Archaeologia Islandica - 01.01.2007, Qupperneq 89

A REASSESSMENT OF THE DEVELOPMENT OF THE CATHEDRAL AT GaRÐAR, GREENLAND
lem, however. Letting the foundation of
the east gable of the church run the entire
length of the gable would make sense if
one meant to add extra strength to the
gable wall in order to secure the stability
of the chancel arch. But the later expan-
sion of the church does not seem to have
increased the width of the chancel arch,
begging the question of why the builders
suddenly decided that further strength-
ening was necessary. That it was indeed
necessary cannot, of course, be discount-
ed completely, but if the middle part of
the foundation here was a later addition,
it is also possible that it represents the
very last phase of the church - a point I
will return to in phase 4.
As mentioned, Norlund believed
the enclosure south of the church to be
absolutely contemporary with Garðar 1.
He gave two main reasons for this, name-
ly that the western and eastern walls of
the enclosure cut under the foundations
of the church. The western one of these
foundations is supposed to cut under the
church foundations east of the proposed
door in the nave, while the eastern one is
supposed to cut under the southern foun-
dation of the south chapel (fig. 4). The
first statement is rather odd: if the west-
ern foundation of the enclosure cut under
the foundation wall of the church, this
would imply that the enclosure actually
predated the church, which is unlikely.
Rather it ought to have been bonded with
the foundation of the church. Looking at
figure 4, an explanation suggests itself.
It is likely that the western foundation of
the enclosure cut under the very south-
ern part of the church foundation, since
this part of the latter seems to be a later
addition - that is the non-coloured part
of the church foundation in figure 4. This
does not prove, however, that the enclo-
sure foundation was bonded with - and
thus absolutely contemporary with - the
foundation of Garðar 1.
The second argument is more
readily understandable. Here Norlund
states that the eastem foundation of the
enclosure (wall A in fig. 6) cuts under the
southern foundation of the south chapel.
However, this only proves that wall A
predates the south chapel, not that it is
contemporary with the chancel of Garðar
1. Norlund does not state anywhere that
wall A is bonded with the south wall of
the chancel, which is the sole argument
that would establish absolute contem-
poraneity between the church and the
enclosure. Unfortunately the issue cannot
be resolved based on the extant material,
and an excavation is necessary to settle
the matter absolutely. Another thing does,
however, point to the enclosure being a
later addition. This relates to the rather
strange way in which wall A meets the
church. Had it been planned from the
outset, it would seem logical to let wall
A meet the south-east corner of the nave
instead of the rather awkward arrange-
ment where the wall meets the church in
the corner between nave and chancel.
A final point to be made here
regards Norlund’s theory of an even ear-
lier church. He argued the presence of an
earlier church due to burials being found
beneath wall A in the south chapel. Hence,
if wall A was indeed erected simultane-
ously with the church, it seems likely that
an earlier church existed. However, if it
cannot be proved that the enclosure walls
are bonded with the walls of the church,
Norlund’s argument for the presence of a
church before Garðar 1 is gone. Summing
up, I would suggest that phase 1 of the
church at Garðar consisted of a church
with a Romanesque plan, corresponding
87