Recent Posts

Read a Random Post



« | Main | »

Don’t Believe It!

By Daniel Hubbard | September 12, 2009

The more research I do the more interested I become in what it is even possible to know. It is not an insignificant question. Genealogists, especially beginners, often oscillate back and forth between believing they know the absolute truth to being totally unsure, depending on whatever information has just turned up.

Proving anything that is not trivial to be true to 100% certainty is never truly possible. Disproving things is much simpler. We prove our genealogical connections with documents and memories. We might support them with a DNA test but are we sure that they are correct? Can we be sure?

Surely the gold standard for proof of parent-child relationship is the DNA test. Not the kind used by genealogists—they show relationships over a great amount of time but without great precision at the level of individuals. If a son and a father take a Y-DNA test, it will, presumably, show that they have the same y-chromosome but so will a potentially huge number of male relatives. A DNA based paternity test can disprove the idea that a man is the father of a child but for a positive test result to be considered conclusive, it only needs to show 99% certainty. Clearly we can’t go back and test every parent in our files to see if every child we’ve assigned to them is biologically theirs but, even if we could, we would not achieve 100% certainty even then. Even with piles of documentation we can never reach the level of certainty of DNA testing.

What then do we mean by proof? Is it all hopeless? Hardly. To take an extreme example, there are many things that are theoretically possible but that are so unlikely that even given a billion times the lifetime of the universe the chance that any one of them will occur is far, far too remote to take seriously. It is theoretically possible that a glass of water left out on a table all day will suddenly boil but it is so magnificently unlikely that to say it is impossible is true in all but the absolutely strictest sense. This is why in law one often hears the standard “beyond a reasonable doubt.” That implies that 100% certainty is not required, it is not possible after all, but only that there is sufficient certainty; ideally just the “unreasonable doubts” remain.

When you start in genealogy, you run into the (in)famous statement “But I read it in a book…” Now, of course, I could add the modern version “But I found it on the internet…” Both imply that seeing something in writing gives absolute certainty. The short version of the problem here is that not everything that gets written down is true, it can’t be. The long version would cover many posts, and perhaps it will.

Not everything written down by record keepers, ancestors and family historians is correct. This isn’t always an easy point to make with people. After all, why write it down if it isn’t true? Why print it if it isn’t thoroughly reliable? There are many, many answers to those questions, ranging from simple error to outright lying. Even a perfect genealogist will produce inaccurate results if only inaccurate information exists.

I like to show that the rule “It’s true because I saw it in writing” isn’t a particularly good rule with a few amusing counterexamples.

Clearly, these records require some supporting information before anything useful can be extracted from them. Then there are all the records that in combination are clearly odd. They might show people aging six years during the ten years between censuses or parents who age five years between the births of children born two years apart.

If such obvious oddities lurk in records, how many errors too subtle for easy detection are hiding?

Twitter It!

Topics: Genealogy, Methods | 1 Comment »

Twitter It!

One Response to “Don’t Believe It!”

  1. Discover, Disbelieve, Disprove, Repeat | Personal Past Meditations- a Genealogical Blog Says:
    February 26th, 2010 at 11:45 am

    […] may not feel so exciting but it can be the right way to go. To prove something absolutely true is not necessarily possible but to prove it false can be trivial. Found a recent immigrant in the first census that could […]