Memory for Problem Solving: Comparative Studies in Attention, Working and Long-term Memory

Living in a complex and dynamically changing environment requires accurate and timely behavioural responses that allow for adapting to such changes. Well-developed perceptual systems provide a continuous flow of abundant and up-to-date information on the changes in the environment, and, thereby, allow for tailoring behavioural responses accordingly. However, issuing any behavioral response would not be possible, if it was not for information processing capacities that link one’s perception and action. Because the information processing capacities of humans and non-human animals are always limited, the available information must be sorted, selected and prioritised at all steps of information processing. The steps of information processing have different names, corresponding to their function. The processes, that support attending to and acquiring the information, belong to attention. The processes, that support working on the acquired input from the environment and comparing it with the information acquired in the past, belong to working memory. And finally, the processes, that supply and update the information acquired in the past for the use in the long term, belong to long-term memory. Attention, working and long-term memory work in concert to harness the flow of information, and to support rapid and flexible adaptation to the changes in the environment. This thesis comprises four empirical papers, in which some aspects of attention, working and long-term memory are compared across five species: the common raven (Corvus corax), the Goffin’s cockatoo (Cacatua goffiniana), the Sumatran orangutan (Pongo abelii), the common chimpanzee (Pan troglodytes) and the human (Homo sapiens sapiens). In the first two studies, chimpanzees, an orangutan, and Goffin’s cockatoos are tested in a novel experimental setup that allows for measuring long-term memory flexibility. Arguably, such flexibility allows for drawing on overlapping past experiences to solve novel problems, even when these experiences conflict with one another. The results suggest that great apes and at least some Goffin’s cockatoos can overcome such conflicts and rely on less salient yet relevant rather than more salient yet irrelevant features of overlapping experiences. In the third study, ravens and humans are tested in a series of novel working memory tasks, completed individually or with a competing partner. Ravens perform better in the social than in the individual tasks, while the opposite is true for humans. Interestingly, ravens seem to handle the increasing difficulty of the task by keeping a steady success rate, perhaps revealing a flexible adaptation to varying demands on working memory in ecological conditions. In the fourth and final study, ravens and humans are tested in another experimental setup, which requires attending to a series of objects. Ravens’ gazes to the objects are half as short as humans’, suggesting a higher speed of perception, and perhaps of cognitive processing