"I thought of the beauty of ruins . . .
of things which nothing lives behind . . .
and so I thought of wrapping ruins around buildings."
                                                                     Louis I. Kahn

Introduction
The story is well known.  Kahn, having built little of note by the age of fifty, spends four months as Architect in Residence at the American Academy in Rome.  During this time he experiences the great ruins of the ancient world and resolves that "the architecture of Italy will remain as the inspirational source of the works of the future."  He returns in 1951 to immediately execute his first major commission, and struggles for the next 23 years to incorporate lessons learned in Italy, Greece, and Egypt.  In the process he redefines modern architecture and becomes the most important architect of the second half of the 20th Century.  He dies at the height of his career after building many of the masterworks of our time: the Kimbell Art Museum, the Laboratories of the Salk Institute, Exeter Library, the Yale Center for British Art, and his monumental projects on the Indian Subcontinent.

The role played by work Kahn could not build is less well known.  Between 1959 and 1961, Kahn used a series of fascinating unbuilt projects - particularly the American Consulate in Angola, the Meeting House of the Salk Institute, and Mikveh Israel Synagogue - to work out and test his new ideas.  In these projects, Kahn developed elements later found in his built work: a configuration of space as discrete volumes, complex ambient light and shadow, a celebration of mass and structure, the use of materials with both modernist and archaic qualities, monumental openings uncompromised by frames, and Kahn’s concept of "ruins wrapped around buildings."  At the end of the 60’s he created what is perhaps the clearest expression of this link to the old world - the Hurva Synagogue for Jerusalem.  Finally, the unbuilt Palazzo dei Congressi in Venice prefigured a significant change in direction, as evidenced by his last major built work, the Yale Center for British Art in New Haven.

This exhibition attempts to shed light on the eternal mystery of how Kahn came to make the architecture he did, by looking in depth at projects left unbuilt.

Subject
The exhibition will feature radiosity-based, hyper-realistic computer graphic renderings of 8 unbuilt masterworks of Louis I. Kahn: U.S. Consulate for Luanda, Meeting House of the Salk Institute, Mikveh Israel Synagogue, Memorial to Six Million Jewish Martyrs, Hurva Synagogue (first, second and third proposals), and the Palazzo dei Congressi.   The exhibition will coincide with the publication of the book Louis I. Kahn: Unbuilt Masterworks by Kent Larson, Monacelli Press (with foreword by Vincent Scully, afterword by William J. Mitchell)

Digital/Analog
The exhibition will incorporate sophisticated digital technology (computer graphic simulation, 3D printing of digital models, and computer vision tracking). It will combine this with a seemingly non-digital interface (physical models moved like chess pieces) and high-resolution analog images (rear-projected 35mm slides).

Exhibit Layout
Physical Models: A central table, 54 inches to a side, will be placed in the center of the exhibit. A 3D printed physical model of each of the 8 unbuilt projects, on bases 3" x 3", will be located, two to a side, at the perimeter of the table. Each model will have a digital tag that identifies it. In the center of the table will be a 3" x 3" depression. When a model is placed in the depression, a sensor will identify the model.

Selection of Views

Projected Plan
When the table identifies that a physical model has been placed in the central receptor, a square image of its schematic floor plan will be projected onto the floor of the exhibit from a projector located at the ceiling with a mirror placed at 45 degrees. Conceptual images of the selected building will be projected on the four 6'8" x 10'-0" rear projection screens.

The schematic plan projected on the floor of the exhibit will contain graphic symbols indicating camera positions for the selected project. A infra-red camera aimed at the table, and linked to a computer vision system, will allow viewers to select the available camera positions. A view is selected when cylindrical object is placed over a camera position symbol. If no view is selected within one minute, the system will automatically scroll through views of the selected project until a view is selected or a new building is chosen.

Interaction
Negotiation will be required by viewers to select both the building to be visualized (through the placing of the physical model into the central position) and to select the views (by selecting the view symbol). This should result in an interesting degree of unpredictability.

Projected Images
An average of four camera positions of each building will be presented - each rendered with front, back, left, and right views - resulting in 16 high resolution renderings per building, or a total of 128 perspectival images for the exhibition. The vanishing point of each will be approximately eye-level, providing the viewer with an appropriate sense of scale.  Four 6'8" wide x 10'0" high rear projection screens will be placed to define the central exhibition space. The 4 mated images will be projected on the rear-projection screens.  Each of the four images will be 4096 by 2731 pixels, for a total of approx. 44,750,000 pixels per camera position. Low resolution examples of image assemblies follow:

                              First Proposal for the Unbuilt Hurva Synagogue, Jerusalem, 1968
 

                         Third Proposal for the Unbuilt Hurva Synagogue, Jerusalem, 1972
 
 

                                The Meeting House of the Salk Institute, La Jolla, California, 1959-65

Film
A 15-minute video, played on a repeating video disc player, will present the first proposal for the Hurva Synagogue for Jerusalem, with a people composited into unbuilt space

Archival Material
Eight computers with 20 inch monitors, one for each building, will present material from the Kahn Archive at the University of Pennsylvania.

