What the studies cannot tell us is who is a Jew.
Jews and their neighbors: The Diaspora
Genetic researchers have not neglected more than 90% of Jews who are neither kohanim nor Levi'im. They began with good reason to suspect that a great deal of mixing had taken place during the millennia of dispersion. People had noticed, after all, that the pale-skinned redheads common in Lithuanian Jewish communities do not look much like petite, dark-haired Jews from Yemen. It was assumed that Jews were bound more by tradition than by genetic kinship, that in the distant past Jewish men had followed opportunity to some far-off city, married local girls, persuaded them to separate the meat and milk dishes and founded new Jewish communities. Moreover, it was believed non-Jewish ancestors had continued to mix into the Jewish community.
The idea that the traditional story - Jews driven into exile faithfully marrying only fellow Jews - might be largely true was startling. And yet, so it seems. There were, of course, times and places where significant numbers of people converted to Judaism. But in the centuries since the beginnings of European Jewry, the best available estimate is that a mere 0.5% of new material entered the gene pool of Ashkenazi Jews in each generation. This is part of a picture of remarkable Jewish genetic continuity emerging from research labs at a dizzying rate. More studies have been carried out on the genetic history of the Jews than on most ethnic groups, perhaps because there are so many Jewish doctors to take advantage of the fabled willingness of Jews to participate in research.
These studies not only show that almost all Jewish populations have origins in the Middle East, but that the DNA of Jews from almost every corner of the Diaspora is more similar to that of other Jews than to any other population. When compared with non-Jewish groups, the closest match is with the Muslims of Kurdistan, not with the European peoples alongside whom Ashkenazi Jews lived for centuries or the Arab neighbors of many Sephardi populations.
Jews and their neighbors: The Middle East
With Jews looking increasingly like a relatively cohesive population largely of Middle Eastern origin, the logical next question is how close a genetic relationship exists with other Middle Eastern groups. A study of Israeli Jews and Palestinian Arabs published in 2000 by Israeli researchers revealed what the authors described as "a relatively recent common ancestry." It was greeted with euphoric proclamations that Palestinians and Jews are "brothers." A closer look at the details of the study gives reason for pause.
The researchers compared Jews and Palestinians to a sample of people from Wales. When compared with the Welsh, Jews and Palestinians did indeed look similar, as they probably would if contrasted with Trobriand Islanders. When the same research team conducted a follow-up study comparing Jews and Palestinian Arabs to Kurds, Armenians, Turks, Syrians, Jordanians, Lebanese and Beduin, they saw a very different picture. Although all Middle Eastern populations have broad similarities, "Jews were found to be more closely related to groups in the north of the Fertile Crescent (Kurds, Turks and Armenians) than to their Arab neighbors."
This could mean that Jews, Kurds, Armenians and Anatolian Turks all carry the genetic markers of ancient indigenous populations of the Fertile Crescent, while Palestinian Arabs and Beduin may largely descend from the Arab conquerors, with their distinctive genetic signifiers. Genetics may eventually provide answers to such questions as what proportion of Palestinian ancestry arrived via earlier or later migrations. So far we have only partial explanations. One of the most compelling studies compared the small Samaritan population in Israel with Druse, Palestinians and Jews from various parts of the Diaspora.
The results appear to corroborate the traditional Samaritan belief that they have lived in Samaria since antiquity and are closely related to the Jews. Only four Samaritan family lineages survive, but of those four male lines, three carry the kohen modal haplotype, while the fourth, the Cohen family of priests, does not. The data indicate that the Samaritans generally married other Samaritans. Y chromosome DNA shows the Samaritan male line to have "a much greater affinity" to Jews than to the Palestinian Arabs who have surrounded them since the Arab conquest.
Studies of Jewish women
The ease of tracing male lineages with the Y chromosome accounts for the large body of research that uses exclusively male populations. However, another technique allows equivalent explorations of the female line.
Every cell of our bodies contains mitochondria, small organelles that generate energy from food. Each mitochondrion harbors its own circular strand of DNA, which both sexes inherit from their mothers, and which is passed on only by women to the next generation.
