Recomposition Based Learning for Promoting Structural Understanding - From Reconstruction of External Representations to Recomposition of Internal Representation-

Abstract

In educational settings, structured external representations, such as diagrams, examples, or concept maps, are often assumed to ensure learners’ understanding. However, recognition of a representation does not guarantee that its underlying meaning has been internalized. Understanding involves not only the elemental meaning conveyed by individual components but also the structural meaning that emerges from their organization; the additional meaning constructed through this organization can be defined as constructed meaning (“constructed meaning = structural meaning – elemental meaning”). To foster this constructed meaning, we propose Recomposition Based Learning, a framework in which learners reconstruct a target structure from predefined components. This process externalizes, compares, and refines learners’ internal representations, thereby promoting structural understanding and metacognitive reflection. By focusing on the manipulation of external representations to refine internal ones, the framework reduces cognitive load while preserving the need for interpretive reasoning. It is formalized as a four-stage cycle: (1) hypothetical recomposition, (2) difference detection, (3) conceptual clarification and completion, and (4) reflective recomposition. This paper outlines the theoretical basis of the framework.