Scientists are locking horns over a Viking map of America. One group claims to have evidence that the document is a fake, but another says that the parchment on which it is drawn predates Christopher Columbus's 1492 discovery of the New World.

The Vinland Map's ink dates to the twentieth century, argues one team — adding that the parchment's age is irrelevant. But other researchers have roundly criticized the ink analysis, saying that the carbon dating of the parchment nudges researchers closer to the conclusion that the map is genuine.

If it is, then the parchment, which depicts Leif Ericsson's journey to what is now Canada in the tenth century AD, is the oldest map of America. It came to light in 1957, when Yale University paid $1 million for it. The idea that the Vikings made it to America before Columbus was new and controversial at the time, although archaeological findings have since confirmed it.

But some have always doubted the map's authenticity, and in 1974 crystals taken from the map were found to contain anatase, a form of titanium dioxide used in white paint which has only been manufactured since 1923.

Now chemists Robin Clark and Katherine Brown of University College London have studied the map's ink with a technique called laser Raman microscopy, which uses laser scattering to reveal chemical composition. The map's lines are made of two different substances, they report1. Their black centres are carbon; their yellowish edges are composed of anatase. Clark says that anatase shouldn't be present in a map from the fifteenth century. He thinks that a forger overlaid a yellow line with a black one to create an aged appearance.

But other experts have blasted these conclusions. They are “singularly unconvincing”, says physicist Thomas Cahill of the University of California, Davis. “There's no example of any forger trying to do double inking.”

Jill Pasteris of Washington University in St Louis, Missouri, a leading authority on laser Raman microscopy, says that the London team has overstated its results, and that the researchers did not even use a microscope to scan the parchment properly for anatase.

“The presence of anatase is by no means conclusive proof that the map is a forgery,” says Jacqueline Olin, a retired Smithsonian Institution chemist who has studied the map for 30 years. More analysis is needed to answer the anatase question decisively, she says.

Olin is part of a team led by Garman Harbottle, a specialist in carbon dating at Brookhaven National Laboratory in Upton, New York, that has dated the map's parchment to between 1423 and 1445 (ref. 2).

“It doesn't prove the map's authenticity, but it's another piece of evidence to be put in the scales,” says Harbottle. “It sets the bar higher for those who want to show that the map's a fake.” Carbon-dating the ink might settle the matter, Harbottle suggests. But current carbon-dating techniques require more material than the ink could provide.