THE CHANGING BRONZE AGETHE FINNISH ANTIQUARIAN SOCIETY, ISKOS 20, 2011

Gods of High Places and Deep Romantic Chasms

Introductory remarks to a study of the landscape situation of Bronze Age sacrificial sites in the Lake Mälaren area

Martin Rundkvist, Vantvägen 9, SE-13344 Saltsjöbaden, Swedenmartin.rundkvist@gmail.com

This paper outlines work in progress with the Bronze Age sacrificial sites of the Lake Mälaren provinces in Sweden. The project’s goals are twofold: a) to understand the landscape rules behind the siting of deposits, and thereby b) to develop predictive model that would allow scholars to find undisturbed Bronze Age deposits without the aid of farmers, dredgers or ditch diggers.

After closer study of nine sites in Uppland and Södermanland provinces in the field and numerous ones in the archives, I have found that the Bronze Age people under study preferred to make sacrifices at wet, high, topographically dramatic and ancestral locations. There are finds from bogs and white-water river gorges, hilltops, a cave and a settlement-site that had once been important. In the rare dry-land deposit locations, eye-catching boulders were sought out.

Known sacrificial sites appear to prefer a location 1.2–1.5 km from settlement-indicating burnt mounds, rock art and the coeval seashore. This means that sacrificial sites are typically part of the same contiguous sightlined landscape room as the homes of the people who frequented them.

Keywords: Bronze Age, Sweden, Uppland, Södermanland, Lake Mälaren, deposits, sacrificial finds, landscape, ritual

But oh! that deep romantic chasm which slantedDown the green hill athwart a cedarn cover! A savage place! as holy and enchanted As e’er beneath a waning moon was haunted By woman wailing for her demon-lover!

Samuel T. Coleridge, from “Kubla Khan”

Introduction

It has long been known that Bronze Age people in North-west Europe preferred wet sites for their sacrifices: streams, rivers, lakes, fens, bogs, even inlets of the sea. This paper reflects the thin-king of a Late Iron Age scholar after a year and half’s study of this subject indoors and in the field. It turns out that be yond the baseline wet-land theme, Bronze Age sacrificial sites in the

159

GODS OF HIGH PLACES AND DEEP ROMANTIC CHASMS

Lake Mälaren area in Sweden are highly diver-se. And the people behind them had a strong predilection for landscape locations that a mo-dern visitor will find dramatic, even sublime.

This latter observation is of course some-what subjective. In the debate over intuitive methods in landscape studies, I side with Andrew Fleming (1999; 2006; 2007) rather than Christopher Tilley (1994; 2010). Personal “phenomenological” impressions are impos-sible to communicate clearly and have an en-tirely unknown relationship to what people ex-perienced during prehistory, and so they have no place in scientific discourse. But as Fleming and Tilley both agree, this is not to say that a landscape archaeo logist can stay indoors. In order to understand a landscape well enough to speak clearly about its characteristics and for-mulate testable hypotheses, an arch aeo logist must traverse it, preferably on foot.

Archaeological sites are commonly sorted into three main categories: settlements, burials and “deposits” (e.g. Malmer 2002). To these, the Bronze Age of southern Scandinavia adds abundant rock art sites and a far rarer class of hilltop sanctuaries. In the Lake Mälaren area, all except the sacrificial sites (or depositional sites, to use a studiedly non-committal term) and hilltop enclosures are readily identified in the field when well preserved and vegeta-tion permitting. While the spatial relationship between settlements, burials and rock art has long been rather well understood (Kjellén & Hyenstrand 1977; Damell 1985; Wigren 1987), the sacrifices are harder to tie into the wider landscape context of the society that produced them as they tend to avoid the other site categories.

All but one of the main categories of Bronze Age site around Lake Mälaren and the adjoining province of Östergötland have recently been the subjects of monographs: settlements (Ullén 1997; Borna-Ahl kvist 2002), burials (Victor 2002; Thedéen 2004), rock art (Hauptman Wahlgren 2002) and hill-top sanctuaries (Olausson 1995). The sacri-ficial sites form the exception. The bronzes themselves received solid treatments long ago

(Ekholm 1919; 1921; Baudou 1960; Oldeberg 1974–76), and since hardly any new finds have been forthcoming, scholars have not pursued that avenue of research further. Sonja Wigren (1987: 53–62) and Susanne Thedéen (2004: 68–82) have however published brief over-views for Södermanland province.

