Hacker News new | past | comments | ask | show | jobs | submit login
Remembering what you Read: Zettelkasten vs. P.A.R.A (zainrizvi.io)
118 points by stanulilic on Aug 24, 2020 | hide | past | favorite | 62 comments

My takeaway from the recent trend in articles talking about how great the Zettelkasten method is:

Document what you do, as if you were describing your work and/or learning to a stranger. That stranger is you, in 6 months to 5 years time.

Maintaining that enthusiasm and finding the time to document is difficult, but the results are very valuable because you are building a "second brain". That extra brain can be indexed, searched, tagged, analyzed, and edited using many powerful tools.

Don't waste energy chasing fancy tools and methodologies without already having a simple workflow in place. In other words, don't go all out learning Emacs+org-mode+org-roam when you already have a directory of text file notes. Once you have a good idea of what works for you, then introduce tools designed to make your life easier. (I say this as someone who uses Emacs+org-mode+org-roam every day)

Oh and read "The Checklist Manifesto" and "How to take Smart Notes".

> How to take Smart Notes

I have read over 100 pages of the book but to be honest, I am disappointed with it. I learn best by examples, the book is failing short of examples on how to use Zettelkasten. My expectation was the book was going to walk me through the process of implementing Zettelkasten like a tutorial, but instead it keeps going back in circles, talking about why I should use Zettelkasten. I am about to finish the book but I still have a lot of questions.

OP here

Yeah, that book was pretty light on practical advice. I didn't find much in there that I hadn't already discovered by reading blog posts about it

One thing that was unique though, was a set of questions to ask yourself while reading to get extract the most insights out of the material.

I posted them here if you're interested:


Genuine curiosity: why do you take the extra effort to choose and then add an emoji at the end of most of the sentences in these tweets?

It doesn't take much effort. Win + '.' opens up the emoji selector, and just start typing a word to search the emojis.

Why bother? Sometimes they help to convey emotion that would otherwise be missing. Other times it helps to break up dryer blocks of text.

That, plus it's fun :)

Weird suggestion that might help:

My second reading of Test Driven Development, I literally did the entire book along with the author, front to back.

That's when it clicked. The actual manual mechanical kinesthetic activity somehow made it work for me.

"That stranger is you, in 6 months to 5 years time."

Seemingly, one of those lessons that must be experienced to be learned.

30 years ago, I added full comments to (my source of) a text editor during a rewrite. Back when we still debated such things, before Code Complete 1st ed.

Thereafter, all maintenance was a breeze, even years later. Like a gift from my prior self.

When I write (code, notes, journals), my headspace is kind of a hybrid of rubber ducking (thinking out loud, like during usability testing) and telling future me a story of me.

OP here

100% agreed on the "don't chase fancy tools" mindset.

When I was taking the Building a Second Brain course, the instructor Tiago Forte had offered us three tools to use for organizing our notes:

- Evernote - Roam - Notion

He personally used Evernote, but he had an interview each with someone to show us how hey used Roam or Notion.

Since all the examples he would show us during the class would be in Evernote, I decided to not bother trying to apply PARA to the other two systems and stick with Evernote until I learned the basics. I could always switch over later, once I grasped the fundamentals better.

I haven't switched over yet, because frankly, Evernote is still good enough for me. The potential switching costs aren't worth it.

> Document what you do, as if you were describing your work and/or learning to a stranger.

I picture it as "doing engineering the way a coroner conducts an autopsy." (Maybe only a helpful framing if you've watched enough CSI, but IMHO a pretty powerful mental lever if you have.)

I love reading and have a frustratingly limited memory. I spoke to my brother about it (who was the better student between us) and he highlighted similar issues for himself which surprised me. His advice was as follows: “accept the fact that your memory sucks and figure out an annotation and note taking system that helps you recall things faster”. You don’t need to remember the notes about a particular topic if you can remember that you took a note about it.

How we implement it:

1- he and I annotate almost everything we read worth annotating (I wont do it for escape reading)

2- our annotations have similar systems across books. Underlines may be something fun or interesting, stars highlight significant points, “Qs” pose thoughtful questions, top left corner is location in plot top right is general theme of the scene etc

One and two allow us to get up to speed on a book we previously read very quickly or find particular sections in an instant.

