How Umberto Eco Wrote a Novel

I tried to outline a novel in a Google Doc. Three hundred pages of nested bullets, color-coded by chapter, a graveyard of links to articles I would never read again. The book never got written. The doc got very, very long.

Then I read about how a man named Umberto Eco wrote actual novels — the ones you have heard of — and the entire premise of "outlining" fell apart for me. He did not outline. He collected cards.

The constraint is not the page count

A novel is too big to hold in your head at once. A chapter is too big. A scene is too big. But a card — one thought, one card — fits inside your working memory the way a single step fits inside a walk.

Eco, an Italian semiotician who also happened to write some of the densest novels of the twentieth century, treated the index card not as a notebook entry but as a unit of thought. A card holds one idea, one source, one quote, one question. The card is small on purpose.

The point is not that the card is "the right tool." The point is that the card is the constraint. The page invites you to ramble. The card refuses.

The work plan is the skeleton, not the table of contents

Before Eco read a single source, he wrote the table of contents for the book he had not yet researched. Title, introduction, sections, subsections, short chapter descriptions. Then he threw it away. Then he wrote it again.

The point was not the artifact. The point was the act of forcing himself to commit to a structure before he had the material to fill it. The skeleton made the reading possible. Without a skeleton, every source is interesting and nothing connects to anything.

This is the move most people skip. They collect cards without a skeleton and end up with a beautiful pile of facts that never becomes a book. The skeleton is what tells you which cards belong and which do not.

One card, one source, one decision

For every source, Eco wrote a bibliographic card. Source details, location, and a one-line note about how important it was. If you cannot write that one-line note, the source is not worth keeping.

For the reading itself, he wrote reading cards — bigger, half a page, several per source. Each card captured: the bibliographic reference, a summary, the quotes (clearly marked as quotes), his own comments (clearly marked as comments, in square brackets and a different color), and an abbreviation pointing back to which section of the skeleton the card served.

The key discipline is that the card forces you to decide. You read, you summarize, you quote, you comment, you tag — all on the same surface. The card refuses to let you be passive. You cannot "highlight and move on." You have to write.

Connection cards are the actual engine

Once you have a hundred reading cards, the obvious move is to sort them by section. Eco did that. But he also wrote a second kind of card: the connection card.

A connection card holds a single observation: "this idea in source A relates to that idea in source B, and the relation is X." A connection card is not a summary. A connection card is a thought. The whole system — all the reading, all the bibliographies — exists to produce connection cards.

This is the move most digital tools skip. You can tag a note, you can sort it, you can search it. But the act of writing a separate card whose only purpose is "this connects to that" is what turns a pile of facts into an argument. The connection card is the card that thinks.

The pile on the floor is a feature

When Eco was ready to write a section, he pulled the cards for that section off the shelf, dumped them on the floor, and wrote. Not the cards in order. Not the cards from one source. The cards for that section.

The card pile on the floor is a physical thing that resists your abstractions. You cannot pretend a card belongs in this section if it is not in the pile. You cannot pretend you have thought about something if there is no card for it. The floor does not lie.

Digital tools try to give you the floor in the form of filtered views and smart folders. They mostly fail. The pile on the floor is a pile on the floor. You cannot multitask your way around the constraint.

The system outlives the project

Eco wrote The Name of the Rose, Foucault's Pendulum, The Island of the Day Before, Baudolino, The Prague Cemetery, and several more novels over thirty years. He used the same card system for all of them.

The system outlived any single project because the system was not designed to produce one book. It was designed to hold a lifetime of thinking. Every card you write is a card you might still need ten years from now.

This is why the constraint matters. A card is small enough that you can keep tens of thousands of them. A page is too big to keep. A document is too rigid to reuse. A card is the unit that survives.

What to do this week

Buy a pack of 4x6 index cards. Pick one project you have been stuck on. Write the table of contents for the project on a single sheet of paper — title, sections, subsections, one-line descriptions.

Then read for twenty minutes with one card per thought. Source on the front, your summary on the back, abbreviation to the skeleton section in the corner. Stop when you have five cards.

Pull the five cards onto your desk. Read them. If you cannot order them into a sentence, you have not read the source yet. If you can, you have the beginning of the project.