It's amazing to me how much depends on search terms. In these early stages of research, when my sources aren't yet settled upon and my ideas feel hopelessly abstract, knowing exactly which search terms will unlock the wonders of archive databases is like knowing which magic words will make a rabbit appear out of a hat.
Today, I quite accidentally said the magic words and poof! bunnies. Well, bunnies in the form of manuscripts and archival documents in places as far flung as Cheyenne and Toledo and Richmond. (Never, of course, Tahiti and Paris and Tokyo, but such is the plight of an Americanist.) Suddenly, this research started to be about possibility again instead of inevitable misstep and error. I stopped being terrified of having the right research plan and concentrated instead on the research. Tomorrow, I'll likely find myself petrified and silent again, but today it was lots of "yes!" and "whoa" muttered under my breath as I stood in at the library terminal.
This dissertation could still be anything. I don't need to know, just yet, what kind of rabbit I seek.