When a 1998 study in Nature announced that people named Cohen share a strikingly similar genetic makeup, I shrugged. As a Cohen, I had assumed that I was related to all other Cohens. After all, the name wasn’t forced on my family by a plantation owner nor is it an Ellis Island abbreviation of Kuznyetshovsky. What did surprise me were the findings revealed a year later by the same researchers: Most members of the Buba clan of the Lemba tribe, a 70,000-strong Bantu-speaking people dispersed throughout southern Africa, are related to us Cohens. Spicing up my personal connection, the Lemba claim to have migrated from Yemen, the home of my father’s parents.

The Cohen link wasn’t exactly news to the Lemba. Despite the dismissiveness of academics and some South African Jews, the Lemba have long claimed that they are black Jews, a lost tribe who 3,000 years ago sailed across the Arabian Sea from their home in Yemen to the African continent, slowly migrating down to what is now Zimbabwe and South Africa. Indeed, they follow some Jewish-like traditions—male circumcision, religious use of a six-pointed star (with an elephant inside it), ritual slaughter of animals, dietary rules that discourage the eating of pork, and marriage within the tribe. Yet many Lemba also have that one decidedly non-Jewish belief: that Jesus Christ is the Lord.

Because I was already headed to Africa on business, I decided to invite myself over to the Lemba’s for shabbat dinner, or something equally cozy, and talk about our mutual ancestors. But first I wanted to know how closely I was related to the Buba clan, so I asked David Goldstein, a geneticist at University College London who led these studies, to analyze my DNA.

Geneticists can determine whether males share common male ancestors because an identical copy of the Y chromosome passes from father to son. In their first Cohen study, Goldstein and his co-workers discovered a unique signature in the DNA of 100 Cohen males they studied—specific mutations on the Y chromosome—which they dubbed the “Cohen modal haplotype.” Based on expected mutation rates, the researchers dated the Cohen modal haplotype back as far as 3,250 years ago (130 generations), approximately the time when Aaron, the brother of Moses, supposedly anointed the Cohen clan as the high priests of the religion. This connection had special resonance for me: My father for most of his career worked as a cantor, the musical counterpart to a rabbi.

Goldstein sent me a sterile Q-Tip in a liquid-filled test tube to collect a DNA sample. I swabbed the inside of my cheek with it and mailed it back. From this sample, Goldstein extracted my Y chromosome’s DNA and commenced his hunt for the Cohen modal haplotype markers. Whereas 56 percent of Lembas from the Buba clan have 17 of the 17 markers that constitute the Cohen modal haplotype, Goldstein found that I have only 16, making me a “one-step neighbor.”

Confused, I asked if being a one-step neighbor meant that one of my forefathers had a mutation in his Y, branching him off the tree that represented the Buba clan of the Lemba and other “true” Cohens in his study? “It can’t be said exactly that way,” Goldstein cautioned, sounding like a rebbe answering a question of Talmudic proportions. “The way that I’d say it that would be accurate won’t be very satisfying for you.”

Goldstein—who has the Cohen modal haplotype himself—added, “We don’t know that there ever was a pure Jewish population.” He urged me to think of my Y chromosome in terms of genealogy, not populations. The essence of his caveat is that all men derive from a common ancestor (say, Adam), and the Cohen modal haplotype simply designates a more recent common ancestor (say, Aaron Cohen). While I suspect my conclusion won’t be very satisfying for Goldstein, I think that, in all likelihood, I’m a mutant Cohen. I was, however, placated to learn how my Y chromosomes compare with other men in Goldstein’s ever-growing database: My markers perfectly match those of three Yemenite Jews from Hadhramout—the very region where the Lemba supposedly once lived.

Steve Jones, another University College London geneticist, argues against the Lemba being a lost tribe “with uninterrupted genetic contact with the Middle East.” The Cohen modal haplotype might well have bounced down the African continent over the generations and been eventually absorbed by the Lemba.

Goldstein says some non-Cohen Jews have taken umbrage at this work, as they (wrongly) inferred that the data say Cohens are genetically superior to other Jews. “Be careful how you write this,” he added. “This cannot be taken as a view of status. Have sensitivities. People get really worked up about these things.”

T o arrange a reconvening of our tribes, I phoned Professor M.E.R. Mathiva, whom one anthropologist dubbed “the spiritual head of the South African Lemba.” A trip to Mathiva’s home in Thohayandou, South Africa, would require a plane flight from Johannesburg, a rental car, and drive through towns called Nobody and Nirvana. I was ready.

But Mathiva declined my self-invitation.

“The Lemba don’t need attention anymore,” said Mathiva, adding that three Yemenite Cohens had visited the previous Friday on a similar mission. Instead, he referred me to a Buba Lemba in Johannesburg, and after a few more phone calls, I had secured an invitation from Peter Mbelangwa to visit him at home.

Mbelangwa, who works as a manager at IBM, welcomed me warmly, if a bit cautiously, to his home in Bramley Gardens, a tony suburb of Johannesburg. He introduced me to his wife and three sons, and we sat in his living room, which was free of the religious artifacts of most American Jewish homes—photos of brides and grooms standing under chuppahs, needlepoints of yarmulke-clad men snapping their fingers in dance, artsy menorahs, and gilded sedar plates.

Probing the Jewish-like “dos and don’ts” of the Lemba, I found as many differences as similarities. Mbelangwa had never heard of the kosher practice of separating milk from meat. Lemba don’t celebrate Passover, which is as important a religious gathering to most Jews as Easter is to Christians. He didn’t know whether it was forbidden to name a child after a living relative, which Jews aren’t supposed to do.

Thinking that some less obvious ancient bonds might exist, I sang a whiny, distinctive Yemenite song. Mbelangwa then sang a Lemba funeral dirge, which couldn’t have been more different. He also played a Lemba instrument for me, a sambira, a gourd that has thin metal reeds sprouting out of it like the fingers of a rake. When he plucked it, it didn’t sound Yemenite to my ear. Mbelangwa’s wife also donned a Lemba dress for me, and some Lemba jewelry. They were beautiful but did not rock my vestigial soul. I asked Mbelangwa’s sons whether I looked like any of their relatives. The boys politely shook their heads no, and I could tell from the slight smiles they thought me a fool.

Mbelangwa and I then discussed his views of South African Jews. “There are Jews who benefited from apartheid,” he said. “And there are those Jews who opened the doors for us during the days of oppression.” Yes, some of his good friends were South African Jews, but he confided that “you don’t close your eyes in front of a Jew.”

That wasn’t exactly the call to brotherhood that I had traveled 7,000 miles to hear, least of all from a Cohen. I understand about having misgivings about your own: I purposefully keep my distance from a few relatives who are only a twig away on the family tree. Still, as my whole Cohen modal haplotype quest began to collapse under its own weight, I had to accept the fact that my DNA wasn’t going to shake hands with Mbelangwa’s over the millenniums. The romance of kinship can mislead, whether it is with a first cousin or, as in the case of Peter Mbelangwa, someone who is an unfathomably distant cousin, 130 generations removed.