Following in Rabelais' Footsteps: Immersive History and the 3D Virtual Buildings Project

John Bonnett

Research Officer

Institute for Information Technology

National Research Council of Canada

46 Dineen Drive

Fredericton, NB

E3B-9W4

In Europe and North America, scholars are only beginning to shed the assumptions of print culture. They are also only beginning to perceive the possibilities of a computerized, electro-magnetic culture. In this first decade of the 21st century, historians face a challenge: how, and to what extent, are we to appropriate the 3D-immersive environment as an instrument of representation and interaction? In many ways, our situation is analogous to that faced by scholars at the close of the Roman Empire. Virtual Reality (VR) — the generation of artificial, three-dimensional, immersive environments by the computer — is the codex of our time. As a medium of communication, it presents new methods of representation, and new forms of narration for historians. As a technology of representation, it affords the opportunity to generate new conventions, for communication, and for documentation. [1]

But in making these statements, I am well aware that I am making the kind of sweeping generalizations that can — and should — provoke complaint from colleagues. After all, who is driving whom here? Why should historians automatically appropriate a technology? Are champions of the computer and VR not evading important questions as to who or what is driving their scholarship? Is technology driving the research agenda, or is it being made to serve a research agenda? Are historians who use IT tying their research and teaching to established, enduring concerns that derive from the philosophy and politics of their discipline? Or, are they letting technology dictate their actions?

In this study, I argue that 3D-immersive environments — Virtual Reality — should not be construed as a thoughtless import from popular culture. On the contrary, I suggest that VR, properly applied, has the potential to heighten the effectiveness of historians, as teachers, as communicators, and as researchers. Specifically, I suggest VR can be used to heighten the critical thinking skills of students. I further suggest that VR will soon enable environmental, urban and cultural historians to produce models of far greater range and sophistication than is possible in print. To support this argument, I present the 3D Virtual Buildings Project as a framework for discussing what is currently possible, and as a prototype for discussing what will soon be possible.

Part One describes the aims and history of the project, and in particular the circumstances in Canada that led to its founding. In Part Two, I provide a description of the project's teaching method. Stated simply, our aim is to help students realize an important concept about historical representations, namely that they are models, models that are imperfect representations of the objects they purport to represent. We believe that students best learn this insight by constructing a historical artifact for themselves. We also believe that 3D objects provide participants with a more efficient way to gain this understanding, when they are used in place of text and number as the raw material for models.

Finally, in Part III, I describe the project's website, which was generated to display multiple historic urban environments. My purpose in this section is two-fold. The first is to demonstrate that historians, be they urban, cultural, or environmental researchers, can produce representations of far greater range than is possible in print. I also present this section to demonstrate a second point, that resorting to virtual environments will demand that historians change their practice, particularly with respect to composition. Some conventions will have to give way, such as single authorship, and the conception of historical models as essentially fixed entitities, immune to revision except under extraordinary circumstances. When we speak of virtual environments, we speak of a potential to represent domains of the past that far exceed print. As Janet Murray and Tom Taylor have noted, virtual reality offers the possibility to generate narratives with tactile, olfactory and auditory information. [2] It further affords the possibility of populating virtual urban systems with avatars, virtual representations of people, that move and interact in an autonomous fashion. It finally suggests the possibility of inscribing laws of physics to govern interaction and movement, similar to laws governing the real world. [3]

My point in outlining these possibilities is to suggest that such complexity in historical representation will demand multiple authorship, in which contributors up-load their contribution to a web-site for peer review, and then posting. It is also to suggest that such representations will be dynamic, susceptible to constant revision. Historians will have to embrace the ethic of open-source. It is finally to suggest that scholars will have to give thought, hard thought, to how information is structured in such an environment. How do we fulfill one of our most basic functions in such an environment, namely the telling of a story? How much constraint do we apply upon the user, thereby affecting his or her ability to navigate and interact with the virtual environment? How much liberty do we provide? In Part III I present the 3D Virtual Buildings Project web-site as an early prototype of how such a system might be constructed, and as a framework for discussing what might be accomplished on future web-sites.

I

The 3D Virtual Buildings Project was founded in 1998, and is the product of a partnership between the National Research Council of Canada, and Industry Canada, the Canadian government's federal department charged with overseeing industry. One purpose of the partnership was to provide graduate students — such as I was at the time — with the opportunity to generate teaching materials that use information technology, and can be disseminated over the Internet. [4] Over its four-year history, the project has also partnered with local educational institutions to assist in the generation and testing of the project's teaching materials, including the University of Ottawa's Institute of Canadian Studies, and three Ottawa-area high schools. In the fall of 2000, the project's initial offering of teaching materials, and its web-site, were placed on the Internet for public viewing and use, and can be accessed either through Digital Collections, the Government of Canada's portal for educational content, or directly via the project's web-site. [5]

The project's aims are three-fold. The first is to provide participants with the skills necessary to generate historic replicas of Canadian cities using 3D computer modeling software. Our teaching materials provide a step-by-step method that enables students to translate information from historical sources into three-dimensional models. We also expose our students to the historical sources they will need to construct their models, and the repositories in which they will find them. Our second aim is to establish a program that has the potential to become national — if not international — in scope, one in which students in Calgary for example can re-create the Calgary of 1905, while participants in Nova Scotia can generate a replica of the Fortress at Louisbourg. Finally, the project's purpose is also to provide a repository for high school and university participants to display their work, a mandate that was fulfilled in 2000, and a topic that I will discuss in greater detail in Part III.

The project was founded, however, not only to generate infrastructure, but also as an attempt to meet identified educational needs. For the past five years or so, a good deal of comment has been generated in Canada's media over the state of historical instruction in Canada, concerns that has been echoed by academics, government policy makers, and members of the business community. While in Canada, as in the U.S., there has been and continues to be controversy over the content of history curricula, there is widespread agreement that Canadians, particularly young Canadians, do not know enough about their collective past, and that new, more engaging, methods need to be devised in order to heighten student interest in history. [6]

Our response to this challenge has been two-fold. One solution, we believe, to heightening student interest in Canada's past is to widen the domain of history from political and social history to include cultural history. Here, we take our cue from the novelist Margaret Atwood, who noted in her 1996 Bronfman Lecture to the University of Ottawa that Canada suffers from a kind of collective amnesia. Few know that Canada enjoyed a vigorous cultural life before the Depression. Most believe that Canada was little more than an outpost of Empire during the 19th century — and the first part of the 20th century — devoted to little more than resource extraction. The stereotype of Canadians as hewers of wood and drawers of water holds such a powerful hold on the collective imagination that it occurs to few to challenge it. [7] They should, and we argue that one way this can be done is to expose Canadian students to the venue in which much of that activity took place, namely the city.

Our second solution is to make students participants in the generation of historical models, textual and three-dimensional. With respect to textual representations, we have produced tutorials to enable students to conduct primary and secondary research. Our program is not concerned with specific content. What we propose instead is that students visit archives and other repositories to engage in primary research, and that they use the building they designate to structure their interaction with the past. For example, if students are interested in the specific history of a structure, they may want to learn more about the technology and architectural conventions that supported its construction. By contrast, students interested in a building that contained a bookstore might want to explore what Canadians were reading during the 19th century. From our standpoint, participating students have leave to explore any dimension of Canada's past, no matter how prosaic, no matter how profound. All we ask is that they use the structure as the organizing principle of their research, be that principle the specific history of the building, the array of activities associated with the building, or significant activities that occurred in proximity to the building. Such constraints provide room to explore everything from the history of science to the history of sport.

