Smakkerup Huse: A Mesolithic Settlement in NW Zealand, Denmark

Author(s): T. Douglas Price, Anne Birgitte Gebauer, Signe Ulfeldt Hede, Charlotte Sedlacek
Larsen, Nanna Noe-Nygaard, Sarah L. R. Mason, Jens Nielsen, David Perry
Source: Journal of Field Archaeology, Vol. 28, No. 1/2 (Spring - Summer, 2001), pp. 45-67
Published by: Boston University
Stable URL: http://www.jstor.org/stable/3181459
Accessed: 14/12/2010 18:27

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=boston.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].

Boston University is collaborating with JSTOR to digitize, preserve and extend access to Journal of Field
Archaeology.

http://www.jstor.org
 45

Smakkerup Huse: A Mesolithic Settlement

in NWZealand, Denmark

T. Douglas Price
Anne Birgitte Gebauer
University of Wisconsin-Madison
Madison, WI

Signe Ulfeldt Hede

Charlotte Sedlacek Larsen
Nanna Noe-Nygaard
University of Copenhagen
Copenhagen, Denmark

Sarah L. R. Mason
University of London
London, Great Britain

Jens Nielsen
Kalundborg og Omegns Museum
Kalundborg, Denmark

David Perry
University of Connecticut
Storrs, CT

Excavationsat the site of SmakkerupHuse on the island of Zealand, Denmark, have

revealedLate MesolithicErteb0lledeposits,dating between5000 and 3900 B.C.Preservation
was excellentand a varietyof bone,antler,wood,and otherplant remainswererecoveredin
the waterloggeddeposits,along with stoneand ceramicartifacts.The remainsdocumenta
diversesubsistencebaseutilizing bothmarine and terrestrialfoods,an elaboratewoodtech-
nology,new artifacttypes,and someof the oldestdomesticcowsin Scandinavia.This report
providesbackgroundinformationon the site, and discussesthe artifactsand plant and ani-
mal remains.The questionof domesticanimals in the Late Mesolithicis addressed.

Introduction remains were recovered, along with many other interesting

The archaeologicalsite of SmakkerupHuse is located at artifacts. These materials were found in waterlogged de-
the headwaters of a former fjord, or inlet, now dammed, posits along a former coastline of the sea. The site, dating
known as the Saltbxk Vig on the NWcoast of the island of between 5000 and 3900 B.C. (calibrated), belongs to the
Zealand, Denmark (FIG. i). Excavations at the site were later Mesolithic, a period known as the Ertebolle.
undertaken as part of the Saltbak Vig Archaeological Pro- The site is important for its 1000-year record of cultur-
ject, a regional investigation of the introduction of agricul- al deposits and the preservationof abundant subsistence re-
ture into prehistoric southern Scandinavia (Gebauer and mains and wooden objects. In addition, SmakkerupHuse
Price 1990). Excavationstook place at SmakkerupHuse in documents some of the oldest domestic cattle in Denmark
1989 and from 1995 through 1997. Some 140 sq m were and a new artifact type-the painted pebble-from the
excavated, in places to a depth of more than 2 m. A total of Mesolithic. This preliminaryreport (a monograph is forth-
more than 225 kg of flint and other worked stone and 150 coming) includes background on the Mesolithic of south-
kg of bone and antler, wood, and nutshell and other plant ern Scandinavia, a history of research at the site, the geol-
46 SmakkerupHuse: A Mesolithic Settlement in Denmark/Price et al.

0 10okm

Figure 1. The location of SmakkerupHuse in NW Zealand, Denmark, and other sites mentioned
in the text.

ogy of the area, the topographic situation of the settle- Pleistocene, and ended with the onset of the farming cul-
ment, the layout and sequence of the excavations, stratig- tures of the Neolithic around 3900 B.C. (TABLE i). The pri-
raphy, the finds, dating, and interpretation. Among sever- mary period of concern for the present discussion extends
al issues that the research raises is the question of how to from 5400 B.C. to 3300 B.C., encompassing the Late
define the Neolithic, discussed in the conclusions. Mesolithic (known as the Ertebolle) and the Early Ne-
olithic (belonging to the Funnel Beaker period, or Tragt-
The Mesolithic of Southern Scandinavia
bager, or TRB) in this area. The Ertebolle is found pri-
The Mesolithic of Europe is the last phase of hunter- marily along the coastal areas of northern Germany, Den-
gatherer adaptations across most of the continent. Because mark, and in the provinces of Scania and Halland in Swe-
of a long history of research, substantial public interest, den. The last two phases of Ertebolle are most relevant to
and excellent conditions for preservation, this period is the site of Smakkerup Huse. The middle Ertebolle phase
perhaps best known in southern Scandinavia. The (4800-4300 B.C.) witnesses the appearance of crude,
Mesolithic of that region began around 8500 B.C. (all dates thick-walled pottery; small, ceramic lamps appear during
in calibratedradiocarbonyears), shortly afterthe end of the the later Ertebolle (4300-3900 B.C.).
 28, 2001 47
Journalof FieldArchaeology/Vol.

Table1. Relativechronologyof southernScandinavia. 1975) are further evidence of increased religious activity
Phases YearsB.C.
and exchange at this time.
Period Facies
Mesolithic
EarlyErteb0lle Trylleskove
5400 B.C.
4800 B.C History of Research
MiddleErtebolle Stationsvej 4300 B.C. The site of SmakkerupHuse was probably first discov-
LateErtebolle Alekistebro 3900 B.C.
ered when the Horve-Varslev railroad was built through
Neolithic
FunnelBeaker(TRB) EarlyNeolithic 3300 B.C this area between 1908 and 1916. Leveling the roadbed
FunnelBeaker(TRB) MiddleNeolithic 0
2800 B.C. across the inlet with earth from both sides of the valley ex-
PittedWareCulture
Single GraveCulture 2400 B.C. posed artifactsof bone and antlerthat were collected by the
Dagger Period 1700 B.C. landowner. The site was first reported to the Danish Na-
tional Museum in 1952, designated as Viskinge sb. 42 (lat-
The presence of extensive shell middens, a diversity of er as sb. 86), as a result of Therkel Mathiassen's (1959) sur-
extraction camps, the remains of a wide range of fish and vey of NWZealand. A visit made during this survey record-
mammals-including seals, dolphins, and whales-and ed a flake axe made from a pointed-butt polished flint axe,
"species-specific"trapping stations show the diversity of an antler axe, the neck of a greenstone axe with an hour-
the subsistence base during the Ertebolle. The importance glass-shaped hole, and four knapped blades, all attributed
of marine resources in the diet is furtherhighlighted by sta- generically to the Stone Age. Bjarne Larsen and Egon
ble carbon isotope values in human bone from this time Iversen, of the ArkaeologiskForening (Archaeological Soci-
(Tauber 1981; Noe-Nygaard 1988; Richards, Price, and ety) in the nearby town of Kalundborg, found a variety of
Koch, in press). The occurrence of cemeteries in Zealand flint and bone materials under a thick, heavily mixed layer
and southern Scania at this time complements a picture of of gravel. Lisbeth Pedersen, director of the Kalundborg og
more sedentaryresidence and suggests increased social and Omegns Museum, visited Smakkerup Huse in 1984 after
ritual complexity (Price 1985, 2000). Significant regional the current landowner, Karl Jensen, reported plowing up
variation in artifacttypes and styles begins in the later part the remains of charcoal and fire-cracked rock. Core and
of the Ertebolle; differences between eastern and western flake axes were collected along with flint debitage. Peder-
Denmark, and among smaller areas within Zealand, have sen and Iversen undertook the first professional excava-
been reported (Andersen 1981; Vang Petersen 1984; Jo- tions at the site in 1985 with five test pits excavated
hansson 1999). Exchange of exotic materials is indicated through layers of heavy gravel. The two tests on top of the
by a varietyof foreign objects in southern Scandinavia.Pot- beach ridge exposed only plow zone on top of gravel and
tery, ornaments made from the teeth of extinct animals, sand. Below the beach ridge, three test pits revealed stratig-
and the "shoe-last" axes of amphibolite, originating in a raphy consisting of the plow zone, a 1-1.5 m thick gravel
Danubian context in Silesia, Poland appearin the Ertebolle layer with heavy iron staining, followed by a grey sand lay-
of Denmark. Bone combs and t-shaped antler axes were er with artifacts. The fids included two core axes, three
copied from LBK farmersto the south. long thin blades, four cores, and 19 flakes. The flint was
The introduction of domesticates, polished stone axes, sharp and without damage from being rolled. Groundwa-
new burial practices, and a new form of pottery marks the ter allowed only a brief glimpse of the cultural layer be-
end of the Mesolithic and the beginning of the early Ne- neath the beach ridge deposits.
olithic (Gebauer 1995). Some coastal sites continued in
use during the earliest Neolithic for specialized hunting Geology and Topography
and fishing until ca. 3300 B.C. (Skaarup 1973; Madsen The local topography and history of deposition at
1987). Changes in settlement and subsistence are reflected Smakkerup Huse is essential to an understanding of the
in sites that are generally smaller in size and located more site and its context. The site is located at the head of the
inland than during the preceding Mesolithic (Madsen and Saltbzk Vig inlet, at the mouth of the Bregninge stream,
Jensen 1982). Stable carbon isotope data indicate a sub- halfway between the opening of the Saltbek inlet to the
stantial reduction in the consumption of marine foods Kattegat to the west and the hills that mark the east end of
(Tauber 1981; Richards, Price, and Koch in press). Offer- the streamvalley (FIG. 2). The sediments in the Saltbxk Vig
ings in bogs and large tombs present new social and reli- region result from four major processes: the deposition of
gious patterns in the earlyNeolithic TRB. Ceremonial sites ground moraine by Late Pleistocene ice sheets, the melting
in bogs (Koch 1998; Bennike and Ebbesen 1986) and sac- of glacial ice leaving deep layers of outwash sand and grav-
rificial deposits of flint axes, amber beads, and copper or- el, the transgression of the sea in the Holocene depositing
naments and axes (Becker 1948; Nielsen 1977; Randsborg marine sand and clay, and the subsequent accumulation of
 Huse:A MesolithicSettlementin Denmark/Priceet al.
48 Smakkerup

Figure 2. The location of SmakkerupHuse at the mouth of the Saltbak Vig. The dark shading shows
present water levels in the Kattegat and the SaltbaekVig. The waters of the SaltbaskVig are maintained
by pumping at -1.2 m below sea level. The lighter shading shows the sea level in the Vig during the late
Mesolithic. The first contour line outside the Saltbak Vig is at modern sea level and the second is at 2.5
m, the approximate sea level during the late Mesolithic. Subsequent contour lines are at 5 m intervals.

organic deposits of peat and organic mud (gytja). Marine been destroyed by wave action and erosion, deep deposits
sediments define the extent of the basin of the Saltbak Vig of waterlain materialwere preserved and became the major
itself, marking the stand of the sea during the Littorina focus of excavation. Excavations in this area continued
(Early Holocene sea) transgressions at a maximum of ap- down to in situ layers from the middle Ertebolle period.
proximately +2.5 m asl (Mertz 1924). Beach ridges were The excavations at SmakkerupHuse involved two weeks in
formed along the coast of the Saltbek Vig at that time. 1989 and more substantial seasons during the summers of
SmakkerupHuse is the name given to a small, higher area 1995-1997, consuming more than 400 person-days (FIG.
defined by the 2.5 m elevation contour, perhaps 1 sq km in 3). In 1989 a 45 m long exploratory trench was excavated
size. This areawould have been an island during periods of by machine from the beach ridge north down the face of
higher sea levels in the late Atlantic and early Subboreal be- the coastal barrierto determine the location and extent of
tween approximately 5000 and 3000 B.C. The Smakkerup the cultural layer and its relationship to past marine trans-
Huse site today is situated along that beach ridge on the gressions. The southern 37 m of the trench contained heav-
south side of the Bregninge A streamvalley near the mouth ily rolled sand and gravel deposits. In the northern, lower
of the SaltbakVig (FIG.2). Directly across the valley on the part of the trench was found a deep layer of gytja contain-
opposite bank of the Bregninge stream lies the late Erte- ing material from the Ertebolle period. This, the same lay-
b0lle settlement of Engelsborg. er found by Pedersen in 1985, contained fresh, black flint
artifactsand an Ertebolle potsherd, along with bones from
The Excavations several different species of animals.
Excavations at Smakkerup Huse were undertaken to Excavations in 1995 began with a machine to remove
look for evidence of an occupation horizon on the coast the railroad bed and plow zone from two areas along the
and to expose the materials preserved in the waterlain de- former beach ridge (FIG.3). In the smaller unit, only three
posits next to the settlement. While the settlement areahad test pits were excavated. In the larger excavation unit, the
 Journalof FieldArchaeology/Vol.
28, 2001 49

3
**- '" +
. -.

