A hypothetical descent from a couple married circa 1800 — one Native American, one of mixed African and European descent — traced to two living siblings and three first cousins in 2026. The family story survives eight generations. The DNA, on average, does not.
Their children carry, on average, 50% Native American, 25% African, and 25% European ancestry — the full starting bundle passed to the next generation.
Each row assumes 6–8 children per couple and marriage to an unrelated spouse each generation — the standard model used to estimate expected autosomal DNA. Native American ancestry (rust) is tracked in full; African and European (gold/teal) dilute the same way.
Inheritance isn't a smooth halving — it's a random reshuffle (recombination) every generation. Two full siblings, descended from the same parents, can inherit very different actual segments. First cousins diverge even more. Below is one plausible outcome for five living descendants who all carry the identical average expectation of 0.39% Native American ancestry.
Oral history can preserve a real relationship in distorted form, absorb an unrelated tradition, or, occasionally, simply be wrong. A DNA result is one more piece of evidence to weigh against the documentary record — not an automatic override of it.
A "great-grandmother" remembered across generations of retelling is very often a 4th or 5th great-grandmother in the paper trail. Each additional generation halves the expected DNA again — the difference between a strong signal and none at all.
By six to eight generations, the average expected share of any single ancestor's DNA — as modeled above — falls below what most consumer tests reliably report, even when the relationship is completely accurate.
Because recombination is random, it's entirely possible to inherit zero detectable DNA from a specific ancestor while a full sibling or first cousin inherits a measurable trace — as shown in Section 03. A negative result for one person is not a negative result for the family line.