🌲 The Value of Consistent Naming Conventions

Naming conventions are critical to my organizational methods, not just because they help keep things in their proper order, but because they telegraph a file's content before I open it. They don't come up often, which is a shame — because in some ways they're as much a cornerstone of personal knowledge management as links, tags, folders, or metadata.

As much as I love folders (and I do love folders!) they're just one tool in my toolbox. Naming conventions are one of the ones I don't talk about as much, mostly because they don't come up in conversation very much — but maybe that means people aren't leveraging them as much as they could be.

There are a couple of different ways to leverage naming conventions for organizing files. I'm less interested in organizing for organization's sake, though. I care about maximizing the amount of information I can get at a glance. That means avoiding exclusively timestamp-based unique filename methods like advocated by some folks in the Zettelkasten community — unless the idea of naming a file makes you completely freeze up, in which case, go ahead. At least a timestamp will sort your files in order and put them together with other things you were thinking about at the time.

But I do believe that you'll be missing out on a valuable tool for sensemaking. For example:

Some people like to prepend their filenames with symbols. I think Bryan Jenks is one of the most well-known proponents of this — you can check out his emoji taxonomy on his public notes page. One thing I'll note, though, is that emojis in filenames make me nervous, though, because they aren't supported by all operating systems and sync solutions. I also personally tend to find them confusing, because I'm not a very visual person and I have trouble remembering the meaning I've given to various symbols. I prefer to use a more "natural" system.

Sometimes, that means prepending with numbers, à la the Johnny Decimal system. This allows for sorting in file systems. I mostly do this with my teaching files, which are naturally organized into numbered units and lessons by my curriculum. I also do it with my newsletters, though — there are lots of files named something like Obsidian Roundup 2022.01.22 or 2021.12.20 Giants. But I can easily tell the difference between those two newsletters because of where I put the numbers. Here are some other consistent naming schema I use for my newsletters:

• Topical research overview: 2022.01.03 Bugs — aka Moth flour, cyanide defenses, & esoteric ways to die.
• A piece of very short microfiction (~100 words): 2022.12.29 Vagrant (MF) — part of Diets and vagrancy: our changing relationships with food.
• The afterword to a piece of very short microfiction, like Mock: 2022.12.08a Sometimes quitting is the right choice.
• A piece of flash fiction (~1000 words): 2021.12.01 The Black Stones of Sanctuary — part of How a building can be too clean for comfort.

This method ensures that my files sort in the order I published them, which makes it easy for me to see gaps and get a sense of my publication pace. When I query my notes for "all stories that mention Surzi," I can see from the titles alone that as of right now, I've already published everything I've written about that character — because they have a numerical date in the file name, instead of just being the title of the story.

I can even differentiate easily between microfiction stories, which are always given a one-word title, longer flash fiction pieces (which are always a couple of words), and novels — which tend to be called something like Temple Mage, Civil Mage, Carmine Mage, Collared Mage, etc. for my high fantasy series, and Charm City Curses or Charm City Conflagration for the urban fantasy series.

As much as I extol the virtues of folders (which you can pry out of my cold, dead hands), I do not actually want to have a folder for every novel universe and a subfolder for every novel (much less chapter!). Consistent naming schema like Civil Mage 01 - Sewers & Temple Mage 01.2 - Desert let me keep all of my novel files in one place, neatly sorted and above all clearly belonging to their respective projects.

But it's not all about sorting. Sometimes the goal is just to know, before you preview or click, what the file is going to look like. Past a certain point, numbers aren't great at that. You have to use another method.

The example I gave back in March was:

MOC: Egypt → LITNOTE: Women in Egypt by Whoever → ZETTEL: Egyptian Princesses had less work but more prestige than their Sumerian counterparts → ARTICLE: Egypt vs. Sumeria — The Role of Princesses in the Fertile Crescent

I called them "increasingly atomic" although I probably should have said "decreasingly atomic." Typically, I start with one broad topic (with a very short name), like Egypt or Phoenicia. Then there's a source note, usually imported from Readwise, and the file name is always {Title} by {Author}. I spin off all the insights I glean from that article into index card style atomic notes consisting of a single claim (as the file name), evidence, and any explanations I need to understand its connection to my other works. This usually winds up looking like:

Then, I can expand that claim out into a complete article, usually after I've added additional evidence to the file. I mark articles with Title Case, which is an easy and natural visual signal that the file contains a product and isn't just a claim that could be slotted easily into a sentence.

But the question that inspired me to write this was actually about organizing worldbuilding articles.

My worldbuilding folder has a pretty flat structure, with notes about fictional geography mixed in with notes about fantasy animals, invented gods, and made-up diseases. Currently, I can differentiate between them pretty easily. Gods in Verraine tend to have names like The Architect or The Realmwalker whereas cities have names like Uskune. Recurring events that happen in those cities are more likely to be called something like Uskune trade fair. Animal names are always lowercase, i.e. yhaoginli or moonfish. At some point, I might start forgetting the difference between euphorigum (a drug) and aetherock (a building material), but since I made sure the names were evocative (one chews the gum to evoke euphoria, one builds with faux rock made from magic known as aether), I expect it to be awhile before that happens.

I'll turn to metadata for sorting when it does, though, and start labeling things as a location or a creature or an organization, but until then, strong intuitive naming conventions are a strong bedrock — usable even in a "flat" (folderless) organizational system.

Flat systems can still have structure. Some people, including myself, prepend important folders and notes with something to make it "sort high," like ! or _. Conversely, you can make things "sort low" with characters like Ω or zz — but I don't like the way it looks (it feels unnatural), so I don't do it often. In fact, I only use this trick when sorting photos, so that the avatars and logos in my photos folder get sorted above my dated vacation photos folders.

Do you have any tricks like that to share? Consistent naming schema of your own? I'd love to see this phenomenon discussed more "in the wild," so feel free to link me to any write-ups you make or are aware of :)