Secrets of Ancient Iceland, Dispatch 1: Digging into the Viking Age

house atop grassy field
Credit Dean Goodman

The medieval turf farmhouse at Glaumbaer in northern Iceland, with the church peeking out from behind. In the hayfield below the medieval house, John Steinberg's crew is investigating a Viking Age longhouse that he discovered in 2002.

Summer 2005

Settled by Viking chieftains around 900, Iceland remained free from outside rule for 300 years. Tradition holds that this independent society was egalitarian, an "almost democracy" that flourished until Iceland was swallowed by the kingdom of Norway in 1262. Penn State anthropologist Paul Durrenberger, however, has a very different view of Iceland's Golden Age.

This summer, a team headed by Durrenberger and UCLA archaeologist John Steinberg are putting Durrenberger's alternate theory to the test. Using a combination of cutting-edge remote sensing tools and traditional trowel-and-shovel archaeology, the researchers are mapping all the Viking Age farms in one valley of northern Iceland.

Writer Nancy Marie Brown is along for the ride. As a volunteer and observer at the dig site, she provides a ground-level (and underground) view of archaeologists at work and the thousand-year-old secrets they are
laboring to reveal. Following are her dispatches from the field in Glaumbaer.

Dispatch 1: Digging into the Viking Age

"Science is at the junction between the expected and the real," Penn State anthropologist Paul Durrenberger told me. "The whole point of it is figuring out what's real and what's not."

I am standing in a hayfield in Iceland in a gale-force wind, one of six volunteers holding down a 100-meter measuring tape with our toes. Bundled in raincoats and stocking caps, we are feeling fortunate that at least the sun is shining. Tourists visiting the Glaumbaer Folk Museum, a collection of historic buildings up on the hill, stop and stare down at
us.

Viking Age turfhouse
Credit Dean Goodman

A reconstruction of a Viking Age turfhouse, showing the building techniques used in Iceland a thousand years ago.

"A lot of people asked me," said Sigridur Sigurdardottir, the museum curator, "'What are those crazy people doing out there in your hayfield?'"

What we think we are doing is taking pictures of a Viking longhouse buried beneath our feet. But at the moment, the only thing that's real is the wind.

Down the line comes the leader of our archaeological crew, John Steinberg of UCLA's Cotsen Institute, and I take two steps back to let him pass. We've been doing this little dance now for five hours. Steinberg is pulling on a length of two-by-four duct-taped to a heavy orange box. Earlier this week he dissected a green plastic watering can and taped
curved pieces of it to the front of the box; now, with its more rounded prow, the box is sliding more freely but Steinberg is still working up a sweat.

Wired to the box is a data recorder carried by Brian Damiata, a UCLA geophysicist. Matching Steinberg step for step, Damiata trips a switch to log every meter. He is careful not to hit the cable with his knee: The bump gets recorded as data. Already a whole day's work had to be thrown out for one such glitch.

man using equipment in field
Credit Gisli Palsson

Sometimes the GPR remote sensing device works better without its cart, but then it takes two people to run it. Here, Dean Goodman records the data at a test site in western Iceland.

Damiata and Steinberg are testing a remote sensing device called Ground Penetrating Radar (GPR). The orange box sends microwaves into the soil and picks up the echoes. These echoes can be translated into animations of what the hayfield might look like if you could slice it horizontally, ten centimeters at a time. The GPR-Slice software was developed by California-based geophysicist Dean Goodman, who has come with us to see if it can tell the difference between plain
dirt and the turf walls of a Viking house.

I'm here dancing in the wind at the invitation of Paul Durrenberger, whose theory about early Iceland this archaeological survey was designed to test. Tradition—based on a romantic reading of the medieval Icelandic sagas—says that Iceland was an "almost democracy" from its settlement by Vikings in 870 until 1262, when it became part of the kingdom of Norway. Durrenberger believes the society was more hierarchical, that landholders controlled all the resources and the rest of the population from the beginning. He details his view in the book The Dynamics of Medieval Iceland (University of Iowa
Press, 1992). If Steinberg and his crew can map all the Viking Age and later medieval houses in the valley, Durrenberger will be able to see from the change in settlement pattern when the political economy of Iceland changed from these chiefdoms to a manorial system.

