R835 and the Uniparental DNA Lottery: The Ancestry Percentage Myth

Uniparental DNA and the Ancestry Percentage Myth

Uniparental DNA frequencies are poor proxies for a population’s overall autosomal ancestry. Founder effects, genetic drift, bottlenecks and unequal reproductive success can dramatically reshape the proportions of surviving Y-chromosome and mitochondrial lineages. Yet those proportions are routinely presented as if they were a numerical inventory of ancient ethnic ancestry.

A Y-chromosome percentage becomes a paternal ancestry percentage. A mitochondrial percentage becomes a maternal ancestry percentage. Sometimes the two are averaged, and the result is presented as the ancestry of the whole population. The arithmetic looks tidy. The biological justification is missing.

Start with R835, an Imperial-period man from Civitanova Marche, Italy. His paternal lineage belongs to the J1-B877 family associated with prominent Ashkenazi Cohen paternal branches. Yet the genome-wide comparison published in my earlier analysis places him strikingly close to both modern Southern Italians and Ashkenazi Jews. This is a concrete ancient individual whose paternal lineage and autosomal profile must be examined together.

R835 brings the distinction into focus. Ireland’s extraordinary R1b concentration demonstrates how dominant paternal lineages can become. The Albanian-derived Arbëreshë show how particular Y lineages can be disproportionately represented in historically documented migrant communities. Ashkenazi maternal founders demonstrate that the same problem applies to mtDNA.

R835: An Imperial Italian Cohen-Associated J1 Lineage With a Southern Italian–Aegean Autosomal Profile

R835 is one of the most important ancient-DNA examples in this entire discussion, and the reason is far more consequential than the discovery of another J1 chromosome in Roman Italy. This Imperial-period man from Civitanova Marche carried a paternal lineage belonging to J1-B877, the Cohen-associated branch documented in published phylogenetic research. Yet his genome-wide ancestry belongs in the Southern Italian–Aegean Mediterranean genetic landscape, with strikingly close affinities to both Southern Italians and modern Ashkenazi Jews. The significance lies in examining those two findings together.

In Sahakyan et al. (2021), Supplementary Table S3 identifies R835 as J1-B877, with an archaeological date of 27 BCE–300 CE and a site at Civitanova Marche. This is not simply a broad J1-P58 assignment. B877 occupies a much more specific position in the paternal genealogy associated with Jewish Cohanim, including Ashkenazi Cohen lineages.

Sahakyan and colleagues explicitly discuss the ancient Imperial Roman representative of this Cohen-associated branch. Earlier research, including Behar et al. (2017), identified Ashkenazi Cohen chromosomes within B877, together with related paternal lines among other Jewish populations. The ancient Roman finding therefore concerns a relevant ancestral branch in the same paternal family, not merely an unrelated ancient man who happened to share the very broad designation J1.

B877 is the assignment in Sahakyan’s phylogeny; it should not be silently equated with every narrower descendant branch distinguished by later sequencing. Nor does the shared ancestral branch prove that R835 himself was the literal genealogical ancestor of living Cohanim. But neither qualification diminishes the central discovery: an ancient Imperial Italian carried a paternal branch directly relevant to the genealogy of modern Cohen-associated J1 lineages.

That observation is already important. What makes it especially damaging to the simplistic equation of J1 with Judean autosomal ancestry is the second half of the evidence: the genetic profile accompanying R835’s Y chromosome does not resemble a distinctly identifiable Judean source population in the available genome-wide comparisons. Instead, his measured affinities are firmly Mediterranean, particularly close to Southern Italian and Ashkenazi reference populations.

The Crucial Finding: A Cohen-Associated J1 Chromosome Inside a Southern Italian–Aegean Genetic Profile

My previous R835 analysis reports a striking set of Global25 distances. Ashkenazi Belarus appears at 0.0235. Campania follows almost immediately at 0.0238, Calabria at 0.0239, Ashkenazi Lithuania at 0.0242 and Ashkenazi Romania at 0.0245.

This is not a picture of an ancient individual’s genome falling exclusively into a separate Levantine population while his paternal lineage happens to appear in Italy. It is an Imperial Italian with a Cohen-associated paternal branch whose whole-genome position overlaps the modern Southern Italian–Ashkenazi genetic neighborhood. Southern Italy’s wider affinities with the Aegean Greek genetic continuum are important to interpreting that location. The exact numerical distances displayed below are to Southern Italian and Ashkenazi references; they should not be misrepresented as separate measured distances to Greek populations.

The distance separating R835 from Ashkenazi Belarus is only 0.0003 smaller than the distance to Campania. Calabria is only 0.0004 farther away than Belarus. These are remarkably close results within the same numerical comparison. Their exact ranking is less important than the unmistakable fact that Southern Italians and Ashkenazim occupy almost identical positions among R835’s closest modern reference populations.

R835’s autosomal DNA is not behaving like a separate, predominantly Judean genetic population that must be explained independently of Southern Italians. The observed profile is entirely consistent with the broader Southern Italian–Aegean Mediterranean genetic landscape, and its near-equal relationship to Ashkenazi references is precisely what makes the case so important for investigating Ashkenazi origins.

There is a critical difference between observing a distant paternal relationship to an ancient population and claiming that the person carrying that lineage possessed the genome-wide ancestry of that population. R835 demonstrates why the distinction cannot be ignored. Even a specifically relevant Cohen-associated paternal branch can be carried by a man whose remaining ancestry is overwhelmingly shaped by the Mediterranean population history visible in the autosomal comparisons.

R835 Cannot Simply Be Reclassified as an Autosomally Judean Migrant

One interpretation of ancient Jewish-associated Y chromosomes found in Roman Italy is that they reflect men arriving from Judea or the wider Levant. Such movement certainly occurred within the Roman Mediterranean. But that historical possibility cannot be used as a substitute for testing the actual genome of the particular man under discussion.

