Bryan Jenks has an interesting approach to tagging notes in Obsidian, which I’m considering to try out.

What tags are

Tags are a way to group notes together. When multiple notes carry the same tag, you can search for them by that tag, you can view them as a group in Obsidian’s graph view, and you can pull them into a Map Of Content (MOC), a note that contains a list of the group of notes in question, using plugins.

With a status type of tag like #todo you could indicate that the note in question contains a task you have assigned to yourself. A tag like #research might indicate that you intend to conduct (or are conducting) research on the topic of the note. #contact could tag a note in which you mention a person you want to reach out to.

To me, status communicates the stage of a note in its content life-cycle, as in: what has been done with this topic and what is the next step? For consistency you may have designed a typical life-cycle for a note, or perhaps even different life-cycles for different kinds of notes. The first status might be #inbox.

Tags add metadata to notes to group them together, meaning that you can view them as a group.

In the case of status tags, these notes may or may not have anything else in common than their status. Their subject matter or context could be entirely unrelated.

I have also used tags to indicate other properties of a note, for example #person, #event, #publication, #title, #channel, #definition, #template, #observation, #opinion, etc.

I’ve used tags to indicate media format, such as #text, #video, #audio, #infographic, #film.

I’ve used tags to indicate media genre, like #news, #documentary, #interview, #quote, #panel, #column.

As you can see, it gets messy. To be honest, I’m still learning how to build a productive process around this kind of tagging.

Keep using hashtags liberally

Now let’s move on to subject matter. All the tags mentioned above are metadata to describe various properties of notes. Some of them describe their status. But when you look at published content, tags are mainly used to describe subject matter, ie. what the content of the item is about.

On the fediverse, #hashtags are even more important than on the conventional web and centralised social media platforms. On the federated web they are the main drivers of search and discoverability.

I would like a set-up, a metadata design that works both for my posts on the fediverse and for my personal knowledge management system.

So I do think there is cause to keep using hashtags liberally when describing content.

Compiling Maps Of Content with Dataview

The other day I learned about Maps Of Content (MOCs), the way Nick Milo has described them (https://www.youtube.com/watch?v=WUq8Pun28FI) in combination with Dataview (https://blacksmithgu.github.io/obsidian-dataview/), a community plugin for Obsidian, as a way to pull together lists of notes with same metadata.

Watching this YouTube video https://youtu.be/zIh1S7ra3aI by Bryan Jenks made me understand that we can not only create MOCs based on tags, but also based on regular internal links.

Bryan explains some advantages to creating MOCs from links. One potential advantage is that it can allow you to simplify your tags set. You could limit it to status tags.

A bigger advantage is that you can allow MOCs to develop more organically, more bottom-up.

Tags and internal links side-by-side

When you write a note, you can create internal links for words and phrases that are key to describing the content of your note. If there are no notes for those linked keywords yet, don’t activate them. On the graph view they will show as uncreated notes with a dedicated colour.

Their size on the graph indicates how many notes reference them. As part of your process, when an uncreated note reaches a certain significance based on the amount of other notes linking to it, you could activate it and even turn it into a MOC if that seems useful. This way, the previously unactivated link would have functioned quite similarly to a subject-matter hashtag.

In conclusion, for now, I will run both types of metadata side-by-side: hashtags as well as internal links. And I’ll create group views, ie. lists, using Maps Of Content and Dataview.