Our students are also encouraged to produce historical representations using threedimensional objects. Here, our aim is not only to encourage student interest in Canadian history, but also to meet a second identified need: the development of critical thinking skills. Our object is to persuade students that historical models are imperfect representations of the past, mediated by historians, and mediated by historical evidence. In Part II, I describe our rationale for adopting information visualization and 3D modeling to induce students to take a more sophisticated view of the strengths and weaknesses of primary sources. I also describe the primary sources students use to generate models of urban environments, and the specific lessons our students learn about the process of historical re-construction as they produce a 3D model.

II

The development of critical thinking skills is an aim that is shared by Canadian, American and international historians. Indeed, at the university level it is a central if not the central benchmark of success in teaching. If our students leave after four years still oblivious to the limitations inherent in all forms of historical representation, then we have failed them and failed ourselves. The key question I would like to pose in this section is whether historians performing this task perform it in the most efficient way. I would suggest that for the most part the correct answer is no.

Consider how most historians make the attempt in the university classroom. They generally employ two strategies, strategies I will refer to as the rhetoric of instruction, and the rhetoric of construction. In the rhetoric of instruction, the professor proclaims the need for new thinking — that a distinction must be made between the actual past and mediated representations of the past —while students are expected to incorporate the new dispensation, and move on. The method is not very effective. At best, students emerge with a shallow understanding of the defects of their previous approach to history, and the merits of the new.

In response to this problem, historians are increasingly turning to a rhetoric of construction. There is nothing particularly surprising about this development. Instructors in a variety of disciplines have turned to constructionism, the teaching philosophy pioneered by Jean Piaget, Seymour Papert, and others. [8] It holds to the view that students do their best thinking in exercises where they are required to model what they know. It is worth asking, however, if historians are employing the rhetoric of construction in the most efficient way. As the communication theorist Harold Innis noted long ago, not all sign systems are equal in their capacity to support higher-order thinking skills. Nor are they equally efficient in representing patterns of economic, cultural or conceptual interest to their users. In the wake of the personal computer, and the development of affordable software that expands the repertoire of historians to include 3D objects, the questions arises: are historians providing students with the best sign system to support the task at hand?

Considering the difficulties students have in retaining, manipulating and analyzing information derived from series of primary sources, I would suggest the answer is no. The magnitude of the problem I think is best perceived if we consider that it can take years for even senior scholars to identify patterns of significance in their documents. And yet we often afford our students as little as two to four weeks to do the same. The problem is further compounded by the referents we ask our students to either perceive or construct. For the most part, they are quite abstract. What real meaning does the decision to use the atomic bomb have to a 19 year old? How is a student to perceive or visualize the formation of the English working class? The task of inducing critical thinking skills in students with the rhetoric of construction is not impossible, but students do not have an easy time abstracting an identifiable historical process from the large quantity of information at hand.

We believe there is a better way to apply this rhetoric. The first innovation we use in our teaching materials is to use the forms of information shown in Figure 1, namely documents that are visual, as opposed to textual. Our second innovation is to provide a concrete referent, namely a historic building that students can see. With respect to primary sources, the benefit of this approach is it enables us to limit the documents to a quantity of three, specifically two photographs, and one fire insurance, or Sanborn, map. [9] With respect to the referent, its benefit is it enables students to more easily understand the relationship between historical representations and the objects they purport to represent. Models are more easily compared with source material, and when possible, the actual building.

Figure 1

Using these materials, we take students through the process of translating the information contained therein into data they can use to construct the structure shown in Figure 2, namely the building of James Hope, a 19th century Ottawa stationer whose business was located one block south of Canada's parliament.

Figure 2

With maps, this process, as shown in Figure 3, is done through direct observation of data (Image A), measurement with a fine-grade ruler to determine map scale and building size (Image B), and juxtaposition with one or more additional sources (Image C).

Figure 3

With photographs, participants isolate a façade of interest, such as the one indicated in Figure 4, and distort it with photo editing software such as Adobe Photoshop so that the wall is aligned onto a flat plane, one that is perpendicular to the line of sight of the modeler, instead of one veering off an angle into a vanishing point. Once the photograph is processed, participants derive a scale for the x and y axes of the photograph, and use a fine-grade ruler and simple math to determine the location of windows, doors, pilasters and other objects associated with the façade of interest.

Figure 4

Once all the information in the source material is converted into useful numeric data, participants then go through the process of constructing the series of 3D objects shown in Figure 5, in order to generate a representation of the Hope Building.

Figure 5

During the process of translation, the project is able to visually demonstrate four specific characteristics of primary sources, characteristics that reveal the challenges historians face while attempting to construct a viable representation of the past. The first lesson students learn in our tutorial is that evidence is subject to misinterpretation. A naïve reader of the map shown in Figure 6, for example, would conclude that the indicated structure — the Merchant's Bank — was 50 feet high. In fact, the cartographers followed a different convention, indicating only the height of the total number of full stories within each structure. If a sloped roof covered a building's top floor, it was defined as a half story, and its contribution to building elevation was omitted. Hence, the 50-foot mark of the Merchant's Bank was actually located at the point indicated in Figure 6. By coming to grips with this feature of Fire Insurance or Sanborn maps, students learn that their initial interpretation of a document may not always match the intent of the original author.

Figure 6

The second lesson our students learn is that evidence is incomplete. Stated more specifically, an item of evidence does not always answer the question a historian would ask of it. In the case of our tutorial, our students run into this problem when they attempt to determine the height of the Hope building. As indicated in Figure 6, the convention followed by the cartographers was to indicate building height in the center of each structure. As Figure 7 indicates, however, such information is not forthcoming in the cartographer's representation of the Hope building. Our students are forced to determine the height of their structure via an alternate method.

Figure 7

Our third lesson is indicated in Figure 8. Sometimes historians face not only incomplete evidence, but scenarios in which there is an absence of evidence. How are our nascent historians to deal with situations in which they are unable to document the appearance of a section of a building?

Figure 8

Here, we emphasize to students that historians face gaps in evidence all the time. We further emphasize that they nevertheless go on to produce fully-formed representations of the past. They do so by applying an accepted solution, namely informing their audience of the problem, and making an informed guess as to the probable content of the gap, based on a reading of the historical context of the time. To reinforce this point, we take students through an exercise in which we pose a simple question: did the unknown façade have windows? We then take them through two lines of reasoning in which the answer is alternatively yes or no, and in which the justification for each answer is based on references to the designs of neighboring structures. After that, we take them through a second round of argument, one in which each position is subjected to criticism. At the end of the process, we press our students to accept a simple point. Namely, that as long as each hypothesis remains within the architectural conventions of the time, we have no definitive basis to label one answer as inherently right, and the other wrong. Students learn the need to live with uncertainty.

Our fourth lesson is that more information can be garnered when an item of evidence is interpolated with other source material. This is the method we use in our tutorial to determine the height of the Hope building. Note in Figure 9 that the map to the left indicates the absolute height of the Hope building's neighboring structures. On the right, the photograph indicates the relative height of all the buildings of interest, including the Hope building.

