Writing kata

One of my perennial issues is getting started. My mind is full of ideas that are like a tangled ball of yarn, and pulling on one idea doesn't result in a single thread coming loose but rather a knotty mess that points to a dozen other things and doesn't know where it wants to go.

If you view the web of knowledge (beliefs, opinions,) one person holds as a system of nodes and edges, then in theory the process of elucidating the graph should be a matter of repeatedly identifying two nodes and an edge, and producing the link between them.

In practice, surfacing the nodes and edges is non-trivial. A node might not be a node, but rather a sub-graph, and by the time you pare it down to the minimum, recursively breaking subgraph-nodes into nodes and edges, you find that a single elucidated edge sounds tautological.

So the process is less "nodes and edges" and more "small walks through constrained subgraphs." Less graph theory and more topology; let's define a neighborhood that is within epsilon of this idea, and explore that idea fully.

Writing kata^[1] are an attempt to unstick this process.

Each working day, I look through my notes and my mind to see what ideas appeal to me, which ones lend themselves to a passionate attack, and constrain my output to no more than 30 minutes effort. If I don't hit a target after 30 minutes, that's useful practice; if I hit a target more quickly than that, so be it. It's not about the time committed but rather about the action performed. It's about a simple, quantized, act of practice that should establish in my brain the habit of identifying a point and getting it out without second guessing, without hesitation.

The goal here is not only to practice skills, but also to get the first draft version of my entire knowledge graph into the world, and thence to get me the opportunity to see it all in one place, and try to figure out what version 2 looks like.

I observe that my favorite blog posts are generally 500–800 words, with few or no subheaders, and so this is my target as well. I tend to ramble up to 3000 or 5000 or more words, and in the absence of an established outline and a clear thesis, this kind of post works against my goals: my default voice is so dry that even I don't want to re-read some of my more logical but dry work.

Today is day 2; if I do this practice every workday for the next year, I should produce between 1/8 and 1/5 of a million words. And at that point, if I can't find anything to say, if I can't identify the core ideas that motivate me, then something will have gone really wrong.