Dogs' use of the solidity principle: revisited

Abstract

A wealth of comparative data has been accumulated over the past decades on how animals acquire and use information about the physical world. Domestic dogs have typically performed comparably poorly in physical cognition tasks, though in a recent study Kundey et al. (Anim Cogn 13:497-505, 2010) challenged this view and concluded that dogs understand that objects cannot pass through solid barriers. However, the eight subjects in the study of Kundey et al. may have solved the task with the help of perceptual cues, which had not been controlled for. Here, we tested dogs with a similar task that excluded these cues. In addition, unlike the set-up of Kundey et al., our set-up allowed the subjects to observe the effect of the solid barrier. Nevertheless, all 28 subjects failed to solve this task spontaneously and showed no evidence of learning across 50 trials. Our results therefore call into question the earlier suggestion that dogs have, or can acquire, an understanding of the solidity principle.

Fig. 1. Photographic depiction of two symmetrical halves of the experimental apparatus from dog’s point of view (a) and schematic layout of the testing room with the apparatus marked in grey (b).

a shows the state of the apparatus after an incorrect choice of the blocked condition. Circles in b give start positions for experimenter 1 and 2 respectively. For each trial, experimenter 1 approached the apparatus from the left or the right, placed the occluder between himself and the dog and, after baiting the apparatus, returned the occluder to its position before returning to his start position behind the partition.

Fig. 2. Percent correct choices for the five sessions of the blocked condition.

Numbers in parentheses give sample sizes (one subject completed only four sessions). The dashed line indicates chance level. Data are displayed as mean and standard error.