Methodological concerns in online translation community research: a reflexive netnography on translator’s communal habitus

ABSTRACT This article attempts to integrate Kozinets’ netnography and Bourdieu’s reflexive sociology as a promising methodology to study the communal habitus of online translation community Fixsub. We closely examine the methodological issues encountered and orientates to address the three interrelated questions: How should a researcher posit his/her scholarly position in an online translation community? How is a researcher–participant relationship built within the virtual settings? How does a researcher balance the traditional dichotomy between subjectivity and objectivity during his/her online participation? We conclude that an online translation community research requires the researcher to have an ongoing immersion within the fieldsite, as well as in a state of participant objectivation to assess his/her own positionality in relation to the object of research and negotiate the mutual shaping between a participant and researcher.


Introduction
The surge of online translation communities and the virtual networks formed through such practices have been attracting a wide range of scholarly focuses. Whether those online translation practices have been studied under the crowdsourcing translation (Losse, 2008), user-generated translation (O'Hagan, 2009;Perrino, 2009), community translation (DePalma & Kelly, 2008) or online activist translation (Baker, 2009;Pérez-González, 2010), the diversity and pervasiveness of digital technologies cannot be ignored. Due to the participatory and interactive nature of the Internet, researchers are further accentuated by the frequent use of an 'immersive' and 'prolonged' ethnographic methodology in the latest online translation studies (Cao, 2015;Li, 2017;Zhang & Mao, 2013). However, despite this interest in exploring the mechanism and management process of online translation communities, or the emergent collective identities, roles and virtual aggregations, little attention has been paid to the methodological challenges that researchers may encounter during the process of virtual participation and immersion. For example, how should a researcher posit his/her scholarly position in an online translation community? How is a researcher-participant relationship built within the virtual settings? And, more importantly, how does a researcher make a balance between the dichotomy of subjectivity and objectivity during his/her online participation? This article attempts to address the above literature gaps by combining netnography (Kozinets, 2010) and Pierre Bourdieu's reflexive sociology (2003) as a promising method to study the phenomenon of the online Chinese translation community Fixsub, seeking to investigate members' communal habitus in building and sustaining the online community. The article predominantly examines the methodological challenges that have been encountered during the fieldwork in Fixsub and highlights the corresponding strategies developed to manage them.

Netnography in online community research
The netnography (i.e., ethnography on the Internet) approach is proposed by Kozinets (2010) and is derived from the traditional anthropological approach of ethnography that has gained considerable popularity in the field of social sciences. It applies the idea of common participative ethnographic procedures to a new condition, namely, online social interactions. Kozinets (2010:, p. 366) defines netnography as a written account resulting from fieldwork studying the cultures and communities that emerge from on-line, computer mediated, or Internet-based communications, where both the fieldwork and the textual account are methodologically informed by the traditions and techniques of cultural anthropology.
The online communications used in netnography covers two types. One is the online communications among the objects (members of the community); and the other is the online communications between the subject (the researcher) and the objects. Referring netnography as a particular method beyond ethnography is significant. Ethnography involves an 'immersive' and 'prolonged' engagement with the members of a social community followed by an attempt to identify and understand their culture through 'thick, detailed, nuanced, historically-curious and culturally-grounded interpretation' (Kozinets, 2010, p. 60). Netnography follows these ideas but at the same time encompasses some distinctive characteristics. Kozinets (2010) suggests four significant differences between a traditional face-to-face ethnography and an Internet-based netnography: alteration (i.e., the nature of interaction is altered both constrainedly and liberatedly by the rules of the technological media), anonymity (i.e., the online medium provides new opportunities for liberating behaviours and creates a new sense of identity and flexibility), accessibility (i.e., the participatory nature of the Internet forms a free-access and democratic ethos across geographical boundaries) and archiving (i.e., the conversations and data facilitated by online media are automatically archived). Traditionally, ethnography emphasises 'authenticity', advocating 'a process of face to face interaction leading to transcription and writing of notes', then 'upon return to the home territory, writing of the ethnography' (Kozinets, 2010, p. 113). Yet it seems that there is no physical fieldsite in netnography. This could be problematic for those ethnographers who pursue the reality of the world. For instance, Stone (1995) questions the contribution and value of conducting ethnography in an online world since everything in the field is produced and mediated by texts. However, the answer to this query lies in the conceptualisation of the physical fieldsite: what netnographers study are not texts online, but online users' interactions through various technological mediated means. As Kozinets (2010, p. 114) argues, ethnographers 'do not merely study the movements of bodies and vibrations in air', they study the 'meanings of acts and utterances'. Similarly, the online community may manifest through textual meanings, but it can also manifest through how its members use a variety of Internet applications to make sense of their daily activities in the community. These non-textual manifestations can give researchers a more complete picture of the phenomenon under study.
