Cognitive Load

Definition:

Cognitive load is the total demand placed on working memory at a given moment. Cognitive Load Theory (CLT), developed by John Sweller in the late 1980s, holds that effective learning requires managing cognitive load: keeping total demand within working memory’s limited capacity so that meaningful processing and long-term encoding can occur.

Also known as: cognitive overload, mental load, working memory load, CLT (Cognitive Load Theory)

In-Depth Explanation

Working memory — the active mental workspace where conscious thought occurs — is severely limited in capacity. George Miller‘s landmark 1956 research established approximately 7 ± 2 chunks as the upper bound for simultaneous items; subsequent research revised this closer to 4 items for truly independent information. When working memory is overloaded, learning breaks down: processing effort goes into managing the overload rather than encoding new information into long-term memory.

Sweller proposed that the total cognitive load experienced by a learner consists of three components:

1. Intrinsic load — the inherent difficulty of the material itself, determined by how many elements must be processed simultaneously to understand it (what Sweller called “element interactivity”). A single isolated vocabulary item has low intrinsic load. A complex grammar pattern requiring simultaneous attention to word order, conjugation, tense, and politeness level has high intrinsic load. Intrinsic load cannot be reduced by better instruction — it is a property of the material — but it can be managed by sequencing simpler elements before complex ones.

2. Extraneous load — cognitive demand imposed by poor instructional design that does not contribute to learning. A flashcard with cluttered layout, ambiguous wording, a distracting image, and three pieces of information where one was needed imposes extraneous load. The learner’s working memory is consumed by parsing the presentation rather than processing the content. Good SRS card design minimizes extraneous load: one concept per card, clear and unambiguous prompt, no irrelevant visual elements.

3. Germane load — the cognitive effort devoted to schema formation: actively building and integrating new information into existing knowledge structures in long-term memory. This is the “productive” load — it is what actual learning feels like when working well. The goal of instructional design is to minimize extraneous load and maximize germane load within the total capacity available after intrinsic load is paid.

For SRS design, Cognitive Load Theory has several direct consequences:

• New item limits per session directly address intrinsic load accumulation. Introducing 30 new vocabulary items in one session pushes total load past what can be meaningfully encoded; the items are superficially processed and forgotten quickly. Most SRS apps allow users to cap new items per day — this is CLT in practice.

• Card atomicity (one concept per card) reduces extraneous load. Compound cards that test multiple things at once require learners to hold several variables in working memory to determine whether they “knew” the card. A single, specific prompt eliminates this ambiguity.

• Session length interacts with fatigue-driven load increases. Working memory performance degrades with sustained demand. Shorter, daily sessions respected by SRS scheduling outperform longer, infrequent sessions partly because per-session cognitive load remains manageable.

In language learning specifically, Cognitive Load Theory connects to Stephen Krashen‘s Affective Filter Hypothesis: high anxiety raises effective cognitive load (anxiety consumes working memory resources), leaving less capacity for processing. A low-anxiety self-study environment like SRS therefore has a cognitive advantage over high-stakes performance situations, independent of any motivational effects.

Common Misconceptions

“Cognitive load theory means keep everything simple.”

CLT says minimize extraneous load (unnecessary complexity from design), not intrinsic load (the actual difficulty of material). Oversimplifying the material itself reduces what there is to learn, which is not the goal. Good SRS cards are simple in form (no cluttered design) but can address genuinely complex content.

“If something feels hard, it’s causing too much cognitive load.”

Germane load — the effort of genuine learning — also feels demanding. Not all cognitive difficulty is harmful. The desirable difficulty of retrieval practice (working hard to recall an answer) is a form of germane load that produces better consolidation. The load to minimize is extraneous: effort spent on poor presentation, ambiguity, or irrelevant information.

“Working memory can only hold 7 things.”

Miller’s “7 ± 2” refers to chunks — meaningful units — not raw items. Expert chess players can hold more board positions in working memory than novices because they chunk pieces into patterns. Similarly, language learners who have internalized vocabulary as fluent chunks can process sentences that would overload a beginner, not because their working memory is larger, but because chunking reduces intrinsic load.

