Sorry, You Can’t Speed Read

When I'm reading Hemingway, I'll go back and read individual paragraphs three times to squeeze the last drop of juice out of them. There are only so many Hemingway paragraphs in existence, so wasting one feels like cramming an entire bar of imported Spanish 70% Artisan Chocolate in your mouth and washing it down with soda. Why would you do that???

Take Sordo on the Hill from For Whom the Bell Tolls. It's just a few pages about these guys shooting at each other, but there's so much crammed in there. You learn an awful lot about that guy, his feelings about this war, why he's fighting it, how he ended up on this stupid hill, and why he's going to go ahead and die there rather than any of his other options. He's a real person with complicated motivations and after a few pages you come away knowing pretty much everything you need to know about him and why he was acting the way he was earlier in the book.

Of course you could skim that scene in a minute flat and get the Hollywood Blockbuster version with the crazy local militia yelling stereotypical catch phrases and dying in a blaze of glory.

But then can you really say you read that book? Wouldn't you still want to go back and read it for real some time?

> Doubling the speed, in our experience, leaves individual words perfectly identifiable — but makes it just about impossible to follow the meaning.

Wow, sorry but this is very misinformed. A lot of people me included can listen to audio at 2x. I constantly listen to podcasts (so not a slow narration to begin with) at 2x in English which is not even my native language and I understand everything and enjoy it very much.

Blind users listen at 4x without loss of understanding.

You just need to train it.

> The way to maintain high comprehension and get through text faster is to practice reading and to become a more skilled language user (e.g., through increased vocabulary). This is because language skill is at the heart of reading speed.

So, in short, moving your eyes over the page faster only helps if your speed of language interpretation can keep up (the nytimes article has a slightly paraphrased version).

I wasn't really impressed by the article. It seemed to add very little to the abstract, just a couple of concrete examples of speed reading techniques (and one accompanying study).

Defining speed reading as 'eye movement speedups' and explicitly excluding 'language comprehension speedups' is fine, but it makes me want an article about what happens when you combine both, which I know for a fact can result in reading speedups by a whole number multiple.

i took a full semester class back in high school where all we did was train speed reading. i was able to get to 800 wpm with 90% comprehension on tests. that was from a baseline start of ~150 wpm with 80% comprehension. (the tests afterwards weren't easy "what was the name of bob's dog?" questions)

i've, of course, lost it all in the past 25 years of no practice, but i remember the keys being swallowing entire lines instead of words, achieving "flow", and dropping the subvocalization of what you're reading. (when you pronounce words in your head)

But some text are written to be skimmed (large parts of many college textbooks for example - the author is paid by the word, so there's often a lot of filler. Most people only need to get squinted with most journal article they meet)

It should be possible to get in the zone and not subvocalize, but I can only do it for very brief periods. And it's something that I can only do for "light reading" rather than when trying to understand highly technical information.

[1] https://en.wikipedia.org/wiki/Subvocalization

The big drawback: we can't read anything that fast. We have to format the content in an specific way that we trained with. For example, reading long lines is a big no no. The idea is to maximize the circular area that you could read on a single eye stroke. With training you make the circular area bigger and bigger.

That is, we need to OCR the book or something that let's us create our own book in digital format(tablet) or decrypt the epub-pdf. That is illegal.

Why people love to speed read things like novels? Because normal reading is boring and too slow. It is like being told to play football at 1km/hour max.

Sure you won't get the full literary experience skimming through War and Peace. But fluffy articles with a single tweet's worth of information content I can "speed read" just fine.

To this day, I can get to pages into a book and begin making major assumptions about the rest of the book (usually it will be close enough), so when my assumptions start playing out, I abandon the book, or skip to the last chapter to find out if I was right (more often than not, the trope plays out.) If I were to continue to read it, it becomes more of a chore than something enjoyable. I don't know if this is a result of my speed reading classes, or not...but it's something I just tied together after reading this.

I do enjoy reading history or things based on opinions, biographies, etc; which are much much harder to speed read. But I do wonder, just how much of speed reading is assumption and reliance on tropes...which underscores two problems, I believe, in both speed reading and in writing.

I find that my reading speed is most heavily effected by how much my eyes have to move to scan the words in the line. The best for me is if a column is narrow enough that I don't really have to move my eyes side to side to get all the words in the column.

As an example, a Thompson Chain Reference Bible has about the perfect size columns (roughly 5-7 words per column) for me to be able to take in essentially the whole line at once, so I can just run my finger down the column and take in words about as fast as I can follow it with my eyes.

A related trick in this regard is one I believe I saw on Tim Ferris's blog which (if I remember correctly) suggested focusing on the 6 words at the beginning of a line and then the 6 words at the end of the line. For text that is too long to take in the whole line all at one glance, this seems to work pretty well for me.

I find that I'm able to read and comprehend comfortably at 650 - 700 words per minute as long as the text is well written and on a subject where I already know all the lingo.