Scholarship is severely hampered by the facts that a) sacrificed objects are hardly ever found any more, and b) during the period when they were found, scholars were hardly ever involved in their retrieval. This is because Swedish Bronze Age sacrificial sites are not identifiable on the surface: most are after all in rivers, lakes and bogs, where few archae-ologists have been able to do any directed large-scale fieldwork. And the main era of wetland reclamation in Sweden ended before World War II (Runefelt 2008). This happened about the time when tractors replaced horses in agriculture, placing the farmer in front of the plough where he can no longer see what it turns out of the ground. Finally, Swedish law effectively prevents the growth of any signifi-cant metal-detector hobby (Rundkvist 2008). To my knowledge, the last time a multi-object non-grave bronze deposit surfaced in the Lake Mälaren provinces was in 1980, when the military cleared the Järvafältet firing range in southernmost Uppland of unexploded am-munition using mine detectors (SHM 31368). Furthermore, data coverage is poor. Digging or dredging in various landscape situations can be seen as a kind of experiment as to whether a Bronze Age deposit will be found. Yet we only have information about a small subset of the cases where something was found, and none about the innumerable experiments that have turned out negative.

My on-going research concerns Bronze Age sacrificial or depositional sites. That is, I aim to produce new knowledge on the land-scape scale: not on the artefact level or on the level of the province-wide distribution map, but on a scale of hundreds of meters, where you can see from one studied landscape feature to another and walk between them in an hour or two. Rather than treating the find context

160

MARTIN RUNDKVIST

as an attribute of each find, my current work views finds as attributes of the places under study. This means that I am primarily inter-ested in finds with a reasonably detailed spa-tial provenance, those that can be tied securely to a place. And I aim beyond the anecdotal, to identify regularities, Bronze Age rules of land-scape.

Ultimately, I envision a predictive model, being a set of analytical tools that would allow archaeo logists to go out into the landscape like homing missiles, as it were, and find Bronze Age sacrificial sites without the aid of farm-ers, peat cutters and dredging crews. Then we could learn what sort of materials and struc-tures those 19th century finders left on site when they selected the objects they handed in to the authorities. And we could get a solid palaeo-ecological background for sacrificial events. With such knowledge, we would be in a much better position to say how Bronze Age sacrifices were performed around that an-cient inlet of the Baltic sea that is now Lake Mälaren.

Ritual and Rationality

As Richard Bradley has argued at length (2005), the distinction between ritual and domestic behaviour is not very helpful when dealing with prehistoric societies. One may easily think that “ritual” equals “irrational” and thus “functionally inexplicable”. Conversely, “domestic” would then equal “functionalist”. But it must be remembered that it is impossible to be more rational than what your level of knowledge about the world allows. This has nothing to do with the once-fashionable epistemological relativism where there was talk of “different ways of knowing”. Simply put, in the pre-scientific era that makes up almost the entire history of human culture, people did not know very well what was real and not. It was extremely difficult for them to determine what sort of actions would produce reliable effects. Most likely, almost all prehistoric action was believed to be functional.

If everyone believes in the Lady in the Lake and atheism is unheard of, then it will ap-pear entirely rational to make sacrifices to her. In fact, doing so may produce solidly benefi-cial effects – not thanks to any divine interven-tion, but because it impresses the neighbours. This view coexists easily with some level of modern-style economic rationality where rare imported goods such as bronze would be un-usually valuable and prestigious and thus apt as sacrificial gifts. And conversely, it means that when we see evidence of people acting in mundane, sensible ways that we can easily ex-plain from a modern functionalist perspective, then we are probably not dealing with behav-iour that prehistoric people saw as belonging to any separate category of its own. If you re-ally believe in gods, then sacrificing to them looks as sensible and/or ritual as digging deep post-holes to keep your house from collaps-ing. With the exception of people clinging to old belief systems, every age acts upon its best available knowledge.

My own interpretation of why the depos-its were made and left in place is that all were certainly left for reasons that appeared rational to people at the time. But very few were left for reasons that make any functional sense to someone with a scientific world-view. A be-lief in the supernatural was clearly involved. We will most likely never know whether we would classify the fictional entities to which the sacrifices were directed as gods, demons, spirits or ancestors. But Tacitus tells us that people believed in gods in 1st century AD Northern Europe, and the Mediterranean writ-ten evidence for godly beliefs at the time of the Scandinavian Bronze Age is extensive indeed.

