Reply to ‘How physical information explains infants’ supposed inner psychological world’

Benton and West are sceptical of our claim that infants possess domain-specific yet integrated knowledge of agents and objects (Liu, S., Karakose-Akbiyik, S., Outa, J. & Kim, M. How physical information is used to make sense of the psychological world. Psychol . 5, 59–73 (2026))1; their Correspondence (Benton, D.

How physical information explains infants’ supposed inner psychological world. Psychol. https://doi.org/10.1038/s44159-026-00628-0 (2026))2 suggests that domain-general learning simulations reproduce infants’ behavioural responses in published studies of early agent and object knowledge3 and infants’ sensorimotor experiences deliver rich, dynamic information that could serve as the inputs to this learning. However, Benton and West’s logic — which connects claims about representation, acquisition and sensory inputs — makes predictions that are not supported by existing data.

Our arguments on the organization and nature of physical and psychological processing in infancy and beyond concern representation, rather than acquisition. We are open to the possibility that generic learning algorithms could produce abstract, structured, specialized representations in infants’ minds. But we take issue with Benton and West’s further claim that because mechanisms of acquisition place constraints on the resulting learned representations, evidence for domain-general learning undercuts the domain specificity of infants’ knowledge.

By their logic, if infants learn about agents and objects via sensorimotor association, their ‘knowledge’ consists of just that: a single web of sensorimotor associations, rather than two distinct systems. This is a preview of subscription content, access via your institution Receive 12 digital issues and online access to articles Prices may be subject to local taxes which are calculated during checkout Liu, S., Karakose-Akbiyik, S., Outa, J. & Kim, M. Psychol. 5, 59–73 (2026).

Psychol. https://doi.org/10.1038/s44159-026-00628-0 (2026). An associative-learning account of how infants learn about causal action in animates and inanimates: a critical reexamination of four classic studies. Gen. 154, 497–521 (2025).

Teleological reasoning in infancy: the naive theory of rational action. Knowing before doing: review and mega-analysis of action understanding in prereaching infants. Bull. 149, 294–310 (2023).