Figure 1.

The North Atlantic.

Based on the saga literature, the Icelandic Commonwealth (AD 930-1264) was a class-stratified society without a formal governmental institution (Durrenberger 1992). Because this associated archaeological record is difficult to identify on the surface (Smith and Parsons 1989), survey is challenging and settlement pattern studies have been of little help in understanding the Icelandic Viking Age Commonwealth or its documented transition from chiefdom to state. To help solve this problem, over the last two years the Skagafjördur Archaeological Settlement Survey (SASS) has developed a series of techniques centered on geophysical prospection that make sub-surface survey more reliable in fertile areas with heavy soil deposition. In the process of reconstructing changes in the settlement pattern, we have identified a long-house at the historic farm of Glaumbær (Figure 1).

The last phase of the structure is best understood, but less than 10% has been excavated. From the test trenches, the structure would appear to be 29 m long from outside to outside, with a 15 cm thick floor of tramped clay and peat ash approximately 27 m long and 1.7 m wide. On either side of the narrow floor are raised 1.8 m wide benches. This is a typical layout for a Viking Age long-house (e.g., Magnússon 1973, Smith 1995, contributions to Fitzhugh and Ward 2000). The outer turf walls are 1.8 m thick and the lower 50-60 cm are extraordinarily well preserved. It would appear that the occupants took advantage of the nearby bog iron as substantial slag deposits have been recovered from the pre-1104 midden.

Figure 2.

Conductivity map of Glaumbaer with main structure outlined. Scale is from violet (4ms/m) to red (11ms/m).

The Glaumbær Folk Museum, one of the oldest standing turf houses in Iceland, is 150 m east of the discovered long-house and about 15 m higher in elevation (Figure 5). Excavations into the associated ash midden, just to the south of the museum structure, indicate that the turf house was occupied immediately after the Hekla 1104 eruption, implying that when the lower, recently discovered structure was abandoned, the occupants moved uphill to the current location. No distinctive artifacts have yet been recovered to demonstrate any connection with the New World. Further excavations are planned.

Glaumbær is mentioned in the Vinland Sagas as the farm settled by Gudrid Thorbjarnadottir and Thorfinn Karlsofni, along with their son, Snorri, after their failed attempt to settle Vinland (Kunz 2000). If the stories are true, it would make Snorri the first European born in the New World and this newly discovered structure would be the dwelling established following their Vinland voyages, just after 1000.

Figure 3.

Resistivity pseudo-section through buried turf wall along UTME 50.

Figure 4.

Resistivity fence diagram.

Figure 5.

Main East-West excavation trench across short axis of long-house.

The long-house was discovered on a relatively flat hay field using the SASS program of coring, conductivity surveying (Figure 2), auguring, resistivity profiling (Figures 4 and 5), and test excavation (Figure 5). We have identified three phases: a component built before 1000 (dated by the Veidivotn-Dyngjuháls volcanic ash layer); a middle phase (after 1000) with a well preserved clay floor and eastern wall; and a final phase that incorporated aspects of the middle phase. The final phase seems to be abandoned by the time the white volcanic ash layer of the 1104 AD Hekla eruption falls because the 1 cm thick tephra overlies the entire structure and landed on already fallen turf blocks (Figure 5). An AMS date of 1,017 ±56 BP (AA46689) from a juvenile bovine mandible recovered on the floor, calibrates to possible calendar age ranges of 983-1034 cal AD at 68% probability with a mid point at AD 1018. This date correlates well with the tephrachronology.