At the northern tip of Newfoundland lie the remains of eight buildings raised in turf and timber by Norse people in the 11th century. It is the only confirmed Norse site in North America, and a bronze pin, a spinning weight, slag and boat nails are what proved it. This article stays in the trenches: who dug when, what the houses were for, what the finds say, how long people stayed, and whether the place fits the Vinland sagas at all. On those last two questions the specialists still disagree.
The site sits at the very tip of the Great Northern Peninsula, looking across the Strait of Belle Isle to the coast of Labrador. The modern village of L'Anse aux Meadows lies about a kilometre north on Medee Bay. The Norse houses face neighbouring Epaves Bay. Parks Canada, the federal heritage and national parks agency, places them "a short distance inland on a narrow gravel terrace by a water-logged peat bog", beside a small stream called Black Duck Brook.
Stand there today and you see hardly a tree. That is misleading. According to the Great Unsolved Mysteries in Canadian History project, balsam fir and spruce once grew just behind the site. Basque and French fishers from the 16th century, and settlers from 1835, cut them for fuel, buildings and fish-drying racks, pushing the forest line back some 15 kilometres. The Norse had building timber close at hand.
The location is still strange. The Norse normally chose sheltered spots, and this one takes the full force of wind and storm. The project offers an explanation: the tip is easy to find from any direction, and anyone coming south along Labrador suddenly sees land to port as well. That is a suggestion, not a finding.
The discovery of 1960 did not come out of nowhere. In 1914 the Newfoundlander William Munn had already pointed to the Strait of Belle Isle. Wallace calls him publisher of the Evening Telegram; the Conception Bay Museum describes a Harbour Grace merchant. The Finnish geographer Väinö Tanner later reached a similar conclusion; the year varies, with Birgitta Wallace giving 1939 in The Canadian Encyclopedia and the project page giving 1944. In 1948 an American army engineer, inspired by Munn's pamphlet, searched about 25 kilometres to the southwest, and in 1956 the Danish archaeologist Jørgen Meldgaard worked in the same area. None of them found the site.
So the credit is shared. The idea was decades old, and the local knowledge lived in the village.
Helge Ingstad, a Norwegian writer and explorer, travelled the coast in 1960 looking for landing places from the sagas. According to Wallace his daughter Benedicte was with him. In the village a local man, George Decker, led him to what Parks Canada calls a group of "overgrown bumps and ridges". That proved nothing yet.
The excavation itself was directed by Helge's wife, the archaeologist Anne Stine Ingstad. Parks Canada puts it more broadly, saying the couple led an international team from Norway, Iceland, Sweden and the United States; Wallace names Anne Stine as leader. According to the project, digging took place in the summers of 1961 to 1964 and 1966 to 1968, usually from mid-June to September. The familiar phrase "eight years" does not match this exactly: it simply counts through from 1960 or 1961 to 1968.
By the end, eight buildings and their surroundings were exposed. The small finds were modest: about 300 objects, roughly half of them left by Indigenous users of the site, and among the Norse pieces mostly fragments of iron nails. Parks Canada calls the slag and boat nails the strongest factor in identifying the site as Norse.
From 1973 to 1976 Parks Canada took over, first under Bengt Schönbäck and in 1976 under Birgitta Wallace. A smaller season followed in 2002, and Jenneth Curtis and Todd Kristensen worked there from 2007 to 2009. Memorial University of Newfoundland also dug test trenches outside the main area.
The decisive step of the 1970s led away from the houses and into the bog next door. Wood had survived there in three layers. The middle one was Norse, with Indigenous finds below and above it. That gave a clean stratified sequence. Those years produced about 2,030 objects, again about half Indigenous. This time the Norse material was mainly wood: carpentry waste and about 50 broken, discarded objects, along with a boat floor plank and a bundle of spruce roots.
Anyone looking for a total find count meets conflicting figures. The World Heritage text speaks of about 800 artefacts, while the project gives over 2,000 for 1973 to 1976 alone. What exactly the 800 refers to is not clear from the text available to me, so I give no single total here.
The buildings are sod over a timber frame, in a manner Wallace assigns to the Icelandic style. Long, narrow fireplaces ran down the middle, giving heat and light and serving for cooking. Our article on the Viking longhouse explains how such houses were built in general. What sets L'Anse aux Meadows apart is the division of tasks. According to Wallace the houses form four groups, and the excavation letters read almost like a floor plan:
Seven buildings therefore run north to south, and the eighth stands alone on the far side of the brook. Parks Canada counts more briefly: three longhouses and five smaller buildings. The pages I checked gave no exact dimensions. Wallace offers a comparison instead: even the smallest hall is twice the size of the house attributed to Erik the Red. Her observation that the largest hall is the most elaborate and probably housed the leader points the same way. Her 2009–10 paper divides the eight differently: three large halls, a small house, a small rounded hut and three pit buildings, meaning small sunken structures.
What is missing tells as much: Wallace points to the lack of animal buildings. The project's own scene of cows, sheep and goats at pasture should therefore be read as a story, not as an excavation report.
