Learning Add-On: the Zettelkasten

In our last post, we covered the big-picture view of “why learn how to learn?” and a bit on principles you can follow to do so. We also introduced a basic note-taking system, making use of “Bib Cards. I had originally intended to give a quick-start guide to the full analog Zettelkasten approach in that post, but it was a tale that grew in the telling, and for reasons of space, I’ve pulled it out and reworked it a bit into this post. As mentioned in that post, if you want to go all the way with the Zettelkasten approach, I recommend Scott Scheper’s Antinet Zettelkasten.

Basic Starter Guide to the Zettelkasten

As I said above, you can go very deep on the Zettelkasten, and it’s a very flexible tool, but it’s easy to get caught up in the details and start spending too much time on managing your notes and not enough on whatever you’re taking the notes to be able to do (generate insights, come up with writing, or whatever). So, I’ve purposely presented a barebones version below that I think hits all of the most necessary components. I’ve also designed this so that if you only want to do step one, you can stop there. If you want a little more, you can add step two, and so on, but it’s not quite mix-and-match. To work right, you’ll have to do all prior steps before introducing a later one. Think of it as “crawl, walk, run,” rather than a menu of interchangeable options.

1. Step One: Bib Cards

So, last time, we introduced the idea of “Bib Cards.” As a quick recap, these are cards that have basic information about the book, article, or other source on the front of the card, and your notes taken while reading on the back. These get collected into your “Bib Box,” which becomes a collection of notes, organized by author and source, which as we said, is already far more useful and organized than the kind of notes scattered among random notebooks we usually default to.

As a reminder, the only required steps for a Bib Card are to include basic bibliographical information of the book or article you’re going to read, like this, and then notes on the back. Here’s an example of the bibliographic information I wrote down for a book I read recently: Friedman, David, Leeson, Peter, Skarbek, David. (2019). Legal Systems Very Different from Ours. Independent. As I said, that’s all you truly need on the front of the card, so if you like, call it good there.

Further Options for the Front of the Bib Card

On the other hand, there are a few other pieces of information you might find it helpful to include, that don’t add that much more work to the process. First, you might Dates Read right under the basic bibliographic information. I find this helpful when referring back to notes taken some time ago to get an idea when I read it, what kind of stuff was on my mind, what I was interested in, how my views have changed, and so on. I usually put something like this: “1) 2025/9/25 – 2025/10/13.” This shows that it was the first time I read the book (the “1)”) and how long it took me (sometimes also useful to know – was this book a long, slow slog, or something I tore through at record pace?).

The next bit that might be helpful is to list your Goal(s). I list one per time read, so continuing my example, I might put something like: “1) Get ideas on alternative ways to understand legal issues.” Again, the number indicates this was my goal the first time reading through, and if I ever come back to it and read it again, I might have a different goal. This is a good place to indicate if you’re merely skimming the book or reading it deeply, which might help you decide later whether it needs a revisit. The point of writing a goal for each time you read is to help you focus on taking your notes and to cut down on “note bloat.” Most books worth reading are chock full of interesting ideas, threads you could follow, points you might contest, and so forth. If you try to take full notes on all of that with every book you read, you’ll get a lot out of those books, but you will read far fewer books – taking notes takes time! Having a clear goal helps you to focus – is that interesting insight serving my goals at this time? Or something to put on the back-burner?

Lastly, for books that I want to engage with fairly deeply, I’ll put an Outline on the front of the card (which sometimes spills over on to the back). Usually, this is just a copy of the table of contents, but sometimes I’ll go deeper (like if the chapters have sections not listed in the table of contents). Copying this out by hand is time-consuming, and so not much worth it for works you’re merely skimming or trying to get the gist of, but if you do want to understand it in more depth, this can be a good first step to understanding the structure of the book. In this instance, the slowing down it forces is a feature and not a bug – if you just read over the ToC, chances are good it won’t much sink in. But if you spend 20 minutes copying it all down, you’ll have time to wonder things like “I wonder what that chapter title means?” or “why are these chapters grouped together into a section of the book?” For this one, I won’t subject either of us to an example, as that would be a whole paragraph of its own.

