Think basics, not exotics
The Legal Genealogist emphasized this basic concept for every-day genealogical research in a post last year.
It’s one of the most basic tenets of medicine.
When you hear hoofbeats, think horses, not zebras.
Often ascribed to Dr. Theodore Woodward, a professor at the University of Maryland School of Medicine in the 1940s, what it means in essence is that when a doctor encounters a basic set of symptoms that look like the common cold, the doctor should think that it is the common cold and — unless there’s good reason to think otherwise — not some exotic illness that appears in that time and place only in the pages of a textbook.1
In other words, we should look for the most likely, most common explanation for what we’re seeing first before heading off to consider the far-fetched exotic possibilities.
And, as I noted last year,2 it’s as applicable to genealogy as it is to diagnosing disease. We need to go with the probable solution and try to rule it out before we go on to the less likely possibilities.
In that post last year, I was focusing on a family tree that tried to assign a child born in April 1845 — whose baptismal record said he was the third child and first son of that mother3 — to my second great grand-aunt, Johanne Rosine Geissler, born in 1833. And, I suggested, the fact that she’d have been just 12 at the birth of that child made this a zebra, not a horse. A 12-year-old mother? Not impossible, but unlikely. A 12-year-old having her third child? That’s a zebra. It’s so unlikely that there almost has to be another explanation: it’s not her kid.4
And this same adage — think horses, not zebras — is so, so, so applicable to all those cases where we try to integrate DNA evidence into our genealogical conclusions.
A case in point came up yesterday on Facebook, in a discussion of the likelihood of Native American admixture results being detected in autosomal DNA testing. In other words, how likely is it that those percentages we see will include Native American ancestry? The concern was that a test wouldn’t show such ancestry even if the person actually did have Native American forebears.
And a friend and colleague, Yvette Hoitink, noted: “I had the opposite case, where one test showed Native American DNA where it is extremely unlikely that I have any and it only showed up on the one test. I am from the Netherlands and have my tree mostly complete to the 1600s, and none of my ancestors even set foot in the Americas until after I was born.”
Yvette, also the holder of certified genealogist credentials from the Board for Certification of Genealogists, wrote up her anomalous results with her tiny smidgeon of supposed Native American DNA on her blog, Dutch Genealogy, in 2014.5 And so many of the responses to that blog post trotted out the zebra: she looked like she had Native American features and, after all, Dutch explorers brought enslaved Native people back to the Netherlands. “It was incredible to see in the comments,” she said, “how many people choose to believe it is more likely that somehow a Native American made it to my father’s village in the Netherlands than that a test result might give an incorrect prediction.”
So you know what happened in the discussion yesterday, right? The very first response to Yvette’s comment on Facebook noted the interrelationship of the Dutch and Native peoples in what had been colonial New Netherland — certainly a possible explanation as to why a tribal member today might show Dutch ancestry. But not an explanation as to why a Dutch person today, with documented Dutch ancestry back before the Dutch settled in the New World, might show Native ancestry.
Leading Yvette to repeat the point we all need to take to heart: “I can’t understand why people keep trying to come up with exotic theories to explain a single uncorroborated data point.”
In other words, why there’s such a focus on zebras rather than on horses.
And, in my experience, when we’re trying to use DNA as part of our genealogical toolbox, trotting out zebras usually comes from one of two causes: either we just don’t understand well enough what the test results mean or we’re so emotionally invested in what we want them to mean that we overlook their real meaning.
For the former, consider my autosomal results against a paper-trail third cousin. We’ve both tested at two different testing companies, and neither of them reports us as a match. It’d be easy to say that the paper trail has to be an error: we must not be third cousins. And I see that kind of a comment all the time: “well, if you really were third cousins, the DNA would have show it.”
Nope. Not so. Up to 10% of third cousins simply will not share enough autosomal DNA in common to be detected by the testing companies.6 This particular third cousin does match all five of my mother’s siblings who’ve tested. He matches my first cousins. He matches my second cousins. He matches my brothers and my sisters. He even matches my nephew. The one family member he doesn’t match is me. And — snort! — I paid for his test.
The fact that I don’t match him is just the horse of recombination7 — the luck of the DNA draw. The notion that the paper trail must be wrong here would be the zebra.
For the latter, well, those percentages are a perfect example. I’ve tested with every major consumer DNA testing company out there. So I could tell you, citing my test results as “proof,” that I am 66.9% Scandinavian. Or I could tell you, citing my test results as “proof,” that I have no detectable Scandinavian ancestry whatsoever.8
The horse in this case is that my DNA suggests I have northern European ancestry. It’s a zebra to say I’m either largely Scandinavian or not Scandinavian at all. The science isn’t good enough — and may never be good enough — to distinguish reliably between admixture results at less than a continental level.9
So when it comes to DNA, we need to stop and think about what the evidence is telling us when we hear those hoofbeats.