History

• 1956: George Miller publishes “The Magical Number Seven, Plus or Minus Two,” establishing the capacity limits of working (short-term) memory and framing it as a fundamental constraint on information processing. [Miller, 1956]

• 1974: Alan Baddeley and Graham Hitch replace the simple “short-term memory buffer” with the multi-component working memory model, providing a more mechanistic framework for understanding why capacity limits exist and how different types of information compete for resources. [Baddeley & Hitch, 1974]

• 1988: John Sweller publishes “Cognitive load during problem solving: Effects on learning” in Cognitive Science, introducing Cognitive Load Theory as a framework for designing effective instruction. Sweller identifies intrinsic and extraneous load and argues that reducing unnecessary load is the primary lever for improving learning outcomes. [Sweller, 1988]

• 1998: Sweller, van Merriënboer, and Paas publish a comprehensive CLT framework adding germane load and providing worked-example and split-attention effects as design principles. This paper establishes CLT as a complete instructional design theory. [Sweller et al., 1998]

• 1990s–present: CLT principles are applied systematically to educational technology design, including SRS tools, e-learning platforms, and multimedia learning environments. John Sweller‘s work directly informs how Anki, Sakubo, and similar tools approach card design guidelines and session structure.

Criticisms

Cognitive Load Theory has been criticized for the difficulty of independently measuring intrinsic, extraneous, and germane cognitive load — the three types in Sweller’s original model. Without operationally distinct measures, the framework risks being unfalsifiable (any outcome can be explained post-hoc in CLT terms). The “redundancy effect” and “expertise reversal effect” — both CLT predictions — show the complexity of applying the theory to real instructional design: what reduces load for novices may increase it for experts. Critics also note that CLT’s origins in text-based instruction may not generalize straightforwardly to naturalistic conversational L2 learning.

Social Media Sentiment

Cognitive load is discussed in language learning communities primarily in practical terms — advice to avoid “overwhelming” study sessions, to limit vocabulary per session, and to use spaced repetition to manage memory load. The CLT framework itself appears in content aimed at language teachers and instructional designers more than at learners directly. The concept resonates widely with learners who have experienced burnout from trying to cram too much input at once, validating the “quality over quantity” approach to study sessions.

Last updated: 2026-04

Practical Application

Cognitive load principles have direct implications for L2 learning design. Keeping vocabulary review sessions short (20–30 items maximum) reduces intrinsic load to a manageable level. Spaced repetition systems like Sakubo distribute load over time, ensuring that the number of items requiring active recall at any session is calibrated to working memory capacity. For input comprehension, learners can reduce extraneous load by controlling contextual complexity (starting with slower, simpler audio before natural-speed authentic input) rather than jumping to fully authentic material before L2 processing has become automatic.

Related Terms

• Working memory
• SRS (Spaced Repetition System)
• Study queue
• Retrieval practice
• Spacing effect

See Also

• John Sweller
• George Miller
• Alan Baddeley
• Long-term memory
• Comprehensible input

Research

• Sweller, J. (1988). Cognitive load during problem solving: Effects on learning. Cognitive Science, 12(2), 257–285. https://doi.org/10.1207/s15516709cog1202_4
  Summary: The foundational CLT paper — introduces intrinsic and extraneous load and demonstrates that instructional design which reduces unnecessary cognitive demands produces measurably better learning outcomes. The primary reference for CLT’s application to SRS and other learning tool design.

• Sweller, J., van Merriënboer, J.J.G., & Paas, F.G.W.C. (1998). Cognitive architecture and instructional design. Educational Psychology Review, 10(3), 251–296.
  Summary: The comprehensive CLT framework that adds germane load and articulates the full three-load model. Provides the most complete theoretical basis for applying CLT to instructional design including worked examples, split-attention effects, and expertise reversal.

• Miller, G.A. (1956). The magical number seven, plus or minus two: Some limits on our capacity for processing information. Psychological Review, 63(2), 81–97. https://doi.org/10.1037/h0043158
  Summary: Establishes working memory capacity limits — the cognitive constraint that makes load management necessary. Foundational paper for understanding why SRS session size limits exist and why atomicity of card design matters.

• Paas, F., Renkl, A., & Sweller, J. (2003). Cognitive load theory and instructional design: Recent developments. Educational Psychologist, 38(1), 1–4. https://doi.org/10.1207/S15326985EP3801_1
  Summary: Reviews a decade of CLT research and its applied consequences for educational technology. Documents specific design principles — minimal redundancy, worked examples, split-attention avoidance — that directly translate to SRS card and session design.

• Baddeley, A.D., & Hitch, G.J. (1974). Working memory. In G.H. Bower (Ed.), The Psychology of Learning and Motivation (Vol. 8, pp. 47–89). Academic Press.
  Summary: The multi-component working memory model that provides the structural explanation for why cognitive load limits exist. Essential for understanding how different types of information (verbal, visual, procedural) compete for separate working memory subsystems during SRS review.