Abstract
With words as data, qualitative researchers rely upon language to understand the meaning participants make of the phenomena under study. Cross-language research requires communication about and between linguistic systems, with language a site of power. This article describes the use of the lingua franca of Tok Pisin in a study conducted to explore the implications of male circumcision for human immunodeficiency virus (HIV) prevention for women in Papua New Guinea. Utilizing a transformational grounded theory methodology, researchers conducted an analysis of data from an HIV prevention study. Researchers then facilitated individual interviews and interpretive focus groups to explore preliminary categories identified during the analysis. Most focus groups and interviews were conducted in the local lingua franca Tok Pisin, which is neither the researchers’ nor most participants’ first language. Audio recordings were transcribed and analyzed. Researchers returned to research participants to discuss research findings and recommendations. Following critical reflection by the authors and further discussions with participants, it was evident that using Tok Pisin enriched the research process and findings. Using the lingua franca of Tok Pisin enabled interaction in a language closer to the lived experience of participants, devolved the power of the researcher, and was consistent with decolonizing methodologies. Participants reported the use of Tok Pisin, em i tasim (pilim) bun bilong mipela, “it touches our bones,” and enabled a flow of conversation with the researchers that engendered trust. It is critical researchers address hierarchies of language in order to enable cogeneration of quality research findings.
Background
Qualitative research seeks to understand the meaning participants make of a specific phenomenon (Mills & Birks, 2014). With words as data, qualitative researchers rely upon language to communicate with research participants about the phenomena under study. When research is undertaken in cross-cultural situations where multiple languages are spoken, the qualitative researcher needs to consider cultural implications of the use of language, pay attention to which language is commonly spoken by and among research participants, and which language can best communicate key concepts (Mertens, 2009). Language provides “primary access to people’s experiences…and the production of these data requires an awareness of the issues involved in language expressions of experience” (Polkinghorne, 2005, p. 139). Language considerations are therefore fundamental to good quality research.
Cross-language research refers to a situation where there is a language “barrier” between a researcher and research participants (Squires, 2009). Cross-language research requires the researcher communicate about and between linguistic systems of meaning with research participants. Both oral and/or written translation can be used in cross-language research and underpins successful research outcomes (Temple & Young, 2004). The importance of how and when to use translation for successful qualitative research has been increasingly explored in methodological literature globally (Björk Brämberg & Dahlberg, 2013; Court & Abbas, 2013; Esposito, 2001; Hsin-Chun Tsai et al., 2004; Larkin, Dierckx de Casterlé, & Schotsmans, 2007; Mafile’o, 2005; Tamasese, Peteru, Waldegrave, & Bush, 2005; Temple, Edwards, & Alexander, 2006).
Language is a site of power in research (Temple & Young, 2004). Arguably, language is central to all types of research. Cross-language quantitative research relies on participants to interpret quantitative questions, decide what they mean, and respond in a way that can be collated as evidence by the researcher (Alvesson & Kärreman, 2000). Research that does not directly use written/spoken language to collect data such as arts-based research using storyboarding (Redman-MacLaren et al., 2017), photo-elicitation (Allen, 2011), music, and dance (Leavy, 2015), all still ultimately rely upon the use of words, and a shared language, to communicate the evidence-gathering processes and the findings and/or interpretation of the research task. Thus, language is critical to the research process.
For the qualitative researcher who uses words as the primary form of evidence gathering and communication, their epistemological position will especially influence the attention given to language and its associated power. Data generated is a combination of the experiences of both participants and researchers (Charmaz, 2014), although it is the participant’s story we as researchers ultimately seek (Court & Abbas, 2013). Recognizing the process of cocreating knowledge can provide opportunities for power sharing (Charmaz, 2014). However, this power sharing is not automatic, with both researchers and research participants able to share or withhold power through the selected use of words and/or language. In situations where both researchers and research participants speak multiple languages, the choice of language used in the research process can be central to whether the research is experienced as authoritarian and oppressive or transformative and power sharing (Hole, 2007; Mafile’o, 2005; Mertens, 2009). The process of negotiating which language is used, who chooses, whether multiple languages can be used, and whether interviewers and interviewees can switch between languages to allow the meaning to be communicated is therefore essential in cross-language research. This is especially pertinent when researching in partnership with Indigenous people (Bird, Wiles, Okalik, Kilabuk, & Egeland, 2009; Lincoln & González, 2008; Mafile’o, 2004; L.T. Smith, 2012; Tamasese et al., 2005).
