John Sweller

John Sweller

John Sweller is an Australian educational psychologist, emeritus at the University of New South Wales, who set out cognitive load theory in 1988. The theory holds that learning is limited by how much can pass through working memory at once, so how material is presented changes how much of it is learned even when the material is identical. It is the sturdiest framework in the modern argument about screens and attention, and none of it is about the internet. Sweller was working on how to teach algebra.

Who they are

Sweller's subject is instruction, not technology. He came at the problem from the direction of problem solving: why students who successfully solve a set of maths problems often learn very little from having solved them, and what a teacher can change about the presentation to alter that. The answer he arrived at is the small opening between what is in front of a learner and what ends up in long-term memory.

Cognitive load theory divides the traffic through that opening into kinds, and the division is the useful part. Intrinsic load is the difficulty of the subject itself, which a teacher cannot wish away. Extraneous load is the demand added by the way the material happens to be presented, and it is the only one of the three a designer can cut for free. The third is the productive effort of building understanding, which competes for the same capacity as the other two. The practical instruction that falls out of this is unglamorous and testable: reduce the extraneous, and more of the fixed capacity is available for the thing you actually want learned.

The capacity limit underneath the theory is not his. It comes from George Miller's 1956 work on working memory and was revised downwards by Nelson Cowan in 2001. What is Sweller's is the argument built on it, that instructional format is not a matter of taste, and that two presentations of identical content can produce different learning because one of them spends the bottleneck badly.

Nicholas Carr is why Sweller's name appears in a general argument about reading on screens. Carr's mechanism for the cost of the web is load: every hyperlink is a decision, and decisions consume a working memory that is very small. Our reading of The Shallows puts the cognitive load and memory consolidation chapters down as the soundest science in the book, and the honest qualifier is that this is because that science was done on classrooms and textbooks rather than on browsers.

The bridge from "presentation format affects learning" to "hyperlinks degrade reading" belongs to Carr. It is a reasonable bridge, and there is evidence on the far side of it: the hypertext comprehension studies, and the screen inferiority meta-analyses, which hold at 54 studies and more than 170,000 participants and are bounded to expository text under time pressure. It is still an argument by extension rather than a finding of Sweller's, and it should be attributed that way. Sweller did not study the internet, did not make claims about it, and is not a party to the argument his theory is being used to settle.

Key papers

Sweller's output is journal articles rather than trade books, so there is no book table on this page. He does have one academic volume to his name, co-authored with two colleagues, and it is a monograph for researchers rather than a general book: Sweller, J., Ayres, P., & Kalyuga, S. (2011), Cognitive Load Theory, Springer, written with Paul Ayres and Slava Kalyuga, then all at the School of Education at the University of New South Wales.

Sweller, J. (1988). "Cognitive load during problem solving: Effects on learning." Cognitive Science, 12(2), 257-285. The founding paper, on problem solving and instructional design.

Sweller, J., van Merriënboer, J. J. G., & Paas, F. (1998). "Cognitive architecture and instructional design." Educational Psychology Review, 10(3), 251-296. The statement of the theory as an architecture: a limited working memory with partly separate channels, and an effectively unlimited long-term memory holding schemas.

Sweller, J., van Merriënboer, J. J. G., & Paas, F. (2019). "Cognitive architecture and instructional design: 20 years later." Educational Psychology Review, 31(2), 261-292. The twenty-year retrospective on the same argument.

These three were checked against publisher and university repository records on 2 September 2026. They were not read in full here, so specific figures from any of them should be taken to the papers before use.

Key ideas

Working memory is the bottleneck for learning, and it is small.

Load comes in kinds. Cutting the extraneous kind is the one free improvement available to anyone designing instruction.

Identical content, presented differently, produces different learning. Format is not decoration.

Solving a problem and learning from solving it are not the same event, which is where the theory started.

None of this was about screens. The extension of load theory to hyperlinks and browsing was made by other people and belongs to them.

Concepts on this site

Cognitive Load, his theory, and the page that carries the detail

Hypertext, where load is proposed as the mechanism for the comprehension cost of links

Screen Inferiority, the narrow surviving version of the reading claim

Working Memory is associated with him in our notes and has no page on this site yet

Sources

  1. Sweller, J. (1988). "Cognitive load during problem solving: Effects on learning." Cognitive Science, 12(2), 257-285.
  2. Sweller, J., van Merriënboer, J. J. G., & Paas, F. (1998). "Cognitive architecture and instructional design." Educational Psychology Review, 10(3), 251-296.
  3. Sweller, J., van Merriënboer, J. J. G., & Paas, F. (2019). "Cognitive architecture and instructional design: 20 years later." Educational Psychology Review, 31(2), 261-292.
  4. Sweller, J., Ayres, P., & Kalyuga, S. (2011). Cognitive Load Theory. Springer. Academic monograph, co-authored.
  5. Carr, N. (2010, updated 2020). The Shallows: How the Internet Is Changing the Way We Think, Read and Remember. Atlantic Books. Full text read, 1 September 2026, for the use made of load theory and for the bridge to hyperlinks, which is Carr's rather than Sweller's.
  6. Delgado, P., Vargas, C., Ackerman, R., & Salmerón, L. (2018). "Don't throw away your printed books: A meta-analysis on the effects of reading media on reading comprehension." Educational Research Review, 25, 23-38. 54 studies, 171,055 participants, g = -0.21, restricted to expository text under time pressure.