Crucial to the feasibility of netnography is the awareness of what can be treated as 'data'. Netnographic research predominantly involves three types of sources: (1) archival data; (2) fieldnotes and (3) elicited data. Archival data refers to the data that may be directly obtained from pre-existing computer-mediated communications created by online users. The researcher may benefit from the abundant transcriptions of public web pages, posted threads, messages and images. Fieldnotes data is created by the researcher, helping to record a researcher's observations and reflections on the community. It requires a researcher to be part of the community. This data differs netnography thoroughly from other academic methods such as content or social network analysis, as a content analyst would not be able to 'read them deeply for their cultural information, pondering them and seeking to learn from them how to live in this community' (Kozinets, 2010, p. 96). The third type of data, elicited data, is the data that the researcher co-creates with the online community members through either personal or communal interactions, such as surveys, interviews and emails. The multiple sources of data in netnography help to take various social domains into deep consideration and improve the reliability and validity of the evidence through triangulation.

Reflexive sociology and its integration with netnography
The impact of Pierre Bourdieu's sociology on translation studies in the last two decades is wide ranging. It has opened up new directions in the study of social-cultural dynamics of both the translator and translation community. Bourdieu's sociology evolved from his criticism of the two traditional opposing ideas of understanding the formation of the social world. He attempted to develop a social theory vis-à-vis both existentialism and structuralism. For Bourdieu, the existentialism ideas are problematic as they hold the idea that social world is built upon the subjective consciousness of human agents and overlook the concept of objective social structure. To solve the drawbacks of the subjectivism perspective, Bourdieu was captured by the idea of structuralism mainly developed by Lévi-Strauss (1979), who proposed that the power of structures works independently from human's consciousness. However, Bourdieu's later ethnographic study on the Kabyle Berbers in North Africa pointed out that people's actions and behaviours do not always follow the rules proposed in structuralists' model. Structuralism seems to overlook that an agent may act strategically and practically rather than solely conforming to external social rules. Informed by both existentialism and structuralism, Bourdieu provided a conceptual model that combines the agent that is free from the idealism of subjectivist accounts, and social space that is free from the mechanistic causality inherent in many objectivist approaches. The key of his social theory is the concept habitus, which refers to a 'system of durable and transposable dispositions' that generate and organise social agent's practices (Bourdieu & Wacquant, 1992, p. 53). It is the embodiment of social structures and the deeply ingrained skills, attitudes, appreciations, and perceptions that an agent processes due to his/her social experience. While habitus is not only embodied in individual agents, it can also refer to the collective system of actions and attitudes. For the purpose of this study, the term communal habitus refers to the collective dispositions of a group of agents in a particular communal culture. Communal habitus denotes the communal history that governs the conduct of action and interaction in a community.
In studying habitus, Bourdieu (2003) proposed a reflexive methodology, namely participant objectivation, in ethnographic research to overcome the dichotomy between agency and structure. He (1992, 253) noted: What I have called participant objectivation (and which is not to be mistaken for participant observation) is no doubt the most difficult exercise of all because it requires a break with the deepest and most unconscious adherences and adhesions.