In this article, we outline an example of cross-cultural qualitative research undertaken in Tok Pisin the lingua franca of the South Pacific nation of Papua New Guinea (PNG). Tok Pisin was not a primary or Indigenous language for either researchers or research participants in the study. Building upon identified challenges of conducting cross-language research, we identify opportunities for participative and power-sharing processes in cross-language research. This article therefore explicitly addresses a gap in the literature about methods for conducting research in cross-cultural and cross-language situation in a language that is not a primary or Indigenous language for either researcher or participant. Rather, the research is conducted in a mutual language that is shared by all—a meeting in the middle.
PNG and Tok Pisin
PNG is a hyperdiverse, lower middle-income country in the South Pacific with a population of approximately 8 million people. PNG gained independence from Australia, as a colonial administration, in 1975. Over 800 separate languages are spoken in PNG, with three national languages: Tok Pisin, Hiri Motu, and English (Muhlhausler & Romaine, 2003). Almost all people speak an Indigenous tok ples language (literally, language of place [ples] of origin; de Groot, 2008). Papuan peoples in the southern region of PNG speak Hiri Motu, and Tok Pisin is spoken throughout much of the remainder of the country. Those with a formal education are usually educated in English. Tok Pisin is a rapidly and continuously expanding language, characterized by regional differences and generational iterations and is increasingly being spoken in the southern region (de Groot, 2008). There are an estimated 3–5 million speakers of Tok Pisin, with up to 500,000 first language speakers of Tok Pisin in PNG (G. P. Smith & Siegel, 2013), including young people with parents from different language groups or who grow up in urban areas. Emerging from plantations during colonial rule in the mid-late 1800s, Tok Pisin has been variously known as NeoMelanesian, Melanesian Pidgin, New Guinea Pidgin, Tok Vaitman, and Tok Boi, with the name Tok Pisin adopted in 1981 (Romaine, 1992). The various names of the language reflect the history of speakers from expatriates to PNG plantation workers and finally to broader PNG. In the early 1900s, Tok Pisin spread into villages and “knowledge of the language was generally accepted as the means of achieving material prosperity…and power” (Muhlhausler & Romaine, 2003). Skills in Tok Pisin created an opportunity for PNG people to communicate with colonial masters, who had control of resources and opportunities not available to village-based people. In intervening years, colonial administrators introduced English, which is now the formal language for education, business, and government. English is less commonly spoken in social contexts, with Tok Pisin recognized as a sign of national identity among PNG citizens (Shelley, 2013). Despite the high status of Tok Pisin previously, those who now speak English are more likely to achieve material prosperity and power as they access education, resources, and opportunities in the rapidly modernizing and globalizing country of PNG.
An Example of Cross-Language Research
In this article, we reflect on an example of cross-language research in which Tok Pisin was used as a mutual language that was shared by researchers and participants, but was not a primary language for any. The transformational grounded theory study explored implications of male circumcision for human immunodeficiency virus (HIV) prevention for women in PNG and was led by MRM (Author 1) for her doctoral research. This study was embedded in a decade-long program of research to investigate multiple aspects of male circumcision for HIV prevention in PNG (MacLaren et al., 2013, Tommbe et al., 2013). Transformational grounded theory is a critical grounded theory methodology that incorporates participatory action research and decolonizing approaches to cogenerate, coanalyze, and act upon data generated (Redman-MacLaren & Mills, 2015). MRM led the study and RT (Author 3) was coresearcher, cultural broker, interpreter, and translator. TM (Author 2) and DM (Author 4) mentored MRM and RT. All authors critically reflected on the research processes throughout.
