didn’t threaten me
tening and abusing a suspect, and I’m currently facing criminal charges. It was with murder, but I was still terrified of him. Every time I changed my number,
a very, very stupid thing of me to do. But back then, I thought it was a matter he somehow found out my new one after a few weeks. He started to hang
of life and death. around in front of the building I live in, and even after being told numerous
times by me and everyone else to stay the hell away from me, I still sometimes
saw him around.
Policeman scenario, non-emotional version I just wanted him to understand that I didn’t want him around me, and
I used to be a police officer in London and I really enjoyed my job. I know a lot since the police wouldn’t help me, I eventually decided to take matters into
of people don’t like cops, but I swear I worked in this job to do something my own hands. I just didn’t want to feel like a helpless victim anymore. I got
good for the community, shield people from harm, and make the world a little into the darknet and hired four guys to corner him when he was lingering
bit more just. around my area again and beat him up thoroughly. I told them to threaten
About a year ago, we caught this young guy a few hours after he had him with further violence should he ever come near me again. I’m not gonna
robbed a bank in South London. He had entered the building with a fake lie, I felt good about it before I heard how badly hurt he was. They had
gun, threatened employees and customers, and left the place with a five-digit kicked his head so hard that it broke his skull. The doctors say that he suf-
sum in cash. He was arrested very quickly, but he had had enough time to hide fered a traumatic brain injury and is in a permanent vegetative state now. It’s
the money before his arrest. We couldn’t find it anywhere – we assumed that it not clear whether his condition will ever improve. I wanted him to get hurt,
was buried somewhere, but even after a vast search, the money stayed gone. but not this badly. This is definitely not what I intended, and I feel terrible
The suspect, obviously, wouldn’t tell us where it was, even though it would about it.
have mitigated his sentence. He thought he could just go to prison for a few
years and get the money at some point afterwards, go abroad, and live happily
ever after. Darknet scenario, non-emotional version
In these days, I was very close to being promoted. I knew that if I could just Your university years are supposed to be the most carefree years of your adult
get the suspect to tell us where the money was, it would look amazing on my life, but for me, it hasn’t been great so far. I’m a student at Bristol, but instead
job record, and I would most likely be one step closer to my dream position as of studying, partying, and dating, the thing that was on my mind for most of
a highprofile investigator. So I did everything I could to get him to talk, but the past year was Josh. We met at a party during Fresher’s week (although it
nothing worked. I was getting frustrated – it would have been better for later turned out that he wasn’t even a student) and started dating soon after-
him, too, if he could just talk! Nothing worked on him, not even subtle threats, wards. Our relationship lasted for about a year. The break-up was very hard,
and I saw my prospects of being promoted retreating into the distance. On the but he hadn’t been a great boyfriend, so I got over it. What I did not get
third day of questioning him, both my colleague and his attorney left the room over, however, was that he owed me a shitload of money. You see, he had a
for a minute or two. I was not supposed to talk to him during this time, but I hard time financially during that year – his parents didn’t support him and
did. he lost his job in June. His landlord threatened to throw him out, and he des-
I openly confess to what I did when I was alone with him: I told him that if perately needed money. We had been together for ten months by then, so I
he didn’t tell us where the money was, I would shoot him in the knee. When trusted him and gave him the 2,000 pounds I had in savings under the condi-
he didn’t respond, I took my gun out of the holster and acted like I was loading tion that he would pay me back as soon as possible. But when we broke up, he
it, while verbally threatening him with the most intense pain he had ever felt in still hadn’t paid me back a penny, and when I texted him after the break-up
his life, the possibility of never walking again, and how much worse life in that I still wanted and needed the money, he just didn’t reply. Over the next
prison would be for someone who couldn’t walk. He started crying, and few weeks, I repeatedly texted and called him, but he blocked me and refused
under his breath, he finally told me the whereabouts of the money. It was a to talk to me.
very, very stupid thing of me to do, and I don’t know why I was thinking It’s not like I desperately needed the money, it’s just that the fact that he
that I would get away with it. I guess I just thought he’d be too scared to simply refused to pay me back made me so angry. It was just so ludicrous
tell anyone, or that no one would believe him over me. I was suspended and disrespectful! The break-up had been bad enough, why would he humili-
from my job the next day for threatening and abusing a suspect, and I’m cur- ate me further? So instead of just reporting him to the police, I decided to
rently facing criminal charges. humiliate him back. I just wanted him to understand that actions have conse-
quences. I just didn’t want to feel like a victim anymore. I got into the darknet
and hired four guys to corner him and beat him up thoroughly. I told them to
Darknet scenario, emotional version
threaten him with further violence should he not pay me back the money as
Your university years are supposed to be the most carefree years of your adult soon as possible. I’m not gonna lie, I felt good about it before I heard how
life, but for me, it’s been hell so far. I’m a student at Bristol, but instead of badly hurt he was. They had kicked his head so hard that it broke his skull.
studying, partying, and dating, the thing that was on my mind for most of The doctors say that he suffered a traumatic brain injury and is in a permanent
the past year was Josh. We briefly met at a party during Fresher’s week, vegetative state now. It’s not clear whether his condition will improve. This is
although it later turned out that he wasn’t even a student. We only talked a definitely not what I intended, and I feel terrible about it.

https://doi.org/10.1017/S136672892400018X Published online by Cambridge University Press