Skip to main content

Tool Use As A Tool For Understanding Causal Cognition In Humans And Corvids

Final Report Summary - CAUSCOG (Tool Use As A Tool For Understanding Causal Cognition In Humans And Corvids)

The human ability to understand causality is at the core of modern civilisation. Our ability to understand ‘how things work’ has enabled us to shape the world in countless ways. Strikingly, a number of non-human animals – such as tool-using apes and corvids (members of the crow family) – perform behaviours that also suggest they understand elements of causality. However, traditionally, we have only had a cursory understanding of what human and non-human causal knowledge comprises, because studies of tool use and problem-solving have largely been descriptive rather than mechanistic.

In our research, by comparing the cognitive mechanisms that corvids, children and adults use to solve similar problems, we aimed to address several key questions such as: (1) how do humans understand the physical world and solve problems? (2) what other ways of understanding causality and problem solving has evolution produced? and (3) what selective pressures lead to the evolution of causal cognition? The overarching aim of our research programme was to create a coherent, experimentally-tested, theoretical framework of the cognitive mechanisms underlying causal knowledge in corvids and humans, both young and adult.

Thanks to this ERC grant, we established a hand-raised population of corvids and studied their development of casual reasoning capacities. We formed connections with schools across Cambridgeshire and beyond to assess the development of causal knowledge in children. Members of our team also conducted fieldwork in New Caledonia, where we have been studying the cognitive abilities of tool-making New Caledonian crows on comparative tasks. The results from these comparative studies inform us about the differences and similarities in the mechanisms that underpin causal knowledge in corvids and humans.

We have published a high number of papers, and have been active in sharing our results and research methods with the public, at a wide range of outreach events. We showed that (1) children possess a core knowledge of the world, as they can distinguish between possible and “impossible” physical situations. Children at the age of 2 can produce new movements after observing correlations between two objects, and by the age of 5, they can recreate novel patterns of movements to solve a task. Around the age of 7, children develop a causal understanding of tool functionality, and by the age of 8-9, they can mentally solve a problem and manufacture a tool to do so. (2) When comparing children and young corvids, we found that birds have a similar core physical knowledge to humans. However, when solving a physical problem, New Caledonian crows, but not children, are biased towards using objects that have previously been rewarded. New Caledonian crows are also incapable of creating novel interventions after observing correlations between two objects, an ability that 2-years old children possess. Yet, like human adults but unlike children, New Caledonian crows can assess when using a tool is necessary. Interestingly, tool-using corvids learn how to manufacture new tools by using other individuals’ tools and recreating their own based on a mental template of the tools, rather than learning directly from observing others, like humans do. Hence, core physical knowledge is common to both humans and corvids, and some key abilities to understanding causality and problem solving seem to have converged between both. (3) Finally, we also investigated which factors affected an individual’s ability to solve a physical task in children and birds. First, we found that species using tools are better at planning the future. Second, we found that self-control can influence an individual’s ability to solve tasks, and having to wait for a more rewarding response affected younger children’s performance more than older children’s. Overall, we made significant steps to understand which cognitive mechanisms underlie causal knowledge, a core ability tool use and problem solving.