TrueSeeker · Verified claim report Case 91e02d1e19 · 2026-07-01

§ Claim under review · Mixed

DNA analysis of human remains recovered from several Japanese battlefield burial sites found that a significant percentage of those buried were female; in some locations, women accounted for roughly 30% to 50% of the individuals examined.

Circulating claim, as submitted.

Verdict

Source exists but framing is misleading

Confidence

Not recorded
§

Summary

The Instagram post's claim that DNA analysis of Japanese battlefield remains showed 30 to 50 percent were female is misleading. There is one real archaeological finding behind it: a 1989 study of about 105 skulls from a head mound linked to the 1580 Battle of Senbon Matsubara estimated that roughly one third were female. That study used skull shape (temporal bone morphology), not DNA, and the remains came from a mound of severed heads collected after the battle, not from a conventional burial of combatants. The "several sites" and "up to 50 percent" figures do not match any verifiable published research. The claim appears to trace back to a passage in Stephen Turnbull's 2010 popular history book *Samurai Women 1184-1877*, which has been repeated verbatim across the internet, often with the same misspelling. Onna-musha and figures like Tomoe Gozen are real parts of Japanese history and legend, but the specific archaeological centerpiece of this post is distorted.

§

The readings

key figures from the evidence
35 skulls

Female skulls identified at Numazu head mound (out of 105)

105 skulls

Total skulls examined at Numazu City Head Mound

§

Why this verdict

There is a genuine underlying finding: a 1989 osteological study estimated that roughly a third of 105 skulls from a head mound linked to the 1580 Battle of Senbon Matsubara were female. However, the Instagram post reframes this as "DNA analysis" across "several battlefield burial sites" with percentages ranging from "30% to 50%," none of which is accurately supported by the primary literature. The "DNA" characterization appears to originate as an error or embellishment in Stephen Turnbull's popular 2010 book and has been amplified through repeated verbatim copying online. The broader historical framing about onna-musha and Tomoe Gozen is directionally reasonable but the specific archaeological centerpiece of the post is materially distorted.
§

Evidence

The Instagram post's central statistic traces to a single passage in Stephen Turnbull's 2010 book Samurai Women 1184-1877, in which he wrote that at the Battle of Senbon Matsubaru between Takeda Katsuyori and Hojo Ujinao in 1580, DNA tests on 105 bodies revealed that 35 of them were female, and that two excavations elsewhere produced similar results . This single paragraph has been reproduced verbatim across dozens of websites, blogs, and social media posts, identifiable because it repeats the misspelling "Senbon Matsubaru" rather than the correct reading Senbon Matsubara, and searching that spelling returns essentially cloned entries .

The underlying study, however, was not a DNA analysis. According to independent tracing by SengokuStudies, there was a study conducted on the remains in the Numazu City Head Mound, but it was neither recent, nor did it involve DNA; in 1988 an excavation was conducted, and in 1989 a comparative osteological analysis of the bones was published . The Grokipedia summary of that 1989 paper describes it as follows: at the Numazu City Head Mound (kubizuka), associated with the severed heads collected after the 1580 Battle of Senbon Matsubara between Takeda Katsuyori's retreating forces and Hojo Ujinao's pursuers, morphological analysis of 105 skulls, focusing on temporal bone traits, identified approximately 35 as female, comprising about one-third of the sample . In other words, sex was estimated from skull morphology, not from genetic material, and the sample came from a head mound (kubizuka) of collected severed heads rather than a battlefield burial in the ordinary sense.

The "two additional excavations elsewhere" that Turnbull references have not been independently verified. On the E-Budo forum, researchers noted they could not locate primary sources for those additional sites and questioned whether the deceased were combatants at all, as opposed to people decapitated in the aftermath of battle.

On the historical figures cited: Tomoe Gozen was an onna-musha, a female samurai, mentioned in The Tale of the Heike, but there is doubt as to whether she existed as she does not appear in any primary accounts of the Genpei War .