If R835 is presented as a recently arrived, predominantly Judean man whose entire autosomal ancestry accompanied his J1 lineage from Judea, the genome-wide evidence creates an immediate problem for that depiction. His reported profile is extremely close to Southern Italians and Ashkenazim, rather than providing a demonstrated, separate ancient Judean genome-wide signature. His modeled deep ancestry also contains substantial Steppe-related and Anatolian farmer-related components characteristic of the broader Mediterranean landscape.

The argument is not that a paternal ancestor could never have traveled from the Levant, nor that one ancient Judean ancestor would necessarily remain detectable as a distinct autosomal component many generations later. Those are different questions. The issue is the unsupported leap from a B877 paternal assignment to the assertion that R835 himself possessed predominantly Judean autosomal ancestry.

That leap is not established by the data. No securely defined ancient Judean reference population has been demonstrated in these particular G25 comparisons as R835’s dominant genome-wide source. Nor do the displayed distances show a distinguishable Judean cluster into which R835 can simply be placed. A claim of absolute genetic incompatibility with every possible Judean ancestor would require more evidence than these comparisons provide; a claim that R835 was predominantly Judean is likewise not established by his Y chromosome.

The broader archaeology and population genetics of the Roman Mediterranean reinforce why this matters. Antonio et al. (2019) documented extensive genetic diversity and movement from across the Mediterranean in ancient Rome. The presence of eastern Mediterranean ancestry in Imperial Italy is not itself an ethnic diagnosis of Judean origin. The Aegean, Anatolia, the Balkans, Southern Italy and the Levant formed historically connected but genetically distinguishable regional contexts, and the correct comparison must examine actual genomes rather than assign identities from haplogroup names.

Sahakyan and colleagues interpreted the relevant Roman-period J1 lineages through Jewish migration from the Levant. Their proposed history deserves examination, but it is not an autosomal demonstration that R835 was predominantly Judean. In particular, the authors’ identification of an ancient B877 carrier cannot be used to bypass the very Southern Italian–Ashkenazi similarities revealed by the genome-wide analysis.

What requires explanation is not merely why Cohen-associated J1 existed in Imperial Italy. It is why an Imperial Italian carrying that ancestral paternal branch also possessed a genome-wide genetic profile so closely aligned with Southern Italians and modern Ashkenazim. That combination directly challenges any account that treats these paternal branches as self-sufficient evidence for a corresponding proportion of Judean autosomal ancestry.

The Autosomal Comparison Defies the Simple J1-Versus-R1b Story

The same published comparison includes R1554, an Imperial-period man from Urbino carrying R1b-U152. His modeled components provide an additional and particularly revealing contrast with R835.

The J1-B877 carrier has only 4.0% modeled Natufian-related ancestry and 3.6% modeled Levant PPNB ancestry in the separate G25 model. Meanwhile the R1b-U152 carrier registers 10.9% and 6.3%, respectively. R835 also carries a substantially higher modeled Steppe component: 24.2% compared with 15.0% in R1554.

Think about what that means for the familiar haplogroup argument. The man carrying the paternal branch most closely associated here with the Cohen genealogy has the lower modeled Natufian and PPNB components, while the man carrying R1b-U152 has higher values for both. The assignments are not Judean ancestry estimates, but their reversal directly demonstrates the absurdity of deriving expectations about an individual’s autosomal composition from broad paternal labels.

Calling J1 a Middle Eastern lineage does not tell us how much Middle Eastern-related ancestry R835 carried across his autosomes. Calling R1b-U152 a European lineage does not remove eastern Mediterranean-related ancestry from R1554. The actual comparison has to be performed. A haplogroup label cannot perform it for us.

R835 therefore supplies both sides of the case in one individual: a documented ancient connection to a Cohen-associated J1 paternal branch and an autosomal profile that must be understood within the Southern Italian–Aegean Mediterranean landscape. Treating the first observation as proof of a predominantly Judean genome means ignoring the second. That is precisely the kind of uniparental ancestry inference this article challenges.

Why Uniparental DNA Is a Poor Proxy for Overall Ancestry

Y-DNA follows the direct father-to-son path. MtDNA follows the direct maternal path: mothers transmit it to their children, but sons ordinarily do not pass it on. Autosomal DNA reaches descendants through many paths involving both sexes, with recombination reshuffling the inherited segments.

Those differences make uniparental DNA a poor stand-in for the entire pedigree. Suppose a man has only daughters. His Y lineage ends with him, but his ancestry does not. His daughters inherit autosomal DNA and can pass it onward. Suppose a woman has only sons. Her sons carry her mtDNA, but that maternal line ordinarily stops with them. Her autosomal ancestry can continue through their children.

Lineage extinction is not the same event as the extinction of someone’s descendants or genetic contribution. Confusing them erases ancestors who are absent from a particular uniparental tree but present elsewhere in the pedigree.

The reverse also happens. One paternal line can expand while its carriers repeatedly marry people with different ancestry. The Y chromosome persists along that line; the original man’s individual autosomal contribution does not remain intact beside it. The same principle applies to a maternal founder whose daughters and their daughters repeatedly have children with men of other ancestry.

A Simple Inheritance Calculation

Consider an explicitly simplified pedigree: one founder appears along one path, and each successive partner is unrelated to that founder and contributes none of the founder’s source ancestry. The expected autosomal share from that individual halves each generation. Along the appropriate uninterrupted paternal or maternal path, the lineage remains identifiable.

The calculation illustrates why an intact paternal or maternal lineage does not preserve an intact package of its founder’s autosomal ancestry. At sufficient genealogical distance, a particular descendant may retain no identifiable autosomal segment from that individual while still carrying the relevant Y or mitochondrial lineage.

