I have been thinking about and writing about personal knowledge management a lot – just not in a concise, blog-forward way. I've had tons of conversations with people in private and in public fora, made reference to personal knowledge management in the context of multiple podcasts, but I've never really put together a solid comp. of thoughts from me regarding how to do things like think about breaking down notes on an article, or the underlying structure that you might want to think of your note taking within. (However I did do a recent article about breaking down interviews for Obsidian and PKM reference: How to work on an interview in Obsidian
Part of the reason for that is once you start talking about knowledge acquisition in the abstract, you start sounding like a pretentious git. You start sounding like someone who wants to sell something, and I'm deeply uncomfortable with that.
I'm not going to try to sell you something. I don't have a $300, six episode course to try and teach you to be a better person who is more creative with perfect recall and an absolute success in your field. I don't dress nearly well enough to pull that off and I would feel really bad about taking your money in exchange for something and that I would otherwise tell you for free.
The Deal
So here's the deal – this is going to be an article that folds together some of my ideas from my notes, some of them literally from cut and pasted conversations (thanks mifga), and it's probably something that will get referenced and expanded as we go forward. Much like my personal Vault, this is just a collection of ideas which are linked to one another in order to make thinking about my thinking possible.
If you would like these ideas, if they are meaningful to you, if they help you out – by all means – let me know.
If they spark questions in you – even better. Write them down for yourself and then let me hear them. I'll write them down for myself and maybe come up with good answers, or perhaps I'll be just as boggled as you are and start searching for answers. (Sometimes I'll tell you that you're asking a crap question because that's the nature of the world. Not everything is gold.)
If you feel the urge to send thousands of dollars in your choice of cryptocommodity or cash, I will have a section at the bottom which instructs you how to do so. Knock yourself out. I prefer my lambos in black and red, my houses in Northern rather than Southern California, and my offshore accounts Panamanian. Just as a note.
Tools for the job
My own preference, at the moment, for organizing my personal information and knowledge base is Obsidian, and you are going to see multiple references to it throughout. You can get it for free, you can use it for free, you can use it completely privately because it doesn't require any shared file space out on the Net – it just works. It is effectively your own private wiki.
Throw in some community plug-ins, and a little creativity and it can be disturbingly customized. Drop a few dollars on the creators and they'll even give you the ability to host your entire Vault publicly in a pretty nice format. Or you can use Jekyll or an equivalent to build your own static website and update it however you like.
I'm not trying to sell you Obsidian. I am telling you it's an easy, cheap, fast way to get your hands dirty and start actually doing something for your personal knowledge management.
Any wiki will do.
It doesn't even, technically, have to be a wiki – you can probably follow most of my advice if you're one of those dirty, filthy org-mode people, and you know who you are, Eric. All you need is an information storage solution that allows you to consciously and actively connect document nodes together with some sort of explicit relationship.
But for me, right now, we are going to be talking Obsidian because it's what I use.
A brief thought about systems
Replace "organization" with "individual" in the description, and knowledge management turns into personal knowledge management (PKM).
On the zettelkasten
A method of organizing ideas, thoughts, notes, and all the rest of that joke based on the work of an obscure German sociologist who liked to put things on 3 x 5 cards and stick them in boxes.
-- me
A lot of people love the zettelkasten method. If you don't know what I'm talking about, do a five minute Google search and you will be inundated with people who will swear in every way that as long as you go through the very precise motions and thought patterns associated with zettelkasten, you will be a productivity God, writing a book in about 13 minutes, being a masterful expert on any subject in a day, and giving orgasms to People of the Appropriate Gender with a mere glance.
These people are liars.
Niklas Luhmann was undoubtedly a brilliant guy. He wrote a lot of stuff. He published a lot of things. He was also born in 1924 and for most of his productive life the most cunning technology he had for organizing thoughts was a card catalog.
The man loved 3 x 5 cards. Who doesn't?
So he came up with an idea to organize the chaotic cascade of ideas that he swam around in by using 3 x 5 cards which were linked to each other by a naming notation scheme.
