A taxonomy of notes
(h/t to Andy Matuschak, Conor White-Sullivan)
HUGE DISCLAIMER — as of Today’s Date Mon, August 10th, 2020, I can confidently declare that;
ALMOST THE ENTIRETY OF THIS ESSAY HAS BEEN RENDERED OBSOLETE.
My use of Roam Research has pretty much UTTERLY changed since I wrote this … and that’s a good thing!
At some point I’m going to come back to this, and do cleanup.
It’s literally here only for its archival purpose. I have left it only because, one day, I want to make the contrast clear between “What I was trying to do THEN” versus “What I am trying to do NOW”.
When v.2 of this comes out, as it will eventually, I will delete this — for now it’s here for archival purposes… that’s my caveat. Read if you want, but take everything with a proverbial grain of salt because most of it, I now have replaced with a completely different workflow, which I will write about at some future, as yet undetermined, date.
— — — — — — — — — — — — — — — — — — — — — — — — —
(original article follows, written in early 2020)
I’m trying an experiment in Roam, and I’m excited to share it, because it’s changing the way I take notes.
It started with reading something that Andy Matuschak wrote, about having a “taxonomy of notes”.
He talks about the different kinds of notes you can take, with the idea being that you slowly build upwards towards more and more structure, as you think about any given topic. (I highly recommend looking at what he has to say on this topic himself, in his online notebook, which he recently made publicly visible in an act of wonderful generosity — see below — and thanks, Andy!).
(Warning: it’s a pretty magical place, and you probably will want to explore this notebook in more detail.. so set aside some time if you can… it’s worth it.)
Andy’s notebook entry
(P.S. — you can click within this rectangle, to view it in a separate tab):
After reading this, I put together some tags that I can use in Roam, inspired by Andy’s taxonomy.
I am now using “[[🌲1]], [[🌲2]], [[🌲3]], and [[🌲4]]” as hashtags, whenever I create a new note.
This is shorthand for what I am calling “Evergreen note - level 1, level 2, level 3, level 4”.
(speed tip: I use the “textblaze” browser extension to make hotkey combos to type those tags … that way, a short keystroke combo generates each of those complete hashtags — complete with square brackets and “evergreen emoji”l — in just a few keystrokes, whenever I want the tag.)
New pages that I write, now have an attribute line — (tags::) — right beneath the page title, and on that line, I use one of those 4 tags to tell me which “class” of evergreen note that page is.
There is a hierarchy here, and you can read about it in Andy’s words in his notebook (above), or you can let me give you my version , below— or both.
Below, I’m going to illustrate the concept using a real-life project that I just started: I am organizing my knowledge about the physiology of sleep.
(P.S. In addition to creating these tags, I created a “process page” or “meta-page” about the whole concept; my process page looks like this:)
- “Level one” evergreen note - aka [[🌲1]]:
Andy calls this a “stub” — I also like to think of it as a “placeholder” or even a “cognitive hail Mary pass”.
This is where you are free-writing and you suddenly realize that something *might* come in handy later as a conceptual placeholder, so you go ahead and make it into a topic, without really even knowing what will connect with that topic.
Often, you just have a hunch that something will be useful. So you set aside some space for it, and let the details arrive later.
Example:
I’m researching sleep — and the first thought I have about it is that it is a “regenerative process”. So on the spot I create a new page called [[regenerative processes]]. That page is an example of a [[🌲1] page.
It’s a stub, and it’s likely to broadly associate with lots of other topics… in the long run, it will be a place where I can associate with other regenerative processes of all kinds, where I can examine the biology of equilibria and homeostasis, dynamic equilibrium, entropy in living systems, and whatever other big topics I might at one point want to connect all this to. But for now, it’s a place-holder, but it’s one that, I know, will grow, the more I write.
Broadly speaking, this is a concept-level page. Think of it as the view from the sky.
2. “Level two” evergreen note -aka [[🌲2]] —
This is the level of neatly defined concepts, or as Andy writes, “terms of art” — that is, the concept-knowledge building blocks that you need in order to start thinking about a complex topic.
In my sleep example, [[Sleep Phases — NREM and REM]], a page in which I map out my definitions, terms, and foundational knowledge — that would be a “level 2” note.
It is a place where I can elaborate on what sleep phases are, because other, more complicated building blocks, are later going to arise from this foundational concept.