Population frequencies introduce another layer of distortion into uniparental DNA comparisons. A founding group may contain an unrepresentative sample of its source population’s paternal and maternal lines. Random survival and reproductive differences then change their relative abundance. Bottlenecks intensify this sampling; subsequent population growth can multiply descendants of the surviving founders. Endogamy can preserve the resulting concentration without making the original sample representative.

This is the uniparental DNA lottery: surviving lines can become profoundly unrepresentative of the full ancestry behind a population. Social power can load the dice, too. The process includes random drift and nonrandom differences in reproductive success; it is not all one mechanism.

Ireland’s Uniparental DNA: What 85.4% R1b Actually Measures

Moore et al. (2006) examined 796 Irish Y chromosomes and reported that 85.4% belonged to R1b3, the study’s M269-defined category. Within it, the Irish modal haplotype and one-step neighbors accounted for 8.2% of the island-wide sample and 21.5% in the northwest. The researchers connected that narrower concentration with Gaelic dynastic history and unequal male reproductive success.

There are two distinct observations here: the dominance of a broad haplogroup and the expansion of a narrower paternal cluster. The entire R1b total must not be attributed to one medieval man. Nor do these data genetically identify a named legendary ancestor with certainty.

What they show is that the paternal distribution can become exceptionally concentrated. The social and demographic history of successful male lines matters enormously to the result.

Now try the familiar ancestry shortcut: because 85.4% of sampled Irish men carried R1b3, therefore 85.4% of their autosomal ancestry came from the original population in which that lineage arose. The conclusion does not follow. It skips migration history, changing populations, intermarriage, and the different inheritance of Y chromosomes and autosomes.

The Irish percentage is a uniparental DNA statistic. There is no automatic conversion into a Bronze Age, prehistoric, or other source-ancestry percentage. Determining those contributions requires genome-wide evidence and appropriate historical comparisons.

A rule that fails for Irish R1b does not become valid when the label changes to Ashkenazi J1. Fifteen percent of men carrying a paternal lineage is not, by itself, fifteen percent of the population’s ancestry from the Levant, Judea, Italy, or anywhere else.

That is the point of the Irish example. The extraordinary frequency makes the distinction impossible to miss: paternal-line dominance is a real demographic outcome, but its numerical magnitude is not automatically the magnitude of a genome-wide ancestry contribution.

Arbëreshë Uniparental DNA: E-M123 Rose to 23 Times Its Tosk Albanian Frequency

The Arbëreshë provide one of the clearest documented examples of why Y-chromosome frequencies cannot be mistaken for overall autosomal ancestry. These Albanian-speaking communities of Southern Italy descend substantially from historical Albanian migrations, especially from the southern Tosk-speaking population. Yet their modern paternal-line distributions can differ enormously from those of the ancestral population.

Consider haplogroup E-M123. In a peer-reviewed study published in the European Journal of Human Genetics, Sarno et al. (2016) examined the Y chromosomes of Arbëreshë communities in Calabria and Sicily alongside Albanian, Greek and Southern Italian reference populations. The study’s Supplementary Table S4 supplies the exact comparison that matters.

Among 104 sampled Tosk Albanian men, E-M123 appeared in just one individual: 0.96%. Among 119 Gheg Albanians, it was found in one man, or 0.84%. Across the combined 223 Albanian reference males, only two carried E-M123, a frequency of 0.90%.

Now compare those figures with the Arbëreshë. In the Calabrian community designated POL_SW, eight of 36 men carried E-M123: 22.22%. In the combined Sicilian Arbëreshë sample, eight of 44 carried it: 18.18%.

That is approximately 23 times the sampled Tosk Albanian E-M123 frequency in POL_SW and 19 times the Tosk frequency in the Sicilian Arbëreshë sample. These figures come from the same published research, use the same paternal haplogroup classification and can be compared directly without pretending that they represent autosomal admixture percentages.

The Critical Point: Arbëreshë Descend From Tosk Albanian Sources, Not a Population With 22% E-M123

The same Sarno study independently examines the Arbëreshë population’s origins. Its simulation-based analysis identifies Tosk Albanians as the most likely source population for the Calabrian Arbëreshë, with comparatively limited local Italian contribution in the best-fitting models. That finding is central: these are not two unrelated populations being compared simply because one happens to have more E-M123.

They are historically and genetically connected populations whose surviving paternal haplogroup frequencies have diverged dramatically. The Albanian source population represented in the study carries E-M123 at 0.96% among Tosks. Yet one of its historically connected descendant communities carries the same broad lineage at 22.22%.

The descendants did not need to replace their Albanian ancestry with an entirely different ancestry to produce this result. A comparatively small number of E-M123-bearing founders, uneven paternal-line survival, reproductive differences, population subdivision, genetic drift and subsequent endogamy are capable of producing a strikingly unrepresentative Y-chromosome distribution.

The study’s statistical analysis strengthens the observation. Sarno and colleagues found that E-M123 was significantly overrepresented among the Arbëreshë relative to the broader Balkan and Southern Italian comparisons, reporting P = 0.0015. The distinction is not merely an impression created by looking at the chart.

Nor was E-M123 uniformly elevated across every Arbëreshë community. Sarno’s two other Calabrian groups, POL_AREA and VAL_CRA, contained no E-M123 carriers in their samples, while POL_SW reached 22.22%. Within a related ethnolinguistic population, different settlements preserved sharply different paternal-line distributions.

This internal variation makes the interpretation even more compelling. An entire population’s deep autosomal ancestry does not simply swing from one ancient geographic source to another because a paternal haplogroup is common in one village and absent from samples in neighboring villages. The different frequencies are observations about surviving male genealogical lines and their local demographic histories.

