Cubes are difficult things to draw

Abstract

The effects of general or specific verbal instructions on the production of a visual realistic drawing were investigated. Children aged 7 and 12 years and adults were asked to draw a cube placed in front of them on a table so that the front and top faces only could be seen. The two groups of children produced more two‐segment drawings with the specific than with the general instruction. The adults drew more two‐segment forms in both conditions. Most of the 7‐year‐olds' drawings were rectilinear. Most of the 12‐year‐olds' drawings contained obliques, and the parallel oblique form was particularly dominant. Although the adults, in contrast to the children, converged the ‘depth’ lines into the distance, only a minority of them did this; a sizable proportion of adults produced rectilinear forms. The data suggest that young children are more capable of visual realism and adults are less capable than the traditional theory of pictorial representation predicts.

Foreshortening and the perception of parallel projections

“…One explanationis that parallel projection drawings may be successful to the extent that they entail properties on the picture surface that concur with knowledge of3-D object structure (Arnheim, 1974;Cox, 1986;Duthie, 1985;Freeman, 1986). Consider the cube drawings in Figure I.…”

“…Research in drawing development also indicates that parallel projections are widespread. Many older children (10 years and above) and adults produce a square with parallel oblique lines when asked to draw a cube (Arnheim, 1974;Caron-Pargue, 1985;Cox, 1986;Freeman, 1980Freeman, , 1986Willats, 1981Willats, , 1984. But the widespread use of parallel projection to represent three dimensions suggests problems for theories which contend that the perception of depth in pictures follows the laws of polar projective geometry.…”

“…We examined examples published by theorists writing about cube drawings, especially those from theorists writing about a knowledge-based account (Arnheim, 1974;Caron-Pargue, 1985;Cox, 1986;Deregowski & Strang, 1986;Freeman, 1986Freeman, , 1987Mitchelmore, 1987;Moore, 1987;Phillips, Hobbs, & Pratt, 1978;Willats, 1984). Only I of 10 cube figures, that ofWillats (1984), had the front-line-to-side-line ratio of 1:1 that a knowledge-based view would suggest.…”

“…Parallel projection draw- ings are not produced according to polar projection rules. Indeed, explanations for the success of parallel projection drawings are often couched in terms of the observer's knowledge about the structure and features of the 3-D object (Cox, 1986;Duthie, 1985;Freeman, 1986Freeman, , 1987, rather than the optic projection from the picture. In the face of this knowledge-based view, we suggest that the perception of parallel projection drawings of cubes is related to how well they reenact what is projected from the 3-D object to a specific vantage point.…”

“…Arnheim (1974) describes this kind of drawing as looking "cube-like" because its properties on the picture surface preserve parallelism, an "essential objective property" of a 3-D cube. Cox (1986) asked 26 undergraduates to draw a cube from memory. All produced a square-with-obliques version.…”