Vol. XV, No. 1
CSA Newsletter Logo
Spring, 2002

Virtual Reality Systems - What Can They Contribute to Archaeology?

Harrison Eiteljorg, II

What is the proper role of Virtual Reality systems in archaeology? That may seem an odd question to ask in this setting, but VR was very prominently featured at the recent Computer Applications and Quantitative Methods in Archaeology (CAA) Conference in Heraklion, Crete. The number of sessions and papers devoted to VR suggests that many in attendance at those meetings believe that VR has an important role to play.1 An article in the New York Times, Sam Lubell, "Virtually Rebuilt, a Ruin Yields Secrets," Thursday, May 2, 2002, p. G6, emphasized the growing use and acceptance of VR.

Returning to the question, "What is the proper role of Virtual Reality systems in archaeology?" I begin with a working definition of Virtual Reality. Some would limit the term to those systems that put the viewer into an immersive environment of some sort and then allow the viewer to navigate through a virtual world, perhaps through time as well as space. The navigation system must provide real- time interaction, meaning that the interaction between user and computer is continuous and instantaneous. Others would expand the definition to include much simpler viewing systems -- as simple as a standard computer monitor -- and to permit interaction between an operator and the virtual world, with the viewer in only a passive role. Although I would prefer a more limited definition, requiring at least that the viewer be in charge of navigation, I think the broadest definition is better for the purposes of this discussion. Therefore, I will take VR to include any system that provides interactive, real-time access to a relatively realistic portrayal of some specific physical realities. The user, in this broad definition, may be either someone trained to operate the system or a naive user, but the definition is not stretched to include prepared sequences that are, in essence, short movies; real-time interaction is a requirement. The level of realism required is, I acknowledge, vaguely defined. I do not think a good, precise definition can be constructed; this is a moving target because the technology is changing so fast. However, the term is meant to indicate realistic textures and coloring, some shadows or shading, though not necessarily cast shadows, and a general sense of realism strong enough to engage viewers. Readers should note that this definition does not include Web access. An assessment of the technology should not be bound by the current limitations of Web-based VR.

Any VR system meeting this definition must maintain the data defining a specific VR world in some format suitable for the task. The more complex and irregular the geometry, 2 the more difficult it is for the software to allow interaction between the user and the system on a real-time basis. As a natural result, complex, irregular geometry is not expected; only the most powerful of computers can handle even moderately complex and irregular geometry. Furthermore, most VR programs were not created to define complex, detailed, precise, irregular models. When such models are required for VR, the models are made in CAD systems and transferred to VR systems -- and are routinely simplified in the process. This is the situation with VR today, though advances in computer technology may certainly change matters.

When dealing with archaeological sites, the problems of complex, irregular geometry are severe. On the other hand, complex geometry can be extremely regular when making reconstructions. For instance, the model of the Colosseum shown in the NY Times article mentioned above is not too complex for VR; the building seems to have been modeled from basic dimensions. Although it also seems somewhat simplified, it is more significant that the model does not include the surviving remains. The irregularity of those remains would make it all but impossible to use the model in a real-time, interactive mode.

Given this dichotomy between models of surviving archaeological material and models of isolated monuments that may be reconstructed from less irregular dimensions (admittedly a dichotomy without a firm dividing line), I should define my use of archaeology in the base question. I am interested here in the presentation of complex archaeological sites. So the question might be restated as "What is the proper role of Virtual Reality systems in the presentation of complex archaeological sites?"

Using this modified definition of the subject matter - complex archaeological sites -- the current state of the technology requires building models with CAD software and then simplifying the geometry for VR programs. Real-time, interactive presentations of the complexity and irregularity encountered in a site, as opposed to a building, is effectively beyond the reach of today's computers. However, if the geometry underlying a VR model has been simplified, an obvious question -- one asked by a member of the audience in one session -- arises: "Can one get from the VR model back to the original archaeological data?" The answer is no. Once the geometry has been simplified in order to permit real-time interaction, there are at least some parts of the original data that have been sacrificed. One might thus compare the creation of a VR model to the creation of a publication; each is a selection from the available evidence, not an attempt to supply access to all the evidence.

VR was not intended to be an appropriate technology for archaeological recording or for access to the full geometry of the models. CAD programs were deigned with those goals in mind, and VR programs were designed to provide interactive, realistic imagery -- generally with the assumption that CAD models could be imported when precise, complex models were required. There is no need to change that symbiotic relationship between CAD and VR programs; the respective roles have developed naturally and make sense. The situation is unlikely to change until advances in the technology permit a true merger of the two program types. 3

If VR is not the scholar's documentary choice, can it be the scholar's publication mechanism? For other scholars? For the general public? For both?

