The short version
Speed reading techniques are the methods sold to lift you past the average 250 words per minute. Most of them train your eyes. Your eyes were never the slow part. The techniques that survive the evidence are previewing, deliberate skimming, changing gear by purpose, and reading a great deal. The rest is theatre.
The test every technique has to pass
Your reading rate is not set by your eyes. It is set by how fast your brain pulls meaning out of a word.
That sentence sorts this whole page. Every technique below gets held up against it, and most of the market falls over.
Start with the mechanics. Your eyes do not glide along a line. They jump, they land, they jump again. Each landing, a fixation, lasts about 250 milliseconds. Each jump, a saccade, takes 25 to 30. Add the jumps up and they come to roughly a tenth of your reading time. The landings take the other nine tenths, and the landings are where the reading happens.
So a method that speeds up the jumps is fighting over a tenth of the clock. A method that makes the thinking quicker changes the picture. That is the test. Does it help the brain, or does it only move the eyes?
Who measured this
In 2016 five researchers published a thirty page review in Psychological Science in the Public Interest. They asked how we read and whether speed reading can help.
The lead author was Keith Rayner. With George McConkie he invented the moving window method, the instrument that measures how much text a reader takes in per fixation. He spent forty years building the science of eye movements in reading. He had no course to sell and no app to defend.
We sell speed reading software. Half of what follows costs us money. Read it anyway.
Get a number before you train
Take the speed reading test. It gives you a speed and a comprehension score together, which is the only pair worth having. Measure now, train, measure again. Otherwise you will end up believing whichever technique you enjoyed most.
This page is the reference. It tells you what each method is and whether it holds up. For drills with reps and timings, go to speed reading exercises. For the plain beginner’s route, start with how to read faster. Everything else sits in our speed reading guide.
The numbers, so you can check ours
An average adult reads about 250 words per minute. Children run near 150.
Normal reading, the sort where you understand and remember what you read, tops out around 300 words per minute. The fastest competent adults reach about 550, and by then they are giving something up.
Every number above 600 that you have seen on a sales page belongs to a different activity with a different name. Hold on to that while you read the rest.
Widening your visual span. Verdict: it does not work.
Take the big one first, because every course sells it and it is wrong.
The pitch runs like this. An average reader takes in one word per fixation. A good reader takes in five. Same eyes, five times the payload. Train the span, and the speed follows.
The numbers are invented.
What a fixation actually holds
Your eye resolves letters across about 20 characters. Outside that the page is too blurred to read, and no drill changes the optics of a retina.
Worse for the sales pitch, the span over which you can identify a word runs about seven to eight characters to the right of where you are looking. That is the word you are on, plus the next one. That is the whole payload.
Rayner and his co-authors were blunt about the rest. Taking in a whole line at one glance is simply biologically impossible.
What that means for your practice
You cannot drill your way to five words a fixation, because you cannot resolve five words. There is nothing there to train.
And even if you could stretch the span, you would be stretching the wrong thing. The eye was never the limit. Long fixations do not happen because your vision failed. They happen because the brain stalled on a word it did not know or a meaning it could not fit. Sharper eyes do not fix a stalled brain.
The detail sits on visual span in reading. If you have been practising two words per fixation, then three, then four, stop. That fortnight is better spent reading a book.
Previewing. Verdict: works.
Previewing means going over a text before you read it, to learn what it is and how it is built.
It is the cheapest technique here and the one most people skip. It costs two minutes. It changes what every following minute is worth.
How to preview
Read the title. Read every subheading, in order. Read the first sentence of the first paragraph and the last sentence of the last one. Look at the tables, the figures, anything set in bold. If there is a summary, read it now rather than at the end.
Ninety seconds for a chapter. Two minutes for a long report. Ten seconds for an email.
Then ask three questions. What is this trying to prove? What do I already know about it? What do I want from it? Then read.
Why it works
The bottleneck is meaning, so previewing attacks the bottleneck. It loads the meaning before you start.
A reader who knows the shape of an argument predicts the next sentence better. Better prediction means fewer stalls. Fewer stalls mean fewer trips back up the page to work out what you missed. You are not training your eyes. You are giving your brain a head start, and the brain is the part that was slow.