The research included individual interviews and interpretive focus group discussions with 67 women and 1 man at two sites in PNG (one urban and one rural). Residents at these sites originated from across PNG and therefore dozens of Indigenous languages were spoken by small groups at each site. Participants who were interviewed or participated in interpretive focus groups were from more than 20 cultural groups and collectively spoke more than 50 languages. Most participants spoke at least three or four languages, which is not unusual in PNG. A few participants spoke Hiri Motu, almost all spoke Tok Pisin, and all with formal education spoke English. None of the interviews or focus groups were conducted in the primary or Indigenous language of any of the participants. Most of the interviews and focus groups were conducted in Tok Pisin with intermittent use of English, at the request of both research participants and researchers in different situations. No translation of Indigenous languages was offered or made available (nor was it expected by the participants) due to the huge variety of Indigenous languages represented at each site.
I (MRM) am an Anglo-Celtic Australian who speaks English as my first language. From 2005, I began to develop language skills in Tok Pisin. During 2010, I lived in PNG where I became proficient in Tok Pisin (verbal and written, largely thanks to RT who continues to instruct me). I have however spoken the Melanesian language of Solomon Islands Pijin since 1992. Solomon Islands Pijin is closely related to Tok Pisin, with shared linguistic structure and many shared words. For 25 years, I have been using Solomon Islands Pijin for public talks, teaching research, and conducting development and public health work. In PNG, I have actively been using Tok Pisin since 2010 for public talks, teaching research, and conducting development and public health work. This includes using Tok Pisin throughout my doctoral research.
I (TM) was born and raised in a bilingual home in Aotearoa, New Zealand—of both Pakeha and Tongan heritage. I spent time during high school in Tonga and focused on Tongan approaches to social work for my PhD, including the role and function of Tongan language as it represents worldviews. I also spent 7 years living and working in PNG where I was surrounded by and have a functional understanding of Tok Pisin.
I (RT) am a Papua New Guinean woman from the province of Enga whose tok ples is Engan. The Engan language is spoken by both of my parents, and it distinguishes my ethnic group from others. My second language is Melpa Tok Ples from the Western Highlands Province. I learnt this language in my childhood days when my father was working as a health worker in this part of the Western Highlands Province. My third language is Tok Pisin, which is a national language of PNG and my fourth language is English. I learnt to speak fluent Tok Pisin as a child and am highly proficient in English. I completed all of my formal education in English, including international postgraduate education.
I (DM) am an Anglo-Celtic Australian who speaks English as my first language. I have lived, worked, and conducted research in Solomon Islands and PNG for over 25 years. I have used Solomon Islands Pijin and PNG Tok Pisin in public talks, teaching, and research. I also have basic proficiency of the Kwaio language, a tok ples from Malaita, Solomon Islands. Linguistically, “meeting in the middle” has been a characteristic of my work and research across Melanesia.
We have included this seemingly indulgent history of our respective her/histories of language acquisition and proficiency because the explication of language proficiency is key to conducting and reporting cross-language qualitative research (Squires, 2009). By having these levels of language knowledge in both Tok Pisin and English, we researchers desired to provide language options and to conduct the research in a language closest to the lived experience of the research participants. This was particularly important given the sensitive nature of the research. The process of both researchers and participants switching between Tok Pisin and English throughout the interviews and focus groups (Auer, 2013) was explicitly encouraged to increase options to express a concept or describe a phenomena using the most representative words and tones available in both languages. The code switching was to indicate shared group membership in a bilingual speech community (Auer, 2013) and enriched the tapestry of the research endeavor.
Method
Following the fieldwork described above, the four authors critically reflected upon their experiences, challenges, opportunities, and methodological considerations of using Tok Pisin as the language where researchers and participants linguistically met “in the middle.” Critical reflections were recorded in memos, notes taken during face-to-face discussions, and written comments in each other’s manuscripts. Following this, three people who had been members of focus group discussions were purposively selected and asked to reflect upon their experience of using Tok Pisin in the research. One had a postgraduate education, the second was currently a university student, and third had a primary school education. All were women and aged between 30 and 45 years. All three were interviewed by RT, a Papua New Guinean woman. Participants were asked: (a) what they could recall about how different languages were used in the focus group; (b) what were the good things and not so good things about the language/s used; (c) if they switched between languages during the focus group, why they switched languages; and (d) if using Tok Pisin affects what and how information was shared with the Australian and PNG researchers and others in the group, including specific terms used? Results of the authors’ critical reflections and the participants’ experiences of the cross-language process are described below.