Note also that “irrational” should not be taken to mean “random”, particularly when we consider that rational behaviour depends on your level of knowledge about the world. Rituals, while irrational to someone with a scientific world-view, are in fact anything but random. It is part of the term’s definition that a ritual is structured, even scripted, and proceeds according to certain rules that allows it to be

161

GODS OF HIGH PLACES AND DEEP ROMANTIC CHASMS

repeated in a recognisable form that the par-ticipants and audience accept as traditional (cf. papers in Kyriakidis 2007). And for this rea-son, archaeologists should not dispense with the concept of ritual action. As I have argued above, almost all human action during prehis-tory was very likely perceived as rational at the time. But much of it is nevertheless likely to have been ritualised.

Retrievable and Irretrievable Deposits

Scholars have attempted to distinguish retrievable deposits, “hoards”, from irretrievable permanent deposits, “offerings” (see Berggren 2009; 2010 for overviews). The idea is that dry-land hoards are buried secretly and temporarily for mundane functionalist reasons, while wetland offerings are disposed of permanently to communicate with the gods and often for reasons of ostentatious display. While this dichotomy is an empirical reality (Levy 1982), it is doubtful if the two classes of find should really be seen as exponents of two different modes of thought when we are dealing with a pre-monetary prestige economy and a pre-scientific world-view (Karsten 1994: 30–31; Bradley 1998; 2005: 145–164). In other words: it is true that some of these finds could have been retrieved, and it is true that we often see different object types in those contexts than we do in bogs and rivers, but it is uncertain (and possibly untestable) whether the two classes of find were really deposited for very different reasons.

I have in fact yet to see a convincing ar-gument for why we should interpret a given retrievable prehistoric metal hoard as mun-dane from a modern perspective. Christoph Huth (2009) makes the valuable point that the metal deposits of the Late Bronze Age and the Viking Period are similar in most respects but have been interpreted very differently. But while Huth hints that he favours a mer-cantile interpretation for both classes of finds, I instead hold the opposite view. Few Viking Period hoards were buried for mundane rea-

sons and even fewer were allowed to remain underground for such reasons.

It is in any case important to inter-pret the reason that a hoard was assembled separately from the reason that it was buried in earth or sunk into water. I see no reason to question the idea that scrap metal hoards were collected for recasting. But keep in mind that the vast majority of the collected scrap did de-monstrably become recast, as seen from the alloy composition of Bronze Age metalwork. This means that the buried scrap-metal hoards that we know of are ones that received unusual treatment and were not allowed to walk the normal path of their kind. While scrap metal hoards by definition contain many fragmented objects, the pieces rarely add up to complete artefacts. Stuart Needham (2001: 288) argues that this may be due to a custom similar to that where an animal is sacrificed and only certain parts are burnt as offerings to a god. Perhaps the scrap metal hoards contain the gods’ share of a much larger collection of objects that were re-cast for renewed use. Joanna Brück (2001: 157) suggests that the dry-land deposits repre-sent metal given to the earth in return for goods taken from the earth, including grain. (Whether or not the earth was envisioned as a personi-fied deity here would be difficult to tell.) This may be so. Note, however, that in most parts of Northern Europe including Scandinavia, it cannot have been evident to people that metal had subterranean origins. Bronze came from the packs of traders, not from the earth like grain did.

Deposit Diversity

A scholar wishing to make general statements about all Bronze Age sacrificial deposits will soon realise that this is impossible because of their diversity. There are many kinds of deposit, and it is highly likely that they follow different landscape rules (cf. Bradley 2000: 53; Fontijn 2002). The Lake Mälaren area is not very rich in finds of this kind compared to e.g. southernmost Scandinavia, and so we cannot operate with too many categories. But

162

MARTIN RUNDKVIST

the following distinctions are in my opinion indispensable.

Wet vs. dry. I assume that wet deposition was intentionally permanent, which is part of what “sacrifice” means. Bronze, unlike stone, retains a visible indication of the environment where it has spent the past millennia. Under wet conditions, the yellow metal sheen of the objects survives intact or darkens partly to black or brown. On dry land, the alloy devel-ops a blanket patina of green verdigris that is sometimes beautifully smooth, in other cases crumbly and porous to the point where an ob-ject’s surface detail is lost. In my opinion, this distinction must have meant something im-portant. I am not however persuaded that the dry-land deposits were buried and allowed to remain buried for reasons that would make ra-tional sense to a modern observer (or, for that matter, to a homo economicus).

Single vs. multiple. As a rule, the finds that mark the sites under study are single objects. Multi-object deposits are rare and tend to con-tain unusual object types.

