On When to Know When to Stop Drawing...

I recently had the honor of working in Tom Beeby’s office on the renovation of one of my favorite buildings: Gordon Bundshaft’s Beinecke Rare Book and Manuscript Library. Often described as a “jewel box,” the building’s façade is a tableau of a white, grey-veined marble cut so paper-thin that it serves as a filter, allowing light into the interior. The result is a daunting and solid exterior that simultaneously guards and creates the soft, luminous interior spaces housing one of the world’s greatest collections of rare books, manuscripts, and documents.

Bundshaft and his team at Skidmore Owings and Merrill communicated the design intent for the library in less than 30 pages. Our construction set for the renovation of the building took more than three times that many pages, structured across several phases of construction.

From the time Beinecke opened in 1963 to the completion of its first major renovation in 2016, the profession of architecture underwent a revolution. Architects shifted from drafting by hand on mylar and vellum to drafting on the computer. While this change altered the profession in ways we as architects are still trying to understand, the computer allows drawings to proliferate at much greater speeds than when they were hand drafted. Iterations can quickly be modeled, and “copy” and “paste” allow for the quick generation and exploration of numerous variations.

The amount of time that architects spend on our jobsites has undergone a corresponding shift. Bundshaft worked out of SOM’s New York Office, meaning that he could quickly hop on Metro-North up to New Haven to inspect the jobsite. At Beeby’s office, we were over nearly a thousand miles away from the building we were working on, communicating our ideas to a local architect who spent time on site. One of my coworkers had never even seen the building we were working on. The removal of the architect from the physical site of her building means that ideas must be conveyed on the page rather than worked out in the field. Not wanting our designs to get lost in translation, our office spent a great deal of time drawing details that we wouldn’t have if we were able to spend more time on the jobsite.

Finally, drawing sets have grown as a result of increasing concerns within the profession about liability. Architects often get sued, and thus carry extremely high levels of professional liability insurance. Additional work and repetition of details therefore serve to protect one’s practice. Not only is it difficult to protect the copyright of one’s designs, but it is even more difficult to ensure that we are protecting the public’s health and safety. We therefore compensate by over-detailing our designs, to ensure the structures are properly built.

The switch from manual to computer drafting, a disassociation between the architect and the jobsite, and increased concerns about liability have each contributed to the inflation of the architectural drawing set. In addition to making the job of being an architect much more laborious and difficult, the seemingly endless proliferation of detail in architectural drawing often raises the question: When do we stop drawing?

Back in Beeby’s office, for instance, I spent over a year drawing one 16’x16’ room.  I drew every detail of the room, going so far as to provide instructions for the woodworker on how to cut his knives for the wainscoting. On the opposite end of the spectrum, when I worked in Joseph Pell Lombardi’s office in New York City, the architects felt that most major decisions would best be made when the existing condition of the structure was exposed and carefully reviewed on the site. These are two very different, but equally valid positions on the role of architectural drawings on the jobsite. Using the Beeby method, one spends an inordinate amount of time drawing details. This can result in burning through the architect’s fees. The architect can also become myopic – one works so long on the details that one loses sight of the project as a whole. The JPL method made for a faster drawing process, but things often slowed down in the field and the overall design intent for a project can be lost in the reactive, day-to-day work on site.

Therefore, how do I model my own practice? Do I draw every detail, as in Beeby’s office? Or do I allow for existing conditions on site to dictate how the project will evolve, as done in JPL’s office? In retrospect, I have discovered that both models are valid and useful for their own merits. In the drawing set I have now produced, I have conveyed my design intent with as many drawings as I felt necessary without burning through time. However, I am sure there will still be many conditions for which I cannot prepare once construction commences. I have spent time detailing the rooms which I am most concerned about, while noting in the drawings potential issues that may arise. For example, one of the first notes on the drawings calls for a core-sample of the existing roof to be done. Once the layers of the roof are exposed, I can then detail how I want to repair/replace the roof. In doing so, I am a bit Beeby, and a bit Lombardi. Have license, will draw.

My bid-set went out this week. Wish me luck.

Gordon Bunshaft in front of the Beinecke

 Inside Beinecke

Inside Beinecke

 Wainscot Detail of St. Chrysostom's

Wainscot Detail of St. Chrysostom's

 Basement, First, Second, & Third Floor Plans of 1304 N. Ritchie Court

Basement, First, Second, & Third Floor Plans of 1304 N. Ritchie Court