'
- - ?- ........ .
-- ..

o N
. +4005
0,

-
1* " "'" " ' ~
."
.t.. .--
.
/
/

'-*---**-*-'........-.. //'*----*-^^^^^ ' . . ........

' ............. '

1989l ..... . . . . . _ 199
LI

""
-
,u / - ?"---- -- -----
-

+ +
I+ 530W
+ +
~0oW
...+
490W
+44oS

...
. .....
............ A

,~~~~~~~~~~-
\H - 7

Figure 3. Plan of the excavations at SmakkerupHuse. The location of the composite section shown in
Figure 5 is marked on this plan with a light gray line (A) in the large excavation units from 1996 and
1997. Dotted lines mark removal of the plow zone by machine while solid lines mark actual excavation
areas. Contour lines mark the moder surface of the area at 20 cm intervals up to the line of the railroad
cut. The grid marks are at 10 m intervals.

floor was cleaned, disturbances drawn, and excavation posits. Excavations were concentrated in the large area to
commenced in square meter units and by naturallayers. In the east almost directly under the former railroad bed and
some areas where the natural level was thick, arbitrary10 just offshore in the waterlain deposits. Excavations in 1997
cm units were removed within the layer. Culturalmaterials were opened to the south of the 1996 unit in hopes of un-
were recorded by layer and level. Sediments were water- covering the beach zone of the settlement and to look for
screened though 4 mm sieves, and sediment samples for an occupation horizon on the land adjacent to the sea. In
flotation were collected in levels with cultural materials. addition, three paralleltrenches were excavatedto the west
One long trench was dug along the upper part of the beach to expose the stratigraphyin this area and to look for more
ridge and was used for water screening; a second, to the cultural deposits.
east and south of the eastern excavation unit, exposed
stratigraphyand facilitated drainage of the excavation area. Stratigraphy
Investigations in 1996 also began with a machine to SmakkerupHuse, and much of the coastline of the Salt-
clear the upper sediments from a large area to the east of bak Vig, was exposed to the sea and subject to marine ero-
the 1995 work. In addition, a trench to the west was made sion during the higher sea levels of the mid-Holocene. It is
from the edge of this new area to follow the coastline de- particularlyfortunate that a small point or projection of the
 Huse:A MesolithicSettlementin Denmark/Priceet al.
50 Smakkerup

coast at the site preserved some in situ deposits along its

lee, or eastern, side. This projection was also the spot used +1 m
for the railroad bed across the Bregninge A valley, taking
advantageof naturaltopography. The terrainat Smakkerup
Huse slopes slightly from south to north toward the stream
valley, and to a lesser degree from west to east. Deposits to
the south are closer to the beach with more stone and re- 1m
deposited artifacts.The layers to the north and partiallyto
the east formed in deeper water and contain more gytja and
in situ artifacts.The slope from the south toward the north
is more pronounced under the railroad bed, at the beach
ridge, and becomes more level to the east.
It is important to remember that the excavatedlayers at '3 m

Smakkerupwere deposited in the water adjacentto the set-

tement. Waterlainsediments vary in texture and composi-
tion depending on their position along the coast and their
exposure to wave action. Thus, a single deposit may change
from organic mud to fine sand to gravel depending upon Years BC 5ooo 4000
the depth and energy of the water along the shore. These Atlantic | Subboreal
deposits are largely natural accumulations with the occa- Kongemose Ertebolle Early Neolithic
sional introduction of artifactsor waste as either "outcast" early I middle I late

(materialseither lost or tossed into the water as rubbish) or

Figure 4. The rise in sea level at Halskov, western Sjaelland, in the mid
redeposited materials eroded from the shore. Many of the Holocene (after Christensen 1997).
coastal sites of the Mesolithic of southern Scandinavia ex-
hibit a common pattern of two major zones of deposition
of cultural material (Andersen 1983). The actual occupa- at Smakkerup Huse, then, is a record of these changes in
tion or living areaof the site contains primarilysmall items the inlet during the middle and later Holocene. The com-
and preservation is usually poor. The second, outcast or posite section from the excavations in Figure 5 shows the
"dump" zone is in the water next to the settlement where major layers at the site and the sequence of deposits is sum-
rubbish was tossed and accumulated in the mud (Andersen marized in Table 2.
1993a, 1995). Items here are generally larger than on the Material for the railroad bed comprises the uppermost
settlement; bone, and occasionally wood, preservation can layer, beneath which is a remnant of the plow zone. Layer
be very good. During the Littorina transgression and re- 3 is a laminated horizon of sand and clay that was deposit-
gression at Smakkerup Huse erosion and redeposition of ed in water with changing current velocity. A thick deposit
materials from the settlement area itself took place. of iron-stained coarse sand, Layer 4, marks a substantial
The stratigraphyat SmakkerupHuse is complex because transgression/regressionevent. This layer contained a mix-
of the sequence of marine transgressionsand regressions in ture of cultural material, often rolled and patinated, that
the middle Holocene. The melting of the Weichselian ice- had been redeposited here. Layer 5 is a complex erosion
cap resulted in rising sea levels continuing through the horizon of grey brown sand and gytja becoming more or-
Holocene (Christensen 1993, 1995). The sea finally ganic toward the coast and is likely the product of a trans-
reached and slightly exceeded modern levels around 4000 gression episode. Layer 6 in this horizon was a practically
B.C. This period, around the time of the transition to agri- sterile brown gytja deposited in deeper water conditions
culture, was marked by a series of transgressions and re- perhaps reflecting the onset of the Subboreal transgression.
gressions as the sea gradually adjusted to modern levels. Layer 7 was deposited as a grey sandy ridge during a pe-
Four episodes of transgression and regression have been riod of decreasing water level, likely a regression episode.
found in western Zealand (Christensen 1997), and three Several facies of this layer could be distinguished as it
are recorded at SmakkerupHuse. The first transgression at changed from more gravel in the upper half (7G) toward
Smakkerup Huse takes place during the High Atlantic more fine-grained sand in the lower portion. The density
around 5000 B.C. (FIG. 4). A second episode dates to the of shell and shell fragments in Layer 7 increased away from
Atlantic/Subboreal transition around 3900 B.C., and the the coast in both the upper and lower sections. This layer,
third to the Subboreal around 3600 B.C.The stratigraphy rich in cultural material, yielded two bones of domestic
 28, 2001 51
Journalof FieldArchaeology/Vol.

Table2. The majordepositsat Smakkerup

Huse alongwith probableassociatedeventsin
the stratigraphic radiocarbon
sequence, dates,and associatedculturalmaterial.
Layer Description YearsB.C.(cal.) Chronology Archaeology
1 Railroad bed - Recent Historic
2 Plow zone - Recent Stray Finds
3 Sand and clay lenses - Tidal water? Sterile
4 Coarse orange sand/gravel - Subboreal regression Erteb0lle/early Neolithic
5 Grey brown sand/gytja - Subboreal regression Late Ertebolle
6 Brown gytja - Subboreal transgression Essentially Sterile
7G Grey sand with gravel 3920-3930 Atlantic/Subboreal regression Late Ertebolle
7 Grey sand/gytja 4220-4090 Atlantic/Subboreal regression Late Ertebolle
8 Brown sandy gytja 4990-4940 High Atlantic transgression Middle Ertebolle
9 Grey sand 5060 High Atlantic transgression
10 Grey sand with gravel - High Atlantic transgression
11 Brown peat 5430 Freshwatermarsh Essentially Sterile
12 Sand/gravel - Pleistocene outwash channel Sterile

cow 14C dated to ca. 3900 B.C. (AAR-3316 and AAR- Extending from the first "shoreline"zone to the north,
3317) a very early Subboreal date. A dog skeleton found in we encountered a number of large flat stone slabs that had
the upper part of this layer was dated to ca. 4100 B.C. originally been placed in the water on the slope of the coast
(AAR-3318). These radiocarbon determinations are dis- as stepping stones or pavement (FIG. 6). There were small-
cussed in more detail in the subsequent section on dating. er stones between the largerslabs in places closer to the for-
Because this layer is the result of regression, it is possible mer shoreline. Similar stone pavements at the shoreline
that at least some of the cultural material therein was sec- have been observed at the sites of Tybrind Vig (Andersen
ondarily deposited. This shallow water deposit (Layer 7) at 1985) and Vanget Nord (Price and Petersen 1987). This
SmakkerupHuse likely corresponds to a similarepisode re- zone also contained a large pair of red deer antlers intact
ported elsewhere in Denmark. Investigations of the oscil- with the top of the cranium. In the northern, deeper por-
lations of the Littorina Sea have reported a period of low- tion of the excavation we encountered several pieces of
er sea level around 3900 B.C. at the time of the transition dugout canoes that must have been abandoned in the wa-
between the Atlantic and Subboreal (Christensen 1997: ter. This northern area,perhaps at the edge of a coastal reed
53). At Smakkerup Huse this transition appears as a dis- bed, was a boat landing area. In this section Layers 9 and
tinct sand deposit between the Atlantic and Subboreal gyt- 10 are not present.
ja layers. The beach deposits of Layer 7 at Smakkerup sit Beneath this intact outcast layer is a deep deposit of es-
on top of layers of Atlantic gytja and underlie a later gytja sentially sterile, brown peat (Layer 11) that formed during
that must relate to a later transgression and thus a higher freshwater marsh conditions in the Saltbak Vig. A radio-
sea level. carbon date from the upper part of this freshwater peat
Layer 8 is an intact deposit of brown sandy gytja con- provided an age of 5430 B.C. (AAR-3783). This peat is
taining more shell toward the bottom of the layer and away more than 1.5 m thick in this part of the site and likely sits
from the coast as well as a largely in situ outcast deposit atop the outwash sand and gravels of the Late glacial peri-
from the middle Ertebolle period. The projectile point od, when the Saltbxk Vig was a major outlet for meltwa-
types and radiocarbon dates on hazel nutshell (4940 B.C., ters from the ice decaying in central Zealand. This layer of
WG-2573) and a bone awl (4990 B.C., WG-2574) confirm outwash sand and gravel (Layer 12) can be seen at the base
this age. This layer is rich in organic materials and con- of the south end of the section where layers were deposit-
tained a great deal of worked wood. Severalzones could be ed directly on the coast. On top of this basal deposit, Lay-
distinguished in this layer. One zone lay along the former er 10 is grey sand and gravel marking the truncation of the
coastline where many of the bone points from the excava- freshwater peat by rising water levels. An oak tree in this
tion were found as well as a number of vertical wooden
layer, likely killed by the onset of the transgression, was ra-
stakes; this may have been a fishing areafor both spearfish- diocarbon-dated to 5060 B.C. (AAR-3782).
ing and weirs. A second zone to the east contained three
large slabs of bark and several pieces of the tree fungus de- Finds
scribed below. In addition, numerous chips from wood- The Late Mesolithic Ertebolle culture of southern Scan-
working, as observed at other Mesolithic sites (Andersen dinavia is characterized by an elaborate blade technology
1995), were present. with distinctive projectile points and core axes. Ground
52 Smakkerup
Huse:A MesolithicSettlementin Denmark/Priceet al.

