OK, all you historians and writers out there. I'm begging for help. What are your solutions? What would you do if this problem were your own? You have a rumor of an original document that would provide vital insights about your main character. That original, however, has disappeared. The archives that said they had it really only had a copy. The microfilm that lists it in its index is only talking about a copy. The people who might have handled the original are all dead. If it exists at all, it is in one of two places -- hidden away in the attic of some distant relative who has no idea what it is, or being held in the bowels of a university library, where the label "being prepared for cataloguing" means "nobody will see this until sometime shortly before Judgment Day." (Which, by the way I do not believe will come on Saturday. That would be too easy a solution.)
What remains? At least five copies that all profess to be accurate -- two published editions, two typescripts, and one Xerox of a handwritten copy, which may or may not be in the handwriting of the original author. I can't say for sure because there are no other examples of her handwriting with which to make comparisons.
All these copies seem to have been made within 5-10 years of each other, although only 2 of them are dated, and all are at least 80 years old.
What's worse, all of them are different, but not in any clearly definable way. Each one contains some unique material and each has omitted different passages, so that there is no way to say that "x" is a copy pf "y". Some of the differences are minor, some are alterations in vocabulary, some are large blocks of missing entries, and some are massive rewrites or creative new passages.
Given all that, how would you decide which is the most accurate source of information? What questions would you ask? What clues would you look for? For example, I immediately thought I would try to determine the earliest version, on the assumption that the farther one moves from the original, the more people have had a chance to mess with it. The problem with that little pearl of historical wisdom is that there is no way to tell which one IS the earliest. All the copies cover the same period of time, 1862 to 1866, but all the copies seem to date from the period between 1900 and 1915.
I can't toss them down a flight of stairs and see which one falls the furthest, as college professors are reported to do with term papers. These are all too heavy. I can't have them draw straws. I don't believe in lotteries. What do I do? Toss me a rope, please, even if it is only one with which to hang myself.