Cosmopolitan Civil Societies Journal, Vol.2, No.3, 2010 1 ISSN: 1837-5391; http://utsescholarship.lib.uts.edu.au/epress/journals/index.php/mcs CCS Journal is published under the auspices of UTSePress, Sydney, Australia Ambiguity, Oscillation and Disorder: Online Ethnography and the Making of Culture Jonathan Paul Marshall University of Technology Sydney Abstract Online life is usually held to present particular problems for ethnography as it is hidden and ambiguous, and boundaries are not clear. However, ethnography and online daily life are similar procedures in which people go about constructing ‘culture’ to make sense of others and interact with a degree of predictability. Ethnographers can learn about culture and society by learning how people themselves go about understanding and making those processes. We further, do not have to expect that the reality we describe will be completely ordered, even though the simplifications of constructing ‘culture’ might make this seem inevitable. Disorder can be socially important. This paper grows out of my work on the internet Mailing List Cybermind between late 1994 and 2005. As I have an anthropological background, I see the basis of ethnography as ‘participant observation’, or living with (as best you can) the people being studied, engaging with their lives, and recording that engagement (generally in notebooks), and this shall be assumed in what follows. Ethnography may make use of surveys, structured interviews, or other techniques but, fundamentally, ethnographic research is about ‘living with’ a group, being open to contingency, to variation in meaning, and to both the commonality and disorder of everyday life. It is this sensitivity to disorder and ambiguity which helps ethnography to work The aims of this paper are: 1) To provide a guide for those beginning online ethnography by reference to my own experience. 2) To show that the supposedly ‘difficult features’ of online research, are features of online social life which anyone online has to deal with, or which emphasise problems already found in offline ethnography. 3) To suggest that ‘ethnography’ and ‘culture’ are imprecise and almost undefinable categories, which are connected to each other, so that people produce culture to gain an understanding of each other and act together. 2 Cosmopolitan Civil Societies Journal, Vol.2, No.3, 2010 ISSN: 1837-5391; http://utsescholarship.lib.uts.edu.au/epress/journals/index.php/mcs CCS Journal is published under the auspices of UTSePress, Sydney, Australia 4) To suggest that disorder and misunderstanding are not just defects to be ignored, but that they have a potent social role, and play a part in human activities, including those of ethnography. I argue that culture is, amongst other things, a set of models about: the nature of reality; what people do or should do; and how we should interact with other people. As ethnographers we construct models (together with other people), which explain others and which guide us in our interactions with them, and these models form the culture of the group for us. ‘Ordinary’ people also produce their culture in a similar manner and face similar problems in trying to live with the people they have to live with. Like any such model, it involves guesses at hidden, ambiguous or uncertain motives and actions, and actions based on those guesses. As such, culture is not an impersonal, or uniform, thing which is ‘out there’, it is situated, varied, often incompatible, and subject to struggle between people. Cultures are perpetually ordered and disordered, and that disorder is part of their dynamics. Problems of Ethnography and its History Historically the method and practice of Ethnography is ambiguous and uncertain. The term both applies to the method and the result (the book or article). As Grossberg writes, ethnography is ‘in the first instance, a certain kind of practice in the field, although it is not clear what sort of practice’ (1989, p. 23). Marcus remarks that anthropology is ‘a discipline that has been remarkably silent in a formal way about [ethnographic] method’ (2007, p. 1128). The standard old guidebook Notes and Queries on Anthropology did not try to define ethnography or fieldwork, it simply listed techniques. It accepted contingency and that much of what we might want to describe would be invisible to the ethnographer, but attempted to deal with this uncertainty by recommending that the researcher distinguish between a ‘fact’, a ‘conjecture’ or a ‘retelling’ (Anon 1951, p. 36ff.). However, these cannot always be distinguished: a social ‘fact’ may make no sense or may not be noticed, without an interpretation, retelling, or conjecture by someone as to the roles, actions and motives of others. There can be, therefore, little ethnographic observation without some construction of ‘culture’, and some push towards the hidden or unseen. Traditionally, within anthropology, a model of culture and society gains its meaning and coherence by fitting (theoretically relevant) observations, conjectures and tellings together, Cosmopolitan Civil Societies Journal, Vol.2, No.3, 2010 3 ISSN: 1837-5391; http://utsescholarship.lib.uts.edu.au/epress/journals/index.php/mcs CCS Journal is published under the auspices of UTSePress, Sydney, Australia and relating them to other observations, conjectures and tellings, and to theory. As a result the role or meaning of any custom or event, is relational, it depends upon the meaning we have given other such customs or events. So in accounting for one ‘thing’ the ethnographer is always led to other ‘things’. Therefore ethnographically produced culture has a tendency to delete inexplicable, ‘unrelated’ and supposedly random events, and produce a coherent ‘system’. When it doesn’t do this, it can produce largely incoherent lists of what people say they do and of tales they tell, as was the case with the early work of Franz Boas. The early American ethnographer Robert Lowie adds that ethnography requires academic culture, or awareness of theory (‘ethnology’) (1937, p. 3-4), which in turn requires professional ethnographies of other places so that, again, meaning and validity is relational. This relationship further implies that ethnography is a cultural act within Academia. Historically the birth of ethnography is connected to the birth of anthropology as a discipline, and marks the decline of comparative or diffusionist studies. Ethnography, in contrast, was marked by specificity of field. Among anthropologists any universalist statement could always be unsettled by a local counterexample. Culture was also made local to justify the boundaries drawn around field sites and to construct a simplicity which excused the ethnographer’s ignorance. Thus, in an oft republished 1960’s text book, John Beattie could write that although he happily described the societies in which he did fieldwork, he would not dream of writing about economics, or government organization in a ‘modern community’ because of its complexity (1966, p. 80). Yet the ‘bounded’ societies he did write about were part of worldwide colonial economic and governmental systems. The boundaries were not ‘real’ but heuristic, and perhaps helped preserve the cultural divisions between academic fields. However, as a result of this boundary making, models of culture became somewhat