GPR radargram
Credit Dean Goodman

A sample GPR radargram from Glaumbaer, after being enhanced by Dean Goodman's special software, GPR-Slice. The dark colors show where the microwaves reflected off something other than dirt. By taking thousands of scans at different settings and angles, the scientists can see the walls, hearth, and stone entranceway of the longhouse that is buried a meter underground.

So far it looks like Durrenberger is right, and the change happened well before 1262. At Glaumbaer, the Viking longhouse was abandoned just after 1104 and a new manor house was built up on the museum's hill, where an 18th-century turf farmhouse now sits. Durrenberger joined the project a week late—after we should have (but didn't) have the bugs worked out of the remote sensing device. A cultural anthropologist whose experience in archaeology is of the traditional trowel-and-shovel
type, he immediately joined a group exploring a farmstead further down the road which might follow the same pattern as at Glaumbaer.

Between the farm buildings and the hayfields is a steeply pitched mound where tradition says turf houses have stood since the Viking Age. On top of the last turf house, abandoned in the 1940s, the current farmers have dumped their refuse, including several hundred sheep that were destroyed in 1991 because of disease. When Durrenberger arrived,
the team had already taken test cores, punching a thin steel tube down into the soil and pulling out a cylindrical sample. By looking for layers of charcoal, bone, or peat ash in the soil samples, they can see when people built the first house on this spot.

"In Iceland you have the date built into the soil," explains Durrenberger, "because you also have layers of volcanic tephra."

Tephra has a different color and texture from soil or sand. When your un the edge of your trowel over it, it rings out, as if you had tapped a glass bottle. It feels grittier, almost spiky on your fingertips or, if you're not certain, on your tongue. Even I could spot the most obvious lines: a chalky white layer from an eruption in 1104 and a honey-brown one from either of two eruptions 2000 and 4000 years ago. If you see no charcoal, bone, or ash above a honey-brown line, you know no people have lived here—when Viking explorers discovered Iceland, they found no indigenous people.

If, on the other hand, you find charcoal, bone, or ash near the white 1104 layer, the spot is worth investigating further.

This mound showed signs of occupation down to at least 1104. When Durrenberger arrived, he began digging a one-meter-square test pit with a well-sharpened shovel.

After a long windy day of digging or dancing with the remote sensing device, we met around the coffee table in the local schoolhouse, which we had turned into an archaeological station. The midnight sun was tinting the mountains pink and, although tired, we weren't yet ready to call it a day.

man peeking out of hole he’s dug in the ground
Photo from 2002 Glaumbaer dig, Courtesy Paul Durrenberger

The hole Paul Durrenberger dug into a farm mound here in 2002 is still the talk of the excavation crew. He started digging another as soon as he arrived this summer.

"What did you find in that hole?" I asked Durrenberger. "What was the most exciting thing?"

"Well, we found fishbone," he answered. "We found lots of shells. Some teeth. A nail. A lot of iron slag. Sheep bones. A piece of coal that had to have been imported, because Iceland has no coal. We found a clothing fastener of some sort made of wire. Looks like a ram's head. And a bone bead. That was interesting. It had two rings carved on it." He was still in layers from the 19th century. He had a long way to go to reach the Viking Age, and "exciting" was not the word to describe his work. "It's the top of a trash heap, that's what it is," he said.

"Archaeology, like any science, seeks to find new things," Durrenberger continued. "If we knew there was a house here before 1100, we wouldn't have to dig.

"By and by we'll have a story to tell about this valley," he added, "and it will be orderly and neat. What people don't usually see is all the confusion and chaos that goes into getting that story."

Dispatch 2: Connect the dots

"If I knew what was in there, I wouldn't have to dig a hole."

My first day with a trowel and shovel, digging toward the Viking Age in a hayfield in northern Iceland, this offhand remark by John Steinberg became my motto.

Steinberg, from UCLA's Cotsen Institute, is hoping to make digging outdated for survey archaeology—and it looks like he might get his wish. He is working with Penn State anthropologist Paul Durrenberger to see the pattern in which this valley was settled; the settlement pattern will help them determine when and why Iceland was transformed from a collection of Viking chiefdoms into part of the kingdom of Norway sometime between 870 and 1262. But to find the pattern efficiently, they need a way to locate and measure all the Viking houses without digging lots of holes.

