Beeminder Forum

Speed reading with Peter Kump


Excellent idea! I have a pair of gloves that I know doesn’t work with touchscreens.

Use their weaknesses against them!


Since your citing Scott Young, it’s probably also worth sharing his criticism of the overhyped promises of speed reading:


So I just did Exercise 3 a couple of times, (first reading a book using finger.)

I’m getting about the same speed (tiny bit faster, but this is an easier book,) but now have the added distraction of a finger zooming around getting in the way, distracting me, not keeping up with my eyes and making my shoulder hurt.
I’ve not noticed any eye regressions, which is what the finger seems to claim to solve.

My initial rate was 421 WPM with good comprehension, and I’m doubtful that I’m going to be able to significantly increase my reading speed. The Scott young article suggested I’m not going to be able to get much above 500 WPM anyway. (Maybe it’s worth it since 420 -> 500 is still an extra 20%)

Anyone further along got any advice, anecdotes or speed improvement numbers?


I read a lot of academic articles. When I’m done with an article (if I think it’s worth remembering) I open Anki and write the author and date on one side (e.g. Yellen 1984) and three sentences on the back - Why the paper is cool. How they pulled it off. And a little mnemonic device connecting the authors and the year to the idea.

I think that this is a pretty important step if you are reading above 800 wpm. You end up not spending a ton of time with the paper so it’s easy to lose track of who said what. The Anki step makes writing much quicker.


I also have serious doubts that it works as often advertised. From what I have read, there seems to be an unavoidable tradeoff between comprehension and speed.

I found this wired article quite informative as it cites some published research: Is speed reading actually possible?


The wired article talks about a “forthcoming research paper” titled “So Much to Read, So Little Time: How Do We Read, and Can Speed Reading Help?”. The paper has been published by now. You can find it here.


The book, and the method, is predicated on the idea that we don’t have to read everything at the same level of depth. Thus, some things we only need to breeze through, other things we need to be able to get more focused with etc etc. It’s not all about speed, though getting comfortable with that broad concept does allow for improvements in speed across the board. I credit my 150 books last year to Kump’s method, no question.


I’d say the benefit I’m getting from Kump is entirely in honing my ability to mostly-read-most-of-it when I’ve decided that all I want from something is the gist of it. As a graduate student in English (and, indeed, as a scholar in English), there are a lot of books on my plate where “a general sense of the whole thing” is much more useful to me than “a careful understanding of the first half” (not to mention “a desire to get around to it eventually”!). However, for the same reason I also have a lot of training in reading very slowly and thoughtfully, cross-referencing texts with each other, taking extensive notes, etc; even when I resolve to skim a book, I find I don’t actually know how to do so.

I’m still iffy on the finger, which I suspect is mainly helpful for folks trying to break the 200wpm barrier, but it’s definitely been helpful to physically practice moving my eye away from information when I’ve gotten ‘enough’ from it.


Progress update:

I’ve now read a bit more of the book and done Exercise 4, and am feeling more positive about the process. I was able to measure a 100WPM increase during that exercise which I found impressive. I’m not sure of the comprehension loss, it doesn’t feel like it increased much, I guess I’d have to re-read to find out.

I have some advice for my past self:

You get used to the finger.
Rest your elbow on something to stop your shoulder having to support your whole arm the whole time.
It’s OK to use more than 1 finger, so you don’t have your other fingers curled under and hitting the book with your knuckles the whole time.
The later exercises provide methods for deliberate practice at increasing your reading speed.

I’ve found the benefit of the finger is that it provides an independent pacing mechanism, it’s hard to make your eyes go at their normal reading speed and make them move faster at the same time, but you can independently move your hand and your eyes, so you can provide useful pacing with your finger, and that by building up a faster hand rhythm you can encourage faster eye movement.

It’s going to take a lot more practice to lock in increases to my base reading rate, but I feel that being able to do that will be beneficial, and it doesn’t stop me being able to read slower if 100% comprehension is important.


I’ve followed the book’s own advice and have started being a bit ruthless about which parts I’m going to carefully engage with and which ones I’m going to get the gist of and blow past, and I have to say that I found the late-teen chapters about the structures of writing to be really helpful; I have formal training in rhetoric and composition, and Kump, I think, does a good job of distilling the functions of sentences and paragraphs into tactical “this is where to look for the important stuff” information. I found it extra helpful for me to apply my formal training not as a way to pick apart the things I’m reading in intense detail, but as a way to confidently direct my attention away from the paragraph much sooner. I’m churning through a lot of academic articles right now for some research papers, and Kump has definitely been helpful.

I’m still skeptical about the finger pacing, though, at least in my case; I feel like my eyes move much faster without the finger.


What was your initial reading rate (WPM)?
I’m also skeptical about the finger, I suspect it might be most useful sub 200(?)WPM if people are lingering on words. I think I will try some of the speed increase drills without it and compare to times I have used it.

