Zettelkasten linking for surprising connections

I’ve been reading about (and watching videos on) Zettelkasten with Obsidian for the past few weeks, and the claim that surprising connections naturally emerge from the system is really appealing to me.

I just don’t understand how people are linking in a way to cause this to happen.

Here’s an example:

Say I’m reading and making literature notes on Kant’s Groundwork for the Metaphysics of Morals. I write down the quote, “Act in such a way that you treat humanity, whether in your own person or in the person of any other, never merely as a means to an end, but always at the same time as an end.”

Later, I convert the famous categorical imperative into a permanent note by putting it in my own words, say:

Always remember that people are conscious beings with emotions, hopes, and dreams worthy of moral consideration. When we use someone to get what we want, this dehumanizes them and treats them like an object rather than a person.

I create some links to obvious other cards with keywords like “moral philosophy” or “ethics.”

Three years and 9,000 cards later, I’ve forgotten I ever made that card.

I’m reading a linguist analyzing the use of language in Joyce’s Finnegan’s Wake and I convert a literature note into a permanent note, something along the lines of:

Historically, language in novels has been seen merely as a means to convey story. Joyce says language use can be the core artistic value of a novel; an end in itself.

Now, for us, I’ve juxtaposed this so that the connection is obvious. In the example of years later and forgetting about the categorical imperative card, I see two options.

  1. My brain still makes the connection, and so I didn’t need Obsidian or Zettelkasten.
  2. I don’t see the connection, and I can’t fathom any sorts of tags/links that would get this very clear connection to naturally emerge from the system.

So, the question is: how do people link to get these sorts of connections to emerge? I would think this is the whole point.


This is a great question I have been struggling with as well. My current thinking, which I am still working out, is to use the elements of information for connectivity to other ideas such as facts, definitions, concepts, principles, and theories. To encourage interesting links between notes, link them to a separate note that captures a principle or heuristic, a construct or concept, or a theory. Another linking place is universal questions or problems (e.g., how should we treat others?). In both of your examples, I see universal questions and key terms which will likely connect to other ideas. Creativity requires pulling things apart to their bare bones so they can be reconstructed in new ways.

Maybe a note operationally defining “means” and another for “ends” and another that juxtaposes them or a compare/contrast that brings those two notes together into a model or concept. Then link them to the universal question of how people should treat others. The means/ends would also link to the Joyce note in the same way which might link to the question of how to use language in novels (assuming there might be other answers).

So far, this approach has led me to create some valuable atomic notes that link things together, but it still leaves me with a lot of meaningless notes until I spend time looking at the graph and piecing together a new idea. But, if I look at a universal question/problem, the ideas are instantly available.


This is “the” question.
Something I find useful are structured notes which help me create indexes of similar ideas with maybe one or two quotes or bullets from each to provide context.

The act of putting these types of notes together (mostly through a bunch of keyword searches based around a topic) introduces serendipity into my thinking, and acts as a source of inspiration when looking for interesting ideas or analogies to link.


I believe this type of plugin, which i believe is still WIP, holds some of the answer: GitHub - SkepticMystic/graph-analysis: Analyse the structure of your Obsidian graph using various analysis techniques

Having the graph analysis open in a side pane generate ideas/associations through different types of relations as you type.

I need to experiment with it so i can not yet claim any “ah ah” moments. Give it a go.


Thank you! I’m glad I wasn’t missing something obvious.

The more I think about this, the more it seems a good “best practice” habit to form is to take any new permanent note and immediately try to phrase it in the most abstract/general form possible.

Then if something genuinely useful looking appears in this process, make another new permanent note linking back.

That’s what I picked up from the Zettelkasten.de forums. If you have the “means” page and the “ends” page, then you link them with a note defining the concept of “means vs ends.” Now, you can use any of these three in different ways. If you uncover something interesting doing this, keep us posted.


Don’t rely on single-degree connections - they are close to useless. They really are, and I’m not kidding because I’ve been in this backlinking thing for close to 3 years moving from Roam and now in Obsidian for a long while. Doing this single-degree thing does allow you to access it in future (given if you ever do, given your inflow of torrents of information and your discipline level) but from experience it builds cobwebs of underutilised information that never see the light of day. It feels good because it is lazy, but ultimately isn’t good for the health of your PKM.

My solution to this:

  • Go at least 2 degrees deep, and make this quickly accessible so that it doesn’t become troublesome. Some use Graph Analysis plugin. I prefer having a panel of the Local Graph in my sidebar, with depth level increased to maximum.
  • Make note titles like APIs. They should contain full opinions/claims/succinct summaries or act like thumbnails to a bigger body of information.
  • Whenever you create a note, always create [[access points]] like these. They will create uncreated nodes that help you connect to this note in future. For example, if you typed [[dehumanisation|dehumanised]], everytime you create a subsequent new note in future that links to [[dehumanisation]], it will show you this old note that you’ve created.
  • A quick way to create new notes, or “thought sequences”. I call this Threading, which I use the Breadcrumbs plugin for. For example, this entire forum thread is a “thread”, sequential ideas that take cues from the previous post, a logical sequence of ideas.
  • Another way to pull up “distant links” is to use the Obsidian Sidekick plugin which you can install through the Obsidian42 BRAT plugin. Or you can install the Various Complements plugin. Basically it highlights partial matches of things that you are typing which can jog your memory and provide possible linkages to your current material.

Here’s a video I made showing my process: How To Use Local Graph View In Obsidian - YouTube It’s a little long, but I promise that it will spur some new ideas in how to structure your work.

The power of the local graph view needs to spread. It’s such an integrated function within Obsidian but so severely underused. It has grown to be the single most powerful function in Obsidian that I cannot do without.

I’m not a big fan of ‘retrospective work’. I hate reviewing, I hate going back and processing things. So much wasted time, and the information is already probably lost in context by the time you come back to it. It’s not only inefficient, it’s inorganic and not how our mind works. This approach makes and forces you to have everything you need on hand, forces you to connect to old ideas immediately upon consumption.

It’s almost like weaving a spider web, and I’m considering calling this “Knowledge Weaving”.