No single method is required. Working novelists split into what George R.R. Martin has called architects and gardeners — planners who blueprint the plot before drafting, and discoverers who follow the story scene by scene, adjusting as it grows. Either produces finished books; the difference is where the planning happens.
The split shows up early, before genre or ambition enters into it. Some writers can't set down a first sentence without knowing how the book ends; others can't move at all until they've stopped needing to know. Neither instinct is a defect, and neither is a shortcut around the other's problems: an outline still has to survive contact with an actual scene, and a discovery draft still has to add up to a shape a reader can follow. What matters for a working draft is recognizing which instinct is doing the driving, then building a process around it instead of borrowing someone else's, chapter by chapter, until the mismatch between method and temperament is what's actually stalling the book.
What Do Architects and Gardeners Actually Mean?
Martin coined the split in a 2022 post on his own blog: architects draft blueprints before laying down a sentence, while gardeners plant an idea and see what grows, discovering the story's shape as they dig rather than mapping it first. He has said he belongs to the second group, even across the decades he spent on his A Song of Ice and Fire novels, where the broad destinations were set early but the route between them kept changing.
Martin described himself as someone who generally knows a story's ending years in advance — sometimes decades — but not the path there. "My stories grow and evolve and change as I write them," he wrote, adding that even when the destination is fixed, "sometimes the ground changes under my feet as the words pour forth." The distinction he draws isn't between writers who plan and writers who don't; it's between planning the destination and planning the route. A gardener, in his framing, still knows roughly where a book is headed. What's undetermined is everything in between — which characters survive a scene, which subplot turns out to matter, which chapter gets cut once the shape of the finished book becomes clear. An architect front-loads those decisions; a gardener defers them until the draft has enough material to decide with.
How Does a Discovery Writer Work Without an Outline?
Stephen King has described writing without an outline as following the book rather than steering it — letting a situation's own logic surface the next scene, then adjusting for whatever contradictions or dead ends that method turns up along the way, rather than solving for them on paper first.
In a 2006 interview with The Paris Review, King said his novels usually start from a situation rather than a plan — Cujo, for instance, began with the idea of a big, friendly dog and a car that breaks down on a lonely road, with no outline behind either detail. "Once you've got that much, you start to see all the ramifications of the story," he said. From there, he added, the material takes over: "The book is the boss. You've got to let the book go where it wants to go, and you just follow along." Trying to force a direction, he said, backfires the way a pitcher trying to steer a fastball does — "you screw everything up."
That method has a practical cost, and King doesn't gloss over it. Partway through Cujo, he discovered that the real incubation period for rabies was too long to fit the timeline his story needed, and had to compress events to make the plot work. "You can never bend reality to serve the fiction," he said — a fix that research done before drafting might have avoided, and a fair example of the trade a discovery writer makes: less planning up front, more repair work once the facts of the story catch up with it.
Working this way, by King's account, comes down to a few habits rather than a single technique:
- Start from a situation, not a synopsis — a set of circumstances rather than a plotted sequence of events.
- Follow the questions the situation raises, rather than answering them in advance on an outline.
- Let facts and contradictions reshape the plot after they surface in the draft, instead of researching everything before the first page.
- Treat a scene that resists the plan as information, not failure — the resistance is usually telling you what the story actually needs next.
What Does a Planned, Architect-Style Draft Look Like?
Novelist Stephanie Gangi took the opposite route while drafting her first novel, building a written project plan before she drafted a page — part outline, part accountability system, part morale device for a book with no external deadline pushing it forward.
In a 2021 essay for Literary Hub, Gangi described assembling the plan piece by piece: "I made a project plan (my map to the future); I bought craft books (I had no idea how to write a novel); I signed up for a writing workshop (for accountability)." Her outline for the story's arc was deliberately simple, closer to a question a reader might ask than a formal beat sheet: "I wrote the arc this way: 1. This happens; 2. And then that happens; 3. Now what happens?" She filled notebook pages with individual "Happenings," then shuffled and reordered the pages until they settled into scenes and chapters — a process closer to sorting note cards on a table than filling in a template.
Gangi's plan also included a faux New York Times review written as a mission statement, and a visible schedule of milestones posted where she'd see it — both meant to keep her moving without an agent, an editor, or a deadline holding her to the work. But she's clear that none of it was fixed in place once the drafting started: the plan's elements "expanded or contracted or disappeared entirely, depending on how useful they proved to be." The plan's job, in other words, wasn't to be followed exactly. It was to give a first-time novelist something concrete to revise against, instead of a blank page and a vague sense of where the book should go.
So Which Approach Should You Use?
Neither camp is the safer bet — Martin, King, and Gangi have all finished and published novels using opposite methods, and none of the three describes the other's approach as the reason a book succeeds or stalls. The more useful question is which approach gets a full draft onto the page for you, since an abandoned outline and a stalled discovery draft cost a writer the same thing in the end: momentum.
A workable test is smaller than a whole novel. Try both methods on a single scene before committing either one to an entire manuscript: sketch it in three lines first, the way Gangi's "this happens, then this happens, now what happens" method suggests, and write the scene from that sketch. Separately, on a different day, write the same scene cold and see what it becomes without a plan behind it. The version that produces pages worth revising is the answer for that project, not a permanent identity. Martin's own account of the ground shifting under a plan whose ending he already knew suggests most working novelists borrow from both approaches depending on what a given book, or even a given chapter, turns out to demand.
None of this is a verdict on talent or seriousness. A detailed outline doesn't make a novel more disciplined, and a discovery draft doesn't make it more inspired — they're just two different places to put the labor of figuring out what happens next. The three writers in this piece put that labor in different spots and finished books either way, which is closer to the actual lesson than any rule about which method is correct.