Ethics Approvals
Human Research Ethics Committees of Pacific Adventist University (PNG), James Cook University (Australia), and the Research Advisory Committee of the National AIDS Council Secretariat, PNG provided ethics clearance for this research.
Results and Discussion
Cross-Language Research as Experienced by Researchers
It is our experience that cross-language research projects provide opportunities for researchers to reflect upon and to be explicit about how the use of language informs trustworthiness, validity, rigor, and interpretation of research findings. As qualitative researchers, we embrace the idea that research is not value-free, and qualitative research in particular is grounded in the social world of the researchers meeting research participants (Hsin-Chun Tsai et al., 2004; Mafile’o, 2005; L. T. Smith, 2012). We therefore caution against the premise that cross-language research necessarily presents a barrier between the researcher and research participants. Our experience is that such a deficit approach may mask the opportunities that cross-language research brings for learning for both researchers and participants.
Cross-language research has provided us as researchers with an opportunity to heighten our awareness of the role of language. Our increased awareness of the use of language between the researcher and the participants has led us to reflect on history and power in PNG. This awareness has driven us, as qualitative researchers, to purposively name and address hierarchies of language that have emerged, and continue to emerge, in PNG (Kovach, 2009; L.T. Smith, 2012; Tamasese et al., 2005). Therefore, cross-language research has brought us an opportunity to more carefully examine the representations of meaning held in language as they manifest in the cross-language “space.”
It is our experience that conducting research across languages heightens the researcher’s status as naive inquirer. We believe that being a naive inquirer is consistent with participative, power-sharing approaches to research (Stringer, 2014). With such an approach, the researcher is there as a partner in the inquiry, with much to learn about the way the participants understand the phenomena being examined (Charmaz, 2014). By working in the shared language that is closest to the lived experience of the research participants, it is our experience that researchers and participants discuss the phenomena (in this case, the implications of male circumcision) at a different, deeper, more reflective level. This deeper, more reflective level would not have been possible had the discussion been in a language further removed from the participants lived experience (Mafile’o, 2005).
Consistent with Chen & Boore (2009), it was our experience that working across, and translating between languages, “relates not only to language but also to culture” (p. 235). With this guiding principle, we, within the research team, constantly gained language and cultural learnings from each other. There were multiple times when we sought clarification and corrections within our research team during data collection, transcription, and analysis. This enabled a more informed understanding and clarity of both the actual words spoken and greater insight and engagement with the worldview of the participants, as represented by the lingua franca language. Consistent with this, the transformational grounded theory that was created from the study included codes and categories in Tok Pisin that were translated into English by MRM and then systematically reviewed by not only fellow researchers but other Tok Pisin speakers for accuracy. Before data were presented back to participants in professional and policy forums, RT, as a PNG woman, reviewed all written quotes to ensure not only technical accuracy of translation of the words but cultural understandings were being communicated through those words. We were therefore cognizant that although cross-language proficiency contributes greatly as a research skill, it does not equate to cultural proficiency or worldview proficiency. This allowed a space within our research team to learn from each other about both language and culture as a basis to humbly engage in the lived experience of research participants.
Cross-Language Research as Experienced by Participants
Participants recalled the focus group discussions consisted of Tok Pisin only, or a mixture of Tok Pisin and English, depending upon the group membership. For some participants, English was a language that could be understood, but Tok Pisin was preferred. One participant reported that an international person talking Tok Pisin was central to their communication—em tuchim (pilim) bun bilong mipela (literally, it touches our bones). Despite discussing a sensitive sexual health research topic, participants report using Tok Pisin helped them more comfortably express what they wanted to say. One married female participant explained that although hearing English terms about sexual health made her feel uncomfortable, …ol dispela Tok Pisin words em (the researcher) usim long em ya ibin orait (all of the Tok Pisin terms used were alright). Mi hamamas tru long harim ol disla kaen bikos em tokim stret (I really appreciated everything we discussed because she [the researcher] told us clearly). The participants highlighted the value of conducting research in a lingua franca to increase clarity of communication in the research process for effective knowledge exchange. “…(MRM) being White wasn’t an issue because the language kind of hide the color and we were able to communicate…speaking the same language kind of put us at the same level and I guess that’s all that matters because we were able to understand each other well so we could communicate better.”
