Your Invented World Needs a Maintenance Schedule

Before I wrote sentences for a living I stamped drawings for a living, and every drawing I stamped for a drainage culvert came with a second document nobody outside the ministry ever read: the maintenance schedule, listing who clears the culvert, how often, and what the fine is if the contract lapses.

Nobody builds a bridge and calls it finished. A bridge that is finished on the day it opens is a bridge that will fail on a schedule an engineer can already calculate, because paint fails, bolts loosen under load cycling, and silt collects at the footings whether anyone is watching or not. The maintenance schedule is not an afterthought to the design. It is the part of the design that admits time is going to keep happening after the ribbon is cut.

What the maintenance ledger is

I keep one page per invented structure of consequence in a draft: a wall, a canal, an enchantment, an institution. The page has three columns. What decays, on what interval, and who is contracted to arrest the decay. A ward around a village decays if nobody renews it. A council's authority decays if nobody enforces its rulings. A canal silts up if nobody dredges it every dry season. None of this needs to appear in the prose. It needs to exist so that when I write a scene set inside that structure, I already know its current state of repair, rather than defaulting to the lazy assumption that everything invented is eternally in the condition I first described it.

The ledger changes how a scene opens. A wall I had described once, gloriously, in chapter two, appears again in chapter nine, and because I know its dredging interval is every three years and nobody has dredged it in five, I do not have to invent a new crisis to make the wall matter again. The wall is already failing on its own schedule, quietly, the way real walls do, and the plot simply has to arrive when the failure becomes visible instead of manufacturing an attack to explain why the wall is suddenly relevant.

A worked example

A council of elders in a manuscript of mine had, in an early draft, unlimited authority that never needed defending, page after page, until a rebellion in chapter fourteen that felt bolted on because nothing had been eroding beforehand. I built a ledger for the council: their authority decays whenever a ruling goes unenforced for one full season, and the enforcement mechanism is a debt registry that requires literate clerks, of which the council has exactly four, all elderly. Once I knew that, I did not need the rebellion to arrive from outside. I went back to chapter six and had one clerk die of ordinary illness, leaving three, and a backlog of unenforced rulings begins accumulating from that point, visible to anyone counting, which the rebels in chapter fourteen turn out to have been doing the entire time. The rebellion stopped being an event and became an arithmetic that had been running since chapter six.

The case against this

The case against maintaining a ledger for every invented structure is that it is a great deal of labour for a novel that might never return to most of these structures, and that a writer who spends an afternoon calculating a council's clerk shortage instead of drafting is, on that afternoon, not writing. I do not dispute the arithmetic of time. What I dispute is the assumption that this labour competes with drafting rather than replacing a worse kind of labour, the kind where you reach chapter fourteen, realise the rebellion has no cause, and spend three revision passes trying to plant one retroactively through the whole manuscript. The ledger is cheaper than that repair, but it is only cheaper if you build it for structures the plot will actually revisit, which is why I build it only for the two or three institutions or objects a story returns to more than once, never for the whole invented world at once.

What decay actually gives a writer

A world with no maintenance schedule can only be interrupted by events, because nothing in it changes on its own, and events always look, on close reading, like the author's hand reaching in. A world with a maintenance schedule changes on its own, at a rate you set once and can calculate forward, which means your plot can be the thing that notices the change rather than the thing that causes it. The culverts I used to sign off on failed, on average, at year eleven, if nobody cleared them, and the ministry knew that number before a single one had failed, because failure that follows a schedule is not a crisis. It is an appointment nobody kept.