Participant observation, as Bourdieu understands it, refers to the conduct of an ethnologist who immerses him/herself in a social universe so as to observe while taking part in it, trying to be both subject and object of the observation and critically reflecting both on the object of research and on him/herself as a researcher. This, for Bourdieu, is arduous to sustain, as he asks how can one be both subject and object, the one who acts and the one who, as it were, watches him/herself acting? Participant objectivation, on the other hand, refers to 'the objectivation of subject and operations of objectivation, and of the latter's conditions of possibility ' (2003, p. 281). In other words, it is the objectivation of the researcher him/herself and the social or institutional conditions of possibility of his/her own research. It helps to objectify the researcher's view and judgment on the subjects through interactions with the objects.
As Bourdieu (2003, p. 282) further explains, participant objectivation aims to investigate not the 'lived experience of the knowing subject', but the 'social conditions of possibilityand therefore the effects and limitsof that experience'. In other words, it is not simply about participating in the cultural group under study, but about the objectivation of the subjective relation to the object, and about the reflexivity of the scholarly distance to the object of research. This objectivation thus requires a researcher to be aware of his/ her own position and social trajectory (e.g. age, gender, nationality, social class, education background, beliefs, etc.) in relation to the object of research, as well as the research conventions underlying the discipline in the specific research environment in which the researcher operates, both of which affect the way in which the researcher constructs and makes meaning of his/her observation.
As researchers inside the academic field, all observers are dually 'playing the game' (i.e. obeying the rules of the field) whether this is acknowledged or not. In studying an online translation community, participant objectivation allows a researcher who is also a professional translator to critically examine and reflect on his/her own habitus acquired in the academic field while studying an online translation community made up of amateur translators. The researcher is able to break with any possible adherences to the community members and understand the patterns of social interactions in a translation community through re-examining some 'taken-for-granted' assumptions. Meanwhile, it is worth noting that such a reflexive approach plays a significant role in identifying Fixsubbers' communal habitus. Although the archival data in the form of computer-mediated text helps identify certain patterns in terms of sustaining the online community, as habitus refers to the generative principle of practice and is 'mostly unconscious' (Bourdieu, 1990, p. 90), social agents sometimes are not aware of what governs their practice and brings it to the order of discourse. Therefore, some habitus can only be identified and interpreted from the point of view of a full insider rather than asking them directly.

A Reflexive netnography of online translation community Fixsub
While the dual authorship of this article continues, the documentation of reflexive fieldwork in Fixsub demands its singular narrator. One of the authors (Sijing Lu) was involved in participating in Fixsub and conducting a self-reflexive account, thus the following paragraphs are presented in the first person singular narrator.

The research fieldsite
In seeking to study Fixsubbers' communal habitus, the selection of research fieldsite 1 makes a 'first break' with my social past, but to a degree, this is still my social present, where I continue watching American TV series or films subtitled by Fixsub in my leisure time since my undergraduate life. I am aware that my preference for and love of watching videos subtitled by Fixsub is crucial to the choice of my fieldsite. I have been following Fixsub's Weibo 2 for five years and have been interested in and curious about their community-building for a relatively long time. This previous watching experience gave me a keen sense of the need to select Fixsub as my target fieldsite. Fixsub ( Figure 1) was established in 2015 and later developed into a non-profit online fansubbing community with more than 6000 translators and 5,000,000 followers on Weibo. Apart from the effect of my possible unconscious adherences to the community, Fixsub was selected also due to its relevance, level of activity and accessibility. The first point is its relevance on translational and communal activities. Fixsub has translated hundreds of audio-visual materials of diverse genres and foreign languages into Chinese. Besides the daily translational activities, Fixsub also focuses on community-building. There are various communal activities that Fixsubbers can participate in, ranging from discussing or simply commenting on films to selecting or voting for the best monthly translators. The second point is its level of activity. Since online communities often have a high turnover rate, the recent and regular communication in Fixsub's multiple social media platforms indicates an active and stable status. Specifically, the conversations in QQ 3 chat rooms established by the group members last from morning until midnight; the official website and Weibo are still being updated regularly. The final aspect is its accessibility for participation. Some online translation communities in China have no longer recruited new members; however, Fixsub was still open for new applications when I joined the community, making my fieldwork activities and participation in the online translation community possible.

