Explore a better way to work – one that promises more calm, clarity, and creativity.

Experiments with the Textbook Method

Tailoring the Textbook Method

I spent this past week experimenting with the textbook method. I began by creating a template — a blank LaTex document — for collecting research notes:

When compiled into a pdf, it looks like this:

My plan was to minimize friction when starting work on a new idea or project — all I have to do now is copy the blank document to a new directory, change the title, and start capturing notes.

With my system in place I could take it out for a spin.

I decided to apply the method to a big hairy graph theory problem that my collaborators and I have been battling for months. This big problem keeps branching off into many promising smaller problems, one of which I have been pursuing recently with my grad student. This sub-problem provided a perfect case for applying the method.

I started a new write-up to capture, in my own words, what we thought we knew so far:

This well-constructed plan worked well…for about twenty minutes.

As I was writing, this process of formalization led me to a new approach to the idea. I quickly generated a new blank document (easy to do now that I have a template) and spent the rest of the afternoon, textbook open in front of me, office whiteboard filling with diagrams, trying to work though the details:

At first, I felt somewhat uneasy about leaving that first document half-written. My task-oriented instinct is to finish each write-up, once started. Instead I had abandoned the document as soon as something more relevant popped onto my radar.

But on reflection, I think what is happening — rapid idea abandonment and spawning —  is exactly what I want from the method. The write-ups, I must remind myself, are not a goal in themselves (most likely, no one will ever see them). They’re instead a tool to induce fast learning, and this fast learning, in turn, increases the rate at which I can explore a problem space — exactly what I need in my research.

To summarize: I’m testing this method in my applied mathematics research, but it’s becoming clear that it should work equally as effectively in most scenarios where you need to master complicated things fast — be it a new programming language or marketing strategy. We’ll see how it holds up as I apply it to multiple concurrent projects and more complicated topics.


This post is part of my Craftsman in the Cubicle series which explores strategies for building a remarkable working life by mastering a small number of rare and valuable skills. Previous posts include:


21 thoughts on “Experiments with the Textbook Method”

  1. Keep this series up please. Note-taking is a hard problem and your practical experiments are very helpful.

    FWIW, some things I’ve tried:

    1. Same as you describe in this point, but using a blank markdown document. (Advantage: even less friction than LaTeX).
    2. Make a note document for each paper read as opposed to each topic or problem; file the notes with the pdf of the paper using the same filing scheme. (Advantage: mirrors an existing system of concept organization; easier to file).
    3. Write down any thoughts that come to mind in Evernote, one thought per Note. (Advantage: No need to organize at all, similar to physical index cards.)

    But the problem I have not cracked is how to conveniently trace similar thoughts as they pop up in various places. I suspect the answer is that there is no way to do so systematically, but I still hope for a system that makes similar thoughts easier to spot.

    Looking forward to more of your experiments.

  2. Thx for the follow up on your textbook method. did I understand it correctly, that you don’t try to capture the information in the pdf but rather collect ideas which come from working through the information? And what happened to your Research-Bible you mentioned a couple of months/years ago? Or the little bets projects?

    I myself have my own Research-Bible (simple word doc) where I enter all the info I get from papers, books, interviews about my topic. Only sorted in broad topics but I like it because while entering the info I a) have to tell it in my works and figure out exactly what the argument/ evidence/ concl is and b) I add any thoughts I have while reading – this often ends with little treaties on various topics in the Research-bible. One of my seminar papers I created just by copy and pasting information and thoughts from the research bible! Added an Intro and Concl and voila.

  3. When I’ve done this kind of thing, I really like MS Word’s Outline view. It is the one thing in Word that I’ve never found a good competitor for. The ability to build arbitrary outline structures and freely re-arrange them. I used it when I wrote my book, and I almost use it when working on a project of any size.

    Believe me, I’ve looked for alternatives (especially with the recent awful WORD updates), but the freeform nature of Outline view is just prefect for this type of thing.

  4. I’ve used a similar technique in my studying, rapidly writing thoughts while reading/learning, and often returning to concepts and writing more to see what was retained/forgotten. I found early on that I had to do the process written though, that there was something about the computer that made everything feel less concrete or that I retained less. I think I’ve gradually started coming around to the electronic form, Evernote certainly keeps everything more organized.

    This was always followed by verbally teaching the concept to a captive audience of my two dogs. It was great for prepping a lecture while under the gun.

  5. I’m a huge fan of *Tex, but I think this is the wrong tool for this job. I hate to suggest a distraction of learning a new tool, but if you use emacs you should look into org-mode for note-taking and document writing.

    You’ll be making tables as part of your textbook method, and equations, and re-arranging and spawning etc. org-mode supports LaTeX equations and text (in fact, you can use it as a simple markup-language to put out a full TeX document).

  6. But the problem I have not cracked is how to conveniently trace similar thoughts as they pop up in various places.

    I know people put a lot of effort into organizing diverse collections of thoughts into something more connected (Steven Johnson has written about this on his blog). Honestly, I don’t typically have that problem. I’m not usually coming up with lots of random ideas from scratch. Typically, I have a collection of well-defined problems I’m working on. For each, I might explore a bunch of techniques, but not so many that I lose track. Sometimes techniques from one help another, but I usually make those connections without the need of outside help.

    My standard strategy in technology-assisted work is to use the minimum amount of technology that still makes things much better than without.

    ’m a huge fan of *Tex, but I think this is the wrong tool for this job.