The Back of the Bib Card: Your Actual Notes

Again, as a reminder, on the back of the card, you’ll want a shortened version of the citation in the top left of the card while it’s still in landscape format, like this: Friedman, D., Leeson, P., Skarbek, D. (2019). Then draw a little box around it. To take your notes, rotate the card 90 degrees to portrait orientation, so that the citation you just boxed in is on the top right side of the card, going sideways, and then starting at the new “top” of the card, perpendicular to the citation. As mentioned last time, rotating the card isn’t strictly necessary, but most notes will be brief enough that you can fit more in with a narrower horizontal and taller vertical.

Whenever you want to write a note on what you’re reading, write the page number (or location reference, for a digital book) in parentheses followed by your thought, like this: “(1176.9) Marketable Torts.” To avoid the note bloat we mentioned above, Scheper recommends that for most books, you should read the whole book without writing anything down, and then write down the one “irresistible idea” that stuck with you from the book. Besides just helping you read faster, the other reason to have this as a rule of thumb is that we have a natural tendency to over-value the new and recent, to assume that it will prove to be more useful than it really turns out to be down the line, and this habit counteracts that somewhat. I have to admit, I struggle with this one, but keeping it in mind as an ideal has certainly cut down on the number of notes I tend to take, which has helped my reading speed back up to closer to what it was before I started all this note-taking craziness.

That said, sometimes, you really do want more notes: maybe you know that this book will be core to a major project you’re working on, or foundational for reading a lot of other books, or it just has more than one thought that grabs you hard enough to think “I better write that down.” In this case, there’s nothing stopping you from setting the book down for a second and writing a note as you go in the same way you did above.

If you end up taking a lot of notes, you might run out of space on your original Bib Card, but luckily, this is another place where the flexibility of the system comes in handy. Once the back of your card is full, just pull out a fresh card, flip it over, write the citation on the top of it, the same way as before, but this time, before boxing it in, add “a,” like this: Friedman, D., Leeson, P., Skarbek, D. (2019)a. There’s no need to put the full bibliographic information on the other side of each subsequent card, as the citation ought to be enough to keep it with its “front” Bib Card whenever its filed. If you really go nuts with the note-taking and need even more cards, repeat the process, but move through the alphabet with each new card: “b, c, d, etc.” You likely won’t need to go past “z” for anything but a book you absolutely must suck every last drop of value out of, but if you get there (I’ve done it once or twice), just go to “aa, ab, ac” and so on.

However many notes you take, once you’re through, you’ll want to put the card(s) into your Bib Box. Get a box suitable for notecards (a shoebox works fine, or whatever you can find at an Office Supply Store) and some alphabetical dividers with tabs. Find the tab that corresponds to your first author’s last name and put the card(s) behind it. If your notes ran to multiple cards, just put the whole stack in place, in order. If any other cards are already in that section, just file them alphabetically by author last name, and if you already have cards by the same author, go to alphabetical by title. Now, if you’re ever like “what did I write down about that book on legal systems by Friedman?” you can flip through your bib box, find your note card, and look over what you wrote. Likewise, if you ever decide to revisit the book, go back and find it, pull it out, and add another round of notes to the back.

Just by doing this, you’ll have a well-organized collection of the notes you’ve taken on the things you read, and you’ll have gotten them into your mind more effectively than if you had made highlights or typed up notes on your Kindle. I’d recommend doing just step one for a book or three at least, or until it starts to feel easy and natural, before considering adding anything else. But if and when you think more might be useful, you can move on to Step Two, the Index.

2. Step Two: Index

Now, you may have noticed a weakness in the above: what if I remember the book’s name, but not the author? Or an idea I’ve read about in multiple books, but I can’t remember their titles, much less authors? As it stands after Step One, our only route back to our notes is the author’s last name (or flipping through the bib box until something rings a bell, which is not a bad thing in itself, as you might stumble on something else useful). Sometimes we might want to find our way to our notes by a different point of entry, and that’s where the Index comes in.