"There are two reasons for not digging holes," explains Durrenberger. "When you dig a hole, you destroy the site. And when you dig a hole it costs lots of money."

two people surveying a piece of land
Credit John Steinberg

Volunteer Suzan Erem holds the stadia rod, the target that the Total Station uses to measure the coordinates of a point, while Northwestern graduate student Doug Bolender records the information on a survey form.

For two weeks, Steinberg and UCLA geophysicist Brian Damiata have been taking pictures down through the grass and soil of this hayfield using a remote sensing device called Ground Penetrating Radar (GPR). GPR acts like a microwave camera, sending electromagnetic waves into the earth and picking up the echoes or reflections. They've processed the GPR data into a series of colorful animations showing what the field would look like if a giant took a trowel and peeled off slices horizontally, centimeter by centimeter.

Steinberg concedes that you do need some imagination to connect the dots left by the microwave reflections into the outline of a house. To learn how to read the GPR images better, the next step is to compare them to what traditional trowel-and-shovel archaeology can say about the site. So one night the sod was lifted off a 750-square-meter section of the hayfield, down to the bottom of the plow zone, and the next morning seven of us were issued shovels.

The Vikings in Iceland didn't build houses of stone or wood: The walls we're looking for are made of blocks of turf—essentially the top of a peat bog—often layed up in a herringbone pattern. For two days we scraped the site, using our shovels like spatulas to lift off an inch or two of dirt at a time, looking for the turf signature. Turf is denser and has a higher organic content than dirt. It is swirled with soft earthen colors: greenish-gray and brown, orange and gold, black and rust-red. Both the texture and the color should help us see the walls. But first we had to clean off the layer of white tephra deposited during an eruption of Mount Hekla in 1104. It was unmistakeable: When I hit it, my shovel would glide like it was greased and the upper strata of soil would just flake off.

When the delicate trowel work began, I was assigned a spot covered by ash from a peat fire. Based on the parts we could see, the west wall of the house should continue underneath the ash. I was to define its edges, looking for where the turf blocks broke off. It wasn't easy to see.

rectangular hole dug in ground, with measuring instrument placed in center
Credit Doug Bolender

The edge of the excavated part of the peat ash pit, showing the colorful strata that formed as ash was dumped into the hole over a period of years.

Peat ash is almost as slippery as volcanic tephra, and it comes in very distinct colors, from pale pink to cranberry, with bright yellow, orange, black, and white mixed in, all in all a good match for the palette the Norwegian painter Edvard Munch used in his famous work "The Scream." Working at it for eight days (the ash was much deeper and wider spread than expected), I became quite fond of the color scheme. All week it was sunny and breezy, the smell of cow manure mixing with that of fresh-cut hay, and the days blended together as I scraped and dug. On my coffee breaks, I would look out across the valley, surprised to see the mountains bright green in the sun, or shaded gray with cloud, or sometimes just as pink as the ash I was so focused on.

Somebody, it seemed, had been dumping peat ash into a pit dug against the outside wall for years after the house was no longer being lived in. The ash had blown over the wall and filled in the crevices between the tumbled blocks of wall and roof inside the house: One section, once cleaned off, looked like a relief map, showing each individual turf block. The house had already completely collapsed before the white tephra from the 1104 eruption covered it.

Another day, as I was concentrating on distinguishing pale pink ash from greenish-gray turf, scraping carefully with my trowel, Doug Bolender, a graduate student at Northwestern University, walked over, looked down, and said, "Nancy found the edge of a wall." When I leaned back and readjusted my gaze, it just popped out: A section of wall about a foot long was left rising above the pit out of which I had cleaned the ash. Right under my hand was a completely straight edge.

Parts of the north and south walls of the house turned out to not be so easy to define, and something very funny was going on with the east wall. To clear up these ambiguous spots, Steinberg and Damiata brought a computer out to the site and tried to align what we'd found with the GPR images. Damiata had been working on the GPR data for a week, adjusting the color scheme and tinkering with how the computer transformed the data into images to highlight different types of reflections. One solution made the walls, which are weak reflectors at their bases, much more clear in the pictures. Another highlighted the floors, which are strong reflectors, Damiata says, probably because they're very compacted, greasy, and gritty with ground-in charcoal. Other strong reflectors are stones and the jumbled bones and trash in the rubbish heap, or midden.

