Rows of wooden pillars : continued fieldwork at the Degeberga 
linear monument, with further parallels and societal implications 
Björk, Tony & Wickberg, Ylva
http://kulturarvsdata.se/raa/fornvannen/html/2015_241 
Fornvännen 2015(110):4 s. 241-256
Ingår i samla.raa.se
Rows of wooden pillars. 
Continued fieldwork at the Degeberga linear monument, with
further parallels and societal implications
By Tony Björk and Ylva Wickberg
Björk, T. & Wickberg, Y., 2015. Rows of wooden pillars. Continued fieldwork at
the Degeberga linear monument, with further parallels and societal implications.
Fornvännen 110. Stockholm.
In 2011, the remains of an unusual linear Iron Age structure were discovered at
Degeberga in Scania. The monument dates from the Late Roman Iron Age in its
earliest phase and from the Migration Period in its latest phase. We identified sim-
ilarities with structures at the cemeteries of Anundshög and Old Uppsala in the
Lake Mälaren area. In 2013, we revisited Degeberga for a second excavation cam-
paign. The picture of the site is becoming clearer; first of all, its date has been
markedly refined. The final destruction or abandonment of the monument
occurred shortly after 600 cal AD. There are notable consistencies with other sim-
ilar sites regarding layout and dating, for example with Önsvala, Färlöv and Tomte-
boda, which are further discussed here. Finally, we attempt to grasp the many
aspects of interpreting these linear monuments. We argue that they represent mul-
tifunctional places of regional importance, reflecting perceptions and goals com-
mon to politically and militarily influential upper stratum groups over a large part
of Scandinavia in the period AD 400–650.
Tony Björk & Ylva Wickberg, Sydsvensk Arkeologi AB, Box 134, SE–291 22 Kristianstad
tony.bjork@sydsvenskarkeologi.se, ylva.wickberg@sydsvenskarkeologi.se
During the excavation of a Bronze Age settle-
ment site in 2011, just west of the village of Dege-
berga in Scania, an unusual ancient structure was
discovered. It was a linear monument with post-
holes, pits and hearths from the Late Roman and
Migration Periods, which were flanked by a hol-
lowed-out ancient road and a cemetery of the
Early and Late Roman Period (Björk & Wick-
berg 2012; 2013). Degeberga is in a dramatic land -
scape, where three kinds of land meet: a domi-
nant ridge, hill slopes and a plain with wetlands.
The remains of the monument are located on one
of the heights in the area, in a widely sightlined
position. As far back as it can be studied on maps
(Buhrman 1684), the site has been a crossroads, a
strategic place in a landscape that was difficult in
parts to pass through.
The excavations of 2011 were a small develop-
ment-led project with limited resources. The Coun-
ty Archaeologist gave permission to depart from
the initial plan and uncover the part of the linear
monument that was within the development
area. During fieldwork, the Bronze Age settle-
ment, the Roman Period cemetery and 130 m of
the linear structure were investigated. Neither of
the linear monument's ends were found. It con-
sisted of large, massively stone-lined postholes,
regularly placed in a long row. They were flanked
by a row of pits with a similar rectangular shape
and orientation, also regularly placed. In 2011,
the row of pits was perceived to have been at least
70 m long and to have had a function connected
with heating. The row of pits ran parallel with the
row of postholes in the northern part of the exca-
Fornvännen 110 (2015)
Art. Björk 241-256_Layout 1  2015-11-26  14:37  Sida 241
vation area, but at the middle of the post row, the
line of pits curved in towards them. At this point,
the row of pits was intersected by a row of hearths;
these were not so regularly placed, forming a
slight curve. In the southern part of the excava-
tion area was part of a filled-up sunken lane that
had run parallel to the rows of posts and pits.
Radiocarbon showed that the curved row of
hearths was the oldest phase, from the Late Roman
Iron Age, while the rows of pits and postholes
belonged to the Migration and Early to Middle
Vendel Periods. No charcoal or carbonised seeds
were found in the postholes that could be attrib-
uted to their phase of construction or use, but we
assumed that they were coeval with the row of
pits, since the two structures related spatially to
each other, and for a large stretch were quite par-
allel. We interpreted the sunken lane as a com-
mon denominator for both phases in the history
of the monuments, probably being coeval or old-
er. It had a north-south orientation and it ex ten d-
ed towards the valley of the Forsakar stream to
the south. Near the Forsakar waterfall there was
probably a ford, providing good crossing to the
foot of the Linderöd ridge (Björk & Wickberg
2013; Wickberg 2014a).
It was clear that we had discovered an unusu-
al kind of ancient monument, even a new type. In
the literature we identified a few sites with simi-
lar structures, both in Scania and in the Lake Mä -
laren area, and we concluded they are evidence
for a radical transformation of the ritual land-
scape during the mid-1st millennium AD (Björk
& Wickberg 2012; 2013).
Continued fieldwork
The 2011 excavations raised several important
questions. The site is affected by ongoing con-
struction and cultivation, making it urgent to se-
cure scientific information. Therefore a research
project was started, funded by the Ebbe Kock
Foundation, the Anders Althin Foundation and
Sydsvensk Arkeologi AB. We resumed excava-
tions in 2013. A full report is available both in
print and online (Björk & Wickberg 2014;
www.sydsvenskarkeologi.se).
The second stage of investigation at Degeber-
ga had the following aims.
• To find out whether there were visible
traces of the sunken lane, or other 
features of the linear structure, on the
steep southern slope towards the For-
sakar stream.
• To determine if the row of postholes sur-
rounded an area, or if it was a directional
monument pointing toward something –
e.g. a grave monument.
• To establish the structure's dimensions,
how long the row of postholes was, 
primarily to the north and south.
• To investigate if it was possible to see
clear changes in the monument over its
various phases. 
• To define the chronology more clearly.
• To better understand the function of the
various parts of the monument. 
• To get a better idea of the size, extent
and use period of the cemetery.
To shed light on these questions we used visual
inspection of the area, magnetometry, metal detec-
tion, excavation and analysis of tree species and
radiocarbon.
