Insight learning, a concept central to understanding intelligence and problem-solving, was defined by Thorpe (2) as “the sudden production of a new adaptive response not arrived at by trial behavior or as the solution of a problem by the sudden adaptive reorganization of experience.” This definition encapsulates the essence of what many refer to as the “aha!” moment in learning – a sudden flash of understanding that leads to a solution, seemingly without trial and error. The nature of insight learning, however, remains elusive, often sparking debate in the fields of psychology and animal behavior.
Early investigations into insight learning often cited examples like Tolman’s maze-navigating rats (3). These studies suggested that animals could develop cognitive maps and find shortcuts, indicating learning beyond simple stimulus-response associations. Insight learning proposes a cognitive leap, where an individual restructures their understanding of a problem, leading to an immediate solution. This contrasts sharply with other forms of learning, particularly shaping, which offers a different perspective on how novel behaviors can emerge.
Shaping, a powerful behavioral technique developed by Skinner (4), explains the acquisition of new behaviors through successive reinforcement. In shaping, complex behaviors are built incrementally, rewarding successive approximations towards a target behavior. This method has been remarkably successful in teaching animals complex actions, such as dolphins performing intricate routines or, as relevant to our discussion, training rooks. Bird and Emery’s work with rooks (Corvus frugilegus) (1), demonstrating tool manufacture and use, initially employed shaping techniques. For instance, rooks were initially rewarded for simply pushing a stone, then for pushing stones near the apparatus, and finally for fetching stones from a distance to operate the device for food. This progression highlights the step-by-step nature of shaping.
However, the application of shaping in explaining complex behaviors does not negate the possibility of insight learning. The challenge lies in differentiating between behaviors solely learned through shaping and those that involve a genuine insightful understanding. Thorpe himself acknowledged the difficulty in establishing clear criteria to distinguish insight learning from shaping. The crux of the issue is that the same observable behavior, like tool use in rooks, could potentially arise from either mechanism. Therefore, evaluating learning processes requires a focus on how a behavior is achieved, not just the final outcome. Observing a rook skillfully bending wire to retrieve food is indeed impressive, but understanding the learning process necessitates considering the entire training sequence and the cognitive mechanisms at play.
The ongoing debate underscores the need for a more rigorous, quantitative approach to studying learning. Instead of relying solely on verbal arguments about whether a behavior is due to shaping or insight, a quantitative framework could assess task complexity and animal behavior more objectively. This shift could reveal that insight learning and operant conditioning, including shaping, are not mutually exclusive but rather interconnected mechanisms within a broader spectrum of learning capabilities. By moving towards quantitative evaluations, we can gain deeper insights into the cognitive abilities of animals, including humans, and refine our understanding of “What Is Insight Learning” and its role in adaptive behavior.
References
1 Bird, C. D., & Emery, N. J. (2009). Insightful tool use and manufacture in untrained crows. Proceedings of the National Academy of Sciences, 106(25), 10370-10375.
2 Thorpe, W. H. (1963). Learning and instinct in animals (2nd ed.). Methuen.
3 Tolman, E. C. (1948). Cognitive maps in rats and men. Psychological Review, 55(4), 189.
4 Skinner, B. F. (1938). The behavior of organisms: An experimental analysis. Appleton-Century-Crofts.
