Archaeologia Islandica - 01.01.2007, Síða 92

Mogens Skaaning Hoegsberg
foundation was indeed a later addition,
one wonders why it would have been
inserted in connection with the exten-
sion of the chancel, since the chancel
arch does not seem to have increased in
width at any point. A final possibility is
that the insertion of the middle section
represents a reduction of the church at a
time when the extended chancel of phase
3 had either been rendered old fashioned,
gone out of use or fallen into disrepair.
Which of these possibilities is the more
likely is difficult to answer. A reduction
of the church could be seen as represent-
ing a desire to keep up with newer archi-
tectural currents, since rectangular stone
churches seem to have been the favoured
type of church building in Norse Green-
land after circa 1300 AD (Arneborg
2004, 251). A reduction could be seen as
a means to keep the cathedral modern.
Conversely it could also be interpreted
as a necessity, brought about by a failing
economy which had rendered the upkeep
of the extended chancel economically
impossible. At this point we know too
little about the economic development at
Garðar to gauge which explanation is the
more likely, provided of course that the
theory is at all valid.
It should be mentioned that the
Danish archaeologist Henrik M. Jansen
suggested that the cathedral might origi-
nally have been a rectangular church
which was later extended with the chancel
and chapels (Jansen 1972, 118). This is
not likely to have been the case. Jansen’s
suggestion is in line with the view put
forward by Norlund in De gamle Nordbo-
bygder ved Verdens Ende (1934) that the
oldest set of Greenlandic stone churches
were rectangular and that the churches
with a Romanesque plan were later (Nor-
lund 1934, 30). Norlund’s view was later
refuted by Roussell (Roussell 1941, 111-
126) and Roussell’s view is generally
accepted today (see below). Summing up,
phase 4, if it existed, represented a reduc-
tion of the church, where the extended
chancel of phase 3 was removed and the
cathedral was turned into a plain rectan-
gular stone church.
Dating
Dating stone built Norse Greenlandic
structures is notoriously difficult, and
this problem extends to the cathedral at
Garðar. As for the church buildings, there
is a general consensus among researchers
today that they followed the Scandinavian
norm. Thus the first generation of stone
churches in Greenland had a traditional
Romanesque plan consisting of a nave
with a smaller chancel, and the later gen-
eration of churches were built as rectan-
gular “gothic” buildings where nave and
chancel were equally wide. Generally, the
rectangular buildings are believed to have
been erected from circa AD 1300 onwards,
while the buildings with a Romanesque
layout are generally placed in the 1 lth to
13th centuries (Arneborg 2004, 251). In
phase 3 the cathedral was obviously dif-
ferent from these two models. Of all the
stone built churches in Norse Greenland
it has the richest architectural solution,
including chapels at the chancel. This can
be attributed to the fact that it was the
cathedral, but it does not go a long way
towards providing a firm dating.
Norlund believed that Garðar
1 (my phase 1) was built following the
establishment of the bishop’s seat, where
it replaced the supposed earlier church.
He noted the likeness between the enclo-
sure south of the church to the layout of
European ecclesiastical buildings - both
in connection with monasteries and cathe-
-V
90