Archaeologia Islandica - 01.01.2004, Page 90

Orri Vésteinsson
his meticulous recording of postholes
and stakeholes, features which have until
recently not received much attention by
Icelandic archaeologists. More impor-
tantly he was the only excavator in the
group to show an understanding of
stratigraphy and consider the possibility
that a building could have more than one
phase. At ísleifsstaðir he defined three
phases of a Viking age long house and at
Snjáleifartóttir he excavated two very
different structures, one superimposed on
the other. In terms of impact it was how-
ever the work of Aage Roussell that has
been most important. He wrote the chap-
ter on the project's results (Roussell
1943c), and excavated Skallakot and
Stöng, the latter achieving almost instant
fame as an extraordinarily well preserved
example of a high-medieval and later,
after a revision of the dating, a late
Viking age farm house. The fact that
Roussell was an architect by training had
a clear impact on both his approach to the
excavations as well as to his interpreta-
tion of the excavated remains. Stöng and
Skallakot, as Roussell presents these
structures, are both logical structural
entities, with a clear design and compre-
hensible structural characteristics. It is
hardly a coincidence that the architect
Roussell excavated such imminently sen-
sible buildings while archaeologists like
Matthías Þórðarson and Juoko Voinmaa -
with less experience in building archae-
ology and probably a more limited inter-
est - excavated buildings that are both
incomplete and largely incomprehensible
(i.e. at Skeljastaðir and Stórhólshlíð).
At Stöng more recent excavations
have shown that undemeath the structure
excavated by Roussell there is another
long-house (Vilhjálmsson 1989) and at
Skallakot re-excavation in 2001
(Gestsdóttir 2002) suggests that Roussell
not only failed to fully excavate the most
recent phase of the building, he also iden-
tified rows of stones from different phas-
es as belonging to the same phase - giv-
ing the impression that the building is
complete where it is not - and added fea-
tures to his plan in areas which had not
been excavated. While the element of
fabrication is not extensive or likely to
radically alter the interpretation of the
building it betrays an attitude to field-
work which has been pervasive for much
of the 20th century. Roussell had
encountered a number of well preserved
farmhouses in Greenland and had reflect-
ed extensively on the nature and develop-
ment of West-Nordic building custom.
His culture-historical and evolutionist
approach made him think in terms of
ideal types and it was these he set out to
find in Þjórsárdalur. It is apparent that he
already had a clear idea of what he was
expecting and he also had the architect's
eye for the structural logic of buildings
which made him expect to find evidence,
where in some cases there was none to be
found.
The excavator therefore has a para-
digm in mind before he begins his exca-
vation and his expectations clearly influ-
ence what is found and recorded. When
the excavation reveals by and large that
which was expected, the failure of every
detail to fit to that expectation is out-
weighed by the elegance of the paradigm
and such details are then either ignored or
simply "repaired" to fit the expected
88