1 Cross-cultural research has become a popular genre in the social sciences. Researching people’s experiences from different socio-cultural, geographical, and linguistic settings offers an insight into different ways of knowing, but also exposes researchers to some challenges (Pereira, Marhia & Scharff, 2009). Researching and translating across cultures is not an innocent act; it requires a constant level of reflexivity and ethical considerations through all stages of the research. In this paper, we discuss the complexities of translating narratives cross-culturally. This is a collaborative effort which started during a discussion at the National University of Ireland Galway Narrative Study Group about the role of translation in our respective work and the problems we encountered. We decided to explore our individual experiences during three research stages—research protocol formulation, field work, and data analysis—through monologues. The three monologues merge in a trialogue that considers the challenges of cross-cultural research.

2 First we follow TK in the steps she took in planning her research focused on translation of the concept of resilience from Anglophone academic frames to the cultural context of Slovenia. TK argues that an autoethnographic approach to questions of translation provides a research tool for an in-depth exploration of the construction of meaning across cultures. Autoethnography helped her to further develop a methodological framework, which evolved around the importance of narrative interviewing in exploring resilience across languages and cultures.

3 AL brings us to the interview moment and discusses the need for and limits of reflexivity when dealing with the power imbalances brought in by multilingual sites and post-colonial research contexts. Her research was conducted in Morocco, and all of the participants were fluent in at least two languages, most of them in more than three. AL herself conducted most of the interviews using her third language. She discusses how interview situations affect self-narrating and the power to present the other differently by discussing three interview scenarios: using an interpreter; interviewing a participant in a language in which the participant has the linguistic advantage over the researcher; and conducting an interview in which, linguistically, the power is with the researcher.

4 Finally, EB considers what happens to politically active women’s narratives of in/security when language is colonised by state-centric definitions. EB concludes that a critical narrative approach to data analysis can help us to understand the productive aspects of power relations within language and translation. A critical narrative approach reveals the connection of women’s experiences and their personal narratives to larger discourses, and therefore puts the meta-micro divide between the personal and the political under scrutiny.

5 The trialogue illustrates that translation of people’s stories is not just a technical act, but bears epistemological implications for the research (Riessman, 2008). Young (2009) claims the process of writing up cross-cultural research goes through a dual transformation: adjusting data to academic jargon and translation into English. The last is not embedded merely in translation of particular words, but also has to represent cultural and local realities to which the language is tied. Research in post-colonial contexts is always influenced by the unequal relationship of the participants and the Western researcher. This means that we must remain cognisant of this imbalance throughout the research process and try to meet the other and listen to her through her own language (Spivak, 2012). In this paper, the three voices recognise that narratives and their meanings are socially constructed and constantly reformed through different stages of a research process.

6 In the context of our research, narratives refer to a method of inquiry and epistemological considerations of the storied understandings of lived life (Andrews, Squire, & Tamboukou, 2008; Tamboukou, 2008). Narratives are born out of experiences and it is through the connection and retelling of events that we make sense of ourselves and our place in the world (Bruner, 1985, 1990, 1991; Labov & Waletzky, 1968; Ochs & Capps, 2001). Narratives, therefore, play a role in constructing and maintaining identities (Ochs & Capps, 1996) and are always in a process of transformation and becoming, always constructed and reconstructed in the relation between research participants and researchers (Andrews, Squire, & Tamboukou, 2008, p. 14).

7 In this article, narratives emerge in translations across languages and cultures, better described perhaps as “transcreation”: editing, reconciling and transmuting language and culture (Mukharjee, 2009, p. 55). Temple and Koterba (2009) assert translation is more than a transfer of meaning from one context to another, as it involves a translation of people’s lives. The researcher is not only a mediator between cultures, but also a re-constructor of these same cultures through the research. Baker (2005) suggests that translators’ behaviours are driven by the stories they believe in and events in which they are embedded (p. 11). This approach thus situates the translator in the heart of the cross-cultural research and contests the idea of a neutral linguist-researcher who is set in-between the cultures. Translation is “inextricably bound to the socio-cultural positioning of the researcher, a positioning, whether intended or ascribed, that will also give a meaning to the dual translator/researcher role” (Temple & Young, 2004, p. 168). It begins when the researcher considers conducting a study across cultures. The location of a researcher and the shaping of a researcher’s viewpoint are influenced by his or her relationship with the audience (Temple & Young, 2004, p. 164) as well as with the participants.

