March 3, 2009
"Knobs, knobs everywhere - just vary a knob to think."
Reading Ch12 of Hofstadter's Metamagical Themas which Scott dropped off in my office this morning.
Hofstadter claims that "Making variations on a theme is really the crux of creativity." (235)
Think of a concept as a machine with knobs on it. Use the knobs to interpolate and extrapolate variations from the original concept. Example 1: Rubik's cube might have a knob of dimensionality that happens to be set at 3 in original. Example2: John Gould's dreamed of turning the listener of music into a conductor through a parametric multitrack playback interface.
But where does the set of knobs come from? What can get varied? Is there a fixed, even finite set? Hofstadter argues that extrapolation needs creative analogy and that changing context and perspective produces new sets of knobs. This shifting of perspective, "nondeliberate yet nonaccidental slippage," is at the core of creative thought. Hofstadter also labels this activity as producing "subjunctive variations on a theme" - I believe this is where Aran Lunzer took his notion of "subjunctive interfaces."
The key quote on reframing: "Context contributes an unexpected quality to the knobs that are perceived on a given concept. The knobs are not displayed in a nice, neat little control panel, forevermore unchangeable. Instead, changing the context is like taking a tour around the concept, and as you get to see it from various angles, more and more of its knobs are revealed." (239) Or: "[Good knobs come from] seeing one thing as something else." (251)
Example 3: Don Knuth's Metafont, which allows typeface designers to create parametric letter definitions. This parameterization is hierarchical: there are typeface-level controls as well as letter-level controls. But H~ argues (convincingly) that all these systems only ever explore sub-spaces. One reason: different styles of letters we all agree on as instances of the letter "A" have very different underlying structure - varying continuous dimensions will never result in such fundamentally different approaches.
"One of Knuth's main thesis is that with computers, we now are in the position of being able to describe nut just a thing in itself, but how that thing would vary." (240) An open question though is how accessible this parameterization process is for creators/designers: describing a parametric space of possible designs is a very different activity from producing a point solution within that space. An avenue for future research is pointed out: given a set of examples, automatically derive the structure of the design space within which they are embedded. That sounds hard, but interesting. "If we wish to enlist computers as our partners in this venture of inventing variations on a theme [...] we have to give them the ability to spot knobs themselves, not just to accept knobs that we humans have spotted." In his words, computers should help us explore the "implicosphere" (implicit counterfactual sphere or sphere of implications) of a concept.
H~ mentions "One Book Five Ways" - a book containing, side-by-side, five different edited versions of a single manuscript. This reminded me of Raymond Queneaus's "Exercises de style" which is made up of 99 variations of a common story.
Are variations fundamentally different from new themes? No, but they tweak less obvious, hidden knobs.
P.S.: Hofstadter needs an editor.