Founder Effects, Genetic Drift and the E-M123 Lineages

Sarno et al. identified two distinct Arbëreshë-specific E-M123 haplotype clusters: one associated with Calabria and another with Sicily. Estimated ages for these clusters were approximately 1,133 ± 314 and 883 ± 228 years before present, respectively, placing their inferred common ancestors before the main historical Arbëreshë migrations into Southern Italy.

That chronology matters because the disproportionate representation of E-M123 cannot automatically be attributed exclusively to drift after settlement in Italy. Part of the founder structure may have developed earlier, among differentiated populations in the medieval Balkans. The researchers also discuss the possibility of distinct source groups and admixture during migration.

But that is precisely the demographic lesson. Founder effects and genetic drift do not have to begin at a modern national border. The founding group can already carry unusual paternal frequencies compared with the broader source population. Subsequent isolation, population growth and unequal family-line survival can preserve or further magnify the difference.

The Arbëreshë are therefore an example of a population whose paternal inheritance records a highly selective survival history. The fact that E-M123 is observed at 0.96% in the sampled Tosk reference and 22.22% in POL_SW does not prove exactly when every change occurred. It establishes the enormous present-day contrast, while the broader historical and genetic evidence identifies the population connection.

The genome-wide evidence is independently relevant. Sarno et al. (2017) used autosomal data and identity-by-descent sharing to investigate recent Balkan connections among Arbëreshë groups. This broader evidence is important because the argument does not depend on claiming that a Y-chromosome frequency alone proves autosomal descent.

The Sicilian Arbëreshë also have a more complex historical and genetic structure, including Greek and local contributions. In particular, Sarno et al. identified a substantial Tosk connection for Piana degli Albanesi, while Contessa Entellina showed stronger Greek affinities. The 18.18% Sicilian E-M123 result is a combined sample, and its population history should not simply be assumed identical to Calabria’s.

Why the Arbëreshë Comparison Matters for Ashkenazi Y-DNA Claims

Now apply the arithmetic commonly encountered in Ashkenazi uniparental DNA discussions. If approximately 15% of Ashkenazi men carry a particular J1 lineage, commentators sometimes treat that as evidence for approximately 15% autosomal ancestry from whichever ancient region they associate with the lineage.

Apply the same method to Arbëreshë E-M123. It would turn 22.22% of surviving paternal lines into 22.22% autosomal ancestry from a hypothesized ancient E-M123 source. Yet we know that the broader ancestral relationship of the relevant communities is substantially Albanian, and that the same haplogroup appears in only 0.96% of the study’s Tosk reference sample.

The method fails because the Y chromosome measures one uninterrupted paternal path, not the full mixture of ancestors behind that path. A founder lineage can multiply dramatically while the descendants’ genome-wide ancestry remains overwhelmingly connected to their broader ancestral population. The founding ancestor’s Y chromosome may survive intact across centuries even as his individual autosomal contribution is repeatedly diluted and recombined.

Both Arbëreshë and Ashkenazi population histories involve founder effects, demographic subdivision, endogamy and the differential survival of paternal lines, although their specific histories are not identical. The mechanisms that can distort uniparental frequencies do not stop operating when the population being discussed is Jewish.

This is the key result: a paternal lineage recorded at less than 1% in the sampled Tosk Albanian source population appears at more than 22% in one connected Arbëreshë descendant community. The enormous change in surviving paternal frequency is not a license to assign an equivalent change in autosomal ancestry.

The Arbëreshë example is not a speculative analogy built from imaginary founder scenarios. It is a quantitative comparison from a peer-reviewed genetic study, complete with sample sizes, observed haplogroup frequencies, source-population modeling and phylogenetic analysis. It demonstrates exactly why surviving Y-DNA branches must not be converted into ancient ethnic ancestry percentages by arithmetic alone.

Uniparental DNA Comparisons: J2 and E Are Not Missing From Southern Europe

There is a separate problem in uniparental DNA debates with the claim that Ashkenazi J2 and E frequencies reveal an unmistakably Levantine pattern: Southern European comparisons contain substantial frequencies of the same broad lineages. Treating Northern or Eastern Europeans as the entire European alternative obscures the question being asked.

My preprint on Ashkenazi ancestry modeling makes the importance of Southern Italian comparisons explicit. The table below returns to Semino et al. (2004), Tables 1 and 2, retaining the different E and J sample sizes.

Data correction: the original Apulian row is 2.3% E-M123 and 11.6% E-M78. Earlier summaries, including my preprint, assigned 11.6% to E-M123; that was incorrect. The values above follow the original table. Semino’s Albanian-Calabrian sample is also distinct from Sarno’s later POL_SW sample.

J2 is 23.2% in the Ashkenazi sample, 29.1% in Apulia and 22.8% in Calabria 1. E-M123 is 11.7% in the Ashkenazi sample and 13.2% in the Albanian-Calabrian sample. E-M78 is also well represented in the Southern Italian samples. Southern Europe cannot simply be removed from the comparison and then declared genetically irrelevant.

J1 is higher in the Ashkenazi sample than in these Italian samples. That difference deserves investigation at the branch level. It does not supply an autosomal percentage, and it does not cancel the J2 and E observations.

The argument is not that matching frequencies prove an Italian origin. It is that these frequencies cannot be presented as something Southern Europe lacks, and that no modern haplogroup percentage can simply be relabeled as the corresponding percentage of a population’s overall ancestry.

MtDNA Is Subject to the Same Uniparental DNA Lottery

Nothing about the maternal line escapes these principles. A mitochondrial lineage can become common because its female carriers survived a bottleneck and left many uninterrupted daughter-lines. Another woman can leave an extensive autosomal legacy while her uninterrupted maternal line disappears.

The classic Behar et al. (2006) study identified four major maternal founder lineages accounting for roughly 40% of Ashkenazi mtDNA. That is a striking concentration of surviving maternal lines. It is not a demonstration that four women supplied 40% of the autosomal genome, nor a count of all women who contributed to the population.

