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

Sunday, February 8, 2015

The Ins and Outs of Evidence for Genealogists -- Part Eight: Conclusions

In this last post in the series on evidence, I have taken some time to reflect on the issues I have raised in previous posts on this subject. It is true that the terms "evidence" and "proof" in their various iterations are firmly entrenched in the genealogical landscape. Even though they share a lexigraphic (See Webster's Revised Unabridged Dictionary, published 1913 by C. & G. Merriam Co.heritage to legal, scientific and other disciplines, the two terms mean something substantially different in the context of genealogy and for this reason alone, they are inappropriately applied to genealogical research.

Both of the words, "evidence" and "proof," are also used in common speech to refer to concepts that are really neither proof nor evidence in any strict sense of either word. As I have pointed out in my previous posts, these terms as used by genealogists were only relatively recently adopted. The rapid adoption of these terms and the accompanying pseudo-legal analysis that seems to accompany them, has not added anything to the level of discourse among genealogists or improved the degree of accuracy obtained by the vast majority of the genealogical community. In my opinion, the use of the terms, especially in conjunction with other legalese, has had the effect of making serious genealogical research even less attractive that it has always previously been to both the academic and non-academic communities.

Trying to stem the tide of the use of a common term is virtually impossible. I do see a glimmer of hope when a widely accepted and authoritative new book comes out and decries their usage. See Anderson, Robert Charles. Elements of Genealogical Analysis. 2014.

Fundamentally, genealogists are concerned with history and historical sources. These historical sources are finite and scattered across the world. The activity of genealogical research is to find these scattered sources, extract the information contained therein about a limited set of ancestral families and then analyze, evaluate and record that information in some reproducible format. This activity is highly personal in nature. There may be opportunities for collaboration, but most genealogists are involved in a solitary pursuit. The fact that genealogists record their information publicly does not particularly imply that they arrived at their conclusions any differently than they would have had they kept the information to themselves.

What the genealogists normally refer to as "evidence" is nothing more or less than the information contained in the sources they are able to locate. It is not "evidence" in either a legal or scientific sense. When genealogists conclude that a certain amount of information "proves" one of their conclusions; their conclusion may be entirely indefensible. Adding the gloss of "primary" and "secondary" evidence adds nothing to the process of evaluation. In the legal sense, primary and secondary evidence implies a level of reliability. If you think carefully about the process of examining genealogical sources, you will come to the conclusion that even a document created at or near the time of an event by a person who participated in the event may be far less reliable than an official document created at a time remote to the event.

For example, not too long ago, I examined a marriage record that gave the ages of the bride and groom. From other documents created at a time many years after the wedding, I found that the groom had lied about his age. This example illustrates the fact that proximity in time to an event does not guarantee accuracy. However, the distinction between primary and secondary evidence in the legal sense would seem to make a contrary argument. In fact, the terms primary and secondary sources are used entirely differently in an academic context. Primary sources refer to those recorded by a person present during the time period of the particular event. Secondary sources however, are sources that interpret or analyze primary sources. See "What it is a Primary Source?" There is a danger that genealogical researchers will apply the concepts of primary versus secondary sources without making a distinction as to the real reliability of the source's origination.

It is perfectly possible to carry on a coherent discussion concerning genealogical research without ever mentioning either evidence or proof. When someone claims to have "evidence" of a certain genealogical fact about their ancestors, they are in fact, only expressing an opinion based on their interpretation of the sources that have examined. The term "evidence" is used merely for the purpose of lending credence to the arguments in support of a particular personal opinion. Likewise, use of the word "proof" in any of its forms implies a defensible conclusion rather than the reality of a personal opinion. Examine the following two sentences as examples of what I am talking about:

  • I have examined several sources and believe that John is my grandfather.
  • I have examined the evidence and proved that John is my grandfather.
These two sentences are absolutely based on exactly the same degree of personal certitude and research. But they convey vastly different impressions of what was done and what was concluded. I can carry that particular example a little further with the following sentence:

I have made a reasonably exhaustive search and have concluded beyond a reasonable doubt that John is my grandfather.

In reality, the three sentences talk about exactly the same activity. Dressing the statement up with additional verbiage adds nothing to the reliability of the conclusion. There is no doubt, that I could do exactly the same reasonably exhaustive search and come up with an entirely different conclusion, even though there was a prior claim of my inability to create a "reasonable doubt." In all three cases, there is no way to determine whether or not I have gussied up my conclusions.

What we call evidence is nothing more or less than a highly personal opinion. This is especially trud when the words "evidence" and "sources" are used interchangeably. Genealogy, when all is said and done, proof is personal opinion supported by sources. To a great extent, the reliability of the opinion is based on the reliability of the sources. However, the evaluation of the sources is entirely under the control of the person expressing the opinion. I could examine the same sources cited by one of my genealogical colleagues and come up with an entirely different personal opinion as to the "facts."

Because genealogists often base their opinions on those of historical records created by people who may or may not have had accuracy as a goal, there will always be a substantial measure of uncertainty and unreliability in the genealogical conclusions reached by any researcher. I make this statement in light of the fact that I've recently investigated information that was provided to me by my great-grandmother who directly recorded information given to her by her father concerning his birth that turned out to be both wrong as to the date and as to the place that the birth occurred.

Are we then cast adrift in a boat without a rudder? Not exactly. If we follow a consistent methodology we will produce reliable results. That is, if the methodology involves an acknowledgment that any conclusion is subject to revision from later acquired sources. This fundamental fact in the life of all genealogists is obscured by the use of inappropriate terminology.