For teachable items - books or notes I don’t want to forget that are nuanced - I transcribe my annotations to OneNote (my favorite note taking app).

For example I read a book about Japanese death poems. There were a few that really resonated to me. I don’t need to ever pick this book up again nor do I need to remember any of them, I just have to remember I wrote them down. Here is one of my favorites:

My heart

is a bottomless river,

a raging torrent.

(Written by a woman who took her husbands life, for taking the life of her brother, moments before she took her own.

Anyway this system works for me and thought I would share.

Wow, I really like how you reserve the corners of the page for specific content.

I also really like writing notes in the book, but I take a picture of each page and stick it into Evernote. On each note I'll write down the book and page number that the note came from in case I want to see it in more context.

That way the notes are all searchable and it really helps when I don't remember which book I read something in

Shout out for Joplin [1] for digital note-taking - coupled with Nextcloud, I have a self-hosted note repository that syncs across my mobile, laptop and desktop. Interface is clean, notes are in Markdown format, and notes can be cross-linked and organised into notebooks. And it's OSS.

Not affiliated, just a fan!


+1 ,but hope it gets bi-directional links similar to roam, obsidian, etc one day.

either way, underrated open source, free, note taking app.

I could be wrong, but I think as long as you have access to the same Dropbox account, you can sync up as many devices as you want. No nextcloud integration required.

Yes, Nextcloud is not required - I'm saying that having that gives me self-hosted synchronisation. You can indeed use Dropbox; there are other options too.

Queue org-mode namedrop in 3.. 2..

Actually, I can just be that guy as I use it myself, and it is great. However I do find myself spending energy I wish I didn’t on meta-things like how to organize it all (files, tags, hierarchy etc). I think it could benefit from having preconfigured setups like doom-emacs or ruby on rails or whatever («convention over configuration»). One for a hardcore GTD, one for zettelkasten, one that is bare bones, etc..

I realize the agenda should make this less important, but I still get caught up in it from time to time (tags? Properties?)..

I haven’t tried org-roam though. I just feel like theres so many moving parts to my setup already. Anyone had any good/bad experiences integrating it into your workflow?

Speaking of org-mode, maybe I should start looking at those scheduled items in my agenda that are now some 600 days+ past its due date...

Totally anecdotal: I've been an org-mode user for years, and have spent some time using RoamResearch directly as my primary notetaking tool.

I _still_ haven't figured out how to use org-agenda properly, and it feels like I'm missing a lot of core functionality as a result.

I tried to get into org-brain for a while. It never stuck for me.

RoamResearch was a lot stickier; I may end up back in it at some point because of the mindshare and effort being poured into it. It's a really great product--my threshold for being willing to do my thinking work inside a program that's finely tuned to distract me (the browser) is high, so this is saying something.

In the meantime, though, I've been giving org-roam some love, more than anything else because it's _not_ in the browser distraction machine (to wit, here I am commenting on HN when I should be billing time...). Getting spun up and fully in the habit of using org-roam took me a couple days of background effort. I find it easy enough to forget it's there, which is good enough for me.

Separately, I suspect people that don't want to have to customize every aspect of their editing experience are almost by definition not emacs users :P

OP here

I wrote this article last May and back then it was unclear whether or not I'd stick with PARA

Here's the update three months later:

The Bad: I’m kinda getting falling behind on some of the proactive archiving/organizing parts of it, especially for my personal notes

The Good: Falling behind where PARA's Just In Time aspects really shine. I can organize just the notes that I’m about to use

The Great: BASB made a night and day difference to my project management style

In the past I organized my notes and todos in a completely different way. I was constantly struggling to find the important bits or would miss important tasks. I'd always be feeling overwhelmed and lost

I still feel like that today, but it's no longer because of my notes. My system is no where near perfect now, but it's waaaaaaay better than what it used to be

I shared more of my thoughts in this tweet thread [1] describing the advice I've been giving to my friends who asked if they should sign up for the Building a Second Brain class (where I learned PARA)

[1] https://twitter.com/ZainRzv/status/1296532147096834048

> I hated the idea of wasting all that time I spent reading

Personally, I've come to the conclusion that I read in order to draw broad patterns, not necessarily to remember all the things in every book - my brain has limited space.

