The Relation between Emotional Processing and Language
Proficiency: A Cross-Cultural Study with Native and
Non-Native Dutch Speakers
Name: Svea Cornelius Mentor: Marte Otten Student number: 10762183
Abstract
To investigate whether processing in a second language is characterized by the same sensitivity to emotional stimuli as native language processing, we conducted an EEG study where participants
had to read insulting and complimenting sentences. Two groups – native speakers of Dutch and German speakers with sufficient Dutch language proficiency – participated in the study. The results
showed that insults and compliments evoked the same increase in N400 (indicating increased semantic processing) and LPP amplitude (indicating emotional processing) in both of the groups. In
contrast to what theories concerning emotional processing assumed, emotional processing in the second language had a highly similar ERP pattern as native emotional processing.
Three years ago I left Germany, the country I grew up in, and moved to the Netherlands. It did not take long until I was fluent in Dutch. Eventually, I started with a bachelor's degree that was taught in Dutch. On a Friday afternoon, I came home with some of my new friends. We wanted to have dinner and celebrate a friend's birthday in my apartment. I had lived in that apartment together with my Dutch landlady for a few months. Usually, I did not invite many friends to my place because my landlady did not want me to. That weekend, however, she had planned to visit an old friend.
Therefore, I invited everybody to my place. What seemed to be a good plan in the first place soon turned out to be a big mistake. My landlady had changed her plans and wanted to visit her friend the weekend after. When she came home and saw me and all my friends in the kitchen she became furious. That was the first time I got into an argument with her. What followed was a monologue of her about what she disliked about me and things I did not do right in her eyes. As a matter of course this monologue was in Dutch. I understood every single word, but somehow all the things she said to me went in one ear and out the other. When she left the kitchen, I did not feel hurt or upset at all. I started to ask myself why it was not as painful as it would have been if everything she said would had been in German, my mother tongue. That was the first time I realized that insults in a language that you acquired later in life do not hurt as much as in your mother tongue. “If you talk to a man in a language he understands, that goes to his head. If you talk to him in his language, that goes to his heart.” This quote of Nelson Mandela perfectly describes what I had experienced that night.
Namely, that emotional processing seems to be enhanced in the first language and reduced in later-learned languages.
This phenomenon has been a topic in research for quite a long time. There are many
different ways to explore the relation of emotional processing and language proficiency. We will try to shed light on the complexity of this relation by approaching it from four different perspectives: the clinical, the introspective, the psycho-physiological, and the cognitive angle. Taking it one step further, in the current study we aim to get a closer look on the brain mechanisms linking language and emotional processing by using the electroencephalography (EEG).
Clinical approach
The clinical approach of the relation between emotional processing and language focuses on affective reactivity. Freud and his disciples were the first ones to notice language disturbances in their multilingual patients (Freud, 1893). What they observed was that some of the patients
switched from their first language to one of their other languages for discussions about difficult and anxiety-producing topics. Marcos (Marcos, 1976) described his bilingual patients as emotionally withdrawn or detached in their second language. He was the one to came up with a possible explanation for this differential affective reactivity in multilinguals. In his eyes, multilinguals, compared to monolinguals, have to put more effort in the way they say things instead of what they want to say when they speak in one of their non-native languages. In contrast to that, monolinguals can use all their energy sources to express their emotions.
These observations show that there is an interaction between language and emotions. However, the findings exclusively stem from patients in the clinical context and it might be the case that the observed behavior differs from how the patients would behave in the natural environment. It is therefore interesting to examine how the relation between emotions and language outside the clinical context looks like.
Introspective approach
In contrast to clinical studies, which are based on observations of the therapist, the introspective approach relies on the speakers' subjective perception of emotionality. Commonly, questionnaires are used to gain insight into people's feelings and thoughts. They constitute a practical way to collect an extensive amount of information from a large number of people which, in the end, increases the results' generalizability.
So far, the largest questionnaire-based study concerning bilingualism and emotions was the 'Bilingualism and Emotions Questionnaire' (BEQ), conducted by Dewaele and Pavlenko (Dewaele, 2004a, 2004b, 2004c, 2006, 2008, 2010; Pavlenko, 2004, 2005). This online questionnaire was
freely accessible for two years and collected data from a large sample of bilinguals and
multilinguals all over the world. An example of what the questionnaire examined was the question, whether the emotional weight of the phrase “I love you” varies between the different languages that multilingual speakers are proficient in. The results show that the majority of people perceived this phrase as being strongest when it is presented in their first language. Moreover, they demonstrated that the mother tongue is significantly more likely to be used when expressing positive or negative emotions. Furthermore, bilinguals and multilinguals rated taboo- and swearwords in their native language as more emotional compared to words in one of their non-native languages.
