The mirror project: A dog training method based on social learning

In the last few years much attention has been paid to social learning in dogs. In the “Do as I do” (Topál et al., 2006) protocol, behaviors are initially taught by traditional techniques. The aim of the Mirror Project is to suggest a training method where dogs reproduce observed behaviors since the beginning. The protocol is formed by 7 phases: 1) spontaneous reproduction of 6 behaviors after demonstration; 2) random reproduction of these behaviors; reproduction of: 3) other known behaviors; 4) sequences of known behaviors; 5) unknown behaviors; 6) mixed sequences; 7) different behaviors referred to the same object. Besides the training process, owners were advised to practice, in daily life, activities thought to be useful in developing dog ability to reproduce observed behaviors: doing things together, encouraging dog interest for owner behaviors, using pointing gestures. The Mirror Project was carried out on 7 dogs. Its effectiveness has been evaluated by behavioral tests. For all behaviors of each phase, the number of exact reproductions of the observed behavior, on a total of 10 (or 12) repetitions, has been counted. The test was considered successful when at least 75% of behaviors were correctly reproduced. A low number of test repetitions were needed to reach the success rate. In detail: phase 1 1.4±0.5; phase 2 1.0±0.0; phase 3 1.6±0.7; phase 4 1.5±0.6; phase 5 1.5±0.5; phase 6 1.4±0.4; phase 7 1.7±0.5. The data suggest that the Mirror Project produced good results, and that the use of social learning should be implemented in dog training.