Here are the previous installments:

The Ins and Outs of Evidence for Genealogists -- Part Seven
The Ins and Outs of Evidence for Genealogists -- Part Six
The Ins and Outs of Evidence for Genealogists -- Part Five
The Ins and Outs of Evidence for Genealogists -- Part Four
The Ins and Outs of Evidence for Genealogists -- Part Three
The Ins and Outs of Evidence for Genealogists -- Part Two
The Ins and Outs of Evidence for Genealogists -- Part One


  1. "If we follow a consistent methodology we will produce reliable results" - sorry, I don't buy this. All we will produce from a consistent methodology is a set of *consistent* results. Whether or not those results are reliable is another matter.

    For me the issues major on things like inappropriate use of "same name, must be same person"; lack of critical analysis of sources; giving up and assuming something must be true because examining it further is too much like hard work, etc.

    Whether or not we use words like "proof", "evidence", etc, doesn't seem to matter to me if the fundamentals are wrong.

    1. I recently came across a source stated to be an autobiography, but when I started to examine the information mentioned in the document, things got really strange really fast.

      I puzzled and puzzled over the discrepancies and described the whole situation and difficulties to one of my teenagers on a long car ride, and while discussing it with him, I finally realized the "autobiography" wasn't. The only explanation for the discrepancies was that it was written by someone else decades after the supposed author's death.

      So, yes, we need a consistent methodology, gathering personal and community histories, for example; but as you say we also need to use critical analysis of the sources and not take them at face value.

  2. Paternity is an interesting final example. Given the realities of human life, conclusive proof of that fact would need to include DNA evidence. Before DNA evidence, establishing paternity has always been something of an art form.

    I’m thinking of the recent DNA test for Richard III. Given the DNA and location of the skeleton and physical proof, statisticians estimated that the chance that it wasn’t Richard III was 6.7 million to 1.

    With odds like those, can we state beyond reasonable doubt that it is Richard III? Probably. When the rest of us do genealogy are we working with those kind of odds? Unlikely.

    Anyway, I’ve enjoyed the series. Thanks!

    1. Amy, you say, "Paternity is an interesting final example. Given the realities of human life, conclusive proof of that fact would need to include DNA evidence. Before DNA evidence, establishing paternity has always been something of an art form."

      The notion of 'final example' is really quite subject to iffy interpretation. Say John Doe believes that he sired several sons and Y-DNA testing of descendants of each son results in a high degree of matching. Such results do not exclude paternity by John Doe's actual father, by a paternal uncle, by a first or second cousin, or of all of the sons by a neighborhood philanderer. Common male ancestor, very probably; but *who* is not proven by such results.

    2. You point out exactly why DNA when used with genealogy requires independent research on the lines.

    3. Geo, by "final example" I meant the example James chose for the final post in the series, nothing more.

      Probably the best example of what you're talking about is, of course, the Jefferson-Hemings family. In that case the DNA evidence alone couldn’t prove Jefferson’s paternity, but the documentary trail for the Jefferson family could, since the records are extensive enough to prove which possible father had access to the mother before the birth of each child.

    4. Amy, the documentation available for the Jefferson plantation and for the Hemings' was quite extensive and detailed.

      It would be quite unusual for there to be documentation that would shed light on the circumstance(s) I posed in the case of the Y-DNA matches, particularly if estate or land records reflected the belief of John Doe that he was the father of his wife's sons.

      Y-DNA testing was not available for the Jefferson-Hemings group.

  3. James, you wrote, "Trying to stem the tide of the use of a common term is virtually impossible. I do see a glimmer of hope when a widely accepted and authoritative new book comes out and decries their usage. See Anderson, Robert Charles. Elements of Genealogical Analysis. 2014."

    One could be more specific about what Anderson actually wrote. In Appendix B he did criticize genealogical use of the term "preponderance of the evidence," but uses the term "evidence" throughout the book and at the end of Appendix B where he gives a very general description of his approach.

    1. I said it was a glimmer of a hope. :-)

  4. Thank you for this interesting and thought-provoking series. I don't have any problem with the use of the word "evidence" but like you I do think the use of the word "proof", especially as it is used in US genealogical circles, is problematic. The word proof is used not just in the law but in science and in common parlance. Proof implies a high degree of certainty, but proof is generally elusive. I think it only confuses people to use the word proof in something like the Genealogical Proof Standard which essentially describes a methodology and not a means of proving or disproving a fact. It would be more appropriate to call it the Genealogical Research Standard. I'm also puzzled by the requirement in the GPS to write a "soundly reasoned, coherently written conclusion". If we were to do this for every single link in our family trees we wouldn't get very far with our research, and I really don't see why writing a lengthy conclusion makes any difference to the quality of the research. Somebody can write a long essay and still come to the wrong conclusion. There is no need whatsoever to write "I have examined several sources and believe that John is my grandfather" or indeed any of the longer versions that you cite. For the average family historian entering a fact into a family history program with links to the sources cited is all that is required, perhaps with a few explanatory notes where further clarification is needed. Following a consistent methodology is no guarantee that someone will produce high-quality research. Reproducibility is the only true test of your research. Can someone else who is researching the same family tree as you locate the sources you have used and perhaps supplement them with other sources and come to the same conclusion as you? Often it is better to approach a research problem from a new perspective.

    1. Very good observations. Thanks for reading the series.

  5. "I believe that John is my grandfather" is problematic on the face of it. But that is because I object to the use of "believe" when the correct word is "think."