Ask HN: Do you guys use speed reading techniques?
There are more books worth of reading than time to do it.
And since improvement on any area (hacking, swimming, etc) requires improved techniques and practice.
I was wondering what speed reading techniques, software, courses, etc. do people in HN use and recommend?
44 comments
[ 5.7 ms ] story [ 96.3 ms ] threadThis obviously doesn't work for novels or dense technical reading, but it is very effective otherwise.
the basic gist of it is this: instead of focusing with your eyes on the start of every line and reading to the end, try and start indented in and read the first word with your peripheral vision. end the line on the second to last word in the same way. as you get good, you progress with this technique.
it does truly work, just using the basics of it speed up read time for me substantially.
however, it makes it hard to use your imagination, so if you're reading for enjoyment, don't do this.
Skimming over dull parts and reading the start and end of paragraphs helps. I've also tried EyeQ and AceReader and felt the latter helped in learning to read chunks of words at a time.
Personally though, the biggest difference was made by starting to use a Sony Reader. It'll make a lot of your otherwise mundanely spent time usable for reading.
The greatest value I gain from books is in identifying how the subject discussed can apply to my own circumstances. That insight often comes to me in the pauses between pages and paragraphs where I find myself thinking on what was just read.
I might be able to absorb a bit more information from speed-reading, but I don't think I could process and internalize the text nearly as well. Since that's why I'm reading the book in the first place, speed-reading seems of little benefit to me.
IMHO most of the content in the books (especially non-fiction) is of no use to me. It's the 5% of insight that is added to my life that adds value and that I really want to think about.
Case in point, I'm trying to learn financial math. I get the concepts pretty well (just parabolic PDEs and similar things), but it would be helpful if I could more easily remember what the boundary conditions for "down and out barrier options" were or what the exact definition of "VaR" is.
Speed reading could help here, though I don't know how to do it.
Read a number of articles on the subject. Write out the definition in your own words or explain it to a friend. Complete a few textbook exercises to calculate the values you want to understand.
Taking the time to internalize a subject is significantly more helpful than quickly scanning words you do not yet understand.
A more traditional programming example is this; if you learn Java you need to understand, but learning the libraries is more a matter of remembering than understanding.
Some subjects are broad, not simply deep. Speed reading (or whatever) could help with the breadth.
Sometimes reading is the best way to learn.
Sometimes re-reading is even better. Do this slowly.
Sometimes reading lots of different sources about the same topic is the best. Do this quickly.
Other times, I learn by listening. Or watching. Or trying. And more often than not, by failing.
Like anything else: Speed reading is a tool. Use it appropriately.
When I read, I just want to read like I naturally do. When I interact with people, I think it may make me act awkwardly if I try to consciously apply some techniques. Of course as in anything there may be a learning curve, after which it pays off to have expended the effort. I suspect I might have been burned by a shortcut technique like this before, although I might never be sure.
Before moving to Japan, I started studying the 3000 kanji characters that are required to be able to read properly. Of course this seems like an utterly monumental task, but never fear someone suggested to me this clever hack called the Heisig method. Instead of learning the characters and how to pronounce them by rote, the Heisig method splits the characters into subparts, and the whole learning process into recognition and pronunciation parts. Like in speed reading, the method tries to make you conscious of your learning process.
Sounds great, right? Perhaps I didn't try hard enough, but in my two years in Japan I was not able to complete that book. Even after coming back, I would still open it and try to proceed, but somehow it feels wrong. Now when I look back, I notice that the actual characters I remember are not from Heisig, but from the rote exercises or the practical usage of characters in our classrooms. This makes me suspect it may be better to let subconscious remain as such, and just concentrate instead on practice. Checking characters as you read, writing them in essays and emails. Interacting with a lot of people to get more comfortable at it. Reading tons of books to become a better reader. When you're really into a book, you might find yourself going faster just to discover what will happen next. Let your subconscious take care of the details.
