PRESENCE IN SHARED VIRTUAL ENVIRONMENTS AND VIRTUAL TOGETHERNESS
Nat Durlach and Mel Slater
Research Laboratory of Electronics
Massachusetts Institute of Technology, Cambridge, MA 02139
The subjective sense of presence is receiving substantial attention from engineers, computer scientists, and psychologists concerned with virtual environments, teleoperators, and human- machine interfaces (see, for example, Slater and Wilbur, 1997, and the references cited in that article). Important unresolved issues concerning presence include (a) the definition of presence, (b) methods of measuring presence, (c) the identification of factors creating (or destroying) presence, and (d) the relation of presence to task performance.
In this note, we ignore these unresolved issues and focus instead on a topic closely related to the sense of presence, namely, the sense of being with someone or, equivalently, the sense of togetherness. This topic is not only of theoretical interest; it has great practical importance. Telecommunication companies would probably invest billions of dollars to know how to artificially effect this sense of togetherness. And airline companies would probably do likewise to prevent this knowledge from being discovered! This notion of togetherness, which is assumed to have varying degrees of realization, is taken as the fundamental variable of interest in this work.
In the following paragraphs, we consider two factors that we believe are particularly relevant to creating a sense of togetherness among individuals who are sharing a common virtual environment: (a) presence in a common virtual environment and (b) communication among participants in this common virtual environment. The note concludes with a few remarks on the special role of touch in creating a sense of togetherness. Further consideration of touch is then presented in our companion paper entitled An experiment on the influence of haptic communication on the sense of being together by Ho et al.
For a number of individuals to feel present in a common environment, they obviously must both share a common environment and have a sense of presence in the place depicted. Presumably, the criteria for establishing a common environment are essentially the same as the criteria for defining a common environment in the real world. That this problem is not entirely trivial, however, is indicated by the extent to which individuals with different backgrounds, interests, viewpoints, and sensitivities can be exposed to the same physical environment and yet come away wondering "what world is that other person living in?" At the purely physical level, one could measure the extent to which the various virtual environments (one for each participant) constitute a common virtual environment by examining the extent to which they appear common when examined sequentially by each of the individual participants.
Given that the environment is a common one, the extent to which the participants feel present in this common environment will depend on the same factors that determine presence in individual virtual environments that are not shared by the participants. Thus, for example, the literature has reported high graphics update rate, low latency, wide field of view, and high degree of interactivity as factors that contribute to a high sense of presence (e.g., Barfield and Hendrix, 1995). Because the environment is shared, however, additional factors will be relevant. In particular, it seems likely that the sense of presence in a shared virtual environment will be increased by fostering interactions with the environment in which alterations of the environment caused by actions of one participant are clearly perceived by the other participants. Even more potent might be interactions with the environment in which the environmental changes are not only perceived by many or all of the participants, but also are the result of collaborative work on the environment by the participants. Thus, for example, the sense of presence in the common environment (and thus the sense of togetherness) might be enhanced by rearranging heavy furniture in the virtual environment which requires cooperative lifting by the participants.
Perhaps the most interesting issues in this area concern the use of avatars and, in particular, how each participant relates to his or her avatar. For example, how is the sense of presence in the common environment and the sense of togetherness influenced by the choice of viewpoint? On the one hand, a participant can assume an egocentric viewpoint, see the environment through the eyes of the avatar, and see one's own avatar in the same manner as one sees one's own body in the real world. Alternatively, participants can assume an exocentric viewpoint and view one's own avatar in much the same manner as one views the avatars of the other participants. In what ways do such choices influence the sense of presence in the common environment and the sense of togetherness? Clarification of these issues obviously requires further study.
The sense of togetherness in the common virtual environment will obviously be enhanced by the extent to which rich, multimodal, real-time, intraspecies communication takes place.
The main form of intraspecies communication, speech communication, carries both abstract and emotional information, and is easily made available to the participants of the shared virtual world. Its only problem is that people are so accustomed to this form of virtual togetherness from ordinary telephone usage that its importance often tends to be ignored.
The visual communication that takes place via facial expression and body posture also plays an important role in the real world. Its use in the shared virtual world is more problematic, however. In particular, one must choose between the use of direct, pass-through video of the participants (comparable to the use of microphones to sense the participants' own voices in the speech- communication case) and the use of special sensing systems to extract the relevant visual information about the participants and the use of special algorithms to generate the appropriate expressions and postures for the avatars (comparable to the use of speech analysis and synthesis systems for creating artificial speech). If one chooses the former approach, then one is faced with the problem of appropriately blending the video output with the graphical images. (Apparently, this problem is much greater in the visual domain than in the auditory domain; in the latter domain, the task of blending real sounds with synthetic sounds is relatively easy.)
Perhaps the channel with the greatest potential for enhancing the sense of togetherness in shared virtual worlds is the tactual channel. Our belief that this is in fact the case is based on two factors.
First, it appears that touching and manipulating objects in virtual environments increases the general sense of presence. Although there are few, if any, experiments that have been conducted specifically to test this notion, it is consistent with (a) the well-known importance of the role played by interaction in creating a sense of presence (b) a variety of anecdotal reports on the impact of haptic experiences in virtual environments, and (c) our own personal experiences in this area. In addition, this idea seems reasonable on theoretical grounds. Touch is not, like audition and vision, a "distance sense": in the natural world, one must be very close to an object in order to be able to touch it. Furthermore, whereas technology has provided all of us with substantial amounts of past experience in hearing and seeing events and objects in the real world at supernormal distances via the telephone and television, it is only recently that one has been able to extend one's ability to touch and manipulate objects at supernormal distances via teleoperators. Both of these features suggest that touching and manipulating objects in virtual environments is likely to increase one's sense of presence.
Second, beyond increasing the general sense of presence. The sense of touch obviously plays a unique and important role in human interaction. Touching is not only closely linked to sexual activity and to notions of closeness and intimacy, but, as evidenced in our language, is often used as a metaphor for emotional impact (i.e., "I was really touched by her story"). Furthermore, as evidenced in the research on social touch, touching plays a role, albeit sometimes subliminal, in a much wider variety of social transactions than is ordinarily appreciated. In general, it seems clear that the inclusion of touching in shared virtual environments will strongly increase the sense of togetherness.
Given that this is the case, the question then arises as to which aspects of touch are the most important to include or, from the viewpoint of VR interface design, in what ways will the sense of togetherness in shared virtual environments depend on the type of haptic interface employed? For example, to what extent can the potential increase in the sense of togetherness be realized by haptic interfaces based solely on vibratory or electrocutaneous displays? To what extent can it be realized solely by the use of force feedback without such displays? To what extent are both interface components needed? More generally, what is the most cost-effective haptic interface for increasing the sense of togetherness in shared virtual environments, and how is the answer to this question influenced by the events taking place in the auditory and visual channels?
A preliminary experiment designed to begin addressing some of these questions is described in the above-mentioned companion paper by Ho et al.
Slater, M., & Wilbur, S. (1997) "A Framework for Immersive Virtual Environments (FIVE): Speculations on the Role of Presence in Virtual Environments," PRESENCE 6, 603-616.
Barfield, W., & Hendrix, C. (1995). "The effect of update rate on the sense of presence within virtual environments," Virtual Reality: The Journal of the Virtual Reality Society, 1(1), 3-16.