Crystal Lee (she/她)

Two opinionated approaches to personal knowledgement management

One of my favorite ways to productively procrastinate is to optimize the way I interact with my computer: create rules to sort files automagically with Hazel, install new Alfred workflows to help me find stuff just a little faster, or play around with my calendar and time blocking. The really fun and overwhelming thing to do is cleaning up my personal knowledge management system (PKM): to make all of my notes and files streamlined so that I can theoretically manage all of my projects a little more efficiently. We can debate whether or not any of this actually helps me, as my partner often ribs me for taking more time to organize and document things than it would be to the thing. That being said, I do think that being able to methodically catalog all of your files can be truly useful, particularly for someone like me who loves to collect stuff like class syllabi.

The two most opinionated approaches I know of (and have used) are the Johnny.Decimal system and Tiago Forte's PARA method.1 Johnny.Decimal creates a personalized Dewey decimal system that optimizes for ease of searchability, as it is a static file hierarchy that prioritizes sequential collection. On the other hand, the PARA method asks you to sort everything you have into four major folders (Projects, Areas, Resources, and Archive) to prioritize ongoing projects over ordered documents. The PARA method instead uses a flexible file structure, allowing you to move projects easily from folder to folder, optimizing for ease of discoverability. The argument that Tiago Forte makes in his incredible book (which is oft-fawned upon by the knowledge management crowd), Building A Second Brain, is that files are only useful if you are able to discover and use them again and again. They shouldn't be static, since it means that they will languish in whatever folder you have them sitting in, never to be used again. The PARA method also explicitly addresses the issue of note-taking, whereas Johnny.Decimal (as far as I know) does not explicitly draw out note-taking as a specific use case. I wholeheartedly agree that many of the notes I've taken have languished in those folders, particularly seminar notes -- this is also the problem that folks using the Zettelkasten system promise to solve, since creating ideas that are linked together increases the possibility of discoverability compared to sequential notes, which very rarely have a life outside that specific folder. How often have you been able to discover notes of things you've thought of a year or even a month ago?

Enter the next big project to feed your productive procrastination! Here's a file hierarchy comparison between my use of PARA and the Johnny.Decimal method -- as promised, one that's much more flexible and interchangeable (so "collaborations" could become a project, whereas "sewing and crochet patterns" could move from an area to the archive). The Johnny Decimal system, by contrast, is much more strict and easy to find: of course the all coursework and TA materials go under "teaching," and fellowship materials under "grant and fellowship apps." But which one has had longevity?


So, to be up front, I started out as a Johnny.Decimal evangelist and slowly moved towards PARA after reading the Second Brain book. Even in describing the method now it seems like the obviously better choice: interchangeable, helps you re-discover things you need, and just more flexible. But as time has gone on, I am realizing that my brain just doesn't work that way, so I've slowly migrated back to Johnny.Decimal with some modifications. This is to say that I might solely rely on PARA for note-taking (digital garden, PKM, whatever you want to call it), where I'll organize all of my markdown notes in the PARA framework, but I think the rest of my files need a much more rigid format so that I can actually find them. I used to also separate my work JD system from my personal JD, but I think it not makes more sense to have one cohesive system (I'm not sure that I can deal with the three-letter prefixes to differentiate between systems). I also don't want to remember two sets of numbers, where 15 could refer to apartment stuff in the personal JD whereas 15 might be sample syllabi. I just want to remember one set of numbers in the hopes that I can actually get to the mythical promised land of being able to recollect immediately: "ah. Taxes for 2023. That's 10.23."

I've started working through the Johnny Decimal workbook to rehaul the system a bit since my original iteration during graduate school, and I'm hoping that it allows me to build a system with more nuance. The thing that I've really come around to this time is the index, where you can keep track of things that are specific to a file (e.g., the number of the car repair person under "car maintenance," for example), so that it doesn't get lost in the ether. I think this is great as a reference note (drawing on the Zettelkasten system distinguishing between fleeting, reference, and permanent notes), but doesn't make room for keeping track of fleeting, literature, and permanent notes. This is where I think the PARA method really shines, but my brain simply cannot handle the constant moving-around between areas and resources, for example.

All this to say that it is very much a work in progress, and I'll update the blog with more insights after I've finished my (very slow) migration back to JD and have more thoughts about using JD alongside Obsidian beyond the index.

I hope this helps you think through how you approach file and note management. Happy organizing!

  1. Miriam Posner has a catalog of other methods and programs that you can use to manage your files, particularly research assets, in this gem of a blog post