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

Tuesday, July 9, 2013

Beyond Evidence to Proof

Let's suppose that you are looking for your great-grandfather and his family and with a little effort, you find an entry in Ancestry.com showing, what you assume to be, the family in the U.S. Census record. Is the entry proof or evidence? What is the difference? And who cares? As usual, there are levels of awareness of different issues in genealogical research depending on the sophistication and background of the researcher.

First of all, it may seem simplistic, but evidence is not proof. What you are looking at when you find a source document is evidence. How you view that document and what you do with it becomes proof. The idea of evidence is as old as mankind. The Bible says, at 2 Corinthians 13:1, "In the mouth of two or three witnesses shall every word be established." Here, the reference to "witnesses" is used in the context that witnesses give testimony or evidence of a fact, but implies the principle that one witness is inherently unreliable and more testimony or evidence is necessary to establish any fact, i.e. prove something.

When you are gathering your family history, do you rely on one witness? What if those witnesses contradict each other?

Among those of a certain level of genealogical involvement, it is common to hear references to a genealogical "proof statement." As defined by the Board for Certification of Genealogists, "A proof argument is a detailed, written explanation of the evidence and reasoning used to reach a genealogical conclusion." Now, I would call this type of document an opening brief. In the context of a court proceeding, when a party files a complaint, opposing parties can file motions to dismiss or take other action. Obviously, there are an extremely complicated set of procedures available to attorneys. But the essence of any motion made in court is that the first party filing the motion makes an argument for that party's position in the lawsuit. The opposing party then has the opportunity to file a responsive motion and in most cases, the first party can file a reply. The judge in the case then makes a decision in the case, either with or without the opportunity for the parties' attorneys to make oral arguments to the court (i.e. the judge).

The essence of establishing proof in the context of a court action is the idea of advocacy. Each party gets the opportunity to present their "case" or position to the court. Their case includes any evidence they possess and any legal arguments they wish to make based on the evidence. Obviously, I could go on for pages and pages with all the nuances of legal motion practice but the idea here is that a motion is simply an argument made by one side of the conflict. Now, back to genealogical proof statements or arguments, the Board for Certification of Genealogists goes on to state,
The final step of the Genealogical Proof Standard requires that “we arrive at a soundly reasoned, coherently written conclusion.” Writing a proof argument satisfies that requirement. Proof arguments are frequently composed with the goal of publishing and disseminating information to invite challenge to a conclusion. Arguments defending our conclusions are also useful for our personal files or for clients. Very often the process of writing a proof argument will reveal to us weaknesses in evidence and logic where additional research is required.
Now, as usual, I would take exception to some of the things that are assumed by this type of procedure. Here are several issues I find with the whole process:

1. No conclusion in genealogy is "final." All conclusions are open to revision with the discovery of more evidence. If I find some evidence and write my "proof argument" there is always the possibility that I will find more evidence or that someone else will also. My conclusions may have to be revised. This is implied by the statement above that the proof statement or argument can "invite challenge to a conclusion."

2. Any "proof argument" just like an opening brief in a law case, is merely the statement of one side of the controversy. The person writing the statement (or the brief) selects those facts that support his or her conclusion and omits or minimizes the facts in opposition. Here, using the word "proof" is misleading. There is no proof, the statement contains argument and evidence. If you were to read the proof argument and believe it to be correct, you might consider the facts to be proved, but see No. 1 that no conclusion is final. In this vein, the very name of this type of argument, a proof argument, gives the impression that what a genealogist does is prove facts to be true. Although, I am sure many people would disagree, I take exception with that whole idea. Using proof in this context gives an impression of finality.

3. As I indicated above, proof involves advocacy and judgment. When I enter a name or event on a family group record or pedigree chart, I am acting as an advocate of that particular fact (evidence). Genealogical evidence is full of contradictions and inaccuracies. But in genealogy, the second part of the proof, the judgement, is missing. There are no genealogical courts and judges out there that will rule on whether or not the researcher's fact or evidence has been proved. I could make a perfectly formatted and acceptable proof argument (to the satisfaction of the Board of Certification of Genealogists or a journal editor) and yet be 100% wrong. The key is in the statement made in the same document quoted above, that states, "Do not omit contradictory evidence. When it exists, it must be presented and discussed." In the context of a court action, the opposing party will be quick to point out any omissions in the facts selected by the attorney in the brief, but in genealogy all of this is missing. The vast majority of genealogical proof arguments go unopposed.

I realize that this whole discussion is lost on most genealogical researchers. Very few genealogists rise to the level of making coherent, written proof arguments. In fact, most genealogists are probably unaware that such a thing exists. I am not advocating the abandonment of proof statements or arguments, in some contexts they may serve a valuable purpose, but I do take exception with the implications of using the concept that something has been proved by making those arguments. I would call them evidence arguments or something similar and leave the idea of formal proof to the scientists and lawyers.

