Kazuo Fujita

Humans learn by observing the behaviour of others, which can lead to more efficient problem-solving than by trial-and-error learning. Numerous studies have shown that animals, other than humans, are also capable of social learning. Dogs, as humans’ closest companion animals, can learn to obtain rewards following behavioural demonstrations by humans. However, it is not known whether cats, who also live with humans, can learn how to solve problems by observing human behaviours. Three experiments were used to investigate whether cats could change their behaviour and gain rewards efficiently by observing a human demonstrating how to obtain food. In Experiment 1, a human demonstrated how to open a transparent drawer and take out the reward inside, but cats did not significantly follow the same method as the human. In Experiment 2a, a transparent tube device was used to make the operation easier for cats. However, cats were not influenced by the human behaviour. As the devices used in these experiments were transparent, meaning that the cats could see the food inside directly, they might have required strong inhibitory control. Therefore, in Experiment 2b the tube device was made opaque, and cats again observed the human demonstration. Nevertheless, the cats were not influenced by the human’s behaviour. The results of these experiments indicate a lack of social learning, including imitation, from human behaviours in cats, at least in these experimental settings with food rewards. Other than their inherent ability, cats’ biological characteristics and the experimental context may have contributed towards the negative results.

Des études récentes ont mis en évidence la capacité chez des espèces de primates non humains à détecter la réciprocité ou la non-réciprocité au sein d’un échange de biens entre tiers. Par exemple, nous avons montré que les singes capucins discriminent entre un acteur humain qui effectue un échange équitable avec un tiers et un acteur qui ne respecte pas la réciprocité de l’échange. Plus spécifiques, les singes étaient significativement moins prêts à accepter de la nourriture de l’acteur non réciproque. Autrement dit, les singes prenaient une décision binaire à s’engager avec un acteur aux dépens de l’autre. Dans une nouvelle étude, nous demandons si les singes feraient confiance aux deux acteurs de façon différente selon le degré de réciprocité montré par ces derniers. Suite à un échange équitable ou non, un des deux acteurs a commencé à transférer des morceaux de nourriture vers le singe. Dans cette situation de délai de gratification, le singe peut commencer à manger la nourriture à n’importe quel moment, mais cet acte termine le transfert ; alors pour maximiser la quantité de nourriture obtenue le singe devrait attendre pendant que la nourriture s’accumule. Les résultats ont montré que les singes étaient moins aptes à effectuer un délai de gratification lorsque la nourriture était transférée par un acteur non réciproque par rapport à son tiers. Cette tendance “impulsive” se manifestait surtout dans une condition ou la quantité totale de nourriture potentiellement disponible était inconnue du singe, c’est-à-dire dans une situation d’incertitude. Nous en concluons que la non-réciprocité d’un tiers induit la méfiance chez les singes capucins.

Memory processing in nonhuman animals has been typically tested in situations where the animals are repeatedly trained to retrieve their memory trace, such as delayed matching to sample, serial probe recognition, etc. In contrast, how they utilize incidentally formed memory traces is not well investigated except in rodents. We examined whether domestic dogs could solve an unexpected test based on a single past experience. In Experiment 1, leashed dogs were led to 4 open, baited containers and allowed to eat from 2 of them (Exposure phase). After a walk outside for more than 10 min, during which time the containers were replaced with new identical ones, the dogs were unexpectedly returned to the site and unleashed for free exploration (test phase). Eleven out of 12 dogs first visited one of the containers from which they had not eaten. In Experiment 2, two containers had food in them, one had a nonedible object, and the last one was empty. Dogs visited all 4 containers and were allowed to eat one of the food rewards in the Exposure phase. In the test phase, unleashed dogs first visited the previously baited container from which they had not eaten significantly more often than chance. These results demonstrate that in an unexpected, test dogs may retrieve “what” and “where” information about seen (now invisible) items from incidental memory formed during a single past experience.

This chapter compares amodal completion and the perception of visual illusions in birds and primates. Amodal completion concerns the perception of partly occluded figures. Visual illusions, in theory, represent cases in which perceptual systems that are adaptively suited to some situations may suffer in other situations. The examples discussed in this chapter suggest that the visual perceptual system of a particular species may be tuned to its ecological niche: where its members live, how they move, what they eat, etc.

Understanding conspecifics’ emotional states is important for managing social interactions. We examined whether capuchin monkeys modify their own behaviors in response to a conspecific’s emotional expressions. Six monkeys saw a demonstrator monkey responding emotionally to an object in a container; the subject monkeys could not see the object. The subjects reached for the container more frequently when the demonstrator showed an emotionally positive expression than when she showed an emotionally neutral or negative expression. This is the first report that New World monkeys are sensitive to the emotional valence of conspecifics’ emotional expressions. The finding is consistent with the hypothesis that monkeys can recognize emotional meanings in others’ expressions, an ability previously attributed only to humans and great apes.