It does something quieter too. It makes you decide, before you spend an hour, whether the text deserves one. Most reading problems are not speed problems. They are the problem of reading the wrong thing all the way to the end.
Skimming. Verdict: works, and it is a real skill.
Skimming means going over a text fast for the gist, without reading every word. Scanning means hunting a page for one thing you already want.
Neither is speed reading. For years both were treated as cheating. They are not.
Rayner’s verdict on skimming, in his own words: a reasonable way to cope with the overwhelming amount of text we have to read. The man who spent forty years measuring reading is telling you to skim.
Skimming is triage
You cannot read everything. So decide what deserves a real read.
Run down the page. Take the headings, the first sentences, the numbers, the names, anything the writer has emphasised. Come out the other end with the shape of the thing and a decision. Read it properly, file it, or bin it.
Skimming costs you the detail. You accept the cost with your eyes open. That is what makes it a skill rather than a con.
The con is skimming with a stopwatch running, sold to you as reading. We hold the two apart on skimming versus speed reading.
Changing gear. Verdict: works, and nobody sells it.
A slow reader has one speed. A good reader has a gearbox.
Ronald Carver measured what readers do at different jobs, and the rates are not opinions.
Memorising a text runs near 140 words per minute. Learning it runs near 200. Reading it with full understanding runs near 300. Skimming runs near 450. Scanning for a target word runs near 600.
Read the last two rows honestly
The industry quotes 450 and 600 and stops there. Look at what Carver measured.
The 450 came from readers hunting for transposed words in a sentence. The 600 came from searching a text for a target word. Those are hunting tasks. They are not reading with comprehension, and quoting them at a customer does not make 600 words per minute a reading speed.
Normal reading, understood and remembered, sits near 300. That is the honest table, and it is more useful than the fantasy.
How to use the gearbox
Ask what you want from the text before you open it. The answer sets the gear.
The gist of a news story, fast. An argument you will have to answer in a meeting, slow, with notes. A contract clause with money in it, twice. A novel you love, whatever pleases you. Nobody hands out medals for finishing a good book quickly.
Dense material with a new idea in every sentence has a natural speed and it is not a high one. Push past it and you are moving your eyes, not reading. Worse, you come away believing you know something you do not.
Reading a lot. Verdict: this is the one that works.
Here is the finding the industry buries.
Speed reading courses do sometimes work. Students come out faster than they went in. Rayner and his co-authors say so, and we are not going to call the gain a fraud.
The gain is real. The explanation is not the one on the sales page.
A course runs for 8 to 12 weeks, and over those weeks it makes the student read a great deal of text. That is the active ingredient. The hand sweeps and the chunking drills were passengers.
Reading a lot is the shortcut. There is no other shortcut.
Nobody can charge you for it, which is why nobody puts it on the poster. An hour a day of real reading, kept up for three months, will do more for your speed than every drill on this page stacked together. That is not a motivational line. It is what was happening inside the courses that worked.
Reading widely. Verdict: works, and it is the same finding.
If the bottleneck is meaning, then what slows you down is meaning you do not have.
An unfamiliar word stalls you. An unfamiliar idea stalls you longer. Give a cardiologist a magazine feature on heart disease and she will go through it fast, because she is not learning anything. She is matching the page against what she already holds. Hand her a paper on medieval land law and she will crawl like the rest of us.
Vocabulary and background knowledge are reading speed. Not a metaphor for it. The mechanism is direct. A word you know well comes out of memory faster, so the fixation is shorter, so the page turns sooner.
So read outside your own subject. Read history, read science, read the newspaper that annoys you. You are not improving your eyes. You are cutting the time your brain will need on the words you meet next year.
Attention. Verdict: works, and it is free.
Give the text a purpose. Write down the question you want answered, at the top of the page, before you start. Then read to answer it.
A reader with a purpose is hunting. A reader without one is receiving. Argue with the writer. Guess what the next paragraph says and check whether you were right.
Set the room up while you are at it. Phone in another room, not face down on the table. Tabs closed. Light on the page. Read in blocks and stand up between them, because a tired reader stalls, and stalls are the expensive part of reading.
This sounds like soft advice. It is the hardest working advice on the page, and it costs nothing.
