The speed reading fallacy: the case for slow reading
nesslabs.com
nesslabs.com
I’m a very fast reader of non-fiction (I have several thousand technical books). However, I don’t really read every book word for word. After reading a couple of books on algorithms for example, it’s possible to skim through the next one only looking for new subjects, better pictures, or new approaches. I also jump around and read the parts I’m interested in. The result is I have some recollection of what’s in most of my more recent books, and if I need to look something up I can often remember where I saw it in the book—like roughly how far into the text it is and where on the page it was.
For technical books, I agree it is possible to skim rather quickly through stuff that I already know with speed reading approach.
This is a question I always ask all speed readers and the answers are always so muddled. I can never understand what they are getting at.
Perhaps I relish a good novel more than most, but to me the idea of speed reading a good novel sounds about as appealing as speed drinking a fine wine.
You may be underestimating how many people do this as well :P
Personally, I would love anything that got me thru books faster. I've been reading A Scanner Darkly for months. Admittedly it's more of a problem of concentration than reading speed, but I do understand the appeal of reading quickly. You'll simply get thru more books.
That being said - I don't understand people who listen to sped up podcasts, as I hate burning thru my favourites before a new episode is out. So I think it's just down to personal preference.
If I am enjoying a book, I can’t imagine having the urge to finish it faster. Or slower. What could possibly be gained?
Edit: also, just some unsolicited thoughts on your struggle to complete a Scanner Darkly. IIRC, A Scanner Darkly is a fairly average length novel, meaning you might expect to finish it in ~10 hours of reading. My personal experience is that it's best to enjoy a novel within a relatively limited time frame (say a couple of weeks) so that the various characters and plot lines stay fresh in your head.
If I were in your situation, I would probably question whether:
A) If I actually like the book. If you are even a quarter of the way through a book and don't like it, nothing wrong with picking up something else. As I mention above, books are for enjoyment. If you don't enjoy it, what's the point?
B) If I perhaps don't have time to read a novel now. Sure, nothing wrong with starting a movie, pausing for dinner and then finishing it afterwards. But there are limits. You wouldn't want to split a 2 hour film into a dozen ten-minute segments. And I would argue, wouldn't you want to split a 10 hour novel into a 60 ten-minute segments either. Just not a good reading experience. Nothing wrong with putting the book on hold and coming back in a few weeks once your schedule has calmed a little.
My business sense tells me there's a potential market for speed hearing here.
Also, I can see how if you have multiple hobbies/kids/etc you'd want to speed up your reading because you have limited time.
My own assumptions are quite different, namely I expect that like most other pleasurable activities, you get the pleasure from the time spent doing it, but that doing more of it.
For example, if two people both spend 5 hours hiking, and one person hikes 10 miles and the other 15 miles, I wouldn't then assume the latter had gotten more enjoyment from his or her hiking. Nor would I assume someone that had eaten a meal with 1000 calories got more enjoyment than someone who had merely eaten 500 calories.
In fiction, the conclusion is always at the end of the book. With non-fiction you sometimes know the conclusion by chapter 4. How would you keep going if you knew who the murderer was in chapter 4 of a mystery?
For some non-fiction, authors rely on a lot of repetition, but they don't always make it clear where the next new topic will begin. In a lecture once you get it you can check out. In a video you can skip to the next one. In a book? I don't possess the right coping mechanisms for books, and I wonder if they even exist.
I'm still learning to give myself permission to stop reading a non-fiction book halfway through when I 'get it'. I have a couple shelf-feet of books in my queue because of it. I think I expect there to be some profound observations they've saved until the end, like the little scenes after the credits in movies, and like movies, the random reinforcement keeps me coming back for more.
I've tried a bit of speed reading, skimming, and only reading the first sentence of every paragraph, but it all feels like I'm short-changing myself (and then truly short-changing myself in the process).
Maybe I should just read summaries and skip these books entirely.
