Y-DNA Ancestral Lines of Ashkenazi Jews

Based upon the methodology described here and using the sample set described here compiled in December 2018, the current Ashkenazi Jewish population includes at least 47 ancestral Y-DNA lines that were identifiable as of that date through Family Tree DNA's Big Y testing.  

An analysis performed in December 2022 using information concerning dating and geographical origins available through FTDNA's Discovery utility indicates that the Ashkenazi branches of 13 of these haplogroups are a level or two downstream from the SNP identified in the 2018 analysis, and that another branch is likely non-Ashkenazi.  The list below is annotated to identify updates to the analysis.

The ancestral lines identified in 2018 are as follows (pages with SNP trees for each cluster are linked below):

As shown on the linked pages, many of these Ashkenazi Y-DNA ancestral lines follow a common pattern.  First, the lines typically date back to about the second half of the first millenium CE, i.e., each line typically split into two or more branches about 1,000 to 1,500 years ago.  For the most part, each of the branches from these ancestral lines includes men who, based upon their close autosomal matches to Ashkenazi Jews, are likely to be descended from Ashkenazi Jews on their direct male lines.  This strongly suggests that the shared direct male ancestors of such branches belonged to the Ashkenazi or proto-Ashkenazi population in the second half of the first millenium CE.  

Second, there is typically substantial branching within those subbranches that presumably dates back to the time when the Ashkenazi population began its massive expansion out of a small bottlenecked population, about 700 to 1,000 years ago.   

Because most of the 47 ancestral lines identified above had branched at least once prior to the post-bottleneck expansion -- and some of the lines had branched several times by that point -- it appears that most of the current Ashkenazi population is descended on their direct male lines from perhaps 120 to 150 Ashkenazi Y-DNA lines that existed as of the time of the bottleneck.  

Y-DNA is, of course, passed down only on the direct male line.  Accordingly, transmission of Y-DNA over a 700- to 1,000-year period would, at an average of 25 years per generation, require 28 to 40 consecutive male generations.  As a result, it is very likely that a significant number of Ashkenazi Y-DNA lines that existed as of the time of the bottleneck no longer exist.  Similarly, it is likely that there are some Ashkenazi Y-DNA lines that survive today only as single branches, and that there are other Ashkenazi Y-DNA lines that are not found in the sample set used for this study.  Further testing and analysis will likely identify some such branches.    

In some instances, the Ashkenazi Y-DNA ancestral lines include branches that share a direct male ancestor who lived perhaps 3,000 to 4,000 years ago.  To the extent that both branches are comprised primarily of Ashkenazi men, this strongly suggests that their shared direct male ancestor belonged to a Jewish or proto-Jewish population in the Near East.

More frequently, however, the Ashkenazi ancestral lines start in the second half of the first millenium CE, without brother Ashkenazi branches that date back to Biblical times. Where such lines include multiple branches that are each comprised predominantly of Ashkenazi men, one can infer that the shared direct male ancestor of those branches likely belonged to the Ashkenazi (or proto-Ashkenazi) population. 

The analysis is more complicated where such lines include multiple branches, not all of which are comprised predominantly of Ashkenazi men.  If there is one Ashkenazi branch and one non-Ashkenazi branch, as a logical matter it would be equally likely that the shared direct male ancestor of the two branches was Ashkenazi or non-Ashkenazi; as the proportion of non-Ashkenazi branches to Ashkenazi branches increases, the likelihood that the shared direct male ancestor was not Ashkenazi increases.  

At one point, it was generally assumed that Ashkenazi Y-DNA clusters commonly believed to have originated in the Near or Middle East -- such as haplogroups E, J1, and J2 -- were of Near Eastern origins, while Ashkenazi Y-DNA clusters commonly found among Europeans -- such as R1a and R1b -- likely entered the Ashkenazi population in Europe.  

Over the past 15 years or so, studies based on more precise Y-DNA testing have shown these assumptions as to the origins of some of these Ashkenazi R1a and R1b Y-DNA clusters to be overly simplistic, and often incorrect.  