unreal, and self-contained. As we shall below, these historical shapings of ethnography as local and integrated, are specifically challenged by online ethnography (among other things). Problems of Culture Famously, after Kroeber and Kluckhohn pronounced that ‘culture is the central concept of anthropology’ (1952, p. 36), they went on to display ‘close to three hundred ‘definitions’’ 4 Cosmopolitan Civil Societies Journal, Vol.2, No.3, 2010 ISSN: 1837-5391; http://utsescholarship.lib.uts.edu.au/epress/journals/index.php/mcs CCS Journal is published under the auspices of UTSePress, Sydney, Australia (ibid, p. 149). This cultural diversity in the meaning of term ‘culture’ shows that variation is not simply an epiphenomenon, but is essential to the cultural process itself. Culture is both diverse and shared, and subject to argument and struggle. Kroeber and Kluckhohn’s own definition did not slow down cultural diversification in the use of the term ‘culture’ either. As Leslie White suggested one problem with the Kroeber/Kluckhohn definition was that it implied that culture is both a real thing and an abstraction, a kind of average behaviour, or an ideal type (1954/1968, p. 17). The well known Saussurean division between the abstraction of langue and the immediacy of parole resonates with this ambiguity. The langue or culture is the ideal, and we use the disordered parole or behaviour to deduce or describe this ideal. Kroeber also implied culture was ‘the superorganic’, important in its own right and, even, independent of the social (1952, pp. 22-3). When arguing against the reduction of social life to race, genetics or individualism, this could be sensible. But seeing culture as a thing in its own right (independent of people and ethnographer) leads us to treat it as a unified whole with one meaning or dynamic as if it were a single text, or work of art, to be interpreted or translated1 . This dismisses the mess of people’s ongoing production of culture and swallows people’s lives in overarching homogenised text; making everything else they do irrelevant (McEachern 1998, p. 258-60). It ignores the areas of variation, or the ‘varieties of opinion’ that Malinowski pointed to, which ‘cannot be reduced or simplified according to any principles’ (1916/1974, p. 252) and thus avoids asking which cultural areas are the more prone to difference or variation and its generation. It separates culture irrevocably from living, surviving and struggle, imposes fixed boundaries, and ignores the ways that boundaries blend into each other. In taking the ragged form of culture seriously, and disagreeing with Geertz that ‘Culture is best seen… as a set of control mechanisms’ (1975, p. 44), I suggest it is useful to consider culture and social life as potentially self-disrupting, or ambiguous and riddled with disagreement. As Strathern suggests, more generally, ‘societies are not simply problem solving mechanisms: they are also problem creating mechanisms’ (1988, p. 33). Similarly, if 1 The master of the latter approach is Clifford Geertz (see in particular Geertz 1975: chapter 15). Treating culture as text may give spurious generality. Geertz can make the astounding claim that all Balinese have the same culture, ‘the same general beliefs… the same broad ideas of how their society is or should be arranged’ (Geertz & Geertz 1975: 2). We would not guess from this, for example, that there were Balinese Muslims (cf Barth 1993). Cosmopolitan Civil Societies Journal, Vol.2, No.3, 2010 5 ISSN: 1837-5391; http://utsescholarship.lib.uts.edu.au/epress/journals/index.php/mcs CCS Journal is published under the auspices of UTSePress, Sydney, Australia culture was a perfect whole of ordering then in theory it would be possible to live our whole life within a planned programme of experience, rather than (more realistically) dealing with ambiguity and having to make things up as we go along (Rosaldo 1993, p. 99-104). The ethnographer may need to be sensitive to these issues and not foreclose in premature wholeness and impersonality. Difficulties of Ethnographic Life Online Over the years, many people have suggested to me that online ethnography is impossible or dubious because of specific problems, which usually include: • The problem of the non local field site • The ethnographer being within their own culture • The hiddenness of people’s actions • The ambiguous identity/biography of participants • The absence of boundaries, and • The nature of online space Many of these criticisms arise in conversation, or are documented in informal forums such as the Association of Internet Researchers Mailing List, AIR-L, but it seems unfair to quote academics in communications which are not as carefully presented as they might be elsewhere, so the reader will have to take my remarks in good faith, and not as a specific refutation of a person’s position. This list is not just a list of problems faced by online ethnography, but constitutes a series of features of, and ambiguities present in, online life which cannot be ignored. They are the existential problems of such a life and everyone faces such issues. Researchers and people in online groups all have to work things out, test hypotheses, make judgments and make their culture. That these issues cannot always be resolved needs to be embraced and explored as part of ethnography, rather than used as reasons for abandoning ethnography. These problems also resemble problems which ethnography faces offline. People can lie and hide anywhere; they can information manage the observer, they can decide the boundaries around the ethnographer, they can interpret events and communications misleadingly and so on, as may have happened with Margaret Mead in Samoa (Friedman 1983). As such the 6 Cosmopolitan Civil Societies Journal, Vol.2, No.3, 2010 ISSN: 1837-5391; http://utsescholarship.lib.uts.edu.au/epress/journals/index.php/mcs CCS Journal is published under the auspices of UTSePress, Sydney, Australia ethnographer’s life should always be ‘critical’ and oscillatory. Boellstorf, speaking of ethnography in Second Life, distinguishes ethnography from an ‘outsider method’ (2008, p. 63), but it is the oscillating movement from insider to outsider, from involved to detached that characterises ethnography. [F]rom the perspective of the ‘marginal’ reflexive ethnographer, there can be no question of total commitment, ‘surrender’ or ‘becoming’. There must always remain some part held back, some social and intellectual ‘distance’. For it is in the ‘space’ created by this distance that the analytic work of the ethnographer gets done. Without that distance, without such analytic space, the ethnography can be little more than the autobiographical account of a personal conversion (Hamersley & Atkinson 1983, p.102). This also is not strange in normal life; the ethnographer may only exaggerate the motion. Choosing a Site The first problem of ethnography, similar online and off, is that of choosing a site, with groups you can work with, which allow you to study your interests, without causing too much disruption2 . In my case, Cybermind was chosen by accident. I originally joined the group to discuss its official topics of ‘the philosophical and psychological implications of subjectivity in cyberspace’, while searching for a field site. However, I found List members so apparently open about their offline lives and their experiences, that it seemed probable that fieldwork would not be particularly disruptive. It was high volume so there was a lot to analyse. I also enjoyed the List which, for me, was important as I was going to spend a lot of time reading it. The particular interests I brought to my work, such as wondering: how social control was performed without physical presence; how List life related to offline life; how communication conventions grew up; and how marked symbols were used; all seemed possible to investigate. Other topics presented themselves for research as I went along, and some topics were abandoned. I discussed the possibility of research with the moderator Alan Sondheim, who asked me to participate on the List before proceeding further. Several months later he asked me to announce my intentions to the other members and the work began. 