Posts ⁂ The Postmodern Terminal >_
How a signaling system from Toyota's assembly lines became a workable method for finishing an 80,000-word dissertation one card at a time.
Kanban began on Toyota's factory floor in the 1940s, where a card (kanban in Japanese) signaled that a workstation needed more material. Software teams borrowed it decades later, and individual writers now borrow it again for projects too large to hold in the head at once. I will argue that kanban works for dissertation and article writing because of two practices, chunking and limiting work in progress, and I will show what each one does to a project that otherwise refuses to move.
The method reduces to one idea. Take a board with a small number of columns, commonly something like Not Started, Drafting, Revising, and Done, and represent each unit of work as a card that travels left to right as it progresses. Chunking, which means breaking a large project into visible units, is the whole discipline underneath.
An 80,000-word dissertation is not a task. You cannot act on it; you can only feel guilty about it. A board holding forty specific cards, each one a subsection, a data chapter, or a round of feedback, is something you can work through one card at a time. The board converts a mood into a queue.
For dissertation writing, cards work best at the level of chapter sections rather than whole chapters. A card that says "Chapter 3" can sit in the same column for three months while telling you nothing at all, which is worse than no card, because it looks like information.
Sharper cards read like this: "Chapter 3: code the interview transcripts," "Chapter 3: draft the findings narrative," "Chapter 2: revise after advisor comments." Each one advances on its own timeline. A glance at the board then shows which chapters are stalled in revision and which have barely started, instead of leaving you with the vague and useless sense that "the dissertation is coming along."
For journal articles, the columns map onto the submission pipeline directly: Drafting, Internal Review, Submitted, Under Peer Review, Revise and Resubmit, Published. One card per article lets you hold several submissions at once without losing track of which one is waiting on you and which is waiting on an editor. Keeping a target journal in view from early in the writing matters here; submission is a design constraint on the draft, not an errand you run once the draft is done.
The second practice is the one people skip, and it is the one that does the work. Kanban caps how many cards may sit in an "in progress" column at any moment, often at one or two.
The constraint is deliberate and it should feel slightly uncomfortable. Writers who let five sections sit half finished tend to finish none of them soon. Writers who complete one card before pulling the next accumulate finished work steadily, and finished work is the only kind that counts toward a defense. A work-in-progress limit turns "I am working on my dissertation" into "I am finishing this one section, right now," and the second sentence is the only one you can act on before lunch.
You need no special software. Index cards on a wall, or a whiteboard divided into columns, work exactly as well as an app. If you would rather keep it digital, Notion and Miro both include kanban board views, and dedicated tools such as Trello follow the same model.
The board itself is not the method. The habit it enforces is: name the next small piece of work, finish that piece, move the card. Everything else, the colors, the automations, the elaborate column schemes, is procrastination wearing a lanyard.
Notion, workspace with built-in kanban views
Miro, collaborative whiteboard for boards and clustering
Zotero, the reference manager to pair with the board
The Postmodern Terminal >_