The utility of mitochondrial DNA (mtDNA) studies is demonstrated by recent findings concerning one of the least known Jewish groups: the Bene Israel, descendents of 6,000 Jews "discovered" on the west coast of India by Jewish traders from Baghdad in the 1830s. They carry the kohen modal haplotype along with other Middle Eastern genetic markers, and have substantial mtDNA found only in the Indian population among whom they lived - always keeping the Sabbath - for hundreds, perhaps thousands, of years. From the genetic evidence, it looks as though a small group of Jews, all or mostly male, arrived on the Indian coast, married local women and built a Jewish community.
Studies of mitochondrial DNA suggest that Jewish communities were often founded by very small numbers of women. One study demonstrates that 27% of Moroccan, 41.3% of Bene Israel and 51.4% of all Georgian Jews are descended from a single female ancestor in each community. The matriarch of 41.3% of the Bene Israel came from a local Indian family. We do not know the origins of that founding mother of the Georgian Jewish community or even whether she was born to a Jewish family, only that she carried a distinctive haplotype found in an area stretching from Sicily to the Caucasus to Iraq.
The Bukharan, Persian, Ethiopian, Iraqi and Ashkenazi Jewish communities all have unusually small numbers of maternal ancestors. This could reflect communities founded by tiny numbers of Jewish women willing to travel with their husbands to far-off lands, or situations like the Bene Israel's, where a community was founded at least in part by a small number of women who married Jewish men and lived as Jews. A bit of light recently has been thrown on part of this picture by a study of the mtDNA of Ashkenazi women, close to half (42%) of whom are descended from one of just four matriarchs. The distinctive, complete sequence haplotypes carried by descendents of these four women are almost unknown in other populations, except occasionally in Jewish communities that trace their origin to the expulsion from Spain.
The evidence indicates that these four female ancestors most probably originated in the Levant, perhaps accompanying their husbands from the Middle East to their new homes. Even Samaritan mtDNA has been examined, showing distinctive patterns about equally different from Jewish and Palestinian comparison groups, and hence with somewhat more mixing than in the male line. Genetics of the "lost tribes" Jews and Christians alike have an endless fascination with stories of the "lost tribes" of Israel turning up in odd parts of the world. Witness the recent enthusiasm about the Bnei Menashe in the hills of eastern India. Few such groups have been studied by geneticists, but when they are, the results can be remarkable.
Perhaps the most surprising story in Jewish genetics involves the Lemba, a Bantu-speaking people of about 50,000 living in southern East Africa. Their appearance and lifestyle are largely similar to other Bantu-speaking groups, with a few notable exceptions: They practice circumcision, have a ritualized slaughter procedure for animals, avoid eating pigs and have a strong tradition that their ancestors migrated from "Sena in the north by boat." When westerners came into contact with the Lemba and noticed the similarity of their customs to Jewish practices, they wondered whether "Sena" might be Sana in Yemen and whether the Lemba were of Jewish descent. A British anthropologist arranged for genetic testing of members of the tribe, finding that 10% of Lemba men carry the kohen modal haplotype on their Y chromosomes. Even more impressive, the Lemba have a priestly clan whom they call the Buba. Fully 52% of Buba men who were tested bear that same marker of kohanic descent.
Although mtDNA testing has not yet taken place, it seems likely that the origins of at least some portion of the tribe date back to the arrival on African shores of male Jews, and their subsequent marriage to local women. Only vestigial Jewish traditions were maintained among a population that is animist and Christian in practice. In no sense - cultural or halachic - can the Lemba be considered Jews today. But their story demonstrates the power of DNA to elucidate the probable history of many populations.
Genetics and identity
What genetic data cannot tell us is who is a Jew. The answers to that question are, variously, halachic, political and cultural. On a purely technical level, there is no genetic screen that can sort Jews from non-Jews.
Population differences do not translate into reliable tests of individual lineage. What genetics can tell us is something about where our ancestors came from - no more. It cannot tell us who we are. Nor can it tell us who we want to become, as individual Jews or as a Jewish people. As new data emerge from genetics laboratories, though, we are likely to learn a great deal more about the history of our people.
Diana Muir Appelbaum is the author of Reflections in Bullough's Pond: Economy and Ecosystem in New England (University Press of New England, 2000), and is working on a book on nationalism. Paul S. Appelbaum is the Elizabeth K. Dollard Professor of Psychiatry, Medicine and Law at Columbia University, and writes about the ethics of genetic testing and research.