Zettelkasten

Why You Should Set Links Manually and Not Rely on Search Alone

In the age of lightning fast full-text search, what are links in the archive of your notes good for?

In short, search queries are deterministic means to get to Zettel notes: they produce the same results every time. Links are hand-picked references. The idea that makes you create a link is unique, and so is the resulting link. It’s much more personal and thus better suited to help you think.

Your personal and spontaneous associations count: they are due to a lot of factors not inherent to the material. Maybe what you’ve heard on the radio at that day comes to mind while you write a new note or read an old one. This association may never repeat in your entire life.

A manual link to another note can solidify this spontaneous association for a long time. You can annotate the link to remind your future self of the context. All of this will probably make sense to you and you only. It’s your second brain, after all, so that’s actually a desired effect.

Links are the strongest tie between notes:

  1. Direct links,
  2. Keywords/tags
  3. Full-text search, tying notes together via similar words

One argument for relying on full-text search is this:

Setting links manually is a waste of time because this links are not different from search results; in fact, they are the same. Some search has produced both the current note and the note it’s linking to in the past. A link is simply the remark that “these two things are of the same kind”. A link saves you from typing in in the search term again. That’s all.

As a side-effect, you can ditch specialized Zettelkasten software. All writing applications support full-text search within a single file. Just put all your notes into one file which becomes a representation of the stream of your knowledge.

We had a great discussion in the comments in the past about this.

I disagree. Here’s why:

Each Zettel note relates to a chunk of information if you keep it atomic, which you should. To keep every chunk in its own Zettel note file is likely similar to your existing mental model of how information work. (You can separate and relate information. Information is not just a big pile but chunked.) Consequently, having a single file for convenience is actually not so convenient.

Also, over time, your writing will evolve. You’re going to use different terms and also think differently.

If you wrote a note 10 years ago, I bet it’s likely written differently than a note on the same topic you accidentally and unknowingly write today. Except for highly technical terms, your concepts will change and evolve. Relying on similarity of terms in the body of a note is not too reliable.

Last, skimming a lot of search results is very taxing cognitively. To find “the right thing” requires you take at least a short look at, in the worst case, every search result. If there was a link from note A to B, you’d immediately know how to follow the current trail of thought.

Of course you always could try to rely on memory to find that thing you know is among the search results. But you may be wrong. Relying on memory is very error prone. That’s why you use a Zettelkasten in the first place, to get peace of mind through externalization of information. So this fake way out is actually a really bad idea.

In the same vein, links structure the mess. Instead of an otherwise unsorted pile of stuff which is your search results, you get access to a pre-formed structure in your Zettelkasten note archive.

As Dan Sheffler put it in the comments of an older post:

When one creates a new note and looks for which notes to connect it too [sic!], the preexisting connections between old notes lead you to more notes than you would get by memory or search alone.

The resulting web of connections is more than just a sequence of searches.

watchmaker
Photo credit: Francis Stewart, War Relocation Authority photographer (NARA record: 8464475) (U.S. National Archives and Records Administration) [Public domain], via Wikimedia Commons

When I write a Zettel, tags come to my mind which I could use for this particular note. While I write, I also remember older Zettel or clusters of related ideas which I would like to look at and evaluate for use as link candidates.

In short: in order to create links, I rely on my memory. But wasn’t that what I’d rather not do according to the previous section?

I use my memory differently in the case of linking from when I put myself in front of my Zettelkasten in order to find a particular information.

  1. The web of notes makes it easier to continue old trails of thought because the structure of the Zettelkasten mimics the structure of the brain.
  2. Conversely, without the internal references (which are links and tags) I’d have a hard time utilizing the Zettelkasten as external memory. The whole writing me, “being” part physical brain, part digital Zettelkasten, will remember worse if the Zettelkasten would not mimic the brain’s structure.
  3. To write about a topic will make it easier to remember past ideas you’ve written about. You don’t exactly continue where you left of years ago, but you may recreate parts of your past automatically.

Even having the mental picture of a book I’ve read in the past is a helpful cue. Using the citation key of my reference manager helps look at a specific subset of my notes.

In any case my memory helps to get a good angle to find useful notes in the archive.

In the worst case, if nothing comes to mind when writing, I can still use the full-text search to poke around until a spontaneous idea comes up.

Search becomes more the source of inspiration than the mechanical tool to retrieve notes. Working with a Zettelkasten requires I am “on”, actively engaging my brain. Links help me take note of my brain’s efforts.

Conclusion

We all rely on search, so why add custom connections manually? Instead, let’s keep it simple and use the search exclusively!

This reductionist argument misses a few points, as I hope to have shown.

Let’s say your Zettelkasten note archive grows to 10’000 items. You may have spent 6 years to reach that point.

It’s unimaginable to work with a single file up to this point, because the notion of a Zettel as a single entity, as the representation of a chunk of information, is useful for your mental model. One big pile of text is unwieldy, it cannot be grasped.

It’s unlikely you write notes today as you did 6 years ago. You use different terms.

Every full text search presents tons of Zettel notes. Because all are seemingly equal (and connected in the same way: not at all), you have to either rely on your memory to distinguish useful from useless notes, or look through all of them every time.

Explicit connections make it easier to work with notes. Creating links is time well spent.

Connections, being links of tags, become something to look for. They, too, are elements of your Zettelkasten. They tie Zettel notes together, making some paths easier to follow than others, thus helping you filter useful from useless.

Have you thought about ditching links in the past? Was it probably because you just don’t seem to have related notes which qualify as link targets?

Share your experience with us in the comments!


Sascha’s Comment: Christian made an important point. It is not only important that you connect. How you connect is important, too.