stone appears as axes, adzes, and other tools. Projectile 502W

420S
weapons had specialized tips of bone, wood, and antler, as
well as of stone. A wide range of fishing gear, including
+i.o a; -
nets, weirs, spears, leisters (spears with multiple prongs),
hooks, and harpoons is known from this period, along j^-S
9'l,9 42 tA' 12 ;:
-5Z ll X X
/' / / / / ?
with dugout canoes and paddles. Pottery appeared after Drain 10
4700 B.C. in the Ertebolle period and takes several forms
including pointed base cooking vessels, small oval bowls +0.0 asl _
that likely served as oil lamps, and small cups (Andersen
1991; Gebauer 1995). llPj Stone
Stone tools define the chronological phases of the old-
er, middle, and younger Ertebolle (Vang Petersen 1984). Wood

Transverse points with oblique angles and soft-hammer 1.0 asl -

blades characterize the Trylleskov, the oldest of the three
Erteb0lle phases on Zealand. Flint points become more
Figure 5. Composite stratigraphyat SmakkerupHuse showing major
regular and symmetrical through the Ertebolle with core strata. See also Table 2.
axes predominant in the Trylleskovphase, replaced by sur-
face retouched flake axes in the Stationsvej phase. Core ax-
es are elaborated with specialized retouched edges in the These points are typical for the middle and late Ertebolle.
Alekistebro phase of the late Ertebolle. Blade knives with The leading and trailing edges of the point are the sharp
transverseretouch are common. Burins, scrapers,and bor- edges of the blade and the sides are retouched. The leading
ers vary proportionately in the different phases. edge is usually broader than the trailing edge and flares out
slightly. Through time the points become more symmetri-
Lithics cal and parallel-sided. More than half of the points were
The vast majority of lithic remains at Smakkerup Huse found in Layers 4 and 5 where they must have been rede-
were pieces of flaked stone but examples of ground and posited from the settlement area. There is a tendency for
polished stone artifactswere also found. Fire-crackedrocks the transversepoints in these upper layers to be more par-
were common and their presence is simply noted here allel-sided as is expected for later Ertebolle materials (Vang
Petersen 1984).
along with the unmodified, flat stone slabs used as step-
The flaked stone axes at SmakkerupHuse are fairly typ-
ping-stones from the shoreline into deeper water (FIG. 6).
ical for the late Mesolithic. Forty-nine flaked flint axes were
FLAKED STONE recovered in the excavations. Nineteen of these are core ax-
Almost 25,000 pieces of flaked flint were recovered es and the rest are flake axes. All of the axes are generally
(TABLE 3), and the assemblage was typical of
the Ertebolle small in size. Core axes are poorly made and often difficult
of a few distinctively Neolithic ar- to distinguish from flake axes. Nine of the core axes were
period with the addition and
tifacts in Layer 4. Flakes and blades are the most abundant fragments, with an equal number of bit and butt ends
artifactsand this pattern is reflected in the number of flake one indeterminable.Three pointed core tools (core borers)
and blade cores as well. Hard-hammer blades are some- were found and are not included here with axes. Six of the
what more common at Smakkerup Huse although soft- core axes exhibit specialized edge trimming that is distinc-
hammer blades dominate in Layer 8. Burned flint pieces, tive of the middle and late Ertebolle. The specialized axes
constitute were more or less equally distributed through the stratig-
exposed to fires in the settlement area, usually surface-re-
between 8 and 15% of the total flint in a layer. raphy. Of the 30 flake axes, the symmetrical,
The proportion of retouched tools in Mesolithic assem- touched axes were the dominant type (12 examples); 5
flake axes were atypical; 4 were of the asymmetrical, sur-
blages in southern Scandinavia is low (Johansson 1999;
Stafford 1999) and our site proved no exception. Re- face-trimmed type; and there was one bit end fragment. In
touched pieces are listed in Table 3 in the categories of re- addition two flake chisels were found. The symmetrical,
touched flakes and blades, points, axes, scrapers, drills, surface trimmed flake axes are the best and most well
notched pieces, and burins. The most common categories shaped flake axes, but even these appear irregular, small,
of the retouched tools are points and axes. Projectile points and of poor quality.
and their preforms are abundant; 365 were found at Table 3 also shows the distribution of flaked stone tools
in the
Smakkerup Huse, almost 60% of the retouched
tools. by layer at SmakkerupHuse. There are no artifacts
 28, 2001 53
Journalof FieldArchaeology/Vol.

50lW
410S

railroad bed material (Layer 1) and relatively few in the known from the middle and late Mesolithic of southern
plow zone (Layer 2) at the site. There are two major zones Scandinavia. In Denmark they are more common in the
of artifacts, each reflecting marine transgression; one in west. Several other worked pieces of stone include an oval
Layers 4 and 5 and a second in Layers 7G, 7, and 8. Layer slab of fine-grained sandstone 126 x 94 x 17 mm with
4 is the coarse sand horizon containing secondarily de- small depressions or cups on one surface and the sides. An-
posited materials largely eroded from the coastline during other was a polygonally shaped slab of sandstone 129 x
the last regression. The large number of projectile points in 114 x 20 mm and the single polished surface was striated
this layer is almost certainly a result of this secondary de- with several lines.
position from the settlement. Layer 5 is a sand and gytja Several fragments of Neolithic polished flint axes are
lens that is likely a mix of late Ertebolle material both out- known from Smakkerup Huse, including two specimens
cast and eroded from the shore. Layer 6, an otherwise ster- found during earlier surveys. Two small fragments of Ne-
ile brown gytja, contained only a few, likely introduced, ar- olithic polished flint axes were found in our excavations,
tifacts. Layer 7/7G of grey sand and gytja is the richest both in Layer4. No other definite Neolithic materialswere
horizon at the site with a large number of artifactsfrom the encountered in the excavations, suggesting at most an
late Ertebolle Alekistebro phase (TABLE i). Layer 8 is an in- ephemeral presence by post-Mesolithic peoples.
tact deposit of brown gytja with artifacts primarily from
the middle, Stationsvej phase of the Ertebolle (TABLEI). Ceramics
There is a clear increase in flake axes from the lower to the While pottery is well-known from the second half of the
upper levels at Smakkerup Huse. Only 3 flake axes were Ertebolle period, after ca. 4700 B.C., only nine sherds were
found in the lowest culturallayer (8) and 13 flake axes were found in our work, largely in the upper layers (5, 7G, and
found in the upper Ertebolle horizon (Layer 7/7G), while 7). The differences between Ertebolle and Early Neolithic
the remainderwere found in the erosion layers above (Lay- utilitarian ceramics are not pronounced (Koch 1998), but
ers 4 and 5). Conversely, core axes are the most common the latter have a flat or rounded base and thinner walls,
axe type in the lowest Ertebolle horizon (n= 11) and few- along with finer temper and a different coiling technique.
er were found in the upper Ertebolle horizon (n= 3). Eight The identification of Ertebolle vs. Early Neolithic pottery
came from the higher erosion horizons (Layers 4 and 5) was a concern at SmakkerupHuse because of the presence
where all levels are mixed together. of the domesticated cows in the same layers. One sherd
came from a lamp of the middle to late Ertebolle, with soot
GROUND OR POLISHED STONE
residue on the inside rim radiocarbon-datedto 4500-4440
The most significant ground stone artifactwas one-half B.C. (WG-2575). A small
cup clearly belongs to the
of a greenstone axe (trind0xe). The bit-end fragment was youngest Ertebolle a form that only appears, for ex-
phase,
105 mm long and 45 mm in diameter.The trindoxe, typi- ample, in the level at the Ertebolle site of Ringk-
upper
cally a cylinder of hard, heavy stone shaped by pecking, is loster (Andersen 1995). An extroverted neck sherd is like-
 28, 2001 55
JournalofField Archaeology/Vol.

Table 3. The number of artifactsper category of flaked stone tools, ordered here in standard
European typology, by layer at SmakkerupHuse. The total number of retouched tools used
to calculate the last column, percent of tools, was 620. Soft blades = soft-hammer production;
hard blades = hard-hammerproduction.
Layer 1 2 3 4 5 6 7G 7 8 9 10 11 12 Totals %Tools
Flakes 0 25 0 3588 4114 43 2171 5703 1643 18 10 51 0 17,366
Soft blades 0 2 0 272 262 1 191 309 143 0 0 4 0 1184 -
Hardblades 0 2 0 195 211 4 159 271 168 1 0 7 0 1018
Totalblades 0 6 0 679 678 8 451 800 412 2 1 13 0 3050
Burnedpieces 0 1 0 444 636 3 582 938 449 40 28 17 0 3138
Ax sharp 0 0 0 7 8 0 9 8 5 0 0 0 0 37 -
Flakecore 0 1 0 35 44 1 23 48 74 2 1 1 0 230 -
Bladecore 0 0 0 4 7 0 6 14 16 0 0 1 0 48 -
Microblades 0 0 0 0 0 0 0 4 2 0 0 0 0 6 -
Retouchflakes 0 0 0 7 11 0 10 15 9 0 0 0 0 52 8.4
Retouchedblades 0 0 0 24 24 0 16 17 22 0 0 0 0 103 16.6
Point 0 2 0 99 76 0 38 87 21 0 0 1 0 324 52.3
Pointpreform 0 0 0 19 6 0 4 10 2 0 0 0 0 41 6.6
Flakeaxe 0 0 0 9 5 0 5 8 3 0 0 0 0 30 4.8
Coreaxe 0 0 0 4 1 0 0 2 6 0 0 0 0 13 2.1
Specializedcoreaxe 0 0 0 2 0 0 0 1 3 0 0 0 0 6 1.0
Polishaxefragment 0 0 2 0 0 0 0 0 0 0 0 0 2 0.3
Scraper 0 0 0 1 3 0 0 1 2 0 0 0 7 1.1
Bladescraper 0 0 0 1 2 0 3 0 1 0 0 0 0 7 1.1
Largeflakedrill 0 0 0 1 1 0 0 4 0 0 0 1 0 7 1.1
Bladedrill 0 0 0 3 5 0 0 4 2 0 0 0 0 14 2.3
Smallflakedrill 0 0 0 1 1 0 0 0 0 0 0 0 0 2 0.3
Notchedblade 0 0 0 0 0 0 0 1 0 0 0 0 0 1 0.2
Burin 0 0 0 2 0 0 2 4 3 0 0 0 0 11 1.8
Burinspall 0 0 0 7 4 0 0 3 0 0 0 1 0 15 -
Microburin 0 0 0 6 2 0 0 2 1 0 0 0 0 11
Totals 0 39 0 5412 6101 60 3670 8253 2986 65 40 97 0 26,723