The various field operations of 2013 took
place within a prospecting area of 65 000 m2. The
fieldwork started with a magnetometer survey,
which unfortunately could not trace the structures
we sought. A metal detector survey was con ducted
next to the cemetery investigated in 2011. This
was also without success. The poor result of the
magnetometer survey was given an explanation
when we dug search trenches to the north of the
Lillafors road. Intensive agriculture has re-sculpted
the terrain and caused extensive erosion.
We dug twelve rather small trenches with a
total extent of 1330 m2 (fig. 1). We also investi-
gated an additional number of features in our old
trenches, which were still open since the excava-
tions of 2011. In all trenches except one, features
of a prehistoric character were found. Note how-
ever that the trenches to the north of the Lillafors
road, next to the central plateau, contained very
few features and artefacts, and we did not find 
242 Tony Björk & Ylva Wickberg
Fornvännen 110 (2015)
Fig. 1. The Degeberga site with trenches and features
from the 2011 and 2013 excavations.
Art. Björk 241-256_Layout 1  2015-11-26  14:37  Sida 242
243Rows of wooden pillars
Fornvännen 110 (2015)
Art. Björk 241-256_Layout 1  2015-11-26  14:37  Sida 243
244 Tony Björk & Ylva Wickberg
Fornvännen 110 (2015)
Art. Björk 241-256_Layout 1  2015-11-26  14:37  Sida 244
Fig. 2. The Degeberga monument as far as it is known
today. Hearths represent phase 1, pit hearths and post
holes represent phase 2.
any additional graves. Agriculture and erosion has
worn down the top of the hill. Postholes and other
features grew more and more shallow the further
north we excavated. Only the bottom part of the
postholes and the pit hearths were left in the nor -
thern trenches. We could nevertheless establish
that the post row continued in a straight line in a
north-north-westerly direction, and that it was
at least 190 m long. We reinterpreted the row of
pits as a row of pit hearths, which had been much
longer than previously believed. We had believed
that its southern part belonged to the eastern,
curved row of hearths from the Roman Period
(fig. 2). New dates clarified that the southernmost
features belonged to the row of pit hearths and that
this structure dates from the Migration Period.
We could follow it for a short distance north of
the Lillafors road, and found that it had been at
least 150 m long in total.
Due to the severe erosion and the fact that the
features gradually became more and more shal-
low to the north, we could not determine whe ther
the rows had continued, changed course or ended
at that point. This remains to be clarified. Nor is
it clear how the rows continued to the south, as
part of this area has been dramatically remod-
elled by a railway embankment in the 19th cen-
tury. Part of it is also in a nature preserve, mak-
ing an excavation permit difficult to get. The row
of pit hearths seems to change course to the east,
but no continuation could be confirmed after the
two last ones, which were placed at almost a right
angle to the rest of the row (fig. 2). As mentioned
above, only the bottoms of the observed pit hearths
were preserved in this south-eastern part of the
area; some of these may have been disturbed by
ploughing and erosion, as they were in the nor-
thern part of the structure.
Analysis of plant macrofossils and tree spe cies
in a selection of soil samples from Iron Age fea-
tures identified charcoal from willow, hazel, birch,
pine and oak, along with fragments of burnt cereal
grains, probably barley. In addition to these plant
species, samples analysed after the 2011 season
contained charcoal from ash and burnt grains of
hulled barley (Hordeum vulgare var. vulgare). There
seems to have been a quite a variety of trees in the
area and the main cultivated crop, or perhaps the
crop selected for ritual use, was hulled barley.
There are a number of source-critical problems,
though. For instance, the selected samples came
from special features, the preservation conditions
of this site were poor and, especially, environmen-
tal issues were not at all decisive when we collect-
ed the soil samples. The primary purpose of the
analysis has so far exclusively been to obtain ma -
terial for dating with as low an intrinsic age as
possible.
Eight radiocarbon analyses were done on ma te-
rial from the 2013 excavation. The results are well
in line with the dates from the 2011 season. The
dates are related to the Late Bronze Age settle-
ment remains and to features included in the lin-
ear structure, which has been dated to a period
extending from the Late Roman to the Middle
Vendel Period. In aggregate, the radiocarbon dates
allow us to determine the phases of the monu-
ment much more clearly than before. An especial-
ly important sample was taken from a cow’s jaw
found in a posthole under a large stone that had
fallen after the post had collapsed or been
removed. A tooth from the jaw was dated to
570–660 cal AD (2 sigma, UBA-24678, 1425±28
BP), that is, the Early or Middle Vendel Period.
Therefore, the posts appear to have been fairly
long-lived; each can be estimated to have lasted
for about 100 years (Wickberg 2014a). The first
phase of the monument consisted of a curved row
of hearths, which was made during the Late Ro -
man Iron Age. The second phase consisted of two
parallel rows of pit hearths and posts, made during
the Migration Period. The structure was aban-
doned, or maintenance ended, some time in the
first half of the Vendel Period, most probably the
early 7th century cal AD (Björk & Wickberg 2014
p. 28 ff).
In summary, we found that the linear struc-
tures extended further north, at least 190 m in to tal.
The row of pit hearths continued south and then
probably turned east. Radiocarbon now gives a
much better idea of the use during the various
phases of the monument. Furthermore, the dating
shows that the rows of posts and pit hearths were
constructed during very short periods, probably
representing one or two brief events. These re sults
are valuable, not only to determine when and
how this particular linear structure was built, but
also to make valid comparisons with other sites,
which we will turn to in the following.
The Degeberga monument in its local context
The Degeberga monument was a highly visible
structure. It was meant to be experienced both
from near and far. About 500 m northwest of the
monument, eleven large flat stone grave markers
(7–18 m in diameter) bear witness to settlement
nearby of the Late Bronze and Early Iron Ages
(Raä Ö Sönnarslöv 247:1–2, 248:1, 249:1–3, 250:
1–3, 251:1–2). Remains of a sunken lane passes
be tween some of these grave monuments (Raä Ö
Sönnarslöv 249:2). A cupmark boulder shows an-
other dimension of the sacral environment (Raä
Ö Sönnarslöv 251:3).
One kilometre to the west of the linear struc-
ture are a burial mound and two flat stone grave
markers (Raä Degeberga 56:1, 58:1–2), and here
we find the 1st millennium AD place name Borr-
åkra. Both of these grave clusters are some ways
up the slope of the Linderödsåsen ridge, which is
partly wooded and abounds in boulders. Nearby
are two large areas with fossil fields, also contain-
ing a number of flat stone grave markers.
