Some people eat, sleep and chew gum, I do genealogy and write...

Thursday, January 21, 2016

Truth, Reality, Proof and Genealogy

I began wondering how many genealogists were concerned about the truth of what they have recorded in their files. By this I mean if they ever wonder if what they have found in the documents is "true" in the sense of corresponding to some historic reality. I think this is different than being concerned about accuracy in reporting or even from having a valid, supported conclusion. I began to write about this subject in a post dated 20 July 2013 entitled, "Is Genealogy the Search for Historical Truth?"

Let me illustrate my concerns with a hypothetical situation.

Let's suppose that my ancestor, John Doe, was born on 31 December 1886. For the purposes of this hypothetical, I am simply making up names and dates. Let's further suppose in the first instance, that for whatever reason, the date of his birth is not recorded at the time. In my hypothetical situation, the birth occurred in Arizona and at that time birth records were not required or routinely maintained. Time passes and Doe is old enough to attend school. Of course, the school would like to know his age. Because of the proximity of his birth to the New Year's date, His family decided to celebrate his birthday on another date and they neglect to tell him of the change. The date recorded by the school is the date the family celebrates his birthday and not his actual birth date.

We come along as genealogists and cannot locate a record of his birth. Of course, since we know from the hypothetical that the date was not recorded, this is not surprising. But because we are extremely diligent genealogists, we do find the school record which we dutifully record as his "birth date." All of the subsequent documents found, including his World War I Draft Registration record and his marriage record etc. all reflect the birth date given to the school.

Here is the conundrum. Do we continue searching for a more reliable birth record, or do we simply assume because of the "overwhelming evidence in the historical record" that the date is correct and leave it like that? I would venture to say that there are very few, if any, genealogists who would trouble themselves over this issue. The lack of knowledge of the status of the "true" date is impenetrable.

This hypothetical situation illustrates an important aspect of historical research: we are not involved with finding the "truth" as much as we are involved in ascertaining the content of available records. In my hypothetical situation, John Doe's existence or identity is not an issue. But what if that were the issue? Let me change the details of the hypothetical a bit. Now, John Doe was born on 31 December 1786 in Rhode Island, rather than Arizona. Again, his birth date is not recorded and there is one more item to consider, there is another individual born about the same time with the same name. I will call my hypothetical ancestor "John Doe 1" and the other person "John Doe 2." Let's further suppose that they live in close proximity to each other in Washington County.

As is often the case, researchers over the years have confused the two John Does and mingled information about the two distinct individuals. In Washington County for that time period, birth dates are most likely compiled from death and marriage records and in many cases, calculated from the ages given in those records rather than actual dates. After a careful examination of the existing records as recorded in the following books, I conclude that there are two individuals with the same name and after even more research, I become convinced that I have positively identified each of them and separated out their respective families.

Arnold, James N. Vital Records of Rhode Island, 1636-1850. First Series First Series. Providence, R.I.: Narrangansett Historical Pub. Co., 1891.
Beaman, Alden G. Washington County, Rhode Island Births, 1770 to 1850: Comprising the Towns of North Kingstown, South Kingstown, Exeter, Westerly, Charlestown, Richmond, Hopkinton, and Containing Nearly 6000 Birth Entries Not Recorded in James N. Arnold’s Vital Record of Rhode Island, Vol. 5, Washington County. Princeton, Mass.: Beaman, 1976.

In my research, I examine the Town Records for each of the towns in the area and examine the land records, cemetery records, church records, and ultimately end up visiting the state and local repositories and examine all of their available records. I also spend weeks searching every available record in libraries such as the New England Historic Genealogy Society Library in Boston, Massachusetts and the Library of Congress. For the purposes of my hypothetical I also do extensive descendancy research and locate descendants of both individuals and request DNA tests from multiple individuals. In the end, I am convinced that I have identified both individuals and I am able to document their respective families. But after all of this research, I still do not have any documentation of their actual birth dates. Their identity as individuals has been ascertained to my satisfaction, but their birth dates remain elusive. I end up my investigations by writing an extensive article for a major genealogical journal which is then subsequently published.