2 As is well known in studies of the offline world, this produces a bias in which the groups studied ethnographically tend to be open or relatively powerless. Elite groups, criminal groups, and tightly bounded closed groups in general, tend to be less studied and present added problems of access and risk, but again closedness presents problems for the group members as well. Cosmopolitan Civil Societies Journal, Vol.2, No.3, 2010 7 ISSN: 1837-5391; http://utsescholarship.lib.uts.edu.au/epress/journals/index.php/mcs CCS Journal is published under the auspices of UTSePress, Sydney, Australia My own experiences have been central to my writings about the group. I was not a dispassionate observer. I participated in arguments, made political comment, had close off- list friendships, felt grief, frustration, boredom, and so on. But this resembled the life that any almost any involved member would lead. The main difference, if any, was perhaps the care and detail with which I tried to write about living on the List, and the degree of self- consciousness about my oscillations. There has never been any overt objection to my presence or questioning on the List, perhaps because my research was open, and my onlist writings fitted in pretty well with the blend of confessional, analysis, and political work on the list. Other List members have also written about their experiences on Cybermind, which adds to the data available3 . Cybermind was a fortunate site, but not an exemplary site, as every site has its own features – specificity is a feature of social life. I did, however, compare Cybermind to my experiences on other Lists, or internet forums, to try and work out what was specific and what was more general. The Ethnographer as Encultured Ethnographers are already in culture. They always have ideas of how the world works, and of how people and their actions should be interpreted. It is usually held that this produces blindness to features of their own social life, and the traditional solution to this problem is making everything strange by going somewhere ‘foreign’4 3 A list of published writings about the List can be found at: http://en.wikipedia.org/wiki/Cybermind . Hence the objection that in studying internet culture we remain within the familiar. However, strangeness alone does not allow us to see clearly. To the contrary, it is only with acculturation, with coming to know the language or idiom, coming to participate within rituals and daily life that the ethnographic venture can take place. Lowie argues that ethnography requires prolonged residence and familiarity as this helps overcome the hidden nature of social life (1937, p. 6), however this, 4 ‘To put it simply, ethnography is always about traversing the distance between the familiar and the strange’ (Grossberg 1989: 23). As a result, ‘ethnography constructs the other, from the very beginning, as different’ (ibid: 24) or, more realistically, involves ambiguous boundaries. Making the other different, does not have to make differences within the culture disappear, as Grossberg suggests. This can occur only if the ethnographer assumes the other is not plural. Sometimes strangeness has been thought to be beneficial as it destabilises the anthropologist but as Ang argues ‘self-destabilization may not at all be necessarily disruptive. On the contrary, it may be a form of confirming the intellectual’s sense of self. Self-reflexivity itself then becomes a hidden form of self-indulgence, a form of reinforcing a secure sense of ‘us’, rather than deconstructing it!’ (1989: 28). 8 Cosmopolitan Civil Societies Journal, Vol.2, No.3, 2010 ISSN: 1837-5391; http://utsescholarship.lib.uts.edu.au/epress/journals/index.php/mcs CCS Journal is published under the auspices of UTSePress, Sydney, Australia in turn, leads to the relative loss of strangeness. So if strangeness is the solution then, it too, is caught in paradox and oscillation. Even so, with interpretive processes generally, our attention tends to be drawn to those aspects of life which we don’t understand, or which seem different. Events that make sense to us are ignored, until we find our understanding or ability to act is inadequate. As Michael Agar writes, culture is ‘what happens to you when you encounter differences’ (1994, p. 20), or perhaps unexpected empathies. This works both ways for the people being studied and for the ethnographer – both can be changed by the encounter, as they both may change their ‘culture’ although in ways they do not always notice. Thus culture, as described, is driven by difference, by error even, rather than by sameness; a ‘symbol’ or ‘custom’ may ‘mean’ one thing, to one subgroup, another thing to another group. People are often strange to each other within the cultural group. When these problems of prior knowledge, ambiguity, hiddenness and difference, are treated as issues for everyone in online life, rather than just for the ethnographer, we can explore how people deal with these features in their social life, without having to resolve them into wholeness or certainty. Incompleteness and Hiddenness Although much is hidden from direct observation online, it is hidden to most people; so this is, again, a feature of online life. However, in any social environment, no one, no matter how much they try, can speak to everyone, or hear or see everything that goes on. People, including the ethnographer, will always be socially placed by other people and thus be reacted to in particular ways and placed in a particular politics. Therefore there are always things they will not observe or be told. Furthermore, no description can ever be full; it always has to be abridged, and the demands of writing for a particular discipline mean that order and regularity will be imposed upon the data. When the ethnographer says that people do certain things or believe certain things, that is an approximation that, at best, will refer to what most of those who are visible to the ethnographer do on certain occasions. This order, while providing comprehension may disorder the reality by deletion. Again, this is part of the nature of social life. The society we live in, is always partly hidden, always out of our complete control or understanding. Our ‘social blindness’ has effects on others as we interact with them and sometimes fail to interact satisfactorily. We do our best to make sense of life Cosmopolitan Civil Societies Journal, Vol.2, No.3, 