Page 1
ARCHAEOLOGIA ISLANDICA 9 (2011) 41-49
Comparison of the distribution of pagan burials in Iceland with medieval
information about the number of farmers in different parts of the country allows a
division of the country into three zones of low, medium and high frequency of
pagan burials relative to the number of settlements. Possible explanations for these
differences are briefly explored. This paper is a product of the project Death and
burial in Iceland for 1150 years and sets out some of the problems it aims to solve.
Orri Vésteinsson, Department of Archaeology, University of Iceland, Reykjavík.
Email: orri@hi.is
Keywords: Viking Age, Burial sites, Distribution, Regionality
It is a well known aspect of the Icelandic
corpus of pagan burials that there are
significantdifferences
geographical distribution (fig.1). In some
regions there are many while in others
there are few or none. There are two
schools of thought to explain this. One
holds that the difference is primarily an
effect of discovery (e.g. Eldjárn 2000,
257-61), that burials are more likely to
come to light in regions where soil erosion
has been active and in landscapes where
roads are liable to built in locations where
pagan cemeteries were situated. The other
suggests that the differences reflect real
variation in burial practice: that pagan
burials are few in some regions because
pagan burial rites were not practiced there
to the same degree as elsewhere (e.g.
Einarsson 1989; 1994, 40-67).
My aim here is not to take sides in this
debate but rather to point out that its
premise is not as straightforward as it has
intheir
been made out to be. The observation that
there are differences in the numbers of
burials between regions is obviously
correct and significant but it fails to take
into account that there were differences in
the number of settlements, and hence
population levels, from one region to the
next. Consideration of these leads perhaps
not so much to a solution to the problem
as it has been defined by earlier debate,
but rather to a more nuanced appreciation
of what such points on the map may
imply, whether in terms of belief systems
or settlement patterns.
The number of settlements is not
known for the pagan period but the
relative carrying capacity of each region
can be gauged from later evidence. This is
not straightforward however and each
type of evidence comes with its own set of
problems. The earliest comprehensive
survey of farms in Iceland is from
1695-96 (Lárusson 1967), complemented
ORRI VÉSTEINSSON
A NOTE ON THE REGIONAL DISTRIBUTION OF
PAGAN BURIALS IN ICELAND
JOURNAL 2011_3.3.qxd:4issue.qxd 15/12/11 14:07 Page 41
Page 2
ORRI VÉSTEINSSON
42
by the first national census in 1703
(Manntal á Íslandi). These data provide a
basis on which to build but they are
obviously problematic in that they
postdate the pagan period by seven
centuries. In some regions at least there
are indications that significant changes in
the number of settlements ocurred in the
interim, suggesting that the 1690s figures
cannot be used without qualification as
proxy data for the Viking age. Earlier
figures are available; tax-paying farmers
were counted by region and quarter in
1311 (DI II, 373-75; IV, 9-10; XII, 20-21)
and in his Íslendingabók Ari provides
numbers of assembly-tax paying farmers
by quarter, said to have been gathered in
preparation for the establishment of a new
diocese at Hólar in 1106 (ÍF I, 23). Both
these sets of figures clearly suffer from
lack of precision, especially Ari’s which is
all in rounded hundreds. Serious doubt
may be cast over the organisational ability
of the Icelandic church around 1100 to
collect this sort of data in anything
resembling an accurate manner. The
figures from 1100 should therfore be
regardedasan informed
accurate enough for its stated purpose – to
split the existing diocese in two, requiring
assessments of the adequacy of revenues
for the new see at Hólar as well as of the
scale of the loss of revenue to Skálholt –
but not as exact numbers based on actual
counting. In addition to the imprecision,
which cannot be quantified, the 1100
figures only cover those householders
who owned the minimum amount of
property to be liable to pay assembly-tax.
It is unknowable how many households
estimate,
Figure 1. The distribution of pagan burials in Iceland. Based on Eldjárn 2000, with data on more
recent finds supplied by Adolf Friðriksson (pers. comm.).
JOURNAL 2011_3.3.qxd:4issue.qxd 15/12/11 14:07 Page 42
Page 3
were exempt from paying the tax and it is
conceivable that this proportion differed
from one quarter to the next.