I think anyone who saw the presentations at the CAA meeting -- especially the presentation about Troy -- would agree that VR technology provides a way to present a site that is visually interesting and that can include a great deal of information. (Peter Jablonka gave two presentations of the VR system that had been developed in cooperation with a museum, and the system was available to conference attendees for experimentation in the exhibits room. The first paper was an invited talk, "How Virtual Dreams Can Become Archaeological Reality;" the second a co- authored paper with Steffen Kirchner and Jordi Serangeli, titled "TroiaVR: A Virtual Reality Model of Troy and the Troad.") Nevertheless, we are just beginning to use this technology, and we need many more examples and experiments to develop our use of and expectations for VR presentations.

In general, I believe the most important area for further development revolves around a rather simple issue, one that I have previously written about ("The Compelling Computer Image - a double-edged sword," Internet Archaeology, Summer, 2000, Issue 8 (intarch.ac.uk/journal/issue8/eiteljorg_toc.html) and "Photorealistic Visualizations May Be Too Good," CSA Newsletter, Vol. XI, No. 2, Fall, 1998 http://csanet.org/newsletter/fall98/nlf9804.html) and one that others have also discussed (Paul Miller and Julian Richards, "The good, the bad, and the downright misleading: archaeological adoption of computer visualization," in Computer Applications and Quantitative Methods in Archaeology: 1994, Huggett and Ryan eds., Oxford, 1995): When presenting information, how do we make certain that the viewer/reader understands the difference between what we know and what we think? The problem is not unique to VR but is made more acute by modern computing technology. If what scholars think about a site is the result of an interpretive process starting with what they know, this corollary is suggested: if interpretations always lie between the presentation and the information, how is the viewer to evaluate what is presented -- or is the viewer meant simply to see and believe?

The question of audience now becomes crucial. The CAA presentations, by and large, were about VR systems intended for general audiences, not scholars. As a result, they were not intended to present the evidence and then lead viewers through chains of reasoning to a final conclusion. The systems were intended to present the final conclusions as rather simple, well-constructed, tidy, and indisputable truths. (Since the presentations at CAA were generally short, the full complexity of the VR systems constructed for general audiences could not be shown. Therefore, the foregoing may be an unfair general statement, but it reflects my understanding, based on those presentations, that the intent generally was to present conclusions, not evidence and argument.)

This leads, in its own turn, to two questions. First, can VR be used to present information in a way that maintains the distinction between evidence and interpretation, without concern for the composition of the audience? Second, should the general public be treated differently from scholars and given only the conclusions, not the evidence and argument leading to them?

Can VR be used to present information in a way that maintains the distinction between evidence and interpretation? I think it can, but only if that goal is clearly stated and considered a true requirement -- and if the technology takes a secondary role, with VR being part of a package, not the whole package. Then presentations might including plans of the actual finds and other information to show the real evidence; for example, it is not impossibly difficult to supply the reasoning that leads from a set of real plans or remains to reconstructed building and street plans and then to a fully reconstructed site. Doing that for each building in a complex site may become burdensome, but scholars should accept nothing less from any publication. VR may then be used to show the reconstructed whole, but many other technologies might be used to present the facts and interpretations leading up to the final reconstruction, even -- perish the thought -- expository text. We have no way to be certain that such a multi-threaded publication is possible until one appears, and it does not seem, judging from the examples shown at CAA, that such publications are at the top of many agendas. On the other hand, the preliminary work by Learning Sites, Inc. -- see www.learningsites.com/NVAP/NVAP_text/NVAPhome2.htm -- on publication of The Prehistoric Settlements On Tsoungiza At Ancient Nemea4 suggests that at least one example of VR as only a part of a far more complete and serious publication is being created, though the Web site has not be updated for some time. At CAA VR was treated as a technology for general presentations rather than scholarly ones, and there was a great deal of interest in what the technology can do, considerably less in the prior issues of what information should be presented, in what way, and in what order.

Should the general public be treated differently from scholars and given only the conclusions, not the evidence and argument leading to them? Most of the CAA presentations were about VR systems intended for a general audience. However, only one seemed intentionally to present real complexity and to give viewers evidence and argument instead of a pre-packaged explanation. Francesca Cantone's presentation (in a workshop and therefore not a formal, titled paper) of the new installation at the museum for the Foce del Sele site near Paestum was unique in that sense. The numerous metopes from the site have not been reconciled by scholars into a coherent program; so the exhibit provides various ways for members of the public to view the metopes, connect them by subject matter, and think about how they may have fit into the architecture of the site. This was not a VR system, but it seemed to me to be the only occasion when what was intended for a general audience gave that audience any credit for intelligence. Instead of a single, simplified, pre-digested explanation of a complex, partly understood phenomenon, this exhibit presented serious information and allowed the public to ask serious questions and to engage with the evidence in serious ways.

