Evidence and Proof
By Phil Albro*

        I suppose that those who take an interest in tracing their family tree would like to be able to prove the correctness of every linkage they report, every date, every statement about an individual. They would like to have personally verified evidence in sufficient quantity that they can say "yes, I'm certain!" about every claimed fact in the tree. But exactly how much evidence is that? It is easy enough to accept that proof derives from evidence, and evidence must be verified as to its existence and nature before it "counts." What is the role of authority in all this? Is an official document irrefutable evidence?

        "The proof of the pudding is in the eating." That old saying sums it up perfectly. "Proof" is subjective. It takes a certain amount of evidence to convince you of something, but a different amount of evidence to convince your brother. One argument may be highly convincing to you, but leave someone else completely skeptical. Some people are highly accepting of authority; they will believe something or not depending on who said it (or published it, or endorsed it.) "Well, she ought to know if anyone does!" can be heard in most discussions of matters of opinion. At the other extreme you have people who will say "You found his tombstone? All that proves is that there's a stone there. Did you dig up the coffin and have the DNA analyzed?" (Even then, it would have to be compared to DNA taken when he was alive, and correctly identified and labeled at the time. Can you prove it wasn't mislabeled?)

        In short, there is no such thing as absolute proof. There is evidence, adequate to convince. That is as close as one can come to the imaginary concept "proof." One can "have ones mind made up" such that no amount of contrary evidence will ever convince him to change it; in that case it is easy to see that the concept of "proof" is simply that, a theoretical concept. We also speak of "the burden of proof", a legal concept. In this case, a statement is made that describes something possible, even probable. We can say "the burden of proof" is on those who would refute it. They must produce evidence indicating that, probable or not, it is in fact wrong. I will return to this "burden of proof" concept, as it seems to be to be highly relevant to genealogical research.

        First, let us look at the issue of evidence . Evidence is worthless if you don't know for sure that it really exists, or can't demonstrate to someone else that it really exists. You can say "Here's what it said on the church record ..." and give some words that you read from your notes. What is your evidence that you didn't misread the record, or make a copying error? For that matter, how do we know you didn't make up the whole story? Before duplicative copiers (photography, Xerox, scanners), people tried to provide this evidence by having more than one person make independent pen-and-ink copies, then comparing them. This is an effective system, if the two copies are indeed independent. The principle is, there is only one way to get it right and an infinite number of ways to make a mistake, so if two versions agree, the odds are they got it right. We would find that method unendurably tedious today; fortunately we have faster methods available.

        You will hear much about primary and secondary evidence. Once again, this is a transfer of a legal concept into the research arena. Primary evidence is the testimony of an "eye witness", someone who was present at the event and in a position to know the facts. Secondary evidence is everything else. Unfortunately, there is a tendency to imply that if evidence is primary, it must be reliable. Ever listen to the testimony of a dozen eye witnesses to the same event at a trial? No two will agree. Ever have your own words quoted back to you by someone who heard them directly? Was it a match? Probably not. People are not recorders and cameras. Our perceptions and what gets stored in memory are colored by everything else happening around the same time, guesses we made to fill in portions of the event we didn't see, and our emotions at the time. The real reason why we should place more confidence in primary than in secondary evidence is simply that it has a higher probability of being correct. There is no certainty, just a higher probability. Even that has no scientific or legal basis beyond the fact that, through the years, we observe it to be the case. "Seeing is believing." Believing, yes. Knowing for sure, no.

        "Weigh the evidence." Good advice. There are two implications there; first, that what is presented as evidence must itself be put to some sort of test to see if it is valid or not. Secondly, there is "weighty" evidence, evidence of a kind that is usually found to be reliable when tested, and there is "light" evidence, evidence of a kind that experience shows often turns out to be spurious and misleading. Whenever there is a simple choice to be made between two possibilities, one in effect adds together the weight of each item of evidence on the side of each possibility. The side with the "heavier" total of supporting evidence is the side that ought to be accepted as most probably true. In law, this concept is the "preponderance of evidence" that decides who wins a civil court case.

