Recognition gradients for random shapes following distinctive or equivalent verbal association training.

The effect of distinctive and equivalent verbal label training on a subsequent test of recognition memory for random shapes were assessed. Shape continua, systematically relating the labeled shapes, were used in the memory test so that gradients of recognition memory were obtained. Distinctive-nonrepresentative-label training produced symmetrical recognition gradients with a single mode at the correct target shape. Equivalent-label training produced symmetrical gradients that were as steep as the distinctive-label gradients, but the mode of the equivalent-label gradients was shifted to a distractor shape that varied from the target in the direction of the equivalent-label shape. The results offered support to a Gestalt/configurational account of the effects of verbal labels on memory for form, but the data were also consistent with an extension of Ellis' conceptual coding hypothesis. It was concluded that, rather than being a competing alternative account, the conceptual coding hypothesis may be viewed as an updated, more analytic statement of the older Gestalt view.