A final grave is one kilometre north-east of
the Degeberga monument. It is a barrow named
Torshög (“Thor’s mound”, Raä Degeberga 15: 1),
which is 15 m in diameter and 2 m high. It dates
most likely from the Early Bronze Age, but has
never been excavated. It is in a widely sightlined
spot along one of the old north-south roads
shown on early maps. The sunken lane and the
nearest graves are located on more level ground
above steep slopes. Below the main hill extends an
undulating landscape of sandy and easily worked
soil. The Degeberga monument was on one of
these hills. The Roman Period cemetery next to it
has yielded no traces of superstructures or other
markers, though of course this does not exclude
the possibility that there once were such.
Both the graves and the linear monument sug-
gest that the strategically positioned and widely
sightlined site must have had great significance
during the Iron Age. It is a place that many people
245Rows of wooden pillars
Fornvännen 110 (2015)
Art. Björk 241-256_Layout 1  2015-11-26  14:37  Sida 245
Fig. 3. Ancient monuments at
Degeberga, mentioned in the
text. A) The linear structure and
Roman Iron Age cemetery. B)
Cemetery with stone ship and
standing stones. C) Viking 
Period/ Early Medieval ceme-
tery. D) Medieval church.
must have passed, probably a crossroads. It is near
watercourses, where many types of landscape meet.
For some reason the importance of the site seems
to have diminished, or the centre of gravity drift-
ed, from the Vendel Period on. About 500 m east
of the linear structure and the cemetery is another
cemetery with standing stones, probably the re-
mains of stone ships (Raä Degeberga 13:1). The
second phase of the linear monument, c. AD
400–600, seems to precede or coincide with the
establishment of this cemetery. Today it retains
only badly damaged stone structures, with the
probable stone ship setting and an additional
eleven scattered standing stones within the cur-
rent village (fig. 3). This cemetery has seen no ex -
cavations, but the monument types belong to the
Migration and Vendel Periods (Carlie 1994). This
suggests that the sacred area, and probably the
settlement, grew or moved somewhat towards
the east during these centuries.
Another 200 m to the east, a cemetery with at
least five inhumation graves from the Viking
Period and probably the Early Medieval Period,
were discovered when a house was built in 1931
(Hansen 1931; Raä Degeberga 73:1). This suggests
a continued growth or movement of the sacred
area towards the east. The High Medieval village
and church in Degeberga were eventually estab-
lished on the slope to the south of the pagan
cemeteries. The ancient settlement sites, the
cemeteries and the grand monument were for-
gotten and the ground was used for pasture and
cultivation for generations. On the oldest maps
of the Degeberga and its lands, from the 18th and
19th centuries, is an evident field boundary in
exactly the same location and on the same orien-
tation as the rows of posts and pit hearths, at least
until the agricultural reform of 1813. After that,
the border was finally wiped out, after having
been a structuring element in the landscape for
about 1300 years (Lantmäteriet, 11-DEG-6). It is
fascinating to see that the linear mark lived on in
the landscape for so long, although the monu-
ment and all awareness of it had been gone for
ages.
Comparison with other sites
The Degeberga structure is difficult to under-
stand when viewed only from the information
gathered on site. It must be compared to similar
structures elsewhere. We have previously empha-
sised the strong similarities in structure and date
between the linear monuments at Degeberga,
Önsvala, Old Uppsala and Anundshög (Björk &
Wickberg 2013, p. 101 ff; see also Larsson 1982;
Sanmark & Semple 2011; Beronius Jörpeland et
246 Tony Björk & Ylva Wickberg
Fornvännen 110 (2015)
Art. Björk 241-256_Layout 1  2015-11-26  14:37  Sida 246
Fig. 4. Map showing the location of the discussed sites
with linear structures. 1) Degeberga. 2) Färlöv. 3)
Önsvala. 4) Tomteboda. 5) Old Uppsala. 6) Anunds-
hög.
al. 2013). With the new dates from the 2013 field-
work at Degeberga and after further research, the
connections between Degeberga and Önsvala
have been strengthened and we have also become
aware of two additional sites: Färlöv in Scania
and Tomteboda near Stockholm (fig. 4).
At Önsvala south of Lund, a cemetery threat-
ened by a gravel pit was excavated in 1968–70
(Raä Nevishög 1). The site is about 500 m north
of the small river Torrebergabäcken and the gra-
ves belong to a period from the Late Roman Iron
Age up to the Viking Period. Some 50 m to the
south-east the cemetery was flanked by a 50 m
long row of twelve rectangular pits or hearths at
almost equal distances to each other. In average
the distance from centre to centre was 4.2 m,
compared to Degeberga where the average was
5.1 m, centre to centre. Two of the pit hearths were
radiocarbon dated to the Late Roman and Mi gra-
 tion Periods (Lu-671, 1590±50 BP; Lu-672, 1600
±50 BP – see fig. 7). About 25 m east of the row of
hearths was a parallel trench, and a further 15 m
to the east was a parallel depression, at least 60 m
long and up to 7 m wide, with a partly sooty fill
that may have been the remains of a road. The
function of these features has not been previous-
ly determined, and their interpretation is uncer-
tain. On the bottom of the presumed road depres-
sion were seven hearths, and the fill contained
pottery ranging from the Neolithic to the Iron
Age (Larsson 1982; 2013; Björk 2005, p. 96 f,
197; Björk & Wickberg 2013).
In Färlöv, north-west of Kristianstad, part of a
cemetery was excavated in 1996–98. Smaller ex ca-
vations have also taken place on three later occa-
sions (Raä Färlöv 1, 2, 4, 7, 166). The cemetery
has proved to consist of several groups of burial
monuments on a ridge alongside an approxima-
tely 500 m long section of an ancient road. The
excavated part of the cemetery (Raä 1) is the
northernmost part of a much larger grave com-
plex. During excavation of a small section of this
northern part, several interesting finds were made
of high status burials and monuments. Among
them were a richly equipped weapon burial from
the Late Roman Iron Age (phase C1b), remains
of two extremely large stone ships from the Ven -
del or Early Viking Periods, and a rune stone of
similar date. The stone ships had been of monu-
mental size, about 50 and 80 m long respectively.