        When we consider evidence we must also be careful about exactly what it is evidence of ! You may say "I have seen a letter from his 8th grade teacher saying he was very intelligent." Then you say where this letter can be found. Four other people, none of them your close relatives or in your employ, go there and also report that they saw the letter, and that it says what you said it says. Now we have justification for believing this evidence. But we have no particular reason to believe he actually was intelligent, merely that his 8 th grade teacher thought he was. Now we need evidence that the teacher (a) always expressed her opinion bluntly and honestly, and (b) was a good judge (by some objective standard) of a student's intelligence. This is an easy error to make – evidence that someone actually said something is not evidence that what they said was true. This is the difference between verified evidence and validated evidence. When you verify evidence, you show that it exists as originally claimed. When you validate it, you show that what it claims is in fact correct and accurate; that is, you provide supporting evidence . Valid evidence is not contradicted by other, equally strong evidence, and violates no rules of logic.

        Let me return to the "burden of proof" concept. All scientific research is based on the Principle of Hamlet and the Principle of Laplace. The former says "All things are possible." The latter says "The more improbable something is, the stronger the evidence you need before you should believe it." If there is general acceptance of some statement about one of your ancestors, no one has ever proposed a contrary statement, and in general there is no dispute, then very little evidence is needed to justify your belief in that statement. The situation could change in the future if someone eventually presents a contrary view, but "sufficient unto the day is the evil thereof" – it seems to me you are justified in saving your finite time and energy for research on those issues that already have conflicts. Let "the burden of proof" be on that imaginary future person who decides to deny what is presently accepted. In fact, you have no choice, since you do not know what arguments this person (who may never exist) will put forth. In short, your efforts at collecting evidence must depend on the probability that what is currently believed or claimed actually occurred.

        If the claim at hand is that Archibald's parents were Theobald and Minerva, and Minerva would have been 60 years old when Archibald was born, you need strong ("weighty") evidence to support that claim before you should believe it, as it is not very probable. I suggest as a definition of "improbable", "likely to be quite rare." We're not into advanced math here. I have seen family trees on the Internet that reported a marriage occurring in 1768, and in the same record, stating that the "groom" died in 1767. I find this sufficiently improbable (especially since they proceeded to have a child in 1769, according to those reports) that I can't really visualize the weight of evidence it would take to convince me that it was all true. I do not feel that I have the "burden of proof" that there is at least one error in this report – in cases like this, the "burden of proof" is on the person who is claiming something that on the face of it is highly unlikely (improbable.)

        Finally, I want to say something about official documents. I could use the example of my father's official birth certificate from the state. The birth record provided by the hospital does not agree with the one provided by the state. The disagreement happens to involve the middle name. Is the official birth certificate "weightier" evidence than the hospital document, because the former is what would count the most in court, toward getting a passport, etc.? The hospital record happens to match what both my father and his parents wrote on every relevant document throughout his life. There are two possible viewpoints – (a) the birth certificate has his "real" name because it would be the version accepted by the law, and (b) the other version is his "real" name because it is how he was known all his life, even by himself. Keep in mind that the intent is to record on a birth certificate the name assigned by the parents; but while the hospital certificate is in fact filled out by the parents, the official birth certificate is not. The attending physician, if any, may be the sole arbiter of what appears there. I would point out that any "official" document can certainly be refuted in court, and can be declared invalid if a sufficient "weight of evidence" is brought forth opposing it. That it was "official" simply identifies who has "the burden of proof", namely those who would refute it.

        The bottom line is, think about who has the "burden of proof" in genealogical issues. I suggest that it is the person or authority who is making the least probable claim. I have seen genealogical researchers dedicate a lifetime to the accumulation of more and more evidence for what was never in contention in the first place, or for what needed very little evidence because it was so highly probable. This serves no useful purpose, and limits what can be accomplished.

~~~~~~~~~~~~~~~~~~O~~~~~~~~~~~~~~~~~~~

Note about the author: Phil Albro is a retired research biochemist who has replaced biochemical research with research into his ancestry. He has published approximately 140 articles in research journals, and sees no fundamental difference between valid biochemical research and valid genealogical research. He teaches at a state university, and his course includes the difference between scientific evidence and legal evidence. He works in a bona fide dungeon, and has a lifelong love of science fiction. He believes that excessive credulity and excessive skepticism can both be detrimental to any kind of research. He invites disagreement, and values the willingness to express it.

[Return to Main Page]

[Return to Research Menu]

Send Email to:     .

 

Copyright © 2000 Phil Albro. Commercial use prohibited.