A Modern Timelessness: Kahn, Piano and the Kimbell Art Museum

In keeping with the ongoing tradition of architectural rivalry with Dallas, the latest addition to Fort Worth's world-class cultural district was unveiled late last week. The Kimbell Art museum, which is housed in what is arguably Louis Kahn's greatest masterpiece built in 1972, has presented the final design of its long-awaited addition realized by Renzo Piano, an architect who has earned the reputation as the world's premier designer of museums. The addition more than doubles the exisiting exhibition spaces, provides additional lecture halls, and most importantly for its Texan patrons, abundant parking below ground. The news follows a rather eventful year for high-design in the metroplex, beginning last fall with opening of two brand new buildings in the Dallas Arts District, the unveiling of the Perot Museum of Science by Thom Mayne (currently under construction), and the recent opening of one of Ricardo Legoretta's stronger works to date, the Fort Worth Museum of Science and History.

Compared to Piano's numerous other museum commissions over the years, Texas has been of special value to his success and development. The state has now four of his projects, including two in Houston (the Menil Collection and Cy Twombly Gallery) and one in Dallas (Nasher Sculpture Center). It was in the Menil in 1986 that Piano established his most recognizable architectural prototype, the multi-layered roof that lightly hovers over a single-story art pavilion and bathes spaces underneath with diffuse natural light. The Twombly across the he street expanded further with the addition of motorized sun controls and fabric, and the Nasher exhibits an innovative eggcrate-like system of sunshades. Having closely followed his career since going to the Menil for the first time while in college, it seems that the Italian has blessed Texas with his best work. It's certainly a bold statement, but it has much to do with particular aspects to which I think Mr. Piano excels that are coincidentally rare in most of his projects elsewhere. Upon examining the drawings and images of his latest plans for the Kimbell, it seems he is reinforcing those aspects further, endowing these projects with a timeless beauty, yet preventing him from doing likewise for other projects of differing scales and function.

Whither the Cardboard Model?

For those fortunate to have attended college on a campus with an architecture school, visitors to the architecture buildings are instantly captivated by the small architectural models built by students. These little cardboard or bass wood miniatures that manifest spatial ideas in physical reality is often the most accessible way for non-architecture laymen to understand a design concept. At the same time, the ubiquitous display of models throughout an architecture school give an impression to laymen that architecture lacks a level of seriousness, since spending time doing arts and crafts is commonly understood as a part of child's play. It gives the impression that architecture students must be having fun while in reality there is nothing more tiring and frustrating than gluing small pieces together with such concentration and precision.

Along with the challenges posed by building the model, there is much expense involved as well. It is easy t0 squander hundreds of dollars for a reasonably detailed model for a school project. For some students, modelmaking is a serious craft, the goal being not to solve three dimensional problems toward improving a project's design, but rather to execute a model with great precision, detail and realism. Lacking fine motor skills, patience, and being naturally stingy, I made models as seldomly as possible, and made them intentionally rough as a tool for spatial exploration and problem-solving. I intended them to be abstract, so that concepts could be more powerfully presented, while also thinking that too much detail winded up giving a model a dollhouse-like character.

The major reason I tried to avoid building too many physical models was the amount of time it consumed. There's no way of rushing a model. One can only cut a straight line with an exacto only so fast and glue can only harden so quickly. And still , there was no way of testing various alternatives without building three separate models. The finished results might be intriguing for onlookers and will better help those who want to discuss the project with the student. But they were a chore to complete, and I knew that model-building skills were among the least important in the expansive practice of architecture.

Therefore, as soon as I enrolled in studio classes that did not require physical models for presentations, I switched to using computers to create virtual three-dimensional models. In addition to ensuring high-quality line drawings of unparalleled accuracy, computer aided design (CAD) could build models very quickly and with infinite precision and verisimilitude. The best part was that it spared me from making trips to the nearby crafts store, thus saving me substantial money. It also helped that computer skills have become essential to getting any entry level job in architecture. As an additional benefit, becoming proficient in computer-aided design served as a gateway to understanding computer animation, graphic design and artistic rendering.

With all these advantages, there was yet a significant drawback to computer models: a stranger could not take digital model and view it from any angle in the physical realm. The closest means of doing this on the computer is for that stranger to muddle through unfamiliar software and navigate in virtual space. Since the model could not come to them, the next best thing was to print on paper as many points of view as possible. A digital presentation often consisted of dozens of different views to help convey the spatial idea. This often made such presentations disorienting to those unfamiliar with the project. The physical cardboard model, by contrast, was a lot more straightforward in that it could be viewed from every angle simultaneously by moving one's eye.

Since finishing architecture school, I've rarely had to dedicate any effort into building physical models. Clients will almost never pay for study models, setting aside money only for presentation models useful in selling the project to potential tenants and investors. Since the latter type requires tremendous amounts of time and high levels of workmanship, they are usually outsourced to an independent model making workshop (which are increasingly found in places like China and India). Architect's fees are inherently tight thus rewarding an efficient use of time and manpower that are antithetical to model making. Computer models suffer from an inaccessible interface for people unfamiliar with the software, making older architects unable to fully appreciate the merits of a design. Until this problem is resolved, whether by rapid prototyping or by creating a virtual reality environment that anyone can physically participate in, non-reimbursed cardboard study models will be used from time to time.

As I've taken on more responsibilities on the design end in my professional projects, I've maintained my aversion towards physical models. Luckily, computer modeling software in the last few years have made significant leaps towards intuitiveness and user-friendliness. They are easier to learn, have more efficient rendering engines, and yield quick results. I particular enjoy using Sketchup to investigate massing, shadows, perspectives, and quick fly-through animations. Although it tries to bridge the art of sketching with modeling, Sketchup is deceptively simple, providing few tools to modify objects and yet it accomplishes better results than more expensive and complicated software. Together with trace paper, the schematic design process goes quickly, and the program's compatibility with standards drafting software allows for little time to be wasted in producing two-dimensional plans and elevations.

With all these new innovations, what will become of the cardboard model? My guess is that it will still be of use so long as computer model remains virtually existent int the hard-drive of a computer. Clients can understand a model over something that is almost unintuitive in the forms of architectural plans. Still, a computer-generated rendering is more seductive, can be more easily edited and obviously ship better than a physical model. I predict that physical models will eventually cease becoming the product of exacto knives and glue. In their place will arise models built by machines following computer generated architectural plans. Such technology is in its infancy now, but will soon permit designers to more productively test alternatives. The architect becomes more empowered when he or she does not have to worry about being a craftsman of miniatures.