This is of interest to me, as I have much practical experience with teaching people to read faster.
I read rather fast, around 500-600 WPM. My comprehension is high, and I am also excellent at "skimming": reading far faster than normal to identify specific parts of the text.
Professionally, I teach people to do better on the law school admission test (LSAT). The exam has a section on reading comprehension. Student must read four dense passages in 35 minutes and answer questions.
Most students complain that they don't have enough time. Invariably, when I test these students, they are reading 200-250 WPM. That's half my speed! At the low end (sub-200) students actually vocalize words - their lips move when they read. Other students, who have enough time, read 280+
For the past year, I have been experimenting with training students to reduce subvocalization and read faster. A sizeable minority of students report very rapid increased, perhaps 60-100 WPM within a week. Perhaps 40%. Many other report improved skimming (important on the test), even if their speed doesn't increase. About half report no improvement, but many of them simply don't try the method, as they are skeptical.
I'm not a believer in 1000+ reading speeds, but I do think many people have a latent capacity to read better. My own results support this, though my research methods leave something to be desired.
Does anyone know of studies that test whether improvements within the normal band of reading speed are possible?
Do you feel that speed reading hurts your enjoyment of fiction? I'm a slow reader and one the one hand I want to improve so that I can read more stuff but on the other hand I feel that it will ruin literature for me. Some sentences lose all impact when you blaze through them at 100KMH.
I can't say if this is generalizable, but it certainly hurts my enjoyment of fiction. I read a metric shitton of fiction. I used to read fiction professionally, in fact. I remain a very slow fiction reader, at least subjectively, based on my assessment of my friends', peers', and family members' fiction-reading speeds. But speed isn't really the point when I read fiction.
Speed reading has its place, but I will never comprehend the desire to apply speed reading to pleasure reading. Those two goals, speed and pleasure, seem orthogonal at best to me.
Speed reading is all well and good when you have more on your desk than you can handle, and you need to get the gist of everything quickly. It's a triage technique.
I'd love to speed read for pleasure. That's because most of my pleasure reading is nonfiction. I don't particularly enjoy the reading; I just like learning new things and having my mind expanded a bit. If I could breeze through nonfiction at 1000wpm with great understanding, I'd be all over that. The ability to do so would be worth thousands of dollars to me.
But in my experience, things like Spritz, Spreeder, or just moving eyes faster are fine when the sentences are short and repetitive: "Most people read at 250wpm. You are now reading at 350 wpm. That is 40% faster than most people. And this isn't even hard, right?" When I try to use such techniques to read about, say, relativity, quantum mechanics, or international development, my comprehension approaches 0.
I have a good reading speed. When I'm caught up in a book, it increases because I want to know what happens next.
The fastest that I ever was timed when I read Clan of the Cave Bear in a single sitting. Afterwards my mother and I worked out that I'd done it at an average of 900 wpm.
So no, I don't find that reading fast hurts my enjoyment of fiction. At least some kinds of fiction.
I had a boss with a very average IQ but who was a competent manager. A couple of times, I flew on business trips with him, and we sat in different parts of the plane. Both times, I would get up to walk around the plane a bit and see him over in his seat going through some recent bestselling novel. Amazingly (to me), I would see him burning through his beach literature at an average rate of about 100 pages/hr, which would be about 600wpm, except that he took breaks. I'd guess that he read at about 700-750wpm. He wasn't trying to impress anyone; he was just sitting by himself reading to pass the time.
This was astonishing to me. I read a lot more than he did and had a much richer vocabulary, yet I wouldn't have been able to follow the story at even half that speed. I eventually asked him about it and reached the conclusion that we read in different ways. I'm bored by books like that and usually read things that I have to think about to understand. If I do read a book like his, I read a bit and drift off thinking about it, then read a bit more and drift off in thought.... He rips through it with so much speed and focus that he apparently "experiences" it like a movie without drifting off into analysis. The speed intensifies the experience for him; it doesn't ruin it.
I've tried to do it but without success. I seem to be trained to chew my food, not inhale it. And he used to tease me about the things I would read for fun, apparently considering them unspeakably boring and nearly indecipherable.
I think there is more diversity in the action of reading than most of us imagine.
The intensification of experience sounds familiar to me. The very average IQ, not so much. ;-)
That said, we undoubtably read different stuff. I like reading fiction that is meant to be enjoyed, and non-fiction that teaches me about various things. But when it comes to philosophy, well, I'm with Dijkstra. About the use of language: it is impossible to sharpen a pencil with a blunt axe. It is equally vain to try to do it with ten blunt axes instead.
I have the opposite experience. I tend to read good fiction -- the nice, trashy, swords-and-sorcery-and-spaceships stuff, not the deep, meaningful War and Peace stuff -- at speeds way above my comprehension level. I enjoy it just fine, my imagination fully engaged, and I'm just accidentally skipping words and sentences and paragraphs in my eagerness to find out what happens next. When I go back and reread a good book, I always encounter material I know I missed on the first time through.
