The present experiments investigated the feature-positive effect with pigeons using spatial positional cue. A horizontal stimulus array composed of seven translucent plexiglasses, square response keys (1.5 × 1.5 cm) was mounted on the front panel. The position of a lit key served as a stimulus (Fig. 1).
In Experiment 1, stimulus control of the positional cue was ascertained using successive discrimination training and generalization procedure (Figs. 2 and 3).
In Experiment 2, for half of the pigeons, responses to displays containing the distinctive feature (“D”) were reinforced (feature-positive, FP), and for the remaining pigeons responses to displays without “D” were reinforced (featurenegative, FN). A particular position of a lit key served as the “D”, and the other position of a lit key served as the common feature (“C”). The position of a lit key was denoted as “1”, “2”, ...., “7” from left to right of the front panel. The “D” was either “1”, “4”, or “7”. Each subject was assigned to one of the three “D” s. A discrimination was trained between a single stimulus and a pair of stimului (Fig. 4), until the discrimination criterion had been attained or 30 training sessions had passed.
On the day following discrimination training completed, a generalization test in extinction was performed for all pigeons. Two out of seven keys were lit simultaneously. The combination of stimuli and the order of presentation of the combination were randomized.
Discrimination performance revealed that FP-subjects localized responses on the “D” and learned the discrimination easily, whearas FN-subjects localized responses on the “C” and did not learn the discrimination (Figs. 5 and 6). Thus, feature-positive effect was observed using positional cue.
All FP-subjects showed ordinary convex excitatory generalization gradients arround the position of the “D”, but only one pigeon in FN condition revealed concave inhibitory gradient. The “D” for this pigeon was “4”, and the discrimination ratio of the pigeon was about. 70 on the final session of the discrimination training. Other FN-subjects showed flat gradients.
In Experiment 3, only the stimulus position of “3”, “4”, and “5” were used. This display was compact and was suggested to show a better stimulus control in Experiment 2. A discrimination training and a generalization test were performed as same as Experiment 2 except for the number of stimulus combination. The discrimination performance of FN-subjects were better than FN-subjects in Experiment 2. Generalization gradients were concave for all FN-subjects.
From the results of Experiments 2 and 3, the followings were concluded : (1) The feature-positive effect was observed using only positional cue in pigeons. (2) Proximity of “D” with “C” is an important variable for the acquisition of an inhibitory stimulus control of “D” in FN condition.
抄録全体を表示