Social Life in Virtual Worlds

Multi-user virtual environments (VE's) of various types are becoming ever more widely used, but we know little about how people interact in them. It has been suggested, for example, that there are 'eroding boundaries between the real and the virtual' (Turkle, 1995: 10), or that new types of 'community' are made possible by the internet (Rheingold, 1993).

My research has tackled these problems from various angles: on the one hand, I think we need participant observation studies; studies which analyze the interactions in populations of existing multi-user virtual environments. On the other, it is necessary to study the interactions between users 'experimentally'; varying the conditions in which participants interact, the technologies they use and they tasks they perform. Finally, research on social interaction in virtual worlds needs to put in the larger context of studies of other technologies and communication networks, and how these have changed our lives.

Below, I will describe some research on multi-user virtual reality ('VR', for a definition, see Schroeder, 1996: 25) and offer some suggestions for how to proceed further. We are only beginning to make sense of this area, but the main argument that I wish to make is that the way forward lies in integrating the results of various approaches. Before we come to this research, let us examine some previous studies of technologies related to multi-user VE's.

Some Research on Computer-Mediated Communication and other Communication Networks

In relation to 'community', it is important to note two points: first, the users and the uses of computer networks are limited (Wellman, 1996: 214-6). Quite apart from the fact that computer-mediated-communication (CMC) plays only a small part in everyday life, people engage in CMC for limited purposes. Second, the social rules which govern these computer networks have their own dynamic. As Wellman puts it: 'CSSN's [computer-supported social networks] are developing norms and structures of their own. They are not just pale imitations of "real life". The Net is the Net'(1996: 231). In other words, CMC cannot just be understood as real-life behaviour transferred onto networks - instead, these networks involve different forms of social behaviour. One implication of this is that it is impossible to apply a concept like 'community' to CMC - even provided that this concept is applicable in the 'real' world.

Other studies have focused on the characteristics of CMC in organizations or medium-size groups instead of the larger 'community'. Many of these have focused either on the depersonalizing effects of CMC (absence of certain cues in communication, or self-interested behaviour due to the absence of face-to-face consequences, etc) or on 'participation-' or 'status-equalization effects', the allegedly less hierarchical or more egalitarian processes in decision-making and the like (for 'participation' see, for example, Straus, 1997, and more generally Bonito and Hollingshead, 1997; for 'status', an example is McKenney, Zack and Doherty, 1992). The problem with these studies, as many of them would acknowledge, is that this effect may wear off over time - or it may intensify over time. The problem, again, is the relation between studies of experimental settings and everyday life; or, in the case of observations of everyday settings, of putting the study of one aspect of people's lives into a larger context.

Turning to research on the relation between social interaction in CMC's and 'real world' sociability, again, there have been a number of studies of this topic. Here, it is sufficient to mention a recent survey finding that CMC sociability does not adversely affect real life social relations. Internet friendships or participation in online 'communities' were found to be a complement to - rather than replacing or taking away from - existing friendships and civic participation (Katz and Aspden, 1997). This finding goes against the widely held view that CMC may become an alternative to, or substitute for, 'real' world social relations.

It is also necessary to mention more general studies of the effect of communication networks on society. Fischer (1992), for example, has argued that the telephone has not fundamentally changed American society; it has at best transformed pre-existing social patterns. In a similar vein, Nye (1997) points out that the introduction of new network technologies has always led to expectations of fundamental social change - such as the end of shopping, the advent of direct democracy, and the demise of direct interpersonal relationships. In fact, he suggests that new network technologies have gone hand in hand with a steady rise in the time spent shopping, created or perpetuated business monopolies and oligopolies, enhanced government surveillance and social control, and left the importance of interpersonal relationships intact.

All in all, we need to be careful about the levels on which we analyze communication or interaction in computer-mediated networks - historical or short-term, micro- or macro-interaction, 'natural' groups or task-related ones, experimental or everyday behaviour, and the like.

Structure and Interaction: The Example of Religion in a Multi-User Virtual World

A good example of a multi-user VE that is somewhere between the level of experimental small group behaviour and the larger virtual 'community' is religion in an online virtual world (for access to, and images from, a range of these online virtual worlds, see http://www.ccon.org). I will not go into specific details of this example here because of the sensitive ethical issues that are raised in research on a semi-private online group (see Schroeder, Heather, and Lee, 1999, for a discussion of these issues and a more detailed analysis). Suffice it to say that the phenomenon in question is an (irregular) weekly church service that is held in an online multi-user desktop virtual world. The service typically involves several avatars standing in a virtual church and engaging in prayer for approximately one hour (with a good deal of before and after socializing) by means of text.

