Gene Expression

The adventure of the Cohen Modal Haplotype

The past 10 years have seen a lot of ups & downs in the Cohen Modal Haplotype (CMH) hypothesis, which suggests some veracity to the Biblical narrative whereby the priestly caste status of the Jewish people is passed down patrilineally from Aaron, the brother of Moses. But could there be multiple Aarons? Possibly. Extended Y chromosome haplotypes resolve multiple and unique lineages of the Jewish priesthood:

It has been known for over a decade that a majority of men who self report as members of the Jewish priesthood (Cohanim) carry a characteristic Y chromosome haplotype termed the Cohen Modal Haplotype (CMH). The CMH has since been used to trace putative Jewish ancestral origins of various populations. However, the limited number of binary and STR Y chromosome markers used previously did not provide the phylogenetic resolution needed to infer the number of independent paternal lineages that are encompassed within the Cohanim or their coalescence times. Accordingly, we have genotyped 75 binary markers and 12 Y-STRs in a sample of 215 Cohanim from diverse Jewish communities, 1,575 Jewish men from across the range of the Jewish Diaspora, and 2,099 non-Jewish men from the Near East, Europe, Central Asia, and India. While Cohanim from diverse backgrounds carry a total of 21 Y chromosome haplogroups, 5 haplogroups account for 79.5% of Cohanim Y chromosomes. The most frequent Cohanim lineage (46.1%) is marked by the recently reported P58 T->C mutation, which is prevalent in the Near East. Based on genotypes at 12 Y-STRs, we identify an extended CMH on the J-P58* background that predominates in both Ashkenazi and non-Ashkenazi Cohanim and is remarkably absent in non-Jews. The estimated divergence time of this lineage based on 17 STRs is 3,190 +/- 1,090 years. Notably, the second most frequent Cohanim lineage (J-M410*, 14.4%) contains an extended modal haplotype that is also limited to Ashkenazi and non-Ashkenazi Cohanim and is estimated to be 4.2 +/- 1.3 ky old. These results support the hypothesis of a common origin of the CMH in the Near East well before the dispersion of the Jewish people into separate communities, and indicate that the majority of contemporary Jewish priests descend from a limited number of paternal lineages.

The point about a number of paternal lineages is a critical part of this paper. Here’s a figure which shows the proportions of Y haplogroups among Jewish populations (the use of the term “Israelite” for a contemporary nationality is novel to me, as they mean what I would have thought would be “Israeli”):


The prevalence of haplogroup J is a confirmation of the fact that Jewish male lineages are predominantly Middle Eastern. The concentration of the Cohen among only a few of these haplogroups is also striking, especially the fact that that bias spans both Ashkenazi (European) and non-Ashkenazi Jews. That suggests that a few male lineages monopolized the status of being a Cohen, which one would expect from the cultural practice (whereby sons inherit the status from their fathers), and, that those lineages predate the expansion of Jews and their subsequent diversification in the Diaspora.

The following table gives a sense of the time depth of these lineages:


The numbers above are thousands of years in the past, specifically the points at which the lineages within a haplogroup coalesce to a common ancestor.

Of course if these are the descendants of Aaron why exactly are their several haplogroups around? One hypothesis could be that J-P58* is Aaronic, but that other lineages have “introgressed” through “misattributed paternity.” The problem with such a model is that you should then see a modal haplogroup, and a a host of of much lower frequency unrelated haplogroups. As it is, some of the haplogroups, such as J-M410*, are at very high frequencies. If introgression occurred, it occurred early enough that we can think of the Cohens as a mix of lineages among the early pre-Diaspora priestly caste. But these authors ran some simulations which suggested this may explanation may need some nuance. They contend:

Thus, there would be a reasonably high probability that more than a single Cohen haplogroup could have survived in the Ashkenazi population since the initial founding of the priesthood ยป3,000 years ago…if we would be willing to accept an initial founding population size of >50 priests. However, our simulation results also suggest that it is highly unlikely that as many haplogroups as we actually observe…would persist under this simple model. Another model that deserves consideration is a metapopulation…in which semi-isolated communities maintain multiple Cohen lineages, each with a certain probability of extinction and replacement. In this model, multiple Cohanim lineages would then persist in the entire population, and new lineages would be expected to accrue among Cohanim over time. The presence of several founding lineages among the Cohanim of this survey–both shared between or specific to the Ashkenazi and non-Ashkenazi communities, as well as highly variable frequencies of these lineages among sub-populations within Ashkenazi and non-Ashkenazi communities…may lend support to a metapopulation model. Mutation alone does not provide an explanation for the multiplicity of Cohanim haplogroups, because the ages of most of these haplogroups predate the foundation of the Jewish people.

The logic of this is simple. Within one population over time most lineages will go extinct through random genetic drift, so that the frequency of one lineage will eventually fix at nearly 100% (this is one where is only one “mitochondrial Eve,” the other lineages slowly disappeared). Within several distinct populations which may have some gene flow the dynamics are more complex, as different independent random walks are liable to render different lineages extinct (though without gene flow fixation would occur more quickly).

Finally, here’s a map of J-P58*’s frequencies among Middle Eastern populations:


Citation: Extended Y chromosome haplotypes resolve multiple and unique
lineages of the Jewish priesthood, Hum Genetics, DOI 10.1007/s00439-009-0727-5

Related: How Ashkenazi Jewish are you?, Jacob’s Legacy: A Genetic View of Jewish History and Ashkenazi Jews are more European in ancestry.