And it worked really well for him. Really well.
No matter how hard you try, no matter how obsessed you are with going through the motions of reproducing Luhmann's zettelkasten method, whether it be getting your own 3 x 5 cards and slip boxes to put them in or just beating on some sort of digital representation until it can use the same two-dimensional hierarchical reference mechanism that a man obsessed with 3 x 5 cards really enjoyed – you'll still be stuck being you.
There are a lot of people who treat the zettelkasten method like a religion and they are cargo cultists.
Don't be a cargo cultist.
That's not to say that I completely discard the ideas of the zettelkasten. I love it. I just use it in an appropriate and intelligent manner. It's a set of guidelines – not a straitjacket. It is limited by the available technology and imagination of a man who is quite dead. (And despite the fact that he died in 1998, he never really warmed up to digital storage mechanisms by the end.)
Things that are useful about zettelkasten include:
The idea of atomicity. The idea that your notes/nodes should be relatively small chunks. They should be easily addressable as a manageable bite to keep in your head all at once. When you reference something, you should be able to reference things at a relatively reasonable chunk size.
Unique naming. You should be able to address your intended link to a unique chunk without having to write too much. Luhmann did this with a relatively clumsy tree index naming scheme. We can generate UID's from the current moment's timestamp without thinking about it.
Notes have relationships. When you refer to the UID of a given note, you establish a meaningful relationship between those two. You're saying "these things are related" and, theoretically, you should be following that intention up by saying how and why they are related.
There are really great things in the zettelkasten method. Don't get tied up in sticking your arms out to the side and making motor noises while staring up at the sky, expecting crates to be falling down full of goodies.
Take it as a guide.
On note-taking my way
What I've found has helped a lot when doing notes on a longer document which has clearly defined sections is to create a note for just the title of the book or article. Then create a note named "name of book, chapter 1" which itself links to the book note. If each chapter has subsections, create a note for "name of book, chapter 1, section 1" and make sure that it links to the chapter note.
Put any metadata about the book itself in the book node, but literally do nothing else.
Now, start reading. When you see something that sparks an idea in your mind, create a heading of the appropriate level in the appropriate lowest level note in that branch, take the smallest quote that provides context, and then write your ideas immediately after the quote.
Rinse and repeat.
Eventually you will have a whole lot of leaf nodes with your ideas dangling off of them with headers in the appropriate locality.
When you're feeling spicy, go back through each of the next level up nodes (chapters if you've been doing notes in sections, etc.) and look at all the back links. Go through them and see if any of the things that live in separate files/separate nodes hook up with each other. If so, create a heading section in that chapter file, maybe even transclude the original two ideas from the lower level files (which is easy, because they have headers that you can link to), and then write down your ideas about how these things are related.
When you're feeling extra spicy, do it at the next level up.
Eventually you will get to the top of the tree and you will be synthesizing ideas from all over the text.
Meta-reasoning about how to link ideas between texts is left as an exercise for the reader.
A variant of this is more focused on atomic notes (Editor's Note: Ironically how I do things myself more often since this was originally written) – but all it really does is move your individual notes on subsections into their own notes/nodes with file names which would've been the header names if you would put them in the subsection files themselves. This looks cooler on the graph but doesn't actually improve locality so much. Still, looking cool on the graph is not entirely useless.
When using this method, your ideas on any particular section of the hierarchy should be individual files which make reference to that level file. So when you're ready to start synthesizing ideas about sections in a chapter, your ideas about things you've said about those sections will link to both the chapter file and those individual ideas from the section.
I do this extremely rarely. In fact, I find it kind of boring to do. (Editor's note: Less true now. Once you get into it, it's like vivisecting a child. Fun!) But I, personally, don't have much use for full on literature reviews. That's not the sort of thing that I need to keep notes on in my day-to-day. My notes are more focused on projects that I'm working on right now in a bottom-up way rather than top-down/bottom up ingestion of stuff that other people's written.