[1] http://ds300.github.io/jetzt/

When I read articles without it (which I do most of the time) I read a bit slower. I get distracted much more often. My mind wanders. It feels a bit more like work to make myself read the next sentence.

For the same reason I prefer audio books to written books, and lectures to textbooks (although I prefer to listen to both at a sped up rate.)

The density of the text matters more than anything. You can not speed read a math textbook at even normal reading speed. You can read a very light or uninteresting article at 200 wpm above normal. I can listen to some audio at 2x speed, but sometimes I slow it down to make sure I have time to process everything.

A few hours ago, at 5am, while checking two weeks worth of email, I was very very drowsy and unfocused and couldn't read fast at all. I went to bed, slept a few hours, and now, somewhat refreshed, I can keep in zone again and read at a good pace. Reading speed for me definitely isn't a constant.

When I want to read faster still, I can make a focused effort to do so. Emphasis on focus. The faster advancing eye movement takes focus. As does trying to comprehend the larger amount of words or (short) sentences that the eye consumed in a pass.

Properties of the text have an effect on success; "speed-reading" a math textbook would be a complete waste of time for me, because the comprehension part requires so much more time and effort than "seeing" and "reading" the words. Same goes often goes for code or highly technical texts.

The layout has another effect. Back to the emphasis on focus. Give me the right font size, short lines, and likewise short paragraphs, and I can read much faster without stumbling. Yes, it is about eye movement, about anchors, about breaking the text into conveniently sized chunks that can be parallel processed. If comprehension for the given text is easy enough, then this can help speed up reading a lot. Still, it takes a bit of a focused effort, which I would call speed reading?

I read the linked article pretty fast. And my speed slowed significantly as I got back here, with the long lines stretched to almost full width of my display. First I stumbled a few times, and that was a sign to slow down.

It is fair to say there's a limit to how fast one can read without sacrificing comprehension, that's for sure. But you can totally make a focused effort to read faster.

It depends a bit on the narrator and the content. I find I often need to start with a new narrator at around 1.75 and then bump it up to 2 when I get more comfortable with their voice. But I would say that provided the speaker is clear and not heavily accented my language processing quite happily works at 2x speech speed. (Perhaps they deliberately speek slower for audio books though?).

Whenever my wife hears it she says it sounds like nonsense to her, so I do think it's something that you get used to over time.

I don't read words allowed or to myself in my head. I look at them and move forward. After a few years of doing this, I've been able to go through novels at surprising speeds. It's like I can parse an entire page and visualize everything without trying.

Also, using my finger to guide me only slows me down.

Research [1] suggests that whilst use of Spritz does negatively impact reading comprehension, that impact is relatively minor. My own personal experience of using Spritz agrees with that finding.

I'm not saying there are no downsides to speed reading, but there are times when you do need to read something quickly and you can sacrifice a small amount of reading comprehension. In those cases, speed reading can be immensely useful.

[1] http://sdk.spritzinc.com/www/1.0/psb/Spritz-PSB-study.pdf

I remember when I was in primary school, I overheard an explanation from one of the teachers about the children in a gifted course: "[They're] wrong: They aren't smarter than you, [they] just learn faster than you."

After a week, my memory of my experience of a book is the same whether I λx→x or λy→y, it's just that one of these methods took less time in absolute terms.

I don't really know what the reading experience is like for you, and I don't know what you call these things, but if I experience books faster than you, you want more time in your life with no other downside that I can see, then I can tell you about my experiences.

Now I say this because I'm certainly biased: I didn't enjoy the article, and where's why I think I didn't enjoy the article:

Here's the statements in the article in "support" of λx→x

 • Professor Treiman concluded that it’s extremely unlikely you can greatly improve your reading speed without missing out on a lot of meaning. What exactly do you mean by "greatly" and "a lot"?.

 • Skim readers spent more time reading text that was earlier in the paragraph, toward the top of the page and in an earlier page of the document. These findings were interpreted as evidence in support of a “satisficing” account of skimming process. And?

 • Have you ever tried listening to an audio recording with the speaking rate dialed way up? Yes, I used to read and transcribe legal dictation at +50% and I got done faster. What exactly do you mean by "way up?".

And here's the statements in the article in "support" of λy→y

 • Participants in a 2009 experiment showed reading half the words distributed relatively evenly throughout the text versus losing the beginning half or the end half had better comprehension. No kidding.

 • You can learn to skim strategically so that you spend more time looking where the more important words are likely to be, and if the words are presented in a stream you may be able to learn which words to focus on and which to ignore. Sounds great.

I was unable to extract anything stronger than these statements, but then I was λy→y. What I nonetheless got out of the article, wasn't convincing to me that I, doing λy→y am somehow missing out on something. Instead, I felt like this was attempting to console people who λx→x, and perhaps maybe they shouldn't feel so bad.

In high school, I took a speed reading course from the teacher who was on the speed reading segment on the 80s show "That's Incredible!". By coincidence, she taught at my high school in Southern California. I feel bad for not remembering her name but it's been well over 20 years ago at this point.