Several things are noteworthy about this church service in the light of the points made earlier:

1. Unlike in online virtual worlds generally, in which there are few pre-assigned social roles, in this religious world the leader (the person who would be a 'priest' or 'minister' in a 'real world' service) is clearly set apart from the 'congregation'.

2. As in other online virtual worlds, there are differences between 'insiders', those who participate regularly and may know each other outside of the services, and occasional participants (Schroeder, 1997). But these differences are also different from the distinction between 'insiders' and 'outsiders' that can be made in other online virtual worlds because in this case, the cohesion among 'insiders' revolves around shared religious sentiment rather than merely around how well the participants know each other.

3. The fixed place and regular time of the church service make this type of social interaction much more patterned than the spontaneous encounters in most other online virtual worlds of this type (which are mainly used for entertainment).

4. This 'patternedness' also applies to the content of the service: the structure of the service with its beginning, middle and end; the way that avatars tend to stand relatively immobile during the service with the leader near the altar and participants in a circle around him/her; the structure of responses to the leader's sermon and the flow of conversations; and finally, the seriousness of the tone of the service.

5. Despite the fact that this group is more cohesive in its social interaction than other groupings in online virtual worlds, these ties are also rather limited insofar as they center on religion - quite apart from the fact that the contacts between members, or the occasions on which the group reaffirms itself in a Durkheimian sense, are infrequent.

This form of virtual religion provides a good illustration of the differences and similarities between CMC and real world behaviour. These include:

1. That the content and the social roles of a 'real' world service are to some extent 'imported' into the virtual world. On the other hand, the virtual service has certain constraints, such as the impossibility of more prolonged or sustained interactions, as well as opening up possibilities, like the more participatory style of interactions compared with a 'real' world service.

2. The emotional intensity or ritual solidarity of a face-to-face world church service can only carry over to a very limited extent into the virtual world. On the other hand, the service in the virtual world does often allow for more candid exchanges about intimate personal matters than a 'real world' service.

3. The service in the virtual world allows access to participants who may otherwise not have access. But the relations among participants will also not be as deeply rooted in the sense of cohesion stemming from social relations around a religious community in the 'real' world.

These points can now be related to the broader issues raised earlier: the importance of the micro-macro linkage to the study of multi-user virtual worlds and the nature of networked computer communication generally. The virtual religious service is not a social phenomenon that can be explained by reference to the 'face-to-face' (or avatar-to-avatar) behaviour within this small group alone. Nor is it useful to treat this group as merely part of a larger emerging 'virtual community'. For in the case of this virtual world, it is necessary to take into consideration:

1. How this religious virtual world fits into the larger set of online virtual worlds and how these, in turn, fit into communication networks and social interaction in the 'real' world.

2. How (or whether) the observations that are made about this small group add up to something more than just 'local' behaviour, and how the more widespread adoption of this type of interaction in the virtual world may have wider repercussions for the practice of religion.

At this stage, we can turn for a different perspective to the experimental study of interaction in multi-user virtual environments. For example, in one such study, we examined small group behaviour by comparing interaction in a multi-user VE with social interaction in a similar real world setting (see Slater, Sadagic, Usoh and Schroeder, 1998). This study found, albeit in an exploratory way, that the participant with greater computational resources (using an 'immersive' VR system with a head-mounted display and a more powerful computer, whereas the other two participants used a 'desktop' VR system with a less powerful computer) was regarded by the other two participants as being the leader of the group, even though in the real world setting none of the three was regarded as the leader any more than another.

Another finding of this study was that users' sense of 'presence' and 'co-presence' - or having a sense of being 'inside' the virtual environment and of being inside the environment with others - were positively correlated (on the notion of 'presence', see Slater and Wilbur, 1997). The study also found that participants were uneasy, for example, about being able to walk through walls and through each other's avatars, perhaps because this fails to correspond with our sense of how the 'real' world works and with how the autonomy of our 'real' bodies should be respected.

The reason for mentioning these findings is that it would be impossible to gain this kind of insight into how multi-user VR technology affects the social relations between participants - or how users respond to the technology in different ways - by means of participant observation alone. To give a concrete example: how could we know about whether the use of different types of technology - as one variable - made a difference to the behaviour of users during the virtual church service? Or again, how could ascertain the sense of 'presence' and 'co-presence' at the church service without testing this in relation to a comparable real world situation - in other words, a 'real' church service? Some Conclusions

It is too early to begin to integrate research on multi-user VE's with much more long-term and large-scale (macro-) studies of the role of communication networks in society such as Fischer's and Nye's. But it is nevertheless important to think about the way these studies might relate to each other. What are the implications, for example, of the finding about 'leadership' in a small group experimental setting for communication networks at large: does the relationship between being recognized as a 'leader' and the use of more powerful VR technology also apply to other communication technologies? Or what, for example, are the implications of this finding for the other, smaller-scale studies that were mentioned about 'participation-' and 'status-equalization' effects: might having a more powerful technology in some way cancel out the these effects - if indeed they can be shown to be significant in everyday settings?