Data collection
The experience of my previous engagement with the people in fansubbing communities has necessarily constituted my present self and 'scholarly habitus' towards the collection of the data in this research. My past encounter with several Chinese fansubbing members (who were my close friends) has rendered me rich knowledge that most Chinese fansubbers relied heavily on QQ chat rooms to deliver daily tasks. I was also aware that fansubbers may establish websites or forums to release recruitment information to the public. The self who now engages in a fansubbing-related research has been constituted by my earlier experience that the most useful data of a fansubbing community can be found in QQ chatrooms and their official websites. Having made this clear, three types of netnographic data are employed in the study. Archival data covers the documents which are shown publicly to all the Internet users without accessible limitations. This includes a set of materials from Fixsub's websites and social media pages (e.g. Weibo and Wechat Official Accounts 4 ). Fieldnotes data refers to the observational and reflexive fieldnotes generated by participant objectivation. I use the mobile software Evernote to record images, texts and screenshots together with my own comments that may worth exploring in every moment. The fieldnotes are primarily based on the messages I read in QQ and community's official website, and also include my personal thoughts and impressions of being a translator in different situations. Recording immediate fieldnotes has been a worthy tool, since it enables me to detect the underlying patterns of certain behaviours in naturally occurring settings and later make sense of the reflexive dimension of the research. Elicited data includes a set of text-interviews with 13 Fixsubbers. 5 Fixsubbers were asked about the demographic information ( Figure 2) and communal attitudes, including what motivated them to constantly work for the community, how they describe their relationships with other members, and what is the most important value in a fansubbing community.
Conducting a reflexive netnography involves unpacking the researcher's decisions regarding what questions to be asked to participants. The choice of interview questions is certainly neither neutral nor insignificant. In fact, the most common question that a non-fansubber would ask was why a fansubber would like to engage in volunteer work without monetary rewards, and this was also what I used to like discussing with my previous fansubber friends during my undergraduate studies. I have to admit, the curiosity about the nature of creating fansubs triggers me again to include this question in my research. Given my participation in Fixsub, I gradually developed close-knit relationships with other members, and then I felt better-placed to think critically about whether this is also the case in other members' experience and my close relationship with others does nothing with my simultaneous role as a researcher. Besides, I was also curious to know what other values they would abide by to keep their virtual long-term relationship, although I need to acknowledge my partially biased view towards my own curiosity through this focus. To avoid over-subjective view in defining interview questions, although I have three major interview questions, in my actual interview, our conversation topics were frequently sidetracked from my designed questions. In this situation, I would let my interviewees freely share their opinions without interrupting. The data was then evaluated through an analytical coding analysis where a series of patterns and themes arise after examination.

Ethics considerations
One of the ethical considerations that arises from this study is how to protect Fixsubbers' privacy and community's confidentiality. To overcome these ethical issues, my first step was to contact the community leaders to ask for permission. The group leaders of Fixsub agreed for my participation, while they required me to ask for other individual members' permissions if further research action is needed. I then sent my participation consent information approved by the Research Ethics Committee of the university to which I was affiliated through QQ group message to other members. Participants were assured that their data was only used for academic purposes, and their pseudonyms would be removed if their communications were quoted. Yet some members did still worry about their privacy and they expressed unwillingness to be observed. In this case, I sent them a 'research withdraw' message to ensure they would not be observed and their relevant information would not be used.
Another ethical challenge is the copyright of online resources. Kozinets (2010, p. 141) notes that 'oftentimes the Internet is used as a type of textual publishing medium, and culture members are fully aware of this public function', indicating that online resources' public nature to all who has access to it. Yet some scholars (Bassett & O'Riordan, 2002) argue that there is clear boundary between public and private spaces on the Internet. Taking these arguments into considerations, I carefully thought about the process and implications of the research data, and considered the data used in this research does not infringe the copyright, because (1) I have contacted group leaders and members to introduce the purpose of this research and the potential use of their community's data; (2) the computer-mediated data including their website pages and social media pages are publicly available to all Internet users and (3) although a part of texts in QQ group may not be available to those outside the community, they are available once you gain the official permission to access the community. In the consent information, I mentioned the members have the right to receive credits for their works such as texts, pictures or videos, and if they require credits they could send me message via QQ and I would then contact them privately and credit them using their pseudonyms to avoid violating their privacy.

