Abstract
The current experiment trained pigeons on a computerised object permanence task, using consecutive visible displacements and thus assessing Stage 5 object permanence. The experiment consisted of four phases where white dot stimuli and a varying number of grey square occluders (doors) were presented, dependent on the phase. Phase 1 (one dot, one door) and Phase 2 (two dots, two doors) were shaping phases to teach the birds to peck the stimuli. Phase 3 (two dots, three doors) and Phase 4 (two dots, four doors) were experimental phases, where two doors occluded the two dots, and additional doors covered empty response holes. Birds were required to peck the doors that occluded the dots, while not pecking doors that occluded empty response holes. Birds’ performance across trials was found to differ depending on the configuration the two dots formed on screen. Repeated training was given to try break the configuration biases, and eventually four birds reached criterion on Phase 3, then three reached criterion on Phase 4. Though birds did succeed in completing the task, analysis of the results suggests that birds were responding by impulsively pecking doors based on accessibility and visibility without regard for the task, and that success was temporarily conditioned and still governed by configuration bias. The importance of considering behavioural explanations for success is highlighted in the current study as, though birds did learn the task, examination of the results does not support conceptual understanding of Stage 5 object permanence in pigeons.