Chronology. The Swedish Bronze Age last-ed some twelve centuries. We must allow for change over this time span and make good use of the typo-chronology established by earlier research.

Functional and material categories. Weapons, ornaments, tools, metalworking de-bris; bronze, stone, bone and pottery: these cat-egories must be important as well.

Although society was in all likelihood per-meated by ritual, my project treats sacrificial sites as worthy of separate consideration and avoids closer study of settlements and burials. Therefore, one common find category must be disregarded: the stray single Bronze Age object without any sign of wet deposition or place-ment at an unusual dry site. This places most stone axes out of the investigation’s bounds, as they tend to be stray finds and by their na-ture cannot retain any information about their depositional environment.

Site Continuity vs. Continuity of Site Selection Criteria

David Fontijn (2002: 260) points out that repeated sacrifices in the same bog or river stretch over centuries presents a bit of a conundrum since the deposits would not have left any visible traces to attract subsequent groups of celebrants. He argues that the explanation is oral traditions about sacrificial sites: they may not have looked like much, but people told and re-told stories about what had once happened there. With Stijn Arnoldussen, Fontijn has later suggested another explanation that appears more likely given the long periods involved: the sacrificial traditions may not have conveyed specific memories of individual sites, but instead transmitted general landscape rules governing sacrifices (Arnoldussen & Fontijn 2006; cf. Fontijn 2007).

In this view, a person who sought an ap-propriate place to sacrifice bronzes might not know whether or not anyone had done so be-fore at the individual site, but would find that it fulfilled traditional ritual demands. The idea might not be “This is a known place where the Lady of the Lake has been contacted before”, but “This is the kind of place where She may be contacted”. Such a perspective might ex-plain the pattern Fontijn (2002: 260–263) sees in Limburg, where Bronze Age sacrifices are placed in strangely unspecific and rather ex-tensive zones in the landscape, not at discrete places. If the landscape rules of sacrifice that we seek here were not strongly determinant, then sacrifices would have tended to spread out. But as Fontijn points out (p. 275), the rule cannot have been as simple as “Any wet place will do”.

Recurrent Site Characteristics

In April and May of 2010 I visited nine sacrificial sites in Uppland and Södermanland provinces, selecting them by the criteria that I had to be able to ascertain their locations closely, the finds should preferably be rather

163

GODS OF HIGH PLACES AND DEEP ROMANTIC CHASMS

rich, and I favoured sites located within walking distance of each other. The winter had been long and cold with abundant snow, and so vegetation was still sparse and much plough soil remained open to field walking. This ensured the best possible conditions for landscape observation. A year later I directed plough-soil metal-detecting at three of the sites: Hjortsberga in Vårdinge, Täckhammarsbro and Lilla Härnevi (and fieldwalking at Täckhammarsbro), without making any relevant finds. I summarise some characteristics of the nine sites in table 1.

Commenting on these sites, one must of course first ask what, if anything, they are rep-resentative of. As discussed above, it is impos-sible to gauge how many single-object sacri-fices we have from the two provinces in ques-

tion: many of the de-contextualised dry-envi-ronment bronzes in the museum stores may have been found at settlements and in graves. Furthermore, some were most likely once part of multi-object deposits. Focusing on definite multi-object bronze deposits including accu-mulated assemblages, I know of 32 sites: 17 in Uppland and 15 in Södermanland. The six multi-object sites I have visited thus represent only 19% of the total. But if we look separately at the subset of multi-object deposits that can be pinpointed to an individual plot of land or better, that is, sites with a known landscape lo-cation that we might point to and visit, the ratio is closer to 29%.

As expected, the people under study here primarily sought out wet sites for their sacri-fices. The two hilltop deposits in Vårdinge par-

Fig. 1. Cave in Pukberget, Österunda parish, Uppland. A per. V–VI spearhead and a horse tooth have been found on rock shelves inside.

164

MARTIN RUNDKVIST

ish are particularly eloquent: they were put in high places, but in bogs up there (which con-tradicts Birgitta Johansen’s 1993 wet vs. high dichotomous model). Yet as the title of this pa-per suggests, it also emerges that Bronze Age people had a strong predilection for landscape locations that a modern visitor would find dra-matic, scenic, even grandiose or sublime.

In their-pre-regulation states, the riv-er-gorge sites of Hyndevadsfallet and Täckhammarsbro would have called to mind Coleridge’s deep romantic chasm. Each was a white-water site at the narrowest point down-stream from a major lake system (Böklin 1961: 27; Damell 1999; Zachrisson 2004). And they were deep as measured in time as well. Sacrifices had begun there in the Middle or Late Neolithic and continued at both sites up into the 17th century.

