The case study of Nadia, an autistic child, documents the phenomenally accurate drawings she made. Nadia's precocious drawing capacity raises questions about the drawing hypothesis adhered to by F. Goodenough and D. B. Harris and Gestalt theories of perception and representation. The argument is made that Nadia drew realistically because she lacked (or did not associate) conceptualization with the activity of drawing. Recent research in drawing shows that conceptual information about the object being drawn can interfere with representational accuracy. Nadia's intuitive grasp of the relationship between a pencil line on paper and the edge of an object in space strongly corroborates the argument of those who claim that outline depiction is the least convention-bound of all the systems of graphic representation. Thus, it is no coincidence that an asocial child such as Nadia should have hit on a method of representation that relies so little on conventional imagery. (Author/RL)