Researchers understanding and using Tok Pisin reportedly increased trust between researchers and participants. …for (MRM) being a person from outside our community collecting data—for me personally, one factor that kind of connected me to her was the fact that she was able to speak and understand Pidgin (Tok Pisin) and I think that was also one key factor in getting the group to connect with her. If I can say that she (MRM) is white but when she speaks the language that we understand, it was more like she was part of us and we were able to relax and just go and speak our mind out and want to talk about the topic that we were discussing and I thought you know it kind of create a bond a connection where it is easy to communicate whatever it is that we have in mind.
Participants were asked to reflect upon their experience of switching between languages during the research discussions. Not all participants experienced code switching with one married woman stating, Em bin tok pisin stret nogat wanpla English em putim insaet (She [MRM] spoke only Tok Pisin, there were no English words included). The context of a group meeting in PNG determines the language ones uses, or at least starts in. In a group with women with predominantly primary school-level education, Tok Pisin is the language researchers started with. In a group of women with more formal education, researchers started in English as a sign of respect and acknowledgment of their level of education. However, English does not always meet the needs of participants. One female with a postgraduate education explained her experience of code switching. “There were times we switch coded in English when we wanted to clarify some ideas that we were discussing about the topic…unconsciously we would keep switching in Pidgin (Tok Pisin) and back again.” When asked to clarify which language the participants switched from, the participant explained there had been switching from English to Tok Pisin. …it was a group activity on doing some drawings as a group work and then we did discussion and as we were sharing, because English is also not our first language and as we were sharing ideas, we tried using Pidgin to help ourselves understand what we were sharing…for me English is like a fourth or fifth language and when we shared issues like that, it was a sensitive one and in some instances, ladies would giggle and also yea, and those things, and they felt comfortable using Pidgin (Tok Pisin) to express because another person would understand it better in our own context so that is why we were switch coding between English and Pidgin (Tok Pisin).
Translating a Worldview
In order to explore the meaning people attribute to the phenomena being researched, the researcher and participants need to be able to communicate about, if not share, a worldview (Mertens, 2009). The ability to provide literal translation does not provide certainty as to the interpretation/translation of meaning. Linguistic competency also requires sociolinguistic and strategic competence (Squires, 2008). In the desire to cocreate culturally relevant health knowledge, we are mindful that MRM initially analyzed the data as a non-Papua New Guinean. Some researchers have experimented using coresearchers and translators from the target population to ensure not only language is translated correctly but also interpreted more consistently with the worldview of research participants. The complexity of qualitative data and the potential for error in translation leads to the recommendation to use a panel of experts to enhance the rigour of the work. This panel of experts should include those with language, cultural, subject and methodological expertise to ensure adequate debate on the issues that impinge on the translation (Chen & Boore, 2009, p. 238).
Language is constantly changing, reflecting the social, cultural, spiritual, and economic conditions in which it is constructed (Holmes, 2013). Change occurs when younger generations express themselves using language, with slang, street-talk, and locally specific words and phrases constantly evolving. There is also some convergence occurring between Tok Pisin and English in PNG (G. P. Smith, 2000). As anyone who does not speak the language of the people with whom they are conducting research knows, language is power. Collecting data in a language that people use on a daily basis is highly desirable. It allows the expression of culturally situated metaphors and stories, allows for mental scaffolding of ideas being discussed, and centralizes the research participants’ experience. Having knowledge of the language, including the changes occurring in language, created opportunities for researchers to share power in cross-language research.
The use of Tok Pisin in this qualitative health research study undertaken in PNG is consistent with decolonizing methodologies that intentionally devolve power, epitomized by the participant’s power to choose to engage in the research using Tok Pisin or English. The devolution of power by offering the nonuse of English (the ex-colonizers language) was commented on throughout the research project, with research participants commonly expressing their appreciation at both field sites. The sentiments expressed were consistent with experience of public health research in neighboring Solomon Islands (Redman-MacLaren et al., 2010). The ability of the researcher from Australia to undertake research in Tok Pisin said more than language proficiency. It also demonstrated a long-term engagement with PNG, a commitment to understanding the experience of PNG people and valuing the expression of experience through a language closer to their lived experience using methods that are locally appropriate (Tommbe et al., 2013).
