Stimulus equivalence and arbitrarily applicable relational responding.

Subjects' responses to nonarbitrary stimulus relations of sameness, oppositeness, or difference were brought under contextual control. In the presence of the SAME context, selecting the same comparison as the sample was reinforced. In the presence of the OPPOSITE context, selecting a comparison as far from the sample as possible on the physical dimension defined by the set of comparisons was reinforced. Given the DIFFERENT context, selecting any comparison other than the sample was reinforced. Subjects were then exposed to arbitrary matching-to-sample training in the presence of these same contextual cues. Some subjects received training using the SAME and OPPOSITE contexts, others received SAME and DIFFERENT, and others received SAME, OPPOSITE, and DIFFERENT. The stimulus networks established allowed testing for a wide variety of derived relations. In two experiments it was shown that derived performances were consistent with relational responding brought to bear by the contextual cues. In contexts relevant to the relation of sameness, stimulus equivalence emerged. Other kinds of relational networks emerged in the other contexts. Arbitrarily applicable relational responding may give rise to a very wide variety of derived stimulus relations. The kinds of performances seen in stimulus equivalence do not appear to be unique.     

The phonetic effect in preschool children: The influence of overt rehearsal and verbal instructions

Preschoolers' serial matching of picture lists composed of either phonetically similar or unrelated items was compared under three types of conditions. The lower retention of phonetic lists had been used as an index of verbal mediation by Conrad (1971, 1972) who found that preschoolers did not show differential retention, and hence no verbal encoding, even when they were required to label overtly. This result was replicated and also obtained in a condition which emphasized the verbal aspects of the task. However, a significant phonetic effect and higher overall retention resulted from a condition which required overt, cumulative rehearsal. Alternative interpretations of the influence of rehearsal on the phonetic effect were offered.     

Recognition memory for active and passive sentences

Data from three experiments are reported in which recognition memory for active and passive sentences is compared. The results consistently show no difference, a result taken to invalidate syntactic memory hypotheses (e.g., the kernel plus tag and markedness hypotheses) which assume that the stored representation of passive sentences is more complex than that of active sentences. Previous reports of superior recall of active sentences are attributed to a reconstructive bias. One form of propositional hypothesis and a verbatim hypothesis are consistent with the data.This research was supported by Grant MH 19859 from NIMH to the first author, and by Rutgers University Research Council grants to both authors.     

Transfer of a conditional ordering response through conditional equivalence classes

Eight adult humans were taught conditional discriminations in a matching-to-sample format that led to the formation of two four-member equivalence classes. When subjects were taught to select one comparison stimulus from each class in a set order, they then ordered all other members of the equivalence classes without explicit training. When the ordering response itself was brought under conditional control, conditional sequencing also transferred to all other members of the two equivalence classes. When the conditional discriminations in the matching-to-sample task were brought under higher order conditional control, the eight stimulus members were arranged into four conditional equivalence classes. Both ordering and conditional ordering transferred to all members of the four conditional equivalence classes; for some subjects this occurred without a typical test for equivalence. One hundred twenty untrained sequences emerged from eight trained sequences for all subjects. Transfer of functions through equivalence classes may contribute to a behavior-analytic approach to semantics and generative grammar.