The Pukberget cave deposit is to my knowl-edge unique in the region, having been found inside a jumble of enormous stone blocks from a collapsed cliff side (fig. 1). A more sublime location is hard to come by: the site’s name means “Devil’s Hill”, and it has probably born it since the Middle Ages. (In 1946, Erik Floderus inevitably envisioned a link between the spearhead from the cave and Odin, though no sign has been found of any Iron Age activ-ity at the cave.) A similar association between spearheads and the interior of hills can be seen in a find from a crevice on Oxeberget Hill near Frändesta in Helgesta, Södermanland (SHM 21687).

Yet there is also a major find from a dry domestic site: Lilla Härnevi (Forsgren 2007; 2008). When interpreting this deposit’s land-scape location, however, we must keep in mind that it is extremely late in the Bronze Age, per. VI. Most of Lilla Härnevi hamlet’s many set-tlement-indicating burnt mounds were in all likelihood completed and abandoned centuries before the bronzes were buried. One has been excavated, its accumulation dated with radio-carbon to a short period about 900 cal BC, per. IV–V (Karlenby 1998: 27–28). The burnt mounds are still visually prominent today, the plough soil around the hamlet rich in fire-

cracked stone and quartz. It looks as if people returned to the ruins of a storied ancient settle-ment site and buried their last multi-period col-lection of mixed bronzes there, right about the time when society left the Bronze Age behind and moved on. At the time, the hoard site was about a kilometre from the receding seashore, which is farther than expected for an inhabited site of the time. The Hjortsberga torque de-posit in Vårdinge, while also quite near burnt mounds, has a different relationship to the set-tlement site, being above and beyond the shore zone where the burnt mounds and graves are. I discovered a cupmark boulder among them.

Note also that though my current work tar-gets dedicated sacrificial sites, there is consid-erable evidence for less ostentatious sacrifices at settlements as well (Borna-Ahl kvist 2002). A few are quite opulent. For example, a wet-patina finely decorated per. I sword pommel was found between a coeval house foundation and a cupmark boulder at Sommaränge skog in Viksta, Uppland (Forsman & Victor 2007). The pommel, however, is an exception. We certainly do not see anything like the Lilla Härnevi deposit even at major well-excavated settlements such as Hallunda in Botkyrka, Apalle in Övergran or Pryssgården in Östra Eneby (Jaanusson 1981; Ullén 1997; Borna-Ahl kvist 2002).

The Swedish sites discussed above dem-onstrate the attractiveness of wet, high, topo-graphically dramatic and ancestral locations to Bronze Age sacrifices. Let me finally point out another class of sacrificial site that, like the wet locations, is also well known from other periods than the Bronze Age. To my knowl-edge, at least five dry deposits from Uppland and Södermanland were found under or in contact with eye-catching boulders. Four of them are multi-object deposits, including the great hoard from near Spelvik church in Södermanland (SHM 813), and these are rare. It has been argued, irrefutably, that such a lo-cation with a prominent and durable marker would make it easy to retrieve the objects. Yet the fact remains that these deposits were never retrieved.

165

GODS OF HIGH PLACES AND DEEP ROMANTIC CHASMS

w

/d

Loca

tion

Con

tent

s, da

te

Dis

tanc

e fr

om b

urnt

m

ound

(k

m)

Dis

tanc

e fr

om

rock

art

(k

m)

Dis

tanc

e fr

om

coev

al

seas

hore

(k

m)

Ref

Lång

bro

in V

årdi

nge,

Sö

Wet

Sm

all b

og o

n to

p of

gl

acia

l esk

er

21 v

ario

us b

ronz

es,

1 ho

ne st

one;

sing

le

depo

sitio

n, p

er. V

I

0.9

1.6

1.4

SHM

267

4, 2

842

Hjo

rtsbe

rga

in V

årdi

nge,

Sö

Wet

Sm

all b

og o

n to

p of

m

orai

ne h

ill, 1

900

m

from

the

abov

e fin

d

1 W

ende

lring

to

rque

, per

. VI

0.3

0.3

0.4

Raä

59;

SH

M 1

3117

Täck

ham

mar

sbro

in B

ärbo

, Sö

Wet

R

iver

Nyk

öpin

gsån

in

a go

rge

whe

re it

exi

ts La

ke L

ångh

alse

n,

orig

inal

ly w

ith ra

pids

5 ax

es, 2

spea

rs, 1

sw

ord;

repe

ated

de

posit

ions

, per

. I,

II, IV

2.6

1.3

2.4

Raä

80,

85;