But I might just not have used it enough, on page 23 in: “Always use your Hand” Kump cautions against being “that person” who decides not to use their finger and suggests you should have done a good hour of just reading with the finger. I’ve probably not done that much (outside of the exercises) yet. I think it’s because most of what I read is on a computer screen and that is really not well positions for finger reading.

Does anyone want to speak up in favour of the finger?


I will defend the finger!

I’ll readily admit I don’t always use it, but whenever I do I notice the difference. Maybe it has something to do with how well you are able to focus. A combination of using the finger and having a metronome in the beginning was really crucial in keeping me moving at a certain speed.

Also, I think it’s not something where the difference is necessarily apparent after doing it one or two times. You have to give it a solid (unambiguous) shot.


Looks like I started at 287 WPM for a scholarly article as my very first rate, 349 WPM for an 18thC novel with the finger in the first exercise. 865 WPM as my initial rate the first time I used an actually easy text for an exercise.

I do find the pacing finger to be more compelling when I’m reading paper books. Unfortunately, my field is computational analysis of literature, so I’m not going to be reintroducing paper books to my life too much.

I’ve found that I was pretty much capable of high speeds from the beginning – I started hitting 1000-2000 WPM in the exercises as soon as I switched to ‘easy’ texts that were actually easy, instead of things I needed to read carefully – but my problem is in remembering that I’ve decided to go fast. So I do also feel like I might be “that person” who is giving up too soon on the finger… if my goal is to practice maintaining speed, rather than reaching it, probably I should practice the tools he suggests for that.

@strickvl, have you been using it on a computer screen, or with paper books? Any tips for getting in a good practice session with it?


I got the Kump book on Kindle, but haven’t actually started to read it yet. When that is said, I see a lot of discussion on the topic of paper books vs computer screens. Personally I avoid reading anything of length on the computer screen due to eye strain issues, although I of course do read a whole lot of shorter stuff.

The longest I read would be feature size articles, but when reading them I tend to feel that I stretch it. For longer reads I prefer either paper books or a Kindle. Luckily my gen 3 Kindle (aka 2010) is still alive and kicking, but I see that the newest (and very expensive version) also has physical buttons. I’m uncertain whether that means you can turn off the touch though.

When I get longer papers or documents that I should read, I typically send them to my Kindle, although that is hard to do with work documents. Doesn’t happen all that often though, and it is possible to print them of course.

Also, I’ve heard of computer assisted alternatives to the finger, typically showing 1 word at the time - see for example

I haven’t tried those apps, but I have tried turning the Kindle font up large. I actually like that, but find that I click a lot to change pages.


I just wanted to bump this conversation to see how you all have been getting on with implementing (or not) some form of speed reading in your daily habits…


I finished Kump’s book, though not all the exercises, and find myself satisfied with the improvement in my reading speed.

The only technique I ever use from the book is that of ‘setting intentions’ before I start to read, which often helps me feel good about skimming something very fast (as opposed to feeling vaguely guilty about how fast I’m going, or, worse, slowing down to take too many notes) when I’m trying to get the gist of something. I also think my fast skimming might be slightly faster from the practice, though I haven’t timed myself.

I think I would have found the techniques more useful if I read books on paper – some of the full-page gestures were indeed very helpful when I tried them on a paper book acquired for the purpose. But, I hate reading on paper, so that wasn’t as useful as it could have been.

I’ve been getting slightly more use out using Spritz in the iOS app “ReadMe!”, though I don’t know yet if I will have patience with its poor UI (and attempts to make me spend money) in the long term. I also doubt it would be useful for anyone who wants to read contemporary fiction, since it doesn’t look like it plays nicely with purchased eBooks. All I want to do is read 18thC novels in astonishing bulk for my dissertation, though, so Spritz/ReadMe! works great to plow through the ones in Project Gutenberg.

Overall, I’m glad I read Kump’s book, and glad that halfway through I gave myself permission not to stick too closely to the exercises and just read it lightly for general information. I’d recommend giving it a try to anyone who wants to think through their reading habits and how to make them faster, and particularly to people who need to get through a lot of physical paper pagers.


I just realised you asked about “daily habits” rather than a general assessment of the book – on that front, I have a goal to read one paper in my field per day (more inspired by Cal Newport’s Deep Work than anything else), and I try to skim these articles with brutal Kump-inspired efficiency. So I do that bit of speed-reading on a daily-inspired moderately frequent basis, but no other “habits” of speed reading.


Cool thanks for that. The two main things, at two or so years out from reading the book and doing all the exercises etc, that I seem to have taken away are:

  1. use my finger when reading. this was the real revelation of the book for me
  2. remember to be conscious about what level I need to read the book / paper – i.e. skimming vs detailed word-by-word


Phillip how do you track this in Beeminder? Do you just set up a manual goal?


Yes, I’d use a manual goal, e.g. to track the number of days or number of practice sessions.

Other people find it useful to use the number of minutes, or number of pages. I increasingly like binary measures; did I do something toward this today, or not?

My actual goal for this was a custom job, where the data value was my words-per-minute score in that session, but that it appears on the graph as a +1, using the ‘nonzero’ aggregation method.