Although the 1311 figures seem to be
based on an actual count, with 264
farmers in one region and 268 in another,
the rest of the numbers are given in
rounded tens and, more worryingly, there
is a mismatch between the totals given for
the quarters and the sums of the numbers
given for the regions within each quarter.1
The mismatch is small enough that it does
not affect the overall credibility of the
1311 figures, which are also in broad
agreement with both the 1100 and 1690s
numbers, but it serves as a warning that
their accuracy cannot be relied on. Like
the 1100 figures the ones from 1311 only
include those farmers who paid tax,
leaving out an uknonwn number of
households which did not. Comparison
with data from around 1700 suggests that
the 1100 and 1311 figures represent
lögbýli, assessed farm units, rather than
households, which in 1703 were nearly
two to each farm.
These figures are therefore primarily
useful as a guide to the relative population
levels between quarters and regions. As
these must ultimately relate to the
productive capacity of the land (and sea) it
is reasonable to assume that the same
proportions applied in the Viking age.
Looking first at the proportion of
pagan burial sites (N=170, see Friðriksson
& Vésteinsson this volume) to the later
population proxies, a well known pattern
emerges. Against both the c. 1100 and
1311 data, the highest proportion of burial
sites is in the North and the lowest in the
West although the difference is greater in
1311 (three times as many sites in the
North than West in c. 1100 as opposed to
four times in 1311). The other main
A NOTE ON THE REGIONAL DISTRIBUTION OF PAGAN BURIALS IN ICELAND
43
Figure 2. Burial sites (find locations) as a percentage of the number of tax-paying farmers by
quarter in c. 1100 and 1311.
1Björn M. Olsen (1907-15, 303-307) tried to argue the mismatch away by suggesting that the largest discrepancy, that of
tax-paying farmers in the south, was due to a failure of a scribe to translate the arabic numeral 500 correctly to long
hundreds, but even if this was right it would not explain the smaller inconsistencies in the figures for the northern and
western quarters.
JOURNAL 2011_3.3.qxd:4issue.qxd 15/12/11 14:07 Page 43
Page 4
difference is that in c. 1100 the East and
the South have an equal proportion while
in 1311 the East is much closer to the
North. The reduction in farmers’ numbers
between c. 1100 and 1311 is 35% in the
East, compared to 20% in the North, 8,5%
in the South and a slight increase in the
West, and while some of this decrease is
undoubtedly real it cannot be precluded
thatthec.1100figuresfortheEastmaybe
an overestimate, possibly resulting from
the bishop having less reliable estimates
for that quarter, always somehow the most
remote and obscure in our medieval
sources. It is certainly difficult to locate so
many abandoned farms in the East (which
can be done in the North) or to suggest
reasons for why so many more eastern
farmers should have slipped below the
poverty line in the 12thand 13thcenturies
(although see Rafnsson 1990 for that sort
of an argument).
Apart from confirming in broad
outlinethe well-known
differences in burial numbers (see fig. 1)
these figures are primarily useful to help
interpretthenextset,presentedinfigure3.
This shows the proportion of burials to the
number of farmers by region in 1311. This
breaks the country down to nine regions,
regional
ORRI VÉSTEINSSON
44
Figure. 3. Burial sites (find locations) as a percentage of the number of tax-paying farmers by
region in 1311 (see also fig.4).
JOURNAL 2011_3.3.qxd:4issue.qxd 15/12/11 14:07 Page 44
Page 5
two
Borgarfjörður, that straddles the boundary
betweenthe southern
quarters. The regional division mostly
follows later county (sýsla) boundaries
except in the West where the northern
border of Borgarfjörður is set at Langá
(bisecting modern Mýrasýsla) and the
southern border of Vestfirðir is set að
Þorskafjarðará (bisecting modern Austur
Barðastrandarsýsla).
