Backwriting: Crossmodal Equivalence of Shape and Slant as a Function of Age

Abstract

The equivalence of tactilely explored and visually perceived objects is well known (Gibson, 1969, 215-231). Similar tactile-visual equivalence occurs with passive tactile input, as in backwriting (Neisser, 1967, 46-85). It is difficult to find systematic explorations of tactile-visual equivalence with the tactile standard passively presented. The present study initiated such systematic exploration. A younger (2to 4-yr.-old, n = 8 ) and an older (7to 13-yr.-old, n = 11) group of children were each presented, in varying order, matching-to-sample tasks requiring discrimination between mirror-image oblique lines oriented 45 to the right vs left or between a triangle vs circle. The stimulus chosen from a given pair to be the standard was presented either visually, or by tactile drawing on a cloth template placed on the back. Tactile and visual stimuli were identical in size, shape, and orientation. Sequential stimulus presentation in the tactile mode by drawing on the back was used instead of simply pressing the stimulus on S's back in order to preserve the analogy between our passive presentation and the more commonly used active sequential search procedure. Matching was always done by selecting one of the visually presented pair. The standard was varied randomly from trial to trial. For the oblique line condition, stimulus locus was varied; either the right, left or middle of the back being used on a given trial. Further, a guess rate was obtained for each stimulus pair by eliminating presentation of the standard on certain trials. The older group made no errors when the standard was visually presented and never made more than 8% errors with a tactile standard. Their guess-rates were at chance levels. The younger group made no errors when the visually presented standard was from the triangle-circle pair but performed at only 74% correct when the visually presented standard was from the mirror-image pair. A Mann-Whitney U test showed this difference as a function of stimulus type to be significant ( p < .02). Younger Ss performed at levels significantly below (largest p < .01) those of older Ss under all conditions except the above mentioned visual-visual, triangle-circle matching task. However, the younger group performed above (largest fi < .03) the guess-rate on all but two of the tasks, viz., tactile-visual matching of oblique lines when presented to the left and when presented to the middle of the back. The latter tasks resulted in high inter-subject variability, some Ss performing above chance and others choosing the mirror-image of the standard. Thus, the well-known tendency for young children to have difficulty discriminating visually presented mirror-image stimuli also occurs when mirror-image stimuli are drawn on S's back. The development of ability to discriminate passively received tactile stimuli parallels that previously reported for tactile stimuli which were actively explored (Rudel & Teuber, 1964). In light of the great emphasis which has been placed on the role of active exploratory movement in the development of tactile perception (e.g., Rudel & Teuber, 1964), the above-chance performances of Ss in the younger group are particularly interesting.