(Presented with all the other Commentaries on the Horne and Lowe article)
[No abstract; these are the first two paragraphs.]
In their impressive article, Horne and Lowe (1996) reevaluate the theoretical import of stimulus equivalence. Their critical analysis seems largely correct. At one place, however, Horne and Lowe attribute the failure of equivalence explanations to the "artificial" character of match-to-sample procedures (Horne & Lowe, p. 238). In our view, the fundamental problem with the equivalence framework stems less from its reliance on artificial situations than from a lack of clear theoretical principles (cf. Harzem, 1995[cf11]1[fn1]). Were such principles available, the artificiality of the procedures involved would not constitute a significant obstacle; after all, most behavioral concepts, including the concept of reinforcement used by Horne and Lowe, have been derived from experimental situations no less artificial than match to sample. The Naming Account
What processes, then, underlie stimulus equivalence? Horne and Lowe develop various possibilities, but not all of them seem equally important or relevant, and some are only partly specified. In the section entitled Other Verbal Behavior and Stimulus Equivalence, for example (Horne & Lowe, pp. 221—222), complex utterances such as "circle goes with the open triangle" are said to determine performance on equivalence tests. However, the utterances in question could not be emitted unless equivalence outcomes had already emerged; emitting "green means up" after up ;ra green forward pairings, for example (Horne & Lowe, p. 221), assumes some form of symmetry in the first place because the names "up" and "green" appear in the utterance in a reversed order.
Address correspondence to François Tonneau, Centro de Estudios e Investigaciones en Comportamiento, Universidad de Guadalajara, Avenida 12 de Diciembre 204, Col. Chapalita, CP 45030, Guadalajara, Jalisco, Mexico (E-mail: ftonneau@udgserv.cencar.udg.mx).