How Does a Turkish Bilingual Alter the Emotional Value Attached to Words in English?

Abstract

A small-scale empirical study of how a Turkish–English bilingual attaches different emotional values to words in her first and second languages, combining a literature review of bilingualism and emotion with participant data analysis and discussion.

Introduction

Emotions are mostly affected by the culture of the bilingual speaker, according to an empirical analysis conducted by Anooshian & Hertel (1994), bilinguals have reported that they have a stronger emotional experience in their first language (L1) than in their second language (L2) or other languages (LX). Could that experience, consistently, cause the effect of carrying the values of certain words from L1 into L2? It is interesting to investigate this emotional value within bilinguals because a multilingual speaker I have experienced these contradicting emotions first hand. This experiment is set up for the module Researching English Language use at the University of Brighton. In order to fully frame and answer the question, the existing literature and research must be covered. This is discussed in chapter 2. To ensure an accurate and fair result, the methodology of any experiment must be considered and carefully planned. The methodology and data analysis in chapters 3 and 4 will cover this in full detail.

Literature review

In this chapter the literature review is evaluated. Concepts are explained in more detail based on literature and other useful sources. It covers two distinct areas: what bilingualism is and different studies that argue the emotional values in L1 and L2. The following sections will deal with these two separately. 2.1 What is bilingualism? The term ‘bilingual’ is used to describe a person who uses two or more languages to communicate (Weinreich, 1953). However, there are various interpretations with regard to attitude, proficiency and use. Degrees of bilingualism range from a minimal proficiency in two languages, to an advanced level of proficiency, thus allowing the speaker to function and appear as a native-like speaker of two or more languages. In order to separate these differences, Weinreich (1953) distinguished between: Subordinate bilingualism: One language is dominant. Words in the non-dominant language are interpreted – and have been learnt – through words in the dominant language. L2 acquired and interpreted through L1. Would tend to characterise low proficiency L2 learners. This could refer to an older generation moving to a country and never fully capture the language as their children. Compound bilingualism: Equivalent words in the two languages have the same meaning (refer to the same concept). Two languages are merged at the conceptual level; one system of meaning. Languages acquired and used in the same contexts. Would tend to be found in minority language communities or with fluent L2 speakers. This suggests participants that use both languages equally and that both languages refer to same concepts of meaning (e.g. German and Dutch have many similarities within the languages and cultures). Coordinate bilingualism: Equivalent words in the two languages have slightly different meanings but refer to different concepts. Languages acquired and used in separate contexts. Languages belong to different communities; little to no interaction between them. Speakers may feel different personalities associated with the languages. Two monolinguals in one person; two systems of meaning. This would characterise an advanced level of proficiency in both languages. Krashen (1982) argues that a multitude of factors contribute to an individuals’ proficiency level: learnings strategies, language transfer, motivation and aptitude of the learner. Krashen believes that motivation and the geographical location of a learner in relation to the L2 language seem to play an important role in determining language success. People move to different countries, because of job opportunities, war, exogamy etc. This can cause varieties in learning and communicating in other languages. These individual differences seem to play an important role in determining language success, since it embraces that learners have different phases in language development. The Turkish participants that live in the UK in this assignment, should identify with the Coordinate bilingualism in Weinreich’s theory. However, the Turkish bilinguals participating from Turkey would fit in the Subordinate bilingualism, as their L1 is dominant and they do not have to rely on their L2 in everyday life. This would suggest that these participants could alter the emotional value easier into their L2, because L1 is more dominant. To give more evidence on this matter, more information is provided in 2.2 of this chapter.