8 The following three parts of the paper discuss the main issues that the three voices encountered when they approached people’s experiences across cultures through narratives and translations. The first section shows how to begin with the research when the words are not translatable or used across cultures.

9 Having moved from Slovenia to Ireland three years ago to undertake a PhD, I found myself caught up in cross-cultural research, academically based in Ireland and practically situated in Slovenia. The research focused on cultural understandings of resilience and social support in Western Slovenia involved, for me, the spatial movement “back home” to do fieldwork research. As Andrews (2006) suggests, going back home did not mean to return to the known, as new experiences had changed me and I was not the same as when I had left Slovenia three years ago; simultaneously I became an outsider in the new culture and at home (Alsop Kraft, 2002, para. 19).

10 Resilience has been one of the most challenging concepts during my research. I have had several difficulties in translating the concept into Slovenian and I was concerned how to present the idea to the participants. I started to visualise what the term means to me personally and search for approaches to elicit its cultural meaning. Interview questions were initially informed by previous qualitative research on resilience, as proposed by Ungar (2012, p. 26) and the Resilience Research Centre. I translated the questions several times as some of the translations did not sound accurate or depict a correct representation. The problem was not only in translation as such, but also in its applicability to the Slovenian cultural setting. The formulation of the question “What would I need to know to grow up well here?” when translated into Slovenian indicates that the answers will be detached from people’s personal experiences and also instigates a debate what “growing up well” means in a particular socio-cultural and personal context.

11 Similarly, the other questions developed by Ungar (2012) and colleagues offer a general guideline how to approach the idea of risk and resilience across cultures, but I personally found them hard to use in the interview process. They seemed to be disassociated and disconnected to people’s lives. Therefore, I recognised a need for more engaging interview questions, which would offer an insight into interviewees’ personal experiences as people’s responses are embedded in the cultural framework they belong to and do not necessarily correspond to the normative views of other cultures.

12 After substantial reading about the topic and making initial decisions connected with the recruitment of the research participants, I realised that I can position myself as a member of the generation that grew up during the transition from Socialist to Post-socialist Slovenia in the 1990s. I started to ask myself what risk and resilience mean to me personally. I also started to consider how I would verbalise the concepts and if the participants from my generation would identify with my experiences. Additionally, in order to combine biography and history, I found crucial the relevance of an autoethnographic approach in which private experiences offer an insight into public matters (Bullough & Pinnegar, 2001, p.15). In short, I realised I had to involve myself in the research process.

13 Autoethnography is a recent and still developing approach to qualitative research. It has emerged as a response to the research of other cultures and shifted to the study of ourselves (Hayano, 1979). It is “an approach to research and writing that seeks to describe and systematically analyse (graphy) personal experience (auto) in order to understand cultural experience (ethno)” (Ellis, Adams, & Bochner, 2011). It is a form of narrative analysis as the researcher aims to produce a story that is relevant to the research interests (McIlveen, 2008). Similarly to narrative inquiry, it provides a convenient tool to investigate the researcher’s meanings of experience (Trahar, 2009). Autoethnography recognises a researcher's influence, stance, and relation to the research topic. Individual experience is recognised as an essential part of a broader social world (Mykhalovskiy, 1996). Andersen (2006, p. 387) argues that the method is used not only to depict emotional experiences, but also to access wider social phenomena through personal data.

14 In order to link personal and cultural understandings of resilience, I incorporated autoethnographic methods into my research project (Adams, 2007) by giving an interview about life challenges and how I coped with them. I was interviewed in English by a colleague who was familiar with my research topic. Autoethnography was used as a tool which helped to minimise translation-related confusions and to develop further interview questions. The questions were initially developed according to the recommendations of Resilience Research Centre (as shown above) and based on critical analysis of national and international literature.