Several of these findings have been replicated in other studies. Puntoni and colleagues (Puntoni, Del Langhe, & van Osselaer, 2009), for instance, asked multilinguals to rate their perceived emotionality to advertising slogans that were presented in all the languages they were proficient in. Overall, slogans were rated as more emotional when they were presented in the mother tongue.
One explanation for this varying perception of emotionality comes from Altarriba (Altarriba, 2003). She suggests that emotional words in the first language are stored at a deeper level than their counterparts in a language that was learned at a later point in life. What makes the emotion words in the mother tongue so special is the fact that they have been experienced in a lot of different
contexts. This leads to the creation of multiple traces in the memory which in turn strengthens their semantic representation. Thus, a later acquired language provides a greater emotional distance than the first language.
However, one limitation of the introspective approach is that self-reports rather provide participant's interpretations of their experiences than a direct access to emotional processing. The psycho-physiological approach, on the other side, can be used as an alternative to measure the relation between emotional processing and language in a more objective way.
While the introspective approach might be prone to biases like the observer-expectancy effect, the psycho-physiological approach relies solely on involuntary and automatic responses that can hardly be influenced by the participant. This approach examines physiological responses and organs that are innervated by the autonomic nervous system (Cacioppo, Tassinary, & Berstson, 2007). One example is the human skin. The electrical conductivity of our skin is particularly sensitive to emotional stimuli like threatening pictures. The perception of that type of stimuli leads to an increase in adrenaline, which in turn leads to sweating, which eventually increases the electrical conductivity (Pavlenko, 2012).
Harris and colleagues (Harris, Aycicegi, & Gleason., 2003) inspected the skin conductance responses of bilinguals while they were reading and listening to taboo words. The results revealed that electrodermal reactivity is higher when the taboo words where presented in the mother tongue. In another study, Harris (Harris, 2004) measured the electrodermal reactivity in bilingual speakers while they were listening to emotional phrases in the two languages they spoke. What he found was that emotion-laden words elicit a higher electrodermal reactivity when they are presented in the mother tongue compared to when they were presented in their second language.
Even though there appears to be a connection between emotional stimuli, language proficiency and physiological responses, we do not know how big the influence of other and possibly even confounding variables is. Heightened levels of arousal, for instance, might also be elicited by greater cognitive effort that is required for a task while lower levels of arousal might stem from task habituation effects (Harris, 2004).
Cognitive approach
Using a cognitive approach might allow for a distinction between the different sources of arousal and remove uncertainties caused by different patterns of emotional responses by taking into account various cognitive processes, such as memory, interference and priming and their interaction with language proficiency.
Anooshian and Hertel (Anooshian & Hertel, 1994) examined whether language proficiency has an effect on memory. The design was simple but powerful: first, the bilingual participants had to rate emotional (e.g. death) and neutral words (e.g. table) - presented in their two languages - for ease of pronunciation and emotional intensity. This part of the study was followed by a surprise recall task where the participants had to recall as many of the previously seen words as possible. What they found was that recall was better for emotional words in the mother tongue. These findings suggest that memory is enhanced for emotional words in the first language.
A commonly used task to measure interference and facilitation effects, is a modified version of the emotional Stroop task: the so called Rapid Search Visual Presentation (RSVP) task. The emotional Stroop task is a task where participants have to identify the print color of emotion-laden (e.g., 'love' in green ink) and neutral (e.g., 'window' in red ink) words. Usually, emotional and in particular negative and taboo words have an interfering effect and slow down the response time. This effect was used for the Rapid Search Visual Presentation task. In the RSVP task, participants sit in front of a screen on which two words are presented in a rapidly and sequentially manner. The first word is either a neutral or a taboo word and the second word is a color word (e.g., blue). The participants' only task is to identify the color word. Colbeck and Bower (Colbeck & Bowers, 2012) used the RSVP task to test whether taboo words are more interfering than neutral words and
whether they would have a less interfering effect in the second language of bilinguals. They found the error rate to be significantly higher when taboo words were presented in the mother tongue compared to the second language. Therefore, it seems that the reduced emotionality in the
participants' second language diminishes the interference effect of the taboo word, which eventually leads to an advantage in processing.
