The Historical Record is not the Last Word -- especially online
by John Cardiff

After exhausting first hand personal knowledge and interviewing surviving family members, genealogists turn next to the historical record -- the birth, marriage and death certificates, wills, cemetery stones, census records, land transaction records, military records, church records, newspaper announcements and obituaries, for information about their ancestors. And therein lies a problem too seldom discussed: more often than we'd like to admit, historical records get it wrong.

The bulk of the historical record was recorded by somebody else, most often a low-ranking government or church official, who wrote down what they heard. Or more to the point, what they thought they heard, sometimes in a noisy room, and/or through a thick accent, when people were impatient to finish the interview. Complicating matters significantly, the official's handwriting was seldom a model of great penmanship and his (seldom her) spelling was less than terrific.

But wait, it gets better worse. Most of this happened when literacy was not a given. (Even into the mid-1800s nearly half the local brides and a third of the grooms signed their marriage records with an X. Some WWI recruits signed in by "making their mark.") Free translation: they couldn't proof-read their own miswritten records.

This was particularly true for those who left the fewest historical records: the least educated and financially disadvantaged. Doubly so when English wasn't their first language.

Complicating matters once again, original historical records are kept under lock and key. Only transcriptions of the original are readily available to family researchers. Alas, transcriptions tend to be created by volunteers who have the time and interest to do the transcribing, but frequently do not have the wherewithal to verify the results of their efforts. Often transcribers lack a working knowledge of the names in the community at the time their source document was generated. Nor are they always aware of writing conventions of the day, which by itself ensures transcription and interpretation errors.

Such errors are sometimes easy to spot -- like those of the 1901 Census takers who consistently misspelled Mabel as Mable, thereby throwing other spellings into question.  The problem is: such easy-to-spot errors are not the biggest challenge. More perplexing: the number of times the birth date provided on the 1901 Census disagrees with the birth date on the corresponding cemetery stone. When historical records disagree, how is a genealogist suppose to proceed?

Step One, of course, is to note the conflict. Step Two (for the lazy) may be to follow the genealogical maxim that the record created closest to the event is most likely correct. Step Two for the rest of us is to keep looking, to find other historical records covering the same event, hoping a preponderance of evidence will eventually lead to a "most likely" conclusion.

(Cemetery stones are considered among the least reliable 'historical records' by some as their data is usually provided by a descendant (long) after the fact, rather than by a participant of the event. In particular it is not unusual to find a stone's birth date is out by a year, due to quick and dirty date calculations. Conversely, since cemetery stones are most often erected on behalf of a near relative, soon after the event, name spellings on stones are frequently considered "gospel" -- much more reliable than newspaper or church records.)

One of the first things I did on the Internet years ago was post data later proven untrue. Within months other genealogy web sites had picked up my error and copyrighted it. Ever since they have been providing my error to genealogists who pay to access their "vast array of historical records." For the last decade, at least once a year someone has contacted me to complain the information I now provide disagrees with this error which they "know" is correct because they paid for access to it.

Genealogists "knowing" errors are correct is a challenge at least as large as errors in the historical record or errors in transcription of the historical record. A little humility may go a long way here. In reality none of us knows what happened a century or two ago. All we know is that we found a source that said...

And, of course, if you are citing or creating a fact without identifying your source 
(as I did for too many years), your own credibility ranks south of that of the 
Flat Earth Society.

 
Copyright 2010-2019 John Cardiff