Figure 9

By placing the two documents together, we are able to ascribe an elevation to the 38' point indicated in Figure 9. From there, as Figure 10 indicates, the process of determining the structure's height resolves down to a simple process of photo distortion and editing, measurement, and math to determine the height of the roof. Using a ruler, we can determine the height of Section A relative to that of the wall by measuring their images on the photograph. Measurement of the endpoints indicated by the arrows in the left photograph of Figure 10 enable us to generate a fraction, one we can multiply by 38' to determine the height of Section A. Reference to a second photograph, indicated by the middle photograph in Figure 10, allows us to repeat the process, this time to determine the height of the roof. Making a second set of ruler measurements, we can produce a second fraction (Height "?"/ Height A). From here on in, it's a case of simple math. The second fraction multiplied by our figure for Height A leads to a final figure for roof elevation. Adding the resulting figure to 38', we were able to determine that the Hope building was approximately 46.82' in height.

Figure 10

In 1999 and 2000, the project was able to test its materials in three Ottawa area high schools, tests that produced positive, but limited, results. Some students, such as Ellora Faruq, in 1999 a Grade 12 student at Sir Wilfrid Laurier High School, did demonstrate an understanding of the core concepts that we were attempting to convey, that historical representations are mediated, and that different historians reach different conclusions on why events occur. In a December 1999 debriefing, which can be viewed on-line (See Figure 11), Ellora recounted that: "I was never approached to do something like this before."

Figure 11

Sadly, comments like Ellora's were not the rule during our two trials. The 3D software we employed — Nemetschek's VectorWorks — proved to be a significant hurdle for most of our students, one that hindered their realization of the objectives we set for them. VectorWorks is a complex CAD (computer aided drawing) modeling package, one that takes two to four weeks of sustained effort to master prior to beginning an independent project. For many of our students, it was not a price they were willing to pay, since they desired a more immediate return for their effort. For participating history instructors, the software was also a source of concern, the cost in time learning the software outweighed the value gained from the exercise. For the present we have an instruction program that is better suited for a fourth year university seminar than a high school class.

In retrospect, it is not surprising that we ran into the problems we did. Researchers in a variety of disciplines — from ethnography to cognitive psychology — are pointing to growing bodies of evidence that suggest 3D immersive environments can enhance creativity and transform the way we see the world. The research also suggests that the difficulty in generating 3D immersive environments — due to the complexity of gaming engines, high end modeling software, and the life — is hindering their widespread exploitation. [11] In short, there are grounds to suggest that the conception of our program is sound. But there are also grounds to suggest that our implementation needs work, and further research. It is this challenge that will govern the future research of the 3D Virtual Buildings Project, and many of my colleagues at the Institute for Information Technology. In the short-term, the project will be exploring the utility of photogrammetric software packages, instruments that translate 2D models into 3D models, and have a shorter learning-curve.

Over the long-term, two other, more enduring, solutions are possible. The first will be the emergence of modeling packages that rely on different, multi-modal, interface paradigms, including voice. The second solution, hopefully, will be a cultural reappraisal of the value of 3D objects as instruments of expression. Consider the distinction most jurisdictions in North America and Europe presently make with respect to the signs at their disposal. Generally, state and provincial authorities invest some 16 to 18 years worth of time and treasure to provide citizens with at least a minimal competence using the technologies of text and number. By contrast, in the public schools in which I operated, an interval of a week was considered a reasonable investment for learning 3D modeling software. Surely there is an argument to be made that civil society and government should provide students with more time to learn the techniques of 3D modeling in particular, and information visualization in general. But such a transformation will not occur until information scientists have developed tools that allow for easier generation of virtual environments. It will also not occur until scholars demonstrate that 3D environments can serve as useful complements to traditional methods of teaching, complements that can enhance a student's capacity to profit from the culture of print, and complements that can enhance a student's capacity to interact with the outside world.

III

We live in interesting times. It is not only our signs of representation that are changing, but also our genres of representation. There is a trajectory guiding the evolution of 21st century forms of story-telling. While its outcome is not inevitable, emerging technologies such as broadband communication, multi-player gaming platforms, and Augmented Reality all suggest that the constraints governing story-telling are changing, and that, in turn, the forms will as well. [12] Indeed, this is already happening. Old forms are being complemented — and at times supplanted — by the new. And at times, they are also being collapsed with the new. The MagicBook is one example where the technologies of the codex and the computer are being merged to form a hybrid, one that simultaneously provides print and virtual reality representations of a story. [13] My primary concern here, however, is a more dominant trend, one which suggests that stories will increasingly be located in three-dimensional, immersive environments, environments in which multiple agents interact, and participate in events that have multiple possible outcomes. This is the stuff of non-linear narration. This is the stuff of computer gaming. The author is not God here, and he or she does not determine the final outcome of the story. That is left to agent interaction, and chance. [14]

Such tumult, of course, can be cast in either a positive or negative light. Indeed, in ancient China the invocation that one's fellow live in an interesting era was construed as something of a curse. But in the present, I would argue that historians should construe the transformation of storytelling simultaneously as a necessary imposition and as an opportunity. It is certainly an imposition. Scholars in our discipline certainly did not drive it. But I believe there are grounds to suggest that they can embrace it, and more to the point that they should embrace it. We do not live in a cultural vacuum. There is no merit in pretending that scholarly practice is — or should be — immune from the ever-shifting patterns of cultural, especially popular cultural, change. Certainly, our mandate is to influence society, critique it, argue with it, and even, at times, transform it. But there is also a mandate to attend to society. There is no merit in becoming a 21st century scholastic, ever insistent on the use of classical latin while the rest of the world has read Cervantes and Dante and moved on to the vernacular. If there is a reason to incorporate 3D environments into the historian's craft, it is in part because the outside world in near and far future will increasingly expect us to mediate our expression within such environments.

But the transformation of storytelling can also be construed as an opportunity, one that affords cultural, urban and environmental historians the chance to re-explore traditional questions, and to represent them in new ways. As the communication theorist Harold Innis noted in his writings on the Oral Tradition, the construct of a story is a heuristic exercise. Our conception of the whole to which a story refers is shaped by the process of assembling its constituent parts. Connections emerge that previously escaped our attention. Conceptions are clarified as they are subjected to the rigors of formalization. Innis argued that a change in the method of narrative construction in turn could transform the author's perception of the object of representation. In his various narratives, this is precisely what happened to the ancient Greeks. A shift in poetic conventions encouraged a subsequent shift in Greek thinking about nature, shifting much of the burden of explanation from religion to science. Natural phenomena were increasingly construed in scientific terms. [15]

Here, I would like to suggest that the computer game and 3D-immersive environments are our new meters of poetry. Their capacity to dynamically display text, number and 3D object, and their potential to transmit new forms of information, be it tactile, olfactory or sonic forms of information, suggest that the two will alter our perception of objects of interest. Indeed, the computer's capacity to represent the transformation of patterns over time has already influenced the thinking of scientists and social scientists, and there is every reason to suggest this trend will continue as 3D virtual environments assume greater prominence as interface paradigms and objects for representation on the computer.