The geographic interpretation has also been contested. Costa et al. (2013) argued for substantial prehistoric European ancestry among Ashkenazi maternal lineages, including the major founders. Whatever position one takes on the geographic interpretation of these uniparental DNA branches, the counted object remains a maternal lineage.

Calling a lineage European does not turn its present frequency into European autosomal ancestry. Assigning it a Near Eastern origin does not make its frequency a Near Eastern autosomal percentage. The uniparental DNA rule does not change with the preferred geographic answer.

Founder Research Makes Maternal-Line Loss Central

Livni and Skorecki (2025) discuss extensive loss of original maternal signatures and propose distinguishing founder from subsequently absorbed mtDNA lineages. Their analysis assigns fewer than 15% of current carriers to absorbed lines and argues for a shared Near Eastern origin of the founding paternal and maternal lines.

The geographic conclusion does not follow from lineage survival alone. A model identifying which lines are early or late addresses timing and demographic persistence. Establishing where their founders came from requires additional evidence. A percentage of current carriers assigned to absorbed matrilines is not automatically the percentage of all historical women who joined the population, or their collective autosomal contribution.

Recognizing extinction while treating the surviving frequencies as an ancestry census would undo the very insight extinction supplies. The surviving maternal tree has missing branches; those missing branches can have living autosomal descendants.

Averaging Y-DNA and MtDNA Does Not Fix the Problem

Consider the claim that if paternal lineages are classified as 60% from region X and maternal lineages as 30% from region X, overall ancestry must be 45% from X. The average is arithmetically correct. Its use as an ancestry estimate is unsupported.

The two inputs are uniparental DNA frequencies from two unusually narrow inheritance systems. Neither input is necessarily a representative estimate of the autosomal contribution of all men or all women. Averaging two biased proxies does not reconstruct the missing ancestors.

Sex-biased admixture can be investigated. But doing so requires an explicit model and appropriate genomic evidence. A simple average of surviving paternal and maternal frequencies cannot substitute for that analysis.

Ashkenazi Founder Branches: A Common Ancestor Is Not an Ethnic Percentage

A downstream branch can reveal much more than the broad label J1, J2, E or R1a. It can identify relationships among carriers and help estimate when their surviving lines converged. That is precisely why broad haplogroup totals should not be treated as a substitute for branch-level analysis.

In Behar et al. (2017), the Ashkenazi Levite-associated R1a-Y2619 lineage has a star-like expansion and an estimated common ancestor around 1,743 years before present. The authors support a Middle Eastern origin while acknowledging that the timing and location of the founder’s acquisition of Levite status remain unresolved.

These findings illustrate both the power and the limits of the evidence. A successful paternal founder can leave many modern male-line descendants. His lineage’s prevalence does not measure his source population’s proportion of every descendant’s genome. A coalescence date is also not a record of a conversion, a migration, or the acquisition of a religious office.

The same distinction belongs in discussions of Cohen-associated J1. Hammer et al. (2009) found multiple paternal lineages among Cohanim and described evidence compatible with an ancient shared paternal history. Tofanelli et al. (2014) examined the limitations of using Y-chromosome and mitochondrial haplotype motifs as diagnostic markers of Jewish ancestry. Neither broad haplogroup membership nor a short matching motif is an ethnic certificate.

More refined sequencing improves the genealogy. It does not change what kind of genealogy is being measured. A precise uniparental DNA account of one paternal path is still not an account of every ancestral path.

R835 matters because the relevant paternal family is observed in ancient Italy alongside the reported Southern Italian–Ashkenazi autosomal affinities. The Roman-to-Ashkenazi paternal connections discussed on this site deserve examination at that level. A surviving paternal relationship does not establish the genome-wide ancestry of every person carrying it.

Wikipedia’s Ashkenazi Y-DNA Claims: Small Early Studies and the Missing Southern Italian Comparison

The central problem with Wikipedia’s treatment of Ashkenazi Y-DNA is not simply that it cites older research. It is that its conclusions about predominantly Middle Eastern paternal ancestry still depend heavily on studies whose comparison populations were poorly suited to distinguishing Levantine origins from Southern Italian and broader Mediterranean ancestry.

The Y-DNA of Ashkenazi Jews section of Wikipedia’s Genetic studies of Jews begins by invoking Hammer et al. (2000) to suggest that most Ashkenazi paternal lineages originated in the Middle East. It then develops that interpretation through early studies by Nebel, Behar, Shen and others, many of which relied on limited Ashkenazi samples, coarse Y-chromosome classifications or European reference populations that did not represent Southern Italy.

This is exactly the methodological problem examined in Part II of my preprint, How Ashkenazi Origins Were Mis-modeled, which systematically reassesses the early Ashkenazi Y-chromosomal literature. The issue is not that these studies failed to collect useful paternal-line data. It is that their reported geographic conclusions frequently go beyond what their population comparisons actually establish.

Wikipedia’s Ashkenazi Y-DNA Table Relies Heavily on Studies From 2001–2009

Wikipedia’s own Ashkenazi haplogroup table lists Hammer et al. (2009), Behar et al. (2004), Semino et al. (2004), Nebel et al. (2001) and Shen et al. (2004). This is the core group examined in Part II of my preprint. Three entries involve particularly small Ashkenazi samples: only 20 men in Shen, 79 in Nebel and approximately 80 in Semino. Behar’s 442 Ashkenazi samples constitute a substantially larger dataset, while Hammer’s 2009 study draws on a much larger collection of Jewish men from across the Diaspora.

But sample size alone does not fix a missing comparator. Increasing the number of Ashkenazi Y chromosomes cannot establish that a paternal pattern is specifically Levantine if Southern Italians, Sicilians and relevant Aegean groups are not directly and adequately represented in the geographic comparison.