number of examples have been reported (Andersen 1983; these pieces had been worked, some just cut away from the
Andersen 1987; Christensen 1990; Gron and Skaarup tree and others sharpened to a point; in some cases the bark
1991; Hartz 1999). A complete large dugout from the site had been removed (FIG. 7). There are three types of cuts
ofTybrind Vig in Denmark (Andersen 1987) measures al- seen on the base of the stakes. One is a single facet from
most 10 m in length and 0.65 m in breadth. The sides of one axe blow with the remainder of the limb snapped off
this canoe are 25 cm high and 1-2 cm in thickness, the bow from the tree. A second type of cutting resulted in a chisel
is a rounded point and the stern is a board inserted at the or v-shaped point on the stake created by two or three
back end of the boat. Near the stern of this canoe there is blows. The third type of base is a neatly rounded point like-
an oval fire platform, similar to the one we have, made
ly resulting from sharpening of the stake after it was re-
from clay and small stones 3 cm thick and 70 x 40 cm in moved from the tree.
size.
Analysis of 32 stakes indicated that all were hazel (Cory-
A fragment of an archerybow was found positioned ver- lus avellana L.), varying in age from 5 to 12 years at the
tically, likely recycled as a wooden stake. Made of elm, one time of cutting, with a preference toward seven year-old
end had been delicately carvedto a point. The piece was 19 branches (FIG. 8). The mean diameter of the stakes is 2.8
cm long, 4.5 cm wide and 2.1 cm thick at the break, with cm with many fragmentes more than a meter long and
a rectangularcross section with oval corners. Similar bows quite straight, often exhibiting the swollen base typical of
are known from several other sites,
including Ringkloster coppiced trees (Rackham 1980; Rasmussen 1990). Hazel
and TybrindVig (Andersen 1985, 1995).
normally does not growth in long, straight sections unless
More than 50 worked wooden stakes of
varying length coppiced. Growth rings showed that 22 of the stakes were
and diameter were found both standing
verticallyand lying harvested at the end of the summer growing season. One
horizontally in the deposits. Such stakes were likely fences branch five years of age was harvested early in the growing
and fishweirs, common at coastal Mesolithic sites in south- season (spring), as was another 10 years old. Harvesting of
ern Scandinavia (Andersen 1995; Pedersen 1997). All of
preferred 7-year-old branches may have been concentrated
56 Smakkerup
Huse:A MesolithicSettlementin Denmark/Priceet al.

tant component of Mesolithic diet and could be stored for

162 months (Kosina 1991).
Both acorn shell and cupule (acorn cap) fragments were
present in wet-sieved samples from Smakkerup, but they
cannot be tied to human activity as they are mostly un-
charred and many are immature. They likely indicate that
oak was growing near the site, but charred acorn shells
were present in very low numbers, raising the possibility of
8
human consumption. It is difficult to assess the importance
of acorns in the diet, however, since the nutshell is very thin
and does not preserve well. Remains of buds and small
4- sterile or immature whole acorns (barely, if at all, exserted
from their cupules) were also present but, as with similar
remains at other Ertebolle sites (e.g., Bokeberg III in Reg-
nell et al. 1995; Mollegabet II in Gron and Skaarup 1991;
5 6 7 8 9101 12
Mason, Hather, and Hillman in press b; Halsskov in
Year of Cutting Robinson and Harild in press; Tybrind Vig in Kubiak-
Martens 1999), their interpretation is problematic. They
Figure 8. Histogram of hazel stake ages at cutting.
may have entered the site with twigs, carried in for some
purpose or they may indicate the presence of nearby, over-
during one time of year, with more opportunistic cutting hanging trees. Quercus wood charcoal is present in some
in other seasons. It is interesting that the main harvest samples, indicating its likely use as fuel.
would have taken place at the end of the fishing season, Crataegus(hawthorn) fruits were recovered via wet siev-
perhaps to cure the wood over the winter before its use in ing, usually in the form of the pyrenes (stones), though oc-
a weir. casional whole fruits are present, sometimes somewhat im-
All of the wood remains included in this analysis docu- mature. Where identifiable to species these appearto be of
ment a long-term practiceof coppicing hazel trees and pur- C. monogynaJacq. Since Crataegusfruits are edible, these
poseful management. Since the maximum nut production may be food remains. They are a frequent find at the Erte-
of a hazel tree does not occur until around fifteen years of b0lle site of Mollegabet II (Gron and Skaarup 1991; Ma-
age, this evidence suggests one goal of husbandry was us- son, Hather, and Hillman in press b) and Ringkloster (An-
able wood in addition to nuts. dersen 1998). Crataegusfruits in Scandinaviatoday are col-
lected in the fall afterthe first frost when the fruit is less bit-
Plant Remains ter.
Most plant remains were hand picked from wet-sieved Other specimens present in the wet-sieved samples are
materials, including an enormous amount of nutshell. One largely unidentifiable, mostly consisting of fragments of
additional archaeobotanical sample consisted of the 1-2 petiole or peduncle (leaf or fruit stalks). It is likely that they
mm fraction of one 10-liter bulk sample. Both sample types derive from one or other of the taxa alreadyidentified. One
are thus highly biased. The archaeobotanicalassemblage is charredfragment of plant tissue was examined under SEM
dominated by hazel, oak, and hawthorn. The wet screen- and identified as parenchyma,possibly deriving from roots
ing also recovered large pieces of fungus and a small frag- or tubers. Such finds are increasingly common on Euro-
ment of wickerwork described at the end of this section. pean Paleolithic and Mesolithic sites (Holden, Hather, and
Thousands of fragments of hazelnut shells (Corylusavel- Watson 1995; Kubiak-Martens 1996, 1999; Mason,
lana L.), very few whole but a number of them burnt, were Hather, and Hillman in press a; Perry1997, 1999, in press;
collected during water sieving at Smakkerup Huse. The Robinson and Harild, in press; Zapata et al., in press), ow-
shells were concentrated in Layers 5, 7, 8, and 11 with the ing to improved techniques of recovery. Other plant re-
highest number in Layer 7. Hazelnut shells are abundant mains, from the 1-2 mm fraction of one 10 liter bulk sam-
at other coastal Mesolithic sites (e.g., Tybrind Vig and ple, included both charred and waterlogged plant remains.
Bjornsholm [Andersen 1985, 1993b]). Special locations The bulk of charred remains were wood and stem or rhi-
for harvesting and roasting enormous quantities of nuts are zome of Equisetum (horsetail) with a small number of
known in northern Germany (e.g., Bokelmann 1991). All charred fragments of Tilia bark. There were several frag-
evidence points to the fact that hazel nuts were an impor- ments of possible parenchyma and cf. Gramineae stem (re-
 28, 2001 57
Journalof FieldArchaeology/Vol.

sembling Phragmites)and one seed each ofJuncus, Galium, While the plant species is unknown, the materialis not bast
and Gramineae. Mats made of Tilia bast are known from (beaten bark), but perhaps heather (P. H. Mikkelsen and
other Ertebolle sites, such as Tybrind Vig, and Phragmites H. Strehle, personal communication 2000) or willow
stems would provide useful thatching material. The pres- (Myrhoj 1997). Heather likely would not have survived
ence of Equisetumis harder to explain, but it is interesting well in marine environment (L. Pedersen, personal com-
that it has been noted also at Mesolithic sites in the Nether- munication 2000).
lands, where it was the most ubiquitous class of vegetative
remains (Perry 1999). Animal Remains
The bulk sample also yielded fragments of wood, bark The faunal remains from 1989 were identified by Tine
of
(notably Tilia), stem, leaf of both dicots and monocots, Trolle-Lassen;the non-fish bones from the excavations in
roots, Bryophytes, bud scales (many of Quercus type), as the 1990s were studied by Hede (1999). The fauna in-
well as fungal sclerotia, small snails, fragmented oyster cludes 16 species of mammals, 5 of birds, and 1 amphibian
shell, fish bones, foraminifera, and insect remains. There (TABLE 4). No human bones were recovered. A total of
were also a large number of uncharred seeds, with abun- 1773 bones, representing at least 70 individual animals,
dant Chenopodiaceae (largely cf. Chenopodiumalbum, also could be identified as to species from the approximately
Suaedamaritima, among other species) and Ruppia mariti- 22,500 fragments that were examined. In general the bone
ma. Other species present in smallernumbers included Ur- material is well preserved but heavily fragmented from
tica, Rubus, Rumex, and cf. Malus. The presence of abun- butchering, the extraction of marrow, and the manufacture
dant seeds ofRuppia, a submerged maritime plant of salt- of bone tools.
water pools, and considerable numbers of foraminifera, The remains of forest game animals are most common
which live largely in sub- or inter-tidal zones reflects the among the mammals and include red and roe deer and wild
waterlain deposition at SmakkerupHuse. Many of the re- boar (FIG. 9). Several fur-bearing taxa are present, includ-
mains found are of ruderal species (characteristicof dis- ing marten, beaver,otter, dog, and wildcat, with similar as-
turbed habitats, in some cases especially maritime habitats, semblages known from other Mesolithic sites: Agernams
and also producers of large numbers of seeds). Most of (Richter and Noe-Nygaard in press), Ringkloster, and Ty-
these species have edible seeds, greens, or fruits, and many brind Vig (Andersen 1985, 1995). While fish (discussed
have been recovered from other Ertebolle sites (Kubiak- below) dominate the marine fauna, sea mammals are rep-
Martens 1999; Regnell et al. 1995). resented by two species of seals (Halichoerusgrypus and
Some of the more unusual finds at the site were pieces Phocahispidae).Less than 1% of the total number of bone
of tree fungus or polypores (Fomesor Polyporusfomentar- fragments could be identified as bird; such a small show-
ius), known in Danish asfrsvamp or tondersvamp(fire fun- ing is common at late Mesolithic sites owing to the fragili-
gus or tinder fungus). This fungus grows commonly today ty ofthin-walled bird bones. Small mammals, no doubt un-
as a thick shelf, white-brown in color, on birch and beech
derrepresented here because of the size of the screen, in-
trees (Mason and Hather in press). Over ten pieces more clude squirrels, hedgehogs, and two kinds of voles. While
than 10 cm in diameter were found here and many pieces the voles may occur naturally (cf. Noe-Nygaard 1995),
were also found at the submerged site ofTybrind Vig of the
squirrels and hedgehogs had economic value (Aaris-
west coast of the Danish island of Funen (Lange 1994). Sorensen and Andreasen 1993).
Fyrsvampis well known as a tinder or punk for use in start- Two domesticated species were present, dog and cow.
ing fires. Its use as tinder is mentioned in the Icelandic Bones from several individual dogs were found, including
sagas and the fungus was included in the fire-making kit of one partially articulatedskeleton. Cutmarks on the last in-
the Iceman (Chapela and Lizon 1993; Poder, Puimpel,and dicated that it was skinned before the carcasswas thrown
Peintner 1994). One thin, flat piece of fyrsvamp had the in the water. Stable carbon isotope analysis of the dog col-
texture of leather, but an amino acid comparison showed
lagen produced a negative value of -10.2 parts per thou-
this object not to be skin. The piece resembles felt; sand (ppt). What is striking about this value-the lowest
perhaps
there was some cloth-like use for the material.
reported for a dog in southern Scandinavia-is that it sug-
A wicker fragment, probably a fish trap,
perhaps a bas- gests an almost exclusively marine diet for this animal. Sta-
ket or a bag, was approximately 35 x 15 mm and consists ble isotope ratios in dogs have been used as a
of six cross-strandswoven with two pairs of thinner, braid- proxy for hu-
man diet in Mesolithic studies (Noe-Nygaard 1988,
ed strands. Each pair of strands is twisted so that it crosses
1995). Three bone fragments were identified as the upper
over between each of the cross-strands. This pattern is
joint of the radius and the lower joint of a metacarpus and
known from fish traps at other Danish Mesolithic sites. a metatarsus from domestic cattle. Identification is
58 Smakkerup
Huse:A MesolithicSettlementin Denmark/Priceet al.