To create your Index, get another box suitable for 4×6 cards, again with alphabetical divider tabs. This time, though, you’re going to pre-populate the box a bit. Behind each letter, put a regular 4×6 card with that letter on the title row followed by (1), like this: “A (1), B(1), C(1), etc.” Now, go through your Bib Box and take each title and add it to the card that matches the first letter of the title, with the citation we used on the back of the card written next to it, like this: on card “L(1),” I’d write on the first line “Legal Systems Very Different from Ours – Friedman, D., Leeson, P., Skarbek, D. (2019).” If I also had another “L” book, I’d add that: “The Long Goodbye – Chandler, R. (1953).” If you’ve already built up a bit of a Bib Box, this might be a bit of a slot to get started, and you don’t necessarily have to try to get every title into the Index now. You might do it “as needed.” Next time you try to find a work and have to flip through the Bib Box to find it, go ahead and add its title to the Index. Then, every time you file a new Bib Card, go ahead and add the title to the Index. Eventually, anything you ever actually make use of will be in there, without giving yourself an evening or a weekend of “Homework” to get through. Now, we can find a particular book’s notes either by remembering its author or its title, which is great, but what about if we want to find it some other way?

Well, you can put more than titles on those cards. I could add to my “L(1)” card “Law” and put a reference to Friedman’s book, but that’s likely not so helpful, since both the title and concept are on the same card. But maybe I want to be able to find it from “Political Science,” so I go over to the “P(1)” card and write “Political Science – Friedman, D., Leeson, P., Skarbek, D. (2019).” Now, I can find it from another, related concept, even if I forgot to go look under Law/Legal specifically. On the other hand, if Political Science (or Law) are topics I’m likely to read multiple works on, it would be awkward to leave blank space on the “P(1)” card, and however much I leave, it might run out. Luckily, notecards are flexible and there’s an easy solution. Pull out a new card, at the top write “Political Science (1),” on it write “Legal Systems Very Different from Ours – Friedman, D., Leeson, P., Skarbek, D. (2019)” and any other books you might want to find on the subject, and then drop that in behind “P(1).” As you continue to add concept cards like this, just file them alphabetically, and expand as necessary (“Political Science (2), Political Science (3),” and so forth). Oh, and one pro tip: if you started a concept on a letter card (e.g. you wrote “Political Science” on “P(1),” but then created a “Political Science (1)” card), cross out the entry on the letter card to indicate you should look for its own concept card.

A quick word: the letter cards (“L(1), P(1),” or whatever) might not seem very helpful for anything other than looking up titles at this step, and you might be tempted to skip straight to concept cards, but they become more useful once we get to Step 3 and start building out the Main Box.

This gives us three different ways to find our way back to notes we’ve taken, and one way to start identifying connections between books that might have bearing on each other. Maybe Legal Systems Very Different From Ours is listed on my “Law” index card, my “Political Science” index card, and on my “Iceland” index card (most of us likely wouldn’t have an Iceland card, but I might, because I have an especial interest in its history and culture).

A quick word of caution here. Many of us who have had experience with digital platforms that allow for tagging might see the concept index cards as a “tag” and might want to come up with every possible tag for a book (so, in addition to an “Iceland(1)” card, I might add “Somalia(1),” and “Gypsy(1),” and “China(1),” and so forth, trying to refer to it from all of the legal systems described in the book!). I don’t recommend this. Most of the time it’s make-work. Likewise, don’t try too much to “anticipate” the need for concept cards, creating blank cards for every concept you might eventually need. Just make them as needed, which might be any time you realized you wanted to find two or more books for the same concept. Also, don’t get too worked up about the organization on your plain letter cards for the index (“L(1)” and the like) – they won’t be in alphabetical order on the card, you’ll end up crossing stuff off when you move it to a concept card, or whatever. Again, index cards are flexible, and the goal here is not to have a perfect system that anyone could use, it’s to have a tool that helps you find your notes when you need them.

As with Step One, you can stop here and already be way better organized than just about everyone, and also as with Step One, if you give this a try, I recommend using it for a while until it becomes comfortable and automatic to add something to your Index whenever you add a card to your Bib Box. It shouldn’t take more than a few minutes per book, so if you can do it whenever you finish a book, it’s nice and smooth. Letting 3 (or 4 or 5 or 10) books pile up before you add them, though, will create friction and make it harder to keep up with it.

3. Step Three: Main Notes

Okay, now we get to what most folks tend to think of as the “point” of the Zettelkasten, but I’ve tried to emphasize all along that the previous two steps are useful, and for many people will do everything they need. This step becomes more useful if you are looking to really take the stuff you’re reading and thinking about and find new connections and insights between it, especially if you plan to write books, newsletters, or articles of your own. What we’re talking about here are “Main Notes,” which are notecards containing your own, more worked-out thoughts, usually in response to something you read.