If a book is skimmable, why bother? Presuming the author has something genuinely interesting and substantive to say, there are often better alternatives. For example, books written by scholars are often preceded by long-form journal articles which more concisely present the material. And if you were mistaken in your belief they had something interesting to say, then you've only wasted 20 pages of your time, not 400; yet you can still say you gave them the full measure of your attention and a fair chance to persuade you.
The courage to stop at a couple hundred pages when you've said everything that needs to be said is not something everyone has.
I think this may be one aspect of why I love 'Refactoring' and hate 'Design Patterns'. It has so much more information in a book half the size.
Also, the authors had to show how these patterns can be implemented in different programming languages (Smalltalk and C++). Fowler could get away with one simple example of a refactoring in Java each time.
Having said that, Fowler's prose is super clear, something that cannot be necessarily said about GoF, (IMO of course)...
The act of reading a book permanently alters how you think a little bit, so the most important factor in deciding whether to abandon a book is whether or not you want to think more like the author does.
Like any other skill, the sort of thinking that any given book is training you to do benefits from repetition. The variation in the different scenarios also helps make this skill more robust so that you can apply it correctly to more diverse situations.
Fortunately, all this means you don’t need to make much of an effort to actually remember anything the book says— the book’s real value comes from subtlety altering the way you think which will happen so long as you’re attending to what you read. The later chapters should be easier and more relaxing to read because you’re familiar with the pattern of the author’s argument and they act more like a mental exercise than anything else.
The really good fiction doesn't always have the conclusion at the end -- you just think that that's the conclusion. Of course this presupposes that there's a sequel that the reader is not aware of just quite yet. (The sequel may not even exist yet, but if the author's narrative extends far enough into the future then you can get that feel.)
I think a decent example of this is the Three-Body Problem, although the first one does have a sort of abrupt ending. A weaker example would be Banks' Culture novels, but that was more about there being a whole world out there -- the actual narrative did usually just end at the end of the book.
I was going to say that for the opposite reason -- I don't think that something being a series or part of a larger whole makes it inherently better (sometimes it hurts; many a great book has been ruined by a sequel).
Some of the best books, however, have the conclusion right at the beginning and all the rest is "how we got there". C.F. much Vonnegut.
However, I do think the "unfinished series" idea has some merit to it.
The quote was:
> We are what we pretend to be, so we must be careful about what we pretend to be.
- Arthur Schopenhauer (1851)
But I wanted to make an additional point myself, which is: the key part of reading for information is not consumption, but rather memory. It doesn't matter how many books you read if you only remember a small percentage of each one. Like any educated person, I've read thousands of books in my lifetime. And yet I could probably only name the details of a few dozen. This strikes me as horribly inefficient.
As such, I've been trying to incorporate Spaced Repetition (with Anki) into the reading process itself. My initial idea is to do two things:
- add specific lines and information into Anki. E.g. quotes, statistics, and so on.
- write a brief one-paragraph summary of each chapter or, in shorter books, each page. Then add this summary to Anki.
So far, I've been able to recall far more information about the books I've read, even if it takes 5-10x times longer to read them. Overall, I'd consider that an effective sacrifice.
It was cool. It advocated multiple forms of meditation for concentration training so you could speed read a book. Using your increased concentration, which was mentioned as super important, you'd do pass 1 of 3 speed reading through using the Visual Method mentioned. Meditation and clearing your head were mentioned specifically to reduce regressing, but it was okay to regress. Pass 2 would involve creating an outline of all the topics and information. Pass 3 would be creating a visual mind map of the information.
Point for me was that speed reading helps only to filter or get a first pass on something, but I doubt anything really complicated can be understood by simply speed reading it. Or even simply reading it once.
I can read just under a couple of pages of a standard paperback in about a minute. It's a pleasant way to spend a few hours.
I don't read non-fiction like that, though. It just doesn't happen, and I don't know how to make it happen.
I think this means you're doing it right! In Polanyian terms, your subsidiary awareness is on the particulars of the typography and the words allowing your focal awareness to be on the fun bit, i.e. on the meaning.
