Cognitive Load

Working memory is the small, temporary system that holds and manipulates information during thinking. It is the workshop where reading, reasoning, mental arithmetic, and problem-solving actually happen. Long-term memory can hold vast amounts of knowledge, but working memory is where any new material must first pass to become learning. Working memory is famously limited. Miller (1956) proposed the magical number seven, plus or minus two. More rigorous later work using tests that prevent chunking places the true capacity closer to three to four items (Cowan, 2001). Information in working memory also decays quickly — within roughly 15 to 30 seconds without active rehearsal. These constraints have a decisive consequence for learning. If instruction, material, or environment overwhelms working memory, comprehension collapses regardless of effort. Effective learning is not about pushing harder; it is about respecting working memory and building richer long-term structures that expand what can be handled at once.

Cognitive Load Theory, developed by John Sweller and colleagues from the late 1980s onward (Sweller, 1988; Sweller, van Merriënboer, and Paas, 1998), distinguishes three types of load imposed on working memory. Intrinsic load is the difficulty inherent in the material itself: the number of elements and how they interact. Learning the meaning of a single word is low intrinsic load; understanding a differential equation is high. Extraneous load is difficulty added by how the material is presented: bad diagrams, unclear language, split attention across sources, distracting media, decorative but irrelevant detail. It contributes nothing to learning and can prevent it. Germane load is the effortful work of building schemas — of forming durable, generalizable structures in long-term memory. This is the useful load, the load that actually produces learning. The goal is not to minimize all load. It is to keep total load within working-memory capacity while shifting as much of it as possible from extraneous to germane. Reduce clutter, then invest the freed capacity in effortful understanding.

Extraneous load is everywhere and often invisible. Common sources include split-attention effects (instructions on one page, diagram on another), redundant material (text repeating what a labeled diagram already shows), unnecessary complexity in language (jargon where plain terms would do), bad interfaces (cluttered dashboards, inconsistent icons), environmental noise (notifications, background conversation, music with lyrics during reading), and emotional load — anxiety about performance itself consumes working-memory resources that would otherwise be available for the material (Beilock and Carr, 2005). Much of the difficulty students blame on the material is really extraneous load. The material was hard enough; the presentation made it impossible.

The KAfrequency approach treats cognitive load as an architectural problem. If working memory is the workshop, the goal is to keep the workshop clean and the tools within reach. A practical ruleset: match the difficulty of new material to current capacity — if a text feels impossible after honest effort, drop a level, learn the prerequisite, then return. Reduce presentation clutter; choose one canonical source per topic and close everything else. Chunk material into meaningful groups so working memory can treat many items as one. Use worked examples early — studying a completed solution before attempting a similar problem reduces extraneous load and builds schemas faster than pure problem-solving (Sweller and Cooper, 1985). Interleave practice once basics are stable; interleaving increases germane load and improves long-term transfer. Protect the emotional envelope — reduce anxiety with clear goals, short intervals, and small early wins. The deeper KAfrequency principle: what we are able to think depends on what we make room for. Sovereign learning is the discipline of clearing the workshop.