Aliette de Bodard
Writers are often placed on a scale between pantsers and plotters: between those who start with an idea and then let the story accrete as they write it, and those who like to plan ahead. I'm most definitely an architect: I prefer to plan ahead as much as possible, and to always have an outline to refer to.
There's no one superior path: you genuinely have to find what works for you and your current project. (Your own writing process is going to evolve and change as time goes on!) With that in mind though, I thought I'd share some of the things that have worked for me when putting together a project. The advice I give here is length-agnostic, but I'm going to talk particularly about novels, since these are the pieces that are in the most need of some kind of structure.
The first step towards a completed novel is getting the idea. The idea doesn't need to be resoundingly original: some subgenres (hard extrapolatory SF, for instance) will attach more weight to this, but what I've found is that a lot of the value lies not in the idea, but in the way that you develop it and breathe life into it. I've also found that one idea in and of itself will seldom carry a story: what I usually need is two or so for a short story, and then an intersection of a growing number, depending on the length of a piece. A typical novel for me will merge three or four ideas, if not more. For my novel The House of Binding Thorns, I smashed together "Paris in the wake of a devastating magical war" and "the Opium Wars with magical drugs", added in a bit of "arranged queer marriage", and sprinkled it with "difficult pregnancy with political ramifications" in order to get the plot started.
Second for me is the research. I find that in genre I can get a lot of worldbuilding help by leaning on historical parallels. I don't copy from history because by and large I don't write historical fiction, but researching historical precedents allows me to have a wealth of details and events to draw on, rather than merely making up things myself. (Experience has shown that when I'm left to my own devices to make things up, I will fall back on media clichés and ingrained prejudices, neither of which are good for writing!) So, for The House of Binding Thorns, I researched daily life in the grand mansions and the poor neighbourhoods of nineteenth-century Paris; the French incursions into pre-colonisation nineteenth-century Vietnam, and of course medical practises and medical issues surrounding pregnancy. (My number one rule for plot: "what's the worst that can happen to this character? Then it needs to happen at some point.")
A fragment of the mindmap for The House of Binding Thorns
When I've done all this, I can start to brainstorm. I've found mindmaps very useful. Scapple and other mindmapping software do them very well, but for me nothing beats a piece of paper and a pen. There's something very freeing about simply going from idea to idea and drawing lines between them: it enables me to forge connections I wouldn't have found otherwise.
Free writing is also quite useful. I'll simply jot down my thoughts in a word document and work my way through a problem by essentially dissecting it in text. My draft documents for The House of Binding Thorns included such gems as "this is just stupid, it can't work that way" and "let's try making A the lynchpin of the pregnancy plotline. Hmm. Maybe not".
Another tool I've used before is index cards. I'll write down particular scenes or key plot points as they occur to me, then clear a large table area and set all of them on there, and try and link them together in order to get a plot.
The goal of any plotting for me is hammering out a rough shape of the book, a skeleton to hang events onto. Around the 25-30% mark, some kind of game-changer has to happen, some kind of large event that has repercussions and highly impacts the plot. And 75% of the way into the book, I'll look at heading into a climax--which means some convergence of threads, some quickening of the pace (whether it's a final battle or a high emotional point or both--really depends on the character arcs!). In The House of Binding Thorns, one character gets kidnapped around the 25% mark, which in turn sets a host of other events into motion (rescuing them, finding out who was responsible, and the political consequences of another character getting wounded in the kidnapping). Theoretically, at 50% I should also be looking at the same kind of game-changing event, but I have to confess the midway point of the plot always ends up taking care of itself in all my projects!
Once I have the skeleton in place, I can draw lines again: basically, I know I have to get from point A to point B, and I can then work out how to do that--drawing the consequences of A, working out what kind of pre-conditions I need for B to happen, etc. Then, if needed, I can go back and tweak A and B as much as needed.
I develop my plots into a lot of details, which I know isn't necessarily usual: for a lot of writers, just having the skeleton (the broad idea of where the book is headed) is enough. I, on the other hand, turn this into a list of scenes. Focusing on the small details helps me know how the scenes fit together best and also gives me better control over the length of the book. (Quick rule for me is 1 scene = 2000 words on average.) Lest this make me sound like a control freak, I'll point out the detailed scene by scene outline only extends so far: the further we are from the beginning, the fewer details I'll have. The last three chapters of The House of Binding Thorns in my original outline were simply "climax, the characters somehow beat the threat" (gotta love that "somehow"!).
When I start writing, it's inevitable that the outline won't survive fire. I'll go one way, discovering things as I write; the outline will remain pointing the original way. So, regularly, I'll update the outline. I'll use brainstorming and mindmapping and free writing again. I was, in fact, brainstorming the ending of The House of Binding Thorns as little as two weeks before actually finishing the first draft of the book, and retconning bits and pieces of the draft in order to make what I had on the page fit what I wanted to do next!
So there you go: this is how I write a piece. Hopefully there is something in here that will be of use to you--whether it's research or mindmaps or something else altogether. Oh, and lest I forget, here are my cardinal rules of writing: keep improving, never give up.[GdM]