Résumé

Inference by exclusion, or the ability to select a correct course of action by systematically excluding other potential alternatives, is a form of logical inference that allows individuals to solve problems without complete information. Current comparative research shows that several bird, mammal and primate species can find hidden food through inference by exclusion. Yet there is also wide variation in how successful different species are as well as the kinds of sensory information they can use to do so. An important question is therefore why some species are better at engaging in logical inference than others. Here, we investigate the evolution of logical reasoning abilities by comparing strepsirrhine primate species that vary in dietary ecology: frugivorous ruffed lemurs (Varecia spp.) and folivorous Coquerel's sifakas, Propithecus coquereli. Across two studies, we examined their abilities to locate food using direct information versus inference from exclusion and using both visual and auditory information. In study 1, we assessed whether these lemurs could make inferences when full visual and auditory information about the two potential locations of food were provided. In study 2, we then compared their ability to make direct inferences versus inferences by exclusion in both the visual and auditory domains. We found that these lemur species can use visual information to find food, but that only ruffed lemurs were also able to use auditory cues, mirroring differences in the complexity of their wild ecology. We further found that, unlike many anthropoid species tested to date, these strepsirrhine species failed to make inferences by exclusion. These results highlight the importance of natural history in understanding the evolution of logical inference and help reconstruct the deeper phylogeny of primate cognition.