This ability, I think, depends just as much on how interesting and enjoyable the content is as it depends on your reading skill.
For more Michael Polanyi I recommend this superb (audio, non-fictional) lecture:
https://www.youtube.com/watch?v=rVx8KhsZYPw
>I don't read non-fiction like that, though. It just doesn't happen, and I don't know how to make it happen.
This may be because the books are boring. For instance, textbooks. Here the reading is mainly about searching for the relevant material, so the focal awareness is on the text itself. However, Feynman's Hairy Green Ball Method seems applicable:
https://www.e-reading.club/chapter.php/71262/21/Feynman_-_Su...
I had a scheme, which I still use today when somebody is explaining something that I’m trying to understand: I keep making up examples. For instance, the mathematicians would come in with a terrific theorem, and they’re all excited. As they’re telling me the conditions of the theorem, I construct something which fits all the conditions. You know, you have a set (one ball)—disjoint (two halls). Then the balls turn colors, grow hairs, or whatever, in my head as they put more conditions on. Finally they state the theorem, which is some dumb thing about the ball which isn’t true for my hairy green ball thing, so I say, “False!”
Not to mention the numerous pop-sci books which are a detailed history of science in disguise. I'm actually abandoning those. I'm tired of reading about the personality quirks or life stories of Bohr, Einstein, Heisenberg, Schrödinger whenever I pick up a book about quantum mechanics.
Life's too short to invest your time (slow read) in filler material.
For reading, generally, I strongly recommend Mortimer Adler's classic How to Read a Book, itself an HN perennial, see https://news.ycombinator.com/item?id=12209446 (or all: https://hn.algolia.com/?q="how%20to%20read%20a%20book").
If you're trying to determine what to read, or want to rapidly get an overview of a book, or are "visually grepping" for specific content, then the tricks of speed-reading, especially of chunking, skimming paragraph ledes, and not vocalising or subvocalising, help.
If you're reading for deep understanding, you'll want to fully absorb the material (and work through the exercises!), though this can also involve several passes -- for textbooks or nonfiction, I'll scan the ToC, and then the chapter (especially for section/subsection titles), then read through in full.
Treating such reading as an inquiry rather than simply as data transfer helps tremendously.
And of course, if you're reading quality literature or poetry, you'll want to savour the language, words, rhythm, meter, and rhyme. Go slow. You'll appreciate it all the more.
Reading plays or dialogue-rich content is a particular skill -- try reading with different intonation or to sort out what possible alternate meanings might exist. Shakespeare is of course excellent for this, though many other plays (or film screenplays) are as well. Realise just how much latitude the actor or director has in translating from words on the page to those uttered on the stage or screen.
Which is true but so what? They are elements of life; sometimes you grab a quick bite on the go and sometimes you sit down to a multi-hour meal, yet both are "eating". Likewise with reading.
See Mortimer Adler's How to Read a Book for an excellent guide. Most of the techniques it describes are ones I've long used -- it's been an excercise largely in validation, though with a few additional tricks and bits.
The issue seems to be that while this dialog is crystal clear when you read it slowly and understand all the words, most users appear to read what they think it says, and select the dialog option according to that.
It's not a wall of text or anything, it's a medium length sentence on the form "Select Yes to [...], press No if you want [...]". I ran it past the support guys and they thought it was clear as well. But apparently it's very easy to get wrong.
In all cases the users go "oooohhh, that's what it's saying" when they get told to read it carefully, and after that they select the correct option and make the program do what they want.
I spent a fair amount of time figuring out the most clear and correct way to present the choices, but apparently I failed hard... oh well, that dialog is getting rewritten tomorrow!
> Give alert buttons succinct, logical titles. The best button titles consist of one or two words that describe the result of clicking the button. [...] To the extent possible, use verbs and verb phrases that relate directly to the alert title and message—for example, View All, Reply, or Ignore. Use OK for simple acceptance. Avoid using Yes and No.