Nonfiction on the other hand -- textbooks and papers and the like, the real information dense stuff, not the light and fluffy pop science books -- I read at a much more plodding pace, carefully reading each word, skipping nothing, and frequently backing up to reread a paragraph.
Does your comprehension suffer when reading at that speed? I can read phrases/short sentences at a time and get a general gist of what is going on. That allows for incredibly fast reading, however it is hopeless if I am reading any kind of technical material.
As another really fast reader: if I get to the end of a sentence/paragraph/page and feel like I must have missed something, I just read it again. If you read really fast, it's not that much of a penalty. It's much like how slow typists will arrow back to correct typos, but fast ones will quickly delete all of the words back to the typo and just retype from there.
This is what I do (I'm OP). Normally, I wouldn't say my comprehension is hurt. But if it is, I just reread. You get a deeper understanding the second time through, and rereading is far faster than reading.
I've zombied myself through a few chapters on countless occasions. Sometimes I get into this mode where I'm just processing information and moving on to the next sentence like it's an entirely new work: process, understand, toss out, go on to next thing.
Sometimes that results in noticing that I don't even remember the last two or three chapters and have to reread.
> Does anyone know of studies that test whether improvements within the normal band of reading speed are possible?
Isn't this trivially true? Did you read 600 wpm at age seven?
I'm not sure I'm willing to call 200 wpms (spending more than a minute reading your ~260 word (according to wc) comment) being "proficient" at reading. Barring any fundamental problems (dyslexia, possibly poor sight) anyone should be able to improve beyond that. If you're reading 200 wpm (with good comprehension) increasing to a rather modest 400 wpm doubles your efficiency! That seems like it should be a worthy goal of any student (because it makes them much more efficient students).
I like it when people mention how it's awesome and how it "totally works"..
I can read fast through light content, but some parts of a text need comprehension, not just to be read. They need a mental effort that you don't exert reading an article about a movie star.
Like many of you, I spend a lot of time reading technical doc. It's not necessarily "complicated", but it certainly is complex. Speed readers don't seem to get the subtlety of this, and when anyone says it works, I'd love to hand them the IEEE articles and the bunch of journals I eat all day long and dare them to just summarize the state of the art part.
As I said, there are parts you can read fast because the content concentration isn't that high, and, thanks to plagiarism in the scientific community, a lot of articles are worded in an uncomfortably "similar" -cough verbatim- way. There are parts though you need to stop. Read, re-read.
You can't speed read parts where a comma matters. Where there are a lot of interconnections between several works, etc..
The reason, I think, some people think speed reading works is that they don't read much and don't read much stuff that matters to be able to tell it doesn't work.
Most people have a short attention span. Heck most wouldn't even be able to read a "paragraph", so I wonder where they had all that "experience" speed reading. That's like claiming to have test driven a Ferrari in a 20 feet track. I'll simply say "Good for you".
Try to post that link to your friends who claim to speed read, most of them won't even read it entirely. Ask them how they liked X part (that doesn't exist). They would probably answer you assuming it's really there. There you go, you proved my point.
My base reading speed for most English-language books and other publications is about 500 words per minute, with good comprehension by test. I have never had any trouble at all finishing all the sections of a standardized test with time to spare, for example, whether the SAT or the GRE or the LSAT. There is a huge published literature on reading skills improvement, for readers of all levels of reading proficiency, and when I was in university I read many books about that topic, including some books that made incredible "speed-reading" claims, to see what I could do to improve my ability to finish my homework while working my way through my university courses. One distillation I have of all the advice I read is that it helps reading speed and comprehension a lot to improve vocabulary. English vocabulary improvement can build on studying common Greek, Latin, and French word roots that show up over and over in English words. Doing that seems to have done the most for my reading speed and comprehension, which was never bad in English. A book I recommend for vocabulary development is English Vocabulary Elements 2nd edition.[1] To bolster my reading speed in other languages, I've also had to focus on reading practice and acquiring reading vocabulary. In Chinese for foreign learners, the materials by the late John DeFrancis, his Chinese Reader series in beginning, intermediate, and advanced levels, are still the best available for that purpose, although they appear to be going out of print.
One thing the anti-speed-reading folks never take into account is comprehension per time. They always look at comprehension for one reading of the material.
My regular reading speed is around 220 wpm. But I can speed read or speed listen at much more than double that. Therefore, I can do TWO speed readings in less time than one slow reading.
So for a fair comparison, you have to look at TWO speed readings vs ONE slow reading. In that case, the speed reading might win on comprehension, even though it was behind after its first reading. Reading stuff twice improves comprehension a lot, and can still be done in less time.
