I was asked about how these records we search came to be. Only in the last few generations (late 1800s to present) has it been common for “regular people” to be literate (able to read and write - some could sign their names, but even that was rare in many communities and they would sign with an X - indicated with “his mark,” written by the officiator of the document).
There are 2 main types of documents: Primary (or Original), and Secondary (or Derivative). Of course, this second category also includes tertiary and beyond - such as “word of mouth,” etc. Primary or Original documents are produced at the time of the event and by someone who was at the event (e.g., birth record signed by the person who delivered the baby, marriage record signed by the Priest or minister who officiated, etc.). We call “vital records” (the records dealing with our important life events - birth, marriage, death) to be Original, as far as the data about the person and event in question are concerned. For example, a birth record for a child probably will have accurate data for the child’s name (though that still may be in error), birth date and place, and mother’s information (though, if she is lying, that is not likely to be known). The father’s data may also be accurate, but if there is no father in the picture, or if someone who is not the father is named as such, then it’s anyone’s guess. Prior to the mid- to late-1990s, DNA was not considered as a method of determining paternity (though it could determine non-paternity in some cases).
OK, a death record is considered a primary record for the death information ONLY. And if it’s a case of a body being discovered sometime after death occurred, even that can be in question. If there is a coroner’s report, that will not accompany the death record (that will usually require a court order to see, though a direct family member may have access, depending on circumstances and jurisdictions). Death records usually have the names and birth places of the decedent’s parents, but this information is accurate only to the extent that the informant knew it. My mother’s death certificate lists “Lalla Trapschuh” as her mother. Mom’s mother was “Emma Hollander.” My father, who knew his mother-in-law very well, provided the information. But Dad was distraught and couldn’t think. Who was “Lalla Trapschuh”? “Lalla” was my Grandmother Emma’s nick-name among the family members. “Trapschuh” was HER mother’s maiden name. There is no such person as Lalla Trapschuh, but consider my confusion if I proceeded to try to find this individual in census records or other such documents! And this happens frequently!
A delayed birth certificate (issued sometime after the birth - often years later and usually because the original certificate contained an error, was never filed, or was destroyed, possibly by courthouse fire) is even more uncertain and would be considered a secondary piece of evidence. My great-grand-uncle and his siblings (except my great-grandma) had delayed birth certificates filed in Milwaukee. They are part of the vital records of Milwaukee and say they were born in Milwaukee (sworn to by their mother . . . who gave birth to them . . . in Bilin, Bohemia - I have their christening records and proof they immigrated to the US after they were born). Hence our skepticism of delayed record filing.
Federal Census records, which we use a lot, are generated once every 10 years. These are supposed to record every individual (of all ages) in the US. From 1850-2010, these included names and other data for each person in the household. From 1790 (the first Census - following the winning of the Revolutionary War in 1783) to 1840, only the heads of household were named. Then there are slash marks to indicate what other people were in the household (grouped by age and gender). But who gave the information could be anyone - a child who was home, a spouse (maybe a second spouse giving information about her husband’s prior marriage, etc.), even the next door neighbor. How accurate are these? It’s anyone’s guess. In 1940 there was a mark placed by the person in the household who gave the information, and it was required that every household be enumerated by information from a household member (no more next door neighbor reports). Census records are kept private for 72 years, so the most recent we have is 1940. The US Census for 1890 was destroyed by fire and water and only a handful of names remains (under 2,000, I think).
There are other forms of Census records - State, Native American, Slave, Military, etc. We use these, when available, but preservation of these was not always a priority and many state and special census records are not on line, but need to be accessed via microfilm through FamilySearch. A note about the Slave Schedules of 1850 and 1840: most slaves had (or used) only first names, so if someone named John Freeman is suspected to have been a slave, it is extremely difficult to identify him. Many think that the surname he used as a freed slave is his former master’s. This did happen, but not always. Once emancipated, John may not have wanted to have been reminded of the person who kept him in slavery and, instead, preferred to adopt a new name, declaring his new status as a free man, hence John Freeman. (The Freeman name in the south, among African Americans, is common and relationships to each other are not at all known, and often unlikely.)
Hope that clears up a little of how these records are viewed, why we use them, and why we are skeptical about accepting everything in writing or on the Internet as being an accurate account.