Quick facts on speed reading

Understanding what doesn't work can help us with what actually does

Could you read three times faster if the words flashed through the same spot, rather than you moving your eyes over the text? That’s the claim of a speed-reading technique called Rapid Serial Visual Presentation (RSVP). The idea goes viral periodically - here’s an Instagram post promoting it, which I was sent at the weekend, for example.

Does it work? Well, if you want to draw your own conclusions, here is the entire text of this article presented at 900 words per minute.

No, it doesn’t work.

The technique itself has been around for over 50 years, so the fact that we’re still reading the old-fashioned way, by moving our eyes, will be proof enough for many of us that there isn’t a clear advantage.

For those that want other evidence, there is plenty of scientific research on RSVP, and speed reading in general. From this, we can take some interesting lessons about what the actual limits on reading speed are, and how reading works.

For most of us, our eye movements are not the limitation on our reading speed. Moving the eyes is probably the first deliberate act each human makes, something we do 3 times a second during our waking lives, and faster than any other movement we make. Getting the visual information into our brains is not where reading slows down.

The real limitation on reading speed is comprehension, such that speed and understanding trade off against each other. If you read too fast you stop understanding or remembering what you’ve read.

One study which directly compared RSVP with reading static text showed that people had the best comprehension when reading the static text (Acklin & Papesh, 2017). Taking the RSVP method and “turning up the dial” to fast and faster speeds, showed that when the speed was the same as normal, reading comprehension was the same. When reading was ‘overclocked’ using RSVP comprehension and memory suffered (cited in Rayner et al’s 2016 review).

Speed-reading in general has been associated with some fantastic claims, of people being able to read at 1000, 2000 or even 20,000 words per minute (at this speed the limiting factor is how quickly you can turn the pages of the book). It’s all bunk. Speed reading claims are plagued by poor measurement practices and bad assessment. Once you test if people understand what they are reading properly, the amazing gains promised by speed-reading techniques and courses turn out to be a mirage.

The whole thing can be summed up in the anecdote about Woody Allen¹ who speed-read War and Peace in an hour and when he was asked about the book could only reply “It’s about some Russians”.

What actually works

Debunking out of the way, there are a few techniques which can actually help improve your reading speed.

First of all, for most of us, the real limiting factor on reading is word-recognition speed. You can improve this by reading more, making yourself more familiar with the words and word arrangements which convey the meaning you try and extract. For the typical adult, with thousands of hours of reading experience already, you’ll need many more hours to get detectable gains in reading speed.

Look, I said there were techniques which worked. I didn’t say they were easy.²

The second technique is to know why you are reading and engage different “gears” depending on what you are reading. If you want to find a specific thing in a text you can scan it until you see a keyword or phrase and then home in on that particular part. Or if you just want the gist of a whole text, maybe skimming it is okay (just bear in mind that you probably do want to end up with something more than Woody Allen got out of War and Peace). And sometimes, for important stuff, you will want to read more slowly than normal. A good scientific paper might take me half an hour to read, but a really good paper will take half a day, because I need to read each word carefully and think about what I’m reading.

The science won’t help you do this kind of reading faster, but it might help reconcile yourself to the speed you have to go when it is worth it.

Finally, one trick which can really reduce reading time is to know when you don’t need to read something at all.

If you’ve really mastered this you won’t get to this point in my post.

To everyone who has, thanks for reading!

---

This newsletter is free for everyone to read and always will be. If you can afford it, feel free to chip in to help me keep writing for everyone: upgrading to a paid subscription (more on why here).

Below, references, further reading and other things I’ve been thinking about.

---

This great, open access, review tells you all you need to know about the psychology of speed-reading:

• 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? How Do We Read, and Can Speed Reading Help? Psychological Science in the Public Interest, 17(1), 4-34. https://doi.org/10.1177/1529100615623267

If you want to try RSVP yourself, here are two demos:

• https://rsvp.n0name.eu/ by Thomas Kolmans (code) - this let’s you upload PDF and EPUB, as well as copy/paste text)

• https://reader.soumendrak.com/ by Soumendra Kumar Sahoo, who has a nice blog about what they learnt vibe coding their RSVP app (“After using it for a day, I realized RSVP isn’t suited for my reading style.”).

This study is referenced in the main post:

• Acklin, D., & Papesh, M. H. (2017). Modern speed-reading apps do not foster reading comprehension. The American journal of psychology, 130(2), 183-199. https://doi.org/10.5406/amerjpsyc.130.2.0183

These are some of the pieces I read for research:

• Claudia Hammond on BBC Future: The tricks that can turn you into a speed reader (2019-12-02)

• BigThink.com: Neuroscience shows that speed reading is bullshit

The wikipedia page has some good links, but is overly credulous with respect to some speed-reading claims:

• https://en.wikipedia.org/wiki/Speed_reading

A skeptical review from 2010, but I don’t believe the science has changed since then:

• Dunning, B. (2010, October 26) Speed Reading. Skeptoid Media. https://skeptoid.com/episodes/229

Other things…

PODCAST: The Cognition Project

Tom Griffiths is a modern legend of Cognitive Science, and has spent over a decade interviewing founders in the field. The first episode is with Jerome Bruner - who had a big life: fighting Nazis, on first name terms with Skinner, led the cognitive revolution, sailed from New York to Oxford to take up a lectureship there, and more! He died age 100 in 2016 (Tom’s interview with him is from 2013).

You can also listen to an interview with Molly Potter who pioneered the RSVP technique which is the subject of today’s newsletter. She talks about the culture of cognitive science in the fifties and sixties, as well as the motivation for her studies.

Link: https://feeds.acast.com/public/shows/the-cognition-project

Catch-up

Other recent posts from me

• The wisdom of deliberative crowds. Pooling answers is good, but collectively agreeing a system is better.

• Interception at 10,000 miles an hour. And what it tells us about the nature of intelligence.

• The effects of political disappointment. Testing how we feel about each other when democracy doesn’t go our way, a work in progress.

• Good bias. Let’s untangle what people mean when they say the B word

• Gambling with research quality. How you get 244 different ways to measure performance on the same test of decision making. And what it means for the reliability of behavioural science

Every prediction market creates a secondary action market

Prediction markets are information aggregation mechanisms, where people provide their estimates of likely outcomes via buying bets on particular outcomes. ACX pointed out in this 2022 post, all information markets create secondary action markets. Indeed, all investments create action markets (i.e. they incentivise you to do things to make sure your investments pay off). That’s why gambling is associated with violence, and here we are, prediction market investors/gamblers threaten journalist:

Emanuel Fabian in The Times of Israel, 2026-03-16: Gamblers trying to win a bet on Polymarket are vowing to kill me if I don’t rewrite an Iran missile story

Me: Prestige dynamics

For the RoRI substack, I’ve written about a paper which models prestige dynamics in economics:

Good investments in the prestige economy

Prestige dynamics are an essential part of scholarly life - for good and for bad. I write about a new analysis from Kurtis Hingl which looks at how the "credibility revolution" in economics affected how that field distributed prestige to researchers, depending on when, and if, they adopted newer, better, methods for making causal inferences.

… And finally

“missing kitten just doesnt care” by @guilherme nunes

END

Comments? Feedback? Cat pics? I am tom@idiolect.org.uk and on Mastodon at @tomstafford@mastodon.online

1. As told by Woody Allen, so probably a gag rather than reporting anything that actually happened

2. Like learning algebra, there’s an easy way and a hard way - and the easy way doesn’t work.