15 I realised that in order to understand the research topic, I have to approach the conversation and its transcription through its meaning and context. The insight into my own experiences helped me to realise that personal narratives provide a solution for translating risk and resilience across cultures. Therefore, in order to elicit personal stories situated in a particular social context and time, the idea of technical, verbatim translation of pre-designed interview questions was replaced by introduction of a vague narrative question. Personal narrative provided a first layer of cultural understanding of coping and resilience. For instance, I became aware that transition from secondary school to university was my personal challenge when growing up. The interviewer decided to explore this experience and asked me which strategies I mobilised in order to cope with the uncertainty:

16 This developed new ideas about risk and resilience, situated in a specific social context and time. Personal narrative became linked with the broader social context, which can be examined through the idea of educational transition and moving to another place in order to have an opportunity to study. Did young people at the time experience similar situations and how did they deal with them? The verbalisation of personal risks and challenges in another language led to the realisation that personal narratives can only be explored in their cultural contexts. I realised that my personal growing up experiences in the region and closeness to the research participants are useful to understand and translate the social context in which people have lived their lives. Bourdieu (1996) found that forced questions instigate artificial answers, and thus the interview process has to be conversational and contextualised. Hence, it came to light that I have to “contextualise” the questions according to participants’ personal story.

17 Within that context I conducted 20 in-depth, semi-structured interviews. I used a narrative approach to interviewing, which provided an insight into personal growing up experiences in the region. This experience-centred approach (Squire, 2008; Hollway & Jefferson, 2001) is based on a search for a story within a person and enables exploration of research concepts by focusing on personal examples. The interview process was initiated with the following open-ended question—“What are your experiences with growing up in the region?”—followed by a semi-structured interview guide which followed the story line of each interviewee. I also used different scenarios, asked about experiences other people they knew had with life challenges, and the narrative of resilience was constructed through such discussions. All those questions became meaningful when they were integrated into people’s personal lives and asked in a culturally relevant way.

18 Baker (2005) claims there is a need to “recognize and acknowledge our own embeddedness in a variety of narratives” (p. 12). This predisposition was applied to a further phase of the research, when a combination of my personal and the interviewees’ narratives of coping and resilience were constantly negotiated and reconstructed through the interview process. The way people express themselves depends on the language they use (Temple, 2009). Even though I am a member of the researched community, the way I speak and use Slovenian does not necessarily correspond to its use by other research participants. Eco (2003) defines the process of translation as negotiation in which a researcher has to decide how to present people’s experiences. Therefore, the role of the researcher is not merely to translate words in a technical, verbatim way, but to consider how to present experiences and people’s lives across cultures. In my own research, the participants’ narratives were constructed through my own and participants’ experiences with risk and resilience. Our positions have developed independently, through embeddedness in different contexts and narratives. My own position stems from academic and personal perspective. As a PhD researcher based in an English-speaking university, I had to join ideas recognised as valid in the academic space with personal stories in Slovenian in the community in which I grew up. Those views were many times challenged, but also reconciled by narratives of the other participants.

19 In order to understand the negotiation of translation better, it is important to unpack the role of the researcher on the field. In the next monologue, AL discusses issues arising when conducting research in multilingual settings. Whether using interpreters or asking participants to self-narrate through translation, presenting narratives of others requires understanding of the research context and reflexivity of the processes.

20 My monologue concentrates on issues of the multilingual field and its effects on the narratives that are produced through interviews. I conducted 24 interviews with women’s NGO activists in Morocco. One of the interviews was conducted in English and two were conducted in Arabic using an interpreter. The remaining 21 interviews were conducted in French. Many of the participants held a linguistic advantage in French, which I believed would shift the power relations in the interview situation (Huisman, 2008).

21 The issues of translation I dealt with during my field research challenged me to question where to locate the voices of the participants in this study. As Young (2009) suggests, translating involves transforming someone or something from a subject to an object. Power must then be understood as part of the dynamics in the interview situation, but also as the power to present oneself or the other. The issue of multilingualism and social understandings of language use and literacy in Morocco are closely linked to the relationship between Orient and Occident but also encompass issues of representation in respect to class, gender, education, and linguistic capabilities (Sadiqi, 2003). The official language in Morocco is Standard Arabic, but Moroccan Arabic and French are widely spoken. While most of the participants had little difficulty in speaking about their activities in French, the narratives were necessarily altered, as they had been transported from their original cultural and linguistic setting in Moroccan Arabic to become “similarly produced” as European feminist narratives (Spivak, 2012, p. 322). Language is not only a system of signs, but as Brockmeier (2008) tells us, “constitutive of our being in the world” (p. 33). Thus, by telling themselves in French, the participants changed the context of their self-narrating. The French narratives of the participants’ agency are no less authentic; they merely show a different version and place them in a transnational context of women’s activism.