I will, I imagine, at some point be elaborating on more advanced hypotheses about the phases of sleep, but I need to define some terms, first, in order to elaborate those hypotheses. These are my level 2 notes.
3. “Level 3” evergreen note (aka [[🌲3]]) —
This is where, I think, things really start to get interesting!
What Andy suggests is that at this level, the notes have titles that are, essentially, short, declarative statements, that can be supported by what you have written or cited elsewhere.
(From my perspective, it’s at level 3 that your own ideas first occur, because if level 1 was about defining areas of the house that you were going to fill, and level 2 was about defining the objects you were going to fill them with, then, at level 3, for the first time, you are working on describing and declaring relationships between the objects in your mental house — or your “magic junkyard”, to borrow a descriptor I first read used by Anne-Laure Le Cunff)
As I build my sleep notes little by little, I might eventually be ready to title a page something like this: [[Sleep is a process-state, during which vital neurophysiological housekeeping processes occur]].
Andy goes on to enumerate some of the keys to writing a good “Level 3” note — that such notes should be “precise” and “narrow”; (in that sense, they function much like the process of questioning by a skilled lawyer in a legal case, where the goal is to be really careful and incremental, carefully making only such claims as you can back up with your work elsewhere.)
Note that a claim can have broad implications, yet still ontologically be “precise” and “narrow” — as is the case with my example… for example “All men are mortal” is a very precise claim — albeit, one with significant implication…
4. “Level 4” Evergreen Note [[🌲4]]—
These notes are the “meta” notes, that start to build on ideas you have developed across multiple lower-levels of notes. Andy compares these notes to “API”s, in that they then become useful “handles” for developing ideas at the highest levels of complexity.
A level 4 note is built on top of one or more level 3 notes.
Here I am just going to quote Andy directly, because his description of this highest-level of note is succinct and clear:
(once again, note that if you want to go directly to Andy’s notes, you can click on his underlined terms below)
When Evergreen notes are factored and titled well, those titles become an abstraction for the note itself. The entire note’s ideas can then be referenced using that handle.
In fact, this property itself functions as a kind of litmus: as you develop ideas in notes over time, and improve the “APIs,” you’ll be able to write individual notes which abstract over increasingly large subtrees (e.g. “Enacted experiences have incredible potential as a mass medium”, “Evergreen note-writing as fundamental unit of knowledge work”).
Some effective note “API design” techniques:
separation of concerns (Evergreen notes should be atomic),
sharp titles (Prefer declarative and imperative note titles to sharpen claims), and
positive framings (Prefer positive note titles to promote systematic theory).
For example, in my own writing about sleep, I might find that [[Sleep is a process-state, during which vital neurophysiological housekeeping processes occur]] (… that was my level 3 example…)
Later I might develop more “level 3” notes, to assert that, e.g.:
[[Dreaming occurs predominantly within the REM phase of sleep]] … and [[REM sleep occurs at the end of each sleep cycle]] …
and finally I think I can synthesize these statements into an exploratory “level 4” statement, which could be [[Hypothesis: The physiologic housekeeping in stages NREM 1–3 includes processes that “prime” the brain to dream in REM, the final stage of the cycle]]
And that, I think, is “level 4”… in which I’ve linked some ideas from my existing level 3 notes, in order to form a new hypothesis, one that is atomic, separate, and positive, and built out of what precedes it.
Conclusions:
(i.e. why does this matter, and why should you bother with the hashtags?)
I started this experiment in order to teach myself to be more deliberate, about the kinds of pages I was creating in my Roam notebook. The simple act of hesitating and thinking about “What level-of-page is this?” has encouraged me to be more thoughtful, in general, about how I organize my knowledge, and more specifically, about what kinds of pages I am creating, and why.
In that sense, you could think of the [[🌲1]] — [[🌲2]] — [[🌲3]] — [[🌲4]] tags as “training wheels for the mind”… I’m not necessarily going to use them forever, as I expect that at some point the “sorting” of pages will become instinctual for me, and then (maybe) my use will fade.
However, there is another value, too — I can search for those hashtags and thereby get a “level-by-level” view of my own notebook-workspace — summarizing at a glance where my broad concepts are, where my “terms of art” are, where my “building blocks” are, and where my “compound building blocks” have taken me.
If this has been useful to you, please let me know! Cheers, and happy note-taking:
Comments are welcome.