SH

M 2

273,

41

77, 2

2228

; Strä

ngnä

s 10

85, M

156

Kris

tineh

olm

in H

elgo

na, S

ö W

et

Riv

er N

yköp

ings

ån

1400

m d

owns

tream

fro

m th

e ab

ove

finds

1 ax

e, so

cket

ed

1.2

1.0

1.0

Raä

173

; Nyk

öpin

g 26

6

Hyn

deva

dsfa

llet i

n Es

kilst

una,

Sö

Wet

R

iver

Esk

ilstu

naån

in a

go

rge

orig

inal

ly

imm

edia

tely

bel

ow a

w

ater

fall

7 va

rious

bro

nzes

, re

peat

ed

depo

sitio

ns, p

er. I

, II,

IV, V

1.3

1.4

2.4

Raä

587

; SH

M 8

234:

15,

1367

1, Ö

rebr

o 36

08;

Dam

ell 1

999;

Zac

hris

son

2004

Käl

by in

Esk

ilstu

na, S

ö W

et

Smal

l bog

, 900

m fr

om

the

abov

e fin

d 2

axes

, pos

sibly

a

dagg

er, s

ingl

e de

posit

ion,

per

. II/

III

0.5

1.5

1.5

Raä

558

; SH

M 3

573,

67

59; B

eckm

an-T

hoor

20

02

Dom

ta v

ad in

Öste

rund

a, U

p W

et

Larg

e fe

n th

roug

h w

hich

floa

ts th

e N

ysät

rabä

cken

stre

am

2 be

lt cu

pola

s, 3

rings

, sin

gle

depo

sitio

n, p

er. V

7.0

2.1

1.8

Raä

83;

Upp

sala

569

0;

Arw

idss

on 1

939

Pukb

erge

t in

Öste

rund

a, U

p D

ry

Cav

e on

clif

f sid

e,

1300

m fr

om th

e ab

ove

find

1 sp

ear,

per.

V-V

I, 1

hors

e to

oth

6.2

2.5

3.2

Raä

62;

SH

M 2

3674

; Fl

oder

us 1

946

Lilla

Här

nevi

in H

ärne

vi, U

p D

ry

Out

skirt

s of m

ajor

BA

settl

emen

t C

. 50

vario

us

bron

zes,

1 le

athe

r ga

rmen

t, sin

gle

depo

sitio

n, p

er. V

I

<0.1

0.

4 1.

0 R

aä 6

9; S

HM

116

35,

1260

7:5,

161

20

Tab. 1. Nine sacrificial sites in Södermanland and Uppland. An exploratory sample.

166

MARTIN RUNDKVIST

Distances

Addressing the relationship between the nine studied sites and neighbouring rock art and settlement sites (cf. tab. 1) is complicated due to uncertain data coverage. Finding rock art is a time-consuming specialised task: any distribution map will always say something about where skilled surveyors have been in addition to about where the rock art is. Still, it is worth noting that the distance between sacrificial site and rock art correlates with the distance between sacrificial site and seashore. This suggests that either a) surveyors have preferred to look for rock art near the Bronze Age sea shore, or b) the data coverage for rock art is roughly even across the region. As for the settlement sites, indicated by burnt mounds, it appears that the surveyors of the Swedish sites and monuments register were highly skilled at identifying them when situated alone or in groups. But picking out a burnt mound or two at a Bronze Age settlement site that has become overlaid by one of the region’s ubiquitous Late Iron Age mound cemeteries is of course vastly harder.

Looking at our little data set, it is worth noting that the median distances from sacrificial site to burnt mound, rock art and seashore are almost the same: 1.2 km, 1.4 km and 1.5 km. This not because the three have any tendency to close co-location at the stud-ied sites. Instead I would suggest that when people selected a sacrificial site, it had to be part of the same contiguous sightlined land-scape room as their home, but not be located too close to the other three categories. To some extent this would happen automatically since neither burnt mounds, rock art nor the seashore can exist in the freshwater wetland locations favoured for sacrifice.

If with a larger database we can con-firm that known sacrificial sites do indeed pre-fer a location 1.2–1.5 km from burnt mounds, rock art and the coeval seashore, then the ques-tion we must ask is, to what extent does it work

the other way around? There are many wetland locations that fulfil those criteria. In most cases we have no idea what is hidden there. Would it really be necessary to drain and trial-trench tens or hundreds of wetlands to find an an-swer?

