Lateral reading
A way of judging an unfamiliar website by leaving it almost at once and seeing what other sources say about it, then returning to read it with that context in hand. It is set against staying on the page itself — scrolling, reading the site's self-description, weighing its look and internal consistency — which tends to be the default and tends to mislead. The move is to find out who is behind a source before trusting what the source says about itself.
An online-information evaluation strategy: when encountering an unfamiliar source, leave it almost immediately and consult what other sources say about it, before returning to read the original. The contrast is vertical reading — staying on the page, scrolling, reading the About section, evaluating internal consistency and design quality, which the empirical evidence suggests is the strategy non-experts default to and the one that produces the most evaluation errors. The techniques expert fact-checkers use to evaluate sources online are not the techniques academic reading instruction trains; the gap between the two accounts for a substantial share of contemporary online misinformation susceptibility on the empirical record.
Etymology§
The term lateral reading in the contemporary information- literacy sense was articulated by Sam Wineburg and Sarah McGrew in the 2019 Teachers College Record paper, on the basis of think-aloud protocols comparing professional fact-checkers, academic historians, and Stanford undergraduates. The corresponding teachable technique had been developed in parallel by Mike Caulfield through the 2016–2017 work on the Web Literacy for Student Fact-Checkers textbook and later the SIFT (Stop, Investigate the source, Find better coverage, Trace claims) mnemonic. The two strands converged in the late-2010s information-literacy literature; lateral reading is the construct, SIFT is the operational mnemonic that carries it into classrooms.
The construct's empirical anchor is the 2019 Stanford History Education Group study. Three groups — professional fact-checkers, academic historians from elite universities, and Stanford undergraduates — were asked to evaluate unfamiliar websites making contested political claims while narrating their reasoning aloud. The fact-checkers performed substantially better than the other groups. The strategy that distinguished them was to leave the page almost immediately upon arriving — opening additional browser tabs to find out who the source was, who had funded it, what other organisations had said about it — and to return to the original page only after this lateral context had been gathered.
The historians and the undergraduates vertically read — they remained on the original page, scrolled for context, examined the site's self-description, looked at design quality, evaluated the writing — and arrived at less accurate evaluations. The historians were, by ordinary measures of expertise, more skilled readers than the fact-checkers; they brought their academic- reading habits to a medium for which those habits were poorly adapted. The fact-checkers' habits were learned on the job in newsroom environments under time pressure; what they produced was a faster, more accurate, less internally elaborate reading strategy.
The implication for teaching is the practical question of how to transfer the fact-checker strategy to students who do not naturally use it. The Caulfield textbook is the worked-out instructional response. The book operationalises lateral reading into four concrete moves — check for previous work, go upstream to the source, read laterally, circle back — each illustrated with real examples and practiced in short exercises. The condensed SIFT mnemonic is the form most widely taught in US library- instruction modules and high-school information-literacy curricula.
The construct's relationship to the transfer problem is the open empirical question of how far the trained habit travels. Initial studies through 2020–2024 suggest that targeted instruction does produce measurable transfer to unfamiliar sources, topics, and (so far less studied) media — but the effect is sensitive to how the instruction is delivered and how much practice the trainees get with worked examples. The question of whether lateral reading transfers from the textual online environment in which it was developed to the increasingly multimedia and AI-mediated information environment of the mid-2020s is one of the active research questions in the Stanford group's ongoing work and in the wider information- literacy literature.