Her technique emphasized exercising the eye muscles, improving peripheral vision and focus. The downside to the technique is that it required one to constantly do these eye exercises to maintain speed, although my peripheral reading ability is still there.

It helps if the writing is clear and flowing.

Of course, speed reading gets some meaning lost. But the very skill of speed-reading amounts to developing a sense of look-ahead: first, intuitively establishing fast anchors in text to actually read a few moments later; second, to expand on this anchors, absorb the information and introduce feedback corrections in look-ahead patterns.

All my reading is "speed reading"; I even can't follow the linear text word by word anymore, this is way too boring. But I can regulate the time distance between look-ahead and absorbing passes, and thus the ratio between speed and meaning extracted. This is invaluable to find something in technical documentation, or prepare for exams.

And I can't listen to audiobooks or lectures; they are sooooo slow...

There are free speed reading software available on the internet which you can use to improve your reading speed.

I would like to recommend spreeder.com

I'm a speed reader. It did involve a sheet of paper that had "eye exercises" to practice but this was a tiny part of the class. The vast majority of the study was on scanning and skimming[1] text for information. It's a different way to read technical/programming books and I find it way superior to word-for-wording books. I can read a book this way several times. I can focus (word-for-word) on areas of weakness and skip areas that I know well. I can get through a large volume (with enormous effort) in a weekend with retention that is significantly higher than what I would have experienced if I had read it word-for-word (at least in part because I'd have only gotten through a small part of the book by then if I was reading word-for-word).

It's not magic, but intuitive, when you stop thinking about "speed reading" as a way to read fiction -- or really as anything other than a different way to consume large volumes of information for learning and it's particularly well suited to the huge technical books that plague our industry. Here's how it works in practice (for me; since I can't speak to anyone else):

- First read: 1 hour - a major part of this is simply determining if the book is worth studying. There's no shortage of huge books on complex topics in our industry and many of them aren't effective at teaching the topic they're trying to convey. If I approached reading word-for-word, I might not discover that until 5 hours later or I might miss some useful sections toward the end. In the word-for-word scenario, my choices now are to drop the book and mark that time as wasted or succumb to "loss aversion" and give up on it after several days when I'm exhausted from reading it. This also allows me to eliminate any parts that cover areas I know well, already, which is a huge problem for me when reading programming books. - Second read: Covers only the parts of the book that were identified in the first read as "worth reading". How long this takes depends a lot on the first part. I still mostly scan/skim but with much more attention to detail. I take a lot of notes here. - Subsequent reads: Deep reading (word-for-word) of truly useful material. Here's where I'm probably going to start doing exercises if the text includes any. This may or may not be done in the order presented in the book (especially for programming books which often introduce many concepts, dig into each, then way later cover "advanced techniques" of specific topics -- I hate this practice)

For a book that has 90% "new and useful content" to me, I can get through it in half the time it would have taken me to read word-for-word but with much higher retention. The reason for the higher retention is because when reading word-for-word I've forgotten significant portions of the early parts of the text weeks later when I've reached the latter quarter of the text. Technical books build on material chapter-by-chapter so losing the earlier parts drastically affects my understanding of the latter parts. When I sit down with a book, I block out my time in advance so that I can get through a "step" in my process from start to finish in order to avoid this problem. If I reach a latter part of the book and feel like I'm not understanding things well (I rarely feel "lost" mainly because of this technique), I consult my notes and know where to look for the information I'm missing. It makes consuming information efficient by reducing wasted time being lost, reading things that are redundant, and mentally organizing the information effectively.

On "speed reading apps" and courses, I think there's one point that is extremely misleading. That's the idea that you can use an app or take a course and over night become a speed reader. My course helped quite a lot, almost immediately, but it took 10 years before I was really good at it. This should surprise nobody. Any adult with children learning how to read understands what a complicated process reading actually is. You have to practice and when you read you have to think about your reading technique in addition to consuming the material. Thinking about it in the way I think about running: every read is about reducing the time/mile, reducing the amount of effort required to do a run, keeping your heart rate in the right place and increasing the distance you can run. You can practice running without thinking of these things, but if you're not deliberate about it, your returns on practice will become less and less.

Lastly -- I can't stress this point enough -- it's terrible for fiction. I hate reading any other way and as a result I simply don't read fiction at all. So I stick with audio books -- having a book narrated is the ultimate in word-for-word immersion. Enjoying the formation of sentences and the poetry of the text is impossible for me when speed reading.

[1] As it was described in the course: "Scanning" is reading parts of the page: headings, first paragraph sentences, intro paragraphs, etc. This gets a rough idea of context. "Skimming" I think is best described as "reading words but not sentences". You rapidly read the important words in every sentence. I can blow through a text this way in 10% of the time it takes to read...with about 1% of the information actually reaching me. It's not a way to learn material, it's a way to identify more deeply what the material is about.

"I took a speed-reading course and read War and Peace in twenty minutes. It involves Russia."