Millions now proudly embrace labels like “Viking,” “Yamnaya,” or “Greek and Balkan.” But leading scientists warn: these test results are not what they seem. The categories they offer are built on assumptions, approximations, and genetic illusions.
According to geneticist Carles Lalueza-Fox, the narratives sold by companies like 23andMe and Ancestry.com often oversimplify, or even distort, what DNA can truly reveal. While customers may feel connected to a grand historical lineage, the science behind these tests is far more complex, and far less definitive, than advertised.
For many, learning they are “part Viking” feels like a personal discovery. But experts argue that these interpretations don’t hold up under scrutiny. The labels used in consumer DNA reports are based on rough estimates of genetic similarity to ancient populations, which themselves are arbitrarily defined. They are neither universal, nor objective, and most importantly, they don’t reflect the real, messy history of human migration and genetic mixing.
Reference Populations Are Scientific Constructs
Ancestry tests rely on comparisons between your DNA and those of so-called “reference populations.” These populations are usually built from ancient genomes taken from archaeological samples, assigned labels like “Romans,” “Steppe nomads,” “Anglo-Saxons,” or “Slavs.” But as Lalueza-Fox explains, these categories are not fixed. Scientists select them based on their specific research goals, and their definitions can vary significantly.
These classifications create the illusion of a static human past, where populations stayed in one place long enough to develop isolated genetic identities. In reality, human movement has been constant. From the Out-of-Africa migration some 60,000 to 70,000 years ago, to the forced displacements of the Atlantic Slave Trade starting in the 16th century, history has been defined by mobility. That mixing makes it nearly impossible to tie a person’s genome to a single place or group with any real certainty.
Even if a person’s DNA shows similarities with ancient genomes from Viking-era Scandinavia, that doesn’t mean they are “part Viking.” According to Lalueza-Fox, these associations are built on statistical approximations, not concrete genealogical links. Genetic ancestry testing offers resemblance, not lineage.
Genetic Diversity Is Extremely Limited
At a glance, human beings appear strikingly diverse. We differ in skin tone, facial structure, height, disease resistance, and more. But genetically, humans are among the least diverse species on the planet. Approximately 99.9% of the human genome is shared by all people. The remaining 0.1% accounts for the small genetic variations that tests analyze.
In population genetics, defining a “population” is also fraught with complexity. In theory, it means a group of individuals that mate among themselves and are distinct from others. But in practice, especially with humans, these boundaries are blurred. No human group is truly isolated, and traits like language or geography don’t always correspond with distinct genetics.
To demonstrate this, geneticists James Kitchens and Graham Coop visualized the scale of human genetic variation. Starting with 2.9 billion nucleotide sites, they analyzed 609 individuals from the Americas and found only 39 million observable variants. Restricting the analysis to common variants, those found in more than 5% of people’s chromosomes, reduced the number to 10 million. That’s a minuscule portion of the genome.
They also found that the vast majority of these variants are not exclusive to any single population. In one example, 99 Utah residents with Northern and Western European ancestry showed 5,726,377 variants, while 96 African Caribbeans from Barbados had 8,018,649. The overlap was massive. According to the findings, even variants considered “regional” are typically present in multiple parts of the world.
Race and Identity in Genetics Are Collapsing Concepts
The outdated concept of “race” still influences how people interpret genetic ancestry, but the science has moved on. While some visible traits, like skin pigmentation, show clear geographic trends, full-genome studies have revealed that these patterns are superficial. According to ZME Science, humans exhibit gradual genetic variation across continents, not distinct racial boundaries.
This shift is reflected in the scientific literature. The use of the term “Caucasian,” once common in population genetics papers, has dropped dramatically, from 12% in the mid-20th century to less than 1% in the early 2000s. In its place, broader continental terms like “European” gained popularity, rising to 42% of papers between 2009 and 2018.
Yet even these continental categories are imperfect. Comparing a relatively homogenous island nation like Iceland to a vast and diverse territory like Russia reveals the limitations of grouping people by nationality. Scientists have adopted different approaches; some define populations by country, others by culture, but both methods come with significant trade-offs.
In non-human species, the situation is clearer. Isolated populations like Madagascar lemurs or the extinct dodo of Mauritius show well-defined genetic distinctions due to physical separation. But humans have always been connected. The concept of “endemism,” where a species is confined to a specific location, simply doesn’t apply to our species.
Instead of broad population models, modern geneticists increasingly work with ancient individuals tied to specific places and periods. This allows them to reconstruct movement and change over time without assigning modern identities to ancient people. According to Lalueza-Fox, this method offers a more accurate, nuanced view of the past, without collapsing it into modern labels.