On the eastern wall, what seemed to the trowel-and-shovel crew to be an entrance lined up rather well with a set of strong, deep reflections that Steinberg interpreted as cobblestones at floor level. Just down the hill, we had uncovered a pile of bones marking the top of the midden. Said Steinberg, pointing from the dirt at his feet to the computer image and back, "If that turns out to be the entrance, and this turns out to be pavement, and there's the midden, well, that's pretty perfect."

person in ballcap excavating hole in the ground
Credit John Schoenfelder

Rita Shepard working away at the bottom of a meter deep test pit.

If it turns out to be perfect, says Durrenberger, "then that means the GPR has accurately told us what's in the dirt. And that's the point of the exercise. Then we can explore other sites without digging holes."

And digging holes is what Durrenberger has spent the last two weeks doing, while I was scraping up pink peat ash. At the first hole, he says, "we hit the white 1104 tephra layer and, as Doug predicted, there was evidence of people living there before that. So then I got assigned a new site to dig another hole in. I'm down 1.8 meters, digging through turf, and I haven't found the 1104 yet."

Haymaking is just about over here in northern Iceland, and as the farmers haul their round bales out of the fields, the 13 crew members of the Skagafjordur Archaeological Settlement Survey have been moving in. My part of the crew took the turf off a 750-square-meter section of hayfield just below the Glaumbaer Folk Museum in mid-July, and began scraping away the dirt to reveal the collapsed walls of a 40-meter-long Viking Age house abandoned before 1104.

Dispatch 3: Seeing the context

excavated patch of farmland
Credit Rita Shepard

The south end of the Viking Age longhouse at Glaumbaer turned out to be the most complex. After scraping off a few more inches of soil, graduate student Tara Carter (right) confirmed that the walls continued on out of the excavated area into the farmer's cabbage patch.

At the same time a small contingent including Penn State anthropologist Paul Durrenberger, one of the principal investigators of this National Science Foundation-funded project, began digging test pits in nearby farm mounds. Day after day, the Glaumbaer group moved dirt, looking first for the chalky white tephra left by the eruption of Mount Hekla in 1104, and then for any sign of peat ash or bone or the mottled earthy colors of a turf wall under the tephra. I worked mostly on my knees in the shallow, wide holes, using a dustpan and trowel. (Or, as my Icelandic friends teased, a teaspoon.)

Other, more practiced diggers dug one meter-square test pits with shovels and screened the dirt looking for artifacts and bone. By studying the sheep bones, for instance, and finding out the ratio of lambs to ewes to wethers, scientists can tell if the household whose trash is being uncovered was a central one, managing its sheep for wool to export, or a dependent one whose herds were used for food.

The weather turned cold and rainy. The mountains disappeared in the fog. The exposed earth of Glaumbaer turned to muck. Still we moved dirt. I began to think the holes were endless, that the full buckets—like those in Walt Disney's "Fantasia"—were a wizard's spell out of whack.

"Same hole, different day," Durrenberger would joke when we met at dinner. Or sometimes, "Same day, different hole."

It always seemed to be he who dug the deepest holes—one went almost 3 meters down, slender as a well, and had to be shored up with two-by-fours. I began to wonder: Was Durrenberger involved with the project just because he's a pro with a shovel?

"Paul's here," explained John Steinberg of UCLA's Cotsen Institute, the leader of our archaeological crew, "because sometimes we get so caught up in the turf and the tephra, in the logic of the archaeology, that we forget what questions we came here to answer."

In long dinner-table discussions, and well into the night in the team's computer lab (a room in a local school), Durrenberger, a cultural anthropologist and ethnographer, tries to put each day's hole into context.

man standing in test pit
Credit John Steinberg

Paul Durrenberger happily dug the deepest and nastiest test pits.

Ethnographers live with, observe, and talk to people to learn why they do what they do. When there are no people left of a culture, such as that of the Vikings in Iceland, there are still two ways to gain ethnographic information. First, you can study similar living cultures, an approach Durrenberger has already taken as far as it will go in work conducted years ago among the Lisu and Shan peoples in Northern Thailand. "The important thing about the society of medieval Iceland," he explains, "is that it was both highly stratified and had no state institutions to make that stratification work. That's why in the long run it failed. From the Lisu we know about societies that have no state institutions but are egalitarian, and from the Shan we know about societies that are highly stratified, but do have state institutions. What we don't have in today's world is a society that is both stratified and not a state, the way medieval Iceland was."