What are your thoughts on optimizing retention at the expense of remembering to see the forest?

100% agreed. It's been a journey for me to get to this point, but Tiago Forte's thoughts were really influential for me here.

Now I try to only save notes that I found particularly surprising or insightful, I'm definitely not trying to save everything in the book. I don't put it in:

* If the fact didn't make me think

* If it doesn't trigger an "aha" moment

* If it was something I already knew,

For example, I'm currently taking notes as I read The Art of Doing Science and Engineering. If you looked at my notes[1] then you could barely tell that it's a book about engineering.

[1] https://twitter.com/ZainRzv/status/1289964671563214848

>>> "... Farnam Street’s tips on remembering what you read. ... writing your notes on the book itself ..."

Almost all of what I read is not on paper. If I'm reading something that I'm serious about remembering, I'll create a precis ("a concise summary of essential points, statements, or facts" [1]). My goal is to reduce the size of the content by 90% or more (10 page article becomes a 1 page precis, 200 page book becomes 20 page (or smaller) precis. The process of creating the precis seems to enhance my recall of the important information, and if I need to refresh my memory of the material, I can do it in a tenth of the time.

[1] https://www.merriam-webster.com/dictionary/pr%C3%A9cis

I do something similar but I keep shortening and shortening my summary until I reach the point in which I have a couple of key points at the top of the note, with everything longer than that somewhere underneath.

That way I can quickly catch a glimpse of what the note offers as soon as I open it, and in case I need to understand the arguments behind those points, all I have to do is to scroll down.

This has proven to be quite effective when I want to reference something I've read about years prior.

That's a great way to learn the content deeply.

I tried doing that and failed miserably however, I just couldn't stick with it, lol. That's where PARA's highlight based, minimal effort note taking system really helped me out.

Is it the best option for everyone? Definitely not. But the system that I can stick with beats one that I won't follow.

The issue at the heart of both the Zettelkasten and the PARA methods seem to be that a person wants their notes to be connected in useful ways. To an extent, "ordinary" knowledge graphs permit this -- where by "ordinary" I mean that the edges are unlabeled. Things are related to other things, but you have to infer the relationship yourself after looking at both of its members.

Graphs with labeled edges seem clearly better. But then what if you want a relationship with more than two members? (Or less! "Not" and "maybe" are two important unary relationships.)

And what if you want a relationship's members to themselves be relationships? This is the nature, after all, of most information. Any but the simplest sentence is a relationship between relationships. "Children ask for cookies because cookies have sugar", for instance, is a "because" relationship bewteen an "ask for" relationship and a "have" relationship.

Once you've reified your relationships that way, you'd like to be able to search intelligently through them. "Ordinary" knowledge graphs let you see all of a thing's children, or parents, or maybe in some cases all it's descendents and ancestors up to some level of recursion. Better would be something that lets you ask targeted questions, like "Show me everything Sally has asked me to do that does not require a computer."

To my knowledge there's only one program that lets you reify your data and query it in the ways I've described. It's called Hode, for Higher Order Data Editor. I wrote it.


That's pretty cool Jeff! Those insights are probably used a lot by researchers working on general AI.

How much work does it take you to add your notes to it? I'm curious about it on two dimensions:

* The time it takes to add a note

* What decisions you need to make to add a note (what to link it to, what not to)


You've zeroed in on the weakest point of the program. Adding notes is mechanically easy but conceptually hard. That's because there are many ways you could choose to represent the same concept.

For instance, suppose "harry gave sally a book" is in your graph. One way you could represent that is "Harry #gave Sally #object a book". Another is "Harry #did (give #object a book) #to Sally". Another is "Harry #did (give #object a book #to Sally)". And actually maybe you'd prefer to represent "a book" as "#a book", so that you can just search for "book".

They're all plausible representations, and all have different implications for how you would run searches. You've got to be consistent in how you represent those relationships, and the program can't help you because I mean it's a database, not a general artificial intelligence.

The strength of the program is that the structure of your index is entirely up to you. But soon that freedom gives way to being bound by the decisions of your former self. And your future self will inevitably understand the space you're representing better than your former self did when it bound you.