2010 9 ISSN: 1837-5391; http://utsescholarship.lib.uts.edu.au/epress/journals/index.php/mcs CCS Journal is published under the auspices of UTSePress, Sydney, Australia as individuals in collaboration and competition, and the methods people use to make sense of failure or misperformance can be studied in themselves and perhaps related to other factors such as status, power, intimacy and so on. Incompleteness is a social fact to be investigated. Folktales of online life notoriously portray it as open to ‘deceit’ and to the perils of hiddenness5 . Again this is not just a problem for ethnographers, as it seems that people in Western cultures are less interested in engaging in a free play of multiplicity, as is often alleged (cf Poster 2001), than they are in uncovering truth and authenticity; although this varies with context, such as the particular internet forum involved, and the levels of collaboration or intimacy expected. Both these varying contexts, and the uncovering of authenticity, are surrounded by conventions which can be investigated. For example, many such conventions relate to the public/private continuum. Accuracy of information about ‘the self’ of others might be graded in terms of its ‘privacy’. ‘Truth’ is usually taken by group members and researchers, to increase in the order of: online in a group, online in a pair (in private email or private ‘room’ etc.), offline via phone, and offline face-to-face. The more intimacy is expected the more deceit or ‘play’ is considered inappropriate (Marshall 2007, pp. 105ff.). It is, however, clear that people can deceive as much in face-to-face, as they can online, but this does not seem to be as high on people’s list of fears. Cybermind tended to be confessional and relatively open. People talked about themselves, and these confessions could be compared to previous discussions and confessions (and frequently were by other people). Long-term residence with the group helps the ethnographer (and group member) to discover, and check up, more about other participants. People also gossiped about their experience with others. Thus if someone travelled to meet someone else and their appearance was not as expected from photographs, then that might be discussed offlist, as would failed relationships, conjectures about the motives of others and so on6 5 We might think of the ongoing tales of gender impersonation, of hidden paedophiles, of Nigerian spam frauds and so on. . Only comparatively rarely did this kind of discussion reach the public face of the list (it was hidden), so the ethnographer, like other involved members, needs to delve into the non- 6 I recently observed some teenagers interact with their friends through Facebook. As a result it would seem important to explore the oscillation they had between using tools on Facebook and telephones. Their Facebook activity spurred a lot of phone calls to check what other people thought was happening, to pass on secret comments and to make private arrangements. Hiddeness and the resulting ambiguity of all communication could not be ignored in an analysis. 10 Cosmopolitan Civil Societies Journal, Vol.2, No.3, 2010 ISSN: 1837-5391; http://utsescholarship.lib.uts.edu.au/epress/journals/index.php/mcs CCS Journal is published under the auspices of UTSePress, Sydney, Australia public side of online life, and the more the group portrays itself as community, the greater the need to go into private settings. This in turn might weaken the group as less activity appears in public and thus there is less in common. So, a way people order and give depth to their social experience may eventually destabilise the group the social experience grows out of (Marshall 2007). To pursue this offlist life, I had regular telephone and face-to-face contacts with some Australian members and regular offlist email contacts with both overseas and Australian members. I also met some overseas members as well. Gossip about Cybermind was a common ‘natural’ feature of conversation, although obviously with an Australian bias, and as Orgad (2005, p. 59, 62-3) reports, offline contact was often more open and full than communication through email. Of course gossip is not necessarily to be believed (people might be lying to the ethnographer for their own political purposes), but this is a problem faced with offline gossip as well. Communication implies the possibility of deceit. People online often seem to realise this, and attempt to work around it. The hidden and transient nature of the population may also affect social behaviour in other ways. Quite often the majority of the population may be ‘lurking’ or invisible, and the population of online groups can change fairly quickly. Snapshots of group composition will be temporary and always on the verge of being out of date. Again, rather than just being an ethnographic problem, this can have effects on the group. For example, group conventions and history may be relatively fragile and little discussed, as most participants may not know what the history refers to, and thus rather than generating closeness and tradition, it generates distance and breakage. Members of the group may have to continually re-earn their status, as newcomers do not know what they have done or, on the other hand, perhaps long time members support each other to reduce this potential loss, thus making it harder for new members to participate. Again, the effects of hiddenness, ignorance and ambiguity need investigating. An existential phenomenon that seems to arise for Westerners from this hiddenness is the state of being suspended between presence and absence which I have called ‘asence’; (Marshall 2007, pp. 89ff.). It is often not clear online whether a person, including yourself, is present or being heard, without the responses of others. There are no markers of being other Cosmopolitan Civil Societies Journal, Vol.2, No.3, 2010 11 ISSN: 1837-5391; http://utsescholarship.lib.uts.edu.au/epress/journals/index.php/mcs CCS Journal is published under the auspices of UTSePress, Sydney, Australia than communication. Even a graphic avatar may not tell you if the person ‘behind’ the avatar is actually there listening to you. Westerners seem to find this uncertainty difficult to deal with; they like confirmation of the other’s presence and of the other’s reaction to them. They may prefer flame directed towards them, which confirms their existence, than being ignored. Other behaviour, such as netsex, may also be permeated by the need to overcome asence (Marshall 2003). Friends in Facebook may mark presence as much as they mark status. Such effects may well differ between groups with different background or different offline culture. As Malinowski wrote, in a more conventional fieldwork context, it is of the nature of human life that as we move towards: the manifold details of personal life [our] methods of observations must become more complex and… [our] results less reliable. This cannot be remedied… The most exact of human observations is only approximate (1932, p. 238). As this implies, the argument that we can never know things entirely accurately, or that we can never give a definitive interpretation of what others are or are saying, is not an argument against trying to know as well as we can. If reality is vague, then that has to be factored into the research. Lack of Boundaries Because people rarely live online in just one internet group, the idea of group boundaries is challenged, and internet ethnography automatically has a tendency to become what Marcus has called ‘multi-sited’ (1999, p. 79ff). Similarly, asence implies that personal boundaries are not firm. Boellstorff, however, celebrates these non-existent boundaries, as he wants to study ‘virtual worlds’ in their own terms, without reference to a ‘real’ world (2008, p. 61ff.). He refers to Tylor’s definition of culture as ‘that complex whole’ and asks ‘what is a virtual world if not a complex whole?’ (ibid. p.66). However, virtual worlds are not ‘whole’ as you cannot eat pixels. There is no ‘virtual world’ without an offline world, which impacts upon it. The ‘virtual’ only exists via the ways people can access it, and via the potentially conflicting cultures they bring to it. Imposing strict borders between ‘virtual’ and ‘real’, is like studying a village without mentioning the contacts the villagers have with other villages, foreign aid, 12 Cosmopolitan Civil Societies Journal, Vol.2, No.3, 2010 ISSN: 1837-5391; http://utsescholarship.lib.uts.edu.au/epress/journals/index.php/mcs CCS Journal is published under the auspices of UTSePress, Sydney, Australia religious movements, governments, forestry companies and so on. For online worlds these externalities provide context, and varying contexts provide the varying meanings for what is done; thus undermining the claims of cultural ‘wholeness’. Assuming the online and offline are connected does not, as Boellstorff implies, mean that one is derivative of the other (ibid. p. 63), or that cultural construction and creativity does not happen online – merely that the borders are imprecise and multi-sited. Some online forums such as Twitter, for example, might take their meaning primarily from offline life, others from online life. These permeable, or ‘asent’, boundaries need to be investigated, rather than discarded. As implied above, the interactions between people on Cybermind which occurred offlist and in other forums was an important part of the List dynamics. For many members, the List intersected beneficially with their offline lives (Marshall 2007, pp. 241ff.). Boundaries between ‘virtual’ and ‘real’ were not firm at all. The internet, and these ongoing boundary ambiguities, draws attention to a central and rarely studied feature of modern life namely the temporary and voluntary associations that people form which may be separated, to some extent, from their daily lives. These can include: ‘support groups’, religious meetings, fitness classes, a group of fellow clubbers, sporting associations and so on. The people a person meets at these various groups may be kept quite separate from each other and, say, from their work life. People in the West seem to compartmentalise their lives as part of their general culture, and perhaps this enables their distinctions between online and offline lives at this moment. As more and more of their friends come online, this distinction may evaporate (as may be the case with Facebook). As it seems impossible to describe a person’s social life by observing their behaviour in one place, we have to recognise these other places, and may try and find out about them as part of our ethnography. Compartmentalisation may also mean that a person or group may act differently depending on the social form of the online group, say whether the online group is connected with work, is connected with neighbours, or is largely anonymous etc. Again, we cannot assume that online activity is separated from offline life. Cosmopolitan Civil Societies Journal, Vol.2, No.3, 2010 13 ISSN: 1837-5391; http://utsescholarship.lib.uts.edu.au/epress/journals/index.php/mcs CCS Journal is published under the auspices of UTSePress, Sydney, Australia Finally, these ambiguous boundaries affect academic discourse. It is much harder to pretend that academic or ethnographic research is completely separate from the group’s own cultural processes. With internet groups the chance increases that people will not only read what is written about them but will argue with it, making references to the same theorists that the ethnographer deploys. These theorists are themselves part of the cultural background; they are not independent of it, or separate from it. Similarly, it may be hard to draw borders of origin around ideas which have grown up for the ethnographer in the course of the fieldwork. Formative emails may include quotes, developments and responses from many participants. Ideas, used in analysis, may have arisen in dialogue without a fixed point of origin, and without it being possible to maintain barriers between what is written professionally and what is written out of life in general. The boundary between ‘us’ as ethnographers and ‘them’ as people is increasingly destabilised, and this should be recognised as part of the social process of ethnography and cultural model making. The group has to be acknowledged as part of the creative process and, this too, might need investigation. Place? Clearly place can no longer be use to draw boundaries around social lives, although place and culture can still have an effect. Online place, is further constituted by texts or graphics, or both, all of which affect the kinds of things that can be done in that space, the way it is used and the way messages will be interpreted. It is always a ‘locale’ to use Gidden’s term (1984, p. 118). This locale gives information about the ways that the space should be used and the kinds of things which should be done. Similarities of place, code for similarity of action and behaviour. People tend to use the conventions of locale to define their internet group, or the internet in general as a specific kind of locale for political and rhetorical purposes, although this does not imply instant unity (Marshall 2001). The social constitution of online space does not mean that the online world is deterritorialised. As well as being immersed in offline networks, people are situated in particular conventional spaces when they access the internet. These spaces can affect the ways they can access the internet, the kinds of things they can access, the speed with which they can respond, and the risks involved in that access. It may also influence whether they are 14 Cosmopolitan Civil Societies Journal, Vol.2, No.3, 2010 ISSN: 1837-5391; http://utsescholarship.lib.uts.edu.au/epress/journals/index.php/mcs CCS Journal is published under the auspices of UTSePress, Sydney, Australia online by themselves, in a private room, in an internet café or surrounded by others including offline friends or kin (see Miller and Slater 2002). Again we can explore, the ways in which the local and the global intersect, conflict, connect and misconnect. The boundaries are by no means firm and, online, they may produce new social forms. Again the ethnographer (and the inhabitant as ethnographer) is faced with boundaries which are constructed, reinforced, permeable, ambiguous, or overcome, and which may be made so, in different ways, by different people and different groups. Archive and Memory That Cybermind had archives, together with my own collection of emails, meant that it was often possible to ‘relive’ events and check if my memories were accurate. This