There is some interesting academic literature on the correlation between perceived intelligence and IQ. I read one paper where they identified the factors behind perceived intelligence, and then tried to get people to fake them to increase their perceived intelligence. IIRC the only factor they were able to successfully fake was eye contact, and the rest either had no effect or else made the people look less intelligent.
As for reading, my biggest problem is that every time I get excited by something I stand up and start pacing back and forth across the kitchen thinking about it. I don't think speed reading will help much with that.
The speed-readers did fine. The control group was confused because every other line was from a different text (two base texts).
http://www.amazon.com/Reading-Power-Flexibility-Sparks-Johns...
which was a refreshing change of emphasis from most other speed-reading books.
Good techniques I learned from various sources were pre-reading (for example, making sure to read the whole table of contents, the whole preface/introduction/foreword, and even the whole index before starting the book proper); focused vocabulary development targeting words with Latin and Greek roots used in the international scientific vocabulary; and daring not to read a whole book if reading one section of it would answer my question.
Good vocabulary development books are
http://www.amazon.com/English-Vocabulary-Elements-Keith-Denn...
and
http://www.amazon.com/English-Words-Latin-Greek-Elements/dp/...
http://www.zapreader.com/reader/index.php
Some older discussion on this is here:
http://news.ycombinator.com/item?id=156464
javascript:var%20sel%20=%20window.getSelection?%20window.getSelection()%20:%20document.getSelection?%20document.getSelection()%20:%20document.selection.createRange().text;%20sel%20+=%20'';sel%20=%20sel.replace(/'/g,'&'%20+%20'apos;');newdoc%20=%20open().document;newdoc.write(%22<BODY><FORM%20ACTION='http://www.spreeder.com/'%20METHOD='POST'><INPUT%20T...);newdoc.forms[0].submit();
(Edit: Might be better to get the link from here - http://spreeder.com/bookmarklet.php )
http://www.slate.com/id/74766/
Intriguingly, this legend was greatly encouraged by JFK's campaign staff:
http://www.slate.com/id/74766/sidebar/74768/
I read the first article you linked to. I measured myself on it and got 630 wpm. (And I backtracked a couple times during the reading process. It's hard to focus when you know you're being measured)
I'm by 40-50% the fastest reader I know, except for my father (we're about equal; I've never measured much, just read over his shoulder). When other people read over my shoulder they rarely get farther than through the left page before I'm turning it; most of them only get through 2/3 of the left page.
In a line such as the following (from the first linked article):
My eyes seem to fix in four or five spots (I can't tell for sure): "several", "this", "and", "seconds", and "you". So I think I can instantly read about 3 words at a time. When I stare at "called" I can read the whole phrase from 'this' to 'fixation' without moving my eyes. I get the sense that other people only read one word at a time. Is that true for you?(Edit: In a monospace font like it appears here, I seem to fixate on every other word, so it's more like 8 fixations across the line. The tighter font on the original page seems to let me read faster.)
I was reading before I turned 2 years old, which is pretty young, I think. I am guessing this had something to do with how fast I read, although genetics may also be related.
Edit 2: I should point out that I've never tried to study speed reading or improve my reading rate. I did play around with Spreeder and I could read every word at 1200wpm, but my comprehension was in the toilet. I played around with setting it at 800 (with 4-5 words per screen) and it worked reasonably well, although I never fully got past the distracting element of it flashing lots of stuff in my face while I was trying to concentrate.
I'm on the other end of the spectrum -- originally grew up in a Punjabi-speaking household (immigrant family in The States), moved to English around the time of Kindergarten.
I think it's got to do with the innate redundancy in grammar. You can focus on blocks of text and only get partial data from each, but reconstruct the entire thing in your mind in real time. With practice, of course.
> ...you end up with about 95 percent of all college-level readers reading between 200 and 400 words per minute.
So you both might be in the <5% group who reads faster than that.