  1. #1 Joshua Zelinsky
    September 29, 2009

    It really shouldn’t surprise anyone at all that the kohanim were a more or less separate population before the dispersal. Historical sources suggest a pretty separate population during the Second Temple period and the major dispersal occurred around the destruction of the Second Temple in 70 c.e.

    The really more interesting bit is how this shows that the founding population was clearly more than one single male. That’s making a lot of Orthodox Jews (and even some Conservative and Reform Jews) a bit unhappy. If one looks at least on the Jewish end of the blogosphere, the Orthodox end seems to be doing a pretty good job ignoring this results or dismissing it for pretty transparent reasons.

  2. #2 Arthur Shippee
    September 29, 2009

    Israelite is NOT for the nationality (which btw would include Israeli Arabs, &c.), but to mean Jew (I suppose Jew by blood, not a convert), and presumably who is not a priest, although that’s not clear, whether it’s inclusive or exclusive of priests.

    “Israelite” I as a historian use for citizens of ancient Israel, so this usage does look a bit funny to me.

  3. #3 Joel Hirsch
    September 29, 2009

    mr. zelinsky:
    for your information, the corresponding author of the study cited above (Human Genetics, aug. 2009), Dr. Karl Skorecki, is an orthodox jew.

  4. #4 thebob.bob
    September 29, 2009

    I wonder if you measured haplotypes in other professions that are passed from father to son in other cultures if you would find similar conservation?

    It’s also interesting that Yemeni, Bedouin and Jordanian men have the highest frequency of the J-P58* haplogroup.

  5. #5 Joshua Zelinsky
    September 29, 2009

    Joel, your point? The fact that it is making a lot of Orthodox Jews unhappy is completely independent of whether one of the authors of the study is frum. It might help to publicize that if one is trying to get people to be more accepting of the data. But beyond that, I fail to see its relevance unless you are pointing out that there are frum people who are reasonable but I don’t think anyone in this thread claimed otherwise.

  6. #6 Paul Ackman
    September 30, 2009

    Regarding “Israelite”
    Most modern Jews are descendants of the tribe of Judah, due to the loss of 10 tribes with the fall of the Northern Kingdom in 721 BCE. Descendants of the patriarch Jacob (aka Israel) prior to that time are typically referred to as Israelites, as they were most often NOT descendants of Judah. Also, in contemporary usage, Jews who are not Cohanim nor Levites (descendants of Jacob’s son Levi, who are not also Cohanim) are often referred to as Israelites. It is unrelated to Israeli citizenship.
    Under Jewish law, a Cohen cannot marry a convert, divorcee’, or widow whose husband was not a Cohen. So if you see a personal ad that says, “No converts, widows, or divorcees,” the advertiser is probably a Cohen.

  7. #7 barry youngerman
    October 1, 2009


    In this context, “Israelite” has a very specific meaning. Basically: a Jew who isn’t a Cohen. Greater detail:

    For certain ritual purposes in Orthodox Jewish communities, all Jewish men are categorized as either Cohen, Levi, or Israel. The first two had major ritual roles when the Temple was standing and continue to have minor roles today. The Cohens performed the sacrifices, the Levites helped them and handled the musical/poetic functions.

    In Biblical terms: the tribe of Levi (descending from Jacob’s son Levi) managed religious matters in general. Those descending from Aaron (merely one of Levi’s many descendants) were the elite priests (Cohanim), performing the sacrifices. The descendants of Levi who are not Cohanim (they don’t descent from Aaron) are just called Levi’im or Levites. I.e. Cohanim are Levites too, but in context there is no confusion.

    Other Jews, whether descended from Jacob’s eleven other sons or descended from converts, are simply called Israelites– meaning, ordinary Israelites or Jews. So, for totally confusing precision: All Cohens are also Levites and Israelites; all Levites are also Israelites. But again, no problem in context to keep the three castes distinct.

    I don’t know whether or not “Levi’im” are included within the Cohen group in genetic studies. The question is rather important.

    Some historians believe that the Cohen lineage was corrupted in Second Temple days (Persian, Greek, and Roman Empires), due to the privileges of the priesthood (the ruling class under the Hasmoneans). Even in post-temple days the minor privileges of a Cohen are greater than those of a Levite, which might have encouraged bogus Cohen lineages. There weren’t and aren’t as many reasons to pretend to be a Levite.

    Not surprisingly, there are usually far more Cohens than Levis in Jewish communities, especially among Sephardi Jews, although one would have predicted the opposite based on Aaron’s two sons and their children and grandchildren compared to thousands of other Levites at the time the Jews entered Canaan, in the Biblical account.

    But unless these studies use only Levites, they show that most Cohanim indeed have quite ancient lineage (pre-Persian Empire). Perhaps the bogus Cohanim faded away after the Temple was destroyed. By the way, as the ruling class the Cohanim tended to be Hellenizers; many had Greek names and educations and were basically “Reform Jews.” Perhaps only the loyal, genuine Cohanim remained within the Jewish fold. And perhaps there are more Cohens only because Levites didn’t bother to maintain their useless caste identity.