The process of tool innovation in young children: An attempt to bridge the Gestalt and the perception-action theories Jiajun Guo a,b,* , Honghong Bai c,d , Xiaofei Long b , Xueyun Su e , Weiguo Pang b a Shanghai Institute of AI Education, East China Normal University, Putuo District, Shanghai, 200062, China b School of Psychology and Cognitive Science, East China Normal University, Putuo District, Shanghai, 200062, China c Behavioural Science Institute & Orthopedagogics: Learning and Development, Radboud University, Postbus 9104, 6500 HE, Nijmegen, the Netherlands d Tsinghua Laboratory of Brain and Intelligence & Department of Psychology, Tsinghua University, Haidian District, Beijing, 100084, China e Department of Early Childhood Education, Faculty of Education & Changning Maternity and Infant Health Hospital, East China Normal University, Putuo District, Shanghai, 200062, China A R T I C L E INFO Keywords: Tool innovation Children Creativity Fixation Flexibility ABSTRACT Background: Tool innovation in young children is an underexplored topic, and the Gestalt and the perception- action theories offer different explanations regarding the process of tool innovation. Aims: We aimed to develop nine tool innovation tasks to investigate the tool innovation process in 3-to 6-year-old children, while providing evidence on the validity of the tasks. Sample: From three kindergartens, 141 Chinese children (77 boys and 64 girls) took part in this research. Methods: Children’s task performance was coded as task success (whether they retrieved the reward) and ideal manipulation (whether they performed a certain manipulation), and children’s task exploration process was coded for fixation (the frequency of repeating the same actions) and flexibility (the frequency of shifting to novel actions). Children’s cognitive flexibility, working memory, inhibitory control, divergent thinking, and general intelligence were also measured. Results: The new tasks showed varied difficulty levels and exhibited high internal consistency, and children’s task performance was positively associated with their individual differences in age, general intelligence, and diver- gent thinking. Children’s behavioral fixation hinders their tool innovation, and this effect was strengthened for easier compared to harder tasks. Children’s behavioral flexibility enhances their tool innovation, and this effect was seemingly stronger for harder compared to easier tasks. Conclusions: The new tasks were robust and suitable for studying young children’s tool innovation. Task difficulty level has a critical role in shaping children’s tool innovation process, which may serve as a mechanism to bridge the Gestalt and the perception-action theories in explaining the process of tool innovation. 1. Introduction Tool innovation, as an important manifestation of human creativity, involves creating or modifying tools to solve novel problems and is crucial for advancing human civilization (Rawlings & Legare, 2021). However, tool innovation is not unique to humans, as instances of such behavior have also been observed in various animal species. For instance, chimpanzees have been seen fashioning spears from sticks for hunting (Gibbons, 2007), and a New Caledonian crow has been observed to manipulate a wire to retrieve objects (Weir et al., 2002). Intriguingly, past research has shown that human children, despite having more so- phisticated brain structures and cognitive functions—such as those supporting language development—do not always outperform these animals on similar tool innovation tasks (Beck et al., 2011; Cutting et al., 2011). This observation directs our attention to an important but often overlooked question: How does tool innovation unfold and develop during early childhood? Understanding this process is crucial for unraveling the unique route of human development, where “under-- capable” children gradually develop into “over-capable” adults, that distinguishes humans from other species. * Corresponding author. Shanghai Institute of AI Education, East China Normal University, RM 213, JunXiu Building, 3663 North Zhongshan Rd., Putuo District, Shanghai, 200062, China. E-mail addresses: jjguo@psy.ecnu.edu.cn (J. Guo), honghong.bai@ru.nl (H. Bai), xflong@psy.stu.ecnu.edu.cn (X. Long), xysu@spe.ecnu.edu.cn (X. Su), wgpang@ psy.ecnu.edu.cn (W. Pang). Contents lists available at ScienceDirect Learning and Instruction journal homepage: www.elsevier.com/locate/learninstruc https://doi.org/10.1016/j.learninstruc.2025.102182 Received 30 June 2024; Received in revised form 29 June 2025; Accepted 1 July 2025 Learning and Instruction 99 (2025) 102182 Available online 14 July 2025 0959-4752/© 2025 Elsevier Ltd. All rights are reserved, including those for text and data mining, AI training, and similar technologies.