In the field of meteorology, Edward Lorenz's modeling of weather systems on the computer led to their subsequent characterization as rule-based systems subject to contingent circumstance. Weather is now seen as an emergent system, one whose ultimate trajectory is determined by its initiating conditions. [16] In the nascent field of artificial life, the behavior of large populations is the object of attention, and its practitioners have used models to produce intriguing explanations for a wide array of social phenomena, from segregated neighborhoods and ethnic genocides, to the disappearance of the Anasazi Indians in New Mexico. [17]

The innovations in storytelling present important implications for historians concerned with the history of spatial environments, and the cultural construction of spatial environments. With respect to the latter group of historians, there is opportunity to build on Tom Taylor's insight that complementary forms of representation can heighten one's insight into the strengths and weaknesses of a text, namely the features of the past that it describes well, and those features it omits. [18] Taylor's work suggests that historians' generation of 3D environments has the capacity not only to heighten their capacity to interrogate texts, but also that of their audience. Consider a hypothetical case in which a historian is exploring competing historical constructions of a particular region of a city, ones differentiated by the class identity of the observer. 3D environments would enable historians to display, and if needs be highlight, the objects or parts of a contested region that figured large in each construction. Conversely, such environments could be used to simultaneously display objects at the margins of each competing construction, thereby facilitating their interrogation, and subsequent problematization or reinforcement. As instruments of representation and instruction, three-dimensional environments have much to offer historians concerned with the discursive construction of space.

With respect to historians of spatial environments, the advent of high-speed Internet research networks, such as CA*Net 4 and Internet 2, offers historians the opportunity to generate information-rich environments. With network lines capable of transmitting gigabytes of information, scholars will have the potential to create models of urban and other spaces comprised of highly detailed buildings, and the objects, profound and prosaic, and people populating them. In turn, scholars will need to consider how a user's access to these environments should be structured. While I believe that historical work in future will appropriate forms and conventions from popular culture — including those relating to computer gaming — I am not advocating their thoughtless incorporation, nor do I believe that it will occur. Instead, historians in near and far future will be faced with an ever-growing repertoire of technologies and narrative forms with which to engage, analyze and represent the past. They will in turn be faced with the need to identify appropriate narrative forms and technologies with specific tasks, an effort that I believe will become increasingly systematized, to the point of forming a new domain of research within the discipline. At times, historians will determine that there is little need to impose restraints on a user's interaction with a virtual environment. They will present users with an environment allowing free navigation and exploration. In other cases, historians will opt to govern a user's trajectory through the environment, by imposing a linear or non-linear narrative structure over a user's visit to virtual space. And finally, sometimes, perhaps a majority of times, historians will opt to avoid 3D environments altogether, choosing to fulfill their objective using the traditional technologies of text and print.

Aside from employing narrative structures to mediate a user's interaction with the environment, historians will also need to consider how their audience interacts with specific objects in the environment. When I speak of information-rich environments, I speak of environments in which the objects provide a pathway to the specific history of the object, the historical context that gave rise to the object, or significant event that occurred in proximity to the object. Hence, if I were to present users with a model of the music store shown in Figure 12, which was located in 19th century Ottawa on Sparks St., it could serve as a portal, one providing information on the specific sheet music displayed, as well as more general histories describing the types of music that Canadians were playing and listening to in the late 19th century.

Figure 12

Finally, the same music store could serve as a point of reference to a significant political event that occurred on Sparks Street, namely the assassination of D'Arcy McGee, a fiery orator, and one of the founders of the Canadian state.

Historians will need to consider how such a vision of interaction and information retrieval is to be realized. In the short term, the process will likely neither be complicated, nor difficult to realize. Users will click objects of interest with their mouse, an interaction that will activate a pop-up window, one containing an array of primary and secondary sources for the viewer to consult. [19] The long-term, by contrast, will offer historians less prosaic means to structure and display their content. Consider the *.bot. A product of the field of artificial intelligence, *.bots act as agents inside the computer. They are designed to assist users seeking to retrieve information, in part by mimicking the human capacity to engage in the give and take of conversation. [20] While current prototypes are nowhere near realizing the benchmarks that AI researchers have set for them, consider what historians might be able to accomplish if they do. A user could approach an avatar, a representation of a historic actor, and begin querying the agent regarding, say, the clothing that it is wearing. The agent would be charged with either answering the question directly — in a fashion analogous to an actor at a heritage site — or conducting a search, to locate the avatars or repositories in the information space containing the requested information. The avatar's final responsibility would be to escort the user to his or her selected information repository.

To say the least, the vision of historical representation I am describing here would be labor-intensive. Consider the Ottawa building shown in Figure 13, 96-98 Sparks St., the locale for dry goods and retail stores during the 1880s.

Figure 13

This one building alone could justifiably be linked to a wide array of literatures, including the history of material culture, through the dry goods and textiles situated in front and inside of the building, and the history of technology, through reference to the horse carriages situated in the street in front. The scale of the task in turn imposes an important challenge for historians: how might such information-rich environments be realized?

Here, I would suggest that historians would do well to follow the example of François Rabelais. In Mikhail Bakhtin's account of the French novelist's career, Rabelais appropriated the methods and narrative forms of popular culture to produce his masterwork, Gargantua and Pantagruel. I have already argued that historians should look to popular culture as a source of inspiration, but Bakhtin's Rabelais and His World is useful in explaining why. In Rabelais' time, there was a far more fundamental task than finding an effective means to communicate with one's contemporaries. The more important object was to locate the cultural ecology in Europe most likely to generate the newest — and widest array — of methods and genres for communication. With them, one could experiment, and find new ways to communicate with one's contemporaries. Without them, one could not. Rabelais' genius in Bakhtin's mind was not to look to Europe's high culture for cultural renewal, but rather to a domain with a large number of interacting agents, which Bakhtin alternately refers to as carnival or the market square. Stated simply, Rabelais looked to popular culture. The insight in retrospect was quite simple. The more agents you have, the more possibilities there are for the populace to locate modes of expression that are adaptive, that give expression to the political, environmental and cultural present, and in turn resonate with the populace, to the point that the given genre or method is selected and used by its members. Rabelais' genius, in turn, was to identify these, or the "cries of Paris and the cries of quacks and druggists," as Bakhtin more colorfully refers to them, and incorporate them into Gargantua and Pantagruel. [21]

The insight by Rabelais and Bakhtin that large networks of agents are fonts of innovation and adaptation is of course hardly exclusive to them. In the theory of natural selection, adaptive combinations of genes are unearthed in the evolutionary process via mutation. Different combinations of genes are distributed amongst different members of the populace, and the most useful sets are selected and distributed via reproduction. [22] There is also growing interest in the business and management literature in user innovation networks, as typified by the Open Source movement in software development. According to the Open Source Initiative:

My point in referring to networks is to suggest they are currently playing a role in modifying the way we tell stories. There are contemporary cries of quacks and druggists to be appropriated, and they are to be found in the world of computer gaming. I am specifically interested here, however, in one cry, the emergence of the concept of the game where the distribution of authorship is assumed from the start. As Warren Spector, the creator of the game Deus Ex notes, the initiative behind this form of story generation did not come from gaming companies. In fact, they initially resisted it:

The situation Spector describes is significant because it is network participants who are changing the nature of the relationship between gaming company and network, and it is network agents that are changing the distribution of authorship, not the company. It is the same process of self-organization, violation and transformation that Bakhtin describes in his account of carnival. [25] I would argue that the product of this process — the distribution of authorship to produce large-scale, content-rich, three-dimensional websites — is one that historians should appropriate for their own purposes, in a manner similar to Rabelais. In 2000, Will Wright, the celebrated developer of Sim City and The Sims, announced his intention to develop an on-line version of The Sims. Influenced by the writings of Christopher Alexander, an architect who advocates the generation of design tools and design principles based on studies of human interaction, Wright plans to stimulate the rise of an on-line urban system capable of supporting as many as 50,000 houses, as well as other types of buildings and civic infrastructure. As with his other games, users will be charged with developing content to inform and entertain other gamers, and content that can be appropriated by other gamers. He has produced a number of development tools to facilitate this process. [26]

Wright presents cultural, urban and environmental historians with a powerful template for content generation and representation, one that I believe can be adopted by historians. I present here the website of the 3D Virtual Buildings Project as an early prototype of how a platform for the display of historic urban environments might be constructed. As with The Sims, and other on-line games, the site is designed to host content posted by project participants. We have placed a set of guidelines on the site's tutorial page that explains the type and format of content eligible for posting, as well as documentation requirements. Participants can transmit the material by file transfer protocol to ftp://3dvirtualbuildings.ai.iit.nrc.ca.

My primary purpose, here, however, is to describe the characteristics of the site that enable participants to display their content in the first place, namely its structure, and its supporting instructional material. My object is also to demonstrate how the ethic of opensource might be adopted as a constituent of historic representation. When I refer to structure, I am referring to the principles governing the spatial and temporal division of our objects of display, namely historic Canadian settlements. I am also referring to the methods we have devised to govern a user's navigation through a historic virtual environment, and his or her interaction with objects within each display, specifically individual buildings.

As I indicated in Part I, one purpose of the 3D Virtual Buildings Project was to provide a platform to display historic urban settlements throughout Canada at disparate points in their history. While the current site, as shown in Figure 14, only shows representations of Ottawa in 1878 and 1912, the architecture of our site has been constructed to support models of other locales at disparate points in their history. But while bringing this goal into effect, the members of the project faced a constraint that had an additional bearing on the way we organized our site, namely limitations imposed by the Internet. Stated simply, the present bandwidth capacity of the Internet makes it impossible to transmit anything except crude representations of entire cities, with buildings reduced to little more than simple blocks. We wanted to represent more detail on our structures, a decision that meant we had to reduce the spatial extent of our 3-dimensional representations.

Stated another way, we had to divide our city of interest — Ottawa — into smaller chunks. In determining how we would do this, we decided to employ the template of our source material: fire insurance or Sanborn maps. There was a sound basis for making this decision. Each series of maps contains a key map, one which represents the entire city, and in turn divides it into smaller sections ascribing a number to each section. The utility of this template is that it remained stable over time. The cartographers who produced the 1912 series of fire insurance maps in Ottawa employed the same template to divide the city as had their predecessors in 1878. Map 104 in the 1912 series refers to the same section of the city as Map 104 in the 1878 series.

Figure 14

Accordingly, by adopting the organizational schema of our source material, we were able to construct a site in which it would become a straightforward process to divide new cities as they were added to our collection, simply by consulting the relevant key map of the series. In principle, by imposing a consistent distribution of space over any given city, the template also makes it easier to follow its transformation over time. Each date in the city's history represented by the site will have a button as indicated on the left in Figure 14, and each button will link to a key map, as indicated on the right in Figure 14. Users wishing to follow the transformation of a specific section of the city will simply need to click the relevant square or squares shaded in yellow to access the desired 3-dimensional representation. And most importantly, adopting the maps' organizational template enabled the project to display reasonably detailed representations of urban space — in this case Sparks Street as it existed in 1878 — as shown in Figure 15.

Figure 15

With respect to navigation, the project did not impose a linear or non-linear narrative structure on all or part of the site. There was no reason to. Our purpose is to solicit history students to post a 3D building, and a related narrative or set of narratives composed in text. Aside from spatial proximity, the project made no effort to provide the site's content with an underlying unity. With no common purpose to unite the project's disparate parts, there was no reason to mediate the user's experience via narration, and we have not done so. Instead, we opted for free navigation. More specifically, we opted for the forms of navigation available to sites constructed in VRML, Virtual Reality Modeling Language, the language commonly used to generate and transmit 3D objects over the Internet.

Streetscapes and other objects represented in VRML require a browser plug-in such as Cosmo Player in order to view them, as shown in Figure 16. Most VRML plug-ins provide two means for users to navigate through a three-dimensional site. The first is to provide a navigation panel such as the one indicated in Figure 16. It operates like the joystick in an airplane, enabling users to move straight ahead and backwards, up and down, and around an object. IT also allows users to tilt their point of view upwards at an angle while on the ground.

Figure 16

The second method is to use the "Viewpoint List". It operates much like an automatic pilot in an airplane, providing a list of destinations for users to visit, such as the list indicated in Figure 17.

Figure 17

If a user selects one of the sites in the menu, such as 95-99 Sparks, the site will then automatically "fly" the user to a set of coordinates entered into the system by the webmaster. Once the flight path shown in Figure 18 is finished, the user is then free to interact with the selected object to retrieve information on the building's history.

Figure 18

The central purpose of interacting with objects in the Virtual Buildings site is to retrieve information regarding the history of a specific structure, or the histories of activities and events associated with it. To facilitate the task of acquiring meaningful content, we felt we had to do two things. The first was to incorporate a practice common to many VRML sites, namely hyper-linking objects of interest. If you click a building, the site generates a pop-up window with relevant content. The second measure was to assist users in locating structures of interest. We assumed most users would be drawn to buildings based on their function, and accordingly we have constructed a database based on building function. Users enter queries based on the types of businesses or organizations located on the website. The site in turn identifies buildings of interest.

In the pop-up window we associate with each structure, we have provided links to both primary and secondary sources. For example, if users want to see a photograph of the Merchant's Bank, the structure shown in Figure 18, they can click the "Photograph" link shown at left in Figure 19.

Figure 19

The initial default image for each pop-up window, however, is a secondary source, a rendition of a highly detailed model, akin to the representation of the Merchant's Bank shown in Figure 20. The reason for its inclusion is that we wanted viewers to visually apprehend 19th century Ottawa, and any city for that matter, on two levels, the streetscape, and the individual building.

Initially, I wanted to produce and transmit detailed streetscapes with a level of detail akin to the model shown in Figure 20. After consulting with other members of the project, however, it became clear that given the bandwidth constraints of the Internet such an option was not feasible. Accordingly, participants are now asked to produce two models, one that is a scaled down version, as in Figure 18, the other a highly-detailed version, as in Figure 20.

Figure 20

The former, as indicated, is made a part of the streetscape, and is represented in VRML. The latter is displayed in isolation, and is provided to give viewers a sense of the building's actual appearance from a variety of angles. The building is not displayed in VRML or any other 3D modeling format. The size of one model would preclude transmission. Instead, aside from the initial graphic, we have opted to use QuickTime VR to represent the model. QuickTime VR is a software package that converts animations of high-end models to a series of still shots that are posted to a user's browser, thereby enabling examination of a model from a variety of angles, as shown in Figure 21. Viewers access a building's QuickTime VR representation by selecting the "3D Structures" link, located on the left margin of the pop-up window. Once the QTVR image is loaded into the window, viewers place their cursor on the display, and shift their mouse to the right or left to rotate the structure.