often produced interesting discrepancies, which could be investigated; something which is impossible in normal fieldwork. In this case, ethnographic notebooks, with their spur of the moment interpretations of events, do not completely replace what happened. However, notebooks are still useful to remind the ethnographer of how they saw and felt events at the time. With archives it is possible to observe what factors are important for how memories, or narratives of things, were constructed, by comparison. For example, I observed that for events in which the group had seemed flooded with flame, the flaming messages often turned out to be quite a small proportion of the total messages. As a result, I could suggest that the perception of flooding occurred because these messages had a coherent tone and subject, while the other messages tended to be fragmented in tone and subject, thus not making as much subjective impact. Times in which the coherence of one set of messages seemed dominant were also the most memorable times on the list, for me and for others, and were often used to justify particular behaviours, conventions, or special times onlist, such as when the death of List co-founder Michael Current was frequently remembered as marking the birth of List community. Inaccurate perceptions can have a big effect on social life. However, reading archives is not equivalent to experiencing life on the group. In reading the archives the ethnographer is missing the varying density of messages, the speed with which they arrive and so on, and hence the ambient moods which influenced how people read the group (whether hurriedly of slowly for example), or the way they interpret the group in terms Cosmopolitan Civil Societies Journal, Vol.2, No.3, 2010 15 ISSN: 1837-5391; http://utsescholarship.lib.uts.edu.au/epress/journals/index.php/mcs CCS Journal is published under the auspices of UTSePress, Sydney, Australia of its dominant messages at that time. Reading 40 messages which have arrived within quarter of an hour is a different experience to reading 40 messages which arrive over a couple of days. Unobserved Observer Fallacy People often suggest it is possible to do online ethnography by ‘lurking’ without participating in the group and thus not altering its behaviour, and indeed some people have announced they have been doing this form of ethnography on Cybermind. However this assumes that reading is the same as participation, and is open to the same objections as arise for reading archives alone. If the meaning of a communication inheres in the interaction and resolution of behaviours, then a person who is only reading cannot check their understandings, so their resolution of communication will be even more tenuous than usual. The researcher’s own categories and formulations largely remain uncontested. Such research seems particularly open to charges of producing distortion through the importation and imposition of external categories, whereas interactive ethnography has a potential towards openness because dialogue often goes in different directions to those intended. Likewise hiddenness suggests that the ethnographer might be able to deceive the group they are investigating and gain perspectives from another social role. This process is likely to reproduce category clichés, as the ethnographer goes about performing those roles, and provoking responses, by the conventions they already ‘know’. Hence they might not learn as much as they expect. On the whole, the greatest objections that people on Cybermind had to any research was when it was unannounced, was announced with no possibility of participation, or when they just discovered the publications. So ethical problems emerge. To avoid these problems and to be open about the research, all my writings about the List, were made available to List members before publication for comment and/or protest. This proved a useful way of getting further information, and correcting overt errors, but clearly could be difficult if the group wanted to remain closed to the outside world, or if people had objected to publication; it is possible openness can lead to closure. Problems of Sourcing 16 Cosmopolitan Civil Societies Journal, Vol.2, No.3, 2010 ISSN: 1837-5391; http://utsescholarship.lib.uts.edu.au/epress/journals/index.php/mcs CCS Journal is published under the auspices of UTSePress, Sydney, Australia This leads to the final problem. If the boundaries between ethnographer and subjects are open, then how do we acknowledge the group in our work? Over the ten years I was researching Cybermind, the issue of how people should be quoted came up many times. While changing people’s names is standard practice in internet ethnography (cf Kendal 2002, pp. 241-2), this practice both denies the obligation, and the role of the group in producing the ideas of analysis. Making people anonymous also brings up questions of intellectual property rights as Gurak points out (1997, p. 137), especially for people who live by payment for ideas and texts. By not attributing statements, ideas or art works, the ethnographer is potentially engaged in theft, and putting hard boundaries in place to support their own livelihood. Removing names suggests that boundaries between ethnographer and subject are easily maintained, which, as we have seen, is debateable. These hidden obligations can be exploited by the ethnographer to make their version of the culture seem radically different from the version espoused by members of the group. By referring to people it becomes easier to trace the way the ethnographer’s ideas have risen out of interactions with the group and owe their existence to the group, rather than to a privileged and autonomous academic tradition. By acknowledging others, those others are able to say that I, as ethnographer, have misunderstood. If identity is made uncertain or people are blended into fictions then this is impossible. While acknowledgement increases the vulnerability of the ethnographer to questioning, it could also increase their commitment to being accurate. Yet, it is clearly possible for identification to lead to problems for the group (Reid 1996), and openness may not always be appropriate if what is discussed is classed as private (and this too may not be clear). Discovering what is sensitive, private, or what should not be reported, is part of ethnographic research, will not always be the same in different places, and is always potentially fraught. We can make informed guesses but ultimately the consequences are unpredictable. Being aware of these problems will probably lead the ethnographer to self- censor to protect the group and group members from the incursions of others – especially if they are open about the identity of the group. This may produce distortions of the culture as it is presented, however the censorship is generally censorship of an interpretation. For example, I may have my own opinions about what actions helped lower the population of Cybermind after the Iraq war. These opinions are not necessarily true, and it certainly does not change the effect of those actions if I report them or do not report them. The place and time for comment would have been when they occurred. Reporting them afterwards in a Cosmopolitan Civil Societies Journal, Vol.2, No.3, 2010 17 ISSN: 