A piece of parchment the size of a placemat cost a benefactor about 300,000 dollars together with two companion volumes, kept chemists, librarians and map historians busy for more than sixty years, and ended in 2021 with one sentence from the responsible curator: "The Vinland Map is a fake." The parchment really does date from the 15th century, and the wormholes really do match a medieval manuscript. What is false is the drawing on it. This article follows the case step by step: what each test could prove, what it could not, and why the Norse voyages to North America remain untouched by any of it.
The Vinland Map is held today by the Beinecke Rare Book & Manuscript Library at Yale University. It shows the Old World, Europe, Africa and Asia, without elaborate decoration and in ink that has since faded. Yale's press text compares its size to a placemat. The wormholes in the parchment have been patched. Two halves are held together by a strip glued to the back; a Danish examination confirmed in 2005 that they are now separate pieces, while leaving open whether they ever formed one sheet.
The map owes its fame to the Atlantic. West of Europe lie Iceland, a Greenland drawn as an island and, to the southwest, a large landmass labelled "Vinlanda Insula"; the spelling "Vinilanda" also appears in the literature. A Latin legend reports the discovery of this island and a visit by Bishop Eirik. If the map had been made around 1440, as its first editors assumed, it would have put North America on a European world map half a century before Columbus.
In 1957 the map surfaced in the European antiquarian trade, bound into a slim volume with the so-called Tartar Relation, a Latin account of a journey into the Mongol Empire in the 13th century. The Italian dealer Enzo Ferrajoli de Ry brought it, with the London bookseller Irving Davis, to the British Museum. The museum did not buy. Accounts differ: the Yale Daily News says the museum rejected it as a forgery, while the US programme NOVA stresses Ferrajoli's refusal to disclose its history. J. Huston McCulloch of Ohio State University, who examined the volumes himself, reports that the recent binding and mismatched wormholes also played a part.
That same year Laurence Witten, a young dealer from New Haven and a Yale graduate, bought the volume for 3,500 dollars and offered it to his old university. Yale noticed what London had noticed: the wormholes in the map did not match those in the Tartar Relation. The answer came in 1958 with a third item, a dismantled part of a copy of the Speculum Historiale, the great encyclopedia of Vincent of Beauvais (books 21 to 24). The Yale librarian Thomas Marston bought it for 75 pounds from the catalogue of Davis & Orioli, Irving Davis's firm, and passed it to the Wittens. Once the pieces were laid together a picture emerged: the map had sat at the front of the Speculum, the Tartar Relation at the back, and the worm tracks ran through them all.
Yale could not raise the price. The banking heir and Yale alumnus Paul Mellon bought the volumes on condition that their authenticity be tested, and later gave them to the university. The Yale Daily News and McCulloch give about 300,000 dollars, while an older web chronology speaks of roughly one million. Mellon also required silence while the scholarly edition was prepared. As a result, hardly any expert outside the small circle saw the map before publication.
On 11 October 1965, the day before Columbus Day, Yale presented the map together with the book "The Vinland Map and the Tartar Relation". The authors were Skelton and Painter of the British Museum and Thomas Marston of Yale. They dated it to around 1440, in the setting of the Council of Basel (1431–1449) suggested by the accompanying manuscripts. According to Yale News, the timing caused outrage among New Haven's Italian Americans: a map disputing Columbus's priority, presented on the eve of his holiday.
Five years earlier Helge Ingstad had located the house remains at L'Anse aux Meadows on the northern tip of Newfoundland. The excavation was still running when the map appeared. So the public received two pieces of Vinland news almost at once, one from the ground and one from a vault. Only one of them held.
As early as 1966 an international conference met at the Smithsonian Institution; its papers appeared in 1971 (edited by Washburn). The objections came at first not from the laboratory but from the humanities.
Greenland looked too good: an island of surprisingly correct shape and orientation, whereas Scandinavian accounts of the time attached it to northern Russia as a peninsula. Critics thought the Latin name form "Erissonius" for Leif Eriksson fitted the 17th century better. Especially telling is the formula "regionumque finitimarum", which styles Bishop Eirik bishop of Greenland "and the neighbouring regions". It resembles a phrase in the work of the scholar Luka Jelić (1864–1922), and Peter Foote showed that Jelić had turned a French scholarly expression into Latin. I know these points second-hand. McCulloch regards some of the objections as weaker than often presented, and none alone would have settled the matter.