    This is probably true. But I don’t want to waste cycles learning a new system when what I have now works just fine (my textbook method notes are written in the form of an academic paper, so it makes sense for me to use what I typically use for academic papers).

    Though I have heard cool things about org-mode.

    This was always followed by verbally teaching the concept to a captive audience of my two dogs.

    My dog has heard more distributed algorithm theory than any non-human should ever be exposed to!

  7. Steven Davis:

    “MS Word’s Outline view. It is the one thing in Word that I’ve never found a good competitor for. The ability to build arbitrary outline structures and freely re-arrange them. I used it when I wrote my book, and I almost use it when working on a project of any size.

    Believe me, I’ve looked for alternatives”

    Have you tried Scrivener? I gave up on word’s outline mode years ago as much too fragile.

  8. I really like the detailed nuts and bolts posts on exactly how to work. Evernote is the tool to use. You can arrange categories by separate notebooks or just tag them. If you want to find old notes you can also keyword search. Brilliant for assembling clips of sections of papers to easily find later by searching author names or keywords or tags or separate notebooks. When you are done export and printout the files for study (only XML-HTML though).

  9. As someone who does come up with lots of random ideas from scratch, I’ve got a trick that uses one of Cal’s tools as an idea holding pin before they’re officially filed and organized.

    In a previous post, Cal talks about the Plan.txt file he keeps on his desktop. I started doing this, too, and it’s where I keep my to-do list for the given day.

    Because new ideas hit at odd times (usually when you’re in the middle of an important task, trying to focus), I dump everything into my Plan.txt file. It’s easy to get to and I don’t have to search through my catalog of idea-collecting Google Docs to find the right place.

    Then, at the end of each day as I’m reviewing my task lists, I empty out the ideas in Plan.txt to the appropriate places.

  10. But the problem I have not cracked is how to conveniently trace similar thoughts as they pop up in various places.

    If I understand correctly, I use Tomboy for this on my laptop, and it works amazingly well.


  11. I am a little confused as to the effectiveness of this technique. Writing down and summarizing things in your own words might help in learning new concepts, but how can be sure that you are actually understanding these new concepts.

  12. Would i be overdoing it if i use the tb method on college readings? I’m in arts and social science faculty and readings are truckloads. Or would QEC suffice? Help pls. -Freshie

  13. I’ve been following your recent “textbook method” posts and I agree. I’m a resident physician and am trying to find a way to maximize not just my retention for what I read, but also my motivation/enjoyment.

    I’m now working on developing a system for reading that meshes with a heavy day-job work schedule yet still is effective continually advancing one’s learning curve in a field. I think a daily habit approach is important (ref: But, as we say in surgical fields, “practice makes permanent; PERFECT practice makes perfect”. So I’m increasingly devoting attention to maximizing understanding, enjoyment, and retention for a field in which studying tends to resemble brute-force memorization.

    I’ve recently been writing summaries in margins, making notes on 3×5 cards, etc., but have strongly considered doing something akin to the textbook method. I’ve been more mentally engaged and it’s more enjoyable, but I’d like a more solidified system.

    Cal’s posts re: the Textbook Method have been compelling. As a credit to this approach, I just finished working a call shift alongside an attending who has a USB drive that is essentially a textbook of topics that, over the years, he’s found relevant enough to include and update. And he seems to aid retention by returning to his “textbook” when he wants to look something up. Especially considering his age and phase of his career (he’s 70 yo and a former dept. chair), his thinking is exceptionally sharp and his retention is quite impressive.

    If any of you are in medicine (or a different field w/ similar learning/work demands), what do you do for pushing your learning curve? Do you think the textbook method is applicable? Is there a better system?

  14. I’m not a regular reader and happened to stop by–what serendipity.

    I’m also newly arrived to the TT, and I’m in science. I’ve spent this past month experimenting with a better workflow for my research and regular literature review (i.e., my learning!).

    Although I admire LaTeX’s purity, I agree with others that Evernote might be better optimized for this task. Evernote interfaces beautifully with Papers and Chrome. When I search for certain topics in Google, Evernote will remind me that I have related notes in my “research” stack and will bring them up for me.

    Currently, I have notes for different research topics. Every abstract or paper that I read, and each decent idea or question that I have, gets added to at least one topic. This changes how I approach my reading; previously, unattached “cool ideas” would be lost to my mental ether within days or weeks. My interests are broader than yours (in that I’m brainstorming for very broad grants right now and trying to strategize research directions), and I think I have more new literature to stay on top of than you do, but I’m also computational/theoretical and appreciate having reminders of things that might later be relevant.

    I’m really looking forward to future posts on this topic.

  15. Small aside: I realized that “the textbook method” as it relates to one’s own research really means just starting the freaking manuscript now. I’ve found that I have a very different perspective on the experiments that I need to run when “forced” to explain steps for the supplementary material v. when I think I’m just experimenting for myself. The former requires much savvier focus.

  16. Great series of blog posts! I’ve been following more-or-less the Textbook method for a few years (I didn’t know this had a name!) — I found it particularly effective for more complex topics that really demand deep reflexion. BUT I really do find LaTeX to be counterproductive in this context. What works best for me is to use Mathematica: its notebook user-interface allows word-processor-like writing (with a math syntax close enough to LaTeX that it’s fast to write and easy to remember), but you also get the power of an extremely concise functional programming language to actually quickly test things out in any mildly computationally-related field. So this approach really gets ideas to stick in the head: not only you write the result, but you also try it out for real. This really makes them stick.


Leave a Comment