1837-5391; http://utsescholarship.lib.uts.edu.au/epress/journals/index.php/mcs CCS Journal is published under the auspices of UTSePress, Sydney, Australia format in which those I might criticise could not reply, could seem like an attack. Whatever I decide to do, I cannot remove the ambiguity or the hiddenness from the account, and indeed the self-censorship, to the extent that it exists, is no more a construction of culture than reporting events would be. These problems are, again, not just online problems but faced by ethnographers generally in an interconnected world. Whereas some problems arise because people online are supposedly too anonymous for us to do fieldwork, here it is recognised that they are perhaps not anonymous enough. At least most people on Cybermind had some knowledge of what social research involved and could give consent to the ethnographer’s presence with some degree of awareness of potential consequences and, as the research was open, did not have to consent. This gives another reason for making papers available before publication; another opportunity is provided for objections. In general, Cybermind members objected to the List’s name and long-time members’ identities being concealed unless they requested, or when their words may be the unconsidered words of a moment. So I worked with this, privately deciding not to name people if the topics being written about could be considered potentially harmful to them and to keep private correspondence and interviews anonymous unless otherwise requested. Finally, given many passing people may not have been aware of the research, I usually quoted such people anonymously. Any practices of acknowledgement would have to vary with the group, and be in consultation with the group, but there are no clear guidelines which can satisfy everybody in every situation; that is the nature of social life. Online Ethnography and Culture: Conclusions Ethnography involves participating in life with the group being studied, recording what happens and interacting with its members, for a relatively long time to gain familiarity with events and local meanings. It may make use of many tools such as: surveys, interviews, spontaneous queries, counting words in texts, comparing notebooks with archives, recording of our own experiences, and so on. Ethnographic practice and convention has to adapt to the group being studied, the kinds of problems the ethnographer is interested in, and to the audience being written for; therefore it cannot always be defined adequately in advance; it is inherently ambiguous. 18 Cosmopolitan Civil Societies Journal, Vol.2, No.3, 2010 ISSN: 1837-5391; http://utsescholarship.lib.uts.edu.au/epress/journals/index.php/mcs CCS Journal is published under the auspices of UTSePress, Sydney, Australia Choosing a group to research is always something of a haphazard matter. It is perhaps best to approach the group with relatively clear topics for investigation while expecting that some will be abandoned and other topics will appear. What is discovered will partly be contingent, depending on the events that happen, and the events people talk about. While the ethnographer may participate passionately, they may find it useful to oscillate, to question themselves – especially about their certainties, and what they find as differences. It is also useful to remember that the ethnographer’s problems in carrying out their work may be similar to the problems that other people have. As all sites are unique and have particular issues, it is therefore useful to have some familiarity with other sites and other ethnographies for comparative purposes. Much of any life, but particularly life online, is hidden. The ethnographer has to be aware of this, and the limits of their knowledge. This is a problem for anyone online, so we can study how people attempt to resolve that problem together with the consequences of that attempted resolution. Rather than being put off by hiddenness, we can investigate the social effects of that hiddenness and people’s fears, uncertainties or certainties about it. This may require that we try and participate in the life around the online group, and move into other spaces online and off. Boundaries are vague and crossed all the time, so it may be useful to investigate those boundaries, to find what we and others bring to the group from offline life, and what we take back to offline life. Online worlds are not entire in themselves, but the level of interdependence may vary with the particular world. We oscillate between involved and detached, between local and other groups. Like everyone we are not just insiders. One further consequence of these blurred boundaries is that we cannot assume that we and our theories, or understandings, are disconnected from the group’s culture. Ethnography is inherently collaborative, and this raises issues around acknowledging the group and its members. In making sense of what people in the group being studied do, how they interact, and how they interpret others, the ethnographer is explicitly constructing a model of ‘culture’. This is a more formalised version of what everyone else in the group has to do. People generate culture in the course of living their lives, in moving from one group to another. The process may be more complex if the ethnographer or person comes from a radically different culture, but in general the problems the ethnographer faces online are the problems the group members also Cosmopolitan Civil Societies Journal, Vol.2, No.3, 2010 19 ISSN: 1837-5391; http://utsescholarship.lib.uts.edu.au/epress/journals/index.php/mcs CCS Journal is published under the auspices of UTSePress, Sydney, Australia have to negotiate. In any case, people in contemporary society may continually move into interaction with new temporary groups. This renders ethnography even more of a social phenomenon. Culture is what enables a person to tell, or to make a case that, an event is significant. As such it involves conjecture about the motives, or behaviour, of others. It tries to make the hidden visible. In this understanding, one event cannot be understood without understanding other events, and hence the culture made by the ethnographer tends to become relatively coherent and possibly hide the actual disjunctures of life. Furthermore, the tools the ethnographer uses to understand events will change the model of culture they come up with. The ambiguity of, and disagreement over, the term ‘culture’ in academia shows that contestation and variation of meaning, is part of culture in general. Variation and argument is important. Simplification will occur, and it is easy to smooth over problems, or to delete events that cannot be integrated into the model and which are thus labelled as random or unimportant, so simplification needs care. Everyday life is messy, and this mess may be creative. As Barth argues, variation or misperformance within a society cannot simply be considered an epiphenomenon, or ‘noise’ produced by an underlying order, and thus irrelevant. Variation ‘should emerge as a necessity from our analysis’ (Barth 1993, p. 4). There is likely to be more than one culture in the group, and cultures are likely not to have neat or logically consistent boundaries. Any boundaries, and the mechanisms whereby they come to exist, have to be discovered. So will