Between structure and flexibility: Testing the limits of human generalization.

A hallmark of human intelligence is the ability to both profit from earlier experience and flexibly adapt in novel ways to unexpected situations. How humans are able to trade-off using earlier experience versus going beyond it remains unknown. Here, we argue that humans extract common structure across tasks and then strategically adapt this common structure in novel situations. We describe three such adaptations of common structure: inversion, recomposition, and tuning. To empirically test this, we developed a novel behavioral task, which required adapting knowledge across tasks in order to generalize appropriately in the test phase. Across three experiments, the need for inversion, recomposition, and tuning was increasingly added to the basic task. Results demonstrate that humans can use all three types of adaptation. We discuss how these findings can inform models of human generalization. (PsycInfo Database Record (c) 2025 APA, all rights reserved).