2.2 What are the emotional values in L1 and L2? Certain emotions are universal and experienced in similar ways in all cultures (Anooshian & Hertel, 1994). Other emotions rely on the way they are perceived by the surrounding society and show substantial cultural differences in the way they are experienced throughout childhood. This suggests that it could cause the bilingual child to become bicultural, as they perceive one culture inside the classroom and another at home, similar to their language learning style. According to Anooshian & Hertel (1994), emotions are more deeply culturally influenced: overall L1 dominates L2 or LX and that words in a second language may have deficiencies in the emotional associations of words acquired in childhood. This suggests that the results of the UK bilingual participants are more closely aligned with the results of the Turkish participants. However, it is also important to contemplate what the participants consider their first language to be. If they consider English to be their L1, then their answers are more closely aligned with the English monolinguals. Conversely, Aycicegi and Harris (2004), discussed that emotion-memory effects are found in both L1 and L2, and that it is in fact stronger in the L2 due to the emotional associations of words acquired in childhood. These findings are aggregated and do not consider the importance of individual differences between participants. Three other studies conducted by Dewaele (2004, 2007, 2008), are important to discuss in order to understand why this empirical analysis was set up. He conducted two online questionnaires about the perceived emotional force of taboo words (Dewaele, 2004), and which language bilinguals prefer to swear in (Dewaele, 2007). The first study showed that multilinguals found the emotional value of the taboo words and swear words were higher in their L1. This effect was reduced for languages that were learnt at a later age. In his second study, Dewaele found that multilinguals swear in their dominant language. The dominant language is determined by the frequency of its use, this could be L1 for some of the participants and L2 for the others. The majority of the participants in his study found that L1 has more strength to express themselves and they felt more emotionally attached when using it (which supports Anooshian & Hertel (1994)). What should be considered is the fact that these studies were conducted in the country of the researcher and not in the countries of the participants’ L1. Therefore, it could mean that taboo words are used in L1 because the meaning has more strength due to the fact that the person that is referred to (or the object) does not understand the language of the user. In addition, it could similarly be considered that it is easier to use taboo words in L2 than L1, because it does not carry the same weight to the speaker. However, this does not mean that it is less harmful to the recipient because the recipient would receive that insult in their L1 and attach their value to it. The opposite of taboo words is one of the strongest emotions that humans in all cultures experience: love. In 2008, Dewaele ran an online questionnaire with 1459 multilinguals on how they experience love by measuring the variables of: age, self-perceived language dominance and context of the acquisition of L2. By answering questions about the emotional value of ‘I love you’ in all the different languages of all the participants, Dewaele found that the majority of the respondents found this phrase conveyed the strongest emotional value in their L1, and that no other language could match to this feeling. When it is the value of love, this research has shown that multilinguals prefer to use their L1. For this reason, I may surmise that the Turkish speakers respond more affectionately about the word ‘peanut’ in English, and less affectionately about ‘a bit of crumpet’ in Turkish.

Methodology

The experiment was presented as an online questionnaire, to investigate whether Turkish bilinguals alter the emotional value attached to words between their L1 and L2. The studies presented in chapter 2 had different variables; they looked at culture, taboo words, and love. I have selected the word ‘fıstık’ which means peanut in Turkish. This word carries the same value as the nut that can be consumed and also to address someone you admire by meaning they are attractive and beautiful. In order to compare it with the English language, I have selected ‘a bit of crumpet’ which in a way carries the same value. There may be a difference in how certain cultures feel about certain emotions, for example ‘peanut’ is seen as positive in Turkey, but negative in Holland. Also, other terms of emotions only exists in one language and not in the other (e.g. ‘delidolu’ in Turkish, literally translated: full of craziness, and coveys the value that someone is extremely fun). These differences demonstrate that it may depend on the language that is spoken. I will investigate if Turkish bilinguals attach the same emotional value for ‘peanut’ and ‘a bit of crumpet’ in L1 and L2. According to the evidence provided in chapter 2, the word ‘peanut’ in English may have a more empathic value for the Turkish speakers as the L1 should have more strength than the L2. Whereas, ‘a bit of crumpet’ may not have any influence in Turkish (L1). Other claims are analysed in this empirical task: The negative control question should be answered with the same result in both languages by all participants. Questions are set out to engage both male and female participants, therefore, the results may vary according to gender. The results will not be divided by gender, they will focus only on how the language is used.