Projects Included

U.S. Consulate for Luanda, Angola, 1959-62 (unbuilt)
Meeting House of the Salk Institute, 1959-65 (unbuilt)
Mikveh Israel Synagogue, 1961-72 (unbuilt)
Hurva Synagogue, 1st Proposal, 1967-68 (unbuilt)
Hurva Synagogue, 2nd Proposal, 1972-73 (unbuilt)
Hurva Synagogue, 3rd Proposal, 1974 (unbuilt)
Memorial to Six Million Jewish Martyrs, 1966-72 (unbuilt)
Palazzo dei Congressi, 1968-74 (unbuilt)

Kent Larson
265 Massachusetts Avenue, N51-336
School of Architecture and Planning
Massachusetts Institute of Technology
Cambridge, Massachusetts 02139
tel 617-253-8799
email kll@mit.edu

copyright  Kent Larson 2000

A new architecture exhibit at the University of Pennsylvania is a superb example of the way architecture should be presented to specialists and the general public alike. The exhibit was prepared by Kent Larson, MIT professor and New York architect, for his exhibition entitled "Unbuilt Ruins: Digital Interpretations of Eight Projects by Louis I. Kahn." The exhibition features Mr. Larson’s work on buildings that were designed by Louis Kahn but never constructed; he has completed CAD models of them, based on the original sketches, correspondence about the projects, study of other Kahn buildings, and the like. The exhibit will be on display in the Fisher Fine Arts Library (a building designed by Frank Furness and recently brought back to l ife by an excellent and expensive refurbishing) at the University of Pennsylvania until May 19. (Unfortunately, hours are limited to 9 am to 5 pm, Monday through Friday.) It will be on display in the Palladio Center, Vicenza, Italy, in the fall of this year. In addition, a Web page concerning the exhibit has been mounted for access at http://architecture.mit.edu/~kll/www_compton/exhibit.html.

Collaborating with Mr. Larson were Lawrence Sass, MIT Department of Architecture; Ron MacNeil, MIT Media Laboratory & MIT Department of Architecture; and Flavia Sparacino, MIT Media Laboratory.

Mr. Larson has built computer models of various Kahn structures, but one of the virtues of the exhibit is that he does not present the museum-goer with computers or computer-like space. Instead, the visitor sees a simple table with the plan of one of the Kahn buildings or complexes on the top of the table. There is room around the table for four or five people on each side, but behind the people are rear-projection slide screens, making a kind of exhibit room of the space within the screens, with the table in the center. On the screens are projected images of the building or complex, the plan of which is shown o n the table. The four images show views ninety degrees apart; they are not wide-angle lens views; so the result is not a complete 360-degree view.

There are several places marked on the plan with red crosses, and the slides show four views of the building or complex taken from one of those spots. Thus, one looks at the plan on the table to orient oneself, then looks around the "room" created by the projection screens to see images taken from the marked location. Taken together, the four images provide a panorama of the space from the marked location.

The images are the key to the exhibit, because, quite simply, they are the best computer-generated images of architectural models that I have ever seen - by far. Mr. Larson used photographs of the materials common to Kahn’s structures as well as photographs of the surroundings of the buildings to help create the images, but he also manipulated the lighting to provide a sense of verisimilitude that is startling. (He even included an image of his daughter, then four years old, in one of the views to provide human scale.)

The images are not on computer screens; they are projected slides derived from computer images. The advantage is not merely size; projectors can enlarge computer images in the same way. The slides are much better than computer-screen images. Having no tell-tale jagged edges, they provide better smooth surfaces, and they do not have the appearance of being computer-generated in any way. In short, they are spectacular. (The images on the Web page listed above are similar but display the aliasing problem - jagged diagonal lines - common to computer images. The slides are much better.)

The exhibit uses computer technology in important ways, but the visitor is not aware of computers or computing.

The plan of the building or complex shown on the table is actually produced by a projector shining onto the table from directly above. The plan shown is therefore adjustable; the projected image can be changed. Museum-goers can do that for themselves. On the table are blocks, one for each of the unbuilt projects, and a visitor simply places the appropriate block in a receptacle in the center of the table to change the visible plan to that of another project. The slides change then as well. Thus, all the projects can be seen in a single "room."

Once the correct plan is on display, one sees the red crosses on the plan that indicate points from which the "photographs" were made. (Calling them images suggests the wrong idea; these seem to be photographs.) A cylinder sitting on the table can be moved to any of the red crosses, and its position will be sensed; once the cylinder is in the proper place, a new set of slides is projected on the screens. At the same time, the plan on the table is adjusted to indicate what part of the building or project falls within the view of the "camera lens." The visitor can understand precisely the views presented.

These are not virtual reality models. Viewers cannot metaphorically wander about the models and look in any direction from any point. Images are available only for the points marked and in the directions selected in advance. Mr. Larson wanted better images than could be obtained from a VR system; so all the images had to be generated in advance. Others might prefer a VR model with lower-quality images in return for the ability to experience views from any position within the model. For me, though, the quality of these images is such that I much prefer this exhibit to a VR model. These seem to be, as I have already said, photographs. In addition, the use of selected points and prepared images means that a museum-goer must use the plans carefully to figure out locations and vantage points; the visitor is obliged to think about what he or she does, as the exhibit planners intended. Of course, there is also no computer-game aspect to this, as there might be with a VR system.

The use of computers to provide the models and the "photographs" but not to be the core of the exhibit is an excellent choice. The power of the technology has been harnessed to provide effective images, and appropriate ways of presenting the information have been used. This is an exhibit to be seen and cherished.

- Harrison Eiteljorg, II