The distinction is particularly stark in Hammer et al. (2000), one of the principal sources behind Wikipedia’s opening interpretation. The paper’s headline dataset contained 1,371 males from 29 populations. However, only 115 of those men were Ashkenazi, and the 81 non-Jewish Italians were treated as a general Italian sample rather than resolved into the Southern Italian regional populations relevant to Ashkenazi origins.

The study used 18 biallelic polymorphisms and broad paternal haplotypes, including the category designated Med. Such a design can investigate shared paternal variation across a broad geographic region, but it cannot establish that chromosomes also occurring in the wider Greco-Roman Mediterranean originated specifically in Judea. Its own discussion recognizes the broad circum-Mediterranean distribution of several important paternal haplotypes.

Wikipedia nevertheless uses this early study to introduce Ashkenazi paternal origins as predominantly Middle Eastern, without foregrounding the geographical alternative that the study did not adequately test: Southern Italy and the Aegean Mediterranean.

Semino (2004): The Critical Southern Italian Comparison Wikipedia Does Not Properly Explain

The exception in this early group of studies is Semino et al. (2004). Unlike the other studies summarized in the early Ashkenazi Y-DNA literature, Semino directly compared Ashkenazi Jews with multiple regional Southern Italian populations, including Apulia, Calabria and Sicily. It also distinguished relevant downstream branches of haplogroups E and J.

That distinction matters enormously. When Ashkenazi Jews are compared with Northern or Eastern Europeans, haplogroups such as J2 and E-M35 appear relatively unusual and can easily be characterized as Near Eastern. But Northern Europeans are not an adequate substitute for Southern Italians. The actual frequency of these lineages in Southern Italy must be measured rather than assumed away.

Semino’s published data show that Ashkenazi Jews and regional Southern Italian populations overlap substantially in major paternal haplogroup frequencies. In the Ashkenazi sample, J2-M172 was recorded at 23.2%. Apulia registered 29.1%, while Calabria 1 registered 22.8%. Those are not negligible Southern European frequencies. They are directly comparable to the Ashkenazi result.

The E-derived branches further demonstrate why regional sampling matters. Ashkenazi E-M123 stood at 11.7%, while the study’s Albanian-Calabrian community registered 13.2%. That Albanian-derived community must be distinguished from the native regional Italian samples, but its result reinforces the variability of paternal-line frequencies among Mediterranean populations. Meanwhile, E-M78 was recorded at 11.6% in Apulia, 16.3% in Calabria 1 and 12.7% in Sicily, compared with 5.2% among the sampled Ashkenazim.

J1-M267 was more frequent in the Ashkenazi sample, at 14.6%, than in the Southern Italian regional samples listed in Semino, including 7.1% in Sicily. This is a legitimate difference requiring branch-level and demographic investigation. It does not erase the J2 and E overlap, and it does not transform the J1 difference into an autosomal ancestry percentage.

The essential observation is that J2, one of the major Ashkenazi paternal haplogroups regularly described as Near Eastern, is present at virtually the same frequency in Calabria and at a higher frequency in Apulia. These directly measured results contradict any presentation implying that substantial J2 frequencies are geographically alien to Southern Italy.

Wikipedia includes Semino in its Ashkenazi haplogroup table, but the surrounding discussion never systematically examines what Semino’s Southern Italian comparisons mean for the predominantly Middle Eastern interpretation. Merely listing an article in the references is not the same as confronting the population comparison it supplies.

The Problem With Pooled Jewish Samples and Broad Haplogroup Labels

Hammer et al. (2009) is another important example. Its study included 1,575 Jewish men across the Diaspora, together with 2,099 non-Jewish men and a particular focus on 215 Cohanim. It is not a study of 1,575 Ashkenazi men. The researchers did distinguish Ashkenazi and non-Ashkenazi subgroups in some analyses, including priestly lineages, so the study should not be misrepresented as having no subgroup information. But the overall dataset and its specialized focus cannot be treated as a comprehensive Ashkenazi-versus-Southern Italian population comparison.

Its major contribution was resolving multiple paternal lineages associated with the Jewish priesthood, including the J-P58 branch underlying the extended Cohen Modal Haplotype. That can illuminate relationships among priestly paternal lines. It does not establish the overall autosomal origin of Ashkenazi Jews or demonstrate that similarly classified paternal branches were absent from Southern Italy and the Aegean.

Broad haplogroup labels create another problem. J2-M172 encompasses many ancient and geographically widespread paternal branches. E-M35 is an even broader designation containing several distinct downstream families. Treating entire haplogroups as inherently Levantine ignores the differentiation produced by prehistoric and historical migrations, including within Southern Europe itself.

A geographically broad paternal haplogroup cannot acquire a uniquely Judean identity simply because a study compared its Jewish carriers mainly with populations from Northern Europe and the Middle East. Geographic specificity has to be established through the relevant downstream branches, properly sampled alternative populations and historical evidence. It cannot be supplied by omitting Southern Italy.

Wikipedia’s Geographic Conclusion Exceeds the Comparisons Behind It

Wikipedia’s Ashkenazi Y-DNA section also repeats estimates of relatively small European paternal contributions, including the 5–8% figure associated with Behar et al. (2004) and earlier estimates near 12.5% or 23% depending on the assumed parental populations. The important point is how these estimates were constructed, not simply their numerical difference.

When a model uses German, Austrian, Russian or other Northern and Central European samples to represent a European parental population, it does not directly test Southern Italian ancestry. A haplogroup that is uncommon among those northern samples but prevalent in Southern Italy can be assigned incorrectly to the non-European side of a two-source calculation.

The problem is especially consequential when the supposedly Middle Eastern categories are broad haplogroups shared throughout the Mediterranean. The analysis can produce a low estimate of European paternal input not because Southern Italian ancestry has been excluded by genetic evidence, but because it was never properly represented among the candidate sources.

