One solution is procedural generation, where a human designer lays down a few rules–a basic city plan, for example–and a computer fills in the rest according to those rules. The result is a complete rendering of a city filled with buildings that plausibly inhabit the space, with a human only having to set up the initial parameters. Consider this reconstruction of Pompeii:

The creators of this video started with street plans and a variety of historically correct architectural models. A computer then generated buildings that fit the excavated ruins, resulting in a city that you can tour virtually. While it undoubtedly has inaccuracies, the result is compelling.

Pompeii is better-preserved than most ancient cities, but you can apply a similar technique to any archaeological site. Archaeologists have partially excavated many biblical cities; they know at least some of the city’s layout. Even if they don’t know the whole thing, they can guess at what features a city of a given size needs; by starting with what archaeologists know, a computer can extrapolate a plausible street plan for the rest of the city. (I suppose that you could run the simulation many times and generate a probability of where a certain building–such as a synagogue–is likely to be.)

These projects don’t often generate interior spaces or simulate objects like furniture, both of which dramatically increase the complexity of the simulation for only a modest benefit. But there’s no reason why we couldn’t model interior spaces. A forthcoming game called Subversion, for example, uses procedural generation on both macro and micro scales: it generates both complete cityscapes and architectural floorplans of the buildings that it creates.

Recreating interiors for ancient houses is fairly straightforward: floorplans weren’t nearly as complicated as they are today. Imagine walking around ancient Capernaum, for example, and visiting the house where people lowered a paralytic through the roof. Architecture plays a crucial role in the story, a role that a virtual-reality model would help illuminate.

Further Reading

1. Procedural, Inc. creates software for procedurally generating cities, both modern and ancient.
2. Rome Reborn from the University of Virginia recreates ancient Rome using a combination of hand-modeled buildings (for thirty models and 250 elements) and procedurally generated buildings (for the remaining 6,750 buildings). Academic papers provide more technical detail, especially the one by Dylla, Kimberly, Frischer, et al. (PDF). They use the Procedural, Inc. software.
3. A Subversion video shows the steps a computer goes through to generate a cityscape.
4. Procedural 3D Reconstruction of Puuc Buildings in Xkipché demonstrates an academic application of the technology applied to archaeology.
5. Magnasanti talks about the “ultimate” SimCity city and was the inspiration for this post.

[T]he game is set in Florence, Venice and Rome over a number of decades leading up to the year 1499. The game’s producer-authors… labored lovingly to re-create the environs as exactly as possible. They hired Renaissance scholars to advise on period garb, architecture, urban planning, weaponry and the like. They took tens of thousands of photographs of interiors and streets. They used Google Earth liberally to piece together the ground-up and sky-down perspectives through which the action flows…. The hazy colors and the distant sound of river birds are uncannily correct. Nowadays, the tourist hordes can blot out all sense of history. Once you’ve navigated it on AC2, when you visit the Ponte Vecchio in person the illusion persists of a highly intensified sense of place. In other words, the video brings the place sharply back to life.

Recreating history comes at a price: the budget for the game is something north of $20 million. I hope that the publishers will find a way to put some of their investment to educational use; I for one would love to visit Renaissance Italy without having to assassinate people once I get there.

Someday I hope to see a recreation of ancient Jerusalem this detailed, though I can’t imagine what kind of game could justify the pricetag. In the future, maybe the cost of creating virtual time travel will drop far enough to be within reach of small schools, companies, or individuals.

(Note: I haven’t played the game and don’t intend to. As you might guess from the title, it appears to involve lots of killing. If you’re OK with seeing that kind of thing, on YouTube one of the developers walks through some of the gameplay.)