MRM found analyzing the data in Tok Pisin kept her close to both the content and the field. It also centralized the oral nature of Tok Pisin as a language. Although it can be written, the predominantly oral use of Tok Pisin means people are less comfortable using its written form. This makes the task of verbatim translation (including consistency in spelling) a challenge for both first-language and second/subsequent language speakers of Tok Pisin. Below is an excerpt from a memo written by MRM during data analysis: LOVING being in the Tok Pisin, it makes it so interesting. (It is) working if I read it aloud as it is such an auditory experience for me (i.e., I can make sense of it hearing and speaking it more quickly than reading it). Interestingly this was also the experience of our colleagues who were doing the interviews in Tok Pisin (for the Male Circumcision study)—they were happy to speak Tok Pisin but didn’t enjoy writing the words down. (MRM, February 18, 2013)
Limitations of Meeting in the Middle
The act of meeting in the middle of undertaking research in a language that is not the primary/Indigenous language of either the researchers and/or participants meant that on occasions we all (researchers and participants) conferred about the best word or phrase to use in Tok Pisin or English (depending on the language the group was being conducted in), especially words related to sensitive sexual health terms. There was a recursive process between participants and researchers, with MRM on occasions conferring with RT in English about the best word or phrase to use in Tok Pisin and RT conferred with MRM about the best word or phrase in English. Before collecting data, MRM and RT discussed the way we would explain the concept of research to village women and decided upon wok painim aut (literally, work to find out). In PNG (as elsewhere), the use of language is socially situated, with educated professionals (also known as “elites” in PNG) often preferring to speak to each other in English. The research process using multiple languages and switching between languages was on occasions messy and unpredictable but did allow for an exploration of the research topic in a richer way.
While we conducted cross-language research—and through Tok Pisin there was a meeting in the middle—we acknowledge that there would have been different ideas, concepts, and/or expressions if participants had used tok ples. Future research that was more localized and ethnic/language/tok ples-specific could explore the topic differently. However, the limitation of a more localized, ethnic-specific approach would be that a different (and perhaps reduced) contribution would be made at a national level, when compared to our broader study.
It is rare to find a situation where a researcher from “outside” is fluent in the language of the communities with whom she is working (Temple & Young, 2004). This situation has many benefits but “does not imply that the final text is nearer ‘the truth’…as epistemology cannot be easily tied to social location” (Temple & Young, 2004, p. 168). This article does not reflect the situation or experiences of all researchers conducting research across language or cultures. Nor does it reflect the cross-language experiences of all participants in the study. However, this article does demonstrate a common experience of researchers and participants in many multilingual contexts, especially in the Pacific where we belong and work. This article therefore explicitly addresses a gap in the literature about methods for conducting research in a non-Indigenous language.
Conclusion
In this article, we have critically reflected upon language and meaning-making issues facing researchers who undertake research in a shared but non-Indigenous language, using the example of research undertaken in Tok Pisin, a commonly spoken lingua franca (bridge) language in PNG. As participatory qualitative researchers, we recognize the complex, yet rewarding process of cocreating knowledge between researchers and participants about a phenomenon. It is critical that researchers address the hierarchies of language when conducting research across language and across cultures and locate the sites of power in knowledge generation when researching across cultures for qualitative health research.
Footnotes
Acknowledgments
Thank you to the women involved in the study who generously shared their experiences of this sensitive health topic—you taught us so much. We thank Professor Jane Mills and Dr Clement Manineng for critical discussions about cross-language research and for reviewing this manuscript. Thank you also to partners at Pacific Adventist University and New Britain Palm Oil Limited, Oro Province, for your support of this research.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Financial support for the research was provided by the National Health and Medical Research Council of Australia (GNT1038200), College of Medicine and Dentistry, James Cook University, and the Far North Queensland Hospital Foundation.