Objectivating the researcher's positionality in access to fixsubbers
After determining the fieldsite, I read the recruitment pages on Fixsub's official website and sent my first email to the group administrator, introducing my personal background, interests and the reason for joining the community. The administrator of Fixsub replied to me within one week and sent me the guidelines for the community entry test, which included a questionnaire testing my language skills, and a clip of a video that needed to be subtitled. As a new candidate, I was given 24 h to complete the entry test following a file package that included Fixsub's Translation Rules, Fixsub's Test Regulations and Fixsub's Scoring Process of Entry Test. My education background in audiovisual translation and my occupational role as a freelance translator helped me successfully pass the community's entry tests and I was then invited into their QQ internship groups. However, I have to admit that, their entry test, which requires a high capacity in interpreting and listening skills, is by no means easy for one who does not have relative training in subtitling. Candidate can also easily fail within a very short time slot. This fact can also be found in their recruitment advertisement, which states that you can often try more than once to pass the test. I assume this test might be an effective way to assess potential members' determination to participate in the group's subtitling activities and the extent of their love of fansubbing.
I spent four weeks carrying out my internship. During my internship, I discovered that the way that fansubbers translate and present subtitles again did not conform to my pre-formed scholarly dispositions. As a freelance translator, the standardised rule in industrial practice is no more than 14 Chinese characters per line, black or white, at the bottom of the screen, and with no personal comments or notes. However, the translation rules in Fixsub are much more flexible and relaxed than in the mainstream industry. Members often use different colours and bilingual subtitles, and they can add notes at the top of the screen or their personal comments after the subtitles. Unlike professional practice where translators are often in an invisible position and subtitles are primarily used for referential purposes, the flexible and relaxed approach to subtitles in Fixsub indicates that fansubs are seen by members as the expressions of their individual voices and viewing experience and as an effective way to foreground their visibility and interactivity. It seems to me that the traditional power structure between authorship and viewership is broken down in a fansubbing community, given the fact that fansubbers in Fixsub have more 'power' in the production process to voice their individuality and interact with audiences. With this in mind, occasionally I would also insert my personal comments or notes in subtitles.
I was warmly welcomed by the administrator and invited to join Fixsub's official QQ groups after internship. However, working as an official member does not mean sustained acceptance within a translation community. In order to guarantee a longlasting commitment to the community, members are required to sign in their attendance on an online application called 'Group Calendar' and keep subtitling to meet the work targets. The group leader has also constructed a 'point system' 6 to ensure predictability on project and workload. Normally, I spent ten to fifteen hours per week to create subtitles and five hours a week to chat and communicate with other members. Establishing a high degree of ordinary sociability and normal social intercourse is necessary in conducting a netnographic research (Li, 2017). Consequently, I worked hard to remain visible in the community and ensure I was frequently involved in community's daily interactions to forge a sustained and trustworthy relationship with the other members. As the most active period in Fixsub is between 6 pm to 11 pm when most members are off duty or come home from school, I thus primarily made myself active in the group during the evening.
