How should be the Zettels connected in terms of their direction ?

ulver48

Hello guys and girls.

I know some things about the Zettelkasten method but I guess I am implementing it wrong, so I would like to have some guidance. What confuses me right now is the direction of connectivity. In other words should the child notes point to the parent nodes, the parent nodes to the child notes or should they be bidirectional ?

Index notes in particular point to entry notes for a particular theme. Arriving on an entry note from there does nothing if it does not point to other notes with links ( parent to child connectivity). Most new notes I create point to parent notes and thus the unfolding of new ideas is a bottom up movement. So there is a parent to child connection from the index to the entry notes and child to parent connections from the bottom. I feel that this method makes searching difficult and both the index and other simple notes are becoming more or less useless. In other words navigation becomes hard.

I use Obsidian and I have a big folder with all my notes in there and indexes in the from "year/month/day/hour/minute - Title". I currently create an new Zettelkasten so I have still time to fix things before it turns out to be a mess.

Thanks for your time

Will

Welcome to the forum. I hope we can help.

@ulver48 said:
I know some things about the Zettelkasten method but I guess I am implementing it wrong, so I would like to have some guidance. What confuses me right now is the direction of connectivity. In other words should the child notes point to the parent nodes, the parent nodes to the child notes or should they be bidirectional ?

Bidirectional.

Index notes in particular point to entry notes for a particular theme. Arriving on an entry note from there does nothing if it does not point to other notes with links ( parent to child connectivity). Most new notes I create point to parent notes and thus the unfolding of new ideas is a bottom up movement. So there is a parent to child connection from the index to the entry notes and child to parent connections from the bottom. I feel that this method makes searching difficult and both the index and other simple notes are becoming more or less useless. In other words navigation becomes hard.

I find I use full-text search more than I use following links as a method to navigate my zettelkasten. With 1600 zettel mousing around, clicking links feels less productive of my time.

I use Obsidian and I have a big folder with all my notes in there and indexes in the from "year/month/day/hour/minute - Title."

This is the best setup, I think. Everything in one folder and all files named with the format.year/month/day/hour/minute - Title in a markdown file format.

jeannelking

@Will said:

Bidirectional.

Yep. Bidirectional 100%. Also, know that bidirectional links will emerge as your zettelkasten evolves. I might start with a parent-to-child connection when I create the card because it's obvious. But as I continue to feed my ZK and it feeds me, I start discovering those bidirectional connections. The more you feed and exercise your ZK, the richer those bidirectional connections become.

For example, the child zettel of elegance comes from its parent, innovation. But it also connects to creativity as well as several subsets within creativity and innovation, as well as categories or ideas not directly related to either. But elegance doesn't have to link to every single thing about innovation or creativity...just the concepts with the most power or juice for connection. This approach may help with the navigation/search issue you raise.

I had to take this approach with my tags. For a while, everything remotely related to creativity got the tag #creativity. Except when I'd search for #creativity it was the equivalent to highlighting an entire book: the search pulled up everything, making it unusable. Judicious removal and application of the #creativity tag has helped make my tag searches much more effective and manageable.

I find I use full-text search more than I use following links as a method to navigate my zettelkasten. With 1600 zettel mousing around, clicking links feels less productive of my time.

I find I use both about 50/50. (For context, I use Archive and physical cards in my system.) I use the links embedded in my Archive cards for quick movement between linked cards. I used text search for when a link isn't embedded but I know there's a connection somewhere waiting to be had.

Will

@jeannelking brings a fresh perspective to the connection question with her creativity and innovation example.

@jeannelking said:
... I continue to feed my ZK and it feeds me, I start discovering those bidirectional connections. The more you feed and exercise your ZK, the richer those bidirectional connections become.

I find now I can spend as much time creating a zettel as I spend in the connection discovery phase. Sometimes I just have to yell "STOP" and move on to creating the next zettel.
As an example a short atomic zettel surrounded with 9 bidirectional connections. It took a long time to search up those connections. Very valuable time I wouldn't trade for some automatically software generated convienience.

I find I use both about 50/50. (For context, I use Archive and physical cards in my system.) I use the links embedded in my Archive cards for quick movement between linked cards. I used text search for when a link isn't embedded but I know there's a connection somewhere waiting to be had.

To be more clear it feels I use about 75/25 search and link connection. (My zetttelkasten is digital.)

Phil

@Will. I found myself inadvertently trying to click on your links in the embedded image. Hahaha.

ulver48

So when you have a new thought and you are transforming it to a permanent note, when you link it to a previous note ( the parent note), you don't necessary add an extra link from the parent to the child. As I see it, these bidirectional links are creates spontaneously as the number of notes increases and as you revisit older permanent notes.

GeoEng51

@ulver48 said:
So when you have a new thought and you are transforming it to a permanent note, when you link it to a previous note ( the parent note), you don't necessary add an extra link from the parent to the child. As I see it, these bidirectional links are creates spontaneously as the number of notes increases and as you revisit older permanent notes.

I use mostly bi-directional links. As @Will has posted in the past, if you include the UID for your current Zettel after its title (say, next line), then when you click on that (self-referential) UID, you see all forward and backlinks anyway (in the note list on the left of The Archive).

I tend not to think about "parent" and "child" relationships. What do those terms actually mean, in the context of a ZK? All zettels are created equal, in my mind, and if you use bi-directional links, all links are also created equal. Let the organic nature of your ZK network prevail!!

