The Sagan Doctrine

I ran across a quote recently that got me thinking about the extraordinary. In an episode of Cosmos Carl Sagan said,

Extraordinary claims require extraordinary evidence.

It is a catchy phrase. I think it is fairly clear what that little statement is trying to get across, though it perhaps isn’t so precise. What is missing are criteria for what might be meant by “extraordinary claims” and “extraordinary evidence.”

Sagan’s phrase seems to be a descendant of a quote from the philosopher David Hume,

A wise man, therefore, proportions his belief to the evidence.

In the hands of mathematician Pierre Laplace the statement became

The weight of evidence for an extraordinary claim must be proportioned to its strangeness.

Sagan’s version is catchier but it is not exactly rigorous. We might think of an extraordinary claim as a claim for something that is highly unusual, maybe even to the point of having never before been observed. If we skip what “highly unusual” might mean in exact terms and simply continue with the idea that probability is important, then we might ask what an extraordinary claim in genealogy would look like.

Extraordinary Claims

I’ve thought about what types of extraordinary claims I’ve seen. I’m sure my categories aren’t exhaustive and some extraordinary claims will fall into more than one category.

One area for extraordinary claims might be the biologically extraordinary. Young children and old women can’t normally become parents and though possible, a man isn’t likely to father a child at 90. Living to well past 100 is so rare that it is extraordinary. Children born to the same mother eight months apart is not exactly common. One hopes that claims of long dead people producing offspring are accidental but I’ve seen them so I’ll put them on the list.

There is also the legally extraordinary. Though there have been places and times when prepubescent children could marry, in most cases that people are likely to deal with, it is not legal. Bigamy falls into this category as well. It certainly happens but the accusation shouldn’t be made without good proof that the simultaneously twice married person isn’t really two people with the same name.

Some things are culturally extraordinary. That is they may be normal at one place and time but put them into another culture and they seem strange indeed. Name changes are common in some cultures. A person’s name might change significantly depending on their stage in life, their occupation or their personal preference. In other cultures, a change in name might be a real oddity. Divorce might be something that just does not happen in one culture and common and accepted in another. A claim of a divorce in a culture that was not even remotely accepting of divorce is probably an extraordinary claim.

“Gravitationally” extraordinary. Gravity is the metaphor for how I think of claims that attract claimants. Claims of descent from the famous and powerful go here simply because such claims are so tempting to some, that they are made far more often than they could possibly be true—like Medieval claims about possession of a piece of the one true cross.

There are historically extraordinary claims as well, such as a claim of descent from a historical figure that left no known children. A claim that a person was born in a place that as far as is known did not yet have settlers of the right ethnic group. Descent from people involved in historical mysteries is another historically extraordinary claim. Who wouldn’t want to be a descent of the Roanoke colonists? Genealogical claims that would change our view of history on a broad scale should be backed by extraordinary evidence.

Extraordinary Evidence

What might extraordinary evidence be? The Laplace version of our phrase actually uses the term “weight of evidence.” There are a few things that can add to the weight of evidence. Like the weight of a child’s bucket at the beach as it is filled with many grains of sand, we can fill our bucket and increase its weight with a lot of independent pieces of evidence. Eventually, the quantity of independent evidence to back the extraordinary claim starts to be come extraordinary.

The quality of the evidence can also nudge it toward or away from the extraordinary. If an extraordinary quantity of evidence is filled with chicken scratch from the margins of a family Bible of unknown provenance, smudged census records and some out-of-focus microfilms of a questionable county history, right or wrong, the scales will tip well away from extraordinary.

Another way to make that child’s beach bucket heavier is to put a big rock into it. If you’re trying to show that John Doe was a bigamist and you have evidence of a John Doe marrying Polly Primary and evidence of a John Doe marrying Sally Secondary, there would probably be no bigger rock, no more extraordinary evidence, than the court records of John Doe’s bigamy trial.

An extraordinary collection of evidence might also contain evidence that is not for the claim but rather that is against simpler hypotheses. Evidence that goes against those hypotheses that Ockham’s razor would otherwise identify as being more likely than the extraordinary claim.

Any responsible search for evidence in support of a claim should include checking for contrary evidence. Perhaps we can say that the evidence takes a step toward the extraordinary by containing a conspicuous lack of contradictory evidence. An absence of evidence that goes against the extraordinary claim helps to make the evidence extraordinary.

As Hume put it over 250 years ago, matching claim and evidence is all a matter of proportion.