I use the theoretical framework of distributed cognition to develop a repeatable procedure that non-artists can follow to model 3-D human figures—manikins—on a computer, which are sufficient for prototyping. The approach was not to train artists to be expert human figure modelers, but rather to derive a distribution of abilities across person and computer such that the system of person-in-interaction-with-technology exhibited expertise. These abilities were discovered though an analysis of two equivalent functional systems: figure modeling on a computer and figure drawing on paper. I report a test of the procedure on a group of non-artists, which yielded a high success rate. This research contributes to our understanding of applying distributed cognition to the design of instructional procedures, and to our understanding of the sciences of the artificial.