Degner and colleagues (Degner, Doycheva, & Wentura, 2012) investigated whether affective priming can occur both in the first and in the second language of bilinguals. To test this they had both French-German as well as German-French bilinguals participate in their study. The stimuli in the task consisted of word couples formed out of either a positive prime word (e.g., sourier,
Lächeln, 'smile') or a negative prime word (e.g., douleur, Qual, 'agony') together with a neutral target word (e.g., chaise, Stuhl, 'chair'), a positive target word (e.g., loisir, Freizeit, 'leisure') or a negative target word (e.g., echec, Misserfolg, 'failure'). The prime word was shown for 150ms and was directly overwritten by the target word that appeared on the screen afterwards. The participants task was to rate the target word as positive or negative. What they found was that participants were faster in rating the target word when it was congruent with the prime word. However, this was only the case when the word pairs were presented in the first language of the bilinguals. Affective priming was not observed in the second language of the participants.
Putting it all together, we can clearly see that there is an interaction between emotional processing and language in terms of cognitive functions. Memory, interference and priming of emotional stimuli are enhanced in the first language. The cognitive approach sees the human brain as a black box from which we can observe different responses resulting from different stimuli. It might be more interesting, however, to really see what is happening in the brain.
Neuroimaging approach
In the current study, we therefore hope to get a deeper inside in the brain mechanisms that are linked to the potential relation between emotional processing and language proficiency by using the electroencephalography (EEG). EEG monitors neural excitation and registers complex patterns of neural activity within a few milliseconds after a stimulus has been presented (Teplan, 2002).
Previous studies have indicated that emotional stimuli elicit more activity in the brain than neutral stimuli (Kousta, Vinson, & Vigliocco, 2009). A recent EEG study (Otten, Mann & van Berkum, 2017) went one step further and looked at the differences between processing of positive and negative stimuli. In their study, they compared insults to compliments. They found that reading insulting sentences evokes higher brain activity than reading complimenting sentences. With the aim of replicating these findings, we chose for insults and compliments as the emotional stimuli in our study. However, our main question is whether emotional processing differs depending on
language proficiency. In order to be able to answer this question we had two different groups of participants taking part in our experiment: Dutch native speakers on one side and German native speakers who had learned Dutch at a later point in life on the other side. Both groups had to read insulting and complimenting sentences in Dutch. For our study of highest importance was the electrical activity of the participants' brain right after the presentation of the insulting or the
complimenting word. We focused on two different components of the event-related potential (ERP) to analyze the brain activity: the N400 and the late positive potential (LPP).
The N400 is a negative brain potential that is peaking around 400 ms after the onset of a stimulus. It is related to semantic processing and therefore it is well suited for the current study. Kutas and Federmeier (Kutas & Federmeier, 2014) showed that an elicitation of an N400 can be seen when people are exposed to an unexpected word in a sentence. These finding are particularly relevant for the current study since participants in our study presumably do not expect to be insulted or complimented. In line with the findings of Otten and colleagues (Otten et al., 2017), another study had found socially threatening words such as personal insults to evoke an N400 (Wabnitz, Martens, & Neuner, 2012). This suggests that the insulting sentences in our experiment will evoke a higher amplitude in the N400 than the complimenting sentences.
The late positive potential (LPP) is a positive component within the ERP with a maximum amplitude between 600 and 1000 ms after stimulus presentation. It is often used in research on emotions. Results of different studies show that emotional stimuli promote a higher LPP than neutral stimuli (Schupp, Cutbbert, Bradley, Cacioppo, Ito, & Lang, 2000; Krolak-Salmon, Fischer, Vighetto, & Mauguiere, 2001). One study that went deeper into this matter, illustrated that negative stimuli evoke larger LPP as compared with positive and neutral stimuli (Ito, Larsen, Smith, & Cacioppo, 1998). This is in line with the findings of the study of Otten and colleagues (Otten et al., 2017). A combination of these findings suggests that reading insulting sentences compared to complimenting sentences will result in a higher LPP amplitude.
amplitudes than compliments. As mentioned above, several studies investigating the relation between emotional processing and language proficiency by looking at cognitive functions, found higher levels of cognitive processing in multilingual speakers when emotional words were
presented in their native language compared to one of their non-native languages. These results led us to assume that we might find similar results when studying the brain and directly looking at its activity. In concrete terms this means that we expect the LPP and N400 amplitudes of the native Dutch participants to be lager than the amplitudes of the non-native Dutch speakers.
Methods
Participants
Forty four participants (30 women, 11 German speakers, mean age 22,69 years, range 19-43 years), that were recruited from the University of Amsterdam participant-pool, took part in our experiment. For their participation, which on average took 1,5 hours, the participants could earn study credits. To guarantee a sufficient level of Dutch language proficiency, the participants either had to be Dutch native speakers or had to have lived in the Netherlands for at least two years. Since the impact the insults would have on the participants was not yet clear, we chose to exclude participants with a history of mental disorders. Moreover, we decided to exclude dyslexic participants. The Ethical Review Board of the Department of Psychology at the University of Amsterdam approved the study. Before starting with the experiment all participants read and signed the informed consent.