Will

@ulver48 said:
So when you have a new thought and you are transforming it to a permanent note, when you link it to a previous note ( the parent note), you don't necessary add an extra link from the parent to the child. As I see it, these bidirectional links are creates spontaneously as the number of notes increases and as you revisit older permanent notes.

You are right because the web of connections (web of ideas) increases in complexity as older notes are linked to newer notes and newer notes get connected with older notes. In a sense, the more the merrier.

I resist the notion of parent note/child note as it tends to solidify some imposed hierarchy. The notes are just proxies for ideas. Ideas connect to other ideas. They nourish each other.

See my diatribe on "permanent note".

Sascha

There are no true parents or children in a heterarchical system like a Zettelkasten.

A connection needs to be contextualised. So, if you don't explain to your future self why there is a connection the value of this connection will deteriorate over time. This is over the long-term very important to create surplus value. In short: Placing a link without explaining why you see the relationship to your future self means that you postpone work that should have done in the moment in which you had this information.

So, I am not answering to you directly. However, I suggest to shift your focus towards the context of the link. Then the question will be answered in a case by case fashion by the content of your notes. Sometimes, bidirectional and sometimes not.

As far as I am using the Zettelkasten, I did not encounter the need to make bidirectionality a point. Automatic backlinking, which is automatic bidirectionality, as a feature of the software or as a habit to look them up, is even diluting the quality of the overall link structure because it increases the amount of links with no context given. (article is coming)

zk_1000

Contextualizing creates very strong connections. I am using hard links for that. Connections to indexes (not index notes) are much weaker than that. Even a heterarchical system allows true parents or children, to some degree. Otherwise it would be pointless to create structure notes.

To me, it looks like you are struggling with your tools. I had this problem, too, because on one hand the Software did not exactly do what i wanted, on the other i was actually able to do something with this Software, that in turn helped me figuring out how the things are working. Software combines theory with practice.

For one, it is a matter of taste, like how many structure notes you create, how strong your connections are, how many words you use in a note, etc. With experience comes fashion and style.

For another, it is not a tool that creates a connection. The first thing to figure out is how your ideas are connected. Nobody can tell you that the cow is connected to the dog, and the chicken to the egg. This is your mind, you decide!

Finally, your tools represent this connection exactly like you think it should. So if you search for it and it becomes difficult to find, thinking about different kinds of connection is not going to help.

What you can do is take a step back and show us a specific problem with the tools you use. What do you want to achieve, how exactly do you do it. You can simplify it / obfuscate it if you want.

Sascha

Even a heterarchical system allows true parents or children, to some degree. Otherwise it would be pointless to create structure notes.

What if the children of the children link to the Structure Notes?

zk_1000

@Sascha said:

Even a heterarchical system allows true parents or children, to some degree. Otherwise it would be pointless to create structure notes.

What if the children of the children link to the Structure Notes?

What you have then is a heterarchical system. I was referring to this:

A heterarchy may be parallel to a hierarchy, subsumed to a hierarchy, or it may contain hierarchies; the two kinds of structure are not mutually exclusive. In fact, each level in a hierarchical system is composed of a potentially heterarchical group which contains its constituent elements.

So, isn't the heterarchy form of a Zettelkasten deriving from ranked structures? My perception on this is that heterarchical links are not very beginner friendly but rather a structure a Zettelkasten grows into.

I haven't given this much thought yet, but for instance i am currently not able to search for heterarchical connections with the tools i use. So right now i can only create those links, but to find notes i can only do so indirectly on subsumed structures.

I am not sure if this needs to be improved, Luhmann was doing fine without it. It's on my todo list.

Sascha

Ah, I re-read your previous comment. I missed "to some degree". I think we mean something similar. I stumbled over the "true parents".

So, isn't the heterarchy form of a Zettelkasten deriving from ranked structures?

No. Structure Notes themselves are functionally not different from other notes. They are linked and they link. However, you can "paint" whatever structure (pattern) you want on the individual structure note. I choose the hierarchical patterns because those are not present to you if you rely on the direct link structure only. (hindsight: Are you refering to Structure Notes?)

But imagine just drawing an image and placing IDs on it. Still a structure note because it offers you a structure in which you can place.

I suspect the difficulties of the poster @ulver48 is symptomatic for the first issues ZK users will have if they do not prepare for increasing complexity. Those problems do not arise miraculously. The main issue with note taking systems is bloat. The antidote to bloat is minimalistic software (to avoid encouragement of using different techniques to solve same or similar problems) and applying a method that takes complexity into account (which is: autoscale for any size of the note archive).

I am not sure if this needs to be improved, Luhmann was doing fine without it.

I don't think this is a good heuristic. Luhmann was a workaholic who dedicated his life to a single purpose. And what if he had a (I suspect better) method with digital tools?

---

I don't want to hijack this thread. So, just some bullet points that connect those kind of issues to the bigger picture. The first signs of issues that will bite quite some people in the ass:

1. Software developers are thinking theoretically on note taking problems and applying theory with insufficient empirical knowledge.
2. People who learned about the method just recently publish about ZK as if they researched it extensively.
3. Echo chambers emerge with their own paradigms giving rise to dogmas.

I saw the same phenomenon in my current field of work (health and fitness). So, I prepared (I hope sufficiently).

Even ZK should be a pretty dry topic there is still too much focus on eye candy and to less focus on the productive basics.