I think in most of the mentioned note taking methods (zettelkasten, digital gardening) spaced repetition is kind of frowned upon in a way. It is seen as a necessity because of the way the educational system requires you to conjure up a certain set of knowledge at a particular time (a test). With zettelkasten and digital gardening (or any other method with similar goals) the process and habits that you acquire over time become more important. The goal is not to optimize for the (short-term) memorization of some knowledge only to shove it aside and move on to the next set of knowledge to memorize for the next test. Rather these methods are trying to design a process by which you can build on ideas iteratively. In order to really embrace these methods we'll have to see changes in the educational system as well or we'll end up with methodologies optimized for the educational system that is currently in place (which is what spaced repetition is in my view).
There's a word for this in German: Bulimielernen (bulimia learning). I.e. learning a large number of facts in a manner that you can barf them up exactly once (in the exam) and forget them immediately after.
Naturally, this only works on knowledge-base exams, not skill-based exams (e.g. math). This also shows up in the results: math exams generally seem to have worse grades than other exams, and math exams seem to have lower thresholds to pass (e.g. 25 % vs. 60 %).
Spaced repetition is not the same as cramming, and spaced repetition (if applied correctly) leads to long-term memory formation.
It's not perfect for every type of information, and there are some aspects of memory (most notably contextual memory) which can make it less effective than one might like. But it definitely is more useful than just for exam cramming...
I'm curious why you think maths is exempt from being 'gamed' per se, by this technique.
I'm also not sure where you're getting "math exams generally seem to have worse grades than other exams" and "lower thresholds to pass" from.
At university I specifically remember some of my friends not understanding the calculus they were doing _at all_, yet they could still answer questions because they rote learned the technique.
Spaced repetition is not the same as short-term memorisation. In fact I'd argue that you haven't fully memorised something with an SRS until the interval between reviews is at least a month or two -- which isn't really "short term" especially when you consider that school terms are also on the order of a few months.
I consider Spaced Repetition a mental model of how we learn and retain knowledge. I think your objection is with flash card systems and I don’t disagree with your assessment.
I think of Spaced Repetition more generally. A paper based Zettelkasten system can be enhanced with a paper based review scheduling system; say 43 folders as described by David Allen in Getting Things Done with a simple SuperMemo algorithm for interval spacing.
When we gain good recall of Zettelkasten structure/content I suspect that we inadvertently mimic Spaced Repetition. I can be convinced otherwise.
If you have poor recall, how do you know what you need to look up? How do you know what entry points you have in your 'zettelkasten'? If you have no intention to recall information, why even bother taking notes? Instead of searching your notes, or following a zettelkasten ID trail, search the internet. We want to be able to _use_ information.
I'm all for the zettelkasten method, for good notetaking habits...however there's also a tendency of many notetakers, or productivity porn on notetaking, to waste so much time 'notetaking' and 'organising' that they never _act_. (I've been very guilty of this in the past).
Recall isn't bad. It's essential. It is, however, not the _only_ thing to use. You want to use spaced repetition for the most important, most used information. Stuff that's a bit more niche, rarely used, yeah just search that.
I wonder why people conflate notes with recall. The root is probably that note taking is introduced in school for things you have to recall in exams later. For me the point is that I do not have to train the recalling because I can search my notes instead.
Maybe taking notes helps me recall things and maybe it doesn't. And maybe writing them gives me better recall than typing and maybe it doesn't.
But, yeah. I rake notes to remember things I need to do, to capture some key points, or to be able to "play back" a meeting/talk/etc. to others in either verbal or written form. This is one reason I usually type notes for anything that I'll want to play back in any detail; it's faster and easier to capture more or less verbatim than writing notes by hand. To the degree that recall is helped, I find it's mostly by "processing" the notes in some form.
Spaced repetition, while reasonably useful, doesn't come without its challenges as well.
For instance, how can you be sure at the outset that you have designed a good prompt [1] that will enable recall versus recognition?
And even once you have designed a good prompt, that brings up the question of epistemology: how can you be sure at the outset that the "information" you wish to ingest is factual and that it will be useful to you later on? It's easy to imagine memorizing a compelling aphorism like:
> The German word for "passion", which literally translates to "enjoyable suffering". Because if you think you’re passionate about something, you must be willing to ask yourself if you’re willing to suffer for it. If you are, that means you are passionate. Otherwise, it’s just a hobby. [2]
But, while beautiful and compelling, it turns out that this definition/translation of the word "Leidenschaft" is misleading: "Leiden" means "suffering" and "schaft" simply means "-ship". This is very similar to the word _passio_ which is the Latin root of the English word (and whose meaning shines through in "The Passion of Christ", for instance).