First, a Word of Caution on What the Zettelkasten is Not

The Zettelkasten is not meant to be an archive of every thought you have had or every thread you might someday want to explore, nor is it meant as a universal encyclopedia or comprehensive personal reference. If you’ve spent time with digital knowledge management tools like Evernote, or Devonthink, or whatever, you’ve likely picked up some habits of saving websites, capturing highlights, jotting down every stray thought where it can be readily searched. We want to set those habits aside for the Zettelkasten, because if you try to include them, it will break you.

Seriously: remember how we talked about it being a feature and not a bug that this approach will take you longer and make you think harder? Well, that remains true so long as you’re working with the things you truly do want to explore in the most depth. Less so when you start using lots of time on marginal stuff that might come in handy some day or you feel like you really should have it for the sake of “completeness” or something. So, don’t put every recipe you encounter in your Zettelkasten (unless you’re a chef, maybe!), don’t put random “fun facts” in there, or feel like you have to flesh out every sub-topic of some field just because you went deep on one of them. As with our first two steps, the goal here is to help you learn, to only add notes as and when they’re needed, and to avoid like the plague “makework” that slows us down and distracts from the learning we’re actually trying to do.

The Physical Set-Up for Your Main Box

As before, you’ll want a box and some divider tabs, but this time, you likely want those divider tabs to be blank or easy to whiteout/overwrite, as we’re not going to be using just the alphabet. In this case, the divider tabs are not strictly necessary, but the more your Main Box grows, the more useful they’ll be. Unlike with the Index, you should likely not do any pre-populating here, though more on that below when we talk about organization.

What to Put on Your Main Notes: Your Own Thinking

Let me explain the core idea of Main Notes before we get into how to organize them. Your “Bib Notes” on the “Bib Cards” are meant to be pretty terse – more pointing out what’s interesting, or what it reminds you of, rather than explaining it or analyzing it. Main Notes are where you can get into doing your own reflection, thinking, analysis, and so forth, but at a minimum, are where you restate some thought from a book in your own words. They don’t need to have publishing-worthy prose or air-tight syllogisms, as these notes are “work product” – they’re a tool to help you think and develop and explore different ideas, so it’s okay if they’re a bit messy. Again, this is not a reference, and it’s not for anyone else.

A word on including quotes on Main Cards. You can copy down full excerpts onto main cards, but they’re not primarily meant to be a reference. Think of it this way: putting a quote on a main card should be more like selecting an excerpt to quote in an article you’re writing than like clipping an article out of the newspaper to keep in your files. If the quote is a clear, succinct summary of one of the author’s main points, or it has beautiful language, or there’s some other reason to focus on the original words in full, go for it. But remember, the point of the Main Box is to develop your thinking.

Here’s an example of how coming up with Main Cards might go: in Legal Systems Very Different from Ours, after looking at several legal systems that have private enforcement of penalties, Friedman lays out the common characteristics as he sees them of what he calls “Feud Law.” Now, right now, this is the only place I’ve really encountered this, and it strikes me as a useful conceptual category to add to my thinking, so I decide to create a Main Note titled “Feud Law.” On it, I restate the characteristics Friedman has identified in my own words, add a citation of the book at the bottom of the card, and then call it good and drop it into my box (see below for where to put it). Later on, though, say I’m reading about how feuds were settled among the tribes of the steppe, and I notice some new wrinkle that seems to contradict the characteristics of Feud Law as generalized by Friedman, well then I might come back and write a new Main Card that’s titled something like “Feud Law Edge Cases?” that is more my own speculation (but probably still with the citation to what made me think of it).

Once again, we have lots of flexibility here, but the core idea is that a Main Card is one or more “thoughts” that you strongly suspect you’ll want to come back to later as you develop your thinking along some lines. Now that we’ve covered what goes on a Main Card, let’s talk about where they go in the box and how we keep track of that.

Organizing Your Main Notes with Addresses and Placement

So, that’s making a Main Card: title, your thoughts, and any citations that they’re building off of, if applicable. The one key piece of a Main Card missing now is its address. Each Main Card gets a unique address. Different approaches to Zettelkasten recommend different ways to come up with the address, but figuring out what address to put on any given new card is the biggest stumbling block for folks new to Zettelkasten, so let’s set that aside for one sec and talk a bit about what an address is doing for us and why it’s helpful, along with what it’s not doing, so that we can approach the “how to number my cards?” question with a little less wailing and gnashing of teeth.