I hope my bias was obvious in the foregoing. I think the public deserves better, more complex and nuanced presentations than it usually gets. Indeed, I am convinced that many members of the public -- certainly not all, but many -- do not want pat, simple answers. They will be more fully and effectively engaged in presentations of archaeological material -- in museums, at sites, or anywhere -- if they are encouraged to ask their own questions and to look carefully at the evidence as they try to answer those questions. It is very difficult to construct presentations that lead people through real archaeological evidence without boring them; I do not mean to suggest the contrary. However, engaging the public in archaeological questions, archaeological data, and archaeological reasoning will give them a far better understanding of what we really know about the past than will providing pat, relatively simple, pre-packaged stories about the past. A better educated public, in turn, benefits the discipline.

Pondering this issue reminded me of a comment about a lecture I gave several years ago to a local society of the Archaeological Institute of America. The person in charge of that local society reported back to the AIA office that he thought there was far too much detail in the talk, but, he was surprised to report, the audience seemed to like it. I found that reaction to be telling; I did give the audience many details, but I also made it clear to them that we were trying to solve a puzzle and that the details, which might be forgotten the next day, were needed to understand that puzzle. I asked them to think, and they "seemed to like it."

Can VR rise to this challenge, the challenge of helping people to see the evidence, the reasoning, and the results of archaeological scholarship? I think so, but only as a part of a larger package, including many different channels, and informed by a clear, uncompromising vision of the aims and intentions of the effort. Perhaps the Nemea example will become such a presentation. In the meantime, VR provides an exciting way for people to see what the technology can do; it will be much more exciting to see what people can do with that technology when the job is more complicated and more serious -- and when the technology has become what it should always be, a means to an end rather than the end itself.

-- Harrison Eiteljorg, II

To send comments or questions to the author, please see our email contacts page.

1. A new relative of VR, called augmented reality, has been described in an article in the April, 2002, issue of Scientific American, "Augmented Reality: A New Way of Seeing," by Steven K. Feiner; a Web site showing one simple example of augmented reality, the Temple of Hera at Olympia, is at http://www.aec2000.it/archeoguide/. Augmented reality does not attempt to replace real-world objects with virtual ones but to provide additional information -- as images or text. This has the distinct advantage of allowing the viewer to compare reality directly with computer modifications when the additions are images. Return to body of text.

2. The phrase "complex and irregular" is important in this context. Very complex geometry that is regular -- a column that is a vertical truncated cone, for instance -- does not present enormous difficulties for computers. Complex shapes, obviously add some difficulty, but it is the irregularity that causes much more difficulty for computers and their programs. A real Greek column, for instance, leans away from the vertical and is cigar-shaped (bulging in the middle) rather than being a simple truncated cone. As a result, what might be a single surface in the simple case becomes many different surfaces, and ordinary orientations do not apply. Similarly, modeling an ancient wall that consists of many blocks almost, but not quite, in the same plane and doing so block-by-block results in a much more complex and irregular model than does treating the wall as a single plane. Return to body of text.

3.The demand for more realistic VR imagery will, in my opinion, keep this merger of VR and CAD from coming quickly. Until VR can provide real-time imagery that is as realistic as the best renderings, I do not expect there to be developments in other directions. Cast shadows, one of the requirements for more realistic images, will devour additional computing power; so the merger, if it comes, will not be in the near future. Return to body of text.

4.Authorship of the material currently on the Web is not stated, but the following statement appears in the introduction, for which the authors were James C. Wright and Mary K. Dabney. "The Nemea Valley Archaeological Project (1984-1999) is directed by James C. Wright and sponsored by Bryn Mawr College under the auspices of the American School of Classical Studies at Athens. Support for the Project was provided by The National Endowment for the Humanities, The National Geographic Society, The Institute for Aegean Prehistory, the Hetty Goldman Fund of Bryn Mawr College, the Dean of the College of Arts and Sciences, Florida State University, and private donors." Return to body of text.

For other Newsletter articles concerning the use of electronic media in the humanities or electronic publishing, consult the Subject index.

Next Article: The Importance of Standard File Formats

Table of Contents for the Spring, 2002 issue of the CSA Newsletter (Vol. XV, no. 1)

Master Index Table of Contents for all CSA Newsletter issues on the Web

CSA Home Page