the effects of online ‘place’ and its nature. This view also tends to help us see that society and culture are not just mechanisms of order (however sophisticatedly homeostatic), in which human behaviour is reduced to the side effects of ambiguously present, or ‘asent’, structures. Societies and cultures can also be mechanisms of disordering. We tend to perceive ‘order’ as imposed upon ‘chaos’, when particular kinds of social order might depend upon, or create, particular types of ‘chaos’. Contemporary global capitalism might depend on its ability to make and distribute garbage, yet the abundance of garbage may threaten it. Cybermind’s ways of making ‘community’ could also undermine that community (Marshall 2007). With this view we might come to see disruption and disorganisation as more than evidence of simple failure or anomie. Rather than simply being a disruption, or something to be controlled, argument, dispute, disorder or 20 Cosmopolitan Civil Societies Journal, Vol.2, No.3, 2010 ISSN: 1837-5391; http://utsescholarship.lib.uts.edu.au/epress/journals/index.php/mcs CCS Journal is published under the auspices of UTSePress, Sydney, Australia conflict can be the site of, or provocation for, cultural creation. Similarly, abandoning an association of culture with order allows us to look at the ways in which societies head for disaster through their modes of ordering (such as extending the order of markets), and help us not to reduce all problems to inadequacies. Furthermore, we don’t have to assume smooth continuity, or complete disjunction between cultural artefacts. Their relationship might be an ongoing point of negotiation and conflict, and might never be entirely resolved. Everyone is jointly engaged in interpretation and making sense of others, in making culture for others, and then using this model as a tool to interact with those others. Everyone is trying to make sense of daily life and trying to make the behaviour of others predictable and to use ‘cultural tools’ persuasively so as to fit into the valued categories which are open to them. As such, the problems of the ethnographer exemplify the same kind of problems that people experience online. They are not problems to be avoided but to be explored, as are the ways that people go about dealing with these problems themselves, as it might tell ethnographers something useful for their research. While, ethnographic models might be more formal than most, and hopefully have a wider than usual view of the issues involved, what ethnographers are doing is not unique. We are all constructing ‘culture’ with each other and for each other; conducting battles, and harmonies for the life of the group. As Hine writes, the position that internet ethnography is particularly unreliable, ‘changes somewhat if we recognise that the ethnographer could instead be construed as needing to have similar experiences to those of informants, however those experiences are mediated’ (2000, p. 10). The next step is to realise the basic similarity of procedures involved, and that the group’s perceived culture will be different for different people; there is no overarching single vision. It is always based in interpretation, experience and conflict. Ethnography is not above, but within, cultural interaction. References Anon 1951, Notes and Queries on Anthropology, 6th Edition, Routledge Kegan Paul, London. Agar, Michael 1994, Language Shock: Understanding the Culture of Conversation, William Morrow, NY. Cosmopolitan Civil Societies Journal, Vol.2, No.3, 2010 21 ISSN: 1837-5391; http://utsescholarship.lib.uts.edu.au/epress/journals/index.php/mcs CCS Journal is published under the auspices of UTSePress, Sydney, Australia Ang, Ien 1989, ‘Beyond self-reflexivity’, Journal of Communication Inquiry, vol.13, no.2, pp.27-29. Barth, Frederick 1993, Balinese Worlds, Chicago University Press. 1995, ‘Other knowledge and other ways of knowing’, Journal of Anthropological Research, vol.51, no.1, pp. 65-8. Beattie, John 1966, Other Cultures: Aims, Methods and Achievements in Social Anthropology, Routledge, London. Boellstorff, Tom 2008, Coming of Age in Second Life: An Anthropologist Explores the Virtually Human, Princeton University Press. Freeman, Derek 1983, Margaret Mead and Samoa: the Making and Unmaking of an Anthropological Myth, Australian National University Press, Canberra. Geertz, Clifford 1975, The Interpretation of Cultures, Hutchinson, London. Geertz, C. & Geertz, H. 1975, Kinship in Bali, University of Chicago Press. Giddens, Anthony 1984, The Constitution of Society, Polity Press, Oxford. Grossberg, Lawrence 1989, ‘On the road with three ethnographers’, Journal of Communication Inquiry, vol.13, no.2, pp.23-6. Hammersley, M., & Atkinson, P. 1983, Ethnography: Principles in Practice, Tavistock, London. Hine, Christine 2000, Virtual Ethnography, Sage, Thousand Oaks, CA. Lowie, Robert 1937, The History of Ethnological Theory, Holt, Rinehart and Winston, NY. Kendall, Lori 2002, Hanging out in the Virtual Pub: Masculinities and Relationships Online, University of California Press, Berkley. Kroeber, A.L 1952, The Nature of Culture, University of Chicago Press. Kroeber, A.L. & Clyde Kluckhohn 1952, Culture, a critical review of concepts and definitions, Papers of the Peabody Museum of American Archaeology and Ethnology, Harvard, vol. 47, no. 1. Malinowski, Bronislaw 1932, Sexual Life of Savages in North Western Melanesia, RKP, London. 1974, Magic, Science and Religion and Other Essays, Souvenir Press, London. Marcus, George 1999, Ethnography Through Thick and Thin, Princeton University Press. 2007, ‘Ethnography two decades after writing culture: From the experimental to the baroque’, Anthropological Quarterly, vol.80, no.4, pp. 1127-1145. Marshall Jonathan 2001, ‘Cyberspace or Cybertopos: The creation of online space’, Social Analysis vol.45, no.1, pp.81-102. 2003, ‘The sexual life of cyber-savants’, The Australian Journal of Anthropology vol.14, no.2, pp.229-248. 2007, Living on Cybermind: Categories, Communication and Control, Peter Lang, NY. McEachern, Charmaine 1998,‘A mutual interest? Ethnography in anthropology and cultural studies’, The Australian Journal of Anthropology, vol.9, no.3, pp.251-64. Miller, Daniel & Slater, Don 2002, The Internet: an Ethnographic Approach, Berg, Oxford. Orgad, Shani 2005, ‘From online to offline and back: moving from online to offline relationships with research informants’ in Christine Hine (ed.) Virtual Methods: Issues in Social Research on the Internet, Berg, Oxford. 22 Cosmopolitan Civil Societies Journal, Vol.2, No.3, 2010 ISSN: 1837-5391; http://utsescholarship.lib.uts.edu.au/epress/journals/index.php/mcs CCS Journal is published under the auspices of UTSePress, Sydney, Australia Poster, Mark 2001, Whats the Matter with the Internet? University of Minnesota Press, Minneapolis. Reid, Liz 1996, ‘Informed consent in the study of on-line communities: a reflection on the effects of computer-mediated social research’, The Information Society, vol.12, no.2, pp.169-74 Rosaldo, Renato 1993, Culture and Truth: the Remaking of Social Analysis, Beacon Press, Boston. Strathern, Marilyn 1988, The Gender of the Gift, California University Press. White, Leslie 1954/1968, ‘On the concept of culture’, in Robert Manners & David Kaplan (eds) Theory in Anthropology: A Sourcebook, RKP, London.