Figure 21

The pop-up window finally enables participants to access written essays and other primary sources through the "Building History" link, as shown in Figure 22. Users can use the "Credits" link to learn the name of the author or authors responsible for the 3D model and associated written content, and the schools or institutions to which they belong.

Figure 22

To assist users in locating objects of interest, we have provided three different instruments that are embedded in our web-site interface. On the left, as shown in Figure 23, is a search engine designed to enable users to locate specific types of structures, such as theaters, tin-smith shops and banks. At this juncture, we've equipped the search engine with a popdown menu listing the types of businesses and organizations currently housed on our site. The options in our menu reflect the structure of the site's database. In future, the search engine will be equipped with a conventional text entry field as the number of buildings on site increase.

The second instrument is a map, also shown in Figure 23. Its purpose is to enable users to determine their location in a given space, specifically the street. The map is also designed to display hits located by the search engine. Each shaded region corresponds with one of the 3D representations shown on the right. When the search engine generates a hit, the shaded region changes color. Users have the option of retrieving information by clicking the shaded regions on each map.

The final instrument is the 3D representation itself. Each structure is designed to turn to a specified color, namely yellow or blue, the background color of each search engine, if it corresponds with a hit.

Figure 23

To locate a structure, the user selects one of the options and presses the "Go" button, located to the right of the menu. The search engine generates a hit list. After the user has made a selection, he or she must click the hypertext link in order to determine its location. Once a hit is selected, the website regenerates the map and the 3D scene, as shown in Figure 24, turning the 2D and 3D representations of the structure yellow.

Figure 24

If a second search is conducted using the search engine below, the representation is generated again, turning the second building or set of buildings blue, as shown in Figure 25.

Figure 25

The viewer is then free to navigate to the structure of choice, click it, and then begin examination of the pop-up window for relevant information. Aside from interacting with content, the project also has an interest in assisting participants with the generation of content. While we do not provide tools as Will Wright does for his games, we do provide tutorials that enable participants to generate 3D objects with off-the-shelf software such as VectorWorks and Adobe PhotoModeler. We also provide tutorials to assist students in the process of primary and secondary research. Figure 26 provides a view of two of our most significant tutorials. To the left is a view of our 3D modeling tutorial, which takes students step by step through the process of model construction, from initial analysis of two photos and a fire insurance map, to the realization of a fully finished model like the Hope building shown in Figure 10. [27] To the right is the tutorial Using Archives, which provides students with an introduction to the concepts associated with primary research. Produced in cooperation with the National Archives of Canada and Emc2 Computer Consulting, the tutorial explains to students in simple English everything from how to use a finding aid, to the location of their nearest archival repository. [28] We have also produced tutorials devoted to the conduct of secondary research, the conversion of high-end CAD models to simpler VRML models, and the generation of a non-European settlement, the First Nation Haida settlement at Ninstints. [29]

Figure 26

Stated simply, our purpose in this section is to facilitate the rapid and easy generation of new 3D content. As I indicated in Section II, we have some way to go before achieving that goal. One reason for our continued emphasis on the issue of 3D content generation, however, is that we would like to make our repository a dynamic, self-correcting representation of the past, one that embodies the ethic of open-source. Consider the model of the Merchant's Bank shown in Figure 27. It was the first model I generated for the project. And, not to put too fine

Figure 27

a point on it, it contains a number of significant errors. While building it in 1998, I based it on a fire insurance map, two historic photographs I was able to locate, and contemporary shots of the structure as it exists today. Observe that neither of the older shots provides a good view of the façade indicated in Figure 28.

Figure 28

I remained dependent on photographs of the contemporary building, a decision that in retrospect was a mistake. Figure 29 shows a photograph I discovered after completing my model. It also shows a contemporary shot of the wall of interest, which I'll label Façade 2. Note first that I incorrectly omitted the door to the right. Based on my view of the left historic photograph shown in Figure 28, I did not believe that the 1878 version included it, an inference that was wrong. I also failed to generate the right number of windows. Based on the contemporary photo, I produced four. Based on information from the historic photo, I should have produced five.

Figure 29

I also relied on the contemporary structure to determine the dimensions for Facades 1 and 2 indicated in Figure 30, a decision that in retrospect was also wrong.

Figure 30

While constructing the model, I failed to properly recognize, as Steward Brand notes, that buildings are not static things. Like us, they change their appearance over time, in order to meet the needs and expectations of their contemporary inhabitants. [30] Figure 29 indicates that the Merchant's Bank — today referred to as Canada's Four Corners, a souvenir shop — was rebuilt, a decision that affected not only the appearance of Façade 2, but also the dimensions of Façades 1 and 2. While constructing the model, I was well aware of the discrepancy between the appearance of Façade 1 in the old and new photograph, but attributed the difference to a distortion in the old photograph, an artefact from the lense in the old camera.

Again, this was a wrong assumption. My point in generating this inventory of error is to suggest that the 3D Virtual Buildings Project and other initiatives could house mechanisms that would effectively make them open-source, with virtual environments subject to constant expansion and correction. I am suggesting that in such a venue, comprised of a network of historians familiar with the source material, that the mistakes I made would have been more easily, and more quickly, identified and rectified. Such an initiative would require at a minimum a peer-review panel to scrutinize new contributions to the site, and suggested modifications. Our colleagues in archaeology are already applying this method of scrutiny to such projects as the Rome Reborn project, and we would do well to follow their example. [31]

An open-source approach to historical representation would also require a higher level of documentation than our project presently employs. Each object associated with a structure — from pillar to pilaster — would require the three-dimensional equivalent of a footnote. Peer review committees — and viewers — would require access to the documentation referred to by a historian to justify his or her construction of a particular object or set of objects. They would further require an inventory, itemizing which attributes of specific objects and buildings are based on documentation, and which attributes are based on the assumptions of the historian. The historian in turn would be required to provide primary sources to demonstrate that their assumption is justified. The documentation process would be important, because it would form the basis for a committee's decision to accept a new submission, to replace an existing model with a new interpretation, or to simultaneously display competing interpretations of historic structures, if extant evidence favors neither proposal.

IV

The vision I have described above would, to say the least, require a huge effort, not only in terms of maintaining quality control, but also in terms of generating a network of historians committed to the site's maintenance, expansion and correction. Such is the price of living in an interesting era. But as I hope I've demonstrated in this article, it is a price worth paying. 3D environments are instruments, and if properly exploited they stand to provide historians with substantial gains in their capacity to teach, represent and analyze the past.

My purpose here first has been to suggest that not all sign systems are equal in their capacity to reveal the grammar of history. If the historical model and the historical events can be construed as signs, the relationship between the two can be construed as their grammar. And here, I have argued that the technology of print is a poor means to represent the syntax of the historical discipline. It is akin to consulting long reams of numbers to follow the trajectory of the stock market. A graph is a far more efficient way to represent and perceive it, and in this study I have argued that three-dimensional objects can fulfill a similar purpose. Students construction of virtual buildings can help them understand the true relationship between historical representation and historical object, by enabling them to see that all historical reconstructions — be they monograph, article or 3D model — are mediated, by the author, and by the evidence.

