Alexander’s generating systems rely on a notion of “rules about the way… parts may be combined.” A logical starting point for designers working generatively is to delineate those rules, but that is hardly ever done in a void. A process of rulemaking often begins analytically, almost scientifically, observing the functioning of an existing system in order to derive elementary objects and rules. In 2- and 3-dimensional formal design, this has been explored in the work of computational design researchers George Stiny and James Gips and their notion of shape grammars.1 Shape grammars are abstractions of geometries, described by a set of rules that define transformations. A simple ruleset might only encode a few recognized shapes and basic transformations, such as scaling or rotating. A more sophisticated ruleset, such as Stiny and William J. Mitchell’s 1978 “Palladian grammar,”2 might be capable of generating complex, recognizable forms, such as floor plans in the style of 16th century Italian architect Andrea Palladio. Derived from analysis of a larger body of work, Stiny and Mitchell’s Palladian grammar contains the potential for new works that would be recognized as members of the same stylistic family as the original villas. More recently, in his 2006 book Shape: Talking about Seeing and Doing,3 Stiny has argued for designing shape grammars as a new pedagogy: “Creative design can be taught like language and mathematics in school, with examples, rules, and practice and the opportunity to experiment freely.”
Many of the practitioners I interviewed use rulemaking techniques in their work. Paul, an architect and technologist, puts the tools of his day job to work toward more visually aesthetic ends as experimental artworks. Like Stiny, he argues that a rule-based logic can be used as an expressive medium in design and art. Describing a specific piece, Paul says, “The logic [here is] starting with a cube and then randomly subtracting cubes from that cube. Starting at the lower-left corner and moving all the way up, each one uses a different set of random cubes… I was interested in testing it out, seeing what graphic outcomes it gave me.”4 Through the iteration of a subtractive rule, and by varying parameters (the size and position of the subtraction), Paul’s system is theoretically able to generate an infinite number of possible outcomes. Another interview subject, Anna, a digital artist and game designer, describes a project, a “generative experiment” that spun off of a larger game they were working on, that took the form of a Twitterbot that periodically posts generative drawings of winged animals. Anna notes, “The interesting thing about that project is the way it works, which is very lo-fi in some ways. It draws each [animal] much like a person would draw something, in that it places individual pixels by rulesets to generate these patterns and these shifts.”5 Anna's analytical ruleset is derived from the visual language of animal life, coupled with a procedure modeled after human drawing. The space of possibilities formed by Anna's ruleset, while also infinite, is intuitively larger than the space of Paul’s cube-based artwork — the generated animals vary in size, color, shape, texture, and pattern as opposed to simply geometric form. In both cases, however, the number of possible outcomes exceeds the ability of a single designer to ever observe. These examples each use pseudo-random selection to choose from generated possibilities. In aggregate, with enough sampling, this will eventually provide a qualitative sense of the possibility space: A rough feel for its size and the variations it permits. However, there are other, more directed ways of traversing a generative possibility space.
1. Stiny, George, and James Gips. "Shape Grammars and the Generative Specification of Painting and Sculpture." In IFIP Congress (2), vol. 2, no. 3. 1971. ↩
2. Stiny, George, and William J. Mitchell. "The Palladian Grammar." In Environment and planning B: Planning and design 5, no. 1 (1978): 5-18. ↩
3. Stiny, George. Shape: Talking about Seeing and Doing, 2006. ↩
4. Skype interview with Paul, August 3, 2017. ↩
5. Skype interview with Anna, July 18, 2017. ↩