Still, a nice technique to keep in your back pocket.
Conversation with mifga on PKM processes and Obsidian
mifga Today at 6:57 AM
Hello SquidLord! I've been thinking about your suggestions for using Obsidian for interviews in the #academia channel and feeling really inspired. That vision of how to work to keep generating, and then folding back into a form as you start to see relationships sounds amazing. I am struggling with getting the right commands/plugins/tools within Obsidian to create that experience, though. Do you have suggestions for how to "do" the action of generating this stuff, while pinning in the right connections that you can see emergent structure? That's maybe a tall order, but maybe if you might just share a thought or two about tags vs double square bracket links (I think the latter is better for emergent structure?) and how you choose what to link to vs transclude? how much time you live in preview vs edit mode with notes like these? What you described is exactly the kind of experience I would love to have, I just haven't QUITE sorted out the mechanics of having the flow state AND structure!
for example, i'm obsessed with your suggestion to use the refactor note plugin, but am missing one or two steps to make this useful: http://discordapp.com/channels/686053708261228577/722584061087842365/815622068757463070
SquidLord Today at 11:48 AM
See – here's the thing. It's not about getting the right commands or plug-ins or tools within Obsidian to do that experience. That experience doesn't come from the tool. It comes from basic interaction with your thoughts using the most basic tools available within the environment itself.
I have been very adamant that I much prefer wikilinks to tags for almost every purpose. Because tags have no context that they are capable of building to describe themselves, the only time I use them as if the tag refers to the format/processing requirements of an entire node.
Let me give you an example.
[[Blog]] vs. #blog.
If I'm writing about blogging or talking about someone's blog or making notes from somewhere else that have the topic of the concept of "blog," there is no question – I use the wikilink. That is a concept, it's something that I'm talking about, it's an idea that I might want to come back to later. It is a referential concept within the text.
When I write a blog post and I want that node to be treated specially, either colored in a specific way or processed by an external tool in a specific way, or in some way need to refer to the nature of that note in a way that will never need to be discussed – then I use #blog. That node IS a blog entry.
As a result, I use very few tags compared to most people active in the PKM space.
The key differentiator is that a tag carries no context. There is no note, no node in the graph, where I get to explicate what a tag is. It only describes in a top-down manner. Sometimes that's what you need – but it's not very often. Tags can't really have a participation in the mesh graph as a whole except as one way in references. And that's a one-way in reference that specifically applies to the whole node. If you want to different sections of a text to have two different tags – well you can't. They need to be separate nodes.
(That's often a good indicator that you need to be using Wikileaks instead of tags, anyway. Things start getting messy from that point.)
There is no real secret. There is just, "hey, I'm going to have to do some work – so I better get ready to do some work."
Periodic content reviews if you're not already doing them. Actually working with your notes, making reference to them, taking a moment to add to them if something new comes up or if you have a new thought. All of this is part of the process.
Because it's not just about notes. It's not just about the files on the disk. It's about the process you went through to get them there. They are an artifactual excreta of the process of thinking. They might be useful in and of themselves – but they aren't the point. The thinking that makes them happen is the point.
Start with asking yourself, "what am I doing this for?" Move on from there and you get better results.
Going back to the re-factoring plug-in, it's very simple. Don't over complicate it. All it lets you do is highlight a section of a note that already exists, hit a button, and have a link created in the node you're standing in which replaces the text as the system moves the text to a new node. Thus you can split up ideas fairly straightforwardly after they've been written if you decide, "hey, this needs to be promoted to its own note" for whatever reason.
Usually this happens because you've written it as part of a larger idea within the section of the header but you realize maybe it deserves a little more first class object nature. Maybe you want it to be treated specially as a particular kind of content. Maybe you just want to see it separated out as a node.
That kind of thinking is important.
mifga Today at 12:50 PM
This is really great. And from now on i shall refer to notes as “artifactual excreta of the process of thinking” — i love that. Is that original or a quote? Sounds luhmannish?