Or if I really really care about something, I would do a slow reading and several speed readings, and I think that's way more effective than doing multiple slow readings, for the same amount of time or less.
Also with speed reading I can do over 500 wpm with very good comprehension. The point where I can't keep up mentally or with my eyes is (after a lot of practice) above 500 wpm. But my regular reading speed, if I just use Kindle app or a paper book, remains under half that.
I can also speed read at over 1000 wpm, for light reading but not dense philosophy. Yes comprehension drops, but two readings at 1000 wpm for light reading may still beat one reading at 500 wpm for the same material. Or FOUR readings at 1000 wpm could beat ONE slow reading, in less time.
It's important to be able to read at many different speeds, and also use several skimming methods, and think of them as different tools in your toolbox, and then figure out which is appropriate for what you want to accomplish. If you always read everything the same way, you're doing it wrong.
> Speed reading anything you need to truly comprehend is probably a bad idea. However, if you have a few documents you need to get through or you're reading something that isn't that important, these methods can still be worthwhile.
If the purpose of reading something is to understand it (aka comprehension), then doesn't this supplied conclusion basically defeat the purpose of anyone trying to attain speed reading?
As in everything, Not Always. Sometimes you just need to get the gist, sometimes the information density is really low, sometimes you just need to get primed so you are more prepared for face to face meeting.
Skimming == Speed Reading, it can be a useful skill even for hard to comprehend documents like scientific papers.
I can "read" a technical book at faster than 1600 WPM.
I don't read all of it, and that makes all the difference.
Most people don't learn Skimming, Scanning, and Skipping.
Skimming is when you go over the page quickly jumping sentence start to start judging the sentence to see if you want/need to read it. If a sentence starts with info you don't need, or already know you jump to the next bit.
Scanning is where you go looking for certain words on the page. I'm reading on Data Storage stuff, and I only care about Raid 5+1 so I scan for those words on the page.
Skipping is like hitting the next chapter button on your Shiny Disc player. If the chapter isn't relevant you skip it. And go on to the next. This is useful for books where 80% is beginner stuff.
This isn't how you "read" Harry Potter. You wouldn't have any fun. So a novel in 90 minutes would have to be for people who don't enjoy reading. I can't imagine that fast. (Like my minds eye, and ear can't do all of the Lord of the Rings movies at 8x)
Like most of us, by which i mean intellectual types who define our selves and worth in part by the relative level of our perceived knowledge, speed reading seems like a holy grail. There's so much out there to read, and not enough time in my life to do it. But it is to us as fad diets/exercise regimes to people who care primarily about their looks/weight.
I say this because its apparent I'm a relatively fast reader, i have to read a lot for work, and i have professional colleagues to compare myself to. In all these speed-reading fads in professional environments, I've never actually met a single person, NOT ONE, who can actually read these materials faster than an intelligent, well read person. (barring perhaps abnormalities like Kim Peek, but newsflash, you know if you're Kim Peek and if you are/aren't, there's not much you can do about it). Get someone into an actual environment where they have to read lots of stuff, have to comprehend it, and its professionally demonstrable, and suddenly all the "speed readers" vanish.
Do you know why I'm a relatively fast reader? I'd say probably: a) genetics b) reading a lot.
b) is about the only thing I've seen that has a big effect and is demonstrable, and is in our control. The fastest readers read a lot. The slowest readers don't. Those who didn't read a lot, and then started reading, got faster.
And barring genetic abnormalities and usual statistical variance, no one I've met, EVER, has been able to read more than 3/4/5/6 hundred words per minute with accurate comprehension.
Which brings us of course, to the comprehension debate. Lets avoid the ridiculousness of the comprehension stats that are usually poorly designed and created by people trying to sell you things, they are worth about as much as fad diet testimonials and figures. And this is where a lot of speed reading salesmen try to get you. "I can read this at 1000 wpm with just slightly less comprehension!". "I can skim and pull out the important parts really fast!".
To which my feelings can be summed up: Anyone can purport to increase their reading speed by including words they didn't read or comprehend in their wpm. Frankly, if you are not reading something with %100 comprehension, you are not reading it. Taking a sample and taking a census are two different things. That you can take a 10% sample in 10% of the time does not make you a "speed-census-taker". Ditto skimming, summarizing, or any other weasel-word used to gloss over the fact that someone is trying to speed up their "reading" by reading or comprehending less.
I'm not saying skimming doesn't exist. I am saying its not the same as reading/comprehending, and that "speed-readers" show heavy drops in comprehension.
My parents signed me up for the Evelyn Wood speed reading course the summer before college. They thought that it might give me an edge over other students.
Speed reading took away all the joy of reading for me. Thought my comprehension was way lower but they give you tests afterward and I scored highly. But studying the tests revealed they were super simple.