22 For two participants, an Arabic interpreter was provided. The situatedness of the interpreter as an insider into the research context cannot be assumed (Temple, 2005), but engaging with the interpreter as a key informant, as Edwards (1998) has suggested, and involving the interpreter in the research and analysis process, can enrich the research. As it is evident in the quotation below, the interviewee was frequently mixing French with Arabic without waiting for translation, as she understood the original question. The interpreter, for her part, was keen to interpret the context as well as mere words. The interview became a three-way conversation, and the line between the narratives of the participant and those of the interpreter became blurred. Although the interviewee spoke Arabic, the context became transnational:

23 This example demonstrates Baker’s (2006) criticism of interpretation as objective bridge-building: the interpreter wishes to ensure the researcher understands the context as well as the words, and she adds her own interpretation of the topic as a third layer onto the conversation. Knowing my interpreter personally helped me distinguish between her words and those of the participant, as did the conversations I had with her regarding the interview. The taped interview nevertheless constitutes a co-produced narrative among the three. Rather than judging the resulting conversation as an invalid research interview, I can use these reflections in enriching my analysis (Ficklin & Jones, 2009).

24 Temple (2005) discusses the problems of different approaches to cross-cultural research and the power relations inherent in them: whether one chooses to include only those who can express themselves directly to the researcher or to use interpreters, and how we finally present the narratives of the participants, all have consequences for our research. The participants’ narratives change according to the language they express them through, as language is central in meaning-making and self-telling (Brockmeier, 2008; Bruner, 2002). Self-telling through translation should not itself be regarded as less true, however, as this would indicate the existence of a “true original,” which especially for people from multilingual backgrounds would be impossible to pinpoint (Slavova & Phoenix, 2011). Power cannot be identified as a static quality that we either have or do not have, but as a dynamic within relationships, including the relationship between researchers and participants (Bacchi & Eveline, 2010).

25 Although most of my interviews were conducted in French, Arabic is present in all of them. Some words and local terminology, such as references to particular laws, are always made in Arabic, but these brief interferences in Arabic were not able to invert the power imposed by the francophonisation of Moroccan political, educational, and social systems. It was hence impossible to escape the post- colonial power structures imposed by language onto my research participants. The eloquence with which many of the participants expressed themselves compared to my own foreign accent and errors of grammar and syntax in French allowed me to feel as if these power relations were balanced. One of the participants, Fatima, also used her own digital recorder to maintain a copy of the interview for herself, turning the mirror back to me as an object:

26 The interview, which I conducted in English, forced me to rethink this idea of shifted power balances. During this interview my own position as a Finn, as a non-native speaker of a dominant language, was tested. On one hand, I found myself constantly reassuring the interviewee that her English was very good, taking on the role of the fluent speaker who holds authority over the fluency of others. On the other hand, my accent became more Finnish and expressions less fluent, as if to highlight my non-native status as an English speaker. Suddenly, I was not sure how to phrase my sentences and pronounce words, as I did not want to sound too “native” and felt that my usual accent was too difficult to understand:

27 This partly conscious, partly unconscious, performance of Finnishness through pronunciation, syntax, and vocabulary gave a different view of the power struggles my research presented. In the interviews I conducted in French, hiding behind a Finnish accent allowed a sensation of being removed from Morocco’s colonial history. Of course, a researcher cannot escape the power relations between languages and their speakers (Lutz, 2011). In English, my Irish lilt, usual fast pace of speaking, and extensive vocabulary put me in a clear position of linguistic advantage. My research project, my funding, my privileged educational background, and my freedom of movement had not changed, but the linguistic experience made me fully appreciate the power position of a Western researcher in a post- colonial African country.

28 The participants’ bilingualism means that they are in possession of narratives of their agency in both languages. Self-telling is greatly influenced by cultural and literary traditions (Bruner, 2002). The way women activists tell their stories also intersects with wider institutionalised meanings of gender (Hemmings, 2011). Thus, the stories told in French must be understood in the political and cultural context of the French language in Morocco and of translational women’s activism with which these narratives communicate. For example, when a participant talks about the feminist identity of her NGO, I must remember that the word does not exist in the Arabic lexicon; the narrative construction of self as a feminist is located in the speakers’ situatedness in a francophone education system and transnational women’s activism. Power relations play out in interview situations and are influenced by the choice of language. As I came to realise, however, participants’ advantaged position in the interview situation is not enough to balance the unequal power relations created by my location in Europe and the power I have in presenting narratives of research participants. Although representing participants through translation may be impossible, knowledge of the context in which the narratives are produced and in which they interact is important (Temple, 2005, para.1.3). Similar questions are evident from EB’s monologue, where she discusses translation and analysis.

