Kristy Balliet and Casey Rehm conclude the Summer 2017 mini-series on key architectural elements with a discussion of the sketch. Baillet begins with an exploration of the qualities of a sketch, noting that its multiple meanings are related to its associations with speed, spontaneity, and incompleteness. She proposes categories and discusses examples, beginning with sketch as discovery, illustrated by Baldassare Peruzzi and Leon Battista Alberti; and sketch as gesture (Eero Saarinen’s TWA Terminal). In the category sketch for observation, Baillet demonstrates how early sketches by Hans Hollein informed works realized decades later. She continues with discussions of sketching as idea/ideology (Yona Friedman), as instruction (Erwin Wurm), as line (Greg Lynn, Lucy McRae and Bart Hess, Thomas Gromas, Ramiro Diaz-Granados, Andrew Zago), sketching for translation (the @thingsihavedrawn project on Istagram). Baillet concludes with her collaboration with Casey Rehm to explore sketching in virtual space, where gestural drawings generate renderings and physical models.
Casey Rehm begins his half of the presentation on the role of sketching with examples of sketching as a way of understanding (Leonardo da Vinci), communicating intent (Henri Labrouste). He discusses how current ideas about neuroplasticity are confirming Johannes Itten’s Bauhaus exercises to train the body and mind together (drawing with both hands simultaneously, drawing a circle in air as precisely as possible). He discusses a wide range of sketching practices in the work of Lebbeus Woods, Dwayne Oyler, Álvaro Siza, and Steven Holl. Rehm cautions that published sketches should be treated with skepticism. With Frank O. Gehry, he shifts graphic sketching to the use of sketch models, tracing them in the design process of Eric Owen Moss, and Michael Maltzan. Another kind of sketch model appears in the versioning design process of MVRDV and Greg Lynn, and Casey Reas. Rehm concludes with a demonstration of sketching by non-humans, specifically a Generative Adversarial Network (GAN) employing aerial photographs that learns how to generate convincingly realistic “aerial” views of imaginary cities.