Cognitive load is the total amount of mental effort being used in working memory. John Sweller’s cognitive load theory distinguishes three types: intrinsic load (the complexity inherent to the material), extraneous load (effort caused by poor instructional design), and germane load (effort devoted to building understanding) [@sweller1988].
Working memory is limited — it can hold roughly four to seven elements at a time. When a lesson, document, or specification overloads working memory through dense vocabulary, tangled syntax, or too many simultaneous concepts, comprehension drops regardless of how correct the content is.
This has direct implications for writing:
- One idea per section reduces the number of elements competing for processing. The style guide’s “one idea per paragraph” and the plain language specification’s “one purpose per section” both follow from this.
- Concrete before abstract works because a concrete example is easier to hold in working memory than an abstract definition. The example creates a mental anchor that the formal definition can attach to.
- Defined terms on first use reduces extraneous load — the reader doesn’t have to pause to figure out what a word means. The write-glossary skill operationalizes this.
- Short sentences aren’t just stylistically preferable — they reduce the number of elements the reader must hold in mind while parsing. Richard Mayer’s segmenting principle is the instructional design version of the same insight [@mayer2009].
Cognitive load theory doesn’t mean “make everything simple.” It means reduce the effort caused by poor design (extraneous load) so the reader can devote their limited working memory to the actual complexity of the material (intrinsic load).
Related terms
- readability — readability measures surface features that contribute to cognitive load, but cognitive load is broader
- document design — good document design reduces extraneous cognitive load through visual structure
- audience — what counts as cognitive overload depends on the audience’s prior knowledge