Finger pacing. Verdict: keep it, but know what it does.
Meta-guiding means running a finger, a pen or a hand along the line so your eye follows it. Every course demonstrates it in the first five minutes, because it produces a result on camera.
Here is the honest version.
It holds your attention. That is what it does, and attention is worth having. A moving pointer gives the eye something to lock on to, keeps the pace steady, and stops you drifting back up the page for no reason.
It does not widen your span, because nothing widens your span. It does not lift the ceiling. It cannot make your brain pull meaning out of a word any faster, and that was the ceiling all along.
Two practical notes. Run a pencil down the margin rather than a finger across the text, because a hand on the page hides the words you were trying to read. And it does not suit a screen. If you read on a monitor, drop it and put the time into reading. Screens bring their own trouble, covered on reading on screen versus paper.
Killing your regressions. Verdict: do not.
A regression is a backwards eye movement. You go back over something you already looked at.
Even skilled readers regress on 10 to 15% of their eye movements. Courses sell cards, shutters and gadgets to stop it, and the pitch is that you are reclaiming wasted time.
You are not. Most regressions happen because comprehension failed. A clause opened, another cut across it, and the sentence closed twelve words later. The eye flicks back to hold the halves together. That flick is the repair, and it is the reader doing the job.
Block the eye and you have not fixed the understanding. You have hidden the evidence, and you pay for it in comprehension.
Regressions fall on their own when you understand more. Preview, read widely, pay attention, and you need the trips backwards less often. The full picture is on reading regressions and eye movements.
Reducing subvocalization. Verdict: limited, and it is not the lever.
Subvocalization is the voice in your head saying the words as you read. You cannot remove it, and you should not want to. It carries the abstract words and it holds long sentences together.
You can turn it down. Read with a purpose, read material you know, and the voice thins out to the words that carry the weight.
What it will not do is make you a different reader. The voice is not the ceiling. Meaning is the ceiling. A reader with a silent head and a thin vocabulary still stalls on every third word, and stalls at the same rate as everybody else.
The theory is on subvocalization. The practical version is on how to stop subvocalizing, and it is honest about how little is on offer.
Word flashing apps. Verdict: no.
The apps flash words one at a time in a fixed spot. Rapid serial visual presentation. Your eyes stop moving and the words come to you.
It does speed up word identification. That part is real, and the eye movements it removes were a real cost.
Then it takes two things away. Preview goes, because there is nothing beside the word to preview. And you cannot go back, because the word you wanted has gone. Comprehension suffers, and it suffers worse the longer the text runs.
The apps speed up the one part that was never the constraint on understanding, and they break the two parts that were. The autopsy is on speed reading apps.
Eye gymnastics. Verdict: no.
Figure of eight tracking. Peripheral vision widening. Drills to make the eye muscles quicker.
Saccades are a tenth of your reading time. Perfect the eye and you have won a rounding error. The span you are trying to widen is set by the optics of the retina, not by practice, so there is no rounding error to win in the first place.
Evelyn Wood launched Reading Dynamics in 1959, and most of this comes from her. The hand sweep. The wide vision. The page at a glance. It sounded right. Then somebody measured it.
What to do on Monday
Four things, in order.
Measure. Take the speed reading test and write down both numbers, the speed and the comprehension.
Preview everything before you read it. Ninety seconds. It works from the first page and it costs nothing.
Decide the gear before you start, and let some texts have a slow one.
Then read. A lot, and widely, for months. That is the technique the evidence supports, and it is the only one nobody can sell you.
If you want the drills sequenced and timed, they are on speed reading exercises, and our speed reading course runs them in order. The honest timeline is on how long it takes to learn speed reading.
The reader who gains most is not the one who drills hardest. It is the one who stops working on their eyes and starts working on what they know.
Sources
Rayner, K., Schotter, E. R., Masson, M. E. J., Potter, M. C., & Treiman, R. (2016). So Much to Read, So Little Time: How Do We Read, and Can Speed Reading Help? Psychological Science in the Public Interest, 17(1), 4–34.
Carver, R. P. (1992). Reading rate: Theory, research, and practical implications. Journal of Reading, 36(2), 84–95.
