Also from the Scholars Talk Writing series, Lillian Faderman‘s system sounds very like John McPhee‘s:

“I developed a system when I was writing my dissertation 50 years ago, and I still use it. Before I start writing a book I need to have at least the illusion that I know how it’s going to end and everything else in between.

So first I need to complete most of my research. Then I review all my research notes — which generally takes several weeks — and I decide what will be useful and where in the book I’ll use it. I’ve concocted an elaborate coding system of numbers and letters, which I give to each idea or fact I anticipate using. Then I put it all into a huge outline with key phrases followed by the numbers and letters that will let me locate the material in my notes. When I was working on The Gay Revolution, an 800-page book, I had a 75-page outline. For my Harvey Milk biography, which is about 250 pages, my outline was 20 pages.

But when I finally start writing I veer away from the outline almost as often as I follow it — which is fine. The outline’s biggest purposes are to be a sort of Dewey Decimal System, to tell me where in my voluminous notes I’ll find things; to remind me of the ideas I want to develop; and to suggest their possible order.”

I wish someone would describe the coding system, any coding system, in more detail. I suppose it’s one of those things that is more useful when worked out for oneself, but some insight into a working system would be helpful in creating/adapting one’s own.

Note that she thinks in terms of Dewey Decimal! Once I discovered the Library of Congress system, I never looked back. It all depends on what you’re used to, I suppose.