We tested horses on the impossible task paradigm, a drawback affecting the decision-making process in animals. We used the direction of the horse’s ear cup as an indicator of its visual attention in terms of visual selective attention when both ears were directed at the same target and the visual differential attention when the ears were directed differentially to the persons and to the experimental tools. We aimed to evaluate whether the latter behavior could be considered a possible support to solve the problem. The visual differential attention was the most frequent behavior when the resource was unreachable, which supports the view that this gesture could be linked to a request for help from humans to find the solution to the task. Our procedure proved to be a useful way to understand how horses try to attract human attention when they are in a restricted environment, a typical situation for horses living in stables.
In order to explore the decision-making processes of horses, we designed an impossible task paradigm aimed at causing an expectancy violation in horses. Our goals were to verify whether this paradigm is effective in horses by analyzing their motivation in trying to solve the task and the mode of the potential helping request in such a context. In the first experiment, 30 horses were subjected to three consecutive conditions: no food condition where two persons were positioned at either side of a table in front of the stall, solvable condition when a researcher placed a reachable reward on the table, and the impossible condition when the food was placed farther away and was unreachable by the horse. Eighteen horses were used in the second experiment with similar solvable and impossible conditions but in the absence of people. We measured the direction of the horse’s ear cup as an indicator of its visual attention in terms of visual selective attention (VSA) when both ears were directed at the same target and the visual differential attention (VDA) when the ears were directed differentially to the persons and to the table. We also included tactile interaction toward table and people, the olfactory exploration of the table, and the frustration behaviors in the ethogram. In the first experiment, the VDA was the most frequent behavior following the expectancy violation. In the second experiment, horses showed the VDA behavior mostly when people and the unreachable resource were present at the same time. We speculate that the VDA could be a referential gesture aimed to link the solution of the task to the people, as a request for help.