The second method is to examine what the culture left behind, such as the longhouse we are excavating at Glaumbaer and the signs of occupation seen in the lines of fireplace ash and other rubbish found in the holes on the other farms.

"When I'm out here using a shovel," Durrenberger said, "I'm seeing the same human systems I am when I'm out talking to people."

For medieval Iceland, Durrenberger has yet another source of ethnographic information: sagas written down in the 12th and 13th centuries that tell of the settlement of Iceland 200 or more years before. "Historians might see in the sagas a series of particular events," said Doug Bolender, a Northwestern University graduate student working on the project. "An ethnographer like Paul reads the sagas to learn about the structure of the institutions in the society."

turf house
Credit Sigridur Sigurdardottir, Byggdasafn Skagfirdinga

The walls of the 18th-century turf house at the Glaumbaer Folk Museum.

In Durrenberger's reading of the sagas, substantial inequality existed early on in Iceland, as opposed to the more egalitarian society that other scholars have described. Steinberg and Bolender asked Durrenberger to collaborate with them in Iceland, since economic inequality can be seen in how the pattern of settlement in an area changes over the years. According to this pattern, the initial settlers in this part of Skagafjord established large farms along the river. After a hundred years or so, Durrenberger says, "they created small farms on the margins of the large ones, renting out parcels to tenants who provided labor at crucial times like haymaking and the fall sheep round-up. This arrangement forced the dependents to provide for their own subsistence during the rest of the year, so that the large land-owners did not have to support them."

So far, the results of our digging confirm Durrenberger's hypothesis: the large farms are the oldest; then smaller farms crop up, clustered around the large ones, between the years 1000 and 1104.
But the Skagafjordur Archaeological Settlement Survey isn't yet over. While this year's field season is at an end, Steinberg and Durrenberger plan to return. And the ambiguous results of some of our holes call for more digging.

For instance, high over the river, halfway to the mountains, just below a trout-filled lake, Steinberg asked Durrenberger to dig yet another hole in a farm mound. According to historical tax records uncovered by Bolender, this site was once a dependent farm called Medalheimur. "The story goes," says Steinberg, "that you can't eat the trout in the lake. Anyone who eats the trout dies. When you get a story like that, and a farm that drops out of the historical record, there's something going on. It's worth checking out."

The hole at Medalheimur isn't deep—just over half a meter down—but its walls are a puzzle of colorful striations. There are lines of pink peat ash, familiar from my eight days with a trowel digging out an ash pit beside the Viking Age longhouse at Glaumbaer (see Dispatch £2). Tephra from the eruption of Mount Hekla in 1104 shows up as a chalk-white line. Protruding from one corner at the bottom are a cluster of charcoal-covered stones that Steinberg and Durrenberger identified as a possible hearth. The hearthstones rest right on a dark line of tephra, which volcanologists say was deposited just after Iceland began to be settled in 870. "This is our time machine," Durrenberger says. "You're looking down at the earliest days of the settlement of Iceland."

Unfortunately, there's no sign in the hole of the bluish-gray tephra layer from the year 1000—so this hearth fire may have burned as early as 870 or as late as 1104. To get a firmer date, Steinberg will send a sheep bone, found just below the hearth, to a radiocarbon-dating lab.

"Until we get the date from the bone, we don't know the age of the house," says Steinberg. "We know it's earlier than 1104. If it's between 1000 and 1104, then it's a hjaleiga—a tenant farm—of Glaumbaer. If it's earlier than 1000, then we've got trouble."

The other farm mound holes seem to support the pattern Durrenberger predicted. Medalheimur might, or might not, fit. "It's the same in any science," Durrenberger says. "You get data to answer questions, but those data raise more questions. That's what keeps it exciting."

Nancy Marie Brown is a former editor of Research/Penn State. Her current book, Voyage of a Viking Woman, is about a woman who explored Greenland and Atlantic Canada and lived at Glaumbaer in Iceland a thousand years ago.

Last Updated July 18, 2005