Dogs learn to solve the support problem based on perceptual cues

Abstract

Numerous recent studies have investigated how animals solve means-end tasks and unraveled considerable variation in strategies used by different species. Domestic dogs (Canis familiaris) have typically performed comparably poorly in physical cognition tasks, but a recent study showed that they can solve the on-off condition of the support problem, where they are confronted with two boards, one with a reward placed on it and the other with a reward placed next to it. To explore which strategies dogs use to solve this task, we first tested 37 dogs with the on-off condition tested previously and then tested subjects that passed this condition with three transfer tasks. For the contact condition, the inaccessible reward was touching the second board. For the perceptual containment condition, the inaccessible reward was surrounded on three sides by the second board, but not supported by it, whereas for the gap condition, discontinuous boards were used. Unlike in the previous study, our subjects did not perform above chance level in the initial trials of the on-off condition, but 13 subjects learned to solve it. Their performance in the transfer tasks suggests that dogs can learn to solve the support problem based on perceptual cues, that they can quickly adopt new cues when old ones become unreliable, but also that some apparently inherent preferences are hard to overcome. Our study contributes to accumulating evidence demonstrating that animals typically rely on a variety of perceptual cues to solve physical cognition tasks, without developing an understanding of the underlying causal structure.

Fig. 1

Layout of the experimental setup. Circles indicate positions of E1 and E2, respectively

Fig. 2

Illustration of the experimental apparatus in the four conditions used in the experiments when seen from above: on–off (a), contact (b), perceptual containment (c) and gap (d). The dashed line indicates the location of the wire mesh fence separating the dog’s area (bottom) from the compartment of E1. Note that for the perceptual containment condition and the gap condition, the inaccessible reward was presented on the same level above ground as the accessible reward, whereas the same was not the case for the on–off and the contact condition

Fig. 3

Percent correct choices for the first three sessions of the on–off condition (a) and for the four conditions during intermixed trials (b). OO on–off condition, CO contact condition, PC perceptual containment condition, GA gap condition. Numbers in parentheses give sample sizes. The dashed lines indicate chance level. Data are displayed as mean and standard error. Stars indicate significant deviation from chance level: *p < 0.05; **p < 0.001