Intraverbal stimulus-response reversibility: Fluency, familiarity effects, and implications for stimulus equivalence.

English-speaking subjects with little knowledge of the French language used a computerized flashcard program, Think Fast, to learn 16 English-French word pairs (intraverbals) by typing one word of the pair when presented with the other word as a textual stimulus. In Phase 1, half of the intraverbals were taught from French to English (FE-1) and half from English to French (EF-1). Then, in Phase 2, training continued with the stimulus and response items of each intraverbal reversed, i.e., cards previously in the FE-1 condition were trained from English to French (EF-2) and cards previously in the EF-1 condition were trained from French to English (FE-2). Feedback was provided throughout the experiment. Reversing the stimulus and response items in Phase 2 significantly reduced rate correct and accuracy scores for eight of the nine subjects. In Experiments 1 and 2, this effect was more pronounced for cards in the EF-2 condition; in Experiment 3, when the criterion for a "correct" response was more lenient, there was no consistent difference between cards in the EF-2 and FE-2 conditions. Symmetry, as indicated by accuracy scores on the first trial in Phase 2, was generally poor: eight of the nine subjects averaged only 29% correct when asked to respond to the reversed relations for the first time. We relate our paradigm and results to recent developments in fluency, verbal behavior, and stimulus equivalence, and provide directions for future research.