Emergent Untrained Combined Functions in Equivalence Classes

Three experiments examined whether each of the functions controlled by different stimuli in an equivalence class would become integrated in a test for transfer of function. In Experiment 1, a paper-and-pencil protocol was used to establish two 3-member equivalence classes (A1, B1, C1 & A2, B2, C2). Nonsense syllables were used for all stimuli except for B1 and C1 which were pictures of blue and green discs, respectively. Participants then were presented with five boxes containing discs of five different colors, including blue and green. Participants were instructed to respond as they felt appropriate. Across all participants varying numbers of blue and green discs were placed on B1 and C1, respectively. Responding at A1 included combinations of blue and green discs by seven out of eight participants. These results were replicated after the classes were extended to include D and E stimuli. In Experiment 2, the overall procedure was repeated, however, there was a pretraining condition in which participants were instructed to select and place discs of their choosing on each of B1 and C1. In general, responding within both classes was consistent with results in Experiment 1. Two participants placed similar combinations of colored discs at A1, D1, and E1 but retained the single colors at B1 and C1. Experiment 3 changed the B1 and C1 stimuli to nonsense syllables and trained the function to placing colored discs at B1 and C1. It was found that results were similar to those found in both Experiments 1 and 2. Results are discussed in the context of procedures used to investigate the emergence of novel behavior.