The whole idea was to build your confidence and convince you it worked so you'd bring them more customers. There was no way it was going to help me in the real world classroom.
But my parents meant well, Evelyn Wood's marketing was really well done. Just found out the courses are still on the market.
> Frankly, if you are not reading something with %100 comprehension, you are not reading it.
This is too dogmatic to be (IMO) a useful definition for the unadorned word “reading”.
After all, what does it even mean to comprehend something 100%? Without being inside the author’s head (and perhaps not even then) it’s impossible to know all of the ideas associated with every word and phrase by the author at that particular moment, or the precise rhythm and intonation the author would use in saying the words aloud, or the precise feelings the author was trying to evoke.
Or for that matter, is it possible to comprehend something 100% without deeply thinking about what it means to the reader, all the ideas and feelings the words (in context with the rest of the world at the moment of reading) dredge up. For instance, I personally can’t read something in a truly deep way, and really know what I think about it, without writing (often lengthy) comments of my own. But I wouldn’t say that shallower types of reading aren’t still “reading”.
Language inherently distorts thoughts, which are impossible to perfectly package and unpackage in a serial format.
And I think that's a fair criticism. We get down into the details and not much can be said with such certainty. Even measuring comprehension as a metric is fraught with dangers, biases, and questions of how to do it.
As many people are pointing out, anything over and above green eggs and ham often requires several re-reads and some digestion at normal speed.
But if we must talk in generalities (and i think that's appropriate for the medium and context of hacker news), if we want to talk about measurable comprehension as its used in speed-reading discussions, what level of it actually constitutes reading, and place it on a scale of 0-100%, then I think its reasonable to consider actual "reading" as arbitrarily close/adjacent to 100%.
This we actually can measure - speech recognition's comprehension rate is a big problem for its implementation.
So we know 80% accuracy is not good enough to produce cogent sentences (1 miss in every 5 words). And we know 1 miss in every 10 still doesn't produce something we'd think of as accurate.
So I'd say you need about 90%+ comprehension to actually comprehend something. But that's leaving a lot of room for error anyway - how much stuff have you read and then thought "wait, what was the modifier to that specific phrase?"
I'm with you on the skepticism thing. When we read, at least when we read anything important, we aren't simply ingesting bits of information. We're usually reading something that was carefully crafted to have a logical flow, where what is presented early supports or leads to what is said later. To understand the structure of a written piece is not easy. For "dense" works (e.g., some philosophy) it might take reading and re-reading several times, jumping forward and backward, to fully figure out what is being said, whether it makes sense, whether it's to be taken as reliable or not. Even that might not be enough. The idea of speedreading for the "main points" is ridiculous. Even less dense works have a logical flow, they aren't just a list of bits of information. If you're going to read critically (i.e, with good comprehension), which is generally the proper way to read, speedreading is useless.
> Frankly, if you are not reading something with %100 comprehension, you are not reading it.
No one reads anything with 100% comprehension, and the assertion that such a thing as perfect understanding exists is absurd. Real meaning of words is fleeting and there certainly isn't any universal truth to every written work.
These new-fad speed reading tools are an alternative method of presenting text, the comprehension of which is a function of the speed setting. At what speed does reading become not reading and something else entirely? I say such a limit doesn't exist.
Maybe such tools aren't for you, not everybody is going to like them but some people find them useful. There certainly will be people that overstate their usefulness, but in the end a perfect scientific description of it isn't necessary.
A tool is useful if the person using it finds it useful, not if somebody can scientifically prove it.
You're a skeptic, so what?
I find it to be a useful tool, not for all reading but certainly for longer blocks of text or books which 100% understanding is less important than completion. I find for me it is easier to stay focused reading more words for longer periods, especially for things which I don't find particularly engaging or valuable, or when my mind isn't running at peak. I'm sure lots of others will find usefulness in it too.
You can speed read some things because they are highly redundant and mostly puffery. You can't speed-read an article which cares more about a clear and concise argument than padding itself with lots of fluff.
Of course, you're more likely to run into the former. The education system seems to lean toward articles people pretend to write, written for people who pretend to read them. I guess speed reading is useful, if it teaches you what teachers will be using to read your work. You'll know why it's so important to put your main argument in the topic sentences; and why flaccid padding in the rest is OK.
> A tool is useful if the person using it finds it useful, not if somebody can scientifically prove it.
This is a horrible criteria. People found blood letting useful as a treatment.
>No one reads anything with 100% comprehension, and the assertion that such a thing as perfect understanding exists is absurd. Real meaning of words is fleeting and there certainly isn't any universal truth to every written work.
"meaning of words is fleeting"
Maybe over long periods of time words change their meaning. That doesn't mean you can't fully comprehend SOME writing had you slowed down.
