As shown in the first section, synthetic worlds offer a very actual and interesting research arena for socio-cultural sciences. In this sense, ethnography is one possibility to gather qualitative research data, as described by Geertz (1973): a socio-cultural fact must be embedded in a dense description of the scene to be understood by the observer, who will explain and interpret social forms of expression.
In this section, this paper will briefly give an overview about ethnographic methods used in synthetic worlds. These methods are based on what I would call “classic” ethnographic methods (participant observation, ethnographic interview, and the like) and which will not be discussed here in detail (more information can be found at: Agar 1996; Girtler 2001; Spradley 1979)
Instead, the objective here is to inform the reader about the wide repertoire of ways to gather ethnographical data for the research of synthetic worlds as well as the resulting (potentially problematic) constellations regarding ethical issues. Other methods used in online research in general are not discussed here - Hine (2005) and Johns, Chen and Hall (2004) offer comprehensive information here.
Methods of virtual ethnography
This type of methods is partly related to the interview and participant observation, but mainly focuses on the analysis of media, media content, ego-documents and other forms of written or multi-media sources (see also Williams 2007). The main difference to all other ethnographic methods is the explicit relationship toward digital media or virtual objects such synthetic worlds, avatars, websites, and others (see also Dominguez and others 2007). As such artifacts have a cultural meaning, there are valuable to the ethnographer. But because of the nature of these cultures and their environment, some established methods (as the ethnographic interview and participant observation) must be considered and applied under new circumstances (Hine 2000).
One example is the analysis of media content. Websites maybe changed continuously, by many co-authors, they can tell stories spread through many different websites and connect a network of distant people and organizations. Another problematic aspect of virtual ethnography is the authenticity of interview partners, the reliability of collected data and so on. Virtual ethnography is not only useful to understand modern societies, but also to discover new forms of culture – like the synthetic worlds.
Ethnography in synthetic worlds as a scientific topic
Matthew Williams (2007) provides a good overview about the debate regarding participant observation in virtual worlds. As these virtual spaces are claimed to have created their own culture, a new form of ethnography was adapted: virtual ethnography. The basic idea behind “virtual ethnography” is that it is a collection of classic ethnographic methods adapted to the synthetic world environment. However theses adaptations may look like and be different from one MMORPG to another, they will alter the way ethnographers collect, analyze and present their data (Markham 2005).
Where ethnographic methods still cause question marks regarding authenticity, this is even more valid for synthetic worlds. Ethnographers must take into account their own cultural background, their language, norms and values before they recount the observations and stories seen and watched in their language. In virtual settings, this is even more complicated because each setting will be different (many sub-cultures, different MMORPGs, …). But MMORPGs offer some opportunities too: the presence felt by the players through their avatars provides a more immersed interaction than in text-only environments (Williams 2007: 12). Depending on the design and code of a synthetic world, avatar gestures, movements, objects and so on can be an important factor and information source through participating observation.
Specific ethnographic aspects in synthetic worlds
One aspect, which is very valuable for any kind of participant observation, is visual appearance of avatars. In contrast to text-based virtual settings (live discussion boards, chats, instant messengers, MUDs, and others), the appearance in 3D worlds is based on visual aspects like clothing, hairstyle and avatar body as well as many other factors. This means, from a technical point of view, that the ethnographer can collect data from the observation of avatars only, even without getting in touch with them (“lurking”). This is partly impossible in text-based environments, because we can only observe contributors, but not listeners.
Gestures and style of movement are other sources for an ethnographic account of a synthetic world, distinguishing so-called avatar tourists from real synthetic world citizens.
Language is a very important aspect of culture, either in text-based settings or 3D synthetic worlds. Emoticons and acronyms are sometimes general (“lol – laugh out loud”), but sometimes very game or group specific. More than in field settings, the presence of the researcher cannot be achieved only by his/her body, but mainly through his/her actions and speech (Markham 1998). This means, the ethnographer must immerse deeply into a new culture to be “in”.