Another structure found was a row of seven large
postholes, extending along a 20 m long wi dely
sightlined spot in the centre of the cemetery. The
postholes measured 0.80-1.32 m in diameter and
0.46 –0.92 m in depth. The distance between them
was about 2.9 m, centre to centre, and they all
247Rows of wooden pillars
Fornvännen 110 (2015)
Art. Björk 241-256_Layout 1  2015-11-26  14:37  Sida 247
Fig. 5. Excerpt of the plan of the cemetery section
excavated at Färlöv in 1997. Black features are large
postholes, grey ones parts of burial monuments,
white ones are other features. Features labeled with a
prone S are modern disturbances. The feature labeled
with a question mark was severely damaged. The
dashed line marks the dividing line between an area
to the west where only the topsoil could be removed,
and an area east of it that where the features were
excavated.
had a massive stone lining (figs 5–6). Charcoal
from one of the postholes was dated to the Early
to Middle Vendel Period (Lu-4283, 1420±70 BP
– see fig. 7). The current road, running in a north-
south direction along the west side of the ceme-
tery, exists on the oldest map of the area from
1684. The rune stone stood beside this road. The
row of posts runs in the same direction as the
road, but at a distance of about 25 m to the east
(Björk 1999; 2005; 2010; Nilsson 2014; Wick-
berg 2014b). Today, the village of Färlöv is dom-
inated by a large Romanesque church with twin
towers. Not very long ago, remains of an Early
Medieval manor were found just west of the
church (Kockum 2009). There are thus several
manifest expressions of status and power in the
area, covering a long period, well into historical
times. The site has a strategic location, and, to all
appearances it had a considerable economic sig-
nificance. Finds of status markers and the monu-
ments on the site are probably the result of a situa-
tion where those who controlled this important
geographic position, at a watershed between dif-
ferent economic zones and important land and
water ways, also controlled and profited on com-
munication and trade in the area (Helgesson 2003,
p. 329 ff).
Degeberga has clear traits in common with
both Önsvala and Färlöv. With Önsvala the com-
mon feature is the row of pit hearths and with
Färlöv it is the row of extraordinarily large post-
holes. The dates of Degeberga, Önsvala and Fär-
löv agree extremely well. The rows of pit hearths
are dated unambiguously to the Migration Period,
while the rows of posts are dated to the Early to
Middle Vendel Period, with dates for the construc-
tion of the monument’s first phase and the aban-
donment of its second phase (fig. 7). There is thus
no doubt that there are strong links between
these sites, both in terms of the regular layout
and in how the pit hearths and postholes are
arranged, as well as in the dates and in what kind
of environments the structures are found.
There is yet another site with a linear struc-
ture that deserves attention in connection with
Degeberga. It is a cemetery from the Migration
and Vendel Periods at Tomteboda just north of
Stockholm (Raä Solna 39), excavated in 2001
(Hamilton 2003). The cemetery was on a south-
facing slope towards Lake Ulvsundasjön, which
is part of Lake Mälaren. 24 graves and other fea-
tures were excavated inclusing mounds, flat stone
grave markers, an unmarked inhumation and a
standing-stone circle (Sw. domarring). They were
richly furnished with beads, tools, combs, weapon-
ry etc. and belong to the period AD 400–800.
During fieldwork, fragments of three picture
stones made from red sandstone were found. At
least one of them has runes on the narrow sides.
The runes are of the early 24-character runic
alphabet, but only three can be read: “E R U”.
This find of early runes on a picture stone is
unique in the Mälaren area. It dates from the
Migration Period. The fragments were spread
out in secondary positions, mostly in the stone
fill of grave markers from the subsequent Vendel
248 Tony Björk & Ylva Wickberg
Fornvännen 110 (2015)
Art. Björk 241-256_Layout 1  2015-11-26  14:37  Sida 248
Fig. 6. Left: section
through feature O71,
one of the postholes in
the row at the Färlöv
cemetery. Photo by
Anders Edring. Right:
Thomas Linderoth
next to the section
through posthole 5250
at Degeberga. Photo
Tony Björk.
Period. The picture stones had been destroyed
already during the Early Vendel Period (Hamil-
ton 2003, p. 9 f). This suggests that there was a
takeover of power at this time by other people
than the descendants of those who had erected
the picture stones (Fischer 2005, p. 226).
A row of postholes was also found at Tomte-
boda. It was 53 m long, oriented west-south-west
to east-south-east, and located in the south-west
part of the cemetery. The row was discovered
when the excavation of the graves had finished
and the entire excavated area was dug slightly
deeper. The postholes were placed regularly about
4.7 metres apart, centre to centre, except for an
249Rows of wooden pillars
Fornvännen 110 (2015)
Fig. 7. Radiocarbon
analyses from Dege-
berga, Önsvala and
Färlöv. Note that the
dates of the pit hearths
at Degeberga and
Önsvala correspond
very well, as do the
dates of the postholes
at Degeberga and Fär-
löv.
Art. Björk 241-256_Layout 1  2015-11-26  14:37  Sida 249
Fig. 8. The Tomteboda cemetery. Black features are Migration Period graves, grey ones Vendel Period graves.
The row of postholes is emphasised with a notional line. Map based on information and maps in Hamilton
2003.
area in the north-western part where recent dig-
ging had probably destroyed three postholes (fig.
8). One of the postholes was covered by a crema-
tion layer belonging to the standing-stone circle,
which the finds date to the 8th century AD (Ha -
milton 2005, p. 41). Three other postholes were
covered by stone markers belonging to other
Vendel Period graves. The postholes in the cent-
ral part of the row were shallower than the rest,
which indicates that they had been damaged in
connection with the construction of graves (Ha-
milton 2003, p. 43). All the postholes were lined
with stones. The largest was 0.70 m deep, with a
massive stone lining. The posthole farthest to the
west was 0.55 m deep and 0.90 m in diameter at
the surface. Almost nothing was found in the
postholes (Hamilton 2003, p. 43).
