Speculative Fiction Is Applied Civil Engineering

In my second year of engineering school we were made to design a footbridge for a load we would never see cross it: a company of soldiers marching in step, a scenario the actual bridge might face once in its working life if at all. You design for the worst plausible case, not the average one, and I did not know yet that this was the only lesson I would carry whole into fiction.

Every structural drawing I ever stamped had a number on it called the design load: the maximum weight, force, or vibration the structure was declared able to withstand, stated before a single beam was poured. Nobody discovers a bridge's design load by waiting for it to collapse. You declare it in advance, you build to satisfy it, and the declaration is what makes the bridge a piece of engineering rather than a large object that happens to be standing up today.

The declared load

A speculative world needs the same declaration, made before you draft the scene that will test it, not discovered afterward by seeing what the plot happened to survive. If your magic system, your government, your family, is going to face a maximum case somewhere in the book, decide now what that maximum case is and what the invented structure is rated for. A monarchy that has only ever been tested by a rival's polite objection has no declared load, and when you finally write the coup in chapter twenty, you will not know whether the monarchy was supposed to survive it, because you never decided what it was built to survive.

I use this literally when I plan a book. Before I draft a single chapter I write one sentence: this family, this institution, this enchantment, is rated to withstand X, and will fail above it. For a novel about a family running an illegal generator repair business under a corrupt licensing regime, my declared load was one clean sentence: the business survives any single inspector's demand for a bribe, and fails if two inspectors from different offices demand bribes in the same week, because the family's cash reserve only covers one. That sentence, written before chapter one, told me exactly what crisis the book was structurally required to deliver, and it arrives in chapter sixteen precisely because I had rated the business for it eleven chapters earlier.

Why this differs from ordinary outlining

Outlining tells you what happens next. A declared load tells you what a structure can survive, which is a claim about capacity rather than about sequence, and capacity claims generate their own sequence once you commit to testing them honestly. An outline can always be revised by adding an event. A declared load cannot be revised without admitting the structure was rated wrong from the start, which is a more useful kind of failure to catch, because it forces you to ask what would actually break this family, this government, this ward, rather than what would be dramatic to have happen to it.

The case against engineering fiction this way

Critics of this approach, and I have heard the argument from writers I respect, say that declaring a load in advance is exactly backwards, that fiction should discover its own necessity in the writing rather than being rated like a highway overpass, and that my method smuggles a plan into a process that is supposed to stay open to surprise. I take the objection seriously because I have watched over-declared drafts turn brittle, every scene arriving exactly on schedule to hit its rated failure point, with no room for a character to do something the structure had not anticipated. My answer is that the declared load is a floor, not a script. I still do not know how the family responds to the second bribe demand, only that the business as designed cannot absorb it. The surprise lives entirely in the response. The failure point itself I want decided, the same way an engineer wants to know a beam's limit before, not after, the marching soldiers arrive.

What the training actually gave me

I did not become a novelist despite my engineering degree. I became one because a structural drawing and a manuscript ask the identical question underneath their different vocabularies: what is this thing rated for, and what happens the day the load exceeds the rating. My professor never once used the word plot. He used the word failure, and meant it as a number you calculate in advance, not a surprise you write around. Every scene I have ever trusted was built on a rating I set months before I wrote it.