29 This monologue focuses on the process of data analysis. It explores what happens to narratives during the analysis process when they are translated multiple times. I also reflect on how meaning is transmitted cross-culturally as translation takes place. In particular, I am interrogating what happens to politically active women’s narratives of in/security during cross-cultural translation. In my interviews I have identified that a shift in meaning often occurs when Anglophone terminology is transferred from an academic Anglophone setting to a local non-Anglophone context. This is done through critical narrative analysis, which combines narrative analysis and critical discourse analysis by, in short, connecting individual’s narratives to larger discourses (Souto-Manning, 2005).

30 My research investigates the relationship between women’s political activism and security. I conducted my field research in Kashmir, 4 the conflict-affected area in northern India. In this project I am working in multiple languages simultaneously. The women participants have Kashmiri as their mother tongue and are mainly schooled in Urdu, though the ones who have university degrees have good knowledge of English. Whilst I do speak basic Hindi and Urdu, as well as even more basic Kashmiri (my mother tongue is Swedish), I am not competent enough to do interviews in Urdu or Kashmiri. The women who use English on a daily basis agreed to have a conversation in English, while the other participants were interviewed with the aid of an interpreter. Working cross-linguistically among these three languages, Kashmiri, Urdu and English, as well as cross- culturally among Kashmir, Ireland and my native country, Sweden, created several obstacles in many obvious ways. In Kashmir, as in other parts of South Asia, English is the elite language, used in political and intellectual circles, and is also the official language in which political affairs are conducted (Mohan, 1989). This English is always mixed with Urdu and Kashmiri. English-language words such as “Security Forces,” “Unidentified Gunmen,” “Military,” and “Army” are used instead of their Urdu or Kashmiri equivalents. This is what I refer to as “the language colonised by state-centric definitions.”

31 To understand the functioning of security in Kashmir I draw on Critical Security Studies, which focuses on discourses and identity (Der Derian & Shapiro, 1989; Williams, 1998). From this perspective, then, security is considered a discursive practice which produces its subjects, subjects of security. Defining and naming what is a threat or a danger involve inclusionary and exclusionary practices, which inform and recreate identities (Stern, 2006, p. 182). Additionally, as the women’s stories in this research demonstrate, security and insecurity are deeply interlinked and cannot be separated. Multiple layers of foreign languages added to the women’s narratives reiterate the state-centric position of security, with the effect of making it visible while enabling us researchers to deconstruct it.

32 The interviews centred on women’s lives as activists and specifically focused on their experience with security and what Kashmiri identity meant to them. Afterward, the data was analysed through critical narrative analysis (Souto-Manning, 2005), placing the security narratives of the interview participants within three spatio- temporal locations: personal life, organisational life, and national life (Stern, 2005, p. 65). These spatio-temporal locations function as frames for the women’s in/security narratives. These frames are ordered by different sets of rules, regulations and discourses, and are regulated by power (Butler, 2009; Wibben, 2011). A narrative approach simultaneously highlights stories of disruption and resistance, thus bringing forward women’s agency. Simultaneously, taking account of narratives in translation, the interviewed women partake in giving meaning to security; when the term “security” travels from the interview sheet created by the critical security scholar to a female activist in Kashmir, though always working in English, the translation from one context to another displaces its meaning.