Finding Your Way from One Idea to Another – Linking and Branching

In most writing, like this blog post, or a book, or whatever else, things are presented sequentially: you start at the beginning and proceed to the end. Some forms of writing will make it easier to jump in and out, to read only one section but not another, and so forth (Christopher Alexander’s books are nice examples of ones designed to be engaged with in multiple different ways, depending on what the reader needs at the time), but the default assumption is that things will be presented in such a way that if you follow along, in order, you’ll be given earlier whatever you need to understand what you get later.

This works really well. There’s a reason pretty much every piece of prose follows this convention, and always has (again, there are some exceptions, of course). But it works really well for finished writing that is meant to share things with others, as opposed to in-process writing that is meant to develop your own thinking. For this kind of “working it out” writing, we might not yet know what ought to come earlier rather than later, which things have more weight or less.

And so, it becomes helpful to be able to identify chunks of that thinking/writing and refer back to it. That’s what we’re doing when we cite references from works we’ve read. That’s what we’re doing when we highlight something. That’s what we do when we put some thoughts in the green notebook instead of the black one, or on a new page of its own rather than on the last page with those other thoughts.

Zettelkasten addresses are an attempt to be a little more systematic about this. We identify a chunk of thinking (in this case, everything on a Main Card) and we give it an address. From this point on, if I want to refer back to that specific thinking (or just find it again), I can use that address. Let’s take my “Feud Law” Main Card from earlier and say that I give it the address 2521. Now, in my Index, I can go to “F(1)” and write “Feud Law – 2521” (unless in my head it makes more sense to put it under “L(1)” for “Law, Feud – 2521”). And maybe I add it to the previously established “Law” concept card in the same way. Hooray, we’ve filed a Main Card!

Now, those in the digital Zettelkasten space would say that with addresses and linking, you have everything you need, especially if you have a program to manage stuff for you. You just give every note an entirely arbitrary address (like the timestamp of when it was created), then make sure to link to that address from anywhere else you might want to find it (maybe you create one or more Index or Concept nodes, which theoretically should be just as easy to keep track of as Index Cards). The trouble here is that as conceptually simple as that is, even with the aid of a program to search and track things and automate them, it very quickly becomes hard to keep track of which isolated nodes need to have a link added to them when some new node is added. You might have dozens, or even hundreds of notes that are more-or-less related to the new note, but which ones deserve a link? How do you find all of them if they have arbitrary addresses and not-obviously-related titels? How do you indicate that they’re kinda related, but not directly? All of which gets us to the other organizational concept of the analog Zettelkasten, which is branching.

One way to visualize how different pieces of knowledge are related is to imagine a tree. Your core, baseline understanding of what the world is and how it works, the ways that you understand what is true or false, worthwhile or not, that’s the trunk of your tree of knowledge. Coming off that are various large boles: humanities, sciences, arts, and so forth. But each of these have their own branches, which each have their own smaller and smaller branches, and so on.

In the same way, as you add ideas to your Zettelkasten, some will have no connection to any other card than that they’re both interesting and worthwhile enough to you to make it into your Main Box (main boles). Others wil be “next door” kind of ideas – things that clearly have something to do with each other, but neither is a direct development or expansion of the other (branches on the same bole). But others will obviously be a follow-on or going-deeper-into of another idea (twigs coming off a branch), and it’s here that the use of cards over other analog means really shines, because cards give us a really easily updatable way of spatially organizing things. In a notebook, once I write one thing on page 1 and another on page 2, I can’t go back and write something else in between them. Oh sure, I can go to page 3 and write in big letters at the top “GOES BETWEEN 1 and 2,” which is somewhat like what folks used to do with commonplace books (though with much more sophistication and variety in technique), but how much more satisfying if I can just insert a new page between them, if it feels necessary? Well, with cards you can, but you are still doing it physically (unlike the admittedly even more flexible digital options).

To capture this branching, we have two rules, one of which follows from the other. The first, more foundational rule is “A card should be filed behind the existing card it is most relevant to.” This is simply a way to make sure that cards in closer physical proximity (tend) to have more to do with each other conceptually, which makes it easier to find related ideas, whether filing a new card, browsing through, or checking to see if you’ve already written something on this kind of thing before.