We suggest that this row of posts was coeval
with the cemetery’s Migration Period phase, since
it demarcated the south-south-western part of
the cemetery at that time. The posts must have
been removed during the Early Vendel Period
since the postholes had been covered by Vendel
Period graves. There seems to have been a con-
tinuation in the distribution and direction of the
later graves, as they continued to follow the di rec-
tion of the row of posts in a west-north-west to
east-south-east direction when laid out. This imp-
lies that they followed a road running in this di-
rection. The broken picture stones were found in
and around several of the graves superimposed
on the postholes. Just to the north-east of it, a
number of small fragments of sandstone of the
same kind were found (Hamilton 2003, p. 25).
These small fragments may represent the spot
were the picture stones were broken apart. They
may have stood in the immediate vicinity. 
It cannot be determined whether the row of
postholes followed a road, like the ones on the
other sites, as no traces of a road were discovered.
We consider it likely. Indications of the will to
communicate with people moving in the land-
scape, on land or on water, can be seen in the sight-
lining of the cemetery, the picture stones, the align-
ment of the row of posts and the distribution of
the Vendel Period graves. The position of the site
was strategic by the side of a bay in one of entry
channels to Lake Mälaren, which was a vital part
of a vast communication network. There must
also have been important transportation routes
250 Tony Björk & Ylva Wickberg
Fornvännen 110 (2015)
Art. Björk 241-256_Layout 1  2015-11-26  14:37  Sida 250
on land in the area. The picture stones and the
runic inscription are unusual finds, and the com-
bination is unique. This speaks of a social envi-
ronment in the area during the Migration Period
which was clearly far out of the ordinary.
Earlier research has pointed out the Solna
area as especially interesting during the Viking
Period, judging by rune stones and hoards of gold
and silver. There are many indications of a cent-
ral importance during the Migration and Vendel
Periods too (Hamilton 2005). After the Viking
Age too, in the Early Medieval Period, the area
seems to have maintained a special position with
the building of a fortified round church in the 12th
century. This was another expression of power,
possibly with offensive intention (Hamilton 2005,
p. 50). Part of a rune stone has been bricked into
the church. But about AD 1250 at the time of the
foundation of Stockholm, there seems to have
been a shift in power, or in its location, and the
Solna area no longer saw expressions of any spe-
cial status (Hamilton 2005, p. 51).
We were intrigued to find an almost exact
parallel of the postholes in the row at Degeberga
with those at Tomteboda (fig. 9). This is the case
both in the size of the postholes and the distance
between them. At Degeberga, the distance from
centre to centre was about 4.6 m. At Tomteboda
it was about 4.7 m. Without a doubt, Tomteboda
must be counted as another example of these lin-
ear structures, just like the ones at Degeberga,
Färlöv, Önsvala, Old Uppsala and Anundshög.
Having said this, we will attempt the difficult task
of a closer interpretation of what these linear
structures represent and what they tell us about
the places where they have been encountered.
The risk of circular reasoning is great, but we ask
the reader to bear with us in stretching the mate-
rial in our attempt, as we are treading new ground.
Fig. 9. Comparison plan showing the almost exact
correlation between the postholes in the Tomteboda
row and part of the Degeberga row. Tomteboda to the
left (unfilled), Degeberga to the right (filled).
Connecting rituals, power and territories
Degeberga and the other sites described have many
things in common. Tab. 1 presents the parame-
ters that we consider central. The sites broadly
follow a common pattern. The rows of postholes
were absolutely straight with a strict geometry,
they were located at large cemeteries (often with
high-status graves) at important roads and they
represent the same timeframe, c. 200–600 AD. 
The rows of posts have flanked roads and must
have symbolised something that was evident at
the time. They do not seem to have functioned as
any kind of fortification, as the posts were sparse-
ly spaced and no traces have been seen of any-
thing that could have linked them together. A
symbolic, probably ritual marking seems more
probable, but a territorial marking could also have
been part of the purpose. The widespread Swe -
dish sites known today share so many close simi-
larities that they must represent a shared idea of
why and how to present and regulate certain pla ces.
Why were the post rows built and what was
their ideational background? We have suggested
that their origin may go back to the so called fire
cult sites, with single or double rows of hearths or
pit hearths, of the Late Bronze and Pre-Roman
Iron Ages (Björk & Wickberg 2013, p. 104 f).
These are found in northern Germany and Po -
land, Denmark and southern Sweden (Heidelk-
Schacht 1989; Thörn 1993; Nordqvist 2005; Hulth
2013). The fire cult sites apparently changed in a
profound way during the 4th and 5th centuries
AD. Degeberga offers the clearest example of this
change in the structure of the sites. Here we see a
251Rows of wooden pillars
Fornvännen 110 (2015)
Art. Björk 241-256_Layout 1  2015-11-26  14:37  Sida 251
252 Tony Björk & Ylva Wickberg
Fornvännen 110 (2015)
Degeberga Färlöv Önsvala Tomteboda Old Uppsala Anundshög
Row of posts X X X X X
Row of pit hearths X X ?
Close to cemetery X X X X X X
High status graves X X X X X
Old road/s – sunken lane/s X ? ? X X
Ford ? ?
?
X
Tab. 1. Comparison of common features at the discussed sites. The question marks represent pro-
bable presence, while blank fields represent absence or unknown.
direct continuity between the early and late
monument type, while several of the other linear
monuments seem to have been newly established
in the Late Roman or Migration Period. 
Under the influence of the Roman Empire,
Scandinavians probably acquired ideas about how
to structure space to achieve certain goals. After
the collapse of the empire, this led to a hybridiza-
tion of the ancient custom of fire cult places. The
tradition lived on in an altered form. Old ways of
constructing rows of hearths and pit hearths in
complex variations of more or less even rows and
clusters were transformed into strict straight lines
followed by equally strict rows of posts. Rows of
wooden poles recall stone pillars or columns in
strict rows. In any case, the role as a demarcation
or frame for the ritual area obviously lived on,
but the meaning and the message these struc-
tures mediated probably changed to some extent.
They seem to have kept their role as an important
component of ritual arenas where collective me m-
ories were created. But the extremely evident re-
gulation suggests that the transformation aimed
to clarify elements of control – that is, political
and economic power with strong territorial
claims. Thereby the layout connected to history,
while it also legitimised a new political order
(Björk & Wickberg 2013, p. 106).