Figure 3 gives a much more nuanced
picture of the distribution but also needs
some qualification. The proportion of
pagan burials to farms is clearly highest in
the Northeast and in Rangárvallasýsla in
thesouth.The
Rangárvallasýsla is however undoubtedly
an effect of soil erosion which has
for each quarterand one,
and western
highfigurefor
misproportionately ravaged that region
over others and is responsible for three
quarters of burial finds there. Everywhere
else it is half or less (Eldjárn 2000, 259)
suggesting that actual burial frequency in
Rangárvallasýsla is significantly less than
in the Northeast, although still substantial.
Within the Northeast there is only one
burial site on the eastern seaboard
(Austfirðir) suggesting that the proportion
for Fljótsdalshérað
although the reduction in farm numbers
between c. 1100 and 1311 probably means
that it was no higher than in Eyjafjörður or
Þingeyjarsýsla. The most surprising figure
in this set is no doubt the very low
proportion of burials in Árnessýsla,
Gullbringu- og Kjósarsýsla. In part this is
again due to zonation splitting the region;
is even higher,
A NOTE ON THE REGIONAL DISTRIBUTION OF PAGAN BURIALS IN ICELAND
45
Figure 4. Burial sites as a proportion of tax-paying farmers by region in 1311.
JOURNAL 2011_3.3.qxd:4issue.qxd 15/12/11 14:07 Page 45
Page 6
there are only two burial sites in the whole
of Gullbringu- og Kjósarsýsla, suggesting
it belongs to the same zone of very low
burial freqency as the western quarter, but
even if Árnessýsla is considered in
isolation(based on
weighingforthedifference)itcomesoutat
only 3,75%, on a par with Húnavatnssýsla
and Skagafjörður, and perhaps not so
differentfrom Rangárvallasýsla.
Skaftafellssýsla large swaths of land have
been lost in glacial floods no doubt
skewing the figure which would otherwise
be higher, presumably higher than in the
West but lower than in the Northeast.
Less dramatic but potentially no less
influential geo-morphological process
like subsidence and coastal erosion may
also affect the distribution in particular
regionsbutthisissueremainsunderstudied
(see Kristiansen ed. 1985 for possible
approaches).
Keeping these qualifications in mind it
is possible to suggest that Iceland can be
divided into three zones of burial
frequency (fig.5). The zone of highest
densityisin the
Eyjafjörður to Fljótsdalshérað, a zone of
lowest density is in the West, from
Reykjanesto Hrútafjörður,
intermediatezone
inbetween,in
Skagafjörður and the southern plains from
Reykjanes to Hornafjörður. The eastern
fjords may belong to this zone or the one
of lowest frequency – the total number of
farms is too low to say with confidence.
Although finds of new burials in the
last 10 years largely confirm these
patterns, with most found in the Northeast,
they also clearly reflect research activity,
1690s figures,
In
Northeast,from
and
regions
an
inthe
Húnavatnssýsla,
which has focused very much on the
Northeast and the Northwest, with two
new burial sites in the Northwest
representing a nearly 30% increase in that
region. Systematic investigation (already
underway, led by Adolf Friðriksson) will
tell to what extent these patterns actually
are real or whether they are predominantly
the effect of the accidents of discovery. In
the safe knowledge that this issue will be
cleared up in the not-so-distant future it is
permissable to present a few speculative
explanations which may give some food
for thought in the meantime.
One possible explanation for the
difference in burial frequency is that it
reflectsactualpopulationlevels.Thatthere
simply were many more people in the
Northeast than in the West in the Viking
age.Itiscertainlypossibletoarguethatthe
difference between the Northeast and the
mid-North and South is due to this factor.
Intensive survey work in the Northeast has
in recent years documented a high number
of farms abandoned gradually from the
10thto the 14thcenturies (Lárusdóttir
2007, Lárusdóttir & Hreiðarsdóttir 2011,
Ólafsson ed. 2008, Vésteinsson ed. 2011).
In the investigated regions there was a
more than 50% reduction in the number of
settlements in this period, and although it
does not follow that the population
reduced at the same rate it is certainly
possible that it reduced by enough to
account for the difference between the
ratios in the Northeast and the mid-North
and South. It is much harder to make this
case for the West, which would have had
to be virtually uninhabited in the Viking
age for this explanation to work. That is
not a likely proposition although a case
ORRI VÉSTEINSSON
46
JOURNAL 2011_3.3.qxd:4issue.qxd 15/12/11 14:07 Page 46
Page 7
could be made that Vestfirðir at least were
much less densely inhabited then than
later.