Stimuli
We constructed 56 insulting sentences ('Your appearance is unpleasant and even a bit repulsive.' ['Jouw uitstraling is onprettig en zelfs een beetje afstotend.']) and 56 complimenting sentences ('You are the most fearless person I know.' ['Je bent de dapperste persoon die ik ken.']). Both the insulting and complimenting sentences mainly focused on describing the reader's personality, appearance, and interaction with others. Overall, the length of the two different types of sentences did not differ
(Minsult = 9.7, SDinsult = 2,4; Mcompliments = 9.6, SDcompliment = 3.0). The critical word for our analysis was the first explicitly insulting (i.e. 'unpleasant'[/'onprettig'] in the example) or complimenting (i.e. 'fearless'[/'dapperste']) word in the sentence. The position of the insulting and complimenting word in the sentence was comparable for the two different types of sentences (Minsult = 6.3, SDinsult = 2.2; Mcompliments = 6.3, SDcompliment = 2.6). The length of the critical word in the two conditions did not differ (Minsult = 8.3, SDinsult = 3.5; Mcompliments = 7.7, SDcompliment = 2.7). For the complete list of the Dutch sentences see Supplementary
Materials.
Procedure
Our experiment had a 2x2 mixed factorial design: Native language (2 levels: Dutch (mother tongue), Germans (later in life acquired language)) * Sentence Message (2 levels: insult,
compliment). The order of the 56 sentences in the insulting and the complimenting sentences block was random. Each sentence was paired with a picture of a person the insult or compliment,
hypothetically, stemmed from. Half of the sentences contained the phrase 'Hi (the participant's name)!' before the insulting or the complimenting sentence began. We split the two big blocks of 56 sentences up into smaller blocks of 4x14 sentences to be able to maintain the participants full attention. Between the two blocks of 56 sentences the participants had a ten minute break where they could drink something and relax a bit.
The sentences were presented word for word. The words appeared automatically after the participants had pressed the space bar. We used a Variable Serial Visual Presentation (VSVP) procedure (Otten, Nieuwland, & van Berkum, 2007a) to make the visual presentation of the
sentences as natural as possible. To calculate the presentation time of each word we used a standard offset of 240 ms for every word and added 25 ms for each of the words letters. Hence, words with more letters were presented for a longer amount of time than words with fewer letters. Words preceding a comma as well as sentence-final words were prolonged with an additional 800 ms. The
critical word in each sentence had a fixed presentation time of 425 ms on the screen.
At the beginning of the experiment, participants were instructed to avoid eye-movements and blinks when sentences were presented on the screen. The participants were able to read the sentences at their own pace by pressing the space bar to start a new sentence. We encouraged them to take small breaks between sentences to blink or move their eyes.
Figure 1. Provides an illustration of the sequence of presentation of the beginning of an insulting sentence.
Manipulation check
After the first three 14-sentence blocks of each of the two big blocks, the participants had to fill in a questionnaire that was designed to rate their mood at that point of time. This questionnaire
contained eight different elements of the construct 'mood': feeling angry, humiliated, embarrassed, worthless, furious, cheerful, happy, grateful [boos, vernederd, beschaamd, waardeloos, woedend, vrolijk, blij, dankbaar]. The scale on which the participants could specify their level of agreement was a 7-point-likert scale, ranging from 1 (strongly disagree ['absoluut niet']) to 7 (strongly agree ['absoluut wel']). Since the total amount of blocks was eight, the participants had to fill in the questionnaire six times. To get a reference to compare the results of the questionnaire with, the participants had to fill in the same questionnaire before they started with the experiment. We used the results of the questionnaire as a manipulation check to see whether there was an actual effect of the insults and the compliments on participants' mood. At the end of the two big blocks, a
of these questionnaires in our analysis.
Electrophysiological recording
EEG was recorded with a Biosemi active-electrode system (Biosemi Inc., Amsterdam, The Netherlands) which included 64 + 6 electrodes (Fp1, AF7, AF3, F1, F3, F5, F7, FT7, FC1, FC3, FC1, C1, C3, C5, T7, TP7, CP5, CP3, CP1, P1 , P3, P5, P7, P9, PO7, PO3,O1, lz, Oz, POz, Pz, CPz, Fpz, Fp2, AF8, AF4, AFz, Fz, F2, F4, F6, F8, FT8, FC6, FC4, FC2, FCz, Cz, C2, C4, C6, T8, TP8, CP6, CP4, CP2, P2, P4, P6, P8, P10, PO8, PO4 and O2; CMS and DRL). By placing two electrodes under and next to the left eye, we were able to register blinks and eye-movements. Offline re-referencing of the EEG signal was obtained by placing two electrodes on the left and right mastoid. The power of the brain signals was increased by a BioSemi ActiveTwo amplifier (-3dB at ~102 Hz low-pass, fully DC coupled) with a sample rate of 512 Hz.