When using spaced repetition systems you must be extraordinarily careful around vetting the content that you are putting into them, and most of the time, you would be better simply "coming back to the same reading" on a given topic (thereby effectively building your own ad-hoc SRS process).
The brain exists in order to recognize patterns, but it also likes to over-fit!
how can you be sure at the outset that you have designed a good prompt[?]
The link to Matuschak that you supply explicitly argues that you need to go back and revise prompts based on developments in your knowledge and new connections found. This includes eliminating prompts that are too easy, reformulating them for greater personal import and adding new ones to shore up areas of weakness.
Interesting read. I have always felt very uncomfortable with lectures where the lecturer stands in the front and everybody listens. In my experience this setup is beneficial for about 10 minutes. After that almost all is forgotten, especially when the lecturer is speeding through a huge topic, because he needs to "get through it", sparing no time for the critical nuances.
The best learning experiences I've ever made were with the "digital gardening" method explained in the article. Especially the process of writing down what I'm trying to understand and putting it in my own words is hugely beneficial. That old trick of Feynman's where you take a topic you don't understand and explain it to someone – be it a person, your personal wiki or just a text note – really works wonders. And revisiting a note and changing it a bit also seems to reflect how we remember things: Not regurgitating them, but reliving them.
Apart from that I found the process of creating my own personal wiki highly rewarding and can only recommend it to others (if you have the time). It's an absolute joy to use something you've made and explain the creation of it to yourself using it.
In my case I've also made the wiki in such a way that I can add exercises to notes and then do spaced repetition on those (which was also a blast to implement). The exercises themselves can be functions so that things get mixed up from time to time. I don't often do the exercises, but the process of creating them adds another layer of understanding from my experience.
Edit: Continuation...
But this is just me getting excited over the personalized wiki I've created.
The article importantly mentions that "there is no one-size-fits-all app" to best understand something, so it takes some time and experimenting to figure out what works best for you (heck, I think I went through at least 5 iterations of note-making systems before I settled on this one for about two years now). And since we're on HN, I think everybody here has the skills to make their own.
I really wish schools would add a dedicated, mandatory subject about learning to learn. It would have saved me a lot of time I wasted back in Uni and afterwards just trying to figure out how to quell this urge to retain information.
I wrote a tool myself :). Like many other wiki tools, at the core of it are a bunch of web pages. In my case each page is a note (which is preferably as atomic of a concept as can be) that consists of a section for the note and a section for exercises (each with an id) related to that note. I can then parse those exercises and feed them into a spaced repetition algorithm and practice web app.
Well in my case it was a mix of using Go for templates and general backend stuff, a couple of React and D3.js apps on the frontend for visualization, a Postgres DB for storing exercise results and that's about it. Single notes for example are simple templates that can be used in Go and I write the HTML raw with heavy use of snippets.
A self-hosted personal wiki sounds like a fantastic idea! Is it just a html 'pedia? How do you host the media? I would love to see the "How this wiki was built" page of your wiki.
As a different suggestion, you can have a wiki on your machine with TiddlyWiki. [1] With that, there’s no need for a server or complex software deployment and maintenance.
True that, albeit building my own wiki that has it's own feel and way of working was worth it to me. I took some inspirations from TiddlyWiki, but this space is really packed with a bunch of great tools. The ones I liked most were:
Pretty much. One repo for the notes, one repo for the logic. I tried to keep it as flexible as possible so that I could theoretically integrate any kind of frontend/backend tech at a later point.
At the moment everything gets hosted locally unfortunately. I was planning on polishing it so that it's easier to setup. Then I can put it on a Show HN in the future as a kind of "how this was built" :) and give some inspiration to other people who thought of building their own.
Another great local-first wiki tool is Obsidian (https://obsidian.md/). Each page is just a markdown file in a dedicated folder. They have backlinks and a bunch of plugins.
A bunch of quotes can be found on the Wikiquote page on Feynman [0] together with the proper references, but these two stand out:
“What I cannot create, I do not understand.”
