Chain-Type Drawings: A Further Note

Abstract

Werner (1957) and Hudson (1962) noted that chain-type drawings which show more of a depicted object than is visible from any point of view and may, e.g., show a cube as a series of squares each representing one of the faces are, in certain cultures, preferred to orthogonal drawings which represent objects as seen from a single point. Deregowski (1969, in press) using Hudson's drawings did not detect such preference and did not find the anticipated superiority of orthogonal drawings as an identification cue. Ic was thought that the essentially bird's eye view of the elephants in Hudson's drawings might be responsible for this and the experiment was repeated using the following drawings of an elephant: front view, side view, chain-type drawing consisting of a front view and rwo side views. A set of 18 assorted animal models was used in the identification task. All Ss used were rural Zambian women of little or no schooling. They were either presented with the three possible pairings of the drawings and asked to choose a berter drawing, an elephant model being on display (N = 27), or given one of the drawings and asked to point to the depicted animal in the array of 18 models (three groups of N = 20). Three Ss when choosing drawings made circular responses and so were excluded. The remainder of choices were analyzed (rank totals: front view 58, side view 42, chain-rype drawing 44; Friedman's Xa = 7.3, p > 0.05). This suggesrs that there ' is no difference between the side view and the chain-type drawings, both of which are preferred over the front view. The frequencies (out of 20) with which misidentifications occurred were: front view 0, side view 4, chain-type drawing 7. The frequencg distributions are significantly different only in the case of the two extreme pictures, the front view and the chain-type drawing (X' = 6.2, p = 0.025). The side view occupies an intermediate position and does not differ significantly from either of the others. Hence, the postulated paradox that the preferred chain-type drawing leads less often to identification than the nonpreferred drawing can only be said to hold if one contrasts the chain-type drawing with the front view drawing and not with the side view drawing.