What are sources anyway?

I recently addressed the issue of citing sources in a blog post entitled "Check your sources or else." It occurred to me that I didn't include anything defining a source. I guess this comes from being so familiar with a term that we don't realize that the whole world does not necessarily know what we are talking about. So I thought to rectify that omission by writing about sources in general and avoiding the topic of how or when to cite them (maybe).

You would think that the definition of a source would be helpful. As with many word definitions, the standard dictionary-type entry merely expresses the idea of the word using synonyms. For example, the Mirriam-Webster definition is as follows:

: someone or something that provides what is wanted or needed
: the cause of something (such as a problem)
: a person, book, etc., that gives information

In each case the word definitions skirt the meaning of the word as it is used in genealogy. The third definition comes the closest, but is lacking in some important respects. Let's see if I can give the meaning of "source" a little more specificity. Here it goes:

A source is a person, book, document or other information provider that is used to establish a name, date, place or event in an ancestor's life.

Obviously, my definition is not much better than the one from the dictionary, but I will expand a little more. For example, if you ask your mother her birthdate, she becomes the source for the information derived from that question. In other words, your mother becomes the "source" of your information. Of course, you can have multiple sources. You may also find a birth certificate for your mother and that also becomes a source for the date of her birth and for any more information contained in the certificate.  If you found a book written about your mother that contained a birth date, that book could also considered to be a source. You can see how this could go on forever.

Now, I should mention an important rule. The number of sources has nothing to do with proving the truth or falsity of any fact. One good source is better than a thousand bad ones. So, there is something we have to do to evaluate a source rather than just pile them up one after another. In genealogy, we sometimes speak of primary and secondary sources. Making this distinction is an attempt to overcome the problem of evaluating the accuracy of any particular source. In genealogy, the definition of a source goes well beyond the idea of just finding a fact and recording where it came from, the researcher has to evaluate each piece of information (fact) in the context it was produced and check it for accuracy and consistency. For example, going back to the birth certificate, what if you find U.S. Census records showing your mother as a child and each time her age is recorded it is different from what she told you and what is on the birth certificate? Which record do you believe?

This issue brings up the idea of a "primary source." Thinking about sources as primary or secondary has some weight in evaluating their accuracy, but only has limited value in proving the correct date. In this context, a primary source is one that was created at or near the time the event occurred and was created by a person who actually witnessed the event or recorded the event in an official capacity every other source however denominated is secondary. In using the above examples, you mother's statement as to her birth date could be considered a primary source, but I would consider it otherwise. She may have been there when she was born, but she had to rely on someone else for information about her own birth.

The mere fact that an event was recorded does not speak to the accuracy of the record. Do I record every source? Yes. But in every case, each source has to be evaluated for accuracy and consistency. Another important fact is that any given record may contain both primary and secondary information. Under the definition of a primary source, a birth certificate would be a primary source for information about the date of the birth signed by the doctor, midwife or other person present at the delivery, but almost everything else would be considered secondary. Researchers commonly become disconcerted and overly concerned about inconsistencies in various records and accounts. Inconsistencies are entirely normal. In the end, it may be necessary to conclude that the exact information concerning an event in an ancestor's life can never be completely and accurately determined. The correct information may never have been recorded. Fortunately, genealogy does not require absolute precision. Sometimes we have a pretty large ballpark.

Now to the question of recording obviously inaccurate sources, what should be done? Let's suppose that I go onto one of the larger family tree programs and find a hundred family trees containing information about my ancestor. Do I record every single one of them? My immediate answer would be no. But the reason is a little more complex. The question would be whether or not you used any of those family trees to add information to your own database. If so, the family tree becomes a source. If not, then they should all be ignored. The reason for ignoring the multiple user submitted family trees is based on the idea that there is nothing new in any of them and referring to each one would be a waste of time. In the legal world, we call this cumulative evidence and there are limits as to how much repetition any court will allow.

What if you find several documents containing birth information? Yes, all of them should be recorded as sources. This is usually necessary because of the presence of other information on the other documents. This does not mean you make every copy of every document into a separate source, it merely means that you continue to record sources even when the data overlaps with other similar documents.

The main problems with sources occurs when seeming otherwise accurate sources have contradictory information. This is where we leave the topic of recording our sources and get into the issue of evaluating their accuracy. Maybe next time.