Even Wikipedia’s own discussion of Hammer et al. (2000) acknowledges that Ashkenazi Jews showed close paternal affinities to Southern European populations, particularly Greeks, and refers to migration through Italy in the development of Ashkenazi communities. Yet these Mediterranean observations do not lead to a systematic examination of the regional Southern Italian alternative within the section’s conclusions.

The central finding of Part II of my preprint is therefore directly relevant: the early literature repeatedly approached Ashkenazi paternal origins through an asymmetric comparison, in which Northern and Central European populations served as the European alternative while Southern Italy was generally absent. Semino et al. (2004) supplied the crucial regional comparison, showing substantial overlap in major Ashkenazi and Southern Italian Y-haplogroup frequencies. Wikipedia reproduces the early findings but does not adequately confront this methodological discrepancy.

The later discovery of specific ancient or modern Jewish-associated paternal branches does not repair that deficiency retroactively. R835, an Imperial Italian assigned to J1-B877, is particularly instructive. His relevant Cohen-associated paternal relationship accompanies a genome-wide genetic profile remarkably close to Southern Italians and Ashkenazim. That is exactly why the geographic history of a Y chromosome must be evaluated alongside the autosomal history of the people carrying it.

Wikipedia is not presenting a comprehensive modern reassessment of Ashkenazi paternal origins against the full Southern Italian and Aegean alternatives. Much of its explanatory structure remains tied to early Y-DNA studies whose population selection was incapable of resolving that particular question. The research can be cited for the actual paternal patterns it measured. It cannot be made to establish geographic exclusivity or an autosomal ancestry percentage that those studies never demonstrated.

Kevin Alan Brook: Surviving Lines Cannot Count All Historical Contributors

Kevin Alan Brook’s The Maternal Genetic Lineages of Ashkenazic Jews (2022) provides a clear example of the problem. The book discusses autosomal genetics and numerous specific lineages, so the criticism should address its actual reasoning rather than pretend it contains no wider genetic discussion. My earlier Brook review covers the broader issues.

On printed page 6, Brook describes his paternal lineage as a “clear case of Ashkenazic descent from Israelite men,” then discusses distant matches among modern Middle Eastern populations. Those relationships are relevant to a geographic hypothesis. They do not, by themselves, identify the ethnicity of the ancient ancestor linking the branches.

“Middle Eastern,” “Levantine,” “Israelite,” “Judean” and “priestly” are not synonyms. Each narrowing introduces a further historical claim. An ancestor shared with modern people in several countries does not arrive with an excavated identity card naming his ancient community.

The sharper uniparental DNA error appears on printed page 8. Brook argues that the limited number and frequency of surviving northern and eastern European-associated paternal lines indicate few historical male contributors, and uses that reasoning to constrain the number of unions producing offspring.

Several contributing men can also carry closely related or indistinguishable Y lineages. Counting the surviving distinguishable branches therefore cannot simply count all the men who contributed ancestry.

The same failure applies to maternal reconstruction. A woman missing from surviving mtDNA branches need not be missing from the ancestry of living people. Her descendants through sons are invisible to a strictly matrilineal census.

This uniparental DNA distinction is not a technical quibble. It changes the historical question the data can answer. Low surviving Y-line diversity can be evidence of founder history and lineage loss. Turning it directly into a small number of all historical male contributors skips the demographic processes under discussion.

Brook also recognizes a likely Roman Italian convert origin for J2-Y33795 on page 8 and rejects an automatic ancient-priest identification for that branch. That is a useful acknowledgment. The same demand for historical specificity should apply consistently when the proposed identification is Israelite rather than Italian.

Seven Uniparental DNA Claims That Do Not Survive Scrutiny

R835 Brings the Uniparental DNA Argument Back to the Actual Genome

R835’s importance extends well beyond the discovery of another J1 chromosome in ancient Italy. Here is an Imperial Italian assigned to the Cohen-associated J1-B877 paternal family whose genome-wide profile is remarkably close to modern Southern Italians and Ashkenazim. That is the central observation. A relevant paternal connection exists alongside a Mediterranean autosomal profile that cannot simply be relabeled as predominantly Judean on the strength of the Y chromosome.

His reported autosomal affinities are especially revealing because they place Southern Italian and Ashkenazi references at almost equal distances. The broader Southern Italian–Aegean genetic landscape provides a context in which that relationship can be investigated. The findings challenge the assumption that an Ashkenazi-associated paternal branch must have traveled through history attached to a fixed and distinctly Judean autosomal background.

Ireland demonstrates extraordinary paternal concentration and the importance of successful male lines. The Arbëreshë provide an even more direct numerical example of founder distortion: E-M123 at 0.96% in sampled Tosk Albanians, rising to 22.22% in one connected Calabrian Arbëreshë community. Ashkenazi maternal-founder research demonstrates why mtDNA requires the same scrutiny.

These observations attack the foundation of the uniparental DNA percentage meme. The problem is not merely that someone chose the wrong geographic label. The statistic itself is being made to answer a different question.

If the claim concerns overall autosomal ancestry, present autosomal evidence. If it concerns a particular ancient population, explain why the references distinguish that population from neighboring and historically connected alternatives. If it concerns a paternal or maternal founder, keep the conclusion at the level the evidence supports.

Neither “Levantine” nor “European” is a magic word that turns a surviving lineage into a representative slice of the entire pedigree. Nor can an ancestry calculator component do that work merely because it has an ancient name, as discussed in my IllustrativeDNA proxy-bias analysis.

Uniparental DNA can illuminate particular genealogical histories. Its frequency can also be profoundly distorted by who survived, who reproduced, and which uninterrupted lines happened to continue. Ignoring those processes does not produce a cleaner ancestry estimate. It produces false precision.