Other explanations must therefore be
sought for the difference between the West
and the rest of the country. One could be
that the difference lies in a different ratio
of cemeteries to settlements, that there
were fewer(and
cemeteries in the West. At a stretch the
Berufjörðurand
cemeteriescouldbemadetofitthiskindof
scenariobuttherestofthesitesintheWest
are too incompletely preserved, reported
and/or understood to support such a
hypothesis. To follow this train of thought
a bit further however it is conceivable that
if cemeteries in the West as a rule served
multiple settlements (as opposed to single
settlements which was the norm in other
thereforelarger)
Hafurbjarnarstaðir
regions – see Friðriksson and Vésteinsson
this volume) then the change in burial
custom following the conversion to
Christianity might not have entailed the
relocation of burial grounds; the pagan
cemeteries might be capped by their
Christian successors. There is only the
thinnest shred of evidence supporting this
notion: the description of an apparently
pagan burial in the churchyard at
Laugardalur in Tálknafjörður in 1603
(Annálar V, 505-506), but the otherwise
complete absence of such indications
makes this hypothesis difficult to sustain.
An aspect of the pagan burial evidence not
considered here but which might be
relevant in this context is the number of
graves in each cemetery. It would support
the notion of fewer cemeteries per
settlement if the western cemeteries were
A NOTE ON THE REGIONAL DISTRIBUTION OF PAGAN BURIALS IN ICELAND
47
Figure 5. Proposed zones of burial frequency in Iceland
JOURNAL 2011_3.3.qxd:4issue.qxd 15/12/11 14:07 Page 47
Page 8
significantly larger than in other parts of
the country. In Berufjörður in the West a
cluster of four cemeteries with 4-8 graves
each ups the total figure of graves for the
West considerably, making the West look
slightly less anomalous than when only
burial sites are considered. At present
however only a handful of pagan
cemeteries have been fully excavated so
most figures of grave numbers are likely to
be underestimated (Friðriksson 2009). At
present it is therefore not possible to argue
that there were regional differences in the
sizes of cemeteries. That, however, is an
issue which ongoing and future research is
set to throw light on. More circumstantial,
but no less intriguing, is the fact that the
ratio of Christian cemeteries to settlements
is significantly lower in the West than in
other parts of the country (Friðriksson &
Vésteinsson this volume). If there were
already fewer cemeteries per settlement in
the West before the conversion that might
explain this pattern in Christian times.
It is also possible that there were
regional differences in burial forms which
affect the distribution. Cremations have
been suggested as their absence from the
Icelandic burial record is somewhat
perplexing given the frequency of this rite
in Scandinavia in the Viking age. If
cremations were widespread they should
however have been detected (a single
unsubstantiated case has been reported –
Byock et al. 2005) unless they were less
likely than inhumations to have included
grave-goods, another proposition that
would be difficult to sustain. Water burials
are a more likely candidate. If they were,
for some reason, more common in the
West than in other parts of the country that
would explain the difference. This would
however be exceedingly difficult to
demonstrate, and the lack of finds of
human bones and artefacts from lakes,
rivers and bogs speaks against it.
A final possibility is that there were
regional differences
frequency but in the prevalence of grave
goods. If grave goods were placed in
graves much more rarely or in much
smaller quantity in the West than other
partsofthecountrythiswouldclearlyhave
affected the archaeological visibility of
pagan burials there because the vast
majority of such burials are identified as
such on the basis of the presence of grave
goods. It is well known that pagan burials
can have no grave goods whatsoever but
the frequency of burials with grave-goods
relative to those without is unknown as
pagan burials without grave-goods have
primarily been identified in the very few
cases where controlled excavations of
whole cemeteries have been carried out.