For example, the R1a-Y2619 Ashkenazi Levite cluster was once hypothesized to have European origins because R1a is very common in Europe.  However, the Eupedia distribution map for R1a-Z93, the upstream cluster to which R1a-Y2619 Ashkenazi Levites belong, shows that R1a-Z93 is most common from the Near East through India.  The 2017 Behar study discussed here confirmed that the R1a-Y2619 Ashkenazi Levite cluster has Near Eastern origins by finding that the R-M582 Y-DNA cluster to which R1a-Y2619 Ashkenazi Levites and their closest matches belong includes several Iranian branches and share a direct male ancestor who lived about 3,000 years ago.

Similarly, because R1b is predominant in Western Europe and the British Isles, it was often assumed that R1b lines entered the Ashkenazi population in Europe in the past millenium or so.  However, there is one branch of R1b -- R1b-Z2103 -- that, as shown by a Eupedia distribution map, is very common in the Near East.  There are two R1b Ashkenazi clusters -- R1b-Y19852 and R1b-FGC14600 -- that belong to the R1b-Z2103 branch and therefore are likely to be of Near Eastern origins.  As discussed below, three of the main R1b clusters discussed in this analysis -- R1b-FGC20759, R1b-FGC8580, and R1b-L4 -- share the same ages and branching characteristics of other Ashkenazi clusters, and therefore are either of Near Eastern origins or reflect a European contribution to Ashkenazi Y-DNA dating back to within centuries of the formation of the Ashkenazi/proto-Ashkenazi population.    

As more testing is done and analyzed, we will have a better idea as to which Ashkenazi Y-DNA lines likely originate in the Near East, and which lines more likely reflect Y-DNA lines that entered the Ashkenazi population in Europe.

Based upon the current sample set, the following Ashkenazi ancestral Y-DNA lines share the characteristic -- common to most of the major Ashkenazi Y-DNA lines -- of initial branching in the second half of the first millenium CE, followed by considerable branching beginning about 700 years ago (red text provides updates based on the 2022 analysis): 

(1) E-Y14891; 

(2) E-Y6923 (2022 analysis: E-Y6940); 

(3) the E-BY932 subcluster of E-PF1975 (2022 analysis: E-Y36800); 

(4) E-BY7450 (2022 analysis: E-BY7500); 

(5) E-Z17697; 

(6) E-BY11082; 

(7) the G-FGC35913 subcluster of G-BY764 (2022 analysis: G-BY764); 

(8) the G-FGC31712 subcluster of G-FGC249; 

(9) the G-L201 subcluster of G-L1324; 

(10) the I-Y11261 subcluster of I-S23612; 

(11) I-Y23115; 

(12) J1-L816; 

(13)-(14) the J1-BY101 and J1-ZS4763 subclusters of J1-L823 (2022 analysis: J1-ZS4763 is largely non-Ashkenazi); 

(15) J1-S12192; 

(16)-(17) the J1-CTS4459 and J1-ZS4307 subclusters of J1-Z18297 (2022 analysis: J1-ZS2616 and J1-ZS2497); 

(18) the J1-ZS10589 subcluster of J1-PF7263; 

(19) the J1-FGC5215 subcluster of J1-FGC5230; 

(20) J1-BY67 (2022 analysis: J1-FGC17481); 

(21) the J1-ZS1682 subcluster of J1-F450; 

(22) J2-L556; 

(23) J2-Z30930; 

(24)-(25) the J2-L254 and J2-FGC30508 subclusters of J2-FGC4992 (2022 analysis: J2-FT69390 and J2-Z38408); 

(26) J2-BY268 (2022 analysis: J2-FGC21083); 

(27) J2-Z43500 (2022 analysis: J2-Z43501); 

(28) Q-Y2198; 

(29) Q-YP1003; 

(30) R1a-Y2619; 

(31) R1b-FGC20759; 

(32) R1b-A11711; 

(33) R1b-FGC8580 (2022 analysis: R1b-FGC8654); and 

(34) R1b-L4.  

(T-Y31472 has two Ashkenazi branches that date back to the first millenium CE (2022 analysis: T-Y105938)).

Many such Ashkenazi Y-DNA ancestral lines therefore appear likely to be Near Eastern in origin, although further information concerning upstream branching would in many instances be necessary to confirm whether the lines entered the Ashkenazi or proto-Ashkenazi population in the Near East or in Europe.

