Stimulus Equivalence: A Laboratory Artefact or the Heart of Language?

This thesis surveys some of the implications of the presented collection of publications, all of which address the phenomenon of stimulus equivalence. Stimulus equivalence (SE) is first operationally defined in terms of Sidman’s trio of criteria: symmetry, transitivity, and reflexivity (Sidman & Tailby, 1982). Then some of its main features – the phenomenon of delayed emergence, the effects of nodes, and the influence of properties of the stimuli used, including nameability and meaningfulness - as exemplified in the empirical studies presented, are evaluated in the light of recent literature. The variety of ways in which SE classes may be formed are described, and the question of when SE relations take effect – during the training of the base relations, or subsequently, or only in the course of unreinforced testing for derived relations – is discussed. The effects of nodal number in multi-nodal linear classes are examined and contrasted with those in serial learning. Some methods of chronometric and protocol analysis, as developed in some of the collected studies, are described, and the outlines of a model of SE class formation they might help to form is presented. The role of naming and of language in general is discussed as a sufficient route to SE class formation, but not one that is perhaps necessary for its laboratory demonstration. The role of SE in the opposite direction, in the ontogeny and phylogeny of language, is considered. Here, besides learned speculation, more empirical studies are awaited, of children, and some new developments in comparative cognition. Highlights are described of the few brain imaging studies implicating SE, following the pioneering empirical study and the earlier review in the presented collection. The survey ends by again extolling the relevance of Tinbergen’s (1963) four levels of explanation in behavioural biology to see the phenomena of SE in appropriate perspective