DNA Testing Giveth, and Taketh Away

In her memoir Futureface: A Family Mystery, an Epic Quest, and the Secret to Belonging, the journalist Alex Wagner condemns the unreliability of ancestral origins data from DNA tests. Her criticism is incisive and worth reading. I reviewed Futureface last year for the New York Times Book Review, and here’s part of what I said:

By now, millions of us have taken ancestry tests. We’ve spit into tubes and allowed our genetic information to be uploaded into databases. In return we’ve had our genomes assigned to different parts of the world. The fine print warns us not to rely on those results, despite their seeming precision: 0.1 percent Oceanian! It’s hard to resist telling friends and family that we’ve turned out to be partly sub-Saharan African or Middle Eastern or Scandinavian. But testing with a different company may yield different results, or we may log into our account one day and find that the allocations have changed. Often the tests create as many mysteries as they solve. Autosomal DNA testing giveth, and autosomal DNA testing taketh away.*

The explanations that testing companies give for these shifts tend to be passive and blandly opaque: Identifying “ancestry-informative markers” depends on “sufficient data” from “reference populations.” Errors are “noise.” In her smart, searching new book, “Futureface: A Family Mystery, an Epic Quest, and the Secret to Belonging,” the journalist and former MSNBC host Alex Wagner recounts being incorrectly assigned enough Scandinavian DNA that one of her grandparents could have been from Sweden.

After the head of one of the testing companies suggests to her that blame for this kind of mistake lies with the consumer, who wants a “yes or no” answer when “science is not that simple,” Wagner lampoons his reasoning, comparing it to a chip company’s blaming its own pernicious fake-cheese coating on “nacho-chip-chomping bubble-butts” with an “insatiable lust for extra nacho-cheese products.”

….[Wagner’s] stinging criticism of the “flossy statistics” set out in ancestry composition results is some of the best I’ve seen on the subject. “Possibly inaccurate to the point of uselessness,” she calls them, observing that the categories themselves are suspect. Some DNA is “classified using political borders,” such as “Irish,” whereas others, such as “South Asian DNA,” are defined by “regional assignments.” The tests also fail to account for the permeability of borders over time. It’s true that the British colonized Burma in the mid-1800s, but there were far earlier arrivals by the Pyu (200 B.C.) and the Mon (A.D. 1000). The sites don’t purport to look back that far, but these historical events still raise the question: “At what point was Burmese blood considered ‘unmixed’ and exempt of outside influence?”

The book is recently out in paperback. Barack Obama recommends it.

Most of us who’ve taken these tests know how quickly these geographical assignments can shift. When I wrote about ancestor hunger for Harper’s back in 2014, my 23andMe test results indicated a teeny amount of North African DNA. Though minuscule, the percentage interested me because it didn’t come from my mom, who had also been tested, so (I said at the time) it must have come from my white supremacist father, an irony he would not have allowed himself to enjoy.

Later that year, I took another 23andMe test with a newer, purportedly more accurate chip, for a research study. The amount of the DNA increased slightly for that test, from 0.1% to 0.6%, and included “broadly Middle Eastern and North African” DNA as well as North African.

(The site also thought I had an identical twin, since I was in the database twice. It kept urging me to connect with my twin.)

In the past year, the North African/Middle Eastern DNA predictions disappeared entirely. I am now said to be 100% European, on both accounts. I know if my father were to learn of this—we’re not in touch—he’d feel that everything is as it should be, everything is as he always knew it to be. As he would see it, the European superiority of his genes remains intact.

The change means little for me in practical terms, apart from having relied on these results and needing to retract what I said in the past. But for some people, like Alex Wagner, these kinds of shifts in prediction can cover enormous segments of DNA. And for anyone whose ability to trace their ancestors was severed by a mass trauma like slavery or genocide, being handed ancestral origins and then having them revoked compounds the insults of the past.

Sites like 23andMe and Ancestry.com can help users solve family mysteries or connect with unknown relatives, but there are many reasons not to put our data into those databases. One less-discussed reason is that the geographical assignments are dubious.

Leaving aside haplogroups, which can speak to migrations over tens of thousands of years, even correct assignments typically only reach back 500 years or so. As Alex Wagner points out, people moved around a thousand years ago. There were wars, empires, conquests, crusades, and migrations.

Ancestry DNA has been running genealogy cruises for a few years, and now a more explicitly genetics-based travel industry has sprung up. 23andMe, touting the popularity of “heritage travel,” is partnering with Airbnb to offer ancestry packages. The announcement came after a survey by the homesharing site that found 57% of U.S. residents responding “would give up drinking alcohol for a year in order to get a free heritage trip.”

Conde Nast Traveler explains how these packages work: “Take a DNA test with 23AndMe, wait three to five weeks for your results, and then Airbnb will suggest a custom mix of home rentals, plus tours and classes from Airbnb Experiences, in the countries your genes trace back to.”