The second rule is “A card’s address should somehow reflect it’s branching.” You can see how this follows from the above: if we file by address, and if we want cards to be behind the already-existing card that the new one is most relevant to, then the address needs to match that placement. There are various ways you can achieve this. For example, you might have card 1 and then card 2, but your new card is more relevant to 1 than 2, so it becomes 1a. Then, later, you write a card that’s most relevant to 1a, so you might number it 1a1. Eventually you might have something like “1, 1a, 1a1, 2, 2a, 3, 4, 4a, 4a1, 4a1a” and so forth. Again, don’t get too hung up on how to number just yet, we’ll talk about that in a second, and I’ll provide two alternatives that both work just fine and are hopefully straightforward enough to get going on.

Okay, so the point behind all of this is that the combination of your address and the physical placement of cards in the box helps you to understand which ideas are related to each other. On a practical level, this can also save you some work in adding things to your index. You don’t have to add every last twig on a branch to your index if you’re reasonably certain that the index pointing you to the fatter part of the branch will get you where you need to go.

How to Assign Addresses to Your Cards

Okay, as I warned above, assigning addresses to cards is the number one stumbling block to those new to using a Zettelkasten, it’s one of the topics most likely to provoke fighting in the fora where these things are discussed, and even if you feel like you get it, it can still get confusing at times. The good news is that it doesn’t really need to be that big of a concern, and you have multiple layers of redundancy protecting you from doing it “wrong” (spoiler alert: there’s not really a wrong way, just ways that work better or worse for you).

To help get around some of the stumbling that this tends to produce, I’m going to provide two options, and give you the basics on how to follow them. I would recommend, though, once you pick one, stick with it, at least until you get a little more established with using your Zettelkasten. The address is mostly arbitrary, and as long as some structure is maintained (meaning related-ish stuff tends to be closer to each other), you’re doing just fine. It’s also worth noting that even if you do “stick with” one of these methods as I’ve recommended, you’ll find that elements of the other one will likely creep in as you do your work, and that’s just fine.

Oh, and one practical concern: where to write the address? The top of the card is most helpful, as skimming through the address is often helpful, and if it’s lower down on the card, that’s harder to do. The benefit to putting it at the right hand top corner is you can write your titles at the left before knowing how you’re going to address your card, and then have room left over for the address (usually). The benefit to putting it on the upper left is that we tend to look at the upper left of pages first, at least as English speakers, and that if your address runs long, it will all be on the top row, and if anything has to carry over to the second row, it’s the title, and as I said, sometimes being able to see the address when thumbing through cards in the box is helpful.

Option 1: Bottom Up – Organic Growth

This approach has some very vocal proponents, and the idea is that it most purely captures the organic growth of your own thinking as it is instantiated into your Zettelkasten. On the plus side, it greatly simplifies getting started, gets rid of “did I put it in the right place?” kind of questions, and might be slightly better at revealing unanticipated links between the stuff you put in your Zettelkasten. On the downside, as your Zettelkasten grows, you will be more dependent on your Index, and less able to rely on being able to just open up your box, look around, and find what you need. Also, you might find it harder to mentally “sort” the stuff in your Zettelkasten, which is not necessary, but can help with its purpose of helping your learn and think, rather than just being a place to keep your previous thinking.

So, how to do it? Easy. Your very first card gets the address 1 (you can do 001 or 0001 if you want to get used to saving space for the address on your cards as they get longer). Your next card gets 2. And so on, until you encounter a card that should go between two existing cards (remember, put every new card behind whatever existing card it has the most relevance to, and if the answer to that is “none of them,” put it in the back of the whole box), and then give that the number of the prior card followed by a letter. Now, as long as you’ve just got “1, 1a, 2, 2a, 3,” this is straightforward enough, but what about when you have something that goes between 1a and 2? In my earlier example, we literally just alternated numerals and letters, but that will tend to multiply the length of your address rather quickly, and possibly unnecessarily. So, instead, ask yourself, “is this ‘going deeper’ on 1 or 1a?” Another way to put it is “does this have just as much to do with 1 as 1a does? Or does it have more to do with 1a than with 1?” If it feels like it’s at the same “level” as 1a, go with 1b. If it feels more like it’s a deeper dive on whatever 1a is about, then make it 1a1, with the understanding that you can still make 1b down the road. If this bit is confusing, see the Top Down approach below for a discussion of “Directories” that might help.

