1960 Volume 10 Pages 43-54
Using a modified Skinner box, seventy-one experimentally naive pigeons were trained on a single monochromatic light stimulus (480, 505, 530, 555, 580, 605, or 630 mμ). One hundred continuous reinforcements for two days and thirty two one minute fixed interval reinforcements for following two days were given to each S. On the fifth day, the generalization test was carried out under extinction. Each S was tested on only one of seven test stimuli (480, 505, 530, 555, 580, 605, and 630 mμ). The extinction criterion was no response for five minutes. Total number of responses up to this criterion was taken as the response measure. The following results were obtained :
1) There was the tendency that the pigeon responds more persistently to longer wave length than to shorter one.
2) In extinction period the response was emitted to all test stimuli.Taking 1) into consideration, the slope of generalization gradient seemed to be nearly flat.
3) There was no difference between extinction process under original stimulus and that under generalized one.
It seemed that the discrepancy between the slope of generalization gradient of the present experiment and those of other studies (6) (9) (10) is due to the difference of experimental operations. Namely, in the present experiment the pecking-key was continuously illuminated by the original stimulus throughout training period, while in the other experiments “blackout” was used intermittently.