12 comments:

  1. "I realize that this whole discussion is lost on most genealogical researchers" - possibly, but as a mathematician by training and IT guy by profession, I like it!

    Re your issues:
    "1. No conclusion in genealogy is 'final.' " Agreed. But the GPS never describes the conclusions as final. Sure there's a final *step* in the process, but as we all agree, since we can go round the loop again when we find new evidence, then it's only the final step on this pass. But certainly, we do need to all understand that we might go round again.

    "2 ... Using proof in this context gives an impression of finality." It might to some people but I'm quite sure that, particularly if we emphasise that we might go round again, then nearly everyone will say "But of course you might change your mind!"

    "3. ... The vast majority of genealogical proof arguments go unopposed" The GPS does talk about ensuring that "the conclusion reflects *all* the evidence" (my emphasis) along with "Resolution of conflicting evidence". So there is an attempt to ensure opposing views are heard. Nonetheless, there is a serious point you make here - in science, a proven theory is taken as read in the next paper to use it. In genealogy, we are continually urged to go back to the primary sources and not take previous conclusions are true. If genealogists were in charge of peer reviewing Einstein's Special Theory of Relativity, they would have said, "Yes, Herr Einstein, but you are building on Sir Isaac Newton's Laws of Motion without proving them for yourself". Not sensible. And the reason that Einstein could use Newton's formulae without proof, was that the peer group review process exists in science, so that science could be as certain as possible that the formulae were true and need not be proved again. Such a peer group review is replicated in the courtroom but seldom in genealogy.

    So certainly I would agree that proof statements, because of the lack of review, lack something. However - the big issue is - what would we call them instead? You say, that you "would call them evidence arguments or something similar and leave the idea of formal proof to the scientists and lawyers."

    But here's the point - no mathematician would consider scientists and lawyers ever created formal proofs. A formal proof is totally 100% true - and this can only be done in maths, and it can only be done there because it's done in a theoretical structure, where we don't have any of this messy real world uncertainty. (Note I do not claim that mathematicians never add 2+2 and get 5 or write the wrong formula down). Scientists don't produce formal, 100% certain proofs about the *real* *world* - the existence of the Higgs Boson (if / when proven) will be a statistical argument that says they are x% certain. But of course, there's still a 100-x% possibility they're wrong. That's not complete and final. Indeed the whole essence of the scientific method is that the discovery of anomalous results will take one round the loop again. Just like in genealogy.

    And as for lawyers? Well, given that the burden of proof is different in civil and criminal courts (usually?), how is that a formal proof (in a mathematical sense)? Well, it is by the tenets of its own discipline. Lawyers and scientists use the term "proof" in the sure knowledge that they have no final proof, and that there are rules of the proof game. Mathematicians and logicians, as guardians of the only genuine proof systems, are generally, conversational wind-ups aside, happy for physicists and lawyers to use the word "proof", because they have these rules. So why can't genealogists use the P-word - providing they have the rules in place?

    Well, I enjoyed your post!

    ReplyDelete
    Replies
    1. Very good points and some of the things I considered as I wrote. Well, I suppose there is no way to avoid using the word "proof" in any and all contexts, but I hope I made the point that making a proof statement does not necessarily "prove" anything except the opinion of the author of the statement. I am not sure anyone can prove a historical fact; we can only gather evidence and argue from that evidence.

      Delete
  2. I anticipated more reaction to this James since it's a very emotive topic.

    I came into genealogy from a scientific background, and so terms like theory, hypothesis, and proof have well-defined and long-standing definitions. I totally agree with Adrian that mathematics is the only discipline that has a real 'proof' concept. The term 'scientific proof' is a contradiction in terms that is used loosely but not in any deep scientific way. It generally means supported by experimental observation (i.e. evidence), but the next set of experiments could still overturn it. Hence, there is a strong parallel with genealogy.

    Each discipline has its own terms and definitions, including science, mathematics, law, and genealogy but those definitions certainly do not match.

    I used to be rather vocal about the term 'proof' being misapplied in genealogy but it is fiercely defended by people with a BCG background. I now just accept it but interpret it differently for a genealogical context, e.g. a reasoned theory that fits all the available evidence (whether direct, indirect, or negative), and fits better than alternative theories.

    This is all very well for professional and experienced genealogists but what about the newcomers, or even the clients of those professional reports? Do they always know and accept that a 'genealogical proof' is not necessarily a description of 'real-life truths'?

    Tony

    ReplyDelete
  3. I write software for managing evidence and proof statements, and one of the points I try to make is there is not 'final proof', just a 'current proof based on current evidence'.

    ReplyDelete
  4. James, your concern about the word proof in the GPS has been raised many times in the genealogical community. However, the phrase "proof argument" is clearly described as the process of gathering evidence, resolving conflicts and articulating the logic used.

    I am a recovering mathematician and scientist :) Thus I would add to Adrian's comments by noting that your belief in the formality of the review process in either law or science and the language chosen to prescribe that process is literally getting in the way.