Got it. Yeah, this is one of those instances where the paradox of choice can kill you.

Just a thought: if I were interested in adopting a tool like that, my chances of successfully sticking with it would go up 100x if there was a set of "default recommended practices" to follow when saving a new note, to cut down on the decision I would need to make.

Then over time, as I got more familiar with the tool, I could always tweak those default recommendations to fit into my own style

And who better to offer those default practices than the tool's creator? :)

(Mechanically, adding a relationship is as simple as typing "/add I #like turtles" and pressing enter.)

sounds very much like a mind map. i don't know if any mind map programs allow that kind of search though.

The thing I can't wrap my head around is how come Amazon does not have an API where I can pull my Kindle highlights. I know there are services out there like readwise.io but they require me to manually go in and "sync" things using my browser.

I would pay money for a service that was able to grab these highlights and put them into a Zettlekasten or P.A.R.A. style repository. Even Toi Lutke made something to help accomplish this... https://github.com/tobi/highlights

I agree that automation would help out here--could you just hit Readwise's 'sync' page with `curl` and a cronjob?

Admittedly I haven’t tried either of these methods, but just from reading this post, they both seem prohibitively over-complex. There’s also no mention of Spaced Repetition, which is unquestionably the best method for remembering things.

Instead of complex note systems, try this: write down quotes, paragraphs, summaries and other related information, add them to Anki / another SRS, and do the cards everyday. That’s it. Simplicity is key to building habits and I’d bet any money that SRS works better anyway.

(OP here) The philosophy behind PARA is different from spaced repetition.

Spaced repetition is super valuable when you want to keep all the information inside your head. PARA is more geared at when you want all the information available when you need it, and requires much less time/effort than Anki (I've tried both).

Not to say that one is better than the other. Sometimes you need to access the information and cannot use the note taking system (e.g. you're a student and have a final exam coming up). But if memorization is not a requirement, then I find these other systems easier to use.

Ah okay, that makes sense. I can see the use case.

Personally, I find they're two tangential, but distinct goals. I use anki for things I want to be able to recall instantaneously, like a foreign language, and I use markdown notes for general reference of media/knowledge I've come across, to record my fresh understanding for later. I don't need these memorized- just an easy reference.

Contrary to OPs opinion, I believe tooling is relevant to the learning process, because an inconvenient system will put you off from taking notes and from coming back to them in the future. The catch is, there is no single tool that will fit everyone, you have to find your own method. My own tooling choices have changed a lot across the years.

When I began to apply a structured method to my studies, my first choice was to use mind maps. You can structure a mind map to resemble the associations you have in your own (duh) mental model of any body of knowledge you might be studying, and then it's easy to navigate a map with your eyes to recall stuff and fill the gaps when things get cloudy in your head. The first digital tool I recall using was CmapTools, and it's still pretty good.

Nowadays I use mostly LiquidText for reading notes, on books and articles (but there are other similar products). It lets you build something similar to a mind map, and it's really convenient to select a piece of text or image on a pdf and make it a node in a map. It's especially convenient if you have an iPad Pro with a Pencil. However, I could still not find any single tool that encompasses all my "learning modes". For instance, if I am playing with some OSS project and want to take note on something interesting, I have to use a different product.

OP here, if the article seemed to imply that tooling is irrelevant, that was not my intent. The most important part of any system is that it be something you'll stick with. I 100% agree with you that if a system puts you off from taking notes than it's bad for you. (I'd add the nuance that what's off-putting for one person might work well for another)

That said, it's dangerous to let the perfect be the enemy of the good. I'm trying not to spend too much time researching new systems and instead focus on benefiting from the good-enough system I currently have. I moved to my current system when I was feeling the scaling pains of my old one, and when I feel those pains from this system it'll be time to search for an upgrade!

Speaking of upgrades, I don't have a good way to take notes from pdfs. Looking at their site, LiquidText seems like it's optimized for the pdf use case. I'm going to dig deeper into it to see if I can read pdfs and take notes in there, and then export those notes into Evernote. Thanks for the tip!

