MotionMic. and the Thinking Stack
How I Built a Knowledge System That Gets Me Out From Behind My Desk — and How You Can Too | Part 1
“Your mind is for having ideas, not holding them.”
— David Allen, Getting Things Done
I recently wrote about losing a research note inside my own knowledge system — a note I had dictated while walking, months before it vanished during an AI-assisted migration.
That article explored the drift.
This one explores the design — the workflow I have been building that makes dictation-while-moving the foundation of a serious research practice, not an afterthought.
Because here is the thing nobody tells you about personal knowledge management: most of the advice assumes you are sitting at a desk. Tiago Forte‘s PARA method is brilliant for organizing what lands in your system. David Allen‘s Getting Things Done is indispensable for processing it once it’s there. But neither says much about where the raw material comes from in the first place — the moment of capture, the conditions under which your best thinking actually happens.
Mine happens in motion. Yours might too.
The Microphone Is Literal
I named this space MotionMic. because the concept is not a metaphor. My primary capture tool is a microphone — voice-to-text dictation on my phone while walking, driving, pacing, standing in line. The ideas that become articles, research insights, and manuscript sections rarely originate at a keyboard. They surface when my body is moving and my mind is free-associating without the pressure of a blinking cursor.
This is not productivity romanticism.
There is research on the relationship between walking and creative thinking — Stanford found that walking increases creative output by an average of 60 percent. But you do not need a study to confirm what you already know from experience: you think differently on your feet than you do in a chair.
The problem is that thinking on your feet produces messy output. A dictated note is not a polished note. It arrives with no structure, no links, no metadata, no filing convention. It lands in the Inbox with whatever title felt right in the moment — “Smith 2025 Report” or “that idea about drift and accountability” — and sits there until you do something with it. A mere note or memo to self.
Here’s a brief overview by Marily Oppezzo’s talk poignantly titled: Go for a walk:
MotionMic.™ is a methodology about designing the bridge between that raw, kinetic capture and a structured knowledge system that can actually use it.
The Stack
My system runs on Obsidian — a plain-text, local-first notebook where every note is a markdown file on my hard drive (and phone via it’s Sync feature).
Think of it as a personal wiki that you own completely, with no cloud dependency and no proprietary format. I chose it because my research data requires privacy, my workflow requires flexibility, and I do not trust any platform that could change its terms or shut down and take my decade of notes with it.
The folder structure draws from Forte’s PARA and Allen’s GTD, adapted into five containers:
Inbox — where everything lands first. Dictations from my phone, quick captures, forwarded articles, half-formed ideas. Nothing stays here permanently. The rule, borrowed from Allen: your inbox is a holding pen, not a filing cabinet.
Contexts — the ongoing domains of my life and work. My PhD dissertation, my consulting practice, my personal records. These are not projects with end dates — they are areas of responsibility that persist.
Projects — active, bounded work with deliverables. A manuscript in progress. A consulting engagement. A conference presentation. Each one has a defined scope and a finish line.
Archive — completed or inactive material. Old projects, retired drafts, reference data I am unlikely to need but do not want to lose.
Resources — reference material that serves multiple projects. My literature note library, idea notes, templates, people files.
The sixth element — the one that does not fit neatly into Forte’s four-letter acronym — is _METAVAULT: the system’s own configuration layer.
Templates, scripts, automation rules, documentation about the vault itself. The infrastructure that makes the content layer work for me, usually as I pace.
From Pocket to Page
Here is what a typical capture-to-output cycle looks like in practice.
I am walking. A thought connects two papers I read last week — something about how organizations describe their systems differently than how those systems actually function. I pull out my phone, open Obsidian mobile, and dictate a rough note.
It lands in the Inbox as drift and accountability connection.md. A heuristic slug for what I think in terms of work-as-done versus work-as-imagined.
Later — maybe that evening, maybe two days later — I sit down and process the Inbox. The dictation note gets cleaned up.
I add YAML frontmatter — metadata that tells the system what kind of note this is, what project it relates to, when it was created. I link it to the relevant literature notes in Resources.
If it connects to an active manuscript, I tag it with the project identifier. Here’s an example of what YAML looks like for a related literature note:
type: litnote
source: zotero
itemtype: bookSection
title: Why is work-as-imagined different from work-as-done?
author: Hollnagel, Wears, Hollnagel
year: "2017"
aliases:
- "[@Hollnagel2017b]"
zotero-key: Hollnagel2017b
date_created: 2017-01-01
project: MotionMic/blog-Thinking Stack
zotero-tags:
- HIERARCHY
- PERFORMANCE
- LEARNING
- KPI
- WAD
- WAI
- BUREAUCRACY
- MEANINGFUL
- MEASURE
date_modified: 2026-03-23T18:07The memo to self has now moved from raw capture to situated knowledge.
It exists in relationship to other notes. It can be found not just by searching for keywords but by following links, querying metadata, and visualizing connections in Obsidian’s graph view:
This cycle — dictate in motion, process at the desk, connect to the network — is the heartbeat of the MotionMic.™ approach. The motion produces the raw material. The desk work refines it. Neither is sufficient alone.
The Time-Productivity Trap
There is a well-known tension in the Obsidian community and the broader PKM space between working IN your system and working ON it. Working in your system means writing, thinking, connecting ideas — the actual knowledge work. Working on your system means configuring plugins, redesigning templates, reorganizing folders, debugging automation.
The second mode is seductive. It feels productive. You are improving the tool, optimizing the workflow, building infrastructure. But if you spend more time polishing the workshop than using it, you have a hobby, not a system.
I track my time now.
Not obsessively — but enough to know the ratio. A good week is fifteen hours in, two hours on. A week where those numbers invert is a signal that I have been tinkering instead of thinking.
The MotionMic.™ methodology is partly a discipline against this trap. If the system’s primary input is dictation-while-moving, then the system must be simple enough that the desk processing step does not expand to fill all available time. Capture should be frictionless. Processing should be systematic. And the system itself should fade into the background, not demand constant attention.
What This Series Covers
The introductory article on Medium told the story of what happens when this system drifts — when the infrastructure degrades silently and a missing note exposes the gap between design and reality.
This piece — Part 1 — is a peak into the big picture architectural design itself.
How the workflow is structured, why motion is the starting point, and what the MotionMic.™ approach looks like in practice.
Part 2, “Sitting Still to Start Moving,” sits with the paradox I am living right now — that building a mobile-first workflow required the most stationary day of work I have had in months. It pulls on David Allen, Tiago Forte, and the philosophy of learning by deliberately breaking things.
Part 3, “The Sandbox and the Safety Net,” goes into the security architecture — how I use AI tools like Claude Code inside this system. You will get the actual prompts, scripts, and configuration files.
If you have ever wanted to let an AI assistant help manage your files but did not trust it enough to try, that piece gives you some guardrails.
The system is not perfect. It drifted once already, and I had to rebuild parts of it from the digital equivalent of a recycling bin.
But the core logic holds: think while moving, capture with your voice, process at the desk, connect to the network.
The system exists to serve the thinking — and the thinking happens best when you close the laptop and go for a walk.
So if you’ll excuse me, I have a note to dictate.