We used the Independent Component Analysis (ICA) to remove noise that stemmed from blinks and eye-movements (Jung, Makeig Humphries, Lee, Mckeown, Iragui, & Senjowski, 2000). A high-pass filter of .1 Hz was applied to the EEG signals. After that, we extracted data epochs that were time locked to the onset of the insulting or the complimenting word. The epochs lasted from 150 ms before until 1500 ms after the presentation of the critical word. We used the prestimulus window from -150 ms to 0 ms as a baseline. The remaining, artifact-free trials were then averaged for the four conditions (Dutch, German, insult, compliment). With regard to the predictions we made, we calculated the mean amplitudes of the ERP in the N400 region (350-450 ms after the onset of the critical word) and in the LPP region (600-1000 ms) for each participant and each condition.
Data analysis
First we checked whether the manipulation in our experiment - the insults and compliments – was perceived as realistic and actually had an effect on the mood of the participants using a paired
sample t-test. After that, we performed our main analysis. Data of participants who had more than 3 electrodes with less than 20 segments were excluded from the analysis. Subtracting these
participants from the whole group of participants, eight German and 27 Dutch participants were used for our calculations. Two separate Factorial Mixed ANOVAs were conducted to measure the influence of the two independent variables (sentence message and mother tongue) on the ERP signal for the N400 and the LPP time window.
Results
Manipulation check
We used a paired sample t-test to test whether the insults and compliments would have an effect on the mood of the participants. For this, we calculated the mean of the three questionnaires that belonged to the insulting and the complimenting block in the Dutch and German participant group and compared it with the mood that was registered before the manipulation had started. Three outliers were deleted before we started with the analysis. The results revealed that participants rated their mood as significantly worse after reading the insults compared to the compliments, t(32) = 3.27, p = .01, Cohen's d = .56.
N400
Contrary to our predictions, insults and compliments did not evoke significant differences in N400 amplitude of the ERP. There was a trend for a main effect of sentence message on the N400
amplitude, F(1,34) = 2.365, p = .13, partial η² = .07. The main effect for mother tongue yielded a non-significant F ratio of, F(1,33) = 1.806, p = .19, partial η² = .05. There was no significant interaction between sentence message and mother tongue, F(1,33) = 1.031, p = .32, partial η² = . 03). These results show that the ERP between the native Dutch speakers and the native German speakers did not differ between 350 and 450 ms after stimulus presentation no matter if they were insulted or complimented.
LPP
The Factorial Mixed ANOVA for the second time window (600-1000 ms after stimulus onset) revealed similar results. There were no significant main effects of sentence message on the ERP (F(1,34) = 0.077, p = .78, partial η² < .01. However, there was a trend to a main effect of mother tongue, F(1,33) = 2.189, p = .15, partial η² = .06). We did not find a significant interaction between sentence message, mother tongue and ERP, F(1,33) = .848, p = .36, partial η² = .03. Therefore, we again see that the ERPs between the Dutch and the German group did not differ within the two different sentence messages in the time window between 600 and 1000 ms after the onset of the critical word. Figure 2 shows an overview of the measured ERPs for the Dutch and the German participants for both the insults and compliments.
German participants Dutch participants
_____________________________________________________________________________
Figure 2. Shows the ERPs evoked by insults (black lines) and insults (red lines) in the Dutch (mother tongue) and the German group (not mother
tongue). Czalle CDuitsers -200 0 200 400 600 800 1000 1200 -10 -5 0 5 10 Czalle CNed -200 0 200 400 600 800 1000 1200 -10 -5 0 5 10 Pzalle CDuitsers -200 0 200 400 600 800 1000 1200 -10 -5 0 5 10 Pzalle CNed -200 0 200 400 600 800 1000 1200 -10 -5 0 5 10 Ozalle CDuitsers -200 0 200 400 600 800 1000 1200 -10 -5 0 5 10 Ozalle CNed -200 0 200 400 600 800 1000 1200 -10 -5 0 5 10 ____ Insults _____ Compliments
Discussion
The present study investigated the effect of language proficiency on emotional processing. Dutch participants as well as German-Dutch bilinguals read insulting and complimenting sentences that were written in Dutch. For our analysis, we focused on two components of the ERP signal that had previously been described as characteristic to processing of emotion-laden words in native
language: the N400 and the LPP (Schupp et al., 2000; Wabnitz et al., 2012). Even though, we did not find an effect of sentence message or language proficiency on the N400 and LPP, we can see a peak in negativity between 350 and 450 ms and a more positive ERP between 600 and 1000 ms after the onset of the critical word (Figure 2). These peaks indicate a potential N400 and LPP effect.