My second purpose here has been to suggest that historians should expand their expressive repertoire to include three-dimensional environments. The basis for this argument in large measure derives from the communication theories of Harold Innis. I am advocating the practice of the Oral Tradition in 3D. Innis argued that historically cultures that adopted new conventions for story-telling often came to see their surrounding environment, past and present, in new ways. I suggested here that historians should do the same. There is risk in investing time to learn to generate and exploit virtual space. But current scholarly interest in the history of spatial environments, and the discursive construction of spatial environments, suggest that such an investment could also produce substantial gains.

In this study, I have also argued that it would be impossible for a single historian to take on the task of generating and maintaining a virtual city all by him or her self. Such a repository requires too much work to generate and maintain. The solution to this quandary lies, I believe, in the realm of popular culture. I have argued that historians should appropriate conventions derived from popular culture, like François Rabelais, and specifically the convention that distributes authorship of a virtual space. Such a practice, I suggest, would enable historians to generate dynamic, self-correcting representations of the past. It would enable them to embrace the ethic of open-source.

It would also enable them, in closing, to build a set of skills that will enable them to prepare for the evolution of the Internet, and the emergence of new forms of storytelling that derive from it. It is hard to overemphasize the importance of broadband communication, and the emergence of wireless as one of its primary delivery systems. The existence of Blackberry and other wireless networks is already prompting the search for new platforms, and new interface paradigms, ones that will transcend the limitations of the cell phone. One possibility in this area is Augmented Reality, or AR. AR is a variant of what is referred to as mixed reality, a mode of representation in which one's view of real space is augmented by virtual objects.

In this vision of computing, users in near future will wear computers with the computational power of today's desktops, and the size of today's personal digital assistants. These computers, in turn, will be connected to wireless networks to access and post information, and to head mounted displays the size of glasses to display information. I mention this newly emerging field because it is already making a contribution to the way we represent the past, and the way we tell stories. In principle, it should be possible in the next 10 to 20 years to produce something akin to the holodeck from Star Trek. We will not be able to interact with objects without the mediation of glasses or gloves. But we should be able to generate representations of ancient Rome or 19th century Paris, and project them onto football fields. [32]

Stated another way, the genres governing story-telling are continuing to evolve, and historians would do well to follow the lead of David Staley, the Center for History and New Media, and the Institute for Advanced Technology in the Humanities in exploring how new technologies might be applied. [33] The risks in so doing can be substantial. Our colleagues often receive little or no credit towards tenure or promotion from their work with information technology. But if this paper has fulfilled its purpose, then it will have illustrated another point: the rewards can be substantial as well.

To read the panel introduction, click here.

Endnotes

1. See Jack Schofield. "Birth of a Medium: The Great Art of the 21st Century will not be on campus but computers. Jack Schofield talks to Henry Jenkins to find out why games are the new Goya." In The Guardian. July 11, 2002. Available On-line: http://www.guardian.co.uk/arts/features/story/0,11710,753613,00.html. [November 5, 2002].

2. Janet H. Murray. Hamlet on the Holodeck: The Future of Narrative in Cyberspace. New York: The Free Press, 1997; Tom Taylor. "Historical Simulations and the Future of the Historical Narrative." This edition of Journal of the Association for History and Computing.

3. Nathalie Ferenc, Ronan Boulic and Daniel Thalmann. "An Informed Environment dedicated to the simulation of virtual environments in urban context." In EUROPGRAPHICS '99. Volume 18 (1999), Number 3. Available online: http://ligwww.epfl.ch/~thalmann/papers.dir/233P-Farenc.pdf. [November 5, 2002]; Frédéric Dan Tran, et al. "An Open Middleware for Large-scale Networked Virtual Environments." In Virtual Reality, 2002. Proceedings, IEEE, 2002.

4. Our initial partner within Industry Canada was SchoolNet: http://www.schoolnet.ca. Digital Collections — another department within Industry Canada — provided our second round of funding.

5. See either: http://www.digitalcollections.ca/E/alphabet.asp, or http://3dlearning.iit.nrc.ca/3DVirtualBuildings.

6. See Jack Granatstein Who Killed Canadian History?, (Toronto: HarperCollins, 1998). Other discussions include "Past Dues", Ottawa Citizen, 19 April 1999; Derek McNaughton, "Teaching of History in 'crisis'," Ottawa Citizen, 1 February 1999; Paule Des Rivières, "Quelle histoire, quels enseignants? Histoire et école," Le Devoir, 1 février 1999; "Quéstions d' historiens: 8 experts se réunissent," Le Devoir, 30 janvier 1999.

7. Margaret Atwood, In Search of Alias Grace: On Writing Canadian Historical Fiction, (Ottawa: University of Ottawa Press, 1997).

8. Seymour Papert, The Children's Machine: Rethinking School in the Age of the Computer, (New York: Basic Books, 1993). Yasmin Kafai and Mitchell Resnick note:

9. For more information on these primary sources, see Alain Rainville. "Fire Insurance Plans in Canada." In Archivist: Magazine of the National Archives of Canada. 1996(111): 25-38. Kim Keister. "Charts of Change." In Historic Preservation. 45(3): 42-49; 91-92. 1993.

10. Interview with Ellora Faruq, Building M-50, National Research Council of Canada, Ottawa, ON, 9 December, 1999.

11. Christine Youngblut. Educational Uses of Virtual Reality Technology. Alexandria, VA: Institute for Defense Analyses, 1998. Dorothy Strickland, Skip Rizzo, Yasmin Kafai, and Larry Hodges. "Games and Cognition." At Entertainment in the Interactive Age. Available on-line: http://www.annenberg.edu/interactiveage/assets/transcripts/gac.html. [November 5, 2002].

12. On this point see the proceedings of the Entertainment in the Interactive Age conference at: http://www.annenberg.edu/interactive-age. [November 5, 2002]. Also see Murray, Hamlet on the Holodeck. Tony Manninen. "Towards Communicative, Collaborative and Constructive Multi-player Games." In Proceedings of Computers Games and Digital Cultures Conference, June 7-8, 2002. Ed. F. Mäyrä. (Tampere, Finland: Tampere University Press, 2002): 155-169.

13. Mark Billinghurst, Hirokazu Kato, and Ivan Poupyrev. "The MagicBook: a transitional AR interface." In Computers and Graphics. 2001. 25(5): 745-755.

14. Warren Spector's game Deus Ex is one example of a game in which different outcomes can be reached based on narrative interaction. See Warren Spector et al., "The Audience Takes Change," in Entertainment in the Interactive Age, January 29-30. 2001, at http://www.annenberg.edu/interactive-age/assets/transcripts/atc.html. [November 3, 2002]. For a treatment of multiple outcomes as expressed in literature, see Gary Saul Morson. Narrative and Freedom: the shadows of time. New Haven: Yale University Press, 1994. Niall Ferguson argues historians should incorporate exploration of counterfactuals in Virtual History: alternatives and counterfactuals. New York: Basic Books, 1999.

