Errors in pigeons\u27 memory for number: Effects of ITI and DI illumination

In Experiment 1 all pigeons were trained to discriminate 2 flashes of hopper light in 4 sec from 8 flashes in 4 sec, at a 0 sec delay. One group of pigeons experienced dark lTl’s (Group Dark) while the other experienced an illuminated lTl (Group Light). All birds were then tested with dark delays of 0, 5, 10, 15, and 20 sec. Analysis showed a significant bias to respond to the comparison correct for small at extended delays, with no difference between groups. In Experiment 2 training was identical to that in Experiment l except that a 5 sec baseline delay was used. The pigeons were then tested at delays of 0, 5, 10, 15, and 20 sec. Again, analysis showed a tendency to choose the comparison correct for small at delays longer than baseline, while at delays shorter than baseline they showed a bias to respond large. No group differences were observed. In Experiment 3, an illuminated Dl was introduced for both groups. Analysis showed a reversal of the biases observed in Experiment 2. At delays longer than baseline a choose-large bias occurred, while at delays shorter than baseline a choose-small bias was observed. Again, there were no group differences. It was hypothesized that illuminating the Dl added pulse counts to the pigeons’ memory for the samples, suggesting that an event switch was not being used, but that the total amount of light in each trial was being summed. The results are clearly inconsistent with the contusion hypothesis and support a subjective shortening account of memory biases for temporal discriminations. However, whether this theory can be extended to include a subjective shrinking of number remains in question