I think a core premise of the parent argument is the existence of '100% comprehension' when reading and believe arguments against this premise are very much not directed towards a strawman.
Sure the meaning of words flows with time, but more importantly our languages are imprecise and subject to much interpretation. How any one person translates the meaning of a word or passage into internal conscious and unconscious understanding is most certainly inconsistent. Ask any hundred academics the meaning of a passage of literature or a poem and you'll get 100 analyses. Ask 100 judges to make a determination of how the law applies to a case and you'll get 100 different judgements (even though the law is supposed to be very precise language by design).
My position is that complete comprehension is absurd therefore all comprehension is incomplete.
Given all comprehension is incomplete, arbitrarily setting limits on levels of comprehension based on tools and reading speed is absurd.
Any comprehension is useful, and doing things quickly is useful therefore a tool that provides some comprehension over meaningfully shorter periods of time for a given block of text has definite utility in some circumstances.
A person can be skeptical of the utility of a tool for their own purposes, but the absolutist opinion of the parent is a silly overstatement of that skepticism.
There's another aspect to the question - how engaging or how people feel about speed reading when it comes to being immersed in a story (assuming you read a novel with that technique, for example). Most of the Speed Reading demos out there focus on reading short paragraphs and that's it. Even if comprehension is not a problem, how fast and how feelings develop in your brain as you read may not be comparable between traditional reading and speed reading. That can make a huge difference on how much you enjoy the piece of text you are reading, not matter if you can still get 100% of it.
There are different levels of comprehension, ranging from internalising everything word for word to the point where you could recite it to high speed scans of the page for a particular word that don't comprehending any of the content. All are useful in different contexts. I don't need "accurate comprehension" to determine which HN articles I might wish to click on.
Ultimately a person incapable of progressing beyond verbalising everything to ensure they internalise every minute, insignificant detail and all the excess verbiage is unequivocally worse off than someone that can in the same period of time grasp the meaning and pertinent details of three texts by parsing them at high speed.
Frankly, I don't understand your point better by slowing down to marvel at your sentence construction[1], and even if I did, my recollection of what you'd written tomorrow would be exactly the same regardless of how quickly I skimmed it.
[1]though just in case I read a second time to see what I'd missed and noticed I'd subconsciously picked up on you beginning a sentence with "Frankly" when drafting my reply...
I don't know. I'd say I read your comment, even though I skipped some parts since you were being repetitive.
For example, one follow-up quoted "Taking a sample and taking a census are two different things", which was from a part of your text I explicitly skipped. (I kept track of what parts I was ignoring).
I would still say I read your comment even though I explicitly skipped that part, did not even alight over the area. I literally 'missed it'.
Likewise, if you look through my posting history I've written plenty that is a bit rambling. I don't expect my readers to read every word. Hell, you probably only need to alight on 20% of the words in this comment to get what I'm talking about...
I don't know, I mean what is "speed reading"? Whenever I "informally" clock myself, I read at around 600 wpm +/- 100 wpm or so. If that's "just reading" surely "speed reading" should be significantly faster, say 1200 wpm at the very least?
[edit: Hm, that might have been a bit optimistic, reading the article I clocked myself at ~420. I've never really tried to time myself on a dense paper -- say the first time I read Fielding's dissertation[1] for example. I doubt I'd have read that at the same speed as a somewhat vapid article on speed reading (as the author suggest, although I'm a little taken aback that email is grouped with news as things not to read carefully -- I guess it's a sadly accurate picture of the current state of most peoples online discourse. Sinser[2] would like a word with you! ;-)
[1] "Architectural Styles and
the Design of Network-based Software Architectures" (aka that ReST thing):
I'd say 1200 wpm is below skimming speed -- and there is something in-between -- but I'm not convinced that it's terribly useful as a skill. Being able to skim an article at great speed can be useful, for deciding if it's worth reading or not. I'm not convinced "speed reading" is very useful. The very term implies you speed through material at a speed greater than comprehension (which I'll loosely equate to "regular reading speed").
I do think it can be useful to train oneself to read faster -- by reading more, and occasionally by "forcing" oneself to read faster. For those that already read at reasonable speeds (~500 wpm?) -- it's probably useful to pick up tricks for skimming, and "backwards reading": starting with the conclusion, and reading articles backwards (conclusion, discussion, introduction -- or: conclusion, introduction, discussion -- and either way, probably conclusion again).
Speed reading is like speed sex... Choose between 1 hour with 1 person or 1 minute with 60 people? Okay I get that you might prefer 20 minutes with 3 people and that's fair.
Everyone can learn to read faster and everyone can claim to understand what they read. However, is that all there is to reading? There are things like imagery, sounds, feelings, etc and they can be invaluable both for the enjoyment axis and for the comprehension axis. Not to mention that I couldn't care less about your comprehension 5 minutes after the fact. Come back to me in a week and tell me how much you remember.
