Eric Wittenberg has an interesting post up in which he raises the issue of when a historian has collected sufficient research material for a given project. While I sympathize with Eric and others in reference to this question I don’t believe the problem is as bad as he makes it out to be. And the reason, as Eric points out, is that you can’t collect all of the relevant information. Perhaps most of what is out there is still lying in shoe boxes in people’s attics. Fortunately, you don’t have to collect all of it. The piece of the puzzle that I wish Eric had discussed in more detail is the importance of analysis in historical interpretation. It’s the analysis that needs to drive the project based on the materials collected. There is always the concern that the analytical aspects of the interpretation are supported by insufficient materials; in the worst case scenario the historian simply begins with a set of assumptions and collects data that supports those assumptions However, as long as the historian remains open to revision there is room for the interpretation to grow and hopefully become more sophisticated. The relevant question is whether data that has the potential to significantly alter the analysis has not been uncovered.
I am quite confident that there is plenty of material on the Crater that I missed in the course of my research. There are no doubt some pretty colorful accounts of Confederates sharing their perceptions of having to fight black soldiers that would enliven my narrative. That said, I am pretty confident that there is little out there that would significantly alter the main analytical points that I make in the manuscript. I decided awhile back that I was not going to try to cram as many voices into the project just for the sake of inclusion. My published work hopefully balances data and interpretation effectively.
In the end there is no answer to the question of when enough is enough. The philosopher Raymond Martin once likened the process to a "Brewmasters’ Nose." You stir the pot around for a bit and if you do it long enough and fail at it a number of times you eventually "know" when it’s ready.
So, my advice to Eric (whatever it’s worth) is to think about the main interpretive points that have been made and whether a sufficient range of materials have been collected which support those points.