From the sleeping places Wallace estimates 70 to 90 occupants. She adds a calculation of the labour: about 60 men would have needed two months to build the complex, 90 men about six weeks. For the early 11th-century Greenland colony she gives only about 400 people. Both figures are estimates, but the ratio makes the point. This was a large, planned undertaking and no chance camp.
The best-known single object is the bronze ringed pin, a dress pin with a loose ring at its head that held a cloak closed. Wallace dates the style to the 10th and early 11th centuries and calls the pin the find most useful for dating. Parks Canada says it helped link the site conclusively to the Norse. That is a lot of weight for one piece of bronze, which is why it matters that it does not stand alone.
The spindle whorl is a small ring-shaped disc of soapstone. Fitted on a hand spindle, it worked as a flywheel, keeping up the twist that turned wool into yarn. With it go a fragment of bone needle, which a project caption places in the longfire of Hall D, and a small whetstone, according to Parks Canada for needles or small knives. Parks Canada describes the needle in one place as possibly for knitting and in another as for sewing, so its use is not settled.
Other pieces include a gilded brass fragment that Wallace identifies as part of a ring, a glass bead and, again according to Wallace, eleven fire strikers of red jasper. Kevin Smith's analysis, reproduced by the project, speaks of ten. According to Wallace, those from two halls are Icelandic, while four of the five from the largest hall are Greenlandic, which may point to a leader from there. Wallace traces two to Notre Dame Bay in Newfoundland, Smith one fragment to north-central Newfoundland, which shows trips around the island.
Do these objects prove women at L'Anse aux Meadows? It is an inference in two steps. Textile work was generally women's work in Norse society, so the tools point to women. Wallace and Parks Canada word it accordingly, with "suggest", and assume a few women among a mostly male group. No find says who they were. The sagas name Freydís and Gudrid, but no link to this spinning weight can be shown.
Hut J is among the most unusual parts of the site. It held a small stone-and-clay bloomery furnace, which won iron directly from ore, with a pit for charcoal beside it. The ore came from the nearby bogs, where bog iron forms in small deposits. An exhibit text from Parks Canada reports that tests linked the iron of a nail from Wallace's dig to slag on site, and the agency calls this the earliest known iron smelting in North America.
Why go to such trouble? The most plausible answer lies in the Hall F shed and the bog wood: ships needed upkeep, and nails rust. A boat plank, boat nails and treenails fit that picture. Our article on shipbuilding shows how such hull parts work together, and the one on blacksmithing covers the anvil work. The iron also serves as proof. The authors of the 2021 Nature study write that the Indigenous inhabitants of the time did not use such metal tools.
The Norse were a brief episode at an old camping place. Parks Canada says Indigenous people used it as early as about 6,000 years ago, while another of its pages speaks of four to five thousand years. Wallace counts five groups before the Norse and one after; the project describes short, seasonal hunting camps. A Groswater harpoon shaft is dated by Parks Canada to 1050–850 BCE, and Dorset people camped on the south shore of the bay until more than two centuries before the Norse. The youngest Indigenous traces reach into the 13th and 14th centuries.
Did they meet here? The Parks Canada and Wallace texts say no: nobody else lived on the site during the Norse period, although visible hearths and tent rings would have shown that people had been there. In 2019 Paul Ledger and colleagues suggested instead that the place could have been a shared zone of interaction. That is a proposal drawn from a model, not the find of a meeting. According to Wallace the sagas' Skrælings have no clear link to this site.
New users came later: a French fishing station on the shore, seasonal until 1904, and from about 1835 the village of the first permanent settlers.
Here the figures diverge most. In The Canadian Encyclopedia Wallace reckons with about a decade. Her reasons are archaeological: the extent of the cultural deposits and middens and the lack of rebuilding. Colleen Batey, reviewing Wallace's book, renders this as little more than a decade and reports that on Wallace's account the site was left in an orderly way and possibly burned by its occupants. A Parks Canada exhibit text gives 10 to 20 years, while another agency page says perhaps only a few years and admits the uncertainty.
The sagas help little. As the Nature paper reads them, one suggests about seven winters and roughly thirteen years between first arrival and final return, the other a minimum of about three years.
In 2019 Ledger, Girdland-Flink and Forbes challenged the narrow span in the Proceedings of the National Academy of Sciences. Their Bayesian model, a statistical way of combining many radiocarbon dates, puts the Norse start between 910 and 1030 and the end between 1030 and 1145. For duration the model gives 0 to 195 years (2 sigma), so a short use is not excluded; the abstract speaks of activity possibly lasting a century. The authors add that this does not imply continuous occupation and think of sporadic visits. They do not assign the newly found bog layer of the 12th to 13th centuries to the Norse, associating it instead with Indigenous use, though its cultural affinity remains uncertain. Headlines claiming the Norse stayed for "centuries" do not reflect this.
The Nature study by Margot Kuitems and colleagues names the root problem: the 55 radiocarbon dates tied to the Norse occupation scatter across the whole Viking Age, because methods of the 1960s and 1970s were imprecise and many samples have an unknown inbuilt age. Their answer was to date three worked pieces of wood to a felling year of 1021. The authors themselves stress that the year is a fixed reference point and that its relation to the rest of the voyages is a matter for future research. No span of use follows from it. On the little we have, "short" is the best-founded description; an exact figure cannot be established.
