抄録
The present study examined transitive responding and a symbolic distance effect during a transitive inference task in rats. Seven three-step barriers (45 cm high in total) with different visual appearances and tactile features were used as discriminative stimuli. Four rats were trained to choose and climb over one of two barriers presented simultaneously to enter a goal box and get food rewards. If a rat climbed the wrong barrier the door connected to the goal box did not open. All four rats acquired concurrent discrimination learning of four tasks (A+B-, B+C-, C+D-, and D+E-), and showed significant transitive responding to item B on the novel BD pair in a probe test. Moreover, two rats acquired concurrent discrimination learning of six tasks (X+A-, A+B-, B+C-, C+D-, D+E-, and E+F-), and one of them showed a symbolic distance effect, in other words a better performance for novel pairs composed of more distant items (e.g., AE) than pairs composed of closer items (e.g., BD) in the transitive series (X-A-B-C-D-E-F). The symbolic distance effect is regarded as evidence that animals solve transitive inference task based on non-deductive processes, because deductive reasoning predicts better performance to item pairs composed of closer items that call for fewer premises to be combined. Therefore, findings in the present study suggest that, as is the case with birds, monkeys, apes, and human beings, rats seemed to learn a transitive task based on some non deductive processes, such as formation of liner representation of items or value transfer between items.