Overall, though, you’re ready to go. Just remember to add the addresses for your Main Cards to your index as you create them. Since your “top level” (“1, 2, 3, 4, etc”) cards don’t have any particular logic to them, you might have to add Index entries further down the branch than in the Top Down approach where the address attempts to convey a bit more structure by itself.

Option 2: Top Down – Categorization

Okay, if one of the main benefits of the bottom up approach is that you can just start, one of the chief benefits of the Top Down approach is that by making a few decisions up front, you can give yourself a framework you might find helpful later. The way you do this is basically by Categorization. Now, I think that Scheper speaks against categorization in what he has to say about branching and addresses, but he still recommends starting with a categorized system, whether he calls it that or not.

Before we talk about what Categories you might choose, and how to translate those into addresses, though, let’s talk about a computer metaphor (which was itself originally a metaphor for paper filing, but hey). Especially if taking the Top Down method, but arguably applicable to the Bottom Up method too, it’s helpful to think about directories or folders. “Directories” is the word used in Linux, and so is more common among folks who know how to and spend their time coding, whereas “Folders” is more common among folks who primarily use their GUI to deal with their computer. In this instance, I like directory better, because it’s technically more accurate (we’re talking about a collection of addresses, right?), but don’t get hung up on the word.

So, we can think of our whole Main Box as one big directory, and then however we divide it up as sub-directories, however we divide each of those up as sub-sub-directories, and so forth. It’s just the same as having a folder on your computer named “Zettelkasten,” then inside it having folders for “Humanities,” “Natural Sciences,” “Formal Sciences,” “Applied Sciences,” and so forth, and on down with folders inside folders inside folders. In the world of Zettelkasten, we can reflect these directory levels with numbers. When I gave my example of the Feud Law card and gave it a number of 2521, I was really looking at my categories: every address in the 2000s is Social Science stuff. Every address in the 2500s is Political Science stuff, and since 2510 was “Politics in the US,” I decided 2520 would be something like “Legal Concepts” and this was my first card on such things, so 2521.

I could just as easily write it 2/5/2/1 or 2-5-2-1 to make clear that later numbers are nest inside bigger numbers, but this saves some space. Also, if I think write a follow up to “Feud Law” (like the Edge cases from earlier), I can make that 2521/1. If it took me more than one card to write down all the edge cases, but they’re all “one idea,” I’d use 2521/1a, 2521/1b, 2521/1c, and so forth. If I want to develop that line “horizontally,” like “some new, different edge cases,” I might go for 2521/2. On the other hand, if I wanted to go “deeper” on original edge cases, maybe something like “hey wait, is this actually an edge case, or is it well-covered?” I might use 2521/1/1 if dealing with one of the original edge cases, or 2521/1a/1 if dealing with one of the new, different edge cases.

I don’t tell you all this saying “this is the one true way” or even that you should do it because it works for me. Instead, I’m just trying to give an example you can work from so you don’t get stuck on coming up with the “right” number. I don’t know if Scheper still recommends this, but when I was getting started, I assigned my top level categories by how Wikipedia divides itself (or at least, used to), with a couple of modifications and an arbitrary number at the thousands level for each: 1000: Humanities, 2000: Social Sciences, 3000: Natural Sciences, 4000: Formal Sciences, 5000: Applied Sciences, 6000: Recreation, Exercise, Games, and Leisure, 7000: The Great Work (Personal Development Stuff). I try to reserve the hundreds level for topics within those big picture topics (so, within Humanities, Language Arts, Philosophy, et cetera), and the tens level for sub-topics within those (within Language Arts, English), and usually don’t number a specific main note until I get to the 1s level, but that’s not necessary. And from there, of course, I can always create follow ons with slashes (or dashes or periods or whatever).

One of the many reasons not to get too worried about getting your categories “right”, is that there’s no such thing as a “pure” top-down approach in Zettelkasten. The whole point is discovery and exploration, which means that your nice, neat taxonomy won’t fit everything you come up with cleanly, and that’s good. That means you’re actually finding linkages between what you’ve read and thought and written, rather than sorting things into the appropriate boxes provided by others.

