Semantic relations among equivalent stimuli were evaluated by lexical decision tasks and a semantic differential. Two training structures, known as "one-to-many" and "many-to-one", were employed in order to generate three classes of equivalent stimuli involving five sets of stimuli: A, B, C, D and E. Stimuli for set B were pictures of faces expressing emotions whereas stimuli for sets A, C, D and E were abstract pictures. For one group of participants, the One-to-many Group, the directly trained relations were AB, BC, CD and CE; for the other experimental group, the Many-to-one Group, the directly trained relations were DC, EC, CB and BA. Two pictures were then presented in sequence as "prime" and "target" in trials in which participants responded whether they recognized or not the target. Responses were faster when prime and target belonged to the same equivalent class, reproducing the semantic priming effect. Abstract pictures were then evaluated through the semantic differential. Evaluations by the One-to-many Group were similar to the evaluations of the faces made by a control group, but those made by the Many-to-one Group were not. The data suggest that the transfer of meaning among equivalent stimuli may depend on the relations which are established by the participants.
Stimuli equivalence; Transfer of meaning; Semantic priming; Semantic differential
