Bob, the CSR are indeed rife with errors, primarily due to errors in the original source documents. When I visited the National Archives many years ago, I read a monograph which detailed how the CSR came into being. Of particular interest were the instructions given to the army of "copyists". They were instructed to copy exactly what the source document said -- obvious errors and all. The only leeway allowed was in the deciphering of damaged documents and near unreadable entries. In those cases they were to use their "diligent best effort" to "faithfully" interpret them.

During years of transcribing the CSR, I have noticed that some copyists used little coded marks to indicate to future researchers that some of the entries they faithfully copied were obvious errors -- things like a small "x" placed over certain letters or entire words. Also, at some point after the CSR extract cards were completed, it is apparent that someone had later added remarks to correct certain errors.

So your point is well taken. In most cases, the CSR is the only game in town when you're researching a soldier (unless you're fortunate enough to discover a diary, letter, pension file, etc.) and can usually be relied on, with the clear understanding that they are extracts from original and occasionally conflicted documents.

The most glaring example of erroneous CSR information involves entries regarding desertion. The problem is not with the copyists, but with the officer or NCO who wrote the original document, as we have discussed on this board many times.

Finally, it should be pointed out that the relatively recent addition of a CSR index -- such as that posted on the National Park Service "Civil War Soldiers and Sailors" website -- has many data entry errors and should never be relied on as documentation. It should only be used as a tool to find one's way to a specific CSR.