In 1967 a first material observation followed. A preliminary examination at the British Museum found that the ink was not conventional iron-gall ink, the mixture of oak galls and iron salt that medieval scribes normally used. According to Wikipedia the Speculum volume had also been rebound with plastic thread, available only since about 1950.
In the early 1970s Yale engaged the Chicago chemist Walter McCrone. Sources date his study to 1972, 1973 or 1974, depending on whether they mean the examination, the report or the publication; the core paper by Walter and Lucy McCrone appeared in the Geographical Journal in 1974. Using polarised light microscopy and microsamples, the team found particles of anatase in the yellowish lines of the ink.
Anatase is a crystalline form of titanium dioxide. It occurs in nature, but in a different form: the particles on the map were, as McCrone described them, rounded and very uniform, about 0.2 micrometres across, like those of the industrially made white pigment that reached the market only in the 20th century. Statements on the exact start range from 1917 to the late 1920s; the sources agree on the century. In 1974 Yale's library announced that the famous map might be a forgery.
The anatase dispute was no sham fight. In 1987 a group led by Thomas Cahill at the University of California, Davis, published measurements by PIXE, a method in which a particle beam makes the elements in a sample emit X-rays. They found only traces of titanium, not the large quantities McCrone's picture implied. Four of 33 lines tested showed none. The new edition of Yale's book in the mid-1990s included a contribution from Cahill but none from McCrone. For a while the map counted as rehabilitated.
More objections followed. Jacqueline Olin argued from experiments that anatase can arise as a by-product when iron-gall ink is made from certain vitriols. Others considered clay minerals, or sand used to dry the ink. A Danish group led by Larsen said in 2009 that it saw no signs of forgery. Kenneth Towe of the Smithsonian insisted that crystal size and shape were the decisive test, one the sand theorists had not made under the microscope. In 2002 Katherine Brown and Robin Clark of University College London used Raman spectroscopy, which identifies crystal forms by their vibrational pattern, to detect anatase in the lines but not in the adjacent parchment. Clark added in 2004 that more than twenty medieval manuscripts had yielded no anatase, and that the ink was essentially a carbon ink. In 2008 Towe, Clark and Kirsten Seaver argued that Cahill's figures were too low by a factor of at least a thousand. I know that figure only from summaries, not from the paper itself.
Each side had a real measurement. The question was what a measurement at individual points says about a whole map. Only an analysis of the full surface could answer it.
In 2002 Donahue, Olin and Harbottle published a date for the parchment in the journal Radiocarbon: 1411 to 1468 at 95 per cent probability, more narrowly 1423 to 1445. That fit the Council of Basel strikingly well, and many headlines made it sound as if the map were now genuine. The authors themselves wrote the opposite: "Our dating research does not prove that the Vinland Map is authentic." What had been dated was the animal, not the drawing.
The first sample gave a value that only nuclear weapons tests of about 1950 to 1963 can explain. So the parchment had been treated in that period with a substance that stricter chemical cleaning could remove. The authors place it around 1958 to 1962, when the map was already on the market. Exactly what the substance was has not been conclusively settled.
A forger who wants to fake a 15th-century map reaches for blank 15th-century parchment if he can get it. Radiocarbon answers only the question of the earliest date at which he could have drawn.
The wormholes above all convinced the first investigators, because they are hard to fake. A test with transparent Mylar film at Yale's Institute for the Preservation of Cultural Heritage (IPCH), reported by Yale News in 2018, confirmed that the tracks in the map and the companion volumes line up. Yale itself concludes that the map once sat at the front of the Speculum, probably on an end sheet.
Raymond Clemens, curator of early books and manuscripts at the Beinecke, drew the decisive line in 2018: the match shows that parchment and books were once bound together, but says nothing about when the map was drawn. A blank leaf at the start of a manuscript gets the same worm tracks as the written pages behind it. Anyone who removes it centuries later and draws on it inherits the holes for free. That a forger counted on exactly this is an obvious thought, but it cannot be clearly proved.