The political and economic connections among
the sites with linear monuments are emphasised
by their many shared traits regarding vital com-
munication routes, chronology and, not least,
high-status environments. It has been suggested
that they had a multifunctional significance,
connecting to a broad spectrum of social, reli-
gious, judiciary and political spheres (Björk &
Wickberg 2013, p. 105 f; Beronius Jörpeland et al.
2013, p. 280; 2015, p. 4 f). 
This touches upon many of the social func-
tions associated with Scandinavian central places
and likewise with the much older continental opp-
ida (Helgesson 2002, p. 22 ff; Fernández-Götz
2014, p. 182). The combination of the aforemen-
tioned traits allows us to view the sites as rooted
in a context of centrality, from a power perspec-
tive. Social power can be described as net works
built out of four overlapping types of power – the
military, political, economic and ideological
(Bossen 2006, p. 95). Several of the ingredients
of such a power network seem to have left lasting
imprints on the sites with linear monuments. 
The construction of the long rows of posts re -
quired an organisation involving many people.
Such a coordination of labour requires trust or
direct control, which in turn points towards an
element of power. The regulated sites can tenta-
tively be seen as a step in the concentration of
power at the transition to the Late Iron Age in
Scandinavia. A common interpretation of the
changes seen at this stage is that military power
became concentrated to warlords and their retinue.
It depended on personal alliances and unions,
e.g. dynastic marriages, representing a step in a
process where stronger and stronger territorial
claims and boundaries were asserted (Ystgaard
2013, p. 45 f, 258 ff).
Hillforts are also relevant when a power per-
spective with strong emphasis on military and
political conditions is adopted. These are rare in
Scania, which gives us no clues, but there are
none close to the sites with rows of posts in Upp-
land either, where hillforts are common. Many
Art. Björk 241-256_Layout 1  2015-11-26  14:37  Sida 252
253Rows of wooden pillars
Fornvännen 110 (2015)
Fig. 10. Part of north-eastern Scania from Gerhard
Buhrman's 1684 map of Scania. Mentioned sites are
marked with circles. Note the positions, listed from
the south, of Degeberga, Lyngsjö, Vä and Färlöv along
the current road 19. The road’s course just north of Vä
was probably broken and changed with the founda-
tion of the town of Kristianstad at royal initiative in
the beginning of the 17th century.
Art. Björk 241-256_Layout 1  2015-11-26  14:37  Sida 253
Norwegian hillforts are also in peripheral loca-
tions in relation to the primary settlements. Thus
it has been suggested that many of them should
be seen as defensive points for local warlords,
rather than as central places in a territory. Their
interpretation is however not clear, and the status
of individual sites probably varied (e.g. Skre 1998,
p. 285 ff; Rundkvist 2011, p. 32 w. refs). Another
point of view is that the Migration Period hill
forts were part of a “landscape of mobility where
transit through it was the most important guid-
ing principle, rather than control of it from one
central point. At the same time, the hillforts may
be understood as an expression of a stage in a pro -
cess where a stronger tendency to asserting terri-
tory develops” (Ystgaard 2013, p. 294 f w. refs,
our translation). An outlook that attaches im -
portance to mobility and transit agrees well with
the observations made around the rows of posts,
which always seem to have a close connection to
important roads.
From a power perspective it has been assum-
ed that Färlöv was closely connected to the major
central place Vä on the Kristianstad plain (Björk
1999; 2003; Helgesson 2002; 2003). With the
observed similarities between Färlöv and Dege-
berga, the latter should also be brought into the
discussion about the political landscape in north-
east Scania. Degeberga and Färlöv are not far apart,
27 km as the crow flies, but located on either side
of the Kristianstad plain. However, there is not only
a geographical proximity. Both sites are located on
the periphery of the plain where the landscape
rises, in zones that are perceived as the begin-
ning of border areas towards other territories.
Moreover, both Degeberga and Färlöv are situa-
ted along the ancient road from northern to
south-eastern Scania (present-day road 19). The
similarities in the layouts, in the topographical
positions and the locations along the same vital
communication link, suggest to us that there was
more than one connection between these places.
The 1684 map shows that Vä is not only located
in a central part of the area, but also at the mid-
point of this important stretch of road (fig. 10).
Another important site on this road is Lyngsjö,
seat of the Gärd herred judicial assembly until 
1765 (Nilson 1984). None of these sites was ran
domly placed in the landscape. They are in strate
gic positions, at crossroads of important commu-
nication routes and often at fords or bridges. All
four sites are on the same vital route in the dis-
trict’s infrastructure – road 19 – which confirms that
this road played an important role for those who
had ambitions to control the area, not only in
recent centuries but also further back in time. The
connection between the road and the sites is so
strong that they appear to be among each other’s
preconditions.
We find just as strong a connection to impor-
tant roads at the grand sites of Old Uppsala and
Anundshög. But these two also have something
else in common, an element not known from the
Scanian sites: the enormous barrows. This extreme-
ly evident monumentality is the particular charac-
teristic of these sites today. At least in the case of
Old Uppsala, the role of the site as the region’s
top-level central place in terms of both politic and
religious functions is undisputed.
Olof Sundqvist (2013, p. 96 ff) has suggested
that the cult of several gods, as known from writ-
ten sources at Old Uppsala, may have been a
strategy of ruling that made it possible to gather
people from a larger area, even though their
regional/local cult practices may have had differ-
ent emphases. A cult site interweaving the cult of
deities with that of dead rulers gave a special legit-
imacy to the power of the leading family. When
considering the establishment of rulers’ seats, with
direct relation to places of worship and burial
that legitimised power, the step is not far to view
these sites, situated in infrastructurally and topo-
graphically strategic locations, as making specific
claims to land areas.
We obviously do not see the same kind of
extreme monumentality at Degeberga and Tomte -
boda as at Anundshög and Old Uppsala. The
regional symbolic language differs however, and
in South Scandinavian terms Färlöv should also
be considered extremely monumental with its
very large stone ships. In any case, there were pic-
ture stones, a stone circle and several other grave
monuments at Tomteboda; standing stones, very
large stone ships, a small mound or flatter grave
marker and a rune stone at Färlöv; and at least
the remains of stone ships at Degeberga. We re-
cognise these monuments from the Lake Mä -
laren area: particularly the large stone ships and
rune stone at Anundshög. They can be consid-
ered manifestations of the same desire to be seen.