Table 4. Species of animal bones identified from excavations at ppt) compared to modern individuals and to prehistoric
SmakkerupHuse. roe deer at SmakkerupHuse (n= 3, mean = 21.76 ppt) are
Minimum
Number of interpreted as a result of increasing forestation of the area
Talon Scientificname Commonname Elements around SmakkerupHuse. This conclusion matches the re-
Mammal Cervuselaphus Red deer 732 sults from investigations in the nearby Amose region
Roe deer 664
Capreoluscapreolus
Susscrofa Wild boar 270 (Noe-Nygaard 1995). The red deer sought most of its food
Bos domesticus Domestic cow 4 in the shaded forest vegetation, while the roe deer browsed
Domestic dog 39 in the more exposed vegetation of the forest rim along the
Canisfamiliaris
Vulpesvulpes Fox 1 coast. The stable carbon isotope values of the wild boar
Felissilvestris Wildcat 10
Martes martes Marten 4 (n= 3, mean = -20.08 ppt) are slightly less negative than
the deer, possibly the result of feeding on vegetation along
Lutra lutra Otter 17
Halichoerusgrypus Grey seal 3 the beach.
Phocahispidae Ring seal 2 The vast majority of the faunal remains (96%) come
Castorfiber Beaver 4
from Layers 5, 7G, 7, and 8, with a pronounced abundance
Sciurusvulgaris Red squirrel 1
Water vole 8 in Layer 7. Animals equally distributed in all layers are the
Arvicolaterrestris
Clethrionomysglareolus? Bank vole? 8 most common species including red and roe deer, pig, dog,
6
Erinaceuseuropaeus Hedgehog otter, and water vole. There is little change in the propor-
Amphibian Bufo sp. Toad 1 tions of red deer, roe deer, and wild boar over time; red is
Bird Anas sp. Duck 1 38-50%, roe is 25-43%, and boar is 10-16% in the four
Swan 3
Cygnussp.
Turpusmenula Blackbird 2 layers. Dog is represented in all four layers and particular-
Tetraourogallus Capercaillie 3 ly in Layer 7G with the articulated skeleton. The cow
Haliaeetus albicilla White-tailed eagle 2 bones are found in the higher Layers 5, 7G, and 7, a cir-
Fish Gadusmorhua Cod 805 cumstance that likely reflects the rough contemporaneity of
Merlangius merlangus Whiting 2
2 these layers and the presence of secondarily deposited ma-
Clupeaharengus Herring
Platichthys Flounder 1708 terials. The fur-bearing species are more common in the
Belonebelone Garfish 620 lower layers, but we note that there are many more identi-
Scomberscombrus Mackerel 81 fied individuals in those layers. While there seems to be
Myoxocephalus scorpius Bullhead 5
50 more variety in Layer 8, given the relatively small sample
Squalesacanthias Spiny dogfish
Eel 157 size, there is little substantive observable change over time
Anguilla anguilla
Rutilus rutilus Roach 16 in the faunal remains at SmakkerupHuse.
Bream 2
Abramis brama
2
The bone treatment and fragmentation pattern at
Luciopercalucioperca Zander
Esoxlucius Pike 101 SmakkerupHuse is consistent with those for prey animals
at other late Mesolithic sites in southern Scandinavia(Noe-
Nygaard 1977, 1988, 1995; Trolle-Lassen 1987, 1992).
Few differences could be seen in these patterns among the
straightforward based on the size and thickness of the
bone. Although it is sometimes difficult to distinguish do- three major species-red deer, roe deer, and wild boar-at
mesticated cattle from Bos taurus, the aurochs (Degerbol the site, with bones reduced to fragments via marrow frac-
less
and Fredskild 1970; Grigson 1974; Rasmussen and turing or tool production. Juvenile animal bones are
Richter 1991; Noe-Nygaard 1995), that animal had been fractured than adult bones. Red deer (Cervuselaphus) did
extirpated on the island of Zealand at least 1000 years pri-
show some difference in the representation of elements;
or to the occupation of SmakkerupHuse (Aaris-Sorensen solid bones with little meat, such as phalanges, and the dis-
a tal ends of limb bones are present but the proximal end is
1980). The cow bones are from adult animals; there is
the is terminat- often missing, likely because of scavenging dogs, who left
complete fusion of the diaphysis, epiphysis
ed and the suture is obliterated. Radiocarbon dates from tooth marks on bone fragments from a range of species.
indicates
two of the cow bones and the articulateddog skeleton are Analysis of bone, teeth, and antler fragments
discussed below in the section on dating. Stable carbon val- that the site was definitely occupied in the summer, fall,
ues for the domestic cow bone collagen (n=2, mean = and winter, and likely year-round (FIG. io). In common
-21.5 ppt) seem to indicate that they were feeding along with Mesolithic sites throughout southern Scandinavia,
the forest rim or in more open grass areas, compared to there are few indicators of spring in the fauna. There areju-
other species. veniles, no more than a year old, of both red deer and wild
The '13C values from red deer (n=4, mean = -22.87 boar that must have been born in the spring or early sum-
 Journalof FieldArchaeology/Vol.
28, 2001 59

and mackerel (Scomberscombrus),and thus may be season-

al indicators. Zander (Luciopercalucioperca)is an indicator
of warmer climate during the Atlantic episode; its present
habitat is in the south of France. The majority of the fish,
such as the small cod and flatfish, are taken in shallow wa-
ters during the summer. The evidence for the importance
of fishing at Smakkerup Huse consists of the fish bone,
oyster shell, the placement of the site near the mouth of the
fjord, and fishing equipment, such as bone fish hooks and
awls. Recall also the dugout canoes and the bones of grey
and ring seals. Large oyster shells, some burnt, were pre-
sent in the cultural layers, clearly collected nearby and eat-
Figure 9. Proportion of estimated minimum number of individuals en at the site. Many other species of shells were also found
(EMNI) by major species at SmakkerupHuse. at the site, but with the exception of oysters and perhaps
the blue mussels are largely naturallypresent in the site.
mer. Antler development in red deer and roe deer indicates
summer and winter occupation. Fully developed red deer BoneandAntlerArtifacts
antler attached to the skull reflects a period from Septem- A number of worked bone artifactswere found includ-
ber through February;roe deer carrydeveloped antler such ing a bird bone awl, a possible needle, 2 perforated tooth
as found at SmakkerupHuse in the summer months, May pendants, a worked tooth, 55 awls (Danish: prene), 6
through October. Fur-bearing animals such as otter and points, 6 fish hooks, 1 possible hair pin, 1 dogfish spine,
beaver are usually hunted in the winter when their pelts are and several pieces of worked antler including fragments of
in prime condition. Age estimates from the mandibles of two retouchoirs and an axe. In addition, one complete rack
wild boar show that they were taken during summer, fall, of red deer antler and a large piece of a single antler, both
and winter. with cut marks, were excavated.Another rackof antler had
been found nearby during the construction of the railroad.
Fish and Shellfish Some of these items are briefly described below to provide
The fish remains comprise 11,140 bones, of which a sense of their form and variety.
9332 were identified as to species or family by Larsen Bone spikes (Danish: prene), asymmetricalthough very
(2000). Nearly all were collected by water sieving with 4 uniform in size and shape, were made from the proximal
mm screens. Thirteen species, including three flatfish, are ends of roe deer metatarsals. The discarded distal ends
represented (TABLE 4). The majority of the bones come make up the principal presence of roe deer metatarsals in
from either small species (e.g., roach, Rutilus rutilus) or the faunal remains at SmakkerupHuse. Although awl-like
small individuals from species that have large adults (e.g., in shape, these objects were used as a spike for a fishing
cod, Gadus morhua). Small fish are often caught uninten- spear or, less likely, a leister, since no leister prongs were
tionally in fish traps as secondary catch (Enghoff 1994). Of found. There were 54 found in a distinct zone parallel to
the identified fish bones, approximately 70% came from the coastline a few meters from the water's edge. The arti-
cod, 18% from various flatfish, 7% from gar, and the re- facts must have accumulated there as a result of spearfish-
maining 5% were other species. The vast majority were ing from the shore. The average length of the 32 intact
saltwaterspecies, including the spiny dogfish, whiting, gar,
specimens is 82 mm and the average weight is 3.9 g.
and mackerel. Freshwater species include zander, roach, Two intact fishhooks and four fragmentary specimens
bream, and pike. were recovered, all with a distinctive shape and carvedfrom
The fish species at SmakkerupHuse are coastal dwellers, bone. The shank of the hook has a small knob of bone for
some living out their lives along the coast in in-shore envi-
attaching the line; the curve of the hook is thicker than the
ronments of the eel grass zone, such as the bullhead shank and ends in a barbless point. Five of the hooks were
(My-
oxocephalusscorpius).Gar (Belone belone) and dogfish are small, ca. 15 mm in length by 7 mm wide, appropriatefor
found around the coast in the summer while
feeding on small fish and eels. The hooks were cut from a thin, flat
smaller fish. None of the identified species are
solely deep- piece of bone between 1.4-1.7 mm thick, perhaps a scapu-
water dwellers. The fish represented are found la. Bone points, however, were made by
primarily grinding and pol-
during the warmer half of the year. Some of the species are ishing a dense piece of bone into a round, narrow, sym-
only found in the summer today in Denmark, such as gar metrical shaft with a sharply pointed end. A total of six
60 SmakkerupHuse: A Mesolithic Settlement in Denmark/Price et al.

Month/Indicator J F M A M J J A O N D

Reddeerantlerdevelopment

Reddeerjuvenile
ontogeny

Roedeerantlerdevelopment

Wildboartootheruption

primefur
Otter/beaver
I I t I I I I

Figure 10. Seasonal indicators at Smakkerup Huse.

were found at SmakkerupHuse of which three were intact. cisions are likely decoration. A third tooth was heavily
A little larger than bone spikes, intact bone points averaged worn, with the root ground on two sides to a point. Its
90 mm in length with a diameter of 7 mm. function is unknown.
The awl of bird bone is an artifact not usually seen on
Zealand; in Jutland, where it is more common, it is nor-
OtherFinds
mally made of swan bone. The specimen from Smakkerup A small amber pendant (ca. 15 x 10 mm) of inverted-T
Huse is bird bone (species indeterminate but not swan), shape, with a groove for stringing near the top of the neck,
polished to a beveled, hollow point on one end and broken was recovered in the excavations. Amber is found occa-
on the other. The preservedlength is 118 mm, the bone di- sionally at Mesolithic sites in the Baltic region, often in the
ameter is 6 mm, and the point is beveled over 39 mm. A form of jewelry. The most unusual find from Smakkerup
pin of bone from Smakkerup closely resembles an object Huse was an oval, white quartzite pebble painted with
reported from atop the head of a female burial from Ved- black bands and three dots (FIG. II). The bands range from
bak (Petersen 1990). Our specimen is 172 mm long and 4-12 mm in width, two running around the pebble, on its
approximately 7 mm in diameter, polished to straighten short axis. A wide line on either side of the pebble connects
and create a point at one end. One of the eight dogfish the two bands. Another wide line runs from front to back
spines showed clear evidence of use as a fine piercing tool over the wider end of the oval pebble, connecting one of
with its point slightly rounded and reduced from use, per- the tranverse bands. Two circular dots, approximately 10
haps from working skin or leather. Similar specimens are mm in diameter, were placed on the corners of the wide
known from other Mesolithic sites (Larsen 2000; Noe- end of the pebble and a third at the narrow end. The char-
Nygaard 1971, 1995). coal-based paint has fragments of charcoal visible under a
Two fragmentaryantler axes, one of indeterminate type, microscope. There are no comparable examples from Scan-
were found at Smakkerup Huse. One is a small piece ap- dinavia and the closest parallelsknown are the painted peb-
proximately 4.5 cm in length and 2 cm wide preserving bles of the Azilian, ca. 4000 years earlier,of southern Ger-
perhaps one-third of the shaft hole. Two joining fragments many and France. In fact several illustrated examples of
are from a typical Ertebolle type, made on the main beam Azilian pebbles closely resemble the pattern seen on our
of red deer antler with the shaft hole near the burr,or coro- piece but are found on smaller, flat pebbles.
net, of the antler (Andersen 1995).
Tooth pendants, most commonly of red deer, are known Dating
from Mesolithic residential sites in southern Scandinavia The artifactsat SmakkerupHuse are late Mesolithic, of
and are also well known from burials (Larsson 1989, the middle and late Ertebolle period, specifically its Sta-
1990; Petersen 1990; Nielsen and Petersen 1993). Such tionsvej and Alekistebro phases, between 4800 and 3900
root B.C. There are only a few items from later time periods,
pendants are either perforatednear the tip of the tooth
or incised with thin lines around the root, presumably to such as the polished axe fragments and a projectile point of
facilitate binding. Two examples, both of red deer, were re- the Pitted Ware Culture found on the surface. Eight radio-
covered at Smakkerup Huse. A canine tooth is perforated carbon dates are listed in Table 5. Five dates were obtained
at the tip of the root with an hourglass-shaped hole ca. 2 from the AMS laboratory at Aarhus University (AAR).
mm in diameter. The second, an incisor, is unusual in that Three of these were samples of fauna and two were pieces
it is both perforated and incised with pairs of lines across of wood from geological contexts. Three experimental
the root between the perforation and the enamel; these in- dates were also obtained from a NEC accelerator mass
 28, 2001 61
JournalofFieldArchaeology/Vol.