Participants, Design and Procedure

Data was collective quantitatively via a questionnaire. This questionnaire will ask the participants to answer questions based on discrete numerical values rather than a qualitative approach such as open questions. Data was gathered anonymously online for multiple reasons: To avoid performance anxiety whereby the participants’ scores were somehow altered by the knowledge that they were being marked or judged To avoid potential grading bias (Malouff, 2008). To ensure that the form of questioning suited the lifestyles and situations of the voluntary participants. A five-minute questionnaire is much more suitable that a one-hour interview, especially when participants reside in a different country and time zone. The names used in this paper are not the real names of the participants, however, they will correspond to their mother tongue. When referred to the groups within this research: English Monolinguals (E), English / Turkish Bilinguals (ET), Turkish / English Bilinguals (TE), Turkish Monolinguals (T). All answers are discussed in English; all original results are in the appendices. In order to analyse the English language use of the Turkish participants that live in England (ET), I will have to compare their answers for the questions in English with English monolingual speakers (E). To analyse their Turkish language use, I will have to ask the same questions in Turkish and compare those answers with Turkish monolinguals (T). To make a distinct separation between Turkish English speakers in England and in Turkey, I will compare their answers with Turkish participants that live in Turkey and speak English (TE). All groups are given the same control questions with the underlying presumption that all participants should answer in the same way: You are in a café and you ordered a salad. The waiter asks you if you would like cucumber in your salad. How would you evaluate: Cucumber You are in a café and you ordered a salad. The waiter recommends you cock-salad. How would you evaluate: Cock-salad The meaning of the word “hiyar” (cucumber) has adopted the same double meaning as the word “cock” in English. The people in Turkey use the word ‘salatalik’ instead of ‘hiyar’, because the second meaning has become more offensive or funny depending on the context (Wikibooks, no date). The word ‘salatalik’ literally means ‘something to put in the salad’. For all four groups separate questionnaires were created. The monolinguals have the exact same questionnaire in their mother-tongues (all questionnaires are provided in appendix 1). According to a study created by Whyatt (2007), it is of importance for bilinguals to start the questionnaires with their everyday language, because it could create a more comfortable and confident environment. Subsequently, they were presented with a warning before switching to the other language and answering the exact same questions (see figure 1). Figure 1 The total of 11 participants were presented information about the study and by clicking next they all agreed to volunteer freely to this experiment (see figure 2, and Appendix 1).

Figure 2 The participants volunteering from Turkey were presented the exact same screen in Turkish. In order to analyse the answers, I require information about the participants regarding: age, gender, first language, and country of birth. All questions are applicable for all genders in order to transcribe their answers. The participants: Age Gender First Language Country of birth Richard (E1) Male English England Adam (E2) Male English England Leyla (ET1) Female English Turkey Ali (ET2) Male English England Berkan (TE1) Male Turkish Turkey Nihal (TE2) Female Turkish Holland (lives in Turkey) Pinar (TE3) Female Turkish Turkey Kamil (T1) Male Turkish Turkey Kenan (T2) Male Turkish Turkey Necla (T3) Female Turkish Turkey Hulya (T4) Female Turkish Turkey Figure 3 All participants were born and raised in the countries corresponding to their L1 language apart from the following two participants. Leyla (ET1) was not born in the UK, she moved to England at the age of 10 and considers English to be her first language. Nihal was born in the Netherlands and moved to Turkey at the age of 19 and has been living there for the last 10 years, and considers herself and her first language to be Turkish. These differences are important to consider when analysing the results Prior to the test, the participants were presented with a warning to prepare them: the number of questions and the amount of time per question (figure 4). However, the bilinguals must switch half way through the test and answer the same questions in the other language. The warning text was repeated in the new language.

