From Uncertainty to Virtual Reality: Knowledge Representation in Rome Reborn

Abstract

Abstract: Graphic representations of ancient Rome have become more visually powerful in the late twentieth and early twenty-first centuries with the innovations afforded by digital technologies, but the use value of these images is under debate today. This paper explores the interplay among different types of knowledge representation, an under-theorized area of research in the digital humanities, in the acclaimed Rome Reborn project, now also known as Ancient Rome 3D in Google Earth. Rome Reborn is perhaps the largest and most complex visualization endeavor in the digital humanities to date. The author of this paper belonged to the original project team (UCLA 1999-2001) and is on the Scientific Committee of the current iteration (UVA). Rome Reborn incorporates distinct classes of knowledge—historical sources, archaeological remains, and deductive logic or inference—as a basis to reconstruct the appearance of ancient Rome’s monuments (mainly temples, public buildings and residential structures), urban infrastructure (streets, aqueducts), and topography (hills of Rome, Tiber River). All forms of knowledge utilized in the making of Rome Reborn are represented by the medium of an interactive virtual reality model consisting of millions of polygonal surfaces with applied colors, textures and simulations of light and shadow effects. This paper will perform autopsy on Rome Reborn and expose its interwoven visual representations of historical, archaeological, and conjectural knowledge. The relationships of secure knowledge representations, which are sparse in the model, to the more prevalent conjectural or speculative knowledge representations will be clarified with the aim of identifying Rome Reborn’s underlying epistemological structure. Analysis of Rome Reborn in this manner holds the potential to advance the methodological discourse in the digital humanities for the visual representation of knowledge when multiple forms of knowledge require systemization and when levels of uncertainty are high.