What started out as a follow-up comment to that left by Michael Hait on the post, Confusion with the Various Definitions of Original Source, got too long for a comment and has turned into this post.
Thank you for contributing to the conversation. Any discussion of sources and their classification as original or derivative is not complete without discussing source provenance. When we trace the incarnations of a source all the way back to the original, we are able to do two things: 1) answer the question “Is there a better source?” and 2) determine independent origin.
Let us take the example of an original census that was microfilmed and then digitized. We are looking at the digital copy and we determine that the image is dark and hard to read in certain areas. We ask ourselves if there is a better source and determine that the problem was likely with the microfilm so if a better source exists it would have to be the original. Let’s say that we are able to consult the original and we can read the problem areas. In this case, the original was the better source.
Let’s change our scenario a little. Same original and two derivatives, but this time we ask to consult the original and it doesn’t exist as it was destroyed after microfilming. In this case the microfilm copy or the digital image is the best source available. Similarly, if the original exists but has deteriorated in the years since it was microfilmed, then the microfilm/digital image would be the better source.
When we have an original, we only have two ways to make a copy of it: 1) photographically reproduce it or 2) manual copy (either written or typed).
To handle the first group, Evidence Explained discusses the concept of an Image Copy. An Image Copy is a derivative, but it can be treated as if it were the original if it 1) is legible and 2) doesn’t conflict with other evidence. A #3 is implied in that all representations of the source back to the original must also be able to be treated as originals. So the microfilm is an Image Copy of the original census and can be treated as an original. The digitized image of the microfilm can also be treated as an original because it is an Image Copy and if we print out that digital image we have made another Image Copy which as long as it passes the tests can be treated as an original. In your example, the death certificate is an original source and a photocopy of it is an Image Copy of an original and can likely be treated as if it were an original.
For the manual copy example, we have more choices: Duplicate Original, Record Copy, Transcript, Extract, and Abstract. The census enumerator goes house to house and records the information that is required. It is the original source. Now the Federal government requires that a copy be sent to them, so the enumerator makes a copy of the original. Technically it is a well-known type of derivative called a transcript, but because it is part of the “official” act of creating the census, then we can treat it as if it was the original and we give it the special name of Duplicate Original. There are known examples where this copy was changed to be alphabetical or only listed people by their initials or mistakes were made because the copying was done column by column instead of row by row, etc. There are even case where the true original was sent to the Federal government and the Duplicate Original stayed at the state level. We try not to get too caught up in the details and are satisfied that it is a Duplicate Original all the while keeping an eye out for irregularities. Other examples of Duplicate Originals are grantor/grantee copies of a deed and counterparts. I think about copies made as part of the same “transaction” when I think of Duplicate Originals.
A Record Copy is also technically a derivative (likely a transcript) but this time an official (likely a clerk) is entering information into a register. As in other cases, mistakes can still be made but because of the authority of the official and their mandate to record the information, we can consider both the original deed and its recording in the deed book as originals. There have also been times when the transaction has been recorded directly into the register in which case it is the original.
Outside of the special cases of Duplicate Original and Record Copy, other manual copying is categorized as transcript, extract, and abstract based on the amount of document that is copied. Transcript is full document (the entire census for a year and state), extract ( a page from the census), and an abstract (just the more important parts: names, places, dates, etc.)
All the above helps us determine the best sources for our analysis.
Let’s briefly talk about independent origin. Let’s say we have 100 derivative sources that are varying generations removed from their original. Two sources agree on one birth year (1880 ) whereas the other 98 give the birth year as (1830). We might be tempted to go with the 1830 year as there exist more sources that agree with that year. But without tracing the provenance, that would be a mistake. Let say that the provenance is traced and all 100 derivatives come from only 3 independent original sources. In fact, all 98 derivatives that agree with 1830 are all from the same original source. When we consult this original source, we see that a mistake was made in the transcription that changed it from 1880 to 1830. The concept of independent origin is important to consider. Evidence Explained reminds us that quantity can never trump quality.
As genealogists, we identify source, the information that they contain, and the kind of evidence that information provides when compared to our research problem. Sources, information, and evidence. Not one is more important than the other. That would be like saying that our eyes or hands or mouth is more important than the others.