Through their sheer sizes, they signal that the
sites were in the hands of people with the wish
and means to display the importance of their
ancestors and themselves. Stone ships of over 40
m in length have a pronounced monumentality
and Torsten Capelle (1986, p. 16) suggested that
they may not primarily have been seen as funer-
ary monuments, but rather as memorials to great
seafarers. We are well into the Vendel Period
when the huge mounds and ship settings are con-
structed on these sites and thus at a stage when
the rows of posts had become outdated. These
sites had entered their next phase of transforma-
tion, through new visual attributes.
Svante Fischer (2005) offers a thought-pro-
voking description of social conditions at a time
just before the erection of the rows of posts. These
may in fact reveal an influence from the Roman
Empire. In Fischer’s interpretation, Germanic
warriors trained in the Roman military machine
and participating in Roman wars brought their
experiences back to their home areas. After the
collapse of the West Roman Empire, a “curious
blend of Roman imperialism and Germanic imi-
tatio Imperii” grew up. This led to a “regression of
society under a criminal social order”, which some
researchers discussing socio-economic organisa-
tion during the Migration Period have called a
“looting economy” (Plünderungsökonomie). Fischer
instead calls it kleptocracy. On the continent, the
empires of the kleptocrats were violent and usu-
ally short-lived. Few survived for more than a
couple of generations (Fischer 2005, p. 119 ff).
Should we see the establishment of the sites with
the rows of posts in this light? As manifestations
of a reestablishment of a stable power structure?
Environments with linear monuments might be
interpreted as a reflection of the establishment,
legitimisation and stabilisation of regional net-
works of power after the tumultuous period fol-
lowing the collapse of Western Rome. This was a
time when the basis for political systems was
transformed completely, with a change in the
supply of various attractive goods, new routes
and new alliances. This was followed by a consol-
idation manifested by large burial monuments,
which may also be seen as an expression of the
emergence of a Nordic identity during the Ven -
del Period, increasingly culturally isolated from
much of Christian Europe (Helgesson 2002, p.
38).
There is much to suggest that the sites we dis-
cuss played a multifaceted role. Some it seems
had several functions, while others appear to
have had fewer. It is clear that the rows of posts
were erected in places that were of vital impor-
tance to control. The similarities between the
sites show that they were built by people with far-
reaching contacts and the same kind of ambi-
tions (Wickberg 2014a). Everything indicates
that they were places of regional importance, even
though they seem to represent different organi-
zational levels and functions. Old Uppsala and
254 Tony Björk & Ylva Wickberg
Fornvännen 110 (2015)
Art. Björk 241-256_Layout 1  2015-12-03  16:11  Sida 254
Anundshög probably represent important power
manifestations at core points in the do mains of
the ruling families. We  that Degeberga, Önsvala,
Färlöv and Tomteboda represented a lower hier-
archical level, controlled by close relatives or
allies of the ruling families, and were key parts of
both the power network and the physical road
network of larger areas. At some of the sites with
Migration Period rows of posts, the power func-
tion obviously lived on with evident political,
economic and ideological attributes long after the
Iron Age (Beronius Jörpeland et al. 2015). The
clearest examples of this are Old Uppsala and
Färlöv where continued elite activity is seen into
the 12th century.
Finally, we believe that our initial interpreta-
tion of the linear monuments (Björk & Wickberg
2013) has been further strengthened by a widened
basis for dating and by additional comparisons
with Färlöv and Tomteboda. Some of the differ-
ences probably reflect varying local circumstan-
ces and differences in the importance of each indi-
vidual site, but altogether the detailed similari-
ties noted cannot be coincidences. We find that the
evidence strongly suggests that the sites repre-
sent multifunctional places of regional impor-
tance, reflecting ideas, perceptions and goals com-
mon among politically and militarily influential
upper-stratum groups, over a large part of Scan-
dinavia around AD 400–650.
Acknowledgements
Once more, we owe a big debt to landowner Mattias
Persson for letting us dig in his fields. Thanks also to
Marie Olsson (Länsstyrelsen i Skåne län), Helén Lilja
(Regionmuseet Kristianstad), Bertil Helgesson, Nils
Johansson and Thomas Linderoth (Sydsvensk Arkeo-
logi AB), Lisa Skanser (Västmanlands läns museum)
and Martin Rundkvist (Fornvännen). Our English was
revised by Judith Crawford, whom we also thank for
her good work.
Bibliography
Beronius Jörpeland, L., Björk, T., Gatti, C., Göthberg,
H., Sanmark, A., Seiler, A., Wickberg, Y. & Wik-
borg, J., 2015. Makt och monument under järn -
åldern. Badelundabygden 33. Västerås. 
Beronius Jörpeland, L., Göthberg, H., Seiler, A. & Wik -
borg, J., 2013. Monumentala stolprader i Gamla
Uppsala. Fornvännen 108.
Björk, T., 1999. For Gods and Men to behold. Excava-
tions at an Iron Age Grave-field in Färlöv, 1996-
1998. Lund Archaeological Review 1998. Lund.
– 2002. Just Another Central Place? A Critical
Description of Vä. Lund Archaeological Review2001.
Lund.
– 2005. Skäran på bålet: om den äldre järnålderns gravar
i Skåne.Dept of Archaeology, University of Lund.
– 2010. Gravfältet i Färlöv. Arkeologiska förundersök- 
ningar och undersökningar 1995– 2005. Rapport Re -
gion museet Kristianstad 2010:4. Kristianstad.
Björk, T. & Wickberg, Y., 2012.Degeberga 26:3, Degeber-
ga socken, Kristianstad kommun, Skåne län. Arkeologisk
undersökning 2011. Sydsvensk Arkeologi rapport
2012:20. Kristianstad/Malmö.
– 2013. Linear Iron Age monuments at Degeberga in
Scania and elsewhere in Sweden. Evidence for a ra d-
ical transformation of the ritual landscape. Forn-
vännen 108. 
– 2014. Degeberga 26:3, Degeberga socken, Kristianstad
kommun, Skåne län. Arkeologisk undersökning 2013.
Sydsvensk Arkeologi rapport 2014:60. Kristianstad/
Malmö.
Bossen, C., 2006. War as Practice, Power, and Proces-
sor: A Framework for the Analysis of War and
Social Structural Change. Otto, T. et al. (eds). War -
fare and Society. Archaeological and Social Anthropo-
logical Perspectives. Aarhus.