§

Findings

What's accurate 4

  • A real archaeological finding does exist: roughly one-third of about 105 skulls at the Numazu head mound associated with the 1580 Battle of Senbon Matsubara were estimated to be female.
  • Onna-musha (female warriors of the bushi class) are a documented category of premodern Japanese women, and some received combat training, primarily for household and castle defense.
  • Tomoe Gozen is a real legendary figure in Japanese literature, celebrated in *The Tale of the Heike*.
  • Mainstream scholarship agrees that women's roles in premodern Japanese warfare have historically been underappreciated in popular accounts.

What's misleading 4

  • **Fabricated methodology (marketing as evidence / temporal overreach):** The post says "DNA analysis," but no DNA study of these remains has been documented. The actual 1988 to 1989 study used skull morphology. Viral versions of the story that name "Kyoto University's Forensic Anthropology Department" and describe "recent DNA analysis" appear to be embellishments not supported by any traceable publication.
  • **Exaggeration / subgroup generalization:** The post refers to "several Japanese battlefield burial sites" with figures of "30% to 50%." The only site with a specific documented figure is Numazu, at roughly 33%. The additional sites referenced by Turnbull have not been located in verifiable form, and the "up to 50%" figure has no clear source.
  • **Context omission (site type):** The Numazu remains come from a kubizuka, a mound of severed heads collected after a battle, not a battlefield burial of combatants. Whether the female skulls belonged to warriors, camp followers, non-combatants killed in the aftermath, or victims of head-taking practices is not established by the osteological data alone.
  • **Species/setting mismatch, loosely applied:** Framing skull-based sex estimation as "DNA" overstates the certainty and modernity of the technique.

? What's uncertain 4

  • Whether "two additional excavations elsewhere" (cited by Turnbull) actually exist in the published archaeological literature.
  • Whether the female individuals at Numazu were combatants, non-combatants killed in the campaign, or a mix.
  • The exact identity of any modern DNA study that would justify the Instagram post's phrasing. No such peer-reviewed study was located.
  • The historical existence of Tomoe Gozen as an individual person, as opposed to a literary figure.
Distortion flags fabrication exaggeration subgroup generalization
§

Sources

3 of 6 linked to records
[1]

Stephen Turnbull, *Samurai Women 1184-1877* (Osprey, 2010)

secondary
https://deremilitari.org/2015/02/stephen-turnbull-samurai-women-1184-1877-eads/ ↗
[2]

SengokuStudies, "Does DNA Prove Female Soldiers at Senbon Matsubara?"

secondary
https://medium.com/@sengokustudies/does-dna-prove-female-soldiers-at-senbon-matsubara-6e6045969528 ↗
[3]

Grokipedia entry on Onna-musha citing the 1989 osteological paper: https://grokipedia.com/page/Onna-musha

unknown
This citation could not be independently verified.
[4]

Military History Matters, "Female Samurai Warriors": https://www.military-history.org/feature/samurai-wars/female-samurai-warriors.htm

unknown
https://www.linkedin.com/posts/axios-security-group1_leadership-historymatters-samurai-activity-7354131043139862528-Z4LY ↗
[5]

E-Budo forum discussion tracing the viral spread of the claim: https://www.e-budo.com/archive/index.php/t-52741.html

unknown
This citation could not be independently verified.
[6]

Wikipedia, "Tomoe Gozen": https://en.wikipedia.org/wiki/Tomoe_Gozen

unknown
This citation could not be independently verified.
How links are chosen. A source is linked only when the address comes from the investigation's own retrieval or from a registry lookup (PubMed, Crossref) that matches the citation's title and year. Author lists shown as registry-verified come from the registry record, not from the report text. Citations that cannot be matched are labeled, never guessed.
This is one case on the record See the full case, browse the archive, and search every checked claim on TrueSeeker Open on trueseeker.com →