Causality
In the first strand we made fascinating discoveries about similarities and differences between causal learning and curiosity in primates and children. Both non-human primates and human children were able to solve problems involving predictive causal inference – to avoid a trap or blockage (Jordan et al 2020; Yuniarto et al 2020). And excitingly, there was evidence that they could learn in the absence of reward, through exploration, suggesting that curiosity about the way the world works may have deep evolutionary roots (Jordan et al submitted; in prep). Yet we also found some differences in performance. In tests that required participants to infer an unseen cause in order to find a reward, we found that four- and five-year-old children outperformed monkeys and apes across two methodologies (Civelek et al 2020; 2021). We found that children consistently performed in line with predictions from computational models that are capable of learning not just specifics, but also overarching rules or hypotheses. However, in one study non-human primates showed no evidence of learning, neither at the specific nor the abstract level (Felsche et al 2019). However, in a second study we found tentative evidence that both chimpanzees and capuchin monkeys could learn the abstract rule governing the distribution of rewards in puzzle boxes (e.g. all boxes contain the same type of reward, or all boxes contain a mixture of rewards). Interestingly, when children outperformed primates their performance was related to explain their solution using language, suggesting that thinking in words plays a powerful role in enabling humans to go beyond their primate relatives in their causal reasoning, and harness their inquiring minds to make new discoveries.
Executive Functions
We developed new methods to allow us to compare executive functions across human and non-human primates, in three domains. We developed a new methodological approach based on human psychometrics for establishing validity (Voelter et al 2018). We used this approach to develop new tests of working memory (Voelter et al 2019; Reindl et al 2021); inhibition (Voelter et al 2022), and attentional flexibility (Reindl et al 2022; Reindl et al in revision). Interestingly in some of the experiments children performed better than chimpanzees by the age of 4-5, such as tests of attentional control (Reindl et al in revision) and working memory (Reindl et al in preparation).
Cross-sectional Test Battery
In a huge coordinated effort by our team, we tested 55 chimpanzees and 185 children on a test-battery, using the 12 new tasks developed in the previous 2 strands that yielded the appropriate signatures in both species (Voelter et al, 2022). Our results overall indicated some unity in the structure of individual differences, consistent with the evidence for adult humans. However, we found no strong evidence for an unequivocal structure in the EF tasks for either preschool children or chimpanzees. Task performance only correlated at low to moderate levels and there was no clear indication of a preference for any one of the models proposed in the literature. One explanation for the pattern of results is that performance in EF tasks in children and chimpanzees is not well explained by the presence of several dissociable underlying capacities. The exercise of executive control might best be understood as a dynamic and distributed process involving heavily overlapping networks. In other words, EF are a “moving target” rather than discrete, domain-independent skills. We found no evidence that these constructs are natural kinds that can be identified by tasks raising the same demand in different experimental contexts.