Suppose opening a document gives me an alert that says, "This document contains macros. Macros may contain viruses that could be harmful to your computer. Do you want to disable macros before opening the file?"
With simple Yes/No buttons, I might hastily jump ahead after reading the first sentence and think that affirming the dialog means, "Yes, I know it contains macros."
Or I might hastily jump ahead after reading the second sentence and click, "No, I'm scared, don't let this malware infect me."
Or I might mistake it for another pop up that happens all the time and also has Yes/No buttons, and not realize until it's too late that I've accidentally enabled macros.
Yes, users should read dialogs, but good UI design can help.
https://developer.apple.com/design/human-interface-guideline...
In fact, many commenters take this speed-reading model of hacker news articles to its logical conclusion.
I then later transfer those index cards into notes on my computer into equivalent "cards", which let me edit or rework the information on the fly.
Sometime, I transfer some of the information into Anki SRS, requiring yet another rework into flash cards, and permanently anchoring the information into my brain.
All told I’d say it’s neither positive or negative. I’ve had to teach myself techniques to slow down especially on technical material but I can get through business style reading extremely fast. Different texts and contexts require different techniques.
I often noticed how they could read MUCH faster than i could, and appeared to be actually retaining the information.
Me personally, if i read too fast im literally just "seeing" the words on the paper pass my eyes. I could read a sentence very fast, out loud, and not understand it.
I've always been a slow reader. My sister took speed reading "lessons"(?) when we were younger, apparently she had a problem with reading so my parents felt the need to enroll her in some type of weekend (speed?) reading program that was held at a hotel conference center in a trendy area of our city... it was probably snake oil because i only recall her going once or twice then never returning (perhaps the program was only a day or two long? even worse? how do you really learn anything?)
And I had her share the speed reading techniques with me in an attempt to learn to read faster myself but never could do so.
I dont think that speed reading is all that its cracked up to be.
Speed reading.
Oh, it works with single words too!
Generally speaking, there's a continuum of activities from glancing to skimming to reading to 'close reading' (a term of art in education). It's not surprising that for most people, there is a tradeoff between speed and comprehension/recall.
EDIT: are the downvotes for the first paragraph or the second? Comments would be more helpful!
1: https://hbr.org/2012/03/hard-to-read-fonts-promote-better-re...
Understanding > Knowledge
Effective =/= Efficient
Etc.
As for the other statements; of course you can't read complex or deep text by glancing at it and taking it in in one line or paragraph. Imagine trying to learn topology or something that way; you will fail. You have to slow down to the point you're working problems on the side in a notebook. Same thing for learning a new language; looking at the grammar table won't help: I have to internalize it by repetition. Deeply aesthetic poetry; you have to vocalize it, or it doesn't even begin to make sense. Similarly, complex fiction you have to slow down to figure out what is going on, sometimes rereading passages or whole chapters. Sometimes this also needs to be vocalized (Modernist types like Joyce & Pound).
But for stuffing clearly written briefing type texts, speed reading is the absolute shizznizzlenuts. That's why Kennedy found it useful; that's pretty much what the man read all day at work. Low end college coursework also fit into this category of reading material. Lots of business stuff also does. Her neuroscience balognia about saccades may be true to some extent, but you can skip over or infer most simple sentences in the exact same way that Gmail and Outlook completes your sentences via semi-markov models.
Again: it works for simple texts. It also probably works better if you're smarter and a better reader in general, and so taking a bunch of average people and drawing conclusions from their results is pretty useless.
I have a relative who is a natural speed reader and who has perfect recall. I realize this is a controversial claim as perfect recall (sometimes called photographic memory) is a myth. It's not a myth more likely extremely rare. I could give her a book around 300 pages she had never seen before. She would read it in about 15 minutes. I then could ask questions about the book and she would answer by exactly quoting sections and citing the page number. So... was this helpful? Well she never went to college or did anything with her life. It was just a weird thing she could do naturally which she didn't exploit except to read and remember precisely absolutely vast numbers of books.