Figure 4 At the end, the monolinguals are thanked for their participation and presented an email address for any other queries relating to the experiment, however the bilinguals are asked the following (see figure 5), and are asked to provide an answer in the language they feel most comfortable with:

Figure 5

Results and Data Analysis

Figure 6 shows all possible answers that participants could select from: In English In Turkish Strong complement (SC) Büyük iltifat Complement (C) Iltifat Very affectionate (VA) Oldukça sevgi dolu Affectionate (A) Sevgi dolu It has no meaning (IHNM) Hiçbir anlamı yoktur Normal (N) Normal Offensive (O) Hakaret Very offensive (VO) Kötü iltifat Threatening (T) Endişe verici Other (Or) Başka Funny (F) Komik Hilarious (H) Cok komik Figure 6 Figure 7 shows the questions and provided answers for all participants. The monolinguals are colour-coded in black. The answers given in English by the bilinguals are blue, and the answers given in Turkish are orange: Question IHNM Or

3 You are walking in the park. A group of men pass you and one of them calls you: 'a bit of crumpet'.

4 You are walking in the park. A group of women pass you and one of them calls you: 'a bit of crumpet'.

5 You are walking in the park. A group of men pass you and one of them calls you: peanut.

Question IHNM Or 6 You are walking in the park. A group of women pass you and one of them calls you: peanut.

7 You are passing a construction site. One of the male construction workers calls you: a bit of crumpet.

8 You are passing a construction site. One of the female construction workers calls you: a bit of crumpet.

9 You are passing a construction site. One of the male construction workers calls you: peanut.

ET2TE1

10 You are passing a construction site. One of the female construction workers calls you: peanut.

11You are in a café and you ordered a salad. The waiter asks you if you would like cucumber in your salad. How would you evaluate: Cucumber"

12 You are in a café and you ordered a salad. The waiter recommends you cock-salad. How would you evaluate: Cock-salad"

Figure7 All the answers provided by the participants are available in the Appendix 3 (sheet 5, or raw data online). All the answers are colour coded in the appendix, which can give a better overview of the languages. Figure 8 shows the optional information provided by the participants at the end of the questionnaire: Adam (E2) Personally I think being called a crumpet has connotations of being quite ugly, whilst peanut would be used to describe someone who is excessively small or has an particular feature that is smaller than normal (e.g. head), hence them both being labelled as offensive. Leyla (ET1) Switching to Turkish was different as 'pogaca' doesn't really have a meaning, but it felt like if someone would say that, it would feel like they were making fun of me for some reason. Ali (ET2) It was a bit strange as they don't have the same meaning in the two languages Berkan (TE1) İngilizce biraz daha farklıydı sanki = The English version felt a bit different. Nihal (TE2) Evet çünkü tamamen kültür farkından dolayı. Bizim kullandığımız kelimeleri onlar aynı anlamda kullanmıyorlar = Yes, it was different due to cultural differences. The words we use, they do not use them the same way. Figure 8 Figure 8 shows an interesting insight into the selection of language used by participants. Even when the participants ended the questionnaire in their L2, and were given free choice of language to answer, all participants chose the language of the country in which they currently reside.