Sources and Evidence Notes

The ancient Y assignment, published population frequencies and founder studies below are separate from the author’s G25 comparisons. G25 distances are reference-dependent analyses and are not measurements of ancient Judean ancestry. R835’s branch assignment, the modern-population distance comparisons and the interpretations of his broader Mediterranean genetic affinities are identified separately so that each claim can be evaluated on its own evidence.

1. Sahakyan et al. (2021). Origin and diffusion of human Y chromosome haplogroup J1-M267. Scientific Reports 11, 6659. R835 identification: Supplementary Table S3. The main text and phylogeny identify J1-B877 as the Cohen-associated branch and discuss its Imperial Roman representative.

2. Parker. Ashkenazi Origins: Ancient DNA and Southern Italian Roots. Source of the reported R835 distances and R835/R1554 calculator comparisons; author’s analysis.

3. Moore et al. (2006). A Y-Chromosome Signature of Hegemony in Gaelic Ireland. American Journal of Human Genetics 78, 334–338.

4. Sarno et al. (2016). Shared language, diverging genetic histories: high-resolution analysis of Y-chromosome variability in Calabrian and Sicilian Arbereshe. European Journal of Human Genetics 24, 600–606. Supplementary Table S4 supplies E-M123 frequencies: Tosk Albanians 1/104 (0.96%), Gheg Albanians 1/119 (0.84%), Calabrian POL_SW 8/36 (22.22%) and combined Sicilian Arbëreshë 8/44 (18.18%). See also the population-source simulations, E-M123 haplotype networks and discussion of founder differentiation.

5. Sarno et al. (2017). Ancient and recent admixture layers in Sicily and Southern Italy trace multiple migration routes along the Mediterranean. Scientific Reports 7, 1984. Genome-wide and identity-by-descent evidence.

6. Semino et al. (2004). Origin, Diffusion, and Differentiation of Y-Chromosome Haplogroups E and J. American Journal of Human Genetics 74, 1023–1034. Tables 1 and 2; direct Ashkenazi comparisons with Apulia, Calabria and Sicily. Different E and J sample denominators are retained in the article.

7. Behar et al. (2006). The Matrilineal Ancestry of Ashkenazi Jewry: Portrait of a Recent Founder Event. American Journal of Human Genetics 78, 487–497.

8. Costa et al. (2013). A substantial prehistoric European ancestry amongst Ashkenazi maternal lineages. Nature Communications 4, 2543.

9. Livni and Skorecki (2025). Distinguishing between founder and host population mtDNA lineages in the Ashkenazi population. Human Gene 45, 201445. See also the institutional abstract.

10. Behar et al. (2017). The genetic variation in the R1a clade among the Ashkenazi Levites’ Y chromosome. Scientific Reports 7, 14969. Includes discussion of Cohen-associated J1-B877 paternal branches.

11. Hammer et al. (2009). Extended Y chromosome haplotypes resolve multiple and unique lineages of the Jewish priesthood. Human Genetics 126, 707–717. Study of Cohanim and additional Jewish and non-Jewish populations, including 1,575 Jewish men across the Diaspora.

12. Tofanelli et al. (2014). Mitochondrial and Y chromosome haplotype motifs as diagnostic markers of Jewish ancestry: a reconsideration. Frontiers in Genetics 5, 384.

13. Brook (2022). The Maternal Genetic Lineages of Ashkenazic Jews. Academic Studies Press. Specific arguments discussed here: printed pages 6–8.

14. Parker (2026). How Ashkenazi Origins Were Mis-modeled: Southern Italian Ancestry, Proxy Bias, and the Illusion of Levantine Intermediacy. Version 2. Part II, Re-assessment of Ashkenazi Y-Chromosomal Structure, examines the early studies’ sample sizes, geographic comparisons, haplogroup resolution and omission of Southern Italian populations. The Apulian E-frequency transcription is corrected in this article.

15. Wikipedia. Genetic studies of Jews: Y-DNA of Ashkenazi Jews. The encyclopedia section directly evaluated in this article, including its Ashkenazi haplogroup table and discussion of early European paternal-admixture estimates, checked October 17, 2026.

16. Antonio et al. (2019). Ancient Rome: A genetic crossroads of Europe and the Mediterranean. Science 366, 708–714. Original ancient-Roman genome study documenting the diverse population structure of Imperial Rome and central Italy.

17. Explore Your DNA. R835 ancient individual record. Sample metadata, autosomal component estimates and G25-based exploratory comparisons for the Imperial individual from Civitanova Marche.

18. Hammer et al. (2000). Jewish and Middle Eastern non-Jewish populations share a common pool of Y-chromosome biallelic haplotypes. Proceedings of the National Academy of Sciences 97, 6769–6774. A total of 1,371 males were tested, including 115 Ashkenazim, 81 Italians and 85 Greeks. The study used 18 biallelic Y markers and did not separate the Italian comparison into Southern Italian regional samples.

19. Behar et al. (2004). Contrasting patterns of Y chromosome variation in Ashkenazi Jewish and host non-Jewish European populations. Human Genetics 114, 354–365. Study of Ashkenazi Y-chromosome diversity, founder effects, population comparisons and paternal-admixture interpretations.

20. Nebel et al. (2001). The Y Chromosome Pool of Jews as Part of the Genetic Landscape of the Middle East. American Journal of Human Genetics 69, 1095–1112. Early comparative analysis including 79 Ashkenazi men, focused on Jewish and Middle Eastern paternal relationships.

21. Shen et al. (2004). Reconstruction of patrilineages and matrilineages of Samaritans and other Israeli populations from Y-chromosome and mitochondrial DNA sequence variation. Human Mutation 24, 248–260. The Ashkenazi Y-DNA comparison summarized by Wikipedia involved only 20 Ashkenazi men.