A note on folders: I think grouping is a good thing alongside linking notes. I like to combine both methods, but instead of sorting notes into folders, I tag them. For personal knowledge bases it's probably fine to "silo" everything, but I found it really useful to add tags to notes when collaborating on the same knowledge base with multiple contributors. That way I don't need to rely on someone elses judgement and can group them from my perspective without colliding with the author. So instead of having folder -> articles, I have article -> tags and an can query by tags.

Emvi [1] has a tag overview page to allow just that, in case you're interested. I'm using it alongside 7 other people, where 3-4 are actively contributing to it (writing).

[1] https://emvi.com/

Do you have private tags visible just to you? Otherwise it seems like the system would get pretty cluttered with everyone's tags

For my work notes I've been copying links to the team docs into my personal note taking system and organizing those as I wished

Also a good way to handle it.

No, Emvi does not support private tags. But usually you won't need many tags to identify it for everyone. I was looking for the fixed IP addresses of our local network recently and couldn't find them by searching for "IP", because it was tagged as "network". Adding IP as a new tag is beneficial for everyone.

You can bookmark, watch, or create a (private) list for stuff you look up frequently.

It seems like the only real difference is where you put the effort. With Zettelkasten (Z), it's while reading, with PARA, it's before and after a project which might need those notes.

The Z method seems more scalable to me, since if I'm reading, I have time to spare.

(OP here) The argument I tried make in the article was twofold:

1) You read waaaay more than you use. So only summarizing what you're about to use in a project takes less time total. (and even then, it's just highlighting passages)

2) The best system is the one you stick with. Yeah, in an ideal world I'd be summarizing my readings right as I'm reading it (clearly I have time to spare in those moments). I tried to do that for a while. But I just couldn't stick with it. It felt too much like homework. It was a chore. The motivation wasn't there. This PARA based system is one I can actually stick with

It is hard enough for me to find the time to do the reading, let alone to also do a ton of writing about it. It especially takes me a long time to write by hand, and is just logistically difficult - where do I put the notebook? do I have to do all my reading at a desk? I can type fast, but this has even worse logistical challenges - do I have to do my reading next to a computer? oh, I'll just go check my email real quick while I'm at the computer...

I sometimes daydream about a dedicated writing machine, something small, but with a full-sized keyboard so I can type quickly, and supporting nothing else but text. Sometimes I honestly think about just getting a typewriter, but it seems silly.

Me too. I looked into electronic typewriters that export to rtf or something but never pulled the trigger. Would be interested in hearing if people do use specialist devices like that.

After writing this comment, I've started wondering about tablets with nice keyboard attachments and only using it with an app like Joplin. Seems like it might be a similar solution. You can install email apps on tablets, but you don't have to.

I'm curious about something. To those who have used advanced note-taking methods over a period of say a few years:

Has the method itself helped you in producing work that, without said method, would not have been possible or would have been much more difficult?

For zettelkasten, the one person for which the method demonstrably led to tangible results was Niklas Luhmann himself. Outside of that one data point, I'm hard pressed to find any other real-life examples.

How does Rich Hickey approach note-taking? Or John Carmack?

Before I moved to a Zettelkasten style of "second brain", I used to dump interesting things I've done or learned into my personal wiki page at work. Stuff like "awk that extracts this from that" or "3 different ways of doing X in Puppet and Ansible".

The problems I had with that was that it was located at work, so I couldn't access it easily off-line. Nor could I search and location a particular nugget of information particularly quickly. It was also separate from my daily journal.

In my current setup, my "3 different ways of doing X in Puppet and Ansible" has links to an Ansible page, a Puppet page, and more. Each of those Ansible and Puppet pages have a list of backlinks, so I now have a list of all the tickets that used Puppet or Ansible, and all the daily notes that involved those tickets. So I can spider through my second brain fairly quickly without searching. Which doesn't sound that important but if you don't know quite what you are looking for, then you can remind yourself by walking through the stuff associated with it.

Not the greatest example, sorry. But my "second brain" is still evolving and getting more useful. I'm excited to see what researchers come up with in this problem space!

I use zettr app for note taking. A private slack (with friends) for para. Took the udemy superblearning course a few years ago and picked up a lot of great tips from that too.

How do you use slack for para?

someone who has done this over a period of years, as compared just reading a lot of books, what is the result?