The danger with this approach is it tends to push your towards completionism and trying to build a reference tool, rather than a tool that tracks and organizes your thinking as it goes. As I mentioned, I used the Wikipedia organization as a starting point and now have lots of sections with nothing in them and I get confused on things like whether stuff on my entrepreneurial work should go in the “Business” section or in the “Self Work” section (it doesn’t matter, so I just picked “Business,” but it caused me some angst at the time).

The good news is that whatever approach you take, Zettelkasten is flexible enough that it’ll still work, and while you might reach a point down the road where you’re like “I wish I had made some different starting assumptions,” it’s likely not worth starting over from scratch.

Wrapping Up: A Flexible, High-Powered Tool

If you’ve made it with me this far, congratulations! I set out to write a “quick and easy” guide to Zettelkasten, but what do you know, it turned into something of a beast, at least as far as blog posts go. At any rate, we’ve now seen an incremental way you can step up your analog note-taking from “just write down your thoughts in a way that makes them easy to connect with citations later” all the way up to “carefully file your notes in such a way that they are more likely to help you think up new ideas and insights.” We’ve seen how the Bib Box, the Index, and the addresses and branching of the Main Box all give different, but interlocking, ways to find, link, and think about the knowledge you’re writing down, which helps to build deeper, richer connections, while also creating the chance for stumbling on the unexpected. Using the full-on Zettelkasten system is a bit of an investment in time and effort, but it’s remarkable for its ability to suppor the kind of overall broad, but deep in places, knowledge that we seek as Renaissance Men.

Appendix: Equipment and Logistical Considerations

Once again, we don’t want to get too hung up on details of implementation, but if you go down the Zettelkasten route, here are a few things that might help:

1. Choosing a Size of Card that Works for You: As I mentioned, I prefer 4×6 cards to 3×5 due to the amount of information you can get on them, but if you need to or just like them better, the smaller cards work just fine (just be prepared for more “follow on” cards, where you continue the same note on multiple physical cards).
2. Choosing Physical Cards that Work for You: Don’t worry too much about the specific cards you use, but you may want to experiment with blank vs lined, to see what you like better. Due to the ongoing encrappification of just about everything, “normal” index cards are much thinner and less sturdy today than they used to be, but if you buy the “thick” ones, they’re really thick, and rather expensive to boot. For my Bib Cards and Index Cards, I use Amazon Basics lined cards (extremely thin, very flimsy, but they do the job fine, and the thinness is almost helpful in using them as bookmarks while I read). For my main cards, I splurge a bit and get “Bullet Cards,” which have the same dotted grid as the Leuchtturm notebooks favored by those who do Bullet Journaling. I like them because they have a few more rows than traditional lined cards, but I can still write more neatly with a guide, and also, when I do want to draw a diagram or use a funky layout, it’s easy. Also, they’re sturdier than the basic lined or unlined ones from elsewhere, but not so thick as the heavy duty ones.
3. Writing Only on the Front: Other than a few special exceptions (Bib Cards and cards I use as my daily log for exercises and practices like meditation), I don’t write on the backs of my cards. This is purely to make it easier to skim through cards in the box merely by flipping through them, or when laying them out on the desk, rather than having to take them out and flip them over.
4. Colors of Pens to Use: Many folks find it helpful to color-code certain aspects of their notes, such as making bibliographical references red and card address references green, so having some different-colored pens can come in handy. I also prefer to make my titles a different color than my main text, but that’s almost entirely an aesthetics thing (Chocolate Brown for titles, Sepia for main text)
5. Writing Cards Away from Your Set-Up: If you have to do a lot of your work “on the go,” it quickly gets infeasible to carry your whole Zettelkasten around, so you’ll want some way to carry around a subset of your cards and some pens, like a small case, box, or folder.
6. Housing Your Zettelkasten: The kind of box you use doesn’t matter much, especially at first, but if you keep with it and your notes grow, you may want more space, and/or something nicer. For my Main Notes and Bib Box, I currently have two Vaultz Index Card Cabinets (one drawer is my Bib “Box”), and for my index, I have an Acrimet “Card File Holder,” with a metal bottom and clear plastic flip-up lid.

All of the above was provided in case it helps and to give you some idea of the kind of stuff that might come up, but really, don’t get worked up about any of this stuff, or spend too much time on it: the main thing is to get to note-taking.