Speed reading promises are too good to be true
sciencedaily.com
sciencedaily.com
It's still a fun thing to try if you haven't done it yet. The article points out what is pretty clearly the biggest problem with these speed readers - going back to re-read is a big UX challenge. My little hackathon app used a "back ten words" button (may have been what Spritz recommended?), but I also thought having the actual paragraphs scrolling above the Spritz text as you read would be a more useful way of solving this problem. If you get lost, skim the paragraphs around what you're reading, and press / click a word you want to start speedreading from again. In more dense material where you need to flip back a page or two, you might need some sort of zoom out functionality as well.
If nothing else, these technologies will hopefully find a use in wearables and other tiny-screen devices.
Perhaps a better approach would be to develop deep learning systems that sort of translate cumbersome writing into linguistic structures that our brains can more easily/quickly digest?
"I took a speed-reading course and read War and Peace in twenty minutes. It involves Russia." -- Woody Allen
It's quite old too.
I took on speed reading in college after realizing I needed to study and learn faster or I would fail miserably. The book I picked up seemed like a lot of common sense steps the first one being to inhibiting subvocalization, the internal voice we are used to hear while reading since we where taught to read out loud. The idea here being that our mind can process information much faster than the speed of speech.
The paper cites several studies that suggest inhibiting subvocalization reduces comprehension. However it takes the result of these studies as fact when they might be flawed. An example in their chapter about the role of phonology:
> First, consider a study in which skilled readers were asked to indicate (with a button press) whether briefly presented words were members of a certain category (e.g., foods; Van Orden, 1987). The people in this study incorrectly responded “yes” to a word that was not a member of the category (e.g., meet for the category food) about 19% of the time if that word was a homophone of a true member of the category: In this example, meet is pronounced the same as meat, which is a food. Incorrect “yes” responses to non-homophone words—for example, melt—occurred only 3% of the time.
I looked a the Van Order paper (http://link.springer.com/article/10.3758/BF03197716) since it shows the highest reading comprehension failure rate (19%). They came up with that number from an experiment studing just 5 male and 5 female undergraduate students (see Experiment 1, page 3)
I think citing this study altogether as a basis of one of their main conclusions (eliminating subvocalization lowers comprehension) is flawed: The study didn't compare regular undergraduates with speed readers (people skilled at comprehending without subvocalizing), it just concluded no subvocalization could lower comprehension by 19%.
The book[1] claimed you could reach up to 1200 wpm by doing all the exercises, I never got to finish all of them.
[1] "Superlectura veloz" (fast super-reading) - Carlos Quiroga
Should it refer strictly to (A) the internal monologue, (B) micro-movements of speaking muscles, or (C) both?
Correct me if I'm wrong, but my suspicion is that parsing and comprehension of text would be impossible without A. But also that A is slowed by B; thus reducing or eliminating B is beneficial.
But I still tend to fall back into it when I'm reading something hard to understand, specially when reading english which is not my native language.
That is also why I watch all my movies at half speed!
Only if the text is exceptionally well-written will I slow it down and really dig into what the writer was trying to convey. It's like a conversation. Some are really worth paying attention to and experiencing, most unfortunately are not, just perform your social duty and then go find something more interesting to do.
Saying more with fewer words takes work, and not everyone does it, or can do it.
In the 4th century being able to read silently was notably strange.
We've refactored the process quite a bit since then:
Splitting individual words. Introducing capitals to highlight paragraph boundaries and proper names. Punctuation to speed up parsing. Inter-paragraph spacing and semantic and syntactic conventions to improve concept parsing. Etc., etc.
I see no reason to believe we are done yet. :-)
I've seen it argued that this is actually a myth, popularized by Nietzsche. See e.g. http://hour25.heroesx.chs.harvard.edu/?topic=reading-silentl...
However, if you naturally skim articles really quickly looking for information that will be of interest to you, then in a sense you are already speed reading.
I'm a huge advocate of speed reading, personally[1]. However, I think the points made in the article have a lot of merit.
Speed reading should be seen as a tool for consuming information, but you have to adjust the way you learn accordingly[2]. I can read, cover to cover, a 1000 page book in a few hours. My strategy for learning from a large book is to read it more than once (often three to four times) to treat the first read as a 30,000 ft view. I don't do included exercises, and I also limit my note-taking. My second read is with pen and paper in hand. I break for exercises and notes, but otherwise continue to read just as rapidly. If I struggled with the concepts, I'll repeat the process.
Years ago, that book would have taken several sessions to complete the first read and the second read would only be over particular sections. I'd be done after weeks, rather than a weekend, which also negatively affected my retention. Reading from day 1 is built upon and by day 19, it's foggy. When I can do that in an afternoon, it's fresh on my mind. And at worst, with several reads, I'm still getting through the material in half the time it would have previously taken me and with solid retention[3].
It's also not something you become good at over night. After I took the course, I could read twice as fast, but not twice as effectively. It requires a lot of practice. When I sit down to read something, I have a goal of learning whatever it is I'm trying to learn, but I have a second goal of consuming that material more effectively. The first few years, I focused on reducing my time spent on the first read. As the years have gone by, I've adjusted my note-taking and methods for retention. I read a 500+ page book (often equivalent text in PDF) nearly every weekend. I pick things I am using today and grab books that build on that.
[1] I took a multi-week course that focused on the typical "eye movement" techniques along with skim/scan techniques. That course started, for me personally, a rapid change in my ability to consume information from text and I still believe it was one of the most important classes I've ever taken. I greatly prefer book reading over courses these days because I can completely control my pace from very rapid to word-for-word.