Influence of sentence message on emotional processing
Contrary to what we had expected and what had been found in previous research on this topic (Otten et al., 2017), reading insulting sentences did not evoke higher N400 and LPP amplitudes than reading complimenting sentences. Although, we did not find a significant effect of sentence
message on the N400 and LPP, our data and Figure 2 show that the amplitudes, evoked by the insulting sentences, were higher than the amplitudes that were evoked by the compliments.
A reason for the absence of a significant difference in amplitudes between Dutch and
German participants in the current study, could be the modest sample size of the German participant group (N=7). Due to this, it is likely that the power of our study was too small and therefore failed to detect a difference between insults and compliments. A second reason could be that the data was too noisy. Almost half of the German participants had to be removed due to bad EEG channels. In order to calculate the appropriate sample size to detect an effect, we conducted a post hoc power analysis. This analysis revealed that a sample size of approximately 18 German participants would be necessary to reach a statistical power of .80.
fact that there might be indeed no difference in processing of any emotional stimuli, may they be positive or negative. Support for this theory comes from Holt, Lynn and Kuperberg (2009). They observed that not only negative stimuli but explicitly positive words can evoke an increase in N400 amplitude as well (Holt et al., 2009). The same applies to the LPP. A recent study demonstrated that, relative to neutral stimuli, negative as well as positive stimuli elicit enhanced LPP amplitudes (Liu, Huang, McGinnis, Keil, & Ding, 2012).
For further research on this matter, it is recommendable to compare the insulting sentences not only with complimenting sentences but also with neutral sentences that do not have any emotional valence. Moreover, it is necessary to use increase the sample size in order to minimize noise in the EEG signal and to boost the statistical power.
Influence of language proficiency on emotional processing
We hypothesized that language proficiency would have an effect on emotional processing.
Considering the findings of the clinical, introspective, psycho-physiological and cognitive approach on the relation of language proficiency and emotional processing, it seemed likely that emotional processing is the highest when words are presented in the first language. Therefore, we expected to see higher N400 and LPP amplitudes in the native Dutch participants compared to the native German participants. However, we found that language proficiency of the participants did not change the way they processed the insulting and complimenting sentences in our study. This raises the question why we did not observe any differences in our study.
One explanation might be that our manipulation was just not credible enough. We used a picture of a person that, hypothetically, insulted or complimented the participants in our study. Since this person was unknown to all our participants, it is possible that they did not take the message of the sentences personally. The experimental setting in which the study took place may have contributed to this effect as well. This might be a reason for why we did not find any
a higher ecological validity, it is of greatest importance to make sure that the insults and compliments are perceived as personal. As a more realistic and credible alternative we thought about the possibility of letting the participants solve a difficult task with other, hypothetical, participants. This may lead to a basic personal relationship between the real and the fake
participant. Being insulted by the fake participants, for example for a bad performance on the task, would probably be more believable for the real participant.
Additionally, it is possible that the participants rated their mood as more negative after the insults and as more positive after the compliments merely due to what seemed more reasonable to them. The fact that the participants had to fill in the questionnaire several times during the
manipulation may have pressured them to adjust their ratings to what they thought would fit to the researchers' expectations. With this in mind, it might be better to only let the participants fill in the questionnaire after but not during the manipulation.
Another point of criticism is that we did not measure Dutch language use in everyday life of the German participants. However, emotional processing of German participants with a high
frequency of Dutch language use, might be more similar to the Dutch participants. This would be in line with what Degner and colleagues (Degner, Doycheva, & Wentura, 2012) found in their study. They saw that participants with a high daily frequency of their second language showed the same automatic processing of emotional stimuli as native speakers, even though the participants did not differ in their level of proficiency in the second language. Further evidence for this theory comes from Van Petten and Kutas (van Petten & Kutas, 1990). They showed that the N400 amplitude is affected by the frequency of a word's usage: less commonly used words elicit a higher N400 amplitude than words with a high frequency of usage. Given that the sample of the Germans had a size of seven participants, even two or three Germans with a high usage of the Dutch language could skew the result. That could be a reason why we did not see any differences in emotional processing between the German and the Dutch participants. In order to control for this moderating effect, it is necessary to ask for participants' frequency of use of their non-native language.