:?

:. .: . : . t.

Figure 11. Five sides of the painted pebble. The object is 7.7 cm by 5.3 cm by 3.5 cm (drawing by Kurt
Petersen).

spectrometer in Middleton, Wisconsin, in January 1998 drowning of the tree by the High Atlantic transgression in
(WG). No stable carbon measurements were made on this area and corresponds very closely with the latest die-
these three samples and no corrections for marine reservoir off of oak trees in the Halskov fjord 35 km to the south
effect have been made for any of these dates. Reported cal- (Christensen 1997). Another determination was made on
ibrations are highest probability values at ? 1 s.d. (Stuiver a small fragment of wood from the peat of Layer 11 at an
et al. 1998). elevation of -0.23 cm asl to obtain an age for the freshwa-
The dates for the two bones of domesticated cow found ter marsh. While there were rising sea levels earlier in the
in Layers 7G and 7 are virtually identical. The boundary Atlantic period, the sea had not, at this time, penetrated the
between these two layers was vague in the field and these Saltbek Vig.
layers may have formed at about the same time. The two
cow bones are among the earliest in southern Scandinavia Discussion
and date from the transition to agriculturein this area.The The excavations at SmakkerupHuse document the na-
Neolithic in western Sjalland begins around 3950 B.C. ture of coastal settlement during the late Mesolithic in
(Fischer in press) but the absence of other Neolithic mate- southern Scandinavia. Although the actual living area has
rials in Layer 7G/7 at SmakkerupHuse suggests that these disappeared,deposits in the water next to the site preserved
cattle come from a very late Mesolithic context. The do- a great deal of information regarding the environment, set-
mesticated dog bone (see above, "Animal Remains") in tlement and subsistence, technology, and other aspects of
Layer 7G is slightly older than the two cow bones and per- human adaptation in the middle Holocene. The stratigra-
haps a bit too old. phy provides a history of environmental development in
Three dates were obtained on materialsfrom Layer 8, of this areafrom the meltwater drainage of decaying ice at the
which those from a hazel nutshell and a bone point arevery end of the Pleistocene to the freshwater marsh and river
similar. The third date came from soot residue on a sherd regime of the Bregninge A in the early mid-Holocene. The
of a ceramiclamp, perhaps from burning oil extractedfrom continuing rise of sea level brought the sea into the river
marine mammals, and is several hundred years younger drainage, creating the fjord at around 5000 B.C., when
(4466 B.C.) and is thus problematic. The remaining two SmakkerupHuse was first occupied. Subsequently a series
measurements were obtained from natural features at the of marine transgressions and regressions over the next mil-
site. There were a number of stumps and trunks of oak lennium created the rich deposits that are described here
trees in Layer 9. One stump at 0.33 m asl yielded a radio- while the settlement area itself was erased by coastal ero-
carbon date (5082-4959 B.C.). This date likely marks the sion. Specific information on the size and nature of the set-
62 SmakkerupHuse: A Mesolithic Settlement in Denmark/Price et al.

Table5. RadiocarbondatesfromSmakkerupHuse.
Labno. Material Layer 3/12C 4CB.P YearsB.C.cal.
AAR-3316 Cow bone 7G -21.4 5059?68 3945-3789
AAR-3318 Dog bone 7G -10.2 5319?65 4174-4044
AAR-3317 Cow bone 7 -21.6 5060?61 3885-3795
WG 2573 Hazelnut shell 8 - 6060 ?60 5004-4902
WG 2574 Bone point 8 - 6100 60 5069-4935
WG2575 Lamp residue 8 - 5650 70 4547-4442
AAR-3782 Oak wood stump 9 -25.4 6140 60 5082-4959
AAR-3783 Wood fragment 11 -28.3 6495 ?60 5412-5371
in peat

tlement, and the numbers of dwellings and individuals at fjord, ratherthan deeper waters. The importance of marine
Smakkerup Huse, will never be known. Nevertheless we foods is reflected in the wide variety of fishing equipment.
know that the site was occupied over a relativelylong span Plant foods are dominated by the ubiquitous presence of
throughout much of the 5th millennium B.C., but whether hazelnuts, which could have been stored for a number of
this occupation was continuous or recurrent is uncertain. months and may have been an important staple in the diet
Elsewhere, settlements in the later Mesolithic were often (Zvelebil 1995). Acorns were present and probably eaten
long term, frequently containing material from both the but the evidence is not yet sufficient. There is evidence for
Stationsvej and Alekistebro phases of the Ertebolle (e.g., various fruits including the white hawthorn and charredre-
Andersen 1993a; Larsson 1997). mains suggest that roots were also part of the diet.
Smakkerup Huse, near the juncture of the freshwater The technology of the late Mesolithic emphasizes
stream and the salt water inlet, was well situated with re- wood, bone, and stone. Because of preservation in water-
spect to a great variety of freshwater and marine resources. logged sediments, parts of the wood and bone technology
The settlement must have been situated directly on the survive to complement stone tools. Hazel stakes, the most
coast of the Saltbxk Vig. Such coastal location is typical for common of the wood assemblage, functioned in fishing
the later Mesolithic (Andersen 1993a, 1993b) and provid- fences or weirs. What is probably a fish trap of woven
ed immediate access to the inlet, good spawning ground stalks, and the elm bow add to our knowledge of Mesolith-
for cod and flatfish, with terrestrialresources nearby.There ic lifeways. Several pieces of dugout canoes were found,
is evidence of year-round settlement at SmakkerupHuse in and numerous wood chips may be in part a product of their
the plant and animal remains. The terrestrialfauna points manufacture. The fyrvamp fungus, likely used as tinder, is
to residence during summer, fall, and winter, and the ma- common at the site. The fyrsvamp "felt" is reported here
jority of fish remains contain smaller fish and species that for the first time. Horsetail, phragmites reeds, and lime
are common along the coast in the summer time. Oysters, tree bark probably represent raw materials; the reeds may
on the other hand, are best in the winter and spring before have been used for thatch and the lime bark for flooring.
they spawn and are not eaten much today in the summer Tools of bone and antler, including carefilly crafted bone
and fall. Leafy plants and seeds were likely collected in the fish hooks, were used for fishing, hunting, and other pur-
summer and hazelnuts and white hawthorn fruit ripen in poses. Antlers were employed as hammers and flint-knap-
the fall. Growth rings in the hazel stakes indicate that some ping tools. Most of the stone tools were of good quality
cutting was done in the spring, completing the annual cy- flint available in the region either on beaches or in the
cle at the site. Ecologically, coppiced woodlands require a ground moraine. Flint blades are the single most important
good bit of year-round attention as they attract deer and component of the lithic assemblage. Many of these are un-
other herbivores that, unless discouraged, will eat the retouched and often not counted as tools in traditional ty-
young shoots. Overall, the argument is strong that the site pology. Nevertheless, these simple blades were the major
was occupied throughout the year. stone tool of the Mesolithic; many types such as scrapers,
The terrestrialfauna is dominated by the three primary borers, and burins are uncommon and clearly not a major
game animals of the later Mesolithic-red deer, roe deer, part of the Mesolithic assemblage. There is in fact a re-
and wild boar-along with several fur-bearing species- markablerarityof stone tools other than blades, projectile
marten, fox, beaver, otter, wild cat, and domestic dog. points, and axes, but a large quantity of debitage suggests
Aquatic remains include a variety of salt- and freshwater that many tools may have been made and used elsewhere.
fish, along with seal and oyster. The small size of the fish Core axes and flake axes are present throughout the de-
emphasizes the use of the littoral zone, likely within the posits and one ground stone axe was recovered.
 28, 2001 63
JournalofFieldArchaeology/Vol.

