Multiple Citations for Individual Elements

Paula, the only "hard and fast rule" is that each individual assertion we make must be individually keyed to a sound source. Sourcing is not governed by "how many." It's not about quantity. It's all about quality.

In the research phase, we gather every relevant piece of information we can find. That’s thorough research. Thorough research is not about "Well, I have x-number of sources for that, so I've got enough to prove my point." No. Thorough research means we use everything relevant that exists for the time and place, because just one new piece of information can overturn everything we already have.

If we conduct our research within the framework of research reports, then each document we transcribe or abstract will have one citation of source for the whole document. If we isolate bits and pieces of information from that document and put each piece into a database, then each piece of information we enter must carry its own specific identity of the source it came from.  

Meanwhile, each time we create a citation, we evaluate the strengths and weaknesses of that source. (The Evidence Analysis Process Map helps here.) We record all those strengths and weaknesses we observe. All sources are not created equal. For our research to be sound, we must make decisions about the reliability of each source. Later, as conflicting evidence begins to accumulate, our observations about the strength and weaknesses of each source will be critical to resolving the conflict.

In the writing phase of our work, each time we make an assertion, we choose the best source or sources to support that assertion. If there are no conflicts after thorough research, then citing one sound source is sufficient. If there are conflicts or controversies regarding a point, then we will likely cite multiple sources for that point and explain why x is reliable but y is not.

You specifically ask about multiple sources for a name. Practically speaking, why would that matter?  Here are three examples:

• If we have seventeen documents for one man and sixteen call him Shadrack while one census calls him Frederick, what is the important issue here?  We have to determine why that oddity exists. Is that Frederick actually the same man? Is the “Frederick” record a derivative rather than an original (typically the case with census records) and the census taker made a copying error? Our evaluation of the situation in which the source was created helps us determine what the name should be or whether multiple men are involved.  Can we just call him “Shadrack Frederick," cite both sources, and move on? No. An error by one record-keeper did not change his identity. It’s up to us, as researchers, to define the error and explain it so that it will not be thoughtlessly repeated by others.
• If we have seventeen documents for Shadrack and five spell his name as Shadrick, does that matter? Not really. That’s just a spelling variant. On the other hand, if we have twelve documents for John Smith and five for “Johann Schmidt” that we think represent the same man, then those variants matter greatly. What implications exist here? Does it point to ethnicity? Was he literate? If so, have we found samples of his actual signatures—not a clerk’s copy of his signature but his actual signature, to see how he self-identified?
• If other researchers argue that Jacob Whozit was actually Jacob ALBERT Whozit, then what document actually uses a middle name for him? Is it actually a document—or is the source an assertion made by someone in a later piece of writing?  Was it created by him or in his presence? Is it a document in which he self-identified as “Jacob Albert”? In a case such as this, the “number” of pieces of evidence can help to resolve the controversy. If, for example, our research uncovers fifty-seven documents created by Jacob, or about Jacob in his lifetime, and every one of them calls him only “Jacob” while none of them call him “Jacob Albert,” then the sheer weight of the numbers in this case can be an important point to make in our argument as to his name.  

Everything we do as researchers boils down to one word: judgment.  There are no magic numbers that relieve of us the need to make judgments. All those rules we hear about, “We must have three sources that all agree to prove a point” are meaningless because three unreliable sources mean an unreliable conclusion.  It's the quality that matters, not quantity.

As new researchers, it is natural to wonder “Do I really need five sources for his name?” We look at the five and we see repetition. We see redundancy. That makes us wonder “How many is enough?”  But accuracy is determined by the weight of the evidence, not the number of pieces of information.