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

Friday, June 14, 2013

Methodology vs. Proof in Genealogy

Proof is a slippery subject. How do I prove anything to a person, if that person does not want to accept or believe the proof I offer? The world if full of people who believe all sorts of things based on their own concept of proof such as Aliens, UFOs etc. So how do we approach the idea of "proof" in genealogy? Doesn't all proof come down to who is convinced of what?

Currently in the greater genealogical community, proof statements are in vogue. The process of producing these "proof statements" is summarized a FamilySearch Research Wiki article entitled
"Evaluate the Evidence":
Evaluating the evidence takes place in several phases. It starts as soon as you find a document that must be evaluated to see if it is relevant to the family you are researching. It continues as you transfer the information from the source to your genealogical records and compose a source footnote. In that phase you are evaluating in two ways: (1) a preliminary evaluation of the reliability of the source, and (2) a comparison and contrast of data on the source with other information about the family to see if it corroborates or contradicts other sources. Later, after research on the family is mostly completed, carefully make a final, well-reasoned re-evaluation of all the sources compared to each other to help you reach a reasonable conclusion and write a proof statement subject to the Genealogical Proof Standard.
There are essentially five elements to the Genealogical Proof Standard:
  • A reasonably exhaustive search has been conducted.
  • Each statement of fact has a complete and accurate source citation.
  • The evidence is reliable, and has been skillfully correlated and interpreted.
  • Any contradictory evidence has been resolved.
  • The conclusion has been soundly reasoned.
See also: Board for Certification of Genealogists (Washington, D.C.). The BCG Genealogical Standards Manual. Orem, Utah: Ancestry Pub, 2000.

Now, we are back to where we started; differentiating between methodology and proof. When I say "methodology" I mean that we are establishing a "fact" by simply looking for "evidence." The "fact" is found by a "reasonably exhaustive search" which is an entirely subjective measure which is further defined as assuming "examination of a wide range of high quality sources." Stripped of the verbiage, all this means is that you look at sources and when you feel you have looked enough, you quit and write about what you have examined. That is a methodology.

Proof means something completely different. I question whether or not the concept of "proof" really has any place in the area of genealogy at all. In one sense, proof is a logically based argument that establishes the truth of a given statement. This is the mathematical approach to proof. In order to establish a proof in this context it is necessary to have a formal system of logic that defines each term used and is rigorous in its application. I don't think genealogy has ever approached this definition of proof. It is a social science based on historical evidence, not mathematics or physical science. The polar opposite definition of proof involves presenting arguments in favor of a proposed fact and convincing the "average" person of its validity or truth, if you will.

The second definition of a proof is a little more closely related to the genealogical concept except the judge and jury are missing. There is no "average genealogist" who can be convinced or not convinced of the truth or falsity of any given fact because the adversarial process of convincing the average person is missing. In genealogy, all of this evaluation is left to the individual researcher. To repeat, there is no genealogical judge or jury other than the inertia of those who might be interested in any given "proof statement." In addition, genealogy is not per se adversarial. Although there are situations where individuals disagree on the evidence, ultimately, there is no way to resolve such disagreements if the individuals never change their opinions.

You could go through each step of the proof standard, including writing your published proof statement and I could come along and do more research and shoot down your whole argument. Genealogy is methodology and cloaking it in the robe of proof does not accurately reflect the processes. Finding a reliable source does not prove anything, it merely adds one more level of evidence. Your conclusions are your conclusions. My conclusions are my conclusions. We may agree that there is no basis for questioning any given proposition, but we may both be wrong. The idea that we have proved anything is an illusion based on semantics not proof.

Time for a hypothetical to illustrate the idea. Ronnie Roe is born and lives with her two parents until they die when she is 40 years old. All of her life she has known that Robert and Jane Roe were her parents and she even has a birth certificate to prove it. Two years after her parents' death, she finds out that contrary to all of the evidence she had, she was adopted. My point: all genealogical proof is methodology and subject to revision. We can set up an arbitrary system of proof that depends on evaluating all of the evidence we have found so far, but we must always be ready to revise our entire set of proof based on subsequent evidence. If you take any other position, you are deluding yourself.

Current proof statements may be widely accepted by the academic genealogists but are only true so far as no one else has challenged them. We are not talking about science here. I can't conduct an independent experiment and validate your findings. In genealogy, I have to rely on your integrity and ability. How do I know when you wrote your genealogical proof statement that you didn't just make up the entire thing or ignore obviously contradictory information? How many such proof statements are there that exist in family trees merely because no one has every challenged them? Genealogy is methodology not proof. All proofs are tentative.

I am sure there is more later especially about the issue of whether or not I have to review all of your sources. Remember, genealogist have long been known to have made up or falsified facts to support their positions. Not to question the integrity of today's experts, but to show that relying on past records is not a science at all.


  1. But then again, the only discipline that has real genuine proof, is mathematics - which sits outside the real world. All else in science is just statistics - the "proof" of the Higgs Boson will be a result from a statistical analysis, and it will say that physicists are X% confident (actually that's not the way it's expressed but it'll do) - and this still means that there's a 100-X % probability that they're wrong.

    In other words, they have a methodology that produces a "proof", that could still be wrong.

    I don't really see that this is dramatically different from genealogy.

    To my way of thinking, it is not a case of "methodology v. proof". The two are different - methodology is how you do something, proof is the output of the methodology. Yes, you are very right to worry about the degree of proof or the quality of the methodology - but what else do we call the output if not "Proof"? It's a simple word and (justified) worries about the standard of proof shouldn't stop us using simple words. After all, if it did, as a mathematician I'd stop the legal profession from using the p-word!


  2. You said "We are not talking about science here. I can't conduct an independent experiment and validate your findings."

    Two years ago this argument may have been mostly true.

    Today, genealogists do have the ability to scientifically verify the results of their paper "proofs". DNA does not lie. Autosomal DNA testing and the tools developed for working with this data is available now. It allows us to scientifically prove the validity of our methodology findings.

    It is not cheap ($100 to $200 range depending on current sales), but it is getting there very fast. It is already less then the cost of traveling to a distant repository.

    It does require the participation of other descendants, but social networking and on-line trees make them easier to find then ever.

    It does require time and does not replace paper based proofs, but it certainly is science and it can produce your "proof".


  3. OK... it get it. From both sides of this argument. But it appears that the one essential ingredient is the idea that genealogical findings can't be mathematically or statistically stated.

    We use words like probably or likely or possibly and so on, but why not apply more concrete evaluation to our results or "proofs"? We can't we quantify ideas like "direct is usually better than indirect" or "original is preferred to derivative" or "primary is more valid than secondary"? Are we so fearful that a statistical or numerical expression of our had work will lessen its value?

    I don't mind the word "proof", but I would prefer "statistical likelihood" better. I have stopped trying to argue my case, and I have begun to find ways to express my results from that point of view.

    I do this with my research using an arbitrary scale that puts my proofs at low percentages, and that's fine by me. DNA is presented the same way. We have a certain statistical likelihood that we have a common ancestor with another person. Why not do the same with our proofs?

    Numbers should not frighten us. They can be of great help to show us how strong our proof might be; how much more we must do to improve it; and provide a more precise way to apply methodology to our hard work.