At this point, most researchers would say, "So what?" or "Who cares?" and move on to another research issue. Why do I give this illustration? The point here is neither simple nor obvious. Let me ask a question. Have I proved that their are two individuals named "John Doe?" Does all my research, including the DNA testing, become proof of their identity? Wasn't my use of DNA conclusive?

Now it is time to move on in my hypothetical situation. Let's suppose that another researcher addresses the same issue. She examines exactly the same documents that I examine and also does her own DNA test. She also visits all the libraries but then comes to the conclusion that the two John Does are actually the same person. She finds the DNA tests to be inconclusive and the records to be ambiguous. She manages to find even more records than I did in all my searching and makes a persuasive argument that there is really only one individual and that there is documentation of yet another individual who is named John Roe who has been confused by other researchers as being a second John Doe.

Where does all this end? The fact is that it doesn't. All historical investigations are open ended. We can make our own conclusions. We can stop doing further research. We can write up our opinions, but ultimately the reality or "truth" remains elusively out of our grasp. What I am confronting has been characterized as the problem of historical truth. Here are a few representative samples of this topic.

Ankersmit, Frank. Meaning, Truth, and Reference in Historical Representation. Ithaca, NY: Cornell University Press, 2012.
Carr, Edward Hallett. What Is History? Camberwell, Vic.: Penguin, 2008.
———. Meaning, Truth, and Reference in Historical Representation. Cornell University Press, 2012.
Francia, Guliano Toraldo di. “Historical Truth.” Foundations of Science 1, no. 3 (September 1, 1995): 407–16. doi:10.1007/BF00145403.
Heikkinen, Hannu;Kakkori. “‘And This Story Is True...’: On the Problem of Narrative Truth,” December 4, 2002.
Kern, E. H. “Historical Truth vs. Historical Validity.” The Boomerang, June 20, 2013. 

Genealogists have, almost entirely, absented themselves from the greater issues of historicity. They have wrapped themselves in a blanket of reasonableness and security by claiming to have spent the time and effort necessary to support their conclusions while ignoring the basic undercurrents of serious historical research. Personal conclusions are not per se validated by adopting any particular formal structure or artificially applied jargon.

If genealogy is to become an academic subject, then genealogists need to abandon the facile and conclusionary and address issues inherent in and central to historical research. For example, the adoption of "legal jargon" adds nothing to the credibility of the work that is done by genealogists and in fact, distracts from the main issues. Genealogical discourse needs to move into the main stream of historical discipline. Here is one place we might start.

UCLA Department of History, National Center for History in the Schools, History Standards, National Standards for History Basic Edition, 1996,, (viewed 21 January 2016).


  1. I would argue that the "jargon" is almost irrelevant, James, but that the process is an important guideline needed by most researchers (I'm sure I've made similar comments here before, but I cannot find them now). We both agree that there are some issues with the adopted terminology, but I believe that those issues are relatively minor (once understood) and that they're far outweighed by the advantages of having documented guidelines. If we come from other backgrounds where analysing evidence and making a considered argument for some cause, mechanism, or perceived reality, then those genealogical research principles make clear sense. But for anyone from a different background then they both help them and ensure some consistency of rigour and thoroughness.

    1. I agree with most of your comment, but I do not concede that jargon is irrelevant. Our methodologies are shaped by our characterization of our work. Words and labels are important indicators of attitude and belief.

    2. I agree that terms such as "proof" and "fact" can be misleading, but that's why I added "once understood" in my post. I admit that I'm not happy with those terms, and I do refuse to use the term "fact", but I've accepted "proof" as defined in a genealogical context without it necessarily conflicting with my scientific/mathematical usage. Maybe I find this relatively easy because the software industry is well-known to have fluid & abstract terminology where everyday words are applied to unrelated and unanticipated situations -- sometimes inappropriately. :-)

    3. Certainly. For example the names of programming languages.