But let's look at these often forgotten parts of reading!
Imagery. IRL you can become over-whelmed by visual stimuli. This means there's a limited bandwidth there. Since we're mostly visual creatures there's a lot of bandwidth there but it's still quite finite(If you want convincing participate in one of those visual research experiments). When you read faster, you should expect more of this bandwidth to be taken up. As a consequence your brain will take shortcuts to minimize load and stress. It may feel like you're seeing it all but you probably aren't. It's the fast system of your brain working. The slow part of your system is taking a vacation.
Yes, you can improve your temporal resolution but not by much. Sounds are a lot like imagery.
Emotions. We are bio-chemical systems and there's a time scale over which emotions happen. They won't take place any faster. If you are like me and you care about emotions when reading, you will slow down (All I get from speed-reading is a fast-beating heart and stress quite like the stress in one of those visual labs).
Consider this. Emotions are very important for long-term memory (which is probably the most important part of reading some kind of technical text). By reading very fast you are cutting your emotions short if they happen at all. You effectively average out your emotions. It might not feel like it because you are still experiencing the peaks and valleys but they are smaller and you'll never know it. Memories are formed better when the emotions are stronger.
Speed-reading is just another phenomenon in a series of maximizations along one dimension by disregarding full the effects it has the other dimensions. Other phenomenons are known as texting, facebooking, etc. Okay, these kinds of things have their place but the more we do them, the easier it is to let them take over our lives. Our brains are not designed that way. I just find it hard to justify calling it "I read x wps" when it's more so "I lossy-scan the page and get the jist of it while stressing over my speed".
Having tried out various of the recent speed reading sites and apps, I honestly don't think they are useful for any dense or difficult material requiring thought to comprehend. I almost always find that challenging text requires nonlinear reading to really get, something not possible with these tools. Perhaps they are useful for fluffy news articles, but not much more.
Does anyone know of any good techniques for improving time to comprehension when reading technical work? I feel like I am very inefficient at getting through math or computer science papers, which obviously resist any kind of normal speed reading.
Instead of increasing bandwidth (since human CPU's aren't following Moore's law), burden writers with improving value per word. Please read faster so I can be less elegant?? How about you stop with 1000 word pieces that can be summarized in 100. Oh, you can't slot enough ads in?
I have to defend speed reading. It is in fact highly useful, and I will claim that it can even aid comprehension.
I have two examples.
(1) Last night I went to a meetup to discuss a certain topic related to religion, atheism, and a certain idea explained by Alain de Botton. So I decided to check out his book about it.
I grabbed my notebook, bought the book on Kindle, and skimmed it in 30 minutes, making notes of what struck me as interesting fodder for discussion. It's a pretty short and breezy book, so this was quite enjoyable.
Then I realized that Erich Fromm probably has something to say about this, so I googled a bit and found that he wrote a book about "Religion and Psychoanalysis," which seemed highly relevant from the preview. So I did the same thing with that.
Before the event I had dinner while looking through my notes and thinking about it. This all enabled me to be decently prepared, get a good start on the material, bring some quotes to the table, and so on.
(2) I do a weekly blog series about museum visits, for fun and to practice my writing. Sometimes I'm pretty clueless about the topic at hand, so in order to keep the blog from becoming totally vague and uninteresting, I skim a book or two. This lets me engage with the museum stuff in a more interesting way. For example last weekend I skimmed Deleuze's book about Francis Bacon, and found a couple of great ideas to build on.
So why would I say that speed reading can aid comprehension? Basically, books are tedious. I mean, how many books do you own that you haven't gotten through or even started? Speed reading, especially in combination with note taking, is a way to start engaging with material without investing a huge amount of time and attention, which are extremely scarce resources for a full time worker.
Maybe I'm just saying it's better than nothing. But I also really feel like the sheer velocity of speed reading or skimming helps me somehow. I don't get stuck as easily. I don't feel the obsessive need to "grok" everything. I don't get bogged down.
Of course this is a staple technique for people in academia, and you don't really need any special tools for it. E-book highlighting can be very useful, but manual note taking works great too.
See, this is the first realistic comment I've seen that matches my idea of "speed reading".
The OP, other articles in the last few days, and several comments seem to rest on a misunderstanding of the term "speed reading". It is not the same as "reading faster", nor is it a method of inhaling the same amount of content in a shorter time. It is, instead, a means of grabbing the key points of a large document in a short space of time, either before studying it properly (so you know its structure and end result), or after studying it properly (so you can quickly remind yourself of the main points).