The ceramics in the later Ertebolle at the site are almost ideas or items borrowed from the Neolithic inhabitants of
indistinguishable from the undecorated sherds of Early Central Europe. Items of source-specific raw material such
Neolithic pottery in this area. In fact, Ertebolle pottery as Danubian shafthole axes (Fischer 1982, in press; Fisch-
outside of a few areasin southern Scandinaviais, common er, Pedersen, and Hald 1997) moved from the Neolithic of
characterizationsto the contrary,not particularlythick and eastern Europe into the Mesolithic of southern Scandi-
is reasonably well made. In addition, there are several dif- navia. Antler and bone items were copied in local material
ferent forms represented, even in the small assemblage at in the Ertebolle and waste from their production is not un-
Smakkerup Huse including small cups, oval lamp bowls, common. Clearlythere was contact between contemporary
and larger jars. Mesolithic and Neolithic societies. These items began to
The perforated teeth, the amber pendant, the painted appear in Mesolithic contexts in the middle and late phas-
pebble, and the hair pin are reflections of aesthetic experi- es of the Ertebolle, more than 500 years before the intro-
ence beyond the necessities of subsistence and daily life. duction of domesticated plants and animals. It is, in fact,
Such objects are known elsewhere from Mesolithic graves remarkablethat domesticates appearso late in the sequence
as well as isolated finds. While the perforated teeth were in northern Europe (Price 2000; Rowley-Conwy 1995b).
likely sewn on clothing and the amber pendant was prob- Reports of cereal pollen in pre-Neolithic contexts (e.g.,
ably a piece of jewelry, the painted pebble remains enig- Kolstrup 1988; Schiitrumpf 1972) are not generally ac-
matic. cepted since pollen may be mobile in sediments and there
There is some suggestion of interaction between the are no charred examples. At the sites of Loddesborg and
folk at Smakkerup Huse and western Denmark, although Vik in southern Sweden, Jennbert (1984) has document-
the site is on the west coast of the island of Zealand. Some ed casts of domesticated cereals in Ertebolle pottery dating
artifacttypes at the site are more common on the island of to ca. 4000 B.C., shortly before the beginning of the Ne-
Funen and the peninsula of Jutland than they are in eastern olithic (Koch 1998).
Zealand and southern Sweden. Trindoxe, for example, are Early domesticated cattle also appear.The late Mesolith-
more common in Jutland than in Zealand. Ertebolle ce- ic site of Rosenhof in Holstein, Germany, (Schwabedissen
ramics in Zealand are normally thicker and coarserthan the 1972) yielded bone of domesticated cow with a radiocar-
examples from Smakkerup Huse, while those from the bon date of ca. 4850 B.C. (Hedges et al. 1993). It is not im-
west exhibit thinner walls and finer construction. The bird possible that this cow was stolen from Neolithic farmersto
bone awl from Smakkerup Huse is largely unknown on the south. Other early cattle from Zealand are known from
Zealand, but more common to the west in Jutland and Fu- Early Neolithic contexts at 0garde (3940-3840 B.C.) and
nen. SmakkerupHuse is remarkablysimilar to the late Er- Snevret Hegn (3800-3650 B.C.) (Koch 1998).
teb0lle site ofTybrind Vig, submerged off the west coast of It is certainly the case that most late Ertebolle coastal
the island of Funen (FIG. i; Andersen 1985, 1987), and oc- sites in northern Germany, Denmark, and southern Swe-
cupied intermittently from the early Ertebolle until the be- den witness a continuation of settlement into the Early Ne-
ginning of the Neolithic. olithic. Bjornsholm, for example, is a huge shell midden in
the central Limfjord area of northern Jutland and dates to
What is Neolithic at Smakkerup Huse?
the late Mesolithic and Early Neolithic (Andersen 1993b;
The presence of cattle bones in Layer 7/7G at Andersen and Johansen 1992). The site is of particularim-
Smakkerup Huse raise the question of what defines the portance because of the presence of midden and settlement
Neolithic. Domestic cattle are "Neolithic" by definition remains from both periods, along with evidence for an
and, at SmakkerupHuse, are radiocarbon-datedto the very earthen long barrow. The Ertebolle layers at Bjornsholm
beginning of the EarlyNeolithic period (3900-3800 B.C.). date from 5050 to 4050 B.C. and the Early Neolithic from
These bones were found, however, in a layer where all the 3960 to 3530 B.C. The Ertebolle levels at the shell midden
artifactual materials belong to the Ertebolle period. The of Norsminde in eastern Jutland date from ca. 4300 to
question then becomes whether this is a late Mesolithic site 3800 B.C. while the subsequent Early Neolithic layers are
with domestic cattle or an Early Neolithic site with a from ca. 3775 to 3425 B.C. (Andersen 1991). The Ne-
Mesolithic assemblage. olithic layer at Norsminde was characterizedby an abun-
This problem is present at other late Mesolithic sites in dance of Funnel Beakerpottery, polished flint axes, and the
southern Scandinavia (Petersson 1999; remains of domesticated animals and plants. In all these
Rowley-Conwy
1995a, 1995b; Schwabedissen 1981) where "Neolithic" cases Neolithic artifactsaccompany the introduction of do-
artifacts have been found in Mesolithic contexts. mesticated species.
Pottery,
t-shaped antler axes, bone rings, and bone combs represent There are only four pieces of Neolithic material at
64 Smakkerup
Huse:A MesolithicSettlementin Denmark/Priceet al.

SmakkerupHuse: two small flakes from polished flint axes particularlylike to thank Finn Petersen, Irene Nielsen, Jor-
found in Layer 4 (the redeposited material from the last gen Christensen, Kurt Petersen, and Leif Hammerlev. A
transgression) that lie stratigraphically above the cow major driving force behind the Saltbak Vig Project and the
bones; and two pieces of reworked pointed-butt polished excavations at Smakkerup Huse has been the director of
flint axes from the site that were collected from the surface the Kalundborg og Omegns Museum, Lisbeth Pedersen,
some years ago. These Neolithic pieces are thus likely later who holds a special place in our hearts and deserves our
than the cattle bones, and we conclude that Smakkerup most sincere appreciation.The Saltbak Vig Archaeological
Huse was essentially a Mesolithic settlement when the Project was originally organized by Lisbeth Petersen, Anne
cows arrived. Birgitte Gebauer, Anders Fischer, and T. Douglas Price.
Smakkerup Huse provides striking evidence of prehis- The Graduate School of the University of Wisconsin kind-
toric life-yet the picture remains ambiguous and puz- ly provided the funding for these excavations.
zling. The Mesolithic peoples at Smakkerup Huse were
technologically sophisticated foragers living 6000 years
ago in sedentary communities, successfully exploiting the T Douglas Price, WeinsteinProfessorof EuropeanArchaeolo-
land and the sea. The evidence tells us a great deal about
gy and Directorof Laboratoryfor ArchaeologicalChemistryat
technology and economy but little about organization and the Universityof Wisconsin-Madison,is interestedin the tran-
even less about ideology. The relationship between for- sition to agricultureand archaeologicalchemistry.Current
agers and early farmers at this critical transition in the pre- projectsincludethe analysisof strontiumisotopesin prehistoric
history of northern Europe is complex and remains a sig- burialsfor informationon migrationand field excavationsat
nificant question for research. late Mesolithicsitesin easternDenmark.Mailing address:
Universityof Wisconsin-Madison, DepartmentofAnthropolo-
Acknowledgments gy, 1180 Observatory Drive, Madison WI 53706. E-mail:
We owe a great debt to many who participated and as- [email protected]
sisted in this excavation. The landowners Karl and Bente Anne Birgitte Gebauer,HonoraryFellowin the Depart-
Jensen were unstinting in their hospitality and willingness ment ofAnthropologyat the Universityof Wisconsin-Madi-
for us to dig large holes in their field. The participants in son, has a degreefrom the UniversityofAarhus, Denmark.
the excavations, devoting energy and enthusiasm, included Her interestsincludeceramicsand theNeolithic, the transi-
Jennifer Derfler, Michael P. Jensen, Kasper Johansen, tion to agriculture,and archaeologicalmethod.
Dorthe Pedersen, Lisbeth Pedersen, Erika Price, Michael Signe UlfeldtHede, a Ph.D. studentat the GeologicalIn-
Stafford, Todd Surovell, and Beth Wormaster. Charlie stitute at the Universityof Copenhagen,is interestedinfaunal
and sequencestratigraphy.Her
Christensen, Soren Andersen, Anders Fischer, and Peter analysis,geomorphology,
Rasmussen visited the excavations and/or provided valued Ph.D. projectinvolvesHolocenesea levelchangesin a near-
assistance in various forms. shorelake basin using stratigraphic,biological,and isotopic
variablesfor palaeoenvironmentalreconstruction. E-mail:
Trolle Bagger of Kalundborg very kindly identified the
[email protected]. ku.dk
stone materials from the excavation. Jan Heinemeier as-
CharlotteSedlacek,a Ph.D. studentat the GeologicalInsti-
sisted with the radiocarbon dates from Aarhus. Helle
tute at the Universityof Copenhagen,studiesfishremains,hy-
Strehle and P. H. Mikkelsen pursued the analysis of the E-mail: charlottesedlacek@hot-
drology,andgeomorphology.
fungus and wicker in the Conservation Department at mail.com
Moesgard in Aarhus. Specialized analyses of the fyrsvamp Nanna Noe-Nygaard,Professorat the GeologicalInstitute
and the painted pebble were done by Arne Jensen of the
of the Universityof Copenhagen,studiesfaunal analysis,
Department of Protein Chemistry at the University of taphonomy,sequencestratigraphy,isotopicgeochemistry of hu-
Copenhagen and Mads Chr. Christensen of the Depart- man beingsand animals, lake development,and the transi-
ment of Conservation at the National Museum of Den- tion to agriculture.Currentprojectsincludethe reconstruction
mark, respectively. of lake levelsand environmentalchangein the LI.Amose
This project was the result of collaboration with the basin of westernZealand. E-mail: [email protected]
Kalundborg og Omegns Museum, our base of operations Sarah L.R. Mason,formerResearchFellowat the Institute
and the source of much of the logistic and technical assis- ofArchaeology,UniversityCollegeLondon,is interestedin the
tance for the project. The Museum's fostering of interna- ethnobotanyand archaeobotanyof hunter-gatherersand wild
tional projects is one of its great strengths. The staff of the plant usefor all periodsthroughoutthe temperatezone.
museum was helpful and hospitable beyond all expecta- Jens Nielsen, archaeologistat the Kalundborgog Omegns
tions and their kindness is greatly appreciated. We would Museum in Kalundborg,Denmark, is interestedin the Stone
 28, 2001 65
Journalof FieldArchaeology/Vol.

