The data referenced herein is derived from this source (https://www.sciencedirect.com/science/article/pii/S0092867423011352 Lazaridis et al). A standard search regarding the ancestry of modern Greeks or Albanians yields various articles—such as those found in Science, Greek Reporter, and bioRxiv—which posit significant genetic continuity dating back to the Bronze Age. These publications often emphasize the "near-mythical" origins of the Greek population, frequently challenging the premises of migration or admixture as if such factors were inherently contradictory to their findings. Asking AI powered tools such as Chatgtp and Gemini will result to a similar driven narrative. For example, Gemini suggests: "Modern Greeks are primarily descendants of Bronze Age Aegean civilizations, specifically the Mycenaeans and Minoans, exhibiting remarkable genetic continuity over thousands of years."
These assertions rely heavily on PCA (Principal Component Analysis) model analysis, such as G25, because on a population distance level, modern Greeks and Albanians do indeed cluster closely with ancient Mycenaeans and Illyrians. However, the research cited above demonstrates that PCA-based models may be fundamentally insufficient and incapable of accurately modeling modern populations using ancient sources.
In the Supplemental Information (Data S1, Section 5) of the research, the methodology behind modern Balkan populations testing using qpAdm is described and the following is written: "First, we attempted to model the ancestry of present-day Balkan and Aegean populations as oneway models with different Balkan Bronze Age-Iron Age populations or with later groups whose ancestry derived entirely from pre-Roman Balkan populations:
Aegean_BA_IA
Croatia_IA
Serbia_BA
Albania_BA_IA
Bulgaria_EIA
CroatiaSerbia_RomanLocal: Individuals newly reported in this study from Serbia and Croatia dated 1-500 CE and modelled as a mixture between Croatia_IA and Aegean_BA_IA in section 4 above (Data S2, Table 6-7)."
If the 1-way models provided a good fit to the data, this would indicate genetic continuity in the Balkans since prior to the Roman period and no significant long-term demographic impact of the Slavic migration or other population movements in the region over the past ~2,000 years. However, all the models failed (Data S2, Table 8) with extremely low P-values (the highest pvalue by far being 0.00059 for present-day Albanians modelled as Albania_BA_IA), strongly rejecting population continuity in the Balkans since pre-Roman times, and documenting a recent history of mixture."
While Slavic admixture among Greeks and Albanians is generally accepted, the specific proportions are frequently downplayed in popular discourse. As seen in the picture below, the research suggests these figures are more substantial than often reported.
However, Slavic admixture alone does not explain the current genetic landscape. If it were the only factor, PCA analysis would show modern Greeks and Albanians shifting much further "North," away from their ancient counterparts. This research highlights a secondary, less-discussed migration event during the Early Roman period that significantly altered the region’s genetic profile.
This migration originated primarily from Western Anatolia, though the study also identifies individuals of Levantine and North African descent inhabiting the Balkans at the time. This resulted in a "Southwestern Anatolian" genetic profile emerging—a majority Anatolian component with minority Levant/MENA mixing—similar to that of modern Cypriots.
According to the research data (Data s2 table 8), the Anatolian-related ancestry is 9% in Romanians, 24% in Bulgarians, 44% Albanians and 11-84% in Greeks (depending on the specific region).
On a PCA plot, a modern population mixed with two very distant source populations will often plot in the geometric center of the two. This creates a misleading impression of continuity:
- The Southern Pull: The Anatolian/Levantine influx shifted the genetic profile of Greeks and Albanians away from their Bronze Age ancestors toward a more "Southern" and "Eastern" position.
- The Northern Pull: The subsequent Slavic admixture pulled the profile back toward the "North."
The result is that these two distinct migration events effectively "cancelled each other out" spatially on the plot. This places modern populations in a similar position to where they were 2,000 years ago, masking massive demographic shifts and making it appear as though no significant mixing ever occurred.
Allele-wise, these are however fairly distinct populations within the overall West Eurasian cluster. The Neolithic Farmer and Caucasus Hunter-Gatherer (CHG) profile of Bronze Age Greeks and Albanians—CHG-like ancestry was already present by 4,000 BCE—shares different alleles from the later Roman-era Anatolian migrants. This is not only due to differences in ancestry, as Roman Anatolians were much more CHG/Zagros-rich and Natufian-shifted, but also because of the divergence caused by the 4,000-year time gap. Similar is the case for the Eastern Hunter-Gatherer (EHG) related component; Mycenaeans and Illyrians possessed it in varying frequencies similar to modern Greeks and Albanians, but the Proto-Indo-European peoples had not yet developed the features associated with such ancestry today, such as specific mutations for pale skin and light hair, unlike the Slavs who possessed such features at a much higher frequency.
Think of it this way: a mother gives birth to two children all by herself, so her kids are 100% what she is. Those children both marry outside the family, but their children—first cousins—eventually marry each other. On a hypothetical PCA, the resulting great-grandchild will plot in between its two parents, similar to its great-grandmother, since her grandchildren are each half of what she is. Despite this similar positioning on a plot, they remain genetically very different from each other.
This could perhaps be why the Balkans, despite their plotting, remain among the most phenotypically diverse regions in Europe. Albanians, Northern Greeks and Bulgarians—despite plotting close to North Continental Italians and generally clustering closer to Central Europe than the Near East on a PCA—remain phenotypically "unpredictable."
As a Greek, I must observe that unlike Italy, there is no prevailing notion of a "northern" versus "southern" look in Greece, despite the genetic distance between different regions of the country being perhaps even more significant. This suggests that the diverse genetic inputs have resulted in a mosaic of traits, preventing the development of a predictable, regional frequency in appearance.
Cephalonia 1959, My grandmother's sister (originally from Lesbos) and her nephew
