Reading Bob Doto’s recent book on his note making and writing practices, he explained how he uses his Luhmann-style alphanumeric numbering of notes as a way to spot where a richer grouping is emerging as a possible starting point for his writing. It made me ask, how to spot similar patterns in my digital notes without that numbering system?

Alphanumeric numbering

The alphanumeric numbering basically works like this: When you start your notes collection, you number the notes from the very first one. Say you start with your first two notes on two different topics, A and B. For Topic A, the first note you give the number 1. Topic B’s first note 2. Then for each new note, you decide where to place it in the existing collection. If you think it’s connected to Topic A, it may become 1a. A few notes down the line on other topics you may make a note you place in 1b. Then some time later you have a note you want to place directly after 1a. but not after 1b, so you number it 1a1, etc. Key here is that the numbering system doesn’t give a premade structure where you need to slot a new note in an existing hole (unlike library numbering systems). The numbering system merely means you must choose a first place where you place a new note, intentionally making a first connection to an existing note and number it accordingly. For physical note making systems on index cards this is I think basically the only way to do it, if you don’t want to use a predefined structure (and you don’t because structure is emergent from working with your notes, a knowledge output, not an administrative tool), and do want a navigable and explorable system.

Alphanumeric numbering to detect growth in your system

At the other end of his PKM system, where Doto creates writings using the material he collected over time, the alphanumeric system helps him in finding potential things to write about. Topics where there are lots of notes, accumulated over some period of time, will show up as denser parts of his alphanumeric system. Say in Topic B’s number 2 branch above, a year later there are just 1 or 2 new notes. 2.1 and 2.2. That is a signal Topic B wasn’t an interest that gained any depth that year. But say that note number 1a is now a grouping of a dozen notes, 1a1-1 and a further forking set below it, and 1a2 has some, and there’s 1a3 with no further notes attached. Clearly something happened there in the intervening time. Your attention attached a range of notes to that first starting point 1a, and perhaps it means you have something to write there to bring those notes together. It’s like a grape bunch on a vine. Vine 2 hasn’t grown any grapes let alone any bunches of them, but Vine 1a has. And there may be other grape bunches, bigger, smaller, elsewhere. You choose a bunch to make some wine, i.e. inspecting and reflecting on that grouping of notes and then writing.

On not having alphanumeric numbering in my system

My notes are not numbered that way, although they do contain a timestamp. They’re not numbered because I have a digital notes collection, in which it is easy to make links between notes. I add the time stamp both for unique titles (‘Some good summarising title 20241018125808’ is different from ‘Some good summarising title 20020424125820’ despite the same title text) and to glance from the title from which period in my activities something comes. My notes always have a link to another already existing note, like the alphanumeric system, I always intentionally choose a first connection. Without it, it is not really a navigable and explorable system. However, over time or at the time of first writing, I may add additional links from that note to others. It is not always visible which link was the first link, the equivalent of choosing the first position in the alphanumeric system.

On the output side this means I cannot easily spot where the density of my notes has grown to a budding grape bunch. I can see which notes are most heavily connected, using the Obsidian graph, but that’s equivalent to pointing a lot to a main branch, and the more I point to a single note the less meaningful it is, as it becomes more of a generic category essentially.

So what are my options to detect emerging dense spots in my notes?

The alphanumeric grape bunch is basically a group of one or more short or longer lineages originating from a single point somewhere in the total.
Those lineages are present in my notes too, but not easy to spot. Sometimes I make deliberate chains (Note B linking to more abstract concept A, and to more tangible example C, making an A-B-C train of thought), but not often. Lineages would be easier to spot if links in notes had metadata, like the time of linking. This would both show in a note the first link (the first place it was put in), and allow across notes the exploration of the sequence in which notes x, y and z got connected by looking at note creation data and incoming linkages. Links are information objects in their own right, and have different aspects like direction(s), character, intensity, a time dimension and a versioning history, none of which is captured in my system (nor in those of others I saw). Links can also grow into new notes. A link between A and B over time and through reflection deepening in meaning, necessitating its own note to express it with enough atomicity.

Might local graphs help?

The local graph of a note in my Obsidian tool can provide a bit of lineage, by looking at the vicinity of a single note limited to incoming links and set to a depth of 2 or higher. It’s not visually obvious though, it doesn’t jump out. In the following graph (incoming links, depth 2, with ‘Links as information object’ as the branching off point) I colored the different branches. Note that the title for each node/note has a timestamp showing how much time there was between one note and the following added to the lineage. Some notes have the same date, meaning they were written in the more or less the same sitting. Also note that some older notes link to some newer notes, indicating I edited the note to include a new link to a newer note)

What also emerges from this image is that several ‘lineages’ connect further ‘down’. Essentially they create a loop, a circle of notes. There are many more, and I think they are a sign of density (one that alphanumeric numbering doesn’t have), yet like lineages not easy to spot. Two examples are shown below. One circle based on a depth of 2, another with depth 3.

While this type of visualisation isn’t useless, it’s also not obviously useful in this case. Would there be other ways to search for things like “lines longer than 3 notes, with more than 1 branching off point with at least 2 notes depth”.

Do you have ways to spot emerging clusters outside of alphanumeric numbering and graphs in your notes? As a suggestion for your potential next writing?