when the change of religion took place, and when Roskilde makes its appearance on the
map.
ROSKILDEThe written descriptions of Harald Bluetooth’s burial at Roskilde in 987 , which have
formed one of the primary arguments for the dating of the town to the tenth century,
have been subjected to a critical scrutiny that concludes that these events can hardly
have taken place at that site (Lund 1998 ). Similarly, the archaeological remains cannot
support a foundation date prior to the year 1000 , as only a couple of objects found
within the limits of the medieval town can be dated to the tenth century. The first time
the town can be said to appear with certainty in documentary sources is in an English
text from c. 1022 (Birkebæk 1992 : 58 ). At this time we also find the first archaeological
finds in the form of coins minted under Cnut the Great. Results from excavations
combined with stray finds and the ecclesiastical topography suggest an extensive settle-
ment, covering a considerable area in the eleventh century.
As is the case with the majority of the early medieval Danish towns, Roskilde was
founded on a navigable waterway at the head of a fjord, but the cathedral and – one
assumes – its associated royal manor were built on a 40 m high hill some 700 m from
the shore. If we add to this the location of the other early churches and an excavated
landing site on the fjord from the eleventh century, it seems that the town at this time
covered an area of perhaps half a square kilometre (Christensen 2000 : 9 – 21 and Ulriksen
2000 : 145 – 98 ). The markedly hilly terrain with watercourses and fords has drawn
natural boundaries between the churches and their adjacent buildings, giving an
impression of a settlement pattern reminiscent of what in northern and western Europe
has been called ‘the eleventh-century agrarian urban landscape’. This consists of several
separate settlements that only in the twelfth century combine to form a cohesive site.
When Roskilde gets its town wall in the middle of the twelfth century, and a true
settlement develops in the area around the cathedral, at the same time the old ‘suburbs’
and landing stage by the fjord are cut off.
The background to Roskilde’s location does not seem to have been the presence of an
existing trading site, nor does trade appear to have played an important role in the first
years of the town. Roskilde belongs to a group of bishoprics founded around the turn of
the first millennium, which the king and Church needed as administrative centres for a
new power structure. First and foremost the desire for good lines of communication
between the disparate parts of the realm, and a literally visible placement of church
buildings in the landscape, seem together to have determined the location of Roskilde,
coupled naturally with its proximity to the old centre at Lejre.
As we have seen, there is an impression of stability about Iron Age and Viking Lejre
until the end of the tenth century. At this time comes the disappearance of some of the
elements that, it is suggested here, were connected to pre-Christian cult, but the settle-
ment continues during the Middle Ages in the form of a manorial farm. The settlement
itself is not abandoned, but to judge from the excavations some of its functions are
transferred to the newly founded Roskilde. In this sense it is appropriate to speak of
pagan Lejre and Christian Roskilde.
–– Tom Christensen––