Indirect Evidence To The Rescue

by Harold Henderson | Aug 25, 2011

How I wish I could talk with my wife's great-great-great grandfather Samuel Wesley Boren (1828-1897) for a few minutes. If he could only tell us who his parents and grandparents were, it would be so great.

That's not going to happen. But sometimes we let this impossible dream shape our atttitude toward the evidence we can find - the censuses, deeds, vital records, probates, newspapers, pension files, letters, and all the rest. If a document doesn't tell us what we want to know, then we drop it and move on.

At some point we can't find any record that names the person or provides the relationship we're looking for. We may call that a "brick wall," but it's one we built ourselves. To change the metaphor, instead of waiting to be told, like a child in great-grandpa's big dark old house, we need to be sifting through the papers and looking for clues, like a grownup in great-grandpa's big dark old house.

In the official language of genealogy, evidence comes in two kinds, direct and indirect. Direct evidence is the kind I'd like to hear from Samuel. If we ask, "On what day did Verena die?" an obituary might tell us, "Verena died 12 March 1879." Of course, the obituary might be wrong, but it does answer the question all by itself. (Of course, a good researcher will not assume that the obituary is correct just because it's direct, but will compare it to other evidence.)

Indirect evidence gives a hint but not the whole answer. If Verena had no obituary, and the local court record said, "Verena's will was proved in court on 15 April 1879," that isn't direct evidence because it does not answer our question directly. Of course we could tweak our question a bit and ask, "Did Verena die before the 1880 census?" Then the court record would be direct evidence for that question. But it remains only indirect evidence of what day she died - some day before 15 April 1879. It might have been a few days or a few years, depending on the situation.

Similarly, if there were no obituary and no court record, and we found a deed in which her son John sold property on 4 July 1880 and described it as "what he had from his deceased mother," that would be direct evidence of her death but only indirect evidence of its date.

Why even bother with this distinction? Because it can shake up our research and improve our results. Most records genealogists deal with were not created to answer our questions. Vital records often come close, but common names can be a problem, and as we move back through the years such records become scarce or were never created in the first place. In general, the more distant the ancestors, the less likely we are to find direct answers to our questions about them, and the more we will need to think in terms of indirect evidence -- piling up clues from which we can prove a conclusion, or at least a probability. Often the two kinds of evidence work together.


Indirect Evidence Locks It In - The Goodrich Case

Sometimes pieces of direct evidence disagree and indirect evidence can settle the problem. Take the case of Chicago inventors and brothers Harry C. and Herman B. Goodrich. They were boys in 1838 when their father Levi died young in frontier Illinois. Their widowed mother Hortense soon died also.

I wanted to learn Hortense's birth name. Records called her both Hortense Barnum and Hortense Miller, and it wasn't obvious which record was right. Analyzing and evaluating conflicting evidence is a subject for another day, but finding more evidence usually helps. And in this case it was indirect.

The answer came from a series of newspaper articles published more than forty years after the event that never named or even mentioned Hortense. Reporters in the area became interested in a near-centenarian old settler named Asenath Miller. Her story (confirmed in other records) was that she had been married twice, first to Simeon Barnum and after his death in 1825 to Frederick Miller, and had had children with both. Some kind of family connection already seemed likely between Hortense and this family, because Frederick Miller had claimed reimbursement from Hortense's estate for care during her last illness.

The articles also identified Asenath as the grandmother of Harry C. Goodrich, by then a well-known inventor and public character. Obviously she had the wrong surname to be Levi Goodrich's mother, who was already accounted for in any case. So she must have been Hortense's mother. And in order for Hortense to have two sons by 1838, she had to have been born before 1825, and therefore she was born a Barnum. This conclusion fit with much of the direct evidence and provided a reasonable explanation why one record miscalled her by her stepfather's surname.


Indirect Evidence And The Genealogical Proof Standard - The Burdick case

In the Goodrich case, indirect evidence pretty much locked in the conclusion. The uncomfortable fact about indirect evidence is that things don't always work out so well. It can be difficult to know when you're done.

Chicago insurance agent Joseph Burdick posed such a problem. For a certification project, I wanted to know whether he was the same "Joseph Burdick" who served on the founding board of a Free Will Baptist church in the city in the 1860s, but whose name never surfaced again in its scanty records. I found several pieces of indirect evidence bearing on the question that suggest the two men were the same:

  • Joseph had lived near the church and several of its members.
  • A few years earlier, before moving to Chicago, he had lived near (and probably associated with) another man who was later on the church board.
  • This sub-denomination was known for its abolitionist and pro-Union stands, and Joseph's parents had joined a like-minded Methodist sect some years earlier.
  • No other Joseph Burdick appeared in Chicago city directories or census records in those years.

None of these clues directly answered my question, but together they made it probable that Joseph the insurance agent was also the board member. But just how probable? And how do we use indirect evidence to turn a probability - a hypothesis, if you will - into a genealogical proof?

The Genealogical Proof Standard is also a subject for another day, but one of its requirements is that we conduct a "reasonably exhaustive search" for evidence both pro and con. Another way to say this is that we try to disprove our favored hypothesis. A third way to say this is that we try to produce a written argument that will convince a more experienced genealogist.

At the time the American Baptist Historical Society records were inaccessible because they were in transit from Rochester, New York, to Atlanta, Georgia. I had ample material on Joseph's life, so I left this potentially interesting episode out of my paper on that occasion. And no, I haven't gotten to Atlanta yet.

The archive may well contain no information about Chicago's short-lived Free Will Baptist church. But it does hold the papers of two men who had some involvement in Baptist "home missions" in the 1860s and who might well have dealt with the church when Joseph was on its board. In my opinion, the search will not be "reasonably exhaustive" -- and my hypothesis unproven -- until and unless those papers are reviewed. Chances are they'll have nothing. Or they might just repeat what I already know, that he was a board member. Or they might refer casually to Joseph as "recently arrived from Vermont," in which case my lovely hypothesis would be in deep trouble. You never know until you look. Judging from the advice ABHS offers, its archivists often deal with genealogists who are looking for well-indexed, easily accessible direct evidence and nothing else.


Not Easy But Worth It

With direct evidence it is easier to know when we're done. When several independent original sources give first-hand information that directly answers the question, and they agree, we're good. Weighing and correlating indirect evidence involves more judgment. It also requires us to explain in a written argument what that judgment was. Ultimately we need to reckon whether our work will convince those who read it, some of whom may have been doing genealogy before we could spell the word.

So indirect evidence is not easy, but it's often what we have to work with. And no amount of copying out vital records can match the thrill of seeing the clues in an indirect-evidence case come together to build a conclusion that no record gives by itself.


More Help With Indirect Evidence

The Board for the Certification of Genealogists has the precise wording of the five-part Genealogical Proof Standard

Every issue of the National Genealogical Society Quarterly (available at any good genealogical library) contains articles that use indirect evidence to answer hard questions of identity and ancestry. They are more complex than my examples, so be prepared to read the articles more than once in order to understand them.


Sign up for a free trial account and begin tracing your family history today.

Start 7-Day Free Trial »