Document Type



In Experiment 1, fourteen 5- to 6-year-old children were taught to speak name N1 in response to visual stimulus A 1, and name N2 in response to visual stimulus A2. Thereafter, 7 children were taught to choose A1 when N1 was spoken and A2 when N2 was spoken; the others to choose A2 when N 1 was spoken and A 1 when N2 was spoken. For both groups, this was followed by a purely visual matching-to-sample task with other stimuli, and by a test of symmetric responding with those other stimuli. Symmetric responding predominated on this test; no differences between groups were found. In Experiments 2 and 3, the same was found after generalization of symmetric vs. asymmetric responding to other spoken names had been obtained. These findings suggest that purely visual symmetric matching to sample does not depend exclusively on preexperimental contingencies involving spoken names.