Conclusions

The work in progress I have outlined in this paper concerns landscape studies of Bronze Age sacrificial sites in the Lake Mälaren provinces of Sweden. The project’s goals are twofold: a) to understand the landscape rules behind the siting of deposits, and thereby b) to develop a predictive model that allows scholars to find undisturbed Bronze Age deposits without the aid of farmers, dredgers or ditch diggers.

At the moment of writing, I have looked at sites in Uppland and Södermanland provinces. (Västmanland and Närke are also on my agenda, though finds are more scarce there.) After closer study of nine sites in the field and numerous ones in the archives, I have found that the Bronze Age people under study preferred to make sacrifices at wet, high, topo-graphically dramatic and ancestral locations. There are finds from bogs and white-water riv-er gorges, hilltops, a cave and a settlement-site that had once been important. In the rare dry-land deposit locations, eye-catching boulders were sought out.

Known sacrificial sites appear to pre-fer a location 1.2–1.5 km from settlement-in-dicating burnt mounds, rock art and the coeval seashore. This means that sacrificial sites are typically part of the same contiguous sight-lined landscape room as the home settlement of the people who frequented them.

*

A shorter version in Swedish of this paper has been published as Rundkvist 2011.

167

GODS OF HIGH PLACES AND DEEP ROMANTIC CHASMS

References

Arnoldussen, S. & Fontijn, D. 2006. Towards familiar land-scapes? On the nature and origin of Middle Bronze Age landscapes in the Netherlands. Proceedings of the Prehistoric Society 72, pp. 289–317.

Arwidsson, G. 1939. Bronsåldersfyndet från Domta vad i Österunda socken. Upplands fornminnesförenings tidskrift 1939, pp. 65–80.

Baudou, E. 1960. Die regionale und chronologische Einteilung der jüngeren Bron zezeit im Nordischen Kreis. Stock holm.

Beckman-Thoor, K. 2002. Skogstorpsyxorna. En föreställ-ning tar sin början. Kulturell mångfald i Södermanland 1. (Ed. A. Åkerlund.) Nyköping. Pp. 45–51.

Berggren, Å. 2009. Offerbegreppet i arkeologin – tolk-ningar och perspektiv. Järnålderns rituella platser. (Ed. A. Carlie.) Halmstad. Pp. 33–49.

Berggren, Å. 2010. Med kärret som källa. Om begreppen offer och ritual i arkeologin. Lund.

Böklin, L. 1961. Täckhammar. Sörmlandsbygden 1961. Pp. 27–34.

Borna-Ahl kvist, H. 2002. Hällristarnas hem. Gårdsbebyggelse och struktur i Pryssgården under bronsålder. Stockholm.

Bradley, R. 1998. The passage of arms. An archaeologi-cal analysis of prehistoric hoards and votive deposits. 2nd ed. Oxford.

Bradley, R. 2000. An Archaeology of Natural Places. London.

Bradley, R. 2005. Ritual and Domestic Life in Prehistoric Europe. London & New York.

Brück, J. 2001. Body metaphors and technologies of transformation in the English Middle and Late Bronze Age. (Ed. J. Brück.) Bronze Age landscapes: tradition and transformation. Oxford. Pp. 149–160.

Damell, D. 1985. Bronsålder i Södermanland. Nyköping.Damell, D. 1999. Hyndevad. Från bergslag och bonde-

bygd 1999. Pp. 27–37.

Ekholm, G. 1919. Uppländska depåfynd från bronsåldern. Upplands fornminnesförenings tidskrift, Vol. 8, No 3: 197–202.

Ekholm, G. 1921. Studier i Upplands bebyggelsehistoria. II Bronsåldern. Uppsala.

Fleming, A. 1999. Phenomenology and the megaliths of Wales: a dreaming too far? Oxford Journal of Archaeology, Vol. 18: 119–125.

Fleming, A. 2006. Post-processual landscape archaeology: a critique. Cambridge Archaeological Journal, Vol. 16, No 3: 267–280.

Fleming, A. 2007. Don’t bin your boots! Landscapes, Vol. 8, No 1: 85–99.

Floderus, E. 1946. Pukeberget i Österunda. Uppland 1946, pp. 7–21.

Fontijn, D. 2002. Sacrificial landscapes. Cultural biogra-phies of persons, objects and “natural” places in the Bronze Age of the southern Netherlands, c. 2300–600 BC. Leiden.

