In the summer of 2005, I received a phone call from a documentary filmmaker. The filmmaker explained that he was seeking historical insight on the results of a DNA test conducted on a celebrity who wished to remain anonymous; he had questions about African history, the slave trade, and Africans in the American South. If, during our conversation, I had interesting enough things to say, he might bring me to New York for an on-camera interview. All of these topics are in my wheelhouse, so I was excited and eager to participate.

The celebrity claimed that her DNA test showed she was Zulu. I knew that Oprah Winfrey had recently announced that she was Zulu, but I remained silent. Was the celebrity’s family a recent arrival from South Africa? Or was she claiming descent from enslaved people who were Zulu? The producer insisted it was the latter. I told him that the result was highly unlikely. Fewer than 2 percent of enslaved Africans who arrived in British North America and the United States hailed from Southeast Africa. Moreover, the biological fact of an individual’s DNA did not easily map onto historical social realities: a cohesive Zulu identity didn’t emerge until King Shaka’s empire in 1816, drawing together what had once been a confederation of disparate clans. The legal slave trade to North America had ended in 1808. It was almost impossible for the celebrity to be descended from enslaved Zulus transported to North America since “Zulus” didn’t even exist when the slave trade ended.

The documentary producer was first unhappy and then exasperated. Clearly, I wasn’t giving him the answers he wanted. He pushed and prodded and asked whether I was sure. I assured him that I was. After an hour, we ended our conversation, and I never heard from him again. One year later, Oprah Winfrey appeared on a PBS documentary, where updated DNA evidence suggested that she had a shared ancestry with the Kpelle people of present-day Liberia.

My close encounter with Oprah left me fascinated with the potential promises and pitfalls of DNA testing as a means of accessing ancestry and identity. Several years later, I helped organize a yearlong symposium through the University of Wisconsin–Madison’s African Studies Program focused on genetics, genealogy, and the African diaspora. We invited geneticists, anthropologists, sociologists, historians, and interested UW community members to take part in this project. In addition to shared readings, documentary film screenings, and scholarly research presentations, we included a “participant observation” component to the symposium. Early in the fall semester, participants drew names out of a hat. Five lucky winners earned the right to learn their supposed African origins from one of the commercial DNA testing companies. The participants took their cheek swabs in front of us, we sent off the samples, and then we awaited the reveal day in the spring.

DNA does not easily map onto historical social realities.

By the big reveal in the spring, I thought that the participants understood that DNA did not provide the straightforward historical and ethnographic conclusions that geneticists frequently claim. In fact, such conclusions are often downright misleading. The company we chose claims that they “find identical matches” for 85 percent of those they test, revealing both an African “country and ethnic group of origin.” However, the proprietary databases of genetic samples used in these evaluations are small, especially for Africa. Moreover, as in Europe or anywhere else, African countries and ethnic groups have histories. Like “Zulu,” they are neither static nor primordial. Historical migration means that a genetic sample found in one place today was not necessarily the place of origin for the person producing that sample, let alone the place of origin of their ancestors.

Nevertheless, it became apparent fairly quickly that our volunteers were taking the testing seriously and were anxious to learn about their African ethnic pasts. The first to reveal was a white professor who predictably showed that his origins resided in western Europe with no African ancestry. A graduate student who identified as mixed race likewise demonstrated an expected mixture of ancestry from all over Europe and several places in West Africa. A professor who identified as African American was excited to learn that her closest genetic markers came from the Kpelle people of Liberia (like Oprah!), and the company provided an elaborate celebratory certificate of attestation for her to frame and put on her wall. Another African American professor was crestfallen to learn that she had no African DNA and that all her ancestry came from Europe. Finally, our “control” subject was a Nigerian, born and raised Yoruba, one of the largest ethnic groups in Africa. Surely that’s where he would match.

When our Yoruba subject read the result, he roared with laughter. On his certificate of authenticity, alongside his very obvious Yoruba name, was affirmation that he was in fact descended from the Fang people of Gabon, more than 1,200 miles southeast of Yorubaland. I can speculate why a Yoruba person’s DNA might match someone’s in Gabon, but there is no way to know how long ago movements that might have allowed for such a match occurred. All we know is that a person suffused in Yoruba history, language, and culture—a person who is Yoruba—shares a fragment of genetic ancestry with someone today living in Gabon. Meanwhile, the DNA company’s interpretation of this genetic fragment completely occludes his real Yoruba identity.

Even when limited to tracing familial relations, DNA evidence can be misleading.

So what are we to make of the ambiguities in using DNA testing to trace lost histories? Are the companies dealing in artifice or fraud? The search for ancestral roots can be empowering, especially for African American and diaspora peoples violently alienated from their homelands and denied “history” through generations of enslavement. The possibility of connecting to even one distant African relative promises a passageway to a people, a culture, a history, transcending slavery and America’s racist past. At the same time, this kind of genetic testing tells us almost nothing about history. Herein lies the contradiction. We want to believe that the people in white lab coats with their beakers and pipettes can provide scientific and historical truths when they can’t even offer us the former. To say that two people share a single strand of DNA with one common relative (among thousands), deep in the indeterminate past, is not an “identical match.” Instead, it is a random fragment elaborated into kin, country, and ethnicity by for-profit companies with no ethnographic or historical expertise.

Even when limited to tracing familial relations, DNA evidence can be misleading. Around five years ago, I received a commercial DNA test as a gift. Despite my skepticism, I was curious to see how the company would chart my ethnicity. Predictably, my DNA matched to regions in England and Germany. But then, last year, something strange happened. I received an email informing me that I had a new DNA match with a “1st–2nd cousin” whose name I didn’t recognize. When I went to “explore” the match, I learned that we share 14 percent of our DNA, slightly above the average of 12.5 percent shared by most first cousins. I was stunned. How could I possibly have a first cousin I never knew existed?

My father was adopted in Florida in the mid-1940s. We never knew his birth parents or anything about their history. Family legend held that his birth mother was a teenaged American who’d had an affair with a German man. The man to whom I matched had a rather uncommon German surname, so I started researching to see whether I could find him. The search has been eventful, if mostly fruitless. However, in the course of my digging, it became apparent that I wasn’t necessarily looking for a “first cousin.” Companies assume lineal connections based on traditional families when they report DNA relationships. They would not, for example, report to me that the man with whom I share DNA might actually be my half uncle (i.e., that he and my dad are half brothers). Yet the relational percentage shared with a half aunt or half uncle is nearly identical to that shared between first cousins. The DNA can only narrow the lineal possibilities; it can’t always tell you exactly how you are related to another person. And the commercial DNA companies have no interest in selling potential infidelity!

I’m still as fascinated with DNA testing as I was in 2005. I love watching Finding Your Roots with Henry Louis Gates Jr. And I still hold out hope that I will figure out the mystery of my long-lost “genetic cousin.” But I look on the interpretive results of DNA testing companies with healthy skepticism. Using the objective cloak of “science,” commercial DNA firms sell individuals and television producers what they so desperately want: evidence to reveal new and lost histories. DNA may have the potential to unlock historical secrets, but only if it is subject to deeper conceptual and methodological scrutiny, just like any other subjective shard of evidence.