The dog that did bark

(Crossposted from The Legal Genealogist)


The Legal Genealogist (Judy G. Russell, JD, CG, CGL) had an absolute ball last night giving a webinar for the Board for Certification of Genealogists, through Legacy Family Tree Webinars, on negative evidence.

It’s available online right now and for a few days will be free to review; afterwards, it’ll still be available for a small fee.


It’s true that negative evidence can be a tough topic, because even experienced genealogists can get confused about the exact distinction between direct, indirect and negative evidence in a particular situation.

negativeDirect evidence, of course, is information that appears to answer a research question by itself.1 The example I used was from the 1850 U.S. census of Yancey County, North Carolina, with the research question of “what did Charles Baker do for a living?” The census reports his occupation as High Sheriff.2 That surely gives us the answer, by itself, and so it’s direct evidence.

But if the research question is “where was the Baker family living when Rebecca was born?”, we have a different situation. The census does tell us she was born in North Carolina,3 so surely that’s where her mother was at that moment — but it doesn’t tell us that’s where the family was living at the time. They could have been just passing through, or visiting with relatives or friends.

So we’d need to combine the birthplace with other evidence of residence to answer the question and that — by definition — makes it indirect evidence: a bit of information that has to be combined with other bits to answer the question.4

Negative evidence is another beastie: it’s a “type of evidence arising from an absence of a situation or information in extant records where that information might be expected.”5 In the words of the Sherlock Holmes story, it’s the dog that didn’t bark — when it should have barked — if what we thought was true was in fact true.6

Sometimes we get confused when we search for something — say, a particular person in a census — and we can’t find him, so we think that’s negative evidence. Nope. That’s just a negative search — we don’t have any basis yet for drawing an inference about what it means that we couldn’t find him. Maybe his entry was misindexed. Maybe the census taker skipped over that household. We just don’t know — and we don’t have a basis for speculating.

And sometimes — sigh — we get confused and think that evidence that negatives a proposition — that tends to disprove it — is negative evidence.

As — sigh — I did when one question came up at the end in the Q&A.

The question was whether DNA results could be negative evidence, and, in my answer, the example I used of a case that could be negative evidence… isn’t.

The example I used was a YDNA test. My Shew line from North Carolina against another Shew-Schuh line from Virginia. We had every reason to believe, based on the paper trail, that the Philip who disappeared from the records of Virginia right around the time that my Philip appeared in the records of North Carolina was one and the same Philip.

But YDNA testing thoroughly disproved that working hypothesis. Looking at just the very top level YDNA results — the haplogroup, or which branch of the male human family tree the test takers are sitting on7 — it’s not so: the Virginia line has a haplogroup of R, and my Shew line has a haplogroup of I — and you can’t have two lines descending from the same common male ancestor with two different haplogroups.

That’s not negative evidence.

It does in fact directly answer the research question: “Are the two Shew lines related through the direct paternal line?” The dog did bark — we got an answer to our research question. And the fact that the answer is in the negative (“no, they’re not”) doesn’t change it from direct evidence to negative evidence.

There may be situations where, perhaps, DNA evidence might be put in the negative evidence category. If so, however, it’s not going to be in any case where the DNA test merely disproves a research theory. Any time a test can directly show a relationship (“yes, two people are related”) or debunk a theorized relationship (“no, two people aren’t related, at least not in this way”), it’s direct evidence — even when the answer is no.


1. See Board for Certification of Genealogists, Genealogy Standards (Nashville, Tennessee : Ancestry, 2014), 66.

2. 1850 U.S. census, Yancey County, North Carolina, population schedule, p. 450(A) (stamped), dwelling 975, family 967, Charles Baker; digital image, ( : accessed 4 Dec 2016); citing National Archive microfilm publication M432, roll 407.

3. Ibid., Rebecca A. E. Baker.

4. See Board for Certification of Genealogists, Genealogy Standards, at 70.

5. Ibid., at 71.

6. See A. Conan Doyle, “The Adventure of Silver Blaze,” Strand Magazine (July-December 1892) IV: 645; digital images, Google Books ( : accessed 4 Dec 2016).

7. ISOGG Wiki (, “Haplogroup,” rev. 4 Oct 2016.

Leave a Reply

Your email address will not be published. Required fields are marked *