The long-term objectives combine basic and applied questions relevant to language and perception. Specifically, the objectives are: (1) to provide a more complete analysis of the conditions producing behavior that seems to emerge without teaching; (2) to determine if procedures which have produced classes of equivalent stimuli for auditory and visual stimuli will produce such classes when three modalities (auditory, visual, and tactile) are involved, i.e., to relate work on stimulus classes to the general topic of perception; (3) to analyze cross-modal transfer for auditory, visual, and tactile stimuli; and (4) to determine if handicapped subjects differ in their ability to form classes of multimodal stimuli. The results of these studies will have important implications for educating both normal and handicapped children. The procedures could be used to determine whether or not children who do not learn well from ordinary teaching methods fail to form stimulus classes for particular modalities or combinations thereof. Teachers might then remedy or circumvent specific deficits. The results may also clarify current debates over the efficacy of multisensory teaching techniques. Perhaps multimodal teaching that produces stimulus classes is effective, whereas techniques which do not produce stimulus classes are not. The methodology is a matching-to-sample procedure that has been used productively for auditory and visual stimuli. Subjects are taught to select comparison stimuli when presented with sample stimuli. They are then tested to determine if any new performances have emerged without teaching them. A child taught to select pictures in response to their spoken names (e.g., choose the picture of a hat when the experimenter says """"""""hat"""""""", and taught to select printed words in response to their spoken names (choose h-a-t when the experimenter says """"""""hat"""""""") will be able to perform untaught relationships with the stimuli. He or she will be able to correctly match the pictures and words and will be able to name both the pictures and the words. More generally, if a subject learns to match two sets of stimuli, A and B, and then learns to match sets A and C, matching sets B and C automatically emerges. Only a few relationships are taught, many more emerge. The members of the stimulus sets form stimulus classes of equivalent, interchangeable stimuli. The proposed research will expand this paradigm to three modalities in order to address the questions noted above.