“Once I asked him to explain to me, so that I can understand it, why spin-1/2 particles obey Fermi-Dirac statistics. Gauging his audience perfectly, he said, ‘I'll prepare a freshman lecture on it.’ But a few days later he came to me and said: ‘You know, I couldn't do it. I couldn't reduce it to the freshman level. That means we really don't understand it.’“
His book Surely You're Joking, Mr. Feynman! also has a couple of anecdotes where he solves people's problems by asking them questions and trying to understand them. Unfortunately I read this some time back and the exact page numbers escape me.
Generally, if you search for Feynman Technique you'll find anything in the direction of "explain it to someone to understand", but I think there is a lot of misattribution going on here. Like Einstein, people love to stick his name on things.
And finally, an indirect, but in my opinion well-summarized reference to the way Feynman learned, was in the Richard Feynman section of Lex Friedman's 118th podcast [1].
Slightly off-topic, but still related: does someone have an explanation about the differences between a wiki and Roam or the Zettelkasten method ? Sure, the way to input content can be different, and Roam gives you more visibility into the way nodes are interlinked. But it seems to me those can be attained with plugins, it's not a fundamental difference in the principles behind.
I was curious so I looked it up. From zettelkasten.de/introduction/ when the author is talking about the principle of atomicity:
> Wikipedia, also, is not a web of thoughts, because you can only link to articles and sections within them, but not to individual thoughts inside the text. None of the addresses matches with any thought. Wikipedia is not meant to be such a thing. Rather, Wikipedia is an encyclopedia with each article containing information on a topic. Wikipedia is not a thinking tool but a tool for information retrieval.
I do both within the same app (Obsidian) with some plugins, so let me explain the difference from my perspective.
Wiki, in my mind, is more suitable for somewhat long and detailed notes about concepts. Random titles from my notes: conspiracy theories, neoliberalism, borders, norms, monopolies, habits, addiction, time management...
Zettelkasten is kind of the opposite of that. It's more about making connections between concepts than thoroughly researching certain concepts. Random titles from my notes: news kills creativity, inequality generates status anxiety which generates ecological damage, relationship between paywalls and echo-chambers...
These connections made in zettelkastens have no place in a wiki page because the point of wiki pages is to have a somewhat detailed explanation of concepts. Zettelkastens are much shorter (usually between 5-10 sentences and 1-3 sources).
Another way to look at it: zettelkasten is for stuff that you write once and rarely have a reason to expand upon (other than to create a new link to something similar), while wiki is for stuff that you know you're gonna have to return to with the goal of adding more stuff.
Thank you for your view, I guess Roam also adheres to the Zettelkasten way and that's why there is much more focus on the interlinking of nodes ?
It seems to me the difference lies more in the wiki software. I was asking the initial question because there are wiki engines with a lot of features, and what you are doing for Zettelkasten can be done in them as well: take for example tiddlywiki (https://tiddlywiki.com/) which organizes nodes in "tiddlers" instead of full-blown articles or pages. You can link to another "tiddler", or you can even include it so its content can be reused elsewhere without having to "click". So it seems to me those new systems are more revealing of todays' wiki engines' limitations than completely new paradigms.
Another difference I just realized: wikis are expected to be read and modified by multiple people, while Roam and Zettelkasten systems are expected to be used by one person only. While not written in stone, I feel like this has an impact on the way content is structured.
I used to write notes to keep me busy and engaged during class and now meetings. It’s almost like an acceptable form of doodling. But I rarely ever go back to my notes. I think everything in my notebook can fit in one page in terms of useful information. I hated when my managers would tell me writing things down is important. If it’s important, I’ll document it correctly, not scribble it in a book.
I agree that there is a difference between simply taking notes and simplifying the content and writing down your own take.
I'd argue that a good note-taker will be doing the latter automatically.
I also disagree with the claim that note taking doesn't enhance learning. It forces one to pay attention rather than daydream and ignore the content entirely. It would be less tiresome, of course, to just pay attention and not take notes, but it does serve that purpose.
Another purpose for note taking is to pass the notes on to someone not present during the class/lecture/talk. For example: I was paid to take notes for others in college.
Another use for notes is to look back on the notes and verify that you have effectively absorbed the material; studying portions that you don't understand fully.
The article is far too dismissive on the value of plain note taking.