Discussion and Development

All participants recognised the negative control answers as: funny, hilarious or normal. However, the responses were not entirely as expected, as the majority of the Turkish speakers answered Normal to Cucumber salad in Turkish. Only T1 and T2 considered it to be Funny. The reason for this could be that the participants considered this to be an official university questionnaire which would not contain disrespectful words and applied the original meaning of the word ’hiyar’ to answer the question in Turkish. Richard and Adam have both responded differently to all questions (except for the negative control). Adam (E2) seems to have a different understanding for the word ‘peanut’ and the expression ‘a bit of crumpet’(figure 11), this could be because he was never introduced these expressions. After the results were collected, I asked my tutor if this could be age related. I was informed that a younger generation might not know the emotional value attached to these expressions, if they have not heard it in their households. For this reason, I will exclude Adam’s results in the further discussion. Berkan’s (TE1) response to the last question was ‘The English version felt a bit different’, however his responses to all questions were identical to those of Adam in both languages. Both participants responded to every question Offensive and Very offensive. Therefore, I will exclude his results in the rest of the discussion. Analysing the results there is one thing that can be said with all certainty. Everybody has different opinions. Where one finds one expression a complement another person finds it threatening, an insult, or thinks it has no meaning. Therefore, I will only discuss questions and answers that did have interesting results where the participants did seem to agree. Comparing the results of Leyla (ET1), Ali (ET2) with the English monolingual Richard (E1): For question 1, all three participants seem to agree that the expression ‘a bit of crumpet’ is affectionate. For question 2, Adam and Ali seem to agree that the word ‘peanut’ has no meaning, however Leyla seems to find this affectionate. This could be because she thinks the word peanut means affectionate in English, or she is altering the emotional value from Turkish to English. Although, she has said to consider her first language to be English, it still could have affected her language use as she moved to the UK at the age of 10. Comparing the results of Leyla and Ali with the Turkish monolinguals: For question 1, both groups (ET / T) seem to agree except for Kenan (T2), he finds ‘pogaca’ (a bit of crumpet) offensive, while the rest agree that it has no meaning in Turkish. This could be because of individual differences, or someone has called him ‘pogaca’ in an offensive situation. However, looking at his other responses, he has chosen ‘it has no meaning’ to the rest of the questions which involve ‘pogaca’. The difference in opinion could be due to many factors, one of which is invalid user input. Overall, the rest of the participants (ET / T) seem to agree that it does not have a meaning. For question 2, both groups seem to agree that the word ‘fistik’ (peanut) is a complement or affectionate, except for Necla (T3), she finds it threatening. This could be because she had a bad experience with this word, or she might have incorrectly selected an answer on the questionnaire. Looking at her other answers, she answered the word ‘fistik’ to be affectionate when she was called out by other women in questions 6 and 10. In this scenario the data suggests that Leyla separates the values in Turkish, which could suggest that unconsciously her L1 is dominant over L2. Comparing the answers between the Bilingual groups ET and TE, it can be said: For question 1 in English, Leyla (ET1) and Ali (ET2) have responded with ‘affectionate’, and Nihal (TE2) and Pinar (TE3) have responded with ‘normal’ and ‘funny’. This shows that the TE group assume it has a meaning, but are not sure, otherwise they could have chosen ‘it has no meaning’ or other. However, in Turkish the responses are interesting. Where the ET group and Pinar (TE3) agree that it has no meaning, Nihal (TE2) finds it ‘affectionate’, when comparing that to the other answers she has given we can see she has chosen the option ‘other’ to all her answers relating ‘pogaca’. This could mean that on its own she finds that it sounds affectionate, or simply pressed the wrong button. For question 2 in English: Leyla responded with ‘affectionate’ and Ali with ‘it has no meaning’, the TE group seems to find it ‘normal’. This could mean that the TE group alters the meaning of the word ‘fistik’ into ‘peanut’, provided by the evidence given in chapter 2. When L1 is more dominant, the emotional value does alter in subordinate bilinguals, however, seems to have no effect on the coordinate bilinguals. The TE group seem able to separate the emotional value of the words from the act of translation. Thus they can translate a word into their L2 and yet maintain the emotional value of the word according to the definition of their L1.