According to the Yale Daily News, a new project began at Yale in 2017. At the IPCH, Anikó Bezur, Richard Hark and the conservator Paula Zyats, among others, worked with methods McCrone had not used: multispectral imaging, surface images under changing light angles, X-ray fluorescence (XRF) across the entire surface, and Raman maps instead of single points. On 21 September 2018 researchers presented first results at a symposium at Mystic Seaport Museum in Connecticut. Hark reported, as Wikipedia summarises the livestream, that the anatase in the ink lines fitted modern manufacture.
That same year Clemens added an argument that needs no chemistry. A map bound into a book loses area in the gutter, the fold between the pages. Mapmakers allowed for that and drew past the edge: "Maps bound into books would be overdrawn to account for the space lost in the gutter." The Vinland Map does not do this. It is also off-centre, and Vinland, Greenland and Iceland look added on the outside rather than planned into the world. According to Yale News the world outline is based on a map by the 15th-century Italian cartographer Andrea Bianco.
I did not find a separate Yale finding about trimming of the leaf in the Yale texts I read. What is documented is the missing overdrawing at the fold and the two separate halves; that the leaves come from the Speculum is a reasoned assumption. Trimming cannot be clearly established from this.
The independent researcher John Paul Floyd offered two further pointers. In 2013 he made public that the map repeats copying errors from an engraving of the Bianco map published by Vincenzio Formaleoni in the late 18th century. He also judged the short note on the back only half genuine. In 2018 Hark supported this idea with measurements showing two different inks.
In September 2021 Yale went public with its result. The new element is coverage. McCrone had tiny samples and Cahill individual points; the Yale team scanned the entire leaf with macro X-ray fluorescence and found titanium in lines and lettering, while iron, sulphur and copper appeared little or not at all. In the Vinlanda Insula area titanium ran high, accompanied by some barium, which fits titanium white of the 1920s. Raman microscopy confirmed anatase spread widely across the map. Under a field-emission scanning electron microscope, particles from the map and the Tartar Relation resembled, according to Yale, a pigment made in Norway in 1923; nothing pointed to a natural origin.
The cross-check came from fifty Central European manuscript fragments of the 15th century in the Beinecke's holdings. They contained far less titanium and far more iron. That disposes of the old objection that titanium simply turns up in any old ink.
For the question of intent, the back of the leaf weighs more. It carries a short Latin note, according to Yale originally in iron-gall ink, reading roughly "second part of the third part of the Speculum": probably an instruction for the binder of a medieval manuscript. Someone overwrote it in titanium-bearing ink, so that it now reads like a binding instruction for the map. A leaf merely mistaken for old would have had no need of that. Hark saw in it "pretty strong evidence of a motive to deceive". In the Tartar Relation, too, infrared light revealed altered text, likewise in modern ink.
Clemens's verdict was: "The Vinland Map is a fake." And: "There is no reasonable doubt here." Fresh radiocarbon measurement in 2018 confirmed that parchment and paper of both manuscripts date from about 1400 to 1460. The team planned scholarly publications, and Clemens hoped for an article in a cartography journal. I did not see a full journal paper by the team during this research; the details here come from Yale's announcements. The map stays at the Beinecke because, as Clemens puts it, it has become a historical object in its own right.
Witten was still writing in 1989 that he had no idea where the map came from beyond Ferrajoli. Ferrajoli himself was later convicted of theft from the library of La Seo cathedral in Zaragoza. Accounts differ on scale and period: Wikipedia gives 700 volumes and the 1950s, another account 583 items lost between 1960 and 1965. I did not find the date or the sentence.
In 2013 Floyd came across the Spanish connection: a Speculum volume with the Historia Tartarorum was shown at an exhibition in Madrid in 1892/93, on loan from the Archdiocese of Zaragoza. No map is mentioned there. This fits the assumption that the two medieval manuscripts came from Zaragoza, with the map as a later addition. The chain of custody is not firmly proved by this.