Age and the recoveryof workedwoodremainsand has exca- Christensen,Charlie

vatedfor manyyears. 1990 "StoneAge Dug-out Boats in Denmark:Occurence,Age,
David Perry(Ph.D., New YorkUniversity)doespaleoeth- Form and Reconstruction,"in D. Robinson, ed., Experi-
mentationand Reconstruction in Environmental Archaeology.
nobotanywith a specificfocus on woodand nut remains,espe- Copenhagen:National Museum, 119-142.
ciallyhazel nutsfrom sitesin NwEurope. 1993 "Land og hav" in S. Hvass and B. Storgaard, eds., Da
klingerI muld.25 arsarkxologiI Danmark.Aarhus:Aarhus
UniversityPress,20-23.
Aaris-Sorensen,Kim 1995 "TheLittorinaTransgressions," in A. Fischer,ed., Man and
1980 "Depauperationof the MammalianFaunaof the Islandof Sea. Oxford: Oxbow Books, 15-22.
Zealand During the Atlantic Period" VidenskabeligeMed- 1997 "Den store havstigningi Storebaelt"in L. Pedersen,A. Fis-
delserfraDanskNaturhistorisk Forening149: 131-138. cher, and B. Aaby, eds., Storebxlti 10.000 ar. Kobenhavn:
Aaris-Sorensen,Kim, and T. N. Andreasen A/S Storebdltsforbindelsen,45-54.
1993 "SmallMammalsfrom Danish Mesolithic Sites"Journalof Degerbol, Magnus, and B. Fredskild
DanishArchaeology 11: 30-38. 1970 The Urus (Bos primigeniusbojanus)and NeolithicDomesti-
cated Cattle in Denmark. Copenhagen: Det Kongelige
Andersen,Knud
1983 Stenalderbebyggelsen i den vestsjellandske
Amose. Copen- DanskeVidenskabemesSelskab,Biologiske Skrifter.
hagen: Fredningsstyrelsen. Enghoff, Inge B.
1994 "Fishingin Denmark during the Ertebolle Period,"Inter-
Andersen,Soren H.
1981 "Ertebollekunst.Nye ostjyske fund af monstrede Erte- nationalJournalof Osteoarchaeology 4: 65-96.
bUlesager,"Kuml 1980: 7-62. Fischer,Anders
1985 "TvbrindVig. A PreliminaryReport on a SubmergedErte- 1982 '"Trade in Danubian Shaft-holeAxes and the Introduction
of Neolithic Economy in Denmark"JournalofDanishAr-
bolle Settlementon the West Coast of Fyn,"Journal
ofDan-
ishArchaeology 4: 52-69. chaeology1: 7-12.
in press "Food for Feasting?An Evaluationof Explanationsof the
1987 "MesolithicDug-outs and Paddlesfrom TybrindVig, Den-
57: 87-106. Neolithisation in Denmark and Southern Sweden,"in A.
mark,"ActaArchaeologica
Fischerand K. Kristiansen,eds., TheNeolithisationof Den-
1991 "Norsminde: A 'Kokkenmodding'with Late Mesolithic mark.Sheffield:SheffieldAcademicPress.
and EarlyNeolithic Occupation,JournalofDanishArchae-
Fischer,Anders,LisbethPedersen,and N. Hald
ology8: 13-40. 1997 "DanubianShaftholeAxes-Long-distance Transportand
1993a "MesolithicCoastal Settlement,"in S. Hvass and B. Stor- the Introductionof Agriculture,"in L. Pedersen,A. Fisch-
gaard,eds., Digging Into thePast. Aarhus:AarhusUniver- er, and B. Aaby,eds., TheDanish StorebxltsincetheIceAge.
sitetsforlag,65-69. Copenhagen:A/S Storebalt Fixed Link, 201-205.
1993b "Bjrnsholm. A Stratifiedk0kkenm0dding on the Central Gebauer,A. B.
Limfjord,North Jutland,JournalofDanishArchaeology 10: 1995 "PotteryProductionand the Introductionof Agriculturein
59-96. SouthernScandinavia," in W. K. Barnettand J. W. Hoopes,
1994 "LystrupEnge,"Archaeologiske UdgravningerI Danmark eds., The Emergenceof Pottery.Washington D.C.: Smith-
1993. Copenhagen:National Museum, 177-178, 279. sonian Institution Press, 99-112.
1995 "Coastal Adaptation and Marine Exploitation in Late Gebauer,A. B., and T. D. Price
Mesolithic Denmark-with SpecialEmphasison the Lim- 1990 "TheEnd of the Mesolithicin EasternDenmark:a Prelim-
fjord Region,"in A. Fischer, ed., Man and Sea. Oxford: inary Report on the Saltbak Vig Project,"in P. M. Ver-
Oxbow Books, 41-66. meerschand P. van Peer,eds., Contributions to theMesolith-
1998 "Ringkloster.ErtebolleTrappersandWild BoarHunters in ic in Europe.Leuven:LeuvenUniversityPress,259-280.
EasternJutland.A Survey,"Journal ofDanishArcheology12: Grigson, Caroline
13-60. 1974 "TheCraniologyand Relationshipsof Four Speciesof Bos.
Andersen,Soren H., and ErikJohansen 1. Basic Craniology:Bostaurusand its Size,"JournalofAr-
1992 "AnEarlyNeolithic Graveat Bjmrsholm, North Jutland," chaeologicalScience1: 353-379.
JournalofDanishArcheology 9: 38-59. Gron, Ole, and J0rgenSkaarup
1991 "MollegabetII-a SubmergedMesolithic Site and a 'Boat
Becker,CarlJohan
1948 "Mosefundnelerkarfrayngre stenalder,studierover Tragt- Burial'fromAr0,"Journal ofDanishArchaeology 10: 38-50.
bager kultureni Danmark, Aarbogerfor NordiskOldkyn- Hartz, Sonke
dighedog Historie1947: 5-318. 1999 "Paddelfundeaus der Mittel- und JungsteinzeitSchleswig-
Holsteins" in T. K. E. Czieslaand S. Pratsch,eds., Den Bo-
Bennike,Pia, and KlausEbbesen
1986 "TheBog Find from Sigersdal"Joural ofDanishArchaeol- gen spannen. Festschriftfur Bernhard Gramschzum 65.
ogy5: 85-115. Weissbach,Germany:Bier & Beran,ArchaologischeFach-
literatur,147-160.
Bokelmann,K.
1991 "Duvensee Wohnplatz 9. Ein praborealzeitlichenLager- Hede, Signe Ulfeldt
1999 "En palaeookologiskrekonstruktionaf den sen atlantiske
platz in Schleswig-Holstein, Offa48: 75-115.
kystbopladsved SmakkerupHuse, SaltbaekVig, NV-Sjael-
Chapela,I., and P. Lizon land, Danmark,baseretpa osteologisk Materiale"unpub-
1993 "Fungiin the Stone Age,"TheMycologist7(3): 123. lished specialethesis, Universityof Copenhagen.
66 SmakkerupHuse: A Mesolithic Settlement in Denmark/Price et al.

Hedges, Robert E. M., R. A. Housley, C. Bronk Ramsey,and G. J. Mason, SarahL. R., and JonathanG. Hather
van Klinken in press "ParenchymatousPlant Remains from Staosnaig,"in S.
1993 "RadiocarbonDates from the Oxford AMS System. Ar- Mithen, ed., Prehistoric Huntersin the Hebrides:the South-
chaeometryDatelist 17"Archaeometry 35: 305-326. ernHebridesMesolithicProject1988-1995. Cambridge:Mc-
T. G. and P. N. Watson Donald Institute for ArchaeologicalResearch.
Holden, G., J. Hather, J.
1995 "MesolithicPlantExploitationat the Roc del Migdia, Cat- Mason, SarahL. R., JonathanG. Hather, and Gordon C. Hillman
alonia"JournalofArchaeological Science22: 769-778. in press a "The Archaeobotany of European Hunter-gatherers--
some PreliminaryInvestigations,"in S. L. R. Mason and J.
Jennbert,Kristine London:
G. Hather, eds., Hunter-gathrerArchaeobotany.
1984 Den produktivagadvan. Traditionochinnovationi Sydskandi- Institute of ArchaeologyOccasionalPublications.
navienfodromkring5300 drsedan.Lund: ActaArchaeologi- in press b "PlantRemainsfrom Mollegabet II,"in J. Skaarupand 0.
ca Lundensia.
Gron, eds., MollegabetII: A Submerged ErtebolleDwelling.
Johansson,Axel Degn
1999 "Ertebollekulturen i Sydsjelland,"Arb0ger 1997: 7-88. Copenhagen:Nationalmuseet.
Mathiassen,Therkel
Koch, Eva 1959 Nordvestsjxllands Oldtidsbebyggelse. Kobenhavn: National
1998 NeolithicBog Potsfrom Zealand,Mon, Lollandand Falster. Museum.
Copenhagen:Nordiske Fortidsminder.
Mertz, E. L.
Kolstrup,Else 1924 Oversigtoverdesen-ogpostglaciale Niveauforandringer iDan-
1988 "LateAtlanticandEarlySubborealVegetationDevelopment mark.Copenhagen:DanmarksGeologiske Undersogelse.
at Trundholm,Denmark"Journal ofArchaeological Science
15: 503-513. Myrhoj,Hannah M.
1997 "Rekonstruktionaf en fiskeruse"in L. Pedersen,A. Fisch-
Kosina,R. er, and B. Aaby,eds., Storebxlti 10.000 ar. Kobenhavn:A/S
1991 "Remarkson the Variationof Fossil and Contemporary Storebxltsforbindelsen,167-169.
Nuts of CorylusavellanaL." in J. Renfrew,ed., New Light
on EarlyFarming:RecentDevelopments Nielsen, Eva K., and ErikB. Petersen
in Paleoethnobotany.
1993 "Burials,People and Dogs" in S. Hvass and B. Storgaard,
Edinburgh:EdinburghUniversityPress, 75-91. eds., Digging Into the Past, Aarhus: Aarhus University
Kubiak-Martens,L. Press, 76-80.
1996 "Evidencefor the Possibleuse of Plant Foods in Paleolithic Poul Otto
and MesolithicDiet from the Site of Calowaniein the Cen- Nielsen,
1977 "Die Flintbeilender friihenTrichterbecherkultur in Dane-
tral Part of the Polish Plain" VegetationHistoryand Ar- mark,"ActaArchaeologica 48: 61-138.
chaeobotany 5: 33-38.
1999 "The Plant Food Component of the Diet at the Late Noe-Nygaard,Nanna
1971 "Spur Dog Spines from Prehistoric and Early Historic
Mesolithic (Ertebolle) Settlement at TybrindVig, Den- Denmark: An Unexpected Raw Material for Precision
mark,"VegetationHistoryandArchaeobotany 8: 117-127.
Tools,"Bulletin of the GeologicalSocietyof Denmark 21:
Lange, Christian 18-32.
1994 "7000 ar gamle svampe, Svampe30: 52. 1977 "Butcheringand MarrowFracturingas a TaphonomicFac-
Larsen,CharlotteS. tor in ArchaeologicalDeposits" Palaeobiology 3: 218-237.
2000 "Enpalaookologisk rekonstruktionaf en sen atlantiskkyst- 1988 "a'lC-valuesof Dog Bones Reveal the Nature of Changes
bopladsSmakkerupHuse, SaltbxkVig, NV-Sjalland,Dan- in Man'sFood Resourcesat the Mesolithic-NeolithicTran-
mark,baseretpa fiskeknoglematerial," unpublishedspeciale sition, Denmark,"ChemicalGeology(IsotopeGeoscience Sec-
thesis, Universityof Copenhagen. tion) 73: 87-96.
Larsson,Lars 1995 Ecological,Sedimentary,and GeochemicalEvolutionof the
1989 "Big Dog and Poor Man. MortuaryPracticesin Mesolith- Late-glacialto PostglacialAmoseLacustrineBasin,Denmark.
ic Societies of Southern Sweden,"in T. B. Larssonand H. Oslo: ScandinaviaUniversityPress.
Lundmark,eds., Approachesto SwedishPrehistory. A Spec-
trum of Problemsand Perspectives in ContemporaryResearch. Pedersen,Lisbeth
BAR International Series500. Oxford: B.A.R., 211-223. 1997 "They Put Fences in the Sea" in L. Pedersen,A. Fischer,
World and B. Aaby, eds., The Danish Storebxltsincethe Ice Age.
1990 "TheMesolithicof SouthernScandinavia,"Journal of
Prehistory4: 257-310. Copenhagen:A/S Storebalt Fixed Link, 124-144.
Perry, David
1997 "CoastalSettlement During the Mesolithic and Neolithic
Periodsin SouthernmostPartof Sweden,"in D. Krol, ed., 1997 "The Archaeology of Hunter-gatherers:Plant Use in the
TheBuilt Environmentof the CoastAreas during the Stone Dutch Mesolithic,"unpublished Ph.D. dissertation.New
YorkUniversity,New York.
Age. Gdansk:UniversityPress, 12-22.
Madsen,Torsten 1999 "VegetativeTissues from Mesolithic Sites in the Northern
Did All the Hunters Go? An Assessment of an Netherlands,"CurrentAnthropology 23: 231-237.
1987 "Where
Epoch-MakingEpisode in Danish Prehistory"Journal of in press "PreliminaryResults of an Archaeobotanical Analysis of
DanishArchaeology 5: 229-239. Mesolithic Sites in the Veenkolonien,Provinceof Gronin-
G.
Madsen,Torsten,and Helle JuelJensen gen, the Netherlands,"in S. L. R. Mason and J. Hather,
Denmark" eds., Hunter-gatherer Archaeobotany. London: Institute of
1982 "Settlementand Land Use in Early Neolithic
AnalectaPraehistorica Leidensia1982: 63-86. Archaeology Occasional Publications.
 Journal of FieldArchaeology/Vol.28, 2001 67

Petersen,ErikBrinch Schiitrumpf,R.
1990 "Nye gravefra Jagerstenalderen"Nationalmuseets Arbejds- 1972 "Stratigraphieund pollenanalytischeErgebnisse der Aus-
mark1990: 19-33. grabung des Ellerbek-zeitlichenWohnplatzes Rosenhof
Petersson,Hikan (Ostholstein),"ArchaologischesKorrespondenzblatt2: 9-16.
1999 "WhereDid All the FarmersCome From?"Journal ofDan- Schwabedissen,Hermann
ishArchaeology 13: 179-190. 1972 "Rosenhof(Ostholstein),ein Ellerbeck-Wohnplatz am ein-
Poder, R., T. Piimpel, and U. Peintner stigen Ostseeufer,"Archaologisches Korrespondenzblatt 2:
1994 "TheIce Man,"TheMycologist8(1): 44. 1-8.
1981 "Erteboile/Ellerbek-Mesolithikum