34 comments
[ 3.2 ms ] story [ 49.2 ms ] thread[1] https://en.wikipedia.org/wiki/Spaced_repetition
Naturally, this only works on knowledge-base exams, not skill-based exams (e.g. math). This also shows up in the results: math exams generally seem to have worse grades than other exams, and math exams seem to have lower thresholds to pass (e.g. 25 % vs. 60 %).
It's not perfect for every type of information, and there are some aspects of memory (most notably contextual memory) which can make it less effective than one might like. But it definitely is more useful than just for exam cramming...
I'm also not sure where you're getting "math exams generally seem to have worse grades than other exams" and "lower thresholds to pass" from.
At university I specifically remember some of my friends not understanding the calculus they were doing _at all_, yet they could still answer questions because they rote learned the technique.
I think of Spaced Repetition more generally. A paper based Zettelkasten system can be enhanced with a paper based review scheduling system; say 43 folders as described by David Allen in Getting Things Done with a simple SuperMemo algorithm for interval spacing.
When we gain good recall of Zettelkasten structure/content I suspect that we inadvertently mimic Spaced Repetition. I can be convinced otherwise.
I'm all for the zettelkasten method, for good notetaking habits...however there's also a tendency of many notetakers, or productivity porn on notetaking, to waste so much time 'notetaking' and 'organising' that they never _act_. (I've been very guilty of this in the past).
Recall isn't bad. It's essential. It is, however, not the _only_ thing to use. You want to use spaced repetition for the most important, most used information. Stuff that's a bit more niche, rarely used, yeah just search that.
But, yeah. I rake notes to remember things I need to do, to capture some key points, or to be able to "play back" a meeting/talk/etc. to others in either verbal or written form. This is one reason I usually type notes for anything that I'll want to play back in any detail; it's faster and easier to capture more or less verbatim than writing notes by hand. To the degree that recall is helped, I find it's mostly by "processing" the notes in some form.
For instance, how can you be sure at the outset that you have designed a good prompt [1] that will enable recall versus recognition?
And even once you have designed a good prompt, that brings up the question of epistemology: how can you be sure at the outset that the "information" you wish to ingest is factual and that it will be useful to you later on? It's easy to imagine memorizing a compelling aphorism like:
> The German word for "passion", which literally translates to "enjoyable suffering". Because if you think you’re passionate about something, you must be willing to ask yourself if you’re willing to suffer for it. If you are, that means you are passionate. Otherwise, it’s just a hobby. [2]
But, while beautiful and compelling, it turns out that this definition/translation of the word "Leidenschaft" is misleading: "Leiden" means "suffering" and "schaft" simply means "-ship". This is very similar to the word _passio_ which is the Latin root of the English word (and whose meaning shines through in "The Passion of Christ", for instance).
When using spaced repetition systems you must be extraordinarily careful around vetting the content that you are putting into them, and most of the time, you would be better simply "coming back to the same reading" on a given topic (thereby effectively building your own ad-hoc SRS process).
The brain exists in order to recognize patterns, but it also likes to over-fit!
[1] https://andymatuschak.org/prompts/
[2] Keller, T., Heller, S., Ruhlman, M. and Jones, D., 1999. The French Laundry Cookbook. New York: Artisan.
The link to Matuschak that you supply explicitly argues that you need to go back and revise prompts based on developments in your knowledge and new connections found. This includes eliminating prompts that are too easy, reformulating them for greater personal import and adding new ones to shore up areas of weakness.
It is difficult.
The best learning experiences I've ever made were with the "digital gardening" method explained in the article. Especially the process of writing down what I'm trying to understand and putting it in my own words is hugely beneficial. That old trick of Feynman's where you take a topic you don't understand and explain it to someone – be it a person, your personal wiki or just a text note – really works wonders. And revisiting a note and changing it a bit also seems to reflect how we remember things: Not regurgitating them, but reliving them.
Apart from that I found the process of creating my own personal wiki highly rewarding and can only recommend it to others (if you have the time). It's an absolute joy to use something you've made and explain the creation of it to yourself using it.
In my case I've also made the wiki in such a way that I can add exercises to notes and then do spaced repetition on those (which was also a blast to implement). The exercises themselves can be functions so that things get mixed up from time to time. I don't often do the exercises, but the process of creating them adds another layer of understanding from my experience.
Edit: Continuation...
But this is just me getting excited over the personalized wiki I've created.