33 In the stories the women told me, the foreign language of English is being domesticated (Temple, 2008). The non-native English-speaking women translate “security” into meaning “the security forces.” Therefore, their answers do not refer to security as the lack of vulnerability but instead what the security forces make them feel; that is, the productive effect of security. For Rabiya, a human rights advocate in her late 20s, security is located on the street:

So, Rabiya understands the word “security” as referring to the security forces present in Kashmir. Another woman I interviewed, Farhana, had a similar understanding of the meaning of “security”: “…security what is they’re securing is my insecurity. … Whenever I confront the so-called security forces [inaudible]. So it gives me a sense of insecurity.” Thus both Rabiya and Farhana understand “security” as a threat. There is a shift in the meaning of the word “security” during the translation from the academic context to the local context. For me, when I was preparing my research and composing the interview sheets, I understood the security/insecurity nexus to involve experiences, representations, and the productive effects of safety, vulnerabilities, and dangers. In Kashmir, where the Indian army is omni-present, “security” is understood as the presence of security personnel. Hence, the word “security” has been domesticated and obtained a meaning that makes sense and is relevant to the local context (Temple, 2008).

34 Other women I interviewed had a different understanding of security. Asifa, in the example below, understands security as her lack of it. Asifa has been engaged in social activism since the 1990s, mainly focusing her work on women’s health. She has never thought of what security means to her: “This is a question I’ve asked many people [laughs]. I’ve never thought what it actually means to me.” After some consideration, she finds, nevertheless, that security means to her:

35 In her narrative, the state of insecurity, the gun culture, and the paramilitary troops discursively reproduce her as a woman, as unmarried, and as vulnerable. She juxtaposes this to her work as a peace activist and thus highlights the contradictions of what she feels when she walks down the streets and pursues her work. This displays the displacement of the meaning of security, from my academic perspective to her real life experience. It is clear then that “changing languages involves “translating lives rather than simply words” (Temple & Koterba, 2009, p. 2).

36 So, the colonisation of language—in this example the multiple understandings of the word “security”—transfers and transforms meaning between multiple languages and cultures. Security is understood in its many literal senses: as “Security Forces” and the lack of vulnerability. The different meanings of security thus highlight the ever-presence of the Security Forces, which in turn creates both a real threat of insecurity as well as discursively producing insecure, gendered citizens. A narrative approach can help in connecting women’s experiences and their personal narratives to larger discourses, and therefore link the meta-micro divide between the personal and the political. Nevertheless, the gap in meaning still prevails. A narrative study highlights how in/security discourses operate on multiple levels: women’s stories disrupt state-centred discourse but also demonstrate how these narratives produce security subjects. This serves to, on the one hand, re-impose state-centric narratives of in/security, but also, on the other hand, provide a stage for women’s activism.

37 In this trialogue, we discussed some complexities of translating meaning across cultures coming from different cultural backgrounds and research interests. The exploration of narratives in multilingual settings is a challenging process which involves methodological and epistemological consideration in order to tackle these complexities during the whole research process.

38 Narration is a relational experience between a researcher and research participants. Both sides enter into research with a narrative in mind. TK’s autoethnographic approach evolved around production of her own interview to get insight into a personal narrative of resilience, while the common narrative of coping and resilience have been constructed through relations and experiences with the research participants. For EB, though the majority of interviews were conducted in English, the translation of the main concept of “security,” from the academic context to the Kashmiri context, demonstrated a displacement of meaning as participants understood “security” differently from how EB initially had defined it.

39 The choice of language in the interview situation has great implications for the power of representation. Even when the participants are in full command of the language of self-narration, the final act of representation remains with the researcher and power relations within cross-cultural research cannot be equalised, as shown by AL. The narratives told by the Kashmiri activist women in EB’s part suggest that translation is always political, but that narrative methods can help link personal stories with political discourse. Translation is not simply about words, but about people’s lives (Temple & Koterba, 2009, p. 2) and researchers will need to focus on how people use languages and represent themselves.

40 The researcher’s role is to reflect constantly on his or her position in the research. As researchers, we do not forget the influences of other narrative contexts that have an impact on our research, but we have to reflect on their influence on the construction of a “common narrative.” Reflexivity cannot begin or end at writing a thesis, but must be an integral part of field research. Change of language in the interview situation in Morocco allowed AL to challenge her idea about language advantage as the only source of power in research interviews. This realisation of the impossibility of equalising power relations in research should not stop us from trying to represent our participants, however, “not because it’s possible, but because one must try” (Spivak, 2001, p. 14).

41 This trialogue produced a conversation about challenges on translations as narratives constructed through all phases of research. The discussion demonstrated how the researchers have not tried to bridge the gap between different cultures and languages, but how we tried to “mind the gap” (Baker, 2005, p. 12) in order to present people’s experiences as fairly and accurately as possible.