Conclusion

To summarise, the findings indicate that this experiment needs more participants, research, time and thought to have more conclusive results. Firstly, it was very difficult to find participants for this project as I had to search in two countries. In addition, I understand why it is of importance to have participants that are in the same age group, background, and similar education in any form of collecting data. These variables can have a drastic effect on the results. Similarly, the participants had too many options to choose from, which did have an impact on the results. Secondly, I had to think of other ways to ask the questions. The questions asked involved people that the participants had no relation to. ‘People walking in the park’ and ‘construction workers’ can be intimidating to people. To make a distinction, I had to ask similar questions involving loved ones. As mentioned before, love is the most effective word with the strongest emotional value amongst humans. Thirdly, it would have been beneficial to put more thought into questionnaire creation and participant selection. With hindsight, it would have been better to choose a small same group, perhaps only English mono and bilinguals living in the UK. In addition, it was wrong to presume that everyone in the UK were familiar with the expressions posed. I could have introduced the expressions at the beginning of the questionnaire, to ensure the expressions were understood. Finally, to answer the question: How does a Turkish Bilingual alter the emotional value attached to words in English? Due to coordinate and subordinate bilingualism, it was expected that the ET group would have been more closely aligned with the English monolinguals, as they consider themselves to be English. Although, research discussed in chapter 2, gave evidence that L1 would dominate the emotional value, because the L1 is felt stronger. However, where Ali showed in his answers that he is more closely aligned with the English monolingual group, which shows he has not only adopted the English language in L1 usage, but also adopted the cultural-based values of the English language. Leyla’s results however, were different, her answers showed an imbalance between the two languages. It could be argued that Leyla moving to the UK at the age of 10 did not in fact adjust her (culturally-based) emotional values, it only changed her usage of the English language (it became her dominant language). In conclusion, I have learned an immense amount throughout this project. The following results were particularly insightful: Leyla’s potential unconscious bias, analysis of the final question where participants all answered in their L1 and that the birth country of an individual has a lasting impact on their cultural values, regardless of when they moved to a different country. The results suggest there is a trade-off between a quantitative and qualitative approach to the questionnaire. Larger groups would enable trends to be spotted in a quantitative approach but a qualitative approach was perhaps more appropriate for this study seen the small sample size. The number of possible answers should also be reduced in future studies, and most importantly, ensuring all participants have knowledge of the question areas (the colloquialisms used) before taking the test. (Word count: 3917)

Bibliography

Anooshian, L. J. and Hertel, P. T. (1994) 'Emotionality in free recall: Language specificity in bilingual memory', Cognition & Emotion, 8 (6): 503–514. Aitchison, J. (2012) Words in the Mind. Hoboken: John Wiley & Sons Press. Aycicegi, A. and Harris, C. (2004) 'BRIEF REPORT Bilinguals’ recall and recognition of emotion words', Cognition & Emotion, 18 (7): 977–987. Aycicegi-Dinn, A. and Caldwell-Harris, C. L. (2009) 'Emotion-memory effects in bilingual speakers: A levels-of-processing approach*', Bilingualism: Language and Cognition, 12 (03): 291. Bamberg, M. (1997) 'Language, concepts and emotions: The role of language in the construction of emotions', Language Sciences, 19 (4): 309–340. Caldwell-Harris, C. L. and Ayçiçeği-Dinn, A. (2009) 'Emotion and lying in a non-native language', International Journal of Psychophysiology, 71 (3): 193–204. Dewaele, J.-M. (2004) 'The emotional force of Swearwords and taboo words in the speech of Multilinguals',Journal of Multilingual and Multicultural Development, 25 (2–3): 204–222. Dewaele, J.-M. (2007) 'Blistering barnacles! What language do multilinguals swear in?!', Sociolinguistic Studies, 5 (1): . Dewaele, J.-M. (2008) 'The emotional weight of I love you in multilinguals’ languages', Journal of Pragmatics, 40 (10): 1753–1780. HARRIS, C. L., AYÇIÇEGI, A. and GLEASON, J. B. (2003) 'Taboo words and reprimands elicit greater autonomic reactivity in a first language than in a second language', Applied Psycholinguistics, 24 (04): Levelt, W.J.M. (2013) A History of Psycholinguis-cs: The Pre-Chomskyan Era. Oxford: Oxford University Press. Malouff, J. (2008) 'Bias in grading', College Teaching, 56 (3): 191–192.) Turkish/slang - Wikibooks, open books for an open world (no date) available at https://en.wikibooks.org/wiki/Turkish/Slang, accessed 27 April 2016. Weinreich, U. (1953). Languages in contact: Findings and problems. Mouton: The Hague. Whyatt, B. (2007) 'Two languages, Two cultures, One mind: A study into developmental changes in the students’ view of language as a tool in cross-cultural communication', Poznań Studies in Contemporary Linguistics, 43 (2)

6. Appendices

The appendices can be accessed by clicking on the icons. The hardcopy will contain the actual pages.

← Back to Papers