Exactly! Phillip Lopate has a short chapter on "research" in "To Show and to Tell," and describes his way of researching like this:
"Here the creative nonfiction writer can follow the journalists' lead. Being trained generalists—that is to say, quick studies who can leap opportunistically on intriguing vignettes and facts, give them a vivid twist, and forget the rest—veteran journalists know that they don't have to become specialists, they just have to absorb enough of the material under scrutiny this week or month to file an interesting story. When you are researching, what you are looking for, subconsciously or not, is the oddity that will spark your imagination—not necessarily the most important detail, but the one that will excite your love of paradox or sense of humor."
This kind of "opportunistic" reading is discredited throughout this HN thread as "not real reading." I get the point—it's not the same as reading carefully, reverently, leisurely. It's a workaday tactic, it's somewhat impious. But for me it's a wonderful and even liberating way to work with texts.
i call selection bias on this. he keeps saying studies have shown. i always use that terminology when i want to prove a point, and i want to shut up the other side without arguing
the linked documents are the book the causes of high and low reading[1]. and the paper by rayner[2]
i don't really have access to most papers, but from my quick search on google scholar i'd say there are way more than a handful of papers on reading comprehension at high speeds.
also, don't forget that you can train the human brain like a muscle. so it's not surprising that reading comprehension on untrained people immediately suffers when you move outside of their comfort zone.
another thing that's ignored is what graeme mentioned about subvocalization. a lot of the speed reading practice aims to remove subvocalization from your reading process. it's the crap they teach us in school.
if you ask me these are all hacks though. our encoding ie. language and text is incredibly inefficient. in theory chinese characters are a much better at encoding, ie. have higher entropy than our english language.
ideally someone would throw a couple million at me, and a couple of neuroscientists, and nano engineers, and we build you the ultimate reading hack(team applications welcome :P)
> my quick search on google scholar i'd say there are way more than a handful of papers on reading comprehension at high speeds.
The Rayner paper is what is called a review. It has 19 pages of references to hundreds of papers on the topic that Rayner has read so the nonexpert does not have to.
Generally, reviews are much better sources than individual papers for the non-expert to read because the author has gone through the literature to understand the feuds and arguments and how they have played out. In my field, I regularly have people try to send me papers that were thoroughly thrashed as wrong over a decade ago. That is generally what happens when a non-expert tries to dive into the primary literature blindly without understanding the general framework and consensus that a good review provides.
All that said, it is very hard for a non-expert to tell whether a review is good or not. It is a conundrum.
I used to speed read fiction. I found that I would get a couple pages ahead of my understanding. As I aged, I realized that understanding a few books deeply is much more important than understanding a lot of books superficially.
Why read Hunter S. Thomson if you can't pull out the few awesome lines. When I raced through Fear and Loathing in Las Vegas, I missed "And that, I think, was the handle—that sense of inevitable victory over the forces of Old and Evil. Not in any mean or military sense; we didn’t need that. Our energy would simply prevail. There was no point in fighting—on our side or theirs. We had all the momentum; we were riding the crest of a high and beautiful wave. . . .
So now, less than five years later, you can go up on a steep hill in Las Vegas and look West, and with the right kind of eyes you can almost see the high-water mark—that place where the wave finally broke and rolled back"
How many more of these did I miss? And this isn't even deep literature.
I am dyslexic, and have struggled with reading all my life. Saccades are physically exhausting for me, and so I get fatigued quickly. I also subvocalise, which slows me down to 180 wpm or thereabouts.
RSVP is helpful because it removes the need for saccades, and so reduce fatigue. They also force you to stop subvocalising. As a result, I am able to read at 600 to 700 with high comprehension now. And over 1,000 to 1,200 wpm with enough comprehension to be useful for things I just need the gist of.
1,200 wpm. It's crazy. I never thought it would be possible. But I suspect that if I continue to use RSVP, it will have a profound effect on my life. There is so much stuff I do not read, because I find the normal experience so painful.
I swear, I will never understand the obsession with speed reading.
Just like your coding speed has little to do with your typing speed (and everything to do with your thinking speed), your reading speed has little to do with reading, and everything to do with the level of comprehension you're looking for.
Need to get the gist of something? Skim it quickly. You don't need "speed reading techniques" for that. Need to understand it better? Read it slower. Great literature? Read it word for word to savor the language.
None of it has anything to do with reading itself. It's just a question of how quickly you can integrate new information into your brain, and how much of it you want to integrate. And nothing's going to change that, unless you've figured out a way to change your IQ. Except maybe some coffee or a good night's sleep.
I increased my reading speed by 50% in one semester of graduate school doing one thing. No Music, No computer and No video. I use to read with music and then I just tried reading 100 pages with music (No vocals mostly Jazz and Classical) and 100 pages without music and I was at least 20% faster. I ended up getting almost 50% faster do to my work load.
Was a graduate student in Theology. On average each class was around 4,000 - 6,000 pages of reading a semester. With 4 classes you could have 24,000 pages to be read. That is around 500 pages a week, not including research and papers etc.