As a deep Internet user, it is not difficult for me to immerse myself into a virtual community like Fixsub. As shown in the demographic information, similar to other Fixsubbers who were born after 1990s in China (known as the 'post-90s generation' 7 ), the media we have consumed from our childhood differs greatly from the older generations. I was 25 when I firstly joined the community, which are among the same age group as most members. Being users of digital and mobile media virtually from birth makes us more familiar and dependent on the social interactions on the Internet. Personally, I spend several hours per day on the Internet to browse or participate in different forms of online communications. This virtual experience may unconsciously enable me to have an embedded understanding of the online society and feel 'well-adapted' to join new online communities including Fixsub. Notably, according to the interview, all the participants were raised up in mainland China, although they may be currently studying in a university overseas. In Bourdieu' sense, agents who share the same social-cultural or educational environment tend to form a series of similar social habits or cultural taste. When I objectivated my own growth trajectory, in spite of the different disciplines we may prefer, it appears that both my participants and I are the products of a same national education system: we completed our nine-year compulsory education 8 and high school in mainland China. Our educational background, particularly the early education, tends to be the same. This educational past presents a set of common dispositions, or in Bourdieu's sense, we are often attributed to 'national character'. It seems that I am 'peer' and 'local' to these members, knowing them better than others from a different nationality or age group. The 'taken-for-granted' dispositions made us understand each other more easily with 'a nod and a wink', or allow us to easily recognise what topics merit discussion or not, or what topics are deemed important or interesting in the community.
Such recognition of socio-cultural past based on my existing knowledge created a quite equal power relationship between myself and other research participants. I found we were more easily to develop intimate friendships through using similar language style, slang, neologism or jokes. They thought of me as someone with the same social standing as them and somehow did not display a sense of distance regarding my simultaneous role as a researcher. In terms of interviews, they were conducted after one year's participation. This is because I need time to build trust and friendship with community members. Trustworthiness may contribute to receiving more detailed and reliable responses. With my time in the community increasing, Fixsubbers gradually no longer treat me as an outsider but freely sharing their opinions or chatting with me. As Bourdieu notes (2003, 205), the social researcher should cultivate the 'craft of sociology' to 'affords one to enter the life of others'. Thus the researcher should choose proper questions that are related to participant's life in order to help interviewees to speak out the true feelings. With this in mind, when I interviewed my participants, I seldom used academic terminologies or terms to give fansubbers a scholarly presupposition that they were chatting with an academic researcher. To minimise my dominance over the interview, I used the colloquial language I would use in a QQ room and also tried to not control the interview but made it more dialogic and have more rounds.
In line with the methodology principle of participant objectivation and make better sense of the interactions within the community, I became a 'real', 'stable' and 'fullyimmersed' fansubber through participating in different types of communal activities. During the entire research period, I did not 'distance' myself purposely due to my dual identity as a researcher and a participant. I realise that the way in which I present my scholarly status has influenced my ability to manage the contexts and the relationship with community members, and thus the quality of data. Therefore, I regularly took part in some communal activities, such as the voting activities for favourite TV drama and best translator of the month. Through these activities, I have gained great insight into the inner mechanism of Fixsub and I have also realised that I should follow their communal norms and conventions, use their words, and experience what they experience in order to gain an authentic sense of the lives of Fixsubbers. However, I was also aware that over-identifying or too much exposure in the community might be problematic and thus influence the results of data (Li, 2017). Over-enthusiasm could arouse their suspicion of my real purpose of joining the community, or influence what they really think. One is advised to adopt a degree of 'marginality' (Walsh, 1998) when undertaking life as an ethnographic researcher. To be fully aware of this, my relationship with other Fixsubbers is mostly built through conducting the same subtitling project together, and our communication and interaction often cover the topics from resolving translation problems to sharing viewing experiences of certain films. Sometimes Fixsubbers' conversations could extend to topics unrelated to translation, such as their offline lives, opinions towards certain governmental regulations or social issues. In these situations, I seldom lead the conversation but listen to them quietly. Doing so contributes to 'secure' my durable relationship with community members but also at the same time decrease the risk of over-reaction.
A set of 'community etiquettes' were summarised premised on the daily participation in Fixsub. These etiquettes helped to guarantee a sustained acceptance in Fixsub: . Do not distance yourself. Be involved and active. Due to a researcher's scholarly status, s/he should not hide and only play a role as a distant researcher. Active participation helps to build a durable and trusting relationship. Some valuable understanding on communal habitus may only be identified through close interactions with Fixsubbers. . Be humble to accept criticism. It is common in the translation community that you are blamed by other members in terms of your translation mistakes. The purpose of this criticism is to produce subtitles of the highest quality and make the community a good place for users to learn foreign languages. . Always be aware of your position. A researcher should avoid using 'scholarly presuppositions' to communicate with fansubbers. Constantly reflecting on yourself is useful to minimise a researcher's sense of alienation and domination.

