The key artistic question regarding the medium of interactive virtual worlds is what is characteristic and unique about it's expressive language. Most important is the volitional and interactive use of the camera available to the viewer. This willful navigation of the world in space, and in time (by multiple visits), makes possible the use of a non-linear experiential topology for the mapping of the development of the piece.
The implications for free navigation are also that these movements and views may instigate events which advance the evolution of the state of the world's appearance, sound, behavior and narrative. Instead of a linear sequence of movement and montage, the world can now be seen as a set of constraints and fields of force which influence the free movement and perception of the identity-camera. In addition, these constraints and fields influence the behavior of environments objects entities and characters, and also the development and unfolding of narrative. Two key aspects of the presence of navigational Will in a non-deterministic world are the ability to trigger events by 'proximity' to the location in space-time of the camera, and by the 'visibility' of aspects of the world relative to the camera.
The concept of spatial navigational freedom is fairly easy to grasp, but it is important to remember that free navigation implies the re-visiting of scenes and experiences, even in different contexts and sessions. Thus navigational freedom is not simply spatial but temporal as well. In addition, two viewers moving through the same world, even at the same time, may view different things and events. The time-space-knowledge continuum is so complex that for the first time what is suggested is the everyday conditions under which we routinely live. However, it is important to note that there is no need to create a verisimilitude to this idea of the 'everyday'. In particular, physics need not be obeyed, nor causality, nor rationality. We have before us a blank canvas 'life', whatever that may mean. What is 'real' is precisely what is present in the volitional 'experience' of your instance of 'the world.'
One aspect of identity which seems to differ from our experience, but which is within the scope of variation of the medium, is the shift in viewpoint from omniscient to subjective, and all points in between. Most dramatically this is reflected in the camera view. If the view is omniscient the camera makes accessible to the viewer all experience and information relevant to the comprehension of the narrative or sensory development. As the point of view becomes subjective the camera might begin to take a 3rd person viewpoint in which the geometric 'body' of the viewer is in camera. Finally, when the point of view becomes entirely subjective the viewpoint moves within the geometric body which is then perhaps invisible to the personal view entirely, as if the identity of the viewer were a disembodied floating subjective 'singularity.'
This brings us to the notion of 'montage' in virtual worlds. It is too complex to fully discuss here, and still not fully understood in any case. However, it is useful to point out that rendering 'switches' provide traditional 'cuts', and camera navigation provides the traditional camera pan tilt track and zoom. Additional new techniques consist of 'scale change' which provides a means to perform a 'continuous cut' by context and size reference, and 'change of identity' which provides a means to smoothly adjust the personal context of perception within the same 'objective' scene. What is awaited is the 'Eisenstein' of interactive virtual worlds to create a fairly complete taxonomy of non-linear montage which defines the medium according to a 'grammar' of viewpoint change in order to suggest narrative development in space and time and identity.
The artistic aspect of interactive virtual worlds which is bound most closely to technical aspects of the medium is the seamless surface of various media types unified by a simulated 3D spatial and temporal experience. This 'sensory surface' consists of sound, music, speech, images, vector graphics, text, video, geometry, and lighting. In particular, since the lights and cameras are virtual and do not exist in time and space it is easy to create lighting and camera angles and positions which would be impossible on real sets.
Finally, an interesting aspect of the medium which may possibly be used is a multi-user 'shared' presence. This is the case that two or more navigating identities may move and communicate within the context of the same 'illusion' of a shared world, much like we do in 'reality'.