SquidLord Today at 12:52 PM
That's just me being pretentious; sorry it's not cooler. :stuck_out_tongue:
mifga Today at 12:53 PM
This repositioned my thinking a bit here. I was already heading in the wikilink style directions. I think i jumped on obsidian originally without much thought beyond “it’s like tomboy notes!” (Which was a small understanding of what i would find.)
SquidLord Today at 12:55 PM
It's probably better to bring the idea that "it's a wiki" to the table. We work with wikis every day. Some of us more than others.
The problem comes when you try to imagine that it is more than that. Not to say that it can't be extremely useful – but don't imagine it's something that it's not. Instead, understand that the value comes from your process, and there are a thousand ways you can run the physical manifestation side of your process.
There is no good replacement for hard work and brain sweat. There are too many people looking for it.
mifga Today at 12:55 PM
I have been using obsidian just the way i had been using Sublime, which was plaintext driven, but each note in a folder, and within each note lots of labeled and dated entries worth of information. Then i noticed the zettelkasten stuff and How to Take Smart Notes and realized i wasn’t building knowledge, i was making little piles and losing information in piles
So i’m trying to start thinking in terms of atomizing everything to single ideas and start to do this process of considering and building connections through notes. But trying to address my instincts to make groupings of things in a single file via transclusion, and start looking at those notes and seeing if i’m actually catching an actual thought, and start adding wikilinks. But i’m still fighting less productive modes.
SquidLord Today at 12:58 PM
Right, there is a difference between collecting piles of stuff and taking stuff you know you're going to want to use later.
For some people, I advise a radical approach to trying to get a handle on "making sense of things." Don't copy anything. Don't put anything in your collection that you didn't write. At most, if you want to get something in, rewrite it in your own words, a phrase at a time.
Not everybody needs that radical an approach, but some people just can't get out of the rat's nest. They just keep collecting little bits of things from other people with the assumption that the more of other people's things they have, the better a person they are by mystical transference. It's like the people who collect first editions of books but would never think of reading one.
It's a hard lesson for some people.
But yes – stop trying to pull everything together at the top. That's the one thing that I really hate most about MOC-theory – the idea that everything needs to be brought together in a singular point which defines an architectural hierarchy and thus once you've achieved that you have true understanding.
That's bullshit of the purest race sublime.
Instead of trying to make groupings of things in a single file via transclusion – write your own ideas that explain why these other two things (or multiple things) are related, and then let that stand. Rinse. Repeat. As you go, you will build up nodes which link outwards to multiple things because your ideas are complex, and then you'll start having ideas about your ideas – which is exactly what you want to happen.
Sometimes you will end up with files which contain a lot of links to other things – but that's fine. The connective tissue between those things will be your own ideas.
mifga Today at 1:01 PM
I think with a few projects i can immediately do the hard approach and should — but some of the urgent composition challenges are harder. But worth starting to find ways to pull out details as i go because these urgent projects probably have the most “waste” of effort (ie i just repeat it again when i urgently approach the topic again). And i’d like to start building on these ideas to go further. But i need to probably move slow and methodical a bit to get used to things.
SquidLord Today at 1:03 PM
For myself, I like to move fast and break things. Mainly because I know that as long as I don't delete content in Obsidian, I can build out from it with new, updated ideas and I've lost nothing. Even if I was wrong.
Even if I was wrong about structure or conductivity, or anything – it's all changeable. That's okay. Things should evolve. I should get better. And if there are bits in there that are still old or not updated or haven't been touched because they haven't been important – that's okay too. If they come up, I'll fix them. If they don't, it's okay.
mifga Today at 1:09 PM
Okay, here is a practical question as i sit down to do this work RIGHT NOW! :slight_smile: i am writing a quick presentation about 3d printing and architecture, mostly applied so drawing on my thoughts and a stack of example projects. So this is actually good stuff to enter in because i’d like to recreate this project with more citations and references to history of model making etc. But my main point is that i have a narrow scope of needs for this information, and basically just need the act of flowing through and generating content, tuning the outline as i go. But, most of these things i’m saying are actually talking about various aspects of 3d printing as a technology and as a process step. So i want to grab some wikilinks out now, but likely after the presentation, i want to return and rethink these things. Is it a good approach to just add a few wikilinks now, knowing that the presentation scope is constrained and the time limit is critical, and then return and improve each of these notes to make them more useful free of context of this use?
