The Basic Uncertainty of all Historical Research

There has been much written about achieving genealogical proof but in many cases, conclusions that a certain opinion or conclusion is "proof" of its accuracy is unwarranted. Let me take an example from a fairly common occurrence. Let's suppose that a person "A" has grown up in a family and has always assumed that he or she is a biological child of his or her parents. Proof of that relationship is a validly issued birth certificate showing that relationship. Because of the popularity of genealogical DNA testing, "A" takes a test and discovers a person who the test indicates is a sibling. As a result, "A" learns that he or she was adopted. But what about the birth certificate? It has been a common practice in some states in the United States to issue a "corrected" or false birth certificate when a young child is adopted. This practice is currently being widely questioned, but there still exists some number of these altered or corrected birth certificates.

The example of an altered birth certificate illustrates a fundamental issue with historical records, There is always a certain measure of uncertainty in any historical research effort. A genealogist who searches historical records for information about a particular family or person will likely discover contradictory documents which, in some cases, cannot be reconciled. Over the years, I have developed some general guidelines concerning the accuracy of any given record. By the way, it is overly simplistic to rely on the commonly disseminated criteria for determining a document's reliability: i.e. proximity of creation to the event and the ability of the observer of an event to accurately record that event. These sorts of seemingly accurate criteria for judging the reliability of a record ignore the basic causes for inaccuracy: faulty observation, mistakes in recording, and misrepresentation. It is also possible that for any given event no record exists. This brings us to Rule Two of the basic rules of genealogical research which can be restated as follows: absence of a record does not imply that the event did not occur. See "A New Rule Added: The Rules of Genealogy Revisited Again."

Genealogists, irrespective of their level of expertise (which includes me) all base their conclusions on the content of the records they are able to find and review. Of course, this assumes that the genealogist has actually spent some time doing research using historical documents and is not just copying the conclusions of others. Although there is some discussion about making "exhaustive searches," the real issue with this admonition is the determination of who becomes exhausted and when that might occur. For example, I may compile a long list of places and historical record collections I have searched, but when I stop searching is always arbitrary and anyone looking at my list could possibly suggest further searches that could be made.

Another example of the issue is the existence of an ancestor's will. Researchers will often assume that inclusion or exclusion of someone from a will is dispositive of a relationship. If a "wife" is mentioned in a will there is often no way to determine if the wife mentioned was the only wife or the one at the time the person made the will. The accuracy of a will can be based entirely on when the will was made. In the United States, particularly since the mid-1900s, people have been encouraged to write a will many years before their expected death date. It is common that these out-of-date wills end up being used in a probate action even despite the lack of application of the provisions. In the past, it is also common to find that a testator has left children or prior spouses out of a will.

So how does anyone ever come to a conclusion about their ancestry? The real answer is that all genealogical research remains a "work in progress." All conclusions, no matter how firmly entrenched in the genealogist's conclusions, should be viewed as temporary awaiting further information. Some conclusions are not ever likely to be revised. For example, if my ancestor was born in 1850, it is fair to assume that he died. But the actual date of death may be subject to revision. Did you catch the restatement of Rule #2?

It is common for online family trees to include substantial amounts of information that lacks any source citations. This automatically calls into question any of the information recorded including the name of the person. But it is just as common for information to be recorded with only one or very few sources. For example, I commonly find a list of sources but none of the sources provides information about the events recorded even by inference such as listing a birth date with only a marriage record attached that make no mention of a birthdate or age. Genealogists have compensated for a lack of records through a number of "rules of thumb" that estimate birth, marriage, and death dates based on the common practices of the time. But in without reliance on a source, these should always be subject to revision.

Another common practice is to estimate a birth date given the date of a church christening. I often find documents that show that a person's christening took place when the person was older and not an infant. In one case recently, I found all the children in a large family christened on the same date.

There should be nothing about this discussion that discourages a researcher from relying on apparently accurate records. But it is important to always remember that the names, dates, and places you have recorded in your genealogical compilation may be subject to revision. Also, opinions and conclusions about events and relationships may differ between equally experienced researchers.

DNA appears to be the ultimate solution to any of these uncertainty issues, but careful, DNA testing, especially for genealogical purposes is based on statistical probability, not absolutes. Close family relationships can be determined with a high degree of accuracy, but the accuracy and reliability decrease with the distance of the relationship asserted. As my first example about birth certificates illustrates, DNA is one more tool for determining relationships but DNA testing will not replace searching historical records.

What about accepting the conclusions of experts? One example is the huge amount of research that has been done by the General Society of Mayflower Descendants or more commonly the Mayflower Society. The Mayflower Society has spent many years and apparently a huge amount of time and money research the descendants of the passengers of the Mayflower. In cases like this, it is very unlikely that a casual researcher is going to encounter any more documentation than has already been found and printed in the Society's books and other publications. But it is possible that an experienced and persistent researcher could dispute one or more entries. But for most purposes, if you find that there is extensive documentation and take the time to understand the conclusions derived from that documentation, you can safely rely on the conclusions. What you should not do is rely on conclusions where the researcher refuses to provide documentation.