Instinct to insight: a variation-based framework to test hypotheses about how animals solve problems.

Abstract

Problem-solving is an integral part of most animals' lives. There are generally four types of solutions animals may use: innate, learned previously, learned de novo or insightful. Identifying the types of solutions animals use can be difficult, especially with the trend of having increasingly difficult requirements to test hypotheses in this field. These requirements often amount to proving a negative, which may be impossible. Therefore, here we develop a novel framework for testing hypotheses that can help distinguish the types of solutions animals may use that does not require proving a negative. This framework is based on distinct patterns of qualitative and quantitative variation between and within individuals. Because this framework does not require knowledge of animal's prior history nor that the problem be evolutionarily novel, it can be used with a variety of animals, experimental designs and settings. We suggest this framework could serve as a valuable tool in expanding how we study animal problem-solving, especially in the types of animals studied. Studying problem-solving in a wide variety of animals would allow us to form a better understanding of the problem-solving abilities different brain sizes and structures confer and, more broadly, the evolution of those abilities.