Taxonomy technique (for hierarchical parent groupings)

I have some taxonomy entries in my notes. It turns out that I end up changing the taxonomy at times (adding/removing common ancestors), and then my cards are no longer reasonably consistent unless I put in a lot of work.

I’ve come up with this method:

taxonomy-index.md: (can become quite large, but that’s kind of the point)

^^Each of line is a branch of the taxonomy tree - and animalia for example links to the page for the taxonomy entry of animals. Order only matters for maintainability - so I sort them alphabetically and that’s good enough.

Now in my file on say, dogs, I can say

bio-taxonomy-dogus-dogigenus.md:

[[bio-tax#^tetrapodia|taxonomy]]

Dogus dogigenus

Woof!

and that generates a mouseover link “taxonomy” that has the hierarchical reference I want.

If I want to insert a common ancestor or take one out, I can do it in the taxonomy index file, using the <option>-click selection for multiple lines at once.

6 Likes

Interesting approach. How beneficial do you find this approach vs using tags or index notes?

What types of taxonomies do you manage, and do you attach a single taxonomy to every note? Or do you allow multiple taxonomies per note, or allow notes to optionally have zero or more taxonomy links?

I’m trying to understand the use case here, and why it is better for you to recreate a strict hierarchy via links instead of using folders to encode the structure.

1 Like

Note that Obsidian already supports nested tags, e.g. #test/mid/end is rendered in the tag panel as:

image

@davecan the benefits of using tags over folders, is that a card can belong to multiple taxonomies, e.g. “Problem of induction” is fundamentally a “philosophical / epistemological” thing, but has also utility in business strategy as well. in my system, I have it tagged as:

#strategy/change #philosophy/epistemology

Yep I understand that but thanks. I’m specifically trying to understand the benefits that @elmsfeuer is finding by creating a hierarchical structure that can be easily recreated using either hierarchical tags or folders. Especially if the notes they are creating each have only one taxonomic location.

There are potentially some interesting uses with Hieronymi’s systems science taxonomy but even here hierarchical tags may be more appropriate. Normally I don’t use content-based tags but have softened my stance to allow certain tags to arise inductively from the notes (e.g. through application of grounded theory, etc) which results in a bottom-up derivation of ideas reflecting my own self-created collection of ideas and patterns and themes identified in my notes, as opposed to applying a pre-defined taxonomy from someone else to my thinking. Because of this it would be difficult to apply this pattern since the taxonomy is not defined in advance.

But since @elmsfeuer stated the taxonomy changes I’m curious how it fits in their workflow, since I would also need any such taxonomy to change as well.

TLDR: In a deep tree that is changing, I want to be able to rapidly inspect the properties of all the parent groups of a note, and keep all the notes ancestors consistent. Less interested in neighboring notes or finding members of categories.

My use case at the moment is animal and plant taxonomy. These trees can be many levels deep, and each layer adds a new “feature” or two (bilateral symmetry, exoskeleton, develops with mouth first or with anus first etc). A quick review of all the ancestral categories provides a nice summary of the properties of the organism.

It turns out there are a few different systems and the organization is in constant flux. Initially, it was all naturalist observations but now there are sequenced DNA genomes and the structure is changing.
Also, for ease of use often not all the ancestors are shown, but an abbreviated version and one skilled in the art fills in the gaps (I am not skilled in the art).
So as my understanding improves, I can quickly change the hierarchy for many cards consistently.

I like to link to each of the parent groups, and preferably not hierarchical (that would be another approach), so that I can use the mouseover feature to quickly see what properties each group has (tags can’t do that afaik).

First, I made many cards with explicit, handwritten parent groups, just to find that I was off, and I had to go revise many files. That’s error-prone and I don’t like it.

My interest is mainly quick review from the leaf node back up the tree; there is, even with the ~~~query syntax, no trivial way to list the members of a group just now. I’ll probably add tags to the cards as well so they are searchable.

I considered making a script that updates the md files (based on an indented list or yaml file or something); it would update the taxonomy line and add the relevant tags, but this works well enough for now.

(Tree of Life Web Project is a nice taxonomy tool)

1 Like

Ok I see. So your need is to reflect a legitimate externally-imposed taxonomy in a situation where use of that taxonomy makes perfect sense. This would be less useful in horizontally hyperlinked notes but I can definitely see it being very useful in a situation like yours. Thanks for the tip.