The only time it’s acceptable to write notes directly into a book is when you’re taking it into an exam. Otherwise, it’s just an act of vandalism!

Lol, two-year-ago me would be scandalized at how I now defile books with my vile notes

Pictures if you want them: https://twitter.com/ZainRzv/status/1289964671563214848

Argh! The horror :)

I like you :)

Can you expound more on why you think this? Do you think it’s okay to write in a book you own and don’t plan to get rid of?

Saying it’s vandalism seems a bit extreme to me—-unless we’re talking about a book borrowed from a friend or a library.

I don’t really write notes in margins much anyway. It never seemed useful to me to bind my notes directly to the pages that they come from. They become too hard to find.

I think it stems from an unreasonable reverence for the printed word. You never own a book, you’re just its custodian for a period of time!

I’m only really calling it vandalism with my tongue firmly in cheek–obviously anyone can do what they like with their own books :)

I totally get it. I used to revere books waaaaay to much to let any blemish willingly tar them.

Now though, I'm delighted to write notes all over

I think it's important to be specific about the different purposes of taking notes. I tend to split it into two spectrums, the reference spectrum and the recall spectrum.

The reference spectrum ranges from looking up a small, discrete nugget of information to restoring a complex mental context. At the "small nugget" end of the spectrum could be a command line invocation for a specific task, or your mother's birthday. In my experience, search works fine for this if you put in minimal effort when recording the information.

At the "complex mental context" end of the spectrum could be a project that you were immersed in six months ago and had to set aside. Documentation for this purpose takes a lot of care; it's essentially writing for another person. If you're not careful and systematic, you can end up accidentally taking a crucial aspect of context for granted and leaving it out. To choose how to structure notes for this purpose, I think of myself as a consumer. I have a lot of experience with consuming information, and I know what I like. I don't like mind maps. I like complete sentences of clear, well-structured prose, with an occasional diagram, lightly hyperlinked where it's helpful. All I need is links, basic formatting, and occasionally embedded media.

The recall spectrum ranges from memorizing specific factoids to trying to cement more complex ideas so they're available for thought later. At the factoid end is anything that could fit on a flash card, such as the definition of a word or, whoops, my mother's birthday, which probably belongs here under recall instead of under reference! At the complex end of the recall spectrum is where I would like to put the best books I read.

Again, when taking notes for recall, I think of what helps me in my experience. I like studying Anki cards for factoid recall. For the complex end, I find the effort that is best rewarded is the work of writing straightforward, well-structured prose. It's very easy to sketch graphs and diagrams that feel meaningful, but the ease is a giveaway that you aren't really mentally engaging. Writing is much harder work, because it forces you to engage with the subject in a deeper way than just splashing a vague intuition down on the page. The value is reflected in consumption: well-written prose is invaluable, while most of the images that feel like they capture part of the mental context when I have it are no help in restoring it when I don't. (Technical subjects are an outlier; diagrams are often helpful.) To be clear, since recall is the object, consumption is secondary; what is most important for me is that the act of writing well-structured prose helps cement the knowledge in my mind. It's just a nice bonus that if it doesn't stick the first time, I'll have exactly the kind of notes that are good for refreshing myself later.

Maybe it's just me, but I think discussions of note-taking techniques and structures are skewed by the desire to create something visually impressive. Nobody who is a professional at communicating information to others is using these techniques. I've never heard anybody say, I need to find a mind map to improve my understanding of the French Revolution, or Franz Kafka, or Apache Kafka. Mind maps and similar techniques have been around for many decades and have yet to prove themselves as a medium for teaching or communication. I'm very skeptical of the idea that they have some great yet-to-be-exploited utility for communication, and I'm also skeptical of the idea that their value for personal use vastly exceeds their value for interpersonal use. But this is subjective, and perhaps not everybody is the same as me.

I really like how you distinguish saving nuggets from complex mental constructs. That's a super valuable call out.

Regarding mind maps: I've personally found them useful for thinking in the moment and brainstorming. They were also helpful for studying for a test back when I was in school (and the concepts it had to include were limited to a single chapter)

They're definitely less useful as a way to transmit information to other people though.

emacs users check org-roam its quite nice

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact