Speed Reading Is Mostly Nonsense — Here's What the Evidence Actually Shows

The speed reading industry is worth roughly $200 million a year globally, built on courses, apps, and books promising to triple or quadruple your reading speed while maintaining comprehension. Evelyn Wood, who founded Reading Dynamics in 1959, claimed to teach students to read 1,500 to 6,000 words per minute — well above the 250 to 300 WPM typical for an educated adult. John F. Kennedy famously attended her course. The programme ran profitably for decades and was eventually acquired by a vendor that still sells a version of it today.

The scientific literature on speed reading, spanning the last fifty years and summarised in a 2016 review by Rayner, Schotter, Masson, Potter and Treiman in Psychological Science in the Public Interest, is blunt: most of the claims made by speed reading programmes are either unsupported by evidence or directly contradicted by it. You cannot, according to the research, read much faster than about 500 WPM while retaining full comprehension. Programmes that claim otherwise are measuring something — but that something is usually skimming, which is a different cognitive activity with a different use case.

That said, skimming is genuinely useful, and there are three techniques that hold up under scrutiny. The honest version of the advice is less exciting than the marketing, but it's also more useful.

Why You Can't Actually Read Faster — the Physiology

Reading involves two constraints. The first is saccades — the rapid eye movements that jump from word to word. Eye tracking data shows that even skilled readers make fixations lasting about 200 to 250 milliseconds on each content word, with regression fixations (going back to reread) making up about 15% of all fixations. You can physically reduce the number of saccades, but you can't reduce the fixation time without losing recognition.

The second constraint is the more fundamental one. Reading comprehension requires the working memory to actually process the words — to identify them, integrate them into syntax, connect them to existing knowledge, update the mental model of what's being communicated. This processing has a roughly fixed lower bound that speed reading techniques cannot shortcut. If you go faster than that bound, you're no longer comprehending what you read. You're doing something closer to visual scanning.

The research is consistent: readers who claim to be reading at 1,000+ WPM are measured, on comprehension tests, to be understanding substantially less than they understood at normal reading speed. The perception of speed is real. The comprehension at speed is not.

The Three Techniques That Actually Work

1. Skimming for structure before reading for depth

For any non-fiction book, the first 20 minutes of your engagement should not be reading. It should be structural analysis. Read the table of contents. Read the introduction and conclusion. Scan the headings of each chapter. Read the first and last paragraph of each chapter. Look at diagrams, pull quotes, chapter summaries.

This gives you the architecture of the argument before you invest in the details. The Mortimer Adler framework from How to Read a Book (1940) calls this "inspectional reading" — the first pass that determines whether the book is worth your time and, if so, which parts deserve deep attention versus skimming.

The payoff: you can often discover, in 20 minutes, that 40% of a book is not useful to you, and that the useful 60% is concentrated in specific sections. You then read those sections carefully, and skim the rest. The total time commitment drops dramatically. Comprehension of the useful parts goes up, because you've approached them with the full structure already in mind.

2. Subvocalisation reduction — carefully

Subvocalisation is the inner voice that reads words aloud in your head. Speed reading programmes pitch the elimination of subvocalisation as a way to double your speed. The research is more nuanced. You cannot eliminate subvocalisation — it's built into how the brain processes language. You can reduce it modestly, and this produces a modest speed gain.

The technique: deliberately keeping your inner voice quieter than your normal reading voice. Not silent — quieter. A whisper rather than a speech. This can push a reader from 250 WPM to maybe 350 WPM without major comprehension loss, which is a real gain.

Past that, trying to silence the inner voice completely starts to cost comprehension. Most readers who try aggressive subvocalisation reduction end up in the skimming zone, where speed is high and retention is poor. The moderate version is the sweet spot.

3. Chunking — reading groups of words rather than word-by-word

Most readers fixate on one word at a time. Skilled readers fixate on groups of two to three words, treating them as a single perceptual unit. This is teachable and produces a genuine speed gain.

The practice: consciously try to absorb phrases rather than words. "In the middle of the night" should be one mental unit, not six. This is easier to describe than to execute — most people slip back into word-by-word reading within minutes. Over weeks of deliberate practice, the habit shifts. The upper bound is still around 500 WPM with full comprehension; chunking is one of the mechanisms that gets you to that upper bound rather than sitting at 250.

What Speed Reading Programmes Actually Teach

Honestly: speed reading programmes do produce measurable improvements in reading speed. They just produce those improvements by turning reading into skimming without telling the student. The student ends their eight-week course able to "read" twice as fast, except that what they're doing with text has changed. They're now sampling text, identifying key phrases, and filling in gaps from prior knowledge. On simple, predictable texts — newspaper articles, pulp fiction, textbook chapters on familiar material — this works tolerably. On dense, unfamiliar, or argumentatively complex material, it doesn't.

The trick is calling this "reading" rather than "skimming." If the course were honest and said "we're teaching you to skim more effectively, and here are the specific types of material where skimming is appropriate" — that would be a useful course. It wouldn't sell as well.

When Skimming Is Appropriate

Skimming is not a lesser form of reading. It's a different activity with its own legitimate uses:

• Evaluating whether a source is worth reading in depth. Inspectional reading, as Adler called it.
• Refreshing material you've read carefully before. You've done the comprehension work; skimming is retrieval.
• News and journalism. Most news articles are inverted-pyramid structured — the key information is in the first two paragraphs, and the rest is elaboration. Skimming produces no real comprehension loss.
• Corporate writing. Most memos, reports and internal documents can be read at 800+ WPM without losing anything, because they're padded. The signal density is low.
• Executive summaries, literature reviews, or secondary sources. Where the author is already compressing, you can compress further.

When skimming is inappropriate: technical material you're trying to learn, primary sources in unfamiliar domains, dense argumentative texts (philosophy, legal writing, literary non-fiction), anything you'll need to remember or apply later, anything aesthetically crafted (novels, poetry).

The discipline, honestly, is knowing which mode you're in. Most people skim things that deserve careful reading, then complain that they can't remember anything. The problem wasn't the memory. The problem was the mode.

What Actually Makes People "Fast Readers"

The people who genuinely read more than average do not read faster. They read more often, select better, and retain more per minute read.

Reading more often — 30 to 45 minutes a day over ten years versus sporadic reading — compounds into a much larger total reading volume than any speed technique can produce. Selecting better — being willing to quit boring or unuseful books, choosing high-signal sources — produces more value per hour read than reading fast through mediocre material. Retaining more — through active note-taking, summary writing, and spaced review — extracts more value from each book read.

If you're genuinely frustrated with your reading output, invest in these three dimensions rather than in speed. The math works out better. A reader going through 30 great books a year at 300 WPM will come out ahead of a reader going through 30 mediocre books at 1,000 WPM every time.

The One Honest Speed Gain You Can Get

If you want to actually read faster, the single most effective intervention is deliberately pushing your normal reading pace by about 10-15% for four weeks. Your comprehension will drop initially. Within two weeks it adapts. Within four weeks you'll be reading at 285 or 295 WPM instead of 250, with roughly equivalent comprehension.

This isn't magic. It's the same adaptation that happens with any skill under slight forcing. Your working memory gets more efficient at what it was already doing. The 15% improvement is modest compared to the 400% gain the Evelyn Wood brochure promised, but it's real, it's sustainable, and it doesn't involve pretending that skimming is reading.

Over a career that involves, say, 3,000 hours of reading, a 15% speed gain gets you an extra 450 hours of reading output. That's meaningful. It just takes honest effort over time rather than a weekend course.