15. See Innis' Introduction and Chapter Four, "The Oral Tradition and Greek Civilization." In Harold Innis, Empire and Communications. (Toronto: University of Toronto Press, 1972, c. 1950): 3-11; 53-84. For a recent interpretation of Innis' treatment of the Oral Tradition see John Bonnett. "Communications, Complexity and Empire: The Systemic Thought of Harold Adams Innis." Ph.D. Thesis, University of Ottawa, 2002. pp. 253-265.

16. James Gleick, Chaos: Making A New Science. (New York: Viking Penguin, 1987): 11-31.

17. Jonathan Rauch. "Seeing Around Corners." In The Atlantic Monthly. 289(4): 35-48. April 2002. Also available on-line: http://www.theatlantic.com/issues/2002/04/rauch.htm. [November 2, 2002].

18. Taylor, "Historical Simulations," in this edition of the JAHC.

19. For a different treatments on interaction with 3D objects and 3D environments, see M. Kallmann, D. Thalmann, "Direct 3D Interaction with Smart Objects." In Proceedings of ACM VRST'99, London December 1999. Available on-line: http://citeseer.nj.nec.com/242852.html. [November 5, 2002]; and John A. Waterworth. "A Pattern of Islands: Exploring Public Information Space in a Private Vehicle." In Eds. P. Brusilovsky, P. Kommers, and N. Streitz. Multimedia, Hypermedia and Virtual Reality: Lecture Notes in Computer Science. Springer-Verlag, 1996. Available on-line: http://www.informatik.umu.se/~jwworth/islands.pdf. [November 5, 2002].

20. Ariana Eunjung Cha. "Web May Hold the Key to Achieving Artificial Intelligence," Washington Post, September 6, 2002. John E. Laird. "Research in Human-Level AI using Computer Games." In Communications of the ACM. 45(1): 32-35. January 2002. Gal A. Kaminka et. al. "GameBots: A Flexible Test Bed for Multiagent Team Research." In Communications of the ACM. 45(1): 43-45. January 2002. For reference on work relating to the creation of functioning agents in robotics, see Curtis Gillespie. "Charmed by Six Feet of Circuitry." New York Times, August 8, 2002.

21. Mikhail Bakhtin, Rabelais and His World. Trans. Hélène Iswolsky (Bloomington: Indiana University Press, 1984, c. 1968): 187-188.

22. M. Mitchell Waldrop, Complexity: The Emerging Science at the Edge of Order and Chaos. (New York: Simon and Schuster, 1992): 165-170.

23. Open Source Initiative. "Open Source Initiative OSI — Welcome." At http://www.opensource.org. [November 2, 2002]. For a treatment of the open source movement from the standpoint of business and management literature, see Eric von Hippel, "Horizontal innovation networks — by and for users," MIT Working Paper, http://opensource.mit.edu/papers/vonhippel3.pdf. [November 2, 2002]. With respect to user-innovation networks, von Hippel writes:

24. Warren Spector et al., "The Audience Takes Change," in Entertainment in the Interactive Age, January 29-30. 2001, at http://www.annenberg.edu/interactive-age/assets/transcripts/atc.html. [November 3, 2002].

25. Bakhtin notes of carnival:

Spector notes that gaming companies are adjusting their practice not out of altruism, but because of economic imperatives. In computing gaming, and indeed in any area of software, a primary goal of developers is to build a network of users who are committed to their product, consumers who will continue to purchase the product as it is upgraded over time. Spector notes that gaming companies have made concessions — including providing open access to their gaming engines, the software that produces games — to maintain the relationship with their respective communities. Games are more expensive, and take longer to develop. Other measures are now required to maintain consumer interest. Companies have also found the practice useful, since it provides training for new generations of young game developers prior to their entry into the job market. See Spector, "The Audience Takes Charge." On the importance of consumer networks in the software industry, see W. Brian Arthur, "Increasing Returns and the New World of Business," in Harvard Business Review, July-August 1996. Available On-line: http://www.santafe.edu/arthur/Papers/Pdf_files/HBR.pdf. [November 3, 2002].

26. Will Wright, "Keynote," in in Entertainment in the Interactive Age, January 29-30. 2001, at: http://www.annenberg.edu/interactive-age/assets/transcripts/willw.html. [November 3, 2002].

27. See http://3dlearning.iit.nrc.ca/3DVirtualBuildings/Title/Title.html. [November 4, 2002].

28. Seehttp://3dlearning.iit.nrc.ca/3DVirtualBuildings/UsingArchives/Title/Title.html. [November 4, 2002].

29. See http://3dlearning.iit.nrc.ca/3DVirtualBuildings/Tutorial.html. [November 4, 2002].

30. See Stewart Brand, How Buildings Learn: What Happens after They're Built. New York: Viking Penguin, 1994.

31. Bernard Frischer, "Virtual Reality and Ancient Rome: The UCLA Cultural VR Lab's Santa Maria Maggiore Project," by B. Frischer, D. Favro, P. Liverani, S. De Blaauw, Virtual Reality in Archaeology, BAR International Series S 843, ed. J. A. Barcelo, M. Forte, and D. H. Sanders (ArcheoPress, London 2000) 155-162 Available online: http://www.cvrlab.org/humnet/smm.html. [November 4, 2002]. Also see "CAA UK 2001 — Panel Discussion: Misrepresenting the Past: Virtual Reality and Archaeology." At http://cs.bris.ac.uk/Events/CAAUK2001/Program/panel.html. [November 5, 2002].

32. For introductions of the field of Augmented Reality, see Steven K. Feiner. "Augmented Reality: A New Way of Seeing." In Scientific American. April 2002. Available on-line: http://www.sciam.com/article.cfm?colID=.... [November 5, 2002]; Ronald T. Azuma. "A Survey of Augmented Reality." In Presence: Teleoperators and Virtual Environments. 6(4): 355-385. August 1997; Ronald Azuma et. al. "Recent Advances in Augmented Reality." In IEEE Computer Graphics and Applications. (November/December 2001): 34-47. For overviews of recent attempts to extend AR to storytelling and history, see Wayne Piekarski and Bruce Thomas. "ARQuake: The Outdoor Augmented Reality Gaming System." In Communications of the ACM. 45(1): 36-38. January 2002. Tobias Hollerer, Steven Feiner, and John Pavlik. "Situated Documentaries: Embedding Multimedia Presentations in the Real World." In Proceedings of the ISWC'99 (International Symposium on Wearable Computers), San Francisco, CA, October 18-19, 1999, pp. 76-86. Steve Benford et al. "Applications and Design Spaces: Unearthing Virtual History — Using Diverse Interfaces to Reveal Hidden Virtual Worlds." In Lecture Notes in Computer Science. 2001 , v. 2201 , pp. 225-232.

33. David Staley. Computers, Visualization, and History: How New Technology Will Transform Our Understanding of the Past. Armonk, NY: M.E. Sharpe, 2003. (Forthcoming); George Mason University's Center for History and New Media is exploring 3D forms of representation in its "Lost Museum". See http://www.lostmuseum.cuny.edu. [November 5, 2002]. Katherine Rinne of the Institute for Advanced Technology in the Humanities is also employing VR in her treatment of the hydrological and hydraulic systems of ancient Rome. See http://www.iath.virginia.edu/waters. [November 5, 2002].

John Bonnett, "Following in Rabelais' Footsteps: Immersive History and the 3D Virtual Buildings Project." © 2003, National Research Council of Canada. All rights reserved.