Did an American Indian tribe descend from early Welsh explorers?

Illustration by Slug Signorino

Cecil replies:

Welsh Indians? Dylan Thomas, Richard Burton, and Cochise? Sorry, ain’t seeing it, and I have a hard time believing anybody ever could. But the legend lives on, one of many farfetched stories about European or even extraterrestrial origins for Native American culture, the assumption evidently being that the real Native Americans were too backward to develop this stuff on their own. The Welsh angle is distinguished mainly by its persistence, having bubbled up like swamp gas for more than 400 years. I don’t expect to stamp it out now, but it lacks any basis in fact.

Owain Gwynedd, who died in 1170, was a real Welsh prince, but his putative son Madoc goes unmentioned in contemporary annals. Later a seafarer named Madoc, not necessarily related and possibly mythical, showed up in medieval Welsh literature. One story has him colonizing an island paradise, location unspecified. But overtly fantastic elements — he also visits a magnetic island — mark this as medieval fiction.

In 1580 Welshman John Dee revived Madoc to counter Spain’s American claims, presenting Queen Elizabeth with “Title Royal” to Madoc’s American discoveries. Dee often let politically motivated conjecture trump rationality; he decried the intrusion of fiction into history but claimed Greenland for Britain because King Arthur had conquered it.

Nonetheless, Dee-inspired travelers soon began reporting Welsh-speaking Indians everywhere. All told, 15 identifiable and at least 5 unidentified tribes, ranging from Peru to Canada, were equated with the Madogwys (Madoc’s people). The tribes generally lived in areas remote from white settlement, unknown except by reputation. Once they became familiar and their Welshness demonstrably absurd, the legend moved on.

Some reports came from known hoaxers, but many were the product of delusion or ignorance. Eyewitness accounts are largely by nonspeakers who mistakenly thought they recognized Welsh. Slightly more credible is Welsh-born pastor Morgan Jones, who in 1686 said that years before he’d met Welsh-speaking Doeg Indians. Other sources, however, say the Doegs spoke an Algonquian language, like most of their neighbors. At best, Jones, who seems to have conflated elements of three distinct voyages, misremembered. At worst, he was his era’s Jim Bakker. His congregation upbraided him for “ill-life and conversation,” the dog.

The legend first attached itself to the Mandan people of North Dakota around 1780. The Mandan were promising candidates because they were reputedly fair-skinned, rich, and civilized, and because of a few linguistic coincidences. Hoping to prove the Madoc connection, Welshman John Evans wintered with the Mandan in 1796-’97. Upon returning he told supporters, “There is no such People as the Welsh Indians.”

End of story? If only. Artist George Catlin, meeting them decades later — and apparently ignorant of Evans’s visit — kept the legend alive, writing that “a great many” Mandan were “as light as half breeds.” Nothing odd there — white men had been pooling genes with Mandan women for nearly a century at that point. True, explorer La Várendrye had noted a “light-colored” minority on first encountering the tribe in the late 1730s. But even Catlin admitted that no more than 20 percent looked non-Indian, easily explicable by normal inter- and intratribal genetic variation. Other reports of blond-haired Mandans may have been misinterpretations of achromotrichia — premature hair whitening, which by Catlin’s estimate affected 10 percent of the Mandan population, including children.

The Mandans’ wealth — gained through trade — and their large walled towns weren’t exaggerated. But so what? If an agricultural people has impressive architecture, they must be from Europe?

The Madogwys, who according to legend alighted in the New World at Mobile, Alabama, of all places, are sometimes associated with hilltop stone “forts” (possibly ceremonial) in and around Tennessee. However, archaeological evidence shows Indians built them 1,000 years before Madoc’s time. Believers cite Roman coins found nearby. Twelfth-century Welshmen carrying second-century Roman change? Here’s a simpler explanation: Roman coins are cheap (a few bucks on eBay) and often collected — and lost — by careless modern children.

Not even believers claim the Mandan language is essentially Welsh — it’s too obviously Siouan. Catlin believed it had Welsh loanwords, giving ten examples in his book. He must have flunked Irony 101, because he proceeds to supply a table comparing Indian languages in which three of his “Welsh” words (and many other Mandan terms) have recognizable cognates in another Siouan language. Any two unrelated languages are likely to have a few matches by chance alone.

Over the years many have wanted this or that American tribe, artifact, etc, to be Phoenician, Hebrew, Trojan, Greek, Roman, or what have you. One bunch of Europeans is now generally conceded to have made a pre-Columbian foray into the new world: the Norse. The Welsh won’t be number two.

Cecil Adams

Send questions to Cecil via cecil@straightdope.com.