Abstract:
The present thesis aimed to test Sidman’s (2000) account of the origins of stimulus
equivalence relations, which states that equivalence is a direct outcome of reinforcement
contingencies. Experiments 1 and 2 tested Sidman’s predictions that a reinforcer common
to two distinct equivalence classes should not become a member of those two classes,
while a reinforcer specific to one class should become a member of that class. Class-
specific reinforcers for two three-member classes (i.e., A1B1C1-R/, and A2B2C2-B2) and
a common reinforcer for another two three-member classes (i.e., A3B3C3-F3, and
A4B4C4-F3) were arranged. Experiments 1 and 2 confirmed the prediction that class-
specific reinforcers can become members of equivalence classes. However, this was not
shown by all subjects. Also contrary to Sidman’s predictions, Experiments 1 and 2 found
that a reinforcer common to two distinct equivalence classes can become a member of
those two classes. Furthermore, subjects in Experiment 2 showed inter-class matching
relations between Class 3 and Class 4 based on equivalences involving R3 where, for
example, they matched Class 4 comparisons to Class 3 samples in the absence of Class 3
comparisons. Yet Class 3 and 4 baseline relations continued to be maintained. Therefore,
Class 3 and 4 stimuli were both equivalent (on inter-class matching trials) and not
equivalent (on baseline trials) which, according to Sidman’s account, is impossible.
Experiments 3 and 4 investigated whether equivalence relations that directly
conflicted with the reinforcement contingencies could emerge. Class-specific incorrect
comparison stimuli were arranged, and relations between contingency-related stimuli and
their class-specific incorrect comparisons were tested. For example, Comparison Bi was
always presented with class-specific incorrect comparisons G1 and Hl. The selection of
BI was reinforced following Samples Al and Cl, but selections of G1 and Hl were never
reinforced. Despite contingencies of extinction for matching relations involving Al, Cl,
Gl, and Hl, subjects consistently matched Al and Cl to G1 and Hl (and the reverse
relations). These outcomes can never be predicted from an account that views equivalence
as having its origins in reinforcement contingencies, and show that Sidman’s account is
implausible: stimulus equivalence does not result from reinforcement contingencies.