[2] I'm speaking purely of technical reading. I do not speed read fiction and my personal experience is that speed-reading fiction is awful. I stick with audio books for that.
[3] I do not have data to back this up. I've been reading large, technical programming books for 18 years, and some of that added speed is that I have a relative understanding of the work before I begin. I know how to read a programming book to maximize my retention and I read programming text to become better at what I do, which is something I've learned how to do from years of practice.
[Edit: Added a note clarifying fiction vs technical]
He may delude himself into believing he's actually getting anything out of the books he's speed reading but I bet if put to the test, his performance in terms of comprehension would be abysmal.
I used to work in Infrastructure in the early 2000s and a coworker and I had a running contest on how many certifications we could amass (my company had a library of reading materials). It was my coworker who pushed me to take a speed reading course. We'd each read 600+ page books covering the material around 3 times (maybe 4, but later reads are not complete reads) over a week while testing each other at work. Both of our pass rates were perfect[1] (Novell CNE, CNA, MCNE, A+, Security+ -- do those still exist? -- and a myriad of other certifications of dubious worth) -- failing meant we had to pay out of pocket, so we passed.
The "first read" is a skim/scan, very superficial read, and it's easy to get through a large book this way, especially if it's material I'm already comfortable with. The 1000 page books are technical books (points I made elsewhere) that aren't all wall to wall content, lots of diagrams, code, etc. I read a book more than once if I want to learn the content (the first read is partly to weed out a book that has no value to me).
At the end of the day, it's the equivalent time it takes to learn the material that matters. After a 2 hour "read" (skim/scan/speed-read) of a 1,000 page book, I'm nowhere near where I'd be if I had read the entire thing, slowly, word-for-word while taking notes and doing exercises. If I feel the book is worth it, I'll read it again, then read parts of it carefully, to the point where I've likely read the book 3-4 times with notes and exercises included. I'm at "equivalent to reading it slowly" by the end of the second read. But that's a low bar for a book of that size. Considering that same exercise would have taken me a month to complete the first read. How much of Chapter 4-6 will I remember on day 30 of multi-evening broken-up reading sessions reading it "slowly"? How likely am I to be burned out by then? By doing it this way, much more of it is fresh in my mind and (most important to me), I'm more likely to devote the time to learning the material knowing it will take a week, not a month+, to get what I want out of the book. After finishing a C++ book recently, I was writing in the language. I'm nowhere near proficient, having only two books and a 10-year gap in my C++ code experience, but "the old way" I would have been more likely to let it sit on the shelf half-read, never completing it. If I did, I wouldn't have been able to write more than a "Hello, World" program without having it pulled up along-side my IDE, constantly referring to bits I hadn't read in over a week.
[1] Pass rates, not scores. There were some tests I squeaked by one question over passing. I scored perfectly on more than a few, though, too.
To be fair (and to handle "bullshit" being called by another person's comments), the books usually cover a topic I am already at least partly (and sometimes very) familiar with. And many programming books contain pages of "just code" that are sparse in text. Books on topics that I am less familiar with I read more slowly, but rarely take more than a few hours (3-4) to get through the first read. I've explained in more detail in other comments, but the first read is a skim/scan without notes and without doing any exercises. It's to get a 30,000 ft view of the book to help identify areas I need to study more deeply.
Subsequent reads I focus on exercises, but hone in on areas I am less familiar with. The second read is much slower than the first. The most recent example I can think was Bjarn Stroustrup's "The C++ Programming Language". I've done little with C++ in about 10 years so this was almost all "new material". I started on Saturday morning and completed the first read in about 3 hours. I spent the day Sunday doing the second read, taking notes, but still without exercises. I spent the remainder of the week on subsequent reading doing exercises in 2-3 hour sessions every night.
This book would have taken me a month to read, the first time, before (I recall a Novell Netware book of 800 pages taking me almost a month to complete). And I would have gotten much less out of it as a result of the gaps of time between topics. I am not presently a "C++ genius" having read the book, but I've written a few programs and am comfortable with the material. And I know I'll read it again and refer to it.
You can't become a proficient speed reader over night or with just "some training" or "a course" without a lot of practice. I love reading programming books of any kind and approach it partly as an exercise to become better at consuming information, so every book I hope to get a little more efficient (though the law of diminishing returns applies).
I've way more than doubled my reading speed from when I took a course 18 years ago. My retention is dramatically higher, as well. The main reason for the increased retention, however, had less to do with the course than a change in the way I consume information from books. If the technical book is worth reading, I read it at least twice. "Read" includes skimming/scanning techniques that I've gotten very good at over 18 years. Thinking "well, I can read 90% faster with these techniques" will result in retention that is very low if you change nothing else. The first read is rapid and haphazard (sometimes just to figure out if it's worth using the book as a source of information). The second read will be honed for specific information (and third and fourth).
That serves two points in favor of retention. By being able to get through a massive text in a weekend with multiple reads (versus a single read over a few weeks), that information hasn't become foggy from multiple long breaks/distractions--it's fresh. My second read is cover-to-cover, but focused with note-taking and doing exercises. Subsequent reads are surgically focused on areas where I am weak and are done more slowly.
Before, my learning was single-pass (with some section/chapter review). Because of the reduced time/effort it takes me read, I can do so in a multi-pass manner with careful focus on learning the material rather than on the act of reading.
"Our testing shows that the retention levels when spritzing are at least as good as with traditional reading and that, with just a little bit of experience, you will retain even more than you did before."
So at least some companies are making claims of this nature.