Buhrman, G., 1684. Karta över Skåne år 1684. (Map of
Scania in 1684.) Original in the Military Archives
of Sweden.
Capelle, T., 1986. Schiffsetzungen. Præhistorische Zeit -
schrift 61:1. Berling / New York. 
Carlie, A., 1994. På arkeologins bakgård. En bebyggelsear-
keolgosisk undersökning i norra Skånes inland baserad
på synliga gravar. Acta Archaeologica Lundensia,
Series in Octavo 22. Lund.
Fernández-Götz, M., 2014. Central Places and the Con -
struction of Collective Identities in the Middle
Rhine-Moselle Region. Popa, C.N. & Stoddart, S.
(eds). Fingerprinting the Iron Age. Approaches to Iden-
tity in the European Iron Age. Integrating South-Eastern
Europe into the Debate. Oxford.
Fischer, S., 2005. Roman Imperialism and Runic Literacy.
The Westernization of Northern Europe (150–800 AD).
Aun 33. Uppsala. 
Hamilton, J., 2003. Tomtebodagravfältet. Uppland, Solna
socken, Polisen 2, RAÄ 39. Rapport UV Mitt DAFF
2003:8. Stockholm.
– 2005. Solnas sista gravar och järnåldern i staden.
Andersson, L. et al. (eds). Uppdrag Arkeologi 2. Stock- 
holms län under dina fötter. Årsbok 2004. Stockholm.
Hansen, F., 1931. Undersökta skelettgravar. Rapport
över gjord undersökning av RAÄ 73, Degeberga sn.
Archive report in Lunds Universitets Historiska
Museum.
Heidelk-Schacht, S., 1989. Jungbronzezeitliche und früh-
255Rows of wooden pillars
Fornvännen 110 (2015)
Art. Björk 241-256_Layout 1  2015-12-03  16:27  Sida 255
eisenzeitliche Kultfeuerplätze im Norden der DDR.
Schlette, F. & Kaufmann, D. (eds). Religion und Kult
in ur- und frühgeschichtlicher Zeit. Berlin. 
Helgesson, B., 2002. Järnålderns Skåne. Samhälle, centra
och regioner.Uppåkrastudier 5. Lund.
– 2003. Central Places and Regions in Scania during
the Iron Age. Some examples. Larsson, L. & Hårdh,
B. (eds). Centrality – Regionality. The Social Structure
of Southern Sweden during the Iron Age. Uppåkra-
studier 7. Lund.
Hulth, H., 2013. Att återvända. Arkeologi i olika skeden
från Södra Gärdet i Ultuna. Arkeologiska för- och slut-
undersökningar, Fornlämning Uppsala 555, Uppsala
stad, Uppland. SAU rapport 2013:6. Uppsala.
Kockum, J., 2009. En medeltida källare i Färlöv. RAÄ 67,
Färlöv socken, Kristianstads kommun, Skåne län. Arkeo-
logisk förundersökning 2009. Rapport Regionmuseet
Kristianstad 2009:53. Kristianstad.
Larsson, L., 1982. Gräber und Siedlungsreste der Jün-
geren Eisenzeit bei Önsvala im südwestlichen
Schonen, Schweden. Acta Archaeologica 52 (1981).
Copenhagen.
– 2013. Rich women and poor men. Analyses of a
cemetery at Önsvala in the hinterland of Uppåkra.
Hårdh, B. & Larsson, L. (eds). Folk, fä och fynd.
Uppåkrastudier 12. Lund.
Nilson, T., 1984. Kring Lyngsjö ting. Gärds härads hem-
bygdsförenings årsbok 1984. Degeberga.
Nilsson, I-M., 2014. Fastigheten Färlöv 82:1, Färlövs sock-
en, Kristianstads kommun. Arkeologisk förundersök-
ning 2012. Sydsvensk Arkeologi rapport 2014:45.
Kristianstad/Malmö.
Nordqvist, B., 2005. Kultplatsen vid Stretered. Arkeolo-
gisk för- och slutundersökning, Västergötland, Kållered
socken, Stretered 1:1. Rapport UV Väst 2005:1. Kungs-
backa.
Rundkvist, M., 2011. Mead-halls of the Eastern Geats.
Elite Settlements and Political Geography AD 375–1000
in Östergötland, Sweden. KVHAA Handlingar, An tik-
variska serien 49. Stockholm. 
Sanmark, A. & Semple, S., 2011. Tingsplatsen som arkeo-
logiskt problem. Etapp 3: Anundshög. Arkeologisk prov -
undersökning – forskning. Badelunda 431, Långby 7:3
och 8:10, Badelunda socken, Västerås kommun, Väst-
manland, 2011. Centre for Nordic Studies, Kirk-
wall.
Skre, D., 1998. Herredömmet. Bosetning og besittelse på
Romerike 200–1350 e.Kr.University of Oslo.
Sundqvist, O., 2013. Gamla Uppsala som förkristen
kultplats: en översikt och en hypotes. Sundqvist,
O. & Vikstrand, P. (eds). Gamla Uppsala i ny be -
lysning. Religionsvetenskapliga studier från Gävle
9.
Thörn, R., 1993. Eldstadssystem – fysiska spår av brons-
ålderskult. Ett försök att spåra kultplatser och kulturin-
fluenser. BA thesis. University of Lund.
Wickberg, Y., 2014a. Degebergamonumentet – spår
efter en stormannamiljö. Gärds härads hembygds-
förenings årsbok 2014. Degeberga. 
– 2014b. Färlöv 1:32 och 76:10. Färlöv socken, Kristian-
stad kommun. Arkeologisk förundersökning 2012. Syd-
svensk Arkeologi rapport 2014:34. Kristianstad/ Malmö.
Ystgaard, I., 2013. Krigens praksis. Organisert voldsbruk og
materiell kultur i Midt-Norge ca. 100–900 e.Kr. Sek-
sjonen for arkeologi og lturhistorie, NTNU Viten-
skapsmuseet. Trondheim.
256 Tony Björk & Ylva Wickberg
Fornvännen 110 (2015)
Art. Björk 241-256_Layout 1  2015-12-03  16:28  Sida 256