The article importantly mentions that "there is no one-size-fits-all app" to best understand something, so it takes some time and experimenting to figure out what works best for you (heck, I think I went through at least 5 iterations of note-making systems before I settled on this one for about two years now). And since we're on HN, I think everybody here has the skills to make their own.
I really wish schools would add a dedicated, mandatory subject about learning to learn. It would have saved me a lot of time I wasted back in Uni and afterwards just trying to figure out how to quell this urge to retain information.
[1]: https://tiddlywiki.com
Logseq (open source): https://logseq.com/
Roam Research: https://roamresearch.com/
At the moment everything gets hosted locally unfortunately. I was planning on polishing it so that it's easier to setup. Then I can put it on a Show HN in the future as a kind of "how this was built" :) and give some inspiration to other people who thought of building their own.
“What I cannot create, I do not understand.”
“Once I asked him to explain to me, so that I can understand it, why spin-1/2 particles obey Fermi-Dirac statistics. Gauging his audience perfectly, he said, ‘I'll prepare a freshman lecture on it.’ But a few days later he came to me and said: ‘You know, I couldn't do it. I couldn't reduce it to the freshman level. That means we really don't understand it.’“
His book Surely You're Joking, Mr. Feynman! also has a couple of anecdotes where he solves people's problems by asking them questions and trying to understand them. Unfortunately I read this some time back and the exact page numbers escape me.
Generally, if you search for Feynman Technique you'll find anything in the direction of "explain it to someone to understand", but I think there is a lot of misattribution going on here. Like Einstein, people love to stick his name on things.
And finally, an indirect, but in my opinion well-summarized reference to the way Feynman learned, was in the Richard Feynman section of Lex Friedman's 118th podcast [1].
[0] https://en.wikiquote.org/wiki/Richard_Feynman#Quotes
[1] https://youtu.be/U_6AYX42gkU?t=452 (with exact starting time)
https://ponder.joeldare.com
Full disclosure; I created it for the purpose.
> Wikipedia, also, is not a web of thoughts, because you can only link to articles and sections within them, but not to individual thoughts inside the text. None of the addresses matches with any thought. Wikipedia is not meant to be such a thing. Rather, Wikipedia is an encyclopedia with each article containing information on a topic. Wikipedia is not a thinking tool but a tool for information retrieval.
Wiki, in my mind, is more suitable for somewhat long and detailed notes about concepts. Random titles from my notes: conspiracy theories, neoliberalism, borders, norms, monopolies, habits, addiction, time management...
Zettelkasten is kind of the opposite of that. It's more about making connections between concepts than thoroughly researching certain concepts. Random titles from my notes: news kills creativity, inequality generates status anxiety which generates ecological damage, relationship between paywalls and echo-chambers...
These connections made in zettelkastens have no place in a wiki page because the point of wiki pages is to have a somewhat detailed explanation of concepts. Zettelkastens are much shorter (usually between 5-10 sentences and 1-3 sources).
Another way to look at it: zettelkasten is for stuff that you write once and rarely have a reason to expand upon (other than to create a new link to something similar), while wiki is for stuff that you know you're gonna have to return to with the goal of adding more stuff.
It seems to me the difference lies more in the wiki software. I was asking the initial question because there are wiki engines with a lot of features, and what you are doing for Zettelkasten can be done in them as well: take for example tiddlywiki (https://tiddlywiki.com/) which organizes nodes in "tiddlers" instead of full-blown articles or pages. You can link to another "tiddler", or you can even include it so its content can be reused elsewhere without having to "click". So it seems to me those new systems are more revealing of todays' wiki engines' limitations than completely new paradigms.
Another difference I just realized: wikis are expected to be read and modified by multiple people, while Roam and Zettelkasten systems are expected to be used by one person only. While not written in stone, I feel like this has an impact on the way content is structured.
I'd argue that a good note-taker will be doing the latter automatically.
I also disagree with the claim that note taking doesn't enhance learning. It forces one to pay attention rather than daydream and ignore the content entirely. It would be less tiresome, of course, to just pay attention and not take notes, but it does serve that purpose.
Another purpose for note taking is to pass the notes on to someone not present during the class/lecture/talk. For example: I was paid to take notes for others in college.
Another use for notes is to look back on the notes and verify that you have effectively absorbed the material; studying portions that you don't understand fully.
The article is far too dismissive on the value of plain note taking.