It is interesting to note that there is one finding that is common to many of the studies of CMC that have been discussed here: namely, that participants have a strong sense of 'being there' or of 'presence' in the computer-mediated environment -whether these are text-based MUD's, fully immersive VE's, or desktop VE systems. What accounts for this 'presence': the interface, or the interpersonal relations within the environment? The study mentioned earlier which compares small group behaviour in a VE with real world behaviour (Slater, Sadagic, Usoh and Schroeder, 1998) can go some way towards answering one aspect of this question, but this would need to complemented by research such as Meyrowitz's (1985) to see how these small-scale changes might fit into the larger transformations brought about by electronic networks in our 'sense of place' in society as whole. If we are then able to establish whether the sense of 'presence' or of 'being there' is more related to the technology (or the interface) or to the 'co-presence' (or the social relations) among those using the technologies - then we shall have a systematic understanding of the social implications of electronic networks.

To summarize: I hope to have shown, by reference to several examples, how the study of social relations in multi-user VE's (and, by implication, in electronic networks generally) requires the combination of several levels of analysis and the combination of a number methods of research. The study of multi-user VR and related technologies is still in its infancy. But there are now many specialized studies of CMC which can usefully draw on a longer tradition of research on the role of communication technologies in society. And even if the strengths of these various areas of research may vary, it is only possible to counteract the current hyperbole surrounding new information and communication technologies by putting research about their implications into context.

References

Bonito, Joseph and Hollingshead, Andrea. 1997. Participation in Small
Groups. In Brant Burleson, editor, Communication Yearbook 20, pp.227-61.

Fischer, Claude. 1992. America Calling: A Social History of the Telephone,
Berkeley: University of California Press.

Katz, James and Aspden, Philip. 1997. A Nation of Strangers. Communications
of the ACM, 40(12) pp. 81-6.

McKenney, James; Zack, Michael and Doherty, Victor. 1992. Complementary
Communication Media: A Comparison of Electronic Mail and Face-To-Face
Communication in a Programming Team. In Nithin Nohria and Robert Eccles,
editors, Networks and Organizations: Structure, Form and Action, Boston:
Harvard Business School Press, pp.262-87.

Meyrowitz, Joshua. 1985. No Sense of Place: The Impact of Electronic Media
on Social Behaviour, Oxford: Oxford University Press.

Nye, David. 1997. Shaping Communication Networks: Telegraph, Telephone,
Computer. Social Research, 64 (3) pp.1067-91.

Rheingold, Howard. 1993. The Virtual Community: Homesteading on the
Electronic Frontier. Reading MA: Addison Wesley.

Schroeder, Ralph. 1996. Possible Worlds: The Social Dynamic of Virtual
Reality Technology, Boulder, CO: Westview Press.

Schroeder, Ralph. 1997. Networked Worlds: Social Aspects of Multi-User
Virtual Reality Technology. Sociological Research Online, 2(4),
http://www.socresonline.org.uk/socresonline/2/4/5.html.

Schroeder, Ralph; Heather, Noel and Lee, Raymond. 1999. 'The Sacred and the
Virtual: Religion in Multi-User Virtual Reality', Journal of
Computer-Mediated Communication, 4(2),
http://www.ascuc.org/jcmc/vol4/issue2/schroeder.html.

Slater, Mel and Wilbur, Sylvia. 1997. A Framework for Immersive Virtual
Environments (FIVE): Speculations on the Role of Presence in Virtual
Environments, Presence: Teleloperators and Virtual Environments, 6(6),
pp.606-13.

Slater, Mel; Sagadic, Amela; Usoh, Martin and Schroeder, Ralph. 1998.
'Small Group Behaviour in a Virtual and Real Environment: A Comparative
Study', BT Presence Workshop, Martlesham Heath, June 11-12, available at
http://www.cs.ucl.ac.uk/staff/m.slater/BTWorkshop/Inhabit/ (also
forthcoming in Presence: Journal of Teleoperators and Virtual Environments).

Straus, Susan. 1997. Technology, Group Process, and Group Outcomes: Testing
the Connections in Computer-Mediated and Face-to-Face Groups.
Human-Computer Interaction, 12 pp. 227-66.

Turkle, Sherry. 1995. Life on the Screen: Identity in the Age of the
Internet. New York: Simon and Schuster.

Wellman, Barry et al. 1996. Computer Networks as Social Networks:
Collaborative Work, Telework, and Virtual Community. Annual Review of
Sociology, 22 pp. 213-38.