Archaeologia Islandica - 01.01.2004, Blaðsíða 89

ICELANDIC FARMHOUSE EXCAVATIONS: FlELD METHODS AND SITE CHOICES
recorded in the field, whether as text or
illustration was made available in pub-
lished form shortly after the excavation
was completed (Bruun & Finnur Jónsson
1909, 1911; Bruun 1928).
This method was to remain the basic
approach in archaeological fieldwork in
Iceland until the 1970s. The principle is
that the structure under investigation was
considered to be temporally finite, with
no history of use, rebuilding or develop-
ment - or at least not with any such his-
tory considered worth studying - and
with a clear-cut break between the use of
the house and the deposits representing
its abandonment and disuse. The archae-
ologist's task was then simply to peel off
those post-abandonment deposits and to
plan what was then revealed. The plan
was the aim of the exercise and also the
principal component of the excavation
archive and subsequent publication.
Nothing reveals better the limited aims of
this approach than the fact that floor lay-
ers were as a rule not excavated and as a
result relatively few artefacts tended to
be retrieved. The non-removal of floor
layers meant that excavators rarely
encountered evidence for earlier building
phases or repairs. It also meant that neg-
ative features, like post-holes and under
floor drains regularly went unnoticed.
When building parts or features were
removed it was always because such
structures were considered to be much
later than the building under excavation
and hence of no interest. As a result they
were normally removed without plans
being drawn or other records made (e.g.
Snjáleifartóttir - Stenberger 1943c and
Reyðarfell - Grímsson 1976).
Matthías Þórðarson was the first
archaeologist to test the limits of this
approach in his excavation of
Bergþórshvoll in 1926-1928. Here
Þórðarson had to dig through multiple
phases of a farm mound in order to get at
the early llth century remains he was
looking for. To his credit he recorded all
the floors he encountered and removed,
drew a plan of each, recorded its depth
and related the artefacts found to each
floor. That said, he chose to limit his
analysis of the stratigraphy to recording
the level, size and shape of the floor
deposits - which were as a rule distinct
and easy to define - and large features
such as pavements and vat-holes, but
ignored everything else, both all other
sorts of deposits (walls, roof collapse,
middens) and features like post-holes and
post pads which presumably were there
to record. The fact that Þórðarson never
published his results (they were sum-
marised by Eldjám and Gestsson 1951)
meant that others were not able to leam
from his difficulties and deep stratigra-
phies were to remain outside the experi-
ence of Icelandic archaeologists until the
1970s.
In 1939 a group of experienced
Scandinavian archaeologists descended
on Iceland to excavate farmhouses in
Þjórsárdalur and Borgarfjörður. In the
group were veteran fieldworkers like
Aage Roussell and Márten Stenberger
who were to have a lasting impact on
Icelandic archaeology as well as academ-
ic discourse on Scandinavian building
custom. Stenberger was no doubt the
most accomplished fieldworker in the
group. This can be seen for instance from
87