Helge Ingstad had a clear answer: this was Vinland. The problem was obvious. The sagas describe a fertile land with grapes, and no grapes grow at this latitude. So Ingstad turned to an older explanation of the word, in which vin means meadow or pasture. Our main article on Vinland deals with that argument, so it need not be repeated here.
The strongest counter-evidence came from the dig itself. The Norse bog layer held three butternuts and butternut wood, including a burl with cut marks from a metal knife. The tree's northern limit is in eastern New Brunswick, about 1,000 kilometres south according to Wallace. Wild grapes grow there too, and both ripen in September. The pieces could not be radiocarbon dated because the wood had been treated; they belong to the Norse use only through their position in the layers. That is good evidence but not watertight. Wallace concludes that the Norse explored the Gulf of St. Lawrence.
On this Wallace builds her model. For her, L'Anse aux Meadows is no farm but a winter base, "a gateway to resources" in the south. Vinland would then be a whole region. Eiríks saga rauða, covered at length in our article on Thorfinn Karlsefni, has two main places: northern Straumfjord with the island of Straumey, and southern Hóp, a coastal spot with wild grain and vines. Wallace maps this pair onto the evidence. She writes that only one saga place corresponds to L'Anse aux Meadows in all respects, namely Straumfjord. She looks for Hóp on the lagoon and sandbar coasts of eastern New Brunswick, especially around the Miramichi and Restigouche. According to Batey she also equates Straumfjord with the Leifsbúðir of the Grœnlendinga saga, and her paper of 2009–10 goes as far as the title "Leif Eriksson's Home in Vinland".
"Karlsefni and his people sailed to the mouth of the river, and called the land Hop. There they found fields of wild wheat wherever there were low grounds; and the vine in all places were [sic] there was rough rising ground."Eirik the Red's Saga, ch. 9 in the numbering of J. Sephton's translation (1880)
Sentences like these also show the limits of the method. The sagas were written down centuries later, and how far their topography can be read as a map is itself disputed. Our article on the Icelandic sagas explains how they came about.
A counter-proposal appeared in 2012. Jónas Kristjánsson and four co-authors argue in Acta Archaeologica that Vinland is Newfoundland and that Straumfjord could be Sop's Arm in White Bay. Wikipedia summarises their reason: only there does a fjord have an island at its mouth, and they read L'Anse aux Meadows as Kjalarnes. The pitfall traps they investigated, however, could not be dated conclusively. Lists of proposals in any case run from the Hudson to the St. Lawrence. In Batey's assessment the gateway concept is now widely accepted. The match to any one saga place remains unproven: no find bears a name, and I know of no archaeologically confirmed Hóp. Official texts smooth this over too. Canada's World Heritage page has Leif wintering at a "Straumfiord", blending the two sagas.
In 1978 the site joined UNESCO's World Heritage List under criterion (vi). The justification calls it "the first and only known site established by Vikings in North America". The protected area covers 7,991 hectares, far more than the excavated ground. Parks Canada's federal directory dates its designation as a National Historic Site to 28 November 1968, as does Wallace. The UNESCO page gives 1975 for its establishment as such, without explaining the difference.
Almost nothing original is on view. The remains were reburied for protection. Beside them stand reconstructed houses, built according to Batey in 1979 and 1980; a rebuilt furnace hut followed in 2011, according to Wallace. Reconstruction and evidence must be kept strictly apart: the furnishings inside are interpretation, even when based on excavation plans.
Who discovered L'Anse aux Meadows? George Decker led Helge Ingstad to the ridges in 1960, and Anne Stine Ingstad directed the dig from 1961. The idea of Norse traces in this region is older.
How many buildings were there? Eight of turf over timber: three large halls and five smaller buildings, among them the furnace hut. Wallace classes three of them as pit buildings.
Which finds prove the Norse origin? Above all slag and iron nails and the bronze ringed pin, along with the building type. Spindle whorl and bone needle round out the picture.
How long was the site used? That is disputed. Wallace reckons with about ten years, Parks Canada in places gives 10 to 20, and Ledger and colleagues consider sporadic use of up to about a century possible.
Can you see the original houses? No. They were reburied. Reconstructed houses stand beside the excavated area.
Vinland: texts, finds and the wine dispute · Thorfinn Karlsefni and Gudrid · Leif Eriksson · Erik the Red · The Greenlanders · Freydís · The Viking longhouse · The Kensington Runestone
Vinland · Thorfinn Karlsefni · Leif Erikson.
All History & Finds articles →
In early October 1263 a few hundred Norwegians fought a Scottish force on a shingle beach in the Firth of Clyde while a …
Read more →Some weapons give their name to an entire age. The long iron sword of the Celts is one of them: after the find-spot of …
Read more →Look up "Lebus" in a history book and you will soon meet Thietmar of Merseburg, said to describe the siege of a …
Read more →Leif Erikson (Old Norse Leifr Eiríksson) was a Norse seafarer from the Greenland colony who, around the year 1000 …
Read more →