An explanation for the difference between our results and previous results. may be the fact that language proficiency might not matter when it comes to processing of complex sentences compared to single words. All the cognitive and psycho-physiological studies we cited in our introduction found an effect of language proficiency on emotional processing of emotional words (Degner et al., 2012; Colbeck & Bowers, 2012; Annoshian & Hertel, 1994; Harris, 2004; Harris, Ayc¸ic¸egi & Gleason, 2003). In our study, however, participants had to read whole sentences. To bring further clarification to the impact of semantic complexity, future work is needed to investigate the relationship between semantic complexity and emotional processing in multilinguals.
Temporally, we can conclude from this study that emotional processing is not dependent on language proficiency. Due to methodological deficiencies this observation can not be seen as conclusive evidence. This results stimulate to further investigate the relation between emotional processing and language proficiency.
References
Anooshian, L., & Hertel, P. (1994). Emotionality in free recall: Language specificity in bilingual memory. Cognition and Emotion, 8(6), 503–514. doi.org/10.1080/02699939408408956 Altarriba, J. (2003). Does carin˜o equal ‘liking’? A theoretical approach to conceptual
nonequivalence between languages. International Journal of Bilingualism, 7(3), 305–322. doi:abs/10.1177/13670069030070030501
Cacioppo, J. T., Tassinary, L. G., & Berntson, G. (Eds.). (2007). Handbook of psychophysiology. Cambridge University Press.
Caldwell-Harris, C., & Ayc¸ic¸egi-Dinn, A. (2009). Emotion and lying in a non-native language. International Journal of Psychophysiology, 71(3), 193–204.
doi.org/10.1016/j.ijpsycho.2008.09.006
Colbeck, K., & Bowers, J. (2012). Blinded by taboo words in L1 but not L2. Emotion, 12(2), 217– 222. doi.10.1037/a0026387
Conrad, M., Recio, G., & Jacobs, A. (2011). The time course of emotion effects in first and second language processing: A cross-cultural ERP study with German-Spanish bilinguals. Frontiers in Pyschology: Language Sciences, 2, 351. doi.10.3389/fpsyg.2011.00351
Degner, J., Doycheva, C., & Wentura, D. (2012). It matters how much you talk: On the automaticity of affective connotations of first and second language words. Bilingualism: Language and Cognition, 15(1), 181–189. doi.org/10.1017/S1366728911000095
Dewaele, J.-M. (2004a). The emotional force of swearwords and taboo words in the speech of 426 multilinguals. Journal of Multilingual and Multicultural Development, 25(2-3), 204–222. doi.org/10.1080/01434630408666529
Dewaele, J.-M. (2004b). Blistering barnacles! What language do multilinguals swear in?! Estudios de Sociolingüistica, 5(1), 83–105.
Dewaele, J.-M. (2004c). Perceived language dominance and language preference for emotional speech: The implications for attrition research. In M. Schmid et al. (Eds.), First language
attrition: Interdisciplinary perspectives on methodological issues (pp. 81–104). Amsterdam, The Netherlands: John Benjamins.
Dewaele, J.-M. (2006). Expressing anger in multiple languages. In A. Pavlenko (Ed.), Bilingual minds: Emotional experience, expression, and representation (pp. 118–151). Clevedon, UK: Multilingual Matters.
Dewaele, J.-M. (2008). The emotional weight of I love you in multilinguals’ languages. Journal of Pragmatics, 40(10), 1753–1780. doi.org/10.1016/j.pragma.2008.03.002
Dewaele, J.-M. (2010). Emotions in multiple languages. London, UK: Palgrave Macmillan.
Freud, S. (1893). The psychical mechanisms of hysterical phenomena. London, UK: Hogarth Press. Grosjean, F. (1994). Individual bilingualism. The encyclopedia of language and linguistics (pp.
1656-1660). Oxford: Pergamon Press.
Hahne, A., Friederici, A. D. (2001). Processing a second language: Late learners' comprehension mechanisms as revealed by event-related brain potentials. Bilingualism: Language and Cognition, 4(2), 123-141. doi.org/10.1017/S1366728901000232
Harris, C. (2004). Bilingual speakers in the lab: Psychophysiological measures of emotional reactivity. Journal of Multilingual and Multicultural Development, 25(2), 223-247. doi.org/10.1080/01434630408666530
Harris, C., Ayc¸ic¸egi, A., & Gleason, J. (2003). Taboo words and reprimands elicit greater autonomic reactivity in a first language than in a second language. Applied
Psycholinguistics, 24(4), 561–571. doi.org/10.1017/S0142716403000286 Holt, D. J., Lynn, S. K., & Kuperberg, G. R. (2009). Neurophysiological correlates of
comprehending emotional meaning in context. Journal of Cognitive Neuroscience, 21(11), 2245-2262.