Non-monetary orientation
The fact that Fixsub attracts thousands of fans to constantly translate the audio-visual materials suggests that there is always a certain commitment that binds individuals to the community. We thus categorised the responses from the interviews to thematic patterns in search of communal habitus that contributes to the sustainability of the community. Seven Fansubbers claimed to be driven by an interest and passion: words such as '热 爱 (passionate)', '感兴趣 (interested)', and '乐趣 (enjoyment)' appeared frequently in the responses. They considered creating subtitles as 'the most interesting thing in the world (P3)', 'a way to kill time (P6)' or 'a daily routine (P13)'. These responses suggest that over a sustained period of time in the community, Fixsubbers gradually developed a type of 'leisure'. They do not see creating subtitles as time-consuming; instead, they consider it as a way to relax and entertain themselves. As Massidda (2012) notes, happiness and fun are crucial elements in fan-subtitlers' integration within the community. Through making contributions enabled by special skills and knowledge, Fixsubbers derive durable pleasure such as feelings of enjoyment, self-fulfillment and self-satisfaction from the process of subtitling.
Six respondents highlighted it was other intangible rewards from participating that draw them closer to the community, such as establishing a deep friendship with likeminded people and acquiring a sense of recognition in the community. For instance, 'I have a sense of belonging when talking with other members who are the same age as me (P13), 'I feel proud to see my name at the end of a film (P4)' and 'I feel satisfied when my work is watched by others (P1)'. While the motivations may differ, we find that community members tend to have a non-monetary orientation with regards to working as a translator. As observed in their daily conversations, strong complaints are frequently found in QQ chat rooms about those outside the group who consider Fixsub as an organisation from which to make money. Fieldnotes regarding frequent discussions in QQ also indicate that Fixsubbers emphasise that their works are altruistic, that their hard work is the result of selfless dedication, and that the idea of making financial profit from works is deplorable. They seem to share a communal habitus that joining the community could go some way towards satisfying their personal desire and needs rather than gaining monetary rewards.

A sense of loyalty
Being loyal to the community and developing a close-knit bond with other members was another communal habitus that we find in Fixsub. When asked to describe the relationships with other Fixsubbers in interviews, terms like '师徒 (mentor and apprentice)' and '好朋友 (good friends)' were mostly used by respondents. The relationship with other members has a great impact on involving them in the community. As P13 mentioned, 'it's easy for people to give up when they are not familiar with the environment and there is no one they know'. During the internship period, members are all allocated with a particular supervisor, who was later called as 'mentor', helping and supporting new members to get involved. Such assumed 'mentor' was described as modest, supportive, friendly and professional, providing constant help for members in not only translation issues, but sometimes extending to other life issues. One respondent (P11) noted in the interview, 'I'm so grateful for my mentor's patience and assistance. Her help fills me with great joy and deep gratitude. I love the community and I'll stick to my work'. In terms of the relationship 'good friend', some participants mentioned they use Fixsub as a site to pour out their personal feelings and emotions that may not be talked about in their offline lives. As shown in Figure 2, the majority of participants in Fixsub are females. This interesting result can also somehow explain why an emotional and intimate relationship are more easily to be formed in Fixsub, as Ubando (2016) in gender studies notes, females are more emotionally expressive than male. Their virtual friendship is strengthened through continuous communication and sometimes this online connection is transferred to real life by meeting up offline.
Kinship was another relationship that draws Fixsubbers closer to the community. We find that Fixsubbers tend to be fully immersed in this family-like environment and most of them are not aware of the existence of such kinship addressing. In their daily conversations within the community, it is common to see them call other members 'sisters', 'brothers' or other terms normally reserved for relatives. Addressing others in such a family-like way seems to have fed into their embodied dispositions and be in accordance with traditional Chinese values, whose system of address is deeply rooted in a social norm that 'attaches great significance to kinship' (Lee-Wong, 1994).