Isolated findsof
gravegoods are as a rule not included in
Eldjárn’s catalogue (although Tyrðilmýri
in the West is one exception, Eldjárn 2000,
120) although he publishes a number of
them as possibles alongside the main
assemblage. The distribution of these
burials largely follows that of the more
securely identified pagan burials and there
are certainly not significant numbers of
them in the West. Systematic recording of
such finds is however only now underway,
as a part of the project Death and burial in
Iceland, and may give grounds for
examining this explanation more closely.
notinburial
burials without
ORRI VÉSTEINSSON
48
JOURNAL 2011_3.3.qxd:4issue.qxd 15/12/11 14:07 Page 48
Page 9
Bibliography
Annálar 1400-1800. Annales islandici posteriorum
sæculorum I-VIII, Reykjavík: Hið íslenzka
bókmenntafélag 1922-2002.
Byock, Jesse; Phillip Walker, Jon Erlandson, Per
Holck, Davide Zori, Magnús Guðmundsson &
Mark Tveskov 2005, ‘A Viking age valley in
Iceland: The Mosfell archaeological project.’
Medieval Archaeology 49, 195-218.
DI : Diplomatarium Islandicum eða Íslenzkt
fornbréfasafn IXVI, Kaupmannahöfn/
Reykjavík 1853-1976.
Einarsson, Bjarni F. 1989, ‘Isländska vikingatida
gravar och grävda gardar. Fakta eller fiktion?’
Hikuin 15, 47-52, 232.
Einarsson, Bjarni F. 1994, The Settlement of Iceland;
A Critical Approach. Granastaðir and the
Ecological Heritage, (GOTARCH, Series B.
Gothenburg Archaeological
Gothenburg.
Eldjárn, Kristján 2000, Kuml og haugfé, 2nd ed. by
Adolf Friðriksson, Reykjavík: Mál og menning.
Friðriksson, Adolf 2009, ‘Social and symbolic
landscapesin Late
Archaeologia islandica 7, 9-21.
Kristiansen, Kristian ed. 1985, Archaeological
Formation Processes. The Representativity of
ArchaeologicalRemains
Prehistory, København: Nationalmuseet.
ÍF: Íslenzk fornrit-, Reykjavík 1933-.
Lárusdóttir, Birna 2007, ‘Settlement organization
and farm abandonment: The curious landscape
of Reykjahverfi, North-East Iceland.’ ed. W.
Davies, G. Halsall & A. Reynolds: People and
Space in the Middle Ages, 300-1300 (Studies in
the Early Middle Ages 15), Brepols, Turnhout,
45-63.
Lárusdóttir, Birna & Elín Ósk Hreiðarsdóttir 2011, ‘Í
þegjanda hljóði. Búsetulandslag frá miðöldum í
norðlenskri sveit.’ ed. O. Vésteinsson, G. Lucas,
K. Þórsdóttir & R.G. Gylfadóttir eds. Upp á
yfirborðið.Nýjar
fornleifafræði, Reykjavík: Fornleifastofnun
Íslands, 117-39.
Lárusson, Björn 1967, The Old Icelandic Land
Registers, Lund: Institute of Economic History
Lund University.
Manntal á íslandi árið 1703, Reykjavík: Hagstofa
Íslands 1924-1947.
Olsen, Björn M. 1907-15, ‘Um skattbændatal 1311
og manntal á Íslandi fram að þeim tíma.’ Safn til
sögu Íslands og íslenzkra bókmennta IV,
Reykjavík: Hið íslenzka bókmenntafélag,
295-384.
Rafnsson, Sveinbjörn
ThesesIV),
IronAge Iceland.’
fromDanish
rannsóknirí íslenskri
1990,
Byggðaleifarí
Hrafnkelsdal og á Brúardölum. Brot úr
byggðasögu Íslands, Reykjavík: Hið ísleska
fornleifafélag.
Stefán Ólafsson ed. 2008, Fornleifaskráning í
Kelduneshreppi
Fornleifastofnun Íslands.
Vésteinsson,Orried.
Investigations in Mývatnssveit, Reykjadalur and
Svartárkot 2010, Reykjavík: Fornleifastofnun
Íslands.
I-II, Reykjavík:
2011,
Archaeological
A NOTE ON THE REGIONAL DISTRIBUTION OF PAGAN BURIALS IN ICELAND
49
JOURNAL 2011_3.3.qxd:4issue.qxd 15/12/11 14:07 Page 49
Download full-text 