Based upon the current sample set, the following Ashkenazi ancestral Y-DNA lines do not evidence the typical pattern of initial branching in the second half of the first millenium CE:

(1) the G-L201 subcluster of G-L1324; 

(2) I-BY424; 

(3) L-PAGES00116; 

(4) the R1a-YP6547 subcluster of R1a-Y2632 (2022 analysis: R1a-BY33370); 

(5) R1a-Y1013; 

(6) R1b-Y19862; 

(7) R1b-Z18106; 

(8) R1b-FGC21047; 

(9) R1b-FGC14600; 

(10) R1b-L408; 

(11) the R2-FGC13201 subcluster of R2-F1092 (2022 analysis: R2-FGC13211); 

(12) T-BY11520; and 

(13) the T-PAGES00113 subcluster of T-CTS8862. 

Accordingly, these Ashkenazi ancestral Y-DNA lines are more likely than the Ashkenazi Y-DNA lines identified in the first table above to have entered the Ashkenazi population in Europe.  However, there is a significant likelihood that further test results for some of these lines may show either or both (1) earlier Ashkenazi branching or (2) upstream matches of Near Eastern origins.   

Observations

1.  The Ashkenazi ancestral Y-DNA lines identified in this analysis through a SNP-based analysis of a sample set compiled based upon autosomal matches coincide to a great extent with the ancestral lines identified by Wim Penninx on his website jewishdna.net through the initial use of STR marker values.  Accordingly, the analysis set forth herein is duplicative to a considerable extent of Wim's painstaking analyses, but it also provides independent confirmation of the validity of his STR-based clustering.  (I thank Wim for his thoughtful comments and suggestions.) 

2.  The term "ancestral lines" is used here to refer to those Y-DNA lines that include two or more branches that are each comprised in significant part of men who are likely of Ashkenazi descent on their direct male lines, as evidenced by their substantial autosomal matches to people known to have four Ashkenazi grandparents.  Most of the ancestral lines discussed in this analysis first branched about 1,000 to 1,500 years ago; several other ancestral lines first branched about 700 to 1,000 years ago, at the time that the Ashkenazi population expanded out of a tight bottleneck.   

3.  The rate of Big Y testing varies considerably among Ashkenazi Y-DNA clusters. This likely results from the facts that, inter alia: (1) men with relatively close matches on Y-DNA STR testing are more likely to do upgraded testing, including Big Y testing; and (2) men in certain Ashkenazi Y-DNA clusters (such as the R1a-Y2619 Ashkenazi Levite cluster to which the companion website levitedna.org is devoted) have made concerted efforts to encourage Big Y testing among other men in such clusters.

As a result, certain Ashkenazi Y-DNA clusters continue to be underrepresented in Big Y test results, making it more difficult to identify and date those clusters. Further testing is likely to flesh out information concerning these clusters.  

The number of Ashkenazi Y-DNA ancestral lines is likely to increase, as men from less common or undertested branches do Big Y testing.

4.  Because the initial sample set was gathered by using autosomal matches to a cohort of Ashkenazi Jews to identify people who were likely of Ashkenazi descent on their direct male and female lines, the sample set will exclude those people who are of Sephardic or Mizrahi descent on their direct male or female lines unless those men also have significant Ashkenazi ancestry.  

Because (1) the Y-DNA clusters studied herein go back far longer than the autosomal matches used to compile the sample set and (2) FTDNA's Y-DNA Haplotree provides information concerning the number of men in each cluster, the Y-DNA clusters identified herein are highly likely to include a substantial proportion of men who are Sephardic on their direct male lines (and some men who are Mizrahi on their direct male lines).  This is true, however, only for Y-DNA clusters that are found in both the Ashkenazi population, on the one hand, and the Sephardic and/or Mizrahi population, on the other hand.  

5.  FTDNA's Y-DNA Haplotree includes the results for a substantial number of Ashkenazi men who have not submitted their data to YFull for inclusion in YFull's YTree.  As a result, (1) the FTDNA Y-DNA Haplotree identifies considerable branching not included in the YFull tree, and (2) YFull's estimates of the time to a shared direct male ancestor for a branch is sometimes based on a small sample size (which increases the margin of error for age estimations).