Fontijn, D. 2007. The significance of ‘invisible’ places. World Archaeology, Vol. 39: 70–83.

Forsgren, M. 2007. Depåfyndet från Härnevi. D. 1, Föremålsförståelse och genusperspektiv med ut-gångspunkt från ett s.k. skrotfynd från yngre brons-ålder i Uppland. Photocopied BA thesis. Stockholm.

Forsgren, M. 2008. Depåfyndet från Härnevi. D. 2, Sammanhang och förståelse av en fragmente-rad bronsdepå i torrmark från yngre bronsålder i Uppland. Photocopied MA thesis. Stockholm.

Forsman, C. & Victor, H. 2007. Sommaränge skog. Begravningar, ritualer och bebyggelse från senneoli-tikum, bronsålder och folkvandringstid. Uppsala.

Hauptman Wahlgren, K. 2002. Bilder av betydelse. Hällristningar och bronsål derslandskap i nordöstra Östergötland. Lindome.

Huth, C. 2009. Ansichtssachen. Spätbronze- und wiking-erzeitliche Schatzfunde und ihre wissenschaftliche Deutung. Historia archaeologica. Festschrift für Heiko Steuer zum 70. Geburtstag. (Eds. S. Brather et al.) Berlin. Pp. 41–54.

Johansen, B. 1993. Skärvstenshögar och sörmländsk bronsålder. Arkeologi i Sverige, Vol. 2: 99–118.

Karlenby, L. 1998. Ett arkeologiskt återbesök i Lilla Härnevi : arkeologisk slutundersökning, Lilla Härnevi 1:5, RAÄ 35, Härnevi socken, Enköpings kommun, Uppland. Uppsala.

Karsten, P. 1994. Att kasta yxan i sjön. Lund.

Kjellén, E. & Hyenstrand, Å. 1977. Hällristningar och bronsålderssamhälle i sydvästra Uppland. Uppsala.

Kyriakidis, E. (ed.). 2007. The archaeology of ritual. Los Angeles.

Levy, J.E. 1982. Social and religious organization in Bronze Age Denmark. An analysis of ritual hoard finds. Oxford.

Malmer, M.P. 2002. The Neolithic of south Sweden. TRB, GRK, and STR. Stockholm.

Needham, S. 2001. When expediency broaches ritual in-tention: the flow of metal between systemic and bur-ied domains. Journal of the Royal Anthropological Institute incorporating Man, Vol. 7: 275–298.

Olausson, M. 1995. Det inneslutna rummet. Om kultiska hägnader, fornborgar och befästa gårdar i Uppland från 1300 f Kr till Kristi födelse. Stockholm.

Oldeberg, A. 1974–76. Die ältere Metall zeit in Schweden 1–2. Stockholm.

168

MARTIN RUNDKVIST

Raä. Site number by parish in the National Heritage Board’s sites-and-monuments register for Sweden.

Rundkvist, M. 2008. För en liberalisering av de svenska metallsökarreglerna. Fornvännen, Vol. 103: 118–122.

Rundkvist, M. 2011. I landskapet och mellan världarna. En inledande studie av bron s ålderns offerplatser i Mälarområdet. Bronsålder. Bronsålder i Stockholms län – aktuell forskning. Rap port från ett seminarium 2010. (Eds. K. Andersson et al.) Nacka.

Runefelt, L. (ed.). 2008. Svensk mosskultur. Odling, tor-vanvändning och landskapets förändring 1750–2000. Stockholm.

SHM. Inventory numbers in the National Historical Museum, Stockholm.

Thedéen, S. 2004. Gränser i livet – gränser i landskapet. Generationsrelationer och rituella praktiker i söder-manländska bronsålderslandskap. Stockholm.

Tilley, C. 1994. A phenomenology of landscape. Places, paths, and monuments. Oxford.

Tilley, C. 2010. Interpreting landscapes. Geologies, to-pographies, identities. Walnut Creek.

Ullén, I. 1997. Bronsåldersboplatsen vid Apalle i Uppland. Uppsala.

Victor, H. 2002. Med graven som granne. Om bronsål-derns kulthus. Uppsala.

Wigren, S. 1987. Sörmländsk bronsåldersbygd. En studie av tidiga centrumbildningar daterade med termolu-miniscens. Stockholm.

Zachrisson, T. 2004. Hyndevadsfallet och den kulturella mångfalden. Om depositioner i strömmande vatten i Södermanland. Kulturell mångfald i Södermanland 2. (Ed. A. Åkerlund). Nyköping. Pp. 18–33.