Different relationships confirm the development of close-knit bonds in the community. Such bonds further facilitate emotional attachment and a sense of loyalty to the community. As one of the authors' participation in Fixsub shows, the relationship between a researcher and other community members has greatly influenced the researcher's personal attitudes towards the community. Through prolonged involvement in various communal activities, a good relationship can be established between the researcher and community members, as they discuss their favourite films or TV dramas together and give each other kindly advice and support. In this sense, community members can help a person to get through some difficult times, such as when one feels bored and need something new to have a 'temporary escape' from the real life. Over time, the relationships with other members were strengthened and such emotional bonds made a person reluctant to leave the community. As some scholars (Li, 2017Melucci, 1995) argued, a certain degree of emotional attachment is required in constructing a 'we-ness' and a sense of community integration in any online field. Close-knit relationships in the form of community engagement can activate a sense of loyalty, which may well encourage members to continue their participation through voluntary efforts.

Teamwork spirits
The idea of team collaboration is highlighted in the community's introduction page on the official website, explaining that the success of the community is reliant on the joint efforts of its members. The term '团队精神 (teamwork spirit)' was also frequently mentioned in responses to the last interview question, where respondents were given with an opportunity to reflect on the aspects that are most important to them during the participation. Some members expressed how much they enjoyed working as a team, for instance: I love the feeling of common cooperation and a group of people discussing a particular TV series. (P11) I love the teamwork mode in the community: we have one group leader responsible for supervising the entire progress and a team of proofreaders responsible for the work quality. (P13) The production of one film involves different types of work, including translating, spotting, editing and so on. Every work is a collaborative effort. (P10) Since every project in Fixsub is completed through teamwork rather than on an individual basis, we find that the specification of roles and tasks in the process of translation create a sense of reliance among the translators. Specific functional division of labour joins translators together and encourages them to rely on each other. By turning translators into specialists with certain skills, a strong degree of interdependence is thus formed among them. The division of tasks and the delegation of the work to different translators make creating subtitles a collaborative activity; each person is responsible for completing a small part of the work that they are good at, and all of the roles rely on one another to create the high-quality subtitles possible. The average time spent producing one 50-minute drama episode in Fixsub is around 24-36 h. Some dramas may even take as little as four hours from production of the subtitles to their release. Such a short production time illustrates the importance of teamwork and how teamwork makes up for individual shortcomings. A member may be assigned as a translator but may not be competent in encoding videos and dealing with other technical-related issues. The translator was dependent on proofreaders to check the accuracy and grammar of translations and reliant on encoders to set up the file format. We thus suggest that, the teamwork spirit, embodied as correlative dependence and mutual engagement, inevitably shapes each user in the interwoven process of participation and ultimately contributes to the development and sustainability of Fixsub.

Conclusion
In this article, we have offered a new perspective by integrating Bourdieu's participant objectivation (2003) with Kozinet's netnography (2010) to investigate communal habitus in sustaining the online translation community Fixsub. On the one hand, the immersive, descriptive, and multi-data properties have made netnography relevant and applicable to the study of online networks of translation activities. On the other hand, Bourdieu's reflexive sociology, notably the idea of participant objectivation, allows a researcher to illuminate the persistent academic dichotomy between objectivity and subjectivity. It calls for objectivation of a researcher and reconfigures the positionality and domination of a researcher in the space of observation. The reflexive engagements with participant objectivation enable us to explore one of the thorny issues facing a netnographer when studying social and cultural groups formed online, that is, how the researcher should posit his/her scholarly position when conducting online communications with research participants. We conclude that a netnographic fieldwork requires continuous involvement of researcher into the community, during which the researcher also needs to be in a reflexive state, negotiating the dual roles of a participant and researcher in the process of evaluating his/her personal relationship with the community. The current study contributes methodologically to future fieldsite studies of various translation networks and communities. Nevertheless, there were inevitable limitations on the scope of the study and the availability of resources and data. This research could be profitably developed and extended to examine other language pairs or audiovisual genres in order to build on and nuance the findings of the present study.