Nobody knows the culprit. In 2004 Kirsten Seaver proposed the map scholar Father Josef Fischer; more recent provenance research and a handwriting assessment make that unlikely, and Floyd sets out explicitly to clear him. Clemens puts it drily: the evidence shows the forgery, but "it doesn't tell us who perpetrated the deception". The motive too, whether money, scholarly dispute or a joke, cannot be clearly established.
No solid argument for the voyages ever rested on it. The evidence comes from three independent directions.
The first is early written sources. Around 1075 Adam of Bremen names an island called Winland, as our article on Adam of Bremen explains. In the 12th century the Icelander Ari Þorgilsson, in his Íslendingabók, the Book of the Icelanders, writes in passing about the first settlers in Greenland:
"They found there traces of human dwellings, both in the east and in the west of the land, along with pieces of boats and stonework, from which it can be understood that the same kind of people had travelled there who settled Vinland and whom the Greenlanders call Skrælings."Ari Þorgilsson, Íslendingabók, ch. 6, from the Guðni Jónsson edition on heimskringla.no, my own translation
Ari explains Greenland by means of Vinland, as if everyone knew the name. The second direction is the sagas: the Saga of the Greenlanders and Eirik's Saga, written down some 250 years after the events and contradictory in details. The essentials are in our pieces on Leif Eriksson and Thorfinn Karlsefni. Icelandic annals also record for 1121 that Bishop Eirik set out to seek Vinland, and for 1347 a ship that had come from Markland. So the map's Eirik is a genuine figure of tradition.
The third direction is the ground. At L'Anse aux Meadows, from 1960, led there by George Decker, Helge and Anne Stine Ingstad found the remains of eight sod buildings in the style of Iceland and Greenland around the year 1000, along with iron slag, boat rivets, a bronze ring-headed pin and a spindle whorl. Three butternuts from the bog come from further south, since the tree does not grow on Newfoundland, and point to voyages into the Gulf of St Lawrence. In 2021 a group led by Kuitems dated three pieces of wood using the radiocarbon jump of 993, left in tree rings by a cosmic event. All three were felled in 1021 and worked with metal. The study appeared in Nature in October 2021, a few weeks after Yale's announcement.
None of these lines needs the map. The archaeologist Birgitta Wallace, who dug at L'Anse aux Meadows, and the historian Karen Kupperman said just that at the 2018 symposium, according to one report. What is proved is voyages and a short-lived base. How far "Vinland" reached and exactly where it lay remains interpretation; more on that in our article on Vinland.
Like the Kensington Runestone, the map promised Scandinavians in America and found eager supporters. In the stone's case it was above all the language that gave away the date. With the map it was the reverse: the material was truly old, and that is exactly what kept the debate alive. Only the combination closed the case: Latin and the map image supplied the first suspicion, chemistry the ink, provenance research the trail.
Is the parchment of the Vinland Map forged? No. It dates from the 15th century and probably sat as a blank leaf in the Speculum Historiale. The ink is modern.
Who forged the map? That is unknown. Josef Fischer was proposed but is now considered unlikely; Ferrajoli's role as seller is certain, authorship is not documented.
What is anatase? A crystalline form of titanium dioxide. Natural anatase exists, but according to Yale the particles found on the map resemble an industrial pigment of 1923.
Can the map be seen? It remains at the Beinecke Library at Yale. Medievalists.net points to a high-resolution image in Yale's online collection.
So did the Vikings reach America or not? Yes. Norse people were on Newfoundland by 1021 at the latest, as wood from L'Anse aux Meadows shows. That holds independently of the map.
Vinland · The Kensington Runestone – Vikings in Minnesota or a Nineteenth-Century Forgery? · Leif Erikson.
Makoto Yukimura's Vinland Saga opens with the Danish assault on London in 1013 and ends on a North American shore. In …
Read more →Monks recorded who attacked their monasteries, and an entire age acquired its villain. Yet in 782 Charlemagne, on his …
Read more →Few periods of European history have been shaped by cinema as much as the Viking Age. Most people meet their first …
Read more →The famous "Viking prayer" – "Lo, there do I see my father …" – is not a prayer handed down from the Vikings. It was …
Read more →