And think horses first.
Not zebras.
Or Native Americans in pre-1600 Dutch villages.
Cite/link to this post: Judy G. Russell, “An adage for DNA too,” The Legal Genealogist (https://www.legalgenealogist.com/blog : posted 23 Jan 2022).
SOURCES
- See “When You Hear Hoofbeats Look for Horses Not Zebras,” Quote Investigator, posted 26 Nov 2017 (https://quoteinvestigator.com/ : accessed 23 Jan 2022). ↩
- See Judy G. Russell, “Think horses, not zebras,” The Legal Genealogist, posted 20 Feb 2021 (https://www.legalgenealogist.com/blog : accessed 23 Jan 2022). ↩
- Evangelische Kirche Rehmsdorf (Kr. Weißenfels), Tauf-Nachrichten, Nr. 4/1845, Johann Julius Kirsch; digital images, “Saxony, Anhalt, Anhalt-Bernburg, Anhalt-Dessau and Anhalt-Köthen, Germany, Lutheran Baptisms, Marriages, and Burials, 1760-1890,” Ancestry.com (https://www.ancestry.com : accessed 20 Feb 2021). ↩
- There is also the minor detail that my Johanne Rosine Geissler died on 3 August 1834 at the age of seven months, one week and three days. She was buried at the same church where she was baptized. Ibid., Evangelische Kirche Ossig (Kr. Zeitz), Todten-Anzeigen Nr. 7 1834, Johanne Rosine Geissler. ↩
- Yvette Hoitink, “My Native American DNA. Say what?,” Dutch Genealogy, posted 27 June 2014 (https://www.dutchgenealogy.nl/ : accessed 23 Jan 2022). ↩
- See generally ISOGG Wiki (https://www.isogg.org/wiki), “Cousin statistics,” rev. 21 May 2021. ↩
- See ibid., “Recombination,,” rev. 27 Oct 2021. ↩
- The first would be my admixture at MyHeritage; the second at 23andMe. ↩
- See Judy G. Russell, “Not soup in 2021 either,” The Legal Genealogist, posted 19 Sep 2021 (https://www.legalgenealogist.com/blog : accessed 23 Jan 2022). ↩
Judy, just FYI, you left out one subgroup of the family of equidae: horses, zebras, & asses. You might want to consider those first.
I’d reserve that for those who don’t think at all. 🙂
And we wonder why so many people believe in conspiracy theories and “alternative facts”. I like what Lynn gives above as a third choice!
Mind you, if you’re anywhere near the BBC Radio Studios, when they’re recording Shakesperean battle scenes, you’re not hearing horses, zebras, or asses, but coconut shells being banged together!
Seriously, though, my cousin found that he had 1% South Asian ancestry (India, Pakistan, Sri Lanka, Bangldesh) when he was tested in 2017. The other 99% was composed of Irish and Western European genes.
I asked him last year if that had produced any actual South Asian relations. He went off and checked and said no South Asian relations had appeared. In fact that 1% was no longer South Asian, but now North American Indian, everything else being unchanged.
His problem now is that he has no excuse to go for a curry!
Looking yesterday at a string of births in one parish with one included that’s from another, two days’ journey away. No grandparents or other family over that way. The only possible connection was the mineral being mined was the same. Maybe they wandered over there one year and then wandered back for the next child. But I am not putting that baptism in my tree until I find some other evidence.
Definitely a zebra.
Whereas, in a herd of zebras you can’t see the one horse.
Most of the DNA companies shove my Cornish and Scottish into Irish, completely obliterating and overwhelming the tiny actual 3% Irish with ten times false Irish.
And then there’s my Wend/Sorb input. MyHeritage and someone else thinks they are Croatian. Well the tribe split way back. Some remained just SE of Berlin and into Saxony. The rest headed further south and they are the ones in a panel.
But with several other possible Slav groups in the area it was great to know which this mob was. A zebra that turned out to be what? an Appaloosa maybe?
I (and my sister) have a definite zebra and I’ve been hunting her for over a year. Our family arrived in New York and Pennsylvania while George III was ruling and our ancestors passed the bare bones down. There’s no doubt of our Dutch, German, French, and generally Western European heritage. There is a documented (in NY State Archives and tribal records due to a lawsuit) Native mother and a Dutch father. But it was a great surprise to learn from mtDNA from two companies that we both have haplogroup L1c3a, which originates from Central Africa within 60,000 years, not the expected European haplogroup. This indicates that a woman on our maternal line, likely within the last 300 years, was “out of Africa”. I can trace to a female (1794-1878) by marriage and death docs. I think I’ve found my zebra.
I wouldn’t regard that as a zebra at all. Huge numbers of Americans can trace their lines back to either Native or African haplogroups because of intermarriage and less formal relationships starting very early in colonial times.