The index card method: outlining a novel one scene at a time

When Vladimir Nabokov died in 1977, he left behind an unfinished novel, The Original of Laura, written in pencil on 138 index cards. It was eventually published in 2009 as a facsimile, with each card reproduced and perforated so readers could rearrange them, just as Nabokov had.

Nabokov worked on cards for most of his career, including on Lolita and Pale Fire. He didn't write from beginning to end; he wrote whatever passage was clearest in his mind, on a card, and slotted it into place. In interviews he described the pattern of a book as coming before the book itself, and the writing as filling in the gaps of a crossword at whatever spot he happened to choose.

You don't need to be Nabokov to borrow the idea. Index cards turn an outline into a set of movable parts, and the scene turns out to be exactly the right size for a part.

Why the scene, specifically?

A chapter is too big to be a unit of planning: it often contains several scenes, and chapters can be split or merged late. A sentence is too small. But a scene, one continuous stretch of action in one place and time, is the smallest unit in which something changes. Something is wanted, something gets in the way, and the situation at the end is different from the start.

That makes the scene the atom of story. Outline in scenes and you can shuffle, cut and combine without breaking anything.

What goes on a card

The single biggest mistake with scene cards is writing only what happens: "Mara visits the archive." That tells you the event but not whether the scene works. A better card answers three questions, adapted from Dwight V. Swain's classic Techniques of the Selling Writer:

• Goal. What does the point-of-view character want in this scene?
• Conflict. What stands in the way?
• Outcome. How does it end, usually worse for the character than they hoped?

So: Mara goes to the archive to find her grandfather's service record (goal). The archivist claims the file was destroyed in a flood (conflict). She finds the file has been checked out, by her own mother, three days ago (outcome: disaster).

That card tells you whether the scene has a pulse. If you can't fill in all three, you've found a scene that may be decoration rather than story.

Scene and sequel

Swain also described a second kind of unit, which he called the sequel: the aftermath of a scene, in which the character reacts emotionally, faces a dilemma, and reaches a decision that sets up the next goal.

Not every novel alternates strictly between the two, but it's a useful lens when you lay your cards out. If you see five action scenes in a row with no sequels, the book may feel breathless, with characters never processing what's happening to them. Five sequels in a row, and it probably feels stalled.

Some writers use two colours of card for this, one for scenes and one for sequels, so the rhythm is visible from across the room.

The shuffle test

Once all your cards exist, lay them out in order and try this: pick up any card and ask, "If I removed this, what would break?"

• If the answer is "nothing," cut the card, or combine it with a neighbour.
• If the answer is "the reader wouldn't learn X," consider whether X could be delivered in a scene that's already doing something else.
• If the answer is "the next five scenes would make no sense," you've found a load-bearing scene. Make sure it's strong.

Then try moving cards. Swap two scenes and see if the story improves. Put a quiet scene before a violent one instead of after. Move a reveal a chapter earlier. The point of cards is that these experiments are nearly free. In a manuscript, moving a scene costs an afternoon of continuity fixes. On a card, it costs a second.

Writing out of order

Nabokov's habit of writing whichever card was clearest is worth borrowing too. If you know exactly how the confrontation in the kitchen plays out, but have no idea about the train journey before it, write the kitchen scene. Your cards will show you where it goes.

Many writers find this is the cure for the blank page. You never have to start at the beginning. You only have to start with a card you can see.

A practical workflow

1. Brain dump. One card for every scene you can imagine, in any order. Don't judge yet.
2. Sort. Arrange roughly in sequence. Group by act or chapter if you know them.
3. Fill the three lines. Goal, conflict, outcome, on every card.
4. Find the holes. Where does the story jump? Write cards for the missing links.
5. Test and shuffle. Cut what's dead, move what's misplaced.
6. Tag. Mark each card with its point-of-view character and its plotline so you can see balance at a glance.

Cards work beautifully alongside a plotting grid: the grid shows the big picture and the cards fill its cells. And if you're starting from nothing at all, the Snowflake Method will produce the scene list you need to start dealing cards.

A novel is too big to hold in your head. A scene is not. Plan in scenes.