Research is seductive. The second-most common form of shrapnel found in soldiers wounded during the battle of Stalingrad was human teeth. D.W. Griffith, in the days before movie cameras had sophisticated irises, created fade-outs by slowly raising a cigar box lid in front of the camera lens. In 1944, as German V1 and V2 bombs began decimating London, Londoners, exhausted and afraid, found that food was surprisingly far easier to come by, and oranges—which hadn't been seen in the city for five years—were suddenly plentiful; the sight of orange peels on city streets was immensely cheering.
Even as a child I loved discovering such things, and now as a writer, I often get to use them. In fact, one of my favorite parts of work is researching things I need to know—the medical capabilities of a German WWI field hospital on the Eastern front, in order to determine what a nurse might have on hand in an emergency, or the daily routine of a WWII London Air Raid Warden, to make a character's life seem realistic, or the ins and outs of movie-making in the 1920s and '30s, to grasp how a director of the time would think. Aside from the fun of learning odd new things and the need to understand my characters, research is important because certain images or facts will flesh out scenes.
I was a history major in college, so I often have a general idea of times I'm writing about, but of course to really understand them I have to research extensively. I'll turn to history books first, but mostly for the footnotes and bibliography; I find that memoirs, letters, and diaries are the best sources, as they're filled with the minutiae that history books often don't have time for: what the weather or cityscape was like, what scents were likely hanging in the air, the compromises large and small that people had to make in order to survive not only bombing but rationing of food and clothing and fuel. The internet has in some ways made this easier, though I still love the old-fashioned thrill of stumbling across unknown books in the stacks, because it's the unknown that often leads to greater understanding.
I tend to write and research in alternate bursts: sketching out a scene, knowing that I'll need realistic details to fill it out, putting the work aside as I go off to find them, then returning to the scene, working them in and often changing many of the events (and the characters' thoughts and feelings) because of them. And I almost always find more material than I can (and should) use; I put most of it in, then revise heavily, using as a general rule the idea that if the details are merely fun, they don't belong; they have to serve the story, not slow it down.
The great danger for me, because I so like research and because the actual writing can be so difficult, is that while researching, I can tell myself that I'm working and therefore don't have to sit down and write that troublesome chapter; I'll get to it when I know a few more facts. And that, of course, might mean reading two or three more books, jotting down notes as I go
This is a real danger. Researching isn't writing, finally, no matter how crucial it may be, since that difficult chapter isn't going to write itself no matter how many fun facts you might have assembled for it. When I find myself putting off writing for too long, I make myself sit down and do it. Still, I find research a crucial part of writing—getting it right is part of the seduction, after all—and when research and writing work together, the results can be wonderful.
During a question-and-answer period following a reading for my most recent novel (part of which is set in Hamburg during its 1943 fire-bombing by the British Air Force) a woman in the audience asked if I'd ever lived through war (I haven't). Later, as I was signing her book, she told me she'd been a child in Hamburg in 1943, and that when she got to the fire-bombing in my book she had to put the book aside. I began to apologize, thinking I must have done a poor job combining research and imagination to create the scene, and that she found the mistakes off-putting. No, she said, it wasn't that.
Had it been traumatic to read, then? I asked.
She said it had.
I began to apologize again, feeling a touch of guilt.
No, she said. It brought back so many memories, I had to stop. But I couldn't stay away for long, because you got it exactly right.
That may have been my proudest moment as a writer, and it fired me up to continue researching for my next book.