Architecture

Architects think about spaces for reasons that differ from those of geographers. They are interested in the physical construction of objects within a space, yes, but also in the functional and aesthetic values they can project through such a construction.

In the real world, buildings are designed to fulfill some purpose. This purpose (and the proposed budget) provides the hard parameters within which architects operate. What about cyberspace, though? It’s clearly a space, but then so is the space of possible solutions to a crossword puzzle; architects would not be concerned with the latter, so why should they be concerned with the former?

The answer is that cyberspace in general, and virtual worlds in particular, share with the physical world features important to architects that more abstract spaces do not. In a good introduction ¹ to these issues, Anna Cicognani derives five interrelated characteristics that describe what makes a space interesting to architects:

• Possibility of action. It must be possible to effect changes of state within the space.

• Livability. It must be possible to dwell in the space. This implies that the space can be organized into structures.

• Construction of communities. The space must be social, in that communities should be able to develop by acting on the space while existing external to it.

• Time organization. Only through alterations to space can changes in time be detected, therefore the space must be organized such that these changes can be perceived in a manner that allows them to be mapped onto some consistent (“actual”) time organization.

• Spatial organization. It must be possible to orchestrate the space into structures to allow it to be worked with and studied.

Architecture is about organizing space (or, more specifically, about arranging things into space), therefore of these conditions the fifth is the most important. Virtual worlds satisfy all the criteria.

Some of the same observations about the nature of virtual spaces that were made by geographers Dodge and Kitchin also show up in architect Peter Anders’s book Envisioning Cyberspace ²; this also has large tracts devoted to virtual worlds, and should be on the bookshelf of anyone who takes the design of virtual worlds seriously. Anders advances a thesis that certain types of space are bad for the well-being of the virtual world as a whole (at least if it’s a textual world). Textual worlds use episodic movement rather than dynamic movement—that is, you move a room at a time ³. This contradicts any illusion of conventional “real” space, but is nevertheless so compelling that it forms the basis for all the features that make the virtual world a “world.” The reason it can do this is because players are willing to discard their disbelief, and buy into it. This means they trust the designer to maintain their faith; if the designer pops them out of their immersion, it amounts to nothing less than a betrayal.

Virtual worlds where players can build, transfer this trust to the builders. Unfortunately, many builders do betray it (largely through ignorance or incompetence). As Dieberger also noticed, there will be rooms with non-intuitive connections, “black hole” rooms you can enter but can’t leave, and areas unlinked from the “orientation space” where players enter the virtual world.

Unsuccessful rooms that people never visit will be abandoned; anyone who does enter them will find nothing of interest—not even other characters—thereby contributing to an overall sense that the virtual world is “empty.” Although the central space is generally occupied, the further away from it you are the less likely you are to meet anyone by chance. This robs the virtual world of the spontaneous interaction that gives it much of its immediacy. Everything seems static.

Anders notes one of the problems is that people will usually teleport to locations if they can, rather than walk there. This means they are unlikely to bump into people on the way, of course, but it also suggests they won’t explore. In graphical worlds, you can “see” into adjacent locations, and go over to look at anything that seems interesting; in textual ones, rooms act as their own boundaries, so players are blind with respect to the contents of adjacent rooms ⁴. If people don’t explore, they have less chance of coming across anything (or anyone) new. Exploration can help socialization.

To give players more opportunity to meet people, Anders therefore makes the following recommendations:

• Restrict teleporting.

• Make all private room, link into the general space.

• Have multiple entrance points, for multiple local hangouts.

Although these only officially work for textual worlds with a lot of building, they’re sound advice for all other virtual worlds, too. The fewer anomalies that players must suffer to experience a virtual world, the better; the more opportunities to meet one another the better, too ⁵.

Anders makes a distinction between textual and graphical virtual world design because he sees the former as using cognition rather than perception; information is conveyed in symbols rather than in senses, and space is inferred rather than experienced. One way to interpret this is that for graphical worlds you feel that you are part of the environment, whereas for textual worlds you feel that the environment is part of you. This may partly explain the difference in attraction of the two systems.

Although real-life architecture and virtual world architecture are not the same ⁶, the desire of virtual world designers to mimic aspects of reality (for reasons of immersion) means that many of the processes involved are the same, particularly where there are social concerns. Architects are interested in virtual worlds, and virtual world designers should be interested in architecture.

Social concerns, of course, arise from the presence of large numbers of people in virtual worlds. Where there are people, there will always be other people willing to study those people. When the people share a space—or perceive that they do ⁷—it’s time to call in the anthropologists.