Historical Theology and Seminar classes have been on average around 10,000 pages of reading. I have had to read over 700+ pages for one class in a week. The issue was I HAD to read and comprehend. This was dense stuff with complex context and thought.
To... read between the lines... on this summary, I think The Zen of Python is applicable here. Readability counts.
One technique that might work would be to "translate" content into a simpler language which is optimized for fast serial consumption with high comprehension. Sure, that's probably impossible to achieve with today's technology, and there's the issue of internationalization (is it a version of an exisiting language with strict rules, or a whole new language altogether?), but those are "just" engineering problems, right?
Or maybe the inverse is really the problem - maybe it's not the language, but the metric we're focusing on. WPM is as poor of a metric for "rate of information intake" as LOC is for programmer productivity.
This is a complete aside, but am I the only one who finds it impossible to read this piece and not be completely distracted by a sudden acute awareness of how I'm reading it?
I for one I'm grateful I came across a speed reading book while in high school. I was reading about 150 wpm by then, after a month going through the exercises I had increased it to around 450 wpm.
Comprehension is the biggest question from skeptics and it is understandable: we are taught to read out loud so our reading speed and comprehension becomes limited to the speed of speech.
Re-learning read (and comprehend) without vocalizing the words in your mind is the hardest part. Once you get that, the sky is the limit.
So, here's something that bugs me. I know I don't read very fast (I got 160 WPM on the Staples.com test), but I find that I read faster than some of my friends and classmates. My reading speed is supposedly slightly higher than that of a third grader, to which I call baloney. So, I HIGHLY doubt that the average college student naturally reads at 450 WPM as stated by the Staples test.
I just don't understand, but I would love some insight on this.
I do not mean to be difficult but what does this mean:
> In the case of Tim Ferriss' technique, he's using ideas grounded in
> science, but I couldn't find research beyond Ferriss' own claims on
> his blog post.
What are "ideas grounded in science" that do not have some body of research/experimental-validation? Am I reading it too literally?
I read pretty quickly but have never expressly timed it (400WPM on Spritz seems a bit slow anecdotally, but I don't comprehend Spritz even on low speeds).
I think what I do is read pairs/tuples of words at once, and then move on to the next set.
Eg, to quote graeme - I read [This is of interest to me,] <move eyes> [as I have much practical experience] <move eyes> [with teaching people to read faster].
I have an RSVP bookmarklet that lets me read selected text at 500 wpm. What is interesting about it is that adaptation is a serious problem. If you don't move your eyes, your vision fades. So I find myself having to consciously have to make small eye movements to prevent that. But other then that it sort of works.
that's how i 'speed read' (note: its not voluntary, i just do that naturally):
- i go through paragraphs very fast, condensing the content and getting the keywords in my head. I read and comprehend about 30% of the paragraph
- if it looks interesting my attention will be caught and ill start reading 100% of the words, rather slowly. in fact, i'll probably start from the beginning. Lot of lines, not much content? That will NOT catch my attention.
Of course, that also means i don't read 99% of the blog posts on HN that are more than a page long anyway. These tend to have a very VERY low content to amount of lines ratio.
I think I'll go even further: people get seduced by thinking they can get "smarter" by ingesting more content, and fall into the speed reading trap. Kinda sad.
The most crucial downside of speed reading for me is that it destroys the tone of a phrase. Splitting a sentence into blocks makes it sound robotic, eliminating the nuances of the author, and quite possibly the intended meaning.
I don't speed read, but I've learned to read faster and faster by just keep doing it. When I was a little kid I used to read a lot faster because I was reading more books back then.
One trick my teacher taught me that's not really speed reading but to get an idea of an essay or an article is to read the first and last sentence of each paragraph.
I read rather fast, around 500-600 WPM. My comprehension is high, and I am also excellent at "skimming": reading far faster than normal to identify specific parts of the text.
Professionally, I teach people to do better on the law school admission test (LSAT). The exam has a section on reading comprehension. Student must read four dense passages in 35 minutes and answer questions.
Most students complain that they don't have enough time. Invariably, when I test these students, they are reading 200-250 WPM. That's half my speed! At the low end (sub-200) students actually vocalize words - their lips move when they read. Other students, who have enough time, read 280+
For the past year, I have been experimenting with training students to reduce subvocalization and read faster. A sizeable minority of students report very rapid increased, perhaps 60-100 WPM within a week. Perhaps 40%. Many other report improved skimming (important on the test), even if their speed doesn't increase. About half report no improvement, but many of them simply don't try the method, as they are skeptical.
I'm not a believer in 1000+ reading speeds, but I do think many people have a latent capacity to read better. My own results support this, though my research methods leave something to be desired.
Does anyone know of studies that test whether improvements within the normal band of reading speed are possible?