Another aspect of ethnographic participant observation is the possibility to wander through the synthetic world. The freedom to walk is bigger than in real world, for several reasons. First, avatars can be moved independently of the researchers own physical conditions. Second, avatars may use other transportation and mobility techniques like flying from house to house, or teleporting from island to island. Third, the observation tools enable to overview a complete scene from a bird-angle, or from a first-person-perspective. The researcher can shift its focus depending on his/her intentions to observe.
Space may play a role in synthetic worlds where public and private space exists. Conversations held within private space may be considered private and need a signed informed consent to be included in an analysis.
A very special aspect of participant observation is to try to follow specific users not only within a single synthetic world (Leander and McKim 2003), but also during their visits to other virtual settings. Moving from one world to another, from one service to an alternative discussion board and so on is a widely observed behavior of users. Simultaneous following is difficult, because the observer don’t have the tools to know where the user actually went, even if he/she is still present in one world, but active in another.
Research tools in virtual settings
Additionally to the described aspects of participant observation, interviewing others or conducting observations needs always – at a certain time – to take field notes. These field notes are the base of the following analysis and conclusion and are therefore very important. In offline settings, taking notes is somewhat critical when, where and how they are taken because they can influence the scene. In virtual settings, this problem can be neglected because the observed or interviewed person can’t look at the researcher’s desk or the desktop.
Cheap and easily available audio and video recording software enables the researcher to record scenes almost automatically and without any big effort. By the way, this method doesn’t disturb the daily life of the observed avatars at all. Nevertheless, an informed consent should be sought.
Interviewing key informants may also be conducted not only by using chats, instant messenger or email software, but also synchronous voice-over-IP (VoIP) tools. The first set of tools (chats, instant messenger, email) is often available within a synthetic world, and is a vital communication channel for all citizens of these environments. The history of a chat can be copied and stored as valuable data, a process, which enables to avoid error-prone and time-consuming transcriptions.
VoIP tools can be used to interact more personally and faster with informants, as well as to learn more about the personality of a user. Such applications are mostly known (by the name, at least), but not always available to the other side (user). That is, that such communication depends on the technical possibilities of an informant.
What might go wrong
As described, virtual ethnography and their main methods are useful to collect data. But there are some potential ethical traps too. One of them is the anonymity of the interview partner or observed user. Due to the fact that in these worlds most of the users play a role and use a fictive (avatar) name and a fictive representation (avatar), we cannot prove their age, their identity and even their existence. This is significant because a researcher could unintentionally conduct interviews or other data collection methods with groups that enjoy special protection like kids. This scenario is not uncommon as kids are one of the main target groups of diverse MMORPGs.
On the other side, how do researchers prove their status in a synthetic world to their informants? It is not possible to show an ID card, and providing an email or a link to a personal and professional website is not a credential, but only information who the researcher says to be. Building a transparent relationship between researcher and informants is critical to every ethnographic study, but also difficult to achieve in a virtual setting.
Another aspect of a potential ethical dilemma is the possibility to observe a scene without informing others about your role. The technical feasibility of an all-mighty researcher, recording everything from voice to image, but without even being spotted, is real and seductive. In offline settings, ethnographers couldn’t walk into an Indian village and remain invisible for the rest of their observation period. Online, the Indian village wouldn’t even know that it is the focus of a study.
The leakage of data is another threat to the online ethnographer. The tools to search, filter and combine data from online studies that involves real persons (even when they are hidden behind avatar names) can be used to reveal the true nature and personal information of informants by third parties. Then, this unintended disclosure may be the base for commercial annoyance, personal harassment or even legal persecution.
Other technical traps include network software and data sniffing. Network analysis tools could compromise researcher’s data: it could be possible to find out informants through social graphs, and then analyzing their social networks. During interviews, private information could be listen to by bystanders or others within a certain range. This can be difficult to avoid if there are no private places to go.