Logical reasoning by a Grey parrot? A case study of the disjunctive syllogism

Abstract

Abstract In Call’s (2004) 2-cups task, widely used to explore logical and causal reasoning across species and early human development, a reward is hidden in one of two cups, one is shown to be empty, and successful subjects search for the reward in the other cup. Infants as young as 17-months and some individuals of almost all species tested succeed. Success may reflect logical, propositional thought and working through a disjunctive syllogism (A or B; not A, therefore B). It may also reflect appreciation of the modal concepts “necessity” and “possibility”, and the epistemic concept “certainty”. Mody & Carey’s (2016) results on 2-year-old children with 3- and 4-cups versions of this task converge with studies on apes in undermining this rich interpretation of success. In the 3-cups version, one reward is hidden in a single cup, another in one of two other cups, and the participant is given one choice, thereby tracking the ability to distinguish a certain from an uncertain outcome. In the 4-cups procedure, a reward is hidden in one cup of each pair (e.g., A, C); one cup (e.g., B) is then shown to be empty. Successful subjects should conclude that the reward is 100% likely in A, only 50% likely in either C or D, and accordingly choose A, thereby demonstrating modal and logical concepts in addition to epistemic ones. Children 2 1/2 years of age fail the 4-cups task, and apes fail related tasks tapping the same constructs. Here we tested a Grey parrot (Psittacus erithacus), Griffin, on the 3- and 4-cups procedures. Griffin succeeded on both tasks, outperforming even 5-year-old children. Controls ruled out that his success on the 4-cups task was due to a learned associative strategy of choosing the cup next to the demonstrated empty one. These data show that both the 3- and 4-cups tasks do not require representational abilities unique to humans. We discuss the competences on which these tasks are likely to draw, and what it is about parrots, or Griffin in particular, that explains his better performance than either great apes or linguistically competent preschool children on these and conceptually related tasks.