The Cognitive Architecture of Digital Externalization

This review is aimed at synthesizing current findings concerning technology-based cognitive offloading and the associated effects on learning and memory. While cognitive externalization (i.e., using the environment to outsource mental computation) is a highly useful technique in various problem-solving tasks, a growing body of research suggests that the offloading of information into the environment (and digital storage in particular) can have negative effects on learning. Based on this review, a model of offloading with cognitive load at its core is developed to summarize when learners offload information. A high intrinsic cognitive load (i.e., a high difficulty), a high extraneous load (i.e., unnecessary design elements), and a low perceived or actual working memory capacity trigger offloading. Crucially, the value attributed to information also affects whether information is externalized. In this model, extraneous cognitive load in the design of technology-enhanced learning acts as a triple barrier: (1) It prevents information from entering working memory, (2) it inhibits information being stored in long-term memory, and (3) it can prevent learners from externalizing information using technology. As a result, in many instances, only the gist of information (or its location) is retained, while learners often gain the illusion of having memorized that information. Furthermore, offloading substantially increases the risk of memory manipulation, potentially posing a societal problem. Consequently, educational approaches should maximize the meaningfulness of the residual information that is often retained in the form of “biological pointers.” In addition, current issues surrounding the use of generative artificial intelligence pertaining to externalization are discussed.