A signature to convince

Index

signature

Spiro vector redrawn from Le Bulletin de la Section d’Histoire des Usines Renault (1985)1

With any new palette you encounter, try to execute your signature. If it comes out looking half-way decent, then the pencil is being splined. That means, your stroke is sampled and only smooth interpolations between some of the pixels you visit are drawn. This is an expensive feature and most palettes nowadays skip it. The result is a tool that cannot really be used for free hand drawing. It's not that the artist suffers from some kind of palsy.
George Francis, Essay on Drawing Palettes (2011)2

A signature is a gesture of the body. Since its physical action is intimately linked to the movement of muscles and flesh, it offers an opportunity to question the very nature of the relationship between vectorial curves and the human body. In his “Essay on Drawing Palettes”, George Francis suggests using ones signature, supposedly our most personal and defining gesture, as a probe to measure the level of sophistication and fidelity of a tracing tool.

Something similar to this happened forty years ago in an operation of translations. Trying to convince the managers of car manufacturer Renault about the potential of the tools that Pierre Bezier and his team were developing for the production of car bodywork, they reproduced the signature of the treasurer of la Banque de France, which was at the time printed on all French bank notes.

Simultanously, mathematician Paul de Casteljau3 was working for Citroën on the same issue. Through the ‘culture du secret’ operating at Citroën, and possibly the natural discretion of de Casteljau, we don't know if in his lab he had performed the same kind of hacking already a few years earlier when he was working with Bernstein polynomials4, a set of algorithms that by then was already 50 years old.

Contrasting with typography or writing, these examples show gestures as the purest expressions of shape freedom. When copies like these signature gestures are used to prove that a technology can be more than a good imitation, that it can be trusted as truly lossless, it then has the ability to be accepted as an original.


  1. Le Bulletin de la Section d’Histoire des Usines Renault, tome 5, juin 1985, n°30, 272-283. https://web.archive.org/web/20040703214430/http://www.gmm.insa-tlse.fr/~rabut/bezier 

  2. George Francis, "Essay on Drawing Palettes" in: LGRU Reader (Brussels: Constant, 2010) http://reader.lgru.net/texts/essay-on-drawing-palettes/ 

  3. Paul de Casteljau https://en.wikipedia.org/wiki/Paul_de_Casteljau 

  4. Bernstein polynomials https://en.wikipedia.org/wiki/Bernstein_polynomial