    When I first started working on forensic cases, I was struck by the fact that the legal process was built on opinion not proof, just as I was struck by the frequency of reading "I would argue ... " by historians.

    Historical schools of thought are fluid. Erroneous legal opinions can prevail simply because the process says "That's final." Introducing new evidence is time sensitive. Even the unraveling of erroneous opinions with DNA science is still a process fraught with imperfect procedures.

    In a recent ruling I read that denounced the practices and arguments of an attorney, I was struck by how methodical and detailed the ruling was. Many rulings I have read are riddled with citations to other court cases to justify a part of the opinion yet does not hang together in a logical conclusion for the case as presented. Still that opinion stands until and unless someone finds the right series of procedures that will untangle the mess.

    However imperfect peer review may be in the genealogical community, it does exist, and as you noted, proof arguments are openly invited to be challenged with evidence and logic.

    Thank you for opening this discussion.

    Sharon Sergeant

    ReplyDelete
    Replies
    1. There is a set of books, often used by attorneys, called "Proof of Facts." What this set of books does is provide attorneys with the elements that need to be covered in a trial in order to comply with the legal requirements defining the elements of proof for any particular claim. Attorneys consider that they have "proved their case" when they provide documentary evidence or testimony that covers each of the necessary elements. As you point out, it is a little bit dangerous to regard this process as proof in any final sense of the word.

      Delete
  5. Debbie Parker Wayne, CG, CGL, posted on her blog the following about the GPS:

    "Conclusions that meet the GPS are always subject to reconsideration when new evidence is discovered.

    The GPS doesn't provide finality, even though that's what many are looking for." She quoted from Donn Devine, CG, an attorney and one of those who worked to develop the GPS when it replaced the use of the term "preponderance of the evidence" in the field in the 1990s.

    As to your point about a perfectly reasoned genealogical conclusion making it past the editors of a genealogical journal, I find it highly doubtful that it would get past the eyes of Tom Jones and Melinde Byrne at the NGSQ; Henry Hoff and Helen Ullmann at NEHGS Register; Laura DeGrazia and Karen Mauer Green at the NYGB Record; or David Greene and his team at The American Genealogist. And yes, all of them are, or have been Board-certified by BCG. And I daresay it may well have trouble getting past a lot of folks who do not want, and/or have not yet, earned the Board's credential.


    Debbie Parker Wayne, "Review: Finding Family: My Search for Roots and the Secrets in My DNA," Deb's Delvings Blog, posted 29 August 2012 (http://debsdelvings.blogspot.com/ : accessed 10 July 2013).

    ReplyDelete
  6. Should have said, "perfectly reasoned, yet flawed, genealogical conclusion" in my previous.

    ReplyDelete
  7. I am not a scientist, although I'm very curious about how things work. I am not a mathematician, although I loved math in school and did very well in my studies of it. I am not a lawyer, although I have been married to one (who is now retired) for a long time.

    I have always been curious about my family's origins and began my genealogy research seriously about twenty years ago. In all of that time I have never seen or heard anything which would lead me to believe that a "genealogy proof" was a "legal proof" or met the same standards. I don't have any genealogical certification, nor do I plan to get one. I try to live up to the highest standards of genealogy research out of respect for my family members who came before me and those who will follow. When I work with other genealogists, I choose to work with those who do careful work and avoid those who start their research one day and, two weeks later, post a seven- or eight- generation family chart.

    Having said this, my personal opinion of the above argument is that it is trying to mix apples and oranges and is, therefore, superfluous. I will just use "genealogy proof" and "legal proof" when appropriate, knowing that they do not meet the same standards.

    At my age, I'm interested in using whatever time I have left to get my work in order so that it is useful to the next generations, and to continue refining the work I've already done.

    ReplyDelete
    Replies
    1. Hmm. So the Emperor has no clothes. Now we just have to figure out who is the Emperor.

      Delete
  8. This is a very interesting topic. Proof is always troublesome. I have posted a blog post on my study of Mastering Genealogical Proof that you might find interesting tying the concept of research and proof to an episode of Law & Order. It's at: http://200inparadise.blogspot.com/2013/07/law-order-research-vs-proof-in-genealogy.html

    ReplyDelete
  9. James, this post comes at a great time in my genealogy journey. I am a young genealogist (30 years old) and have only been researching my family history seriously for a few years, so these are just comments from a young rookie. :)

    I am constantly torturing myself over the "proof standard." I find that I *want* things to be final, but come across so many conflicting records and dates, that I realize it is impossible. I turned to my husband the other day after making a new discovery on something I thought was "proven" and I said, "All this work I'm doing...it's all just a big guess, isn't it?"

    After I read your post, I realized that I should not stress myself looking for "final" information. Just because something has been "proven" does not mean it is correct and that's ok! There will always be another crop of evidence to call everything I know into question. Sound arguments as to why I believe what I believe will have to suffice.

    Appreciate your post!

    ReplyDelete