"Studies show that people who read at or above the college level all read at about the same speed when they read for pleasure."
"the fastest college-level reader will read, at best, twice as fast as the slowest college-level reader."
The article isn't talking about 95% of college-level readers, it's talking about the best college-level readers. The above posts (as well as my observations outlined below) refute it quite well.
I am able (obviously) to skim read things quickly - I would read slower when reading for pleasure though.
I've never tried speed reading techniques, mainly because the one person I talked to about it said they found it hard to 'switch off' speed reading. e.g. I was asking him about 'Watchmen', and he couldn't remember some of the stuff I was talking about - he thought it was because he speed-read it.
I'd rather enjoy reading than focus purely on speed.
My wife's story sounds much like lincolnq's. She taught herself to read by the time she was two, and didn't use the clunky phonetic system that most of us (myself included) are taught in school. The article suggests that these people aren't really reading, but in fact skimming. Well, if you define reading as saying each word out loud in your head, then no, they aren't reading. They're assimilating the content in a much more efficient manner that doesn't result in the loss of comprehension suggested by the word "skimming". I don't know about you, but that meets my definition of the word "reading".
I love to read, and when I read for fun I start at the beginning and read to the end. This doesn't work when you have 500 pages to read spread across 4 classes.
Assuming you know what book you're reading and why you are reading it, and that you've gone through the table of contents so you know the shape of the book, let's say you're starting a new chapter:
Read the intro Read the conclusion *Read section headings and topic paragraphs -Hopefully now you know where you need to read more carefully and what you can skim
Sometimes you'll get sucked into what you are reading because it is so damn interesting. You could view this as negative, but I think that is positive. If you are really loving what you're reading, you're going to learn it better than you will if you don't care.
Anyway, I've given some thought to bondafide speed reading, but the above works well for me when I need to get through a bunch of pages and know I need to learn it well.
edit: I've gone back and read your question, sorry if my post response wasn't what you were looking for at all.
But I do speed-read code. When coding, I am bottlenecked not by my ability to program, but by my ability to come up with ideas. And one of the best ways to come up with ideas is to read other peoples' code.
So, as a video encoder developer, I've gotten into a habit of reading through the entire codebase of other video encoders and decoders. I've gotten it down to almost a science where I can sprint through an entire codebase--understanding most of the basic structure and spotting anything "interesting"--at a few hundred lines per minute. I went through the entire libavcodec H.264 encoder proposal ( http://research.edm.uhasselt.be/~h264/ ) in just 7 minutes. I went back later and spent over an hour reading it--and found I missed absolutely nothing of note. And then I stole its strategy for level-code VLC tables for x264.
Other codebases I've read include most of libavcodec's MPEG-related code, dirac, schroedinger, a lot of Intel IPP stuff, libmpeg2, and some proprietary stuff I've had access to from time to time. And probably lots of stuff I forgot.
It is quite easy to speed-read code, of course, if you already know everything the code is going to do, and you're only interested in the implementation or algorithm.
Another thing I do is read changelogs going back years (especially svn/git logs). They often offer even more insight than the code itself.
"The best way to read quickly is to read lots. And lots. And to have started a long time ago."
http://www.marginalrevolution.com/marginalrevolution/2006/12...
I do find that for material that's heavier going my retention is worse than I'd like it to be. I don't know whether that's just because I'm unrealistically optimistic about what ought to be achievable, or whether my reading speed is tuned for easier stuff because, e.g., in those formative childhood years most of what you read is relatively easy, syntactically at least. (For stuff that's conceptually difficult, the prescription is the same whatever your reading speed: put the book down and think/scribble/experiment, read multiple times, force yourself to express the key ideas in your own words, etc.)
Especially when I'm trying to cover a new field (right now, dsp and speech recognition) I just follow a scorched earth reading process where I read a few dozen papers and articles. In the beginning none of it makes sense, but a day/week/month later things just click, and the whole thing makes sense.