Paired Associate Learning, Mediated Generalization, and Stimulus Equivalence

When Murray Sidman presented the modern terminology of stimulus equivalence in 1982, he did so in the context of other traditions studying derived responding. In his early papers on stimulus equivalence, Sidman explicitly address studies on mediated generalization that appear within the paired associate learning tradition. Typically, such studies involve teaching human participants interrelated lists of paired nonsense syllables, and then assessing performance on lists of pairs only indirectly related by training. These types of experimental preparations have never been particularly prevalent in behavior analysis, but rather was developed and thrived under the influence methodological S-O-R behaviorism. The number of scientific publications on paired associate learning, as well as mediated generalization, has declined from a peak in the late 1960’s and the term seem to live on today mainly in multi-method test batteries assessing the effects of, among other things, pharmacological agents, disease, and brain-injury. The number of scientific publications on stimulus equivalence, on the other hand, have increased following Sidman’s introduction of the modern terminology and has since early the 1990’s remained relatively stable. The current paper will discuss the fate of paired associate learning and mediated generalization and explore the possible relevance of this historical development for modern stimulus equivalence research.