Ito, T. A., Larsen, J. T., Smith, N. K., & Cacioppo, J. T. (1998). Negative information weighs more heavily on the brain: The negativity bias in evaluative categorizations. Journal of
Jung, T.-P., Makeig, S., Humphries, C., Lee, T.-W., Mckeown, M.-J., Iragui, V., Sejnowski, T. J. (2000). Removing electroencephalographic artifacts by blind source separation.
Psychophysiology, 37(2), 163-178.
Keysar, B., Hayakawa, S. L., An, S. G.(2012). The Foreign-Language Effect: Thinking in a Foreign Tongue Reduces Decision Biases. Psychological Science, 23(6), 661-668.
doi/abs/10.1177/0956797611432178
Kousta, S., Vinson, D., Vigliocco, G. (2009). Emotion words, regardless of polarity, have a processing advantage over neutral words. Cognition, 112(3), 473-481.
doi.org/10.1016?j.cognition.2009.06.007
Kovelman, I., Baker, S. A., & Petitto, L.-A. (2008). Bilingual and Monolingual Brains Compared: A Functional Magnetic Resonance Imaging Investigation of Syntactic Processing and a
Possible “Neural Signature” of Bilingualism. Journal of Cognitive Neuroscience, 20(1), 153-169.
Krolak-Salmon, P., Fischer, C., Vighetto, A., & Maugiere, F. (2001). Processing of facial emotional expression: Spatio-temporal data as assessed by scalp event-related potentials. European Journal of Neuroscience, 13(5), 987-994. doi:10.1046/j.0953-816x.2001.01454.x
Kutas, M., & Federmeier, K. D. (2011). Thirty years and counting: Finding meaning in the N400 component of the event-related brain potential (ERP). Annual review of psychology, 62, 621-647. doi.org/10.1146/annurev.psych.093008.131123
Liu, Y., Huang, H., McGinnis, M., Keil, A., & Ding, M. (2012). Neural substrate of the late positive potential in emotional processing. Journal of Neuroscience, 32(42), 14563-14572.
doi.org.10.1523/JNNEUROSCI.3109-12.2012
Marcos, L. R. (1976). Bilinguals in psychotherapy: Language as an emotional barrier. American Journal of Psychotherapy, 30(4), 552-559.
Otten, M., Nieuwland, M. S., & van Berkum, J. J. A. (2007a). Great expectations: Specific lexical anticipation influences the processing of spoken language. BMC Neuroscience, 8(1), 89.
doi:10.1186/1471-2202-8-89
Otten, M., Mann, L., van Berkum, J. J. A. & Jonas, K. J. (2017). No laughing matter: How the presence of laughing witnesses changes the perception of insult. Social Neuroscience, 12(2), 182-193.
Ożańska-Ponikwia, K. (2013). Emotions from a bilingual point of view: Personality and emotional intelligence in relation to perception and expression of emotions in the L1 and L2.
Newcastle upon Tyne, United Kingdom: Cambridge Scholars Publishing.
Pavlenko, A. (2004). ‘Stop doing that, ia komu skazala!’: Language choice and emotions in parent– child communication. Journal of Multilingual and Multicultural Development, 25(2), 179– 203. Pavlenko, A. (2005). Emotions and multilingualism. Cambridge, UK: Cambridge University Press. doi.org/10.1080/01434630408666528
Pavlenko, A. (2008). Emotion and emotion-laden words in the bilingual lexicon. Bilingualism: Language and Cognition, 11(2), 147-164. doi.org/10.1017/S1366728908003283
Pavlenko, A. (2012). Affective processing in bilingual speakers: Disembodied cognition? International Journal of Psychology, 47(6), 405-428.
doi.org/10.1080/00207594.2012.743665
Puntoni, S., De Langhe, B., & Van Osselaer, S. (2009). Bilingualism and the emotional intensity of advertising language. Journal of Consumer Research, 35(6), 1012–1025.
doi:10.1086/595022
Teplan, M. (2002). Fundamentals of EEG Measurement. Measurement Science Review, 2(2), 2-12. Schupp, H. T., Cuthbert, B. N., Bradley, M. M., Cacioppo, J. T., Ito, T., & Lang, P. J. (2000).
Affective picture processing: The late positive potential is modulated by motivational relevance. Psychophysiology, 37(2), 257-261.
Van Petten, C., Kutas, M. (1990). Interactions between sentence context and word frequency in event-related brain potentials. Memory and Cognition, 18(4), 380-393.
Wabnitz, P., Martens, U., & Neuner, F. (2012). Cortical reactions to verbal abuse: Event-related brain potentials reflecting the processing of socially threatening words. Neuroreport, 23(13), 774-779. doi:10.1097?WNR.0b013e328356f7a6