It’s “travel as unique as your DNA,” according to 23andMe. The company concedes that “you don’t even need to travel to explore your roots,” but has a pitch for non-travelers, too: “Through Airbnb Experiences, you can still connect with your ancestry in a meaningful way right in your hometown.”

I’m not maligning ancestor-oriented travel. It’s completely worthwhile to visit places where our recent or remote ancestors lived, as long as it’s not some triumphalist, fetishistic excursion. Wagner’s wanderings in Futureface may be the best parts of her book.

But it takes a lot of chutzpah to sell travel packages to places that might get edited out of the traveler’s ancestry composition map with the next update.

In 2014, the novelist and critic Laila Lalami spoke to me for my short-lived Family Tree interview series, which I mentioned in an earlier newsletter. Lalami’s* mom was adopted, and not knowing the origins of her maternal ancestors eventually led a reluctant Lalami to genetic testing. What she learned, she wrote in the New York Times magazine, was unsatisfying.

Lalami is one of the best writers I know, and also a friend. We became friends before I realized how talented she is. I always liked and admired her writing, but I revere her last novel, The Moor’s Account. It’s one of my favorite books. If you haven’t read her work, she has a new book out, The Other Americans, and it’s a good place to start. Her essays are also very worth looking up.

Here are some of my favorite parts of our interview:

Maud Newton: In a Lives piece for the New York Times Magazine, you write that your mom was left in a French orphanage in Fez in 1941, and that, over the years, you had many theories and stories about how she might have ended up there. Your thirst for the truth eventually led you to take a genetic test, but in the end, science couldn’t give you the kind of answers you were seeking. “Only stories could,” you said. Do you think the mystery of your mother’s origins is part of the reason you’re a writer?

Laila Lalami: I think it certainly played a part. When I was growing up, I could never shake the feeling that there was something different about my family. All my friends had maternal aunts, uncles, cousins, grandparents, whereas my extended family consisted solely of relatives on my father’s side. We also did certain things differently at home, like sing French lullabies instead of Arabic ones, or eat pain perdu with mint tea—habits my mother brought with her from the French orphanage. Being different meant that I became more sensitive to detail, more attuned to all the ways in which a person belongs to or is held apart from a group.

For me, the desire to write came from my love of books and my need to tell stories. But I think there’s a connection between feeling like you’re different and wanting to tell a story. When you write you can, at least temporarily, tame that feeling of difference.

MN: Mustafa, the narrator of The Moor’s Account, has been taken from his homeland, stripped of his name, enslaved, and carried along on Narváez’s 1527 expedition to La Florida. On his journey, and when he is captured at his destination, he sustains himself in part on memories of the stories that his mother told him as a boy. His own memoirs are in a sense a continuation of that tradition. “What each of us wants, in the end,” he writes, “whether he is black or white, master or slave, rich or poor, man or woman, is to be remembered after his death. I am no different.” It’s such a wonderful, nuanced novel, and one that underscores the arbitrariness of whose stories are preserved to time and whose are forgotten.

LL: Although I didn’t realize it at the time I started working on the novel, I think one of the things that interested me most was exploring the ways in which some stories are celebrated, propagated, and even canonized, while others are passed over or forgotten.

In history, the stories that prevail aren’t necessarily the best or truest; they’re the ones told by the most powerful people. The beauty of fiction is that it does not obey that pattern. A writer can choose whichever perspective she deems to be the best for the story.

MN: I’m interested in repetition in families, in echoes or the lack thereof down through the generations. Do you think about shared traits or tendencies when you look at your own family?

LL: Yes. Because I know so little about my mother’s ancestry, I often find myself wondering whether certain traits are inherited from my father or my mother. And I do this with my daughter, too. For instance, no one in my family (or my husband’s family) is a musician, but our daughter has always had an aptitude for it. So I always wonder if it’s not because of a distant relative on my mother’s side.

MN: Those of us who attempt to trace our ancestry: what do you think we’re seeking?

LL: Ourselves.

Right now I’m planning to send Ancestor Hunger dispatches on the first and third Tuesday of each month. If you’re wondering what to expect in future installments, I put together an About page. I’ll let you know if I’m thinking of picking up the pace.

All good wishes until next time,


* This sentence was cut from the published review, and I understand that is a weird sentence, but I still like it, so I put it back for newsletter purposes. Because I can!

** I know we’re calling everyone by their first names now, because why stand on ceremony, and surnames can themselves be problematically patriarchal, and probably there are other reasons that escape me, but I’m old enough to remember a few decades ago when we decided it was a bad idea to refer to brilliant notable women by their first names, because it robbed them of the kind of gravitas that notable men tend automatically to be thought to have. I don’t like to be gratuitously formal, but more than that, I don’t want to be minimizingly familiar when I’m talking about writers and ideas I take seriously. I’ve experimented with the other way, and it just doesn’t feel right to me. So, in this context: Lalami. As people used to say a billion years ago on the pre-blog internet, YMMV!