The late Edmund Morgan on writing and conducting research:
I do not employ much of a method in doing historical research except to read indiscriminately everything I can lay my hands on that may relate to whatever topic has excited my curiosity. I have no system….
The reason I do not like to talk about my own research while doing it is that I lose the impulse, the necessity, of putting what I think I have found in writing. I firmly believe that too much discussion amounts to “talking your book away.” But talking about other ideas excites my own thought processes.
If you have studied any part of history enough to be curious about it, enough to want to do some research, you already are aware of the generally accepted views, the orthodox views, the controversies among the experts in the field, what is taken for granted and what is in dispute. You want to learn a little more about some question, and you go to the source materials that are presumably the foundation of the orthodox views. You come across something that you had not known about, something that surprises you a little. Cultivate that surprise. Do not say to yourself, “Oh, I didn’t know that,” and go on with your reading. Stop right there. Ask yourself, Why did I not know that? Is it contrary to what I had been led to expect? Is it because I did not know enough? Or is it because the people who crafted the orthodox interpretations did not know enough? Or perhaps their angle of vision was limited by what came before….
I want to offer a couple of other pieces of advice. First, and probably most idiosyncratic, try to forget philosophies of history and theories of historical causation: Marxist, Straussian, postmodern, or whatever. You probably have one, conscious or unconscious, but try not to let it get in your way. Cultivate that surprise when the documents don’t seem to support your views. Next, try to keep your research and your writing together. Don’t wait until you think you have entirely completed your research before beginning to write. As soon as you begin to see connections between things that you had not noticed before, start writing what you think you have found out about them, even if these writings seem fragmentary. Don’t get too systematic. Don’t make elaborate outlines with headings and subheadings. Don’t spend a lot of time arranging your notes. Stop stalling and start writing.