For example, maybe right now each slide is a note, but as soon as i can spend more time, then rip out ideas from these slides, drop the slides, and have the better marked up note in my system?
SquidLord Today at 1:18 PM
Sure. Just get it into the system now, make sure that you have it broken up properly by section with headers so that you can refer to those headers later. Then set sometime on your schedule to settle in with it and really start breaking it out and may be thinking about having thoughts about what you did.
I'd say keep the slides as individual notes, but have some thoughts about your slides which themselves are individual notes but make reference to the slide link. Now you're starting to build a mesh of ideas that interrelate.
Stop thinking about hierarchy, stop thinking about "a better marked up note," and start thinking about the idea that it exists to capture your ideas at a moment in time which you hope to have more thoughts about.
Or, more prosaically, think about your whole Vault as sending notes to a future you. What would future you care about when it comes to this stuff? Are they really going to care whether section 1.1.1.2 is properly linked to Figure 3? Or are they going to care about you telling them, "[[This Doc#section 1.1.1.2]] ended up being elementally wrong thanks to research from [[This Other Guy]] in [[This Other Document]], although [[This Doc#Figure 3]] still holds true because…"
mifga Today at 1:47 PM
I really appreciate your time here, this has been hugely helpful. I really think this is going to work for me.
oh, one slightly lame note. i'm curious if you use frontmatter (for aliases, etc.) and if you use templates, or just drop the wikilinks inline or at the base of the note.
SquidLord Today at 1:49 PM
And if it doesn't – you can try something else.
mifga Today at 1:49 PM
that is probably me worrying about the wrong stuff,
thanks again -- both the concept of just thinking of this as a wiki and as notes to a future self were both really helpful.
SquidLord Today at 1:51 PM
I use a fair number of templates, mainly because I'm too lazy to type the same things over and over. Aliases get pretty good play because I really don't like to have two separate entries for plurals and because I often have to refer to a number of things that are acronyms, so it makes it easier to find all examples and use them in an appropriate way.
I also, in a method which no one else on earth seems to use, leverage the zettelkasten style no creation to specifically write my personal ideas and thoughts (and keep them in their own directory) while using un-zettel-prefixed notes in general to contain more general ideas. Really just as a convenience to make it blatantly obvious to me in the process of doing things when I am writing my own thoughts and when I'm just creating a node which specifically talks about something that aren't my own thoughts. (As you can imagine, there is some bleed – but all systems that involve human thought are fuzzy.)
mifga Today at 1:53 PM
heh, I'd like to see where my process evolves for these kinds of elements over time. right now I'm picturing me doing things one way -- but I'm not yet doing those things.
SquidLord Today at 1:54 PM
There's only one way to find out what works – do things and fail. If you're not failing, you're not trying to do enough stuff. That's not to say that you should be failing all the time; eventually you should find stuff that works for you most of the time. But if you're not failing sometimes, forever, you're not really learning anything new. Nothing is changing for you. What's the point of that?
mifga Today at 1:55 PM
Thanks! i'm off to the races....
SquidLord Today at 1:55 PM
Go break some stuff.
more conversation with mifga about old projects and their usefulness
mifga Today at 2:30 PM
I generate so much writing, so many presentations and transcripts, and so many of these results languish in folders, and are never seen again by anything beyond the filesystem.
SquidLord Today at 2:33 PM
Ayup. But here's the thing – how many of them really need to be seen again, ever? For the most part, your projects are one and done things. It's not the projects themselves that are particularly useful. It's the things you accumulated along the way in terms of mental architecture. It's the ideas you had that led to that presentation coming together, the idea that you just couldn't get to fit in but don't necessarily want to forget ever happened. The chunk of research that ended up being two slides but took you two weeks to actually put together to the point you understood it well enough to be two slides.
Those are the sorts of things that are important enough to send notes to yourself in the future about. The presentation or transcript itself? Maybe you want to refer to having done it, but that's about it.
mifga Today at 2:33 PM
Hmmm, good point.
Writing is not necessarily a high yield knowledge building process. And it might be overly fussy to spend time trying to turn all of the elements into useful stuff.
SquidLord Today at 2:35 PM
Honestly, most of your old projects are really only important insofar as you have new ideas about them. Which makes them useful for keeping as a first-class node that you can link an idea to, but that's about it. If you suddenly realize there's something in there that you want to break out so that you can think about it separately – that's easy to do. You go in, extract the block or make a brand-new node and attach it, and then start writing new nodes that attach to it. It's simple, it's easy, it's not complex. It's just a matter of doing.
Epilogue
Hopefully some of this has been interesting and maybe even useful to you.
If you've been bored to tears, if you have been struggling word after word, suffering as you plotted on through the mud – congratulations, you've made it to the end. I don't know what I would do without you. Also, you might be a masochist.
If you enjoyed yourself, if you have learned something, if you had an idea that you never stumbled over before and it cast light on a dark corner – congratulations, you've made it to the end. Let me know down below, wherever you read this, how it helped you out.
If you hate every word that comes out of my mouth and you think that this text is an abomination in the sight of God and man – congratulations, you're right. Tell me where I got things wrong, guide me to the path of righteousness, and preach to me of the path I should walk. I may not do it but I will take a certain amount of gratification in your efforts to do so.
If you’ve enjoyed what you’ve seen here today and want to support my continued efforts to bring engineering, art, and the occasional philosophical divergence to the masses, please feel free to send me a tip. Or thousands of dollars, I’m really not that picky. It’s through your efforts that content like this gets created.
Check out my portfolio if you'd like to get in touch or see more of my work.
- PAYPAL: http://paypal.me/SquidLord
- CASH APP: https://cash.app/$SquidInks
- BITCOIN: bc1qy9wkj3g6ueugu0tjuwmh6dfxxn590r3p9tj9qt
- ETHEREUM: 0xcca4C3A3024d7496bDe06f6Fd865BfD5992Bf10B
- BAT: 0xcca4C3A3024d7496bDe06f6Fd865BfD5992Bf10B
- STELLAR: GBI6AMTGCBUPRBKWY3V246FDBZJILGOZHNOYWJXZNRBLACB6ZWAQTIYW
- ZCASH: t1dhcviaYgQuX4z91jhkjHrmkWNf5agJCNc
- MONERO: 48JHK6G4FyvSktU45t5ra7SALwFJAMEebEShPmk1K7xd95p4mYNJMnRZ27JBmcc97VhrJqEDmCyHtVP8iDTeNAJE1VnaCN8
- PO.ET: 0xcca4C3A3024d7496bDe06f6Fd865BfD5992Bf10B
- TRON: THKkzVBLz4V14dmvpDxkag8piKuJ74aATa
- XRP: rHvFZbRtsW2975zQXYTHDz7VAXPSdozoEs
- LBRY: https://open.lbry.com/@LexTenebris:9
- STEEM: @lextenebris
- HIVE: @lextenebris
(Yeah, I don't care, I'll take any money you've got.)
As a special treat for those who made it this far, let me show you what the local graph of this particular article looks like in my personal Obsidian set up. Red nodes are blog posts, green nodes are zettelkasten-nodes which are explicitly my own thoughts, dark grey are nodes which haven't been fully created yet, and white nodes are places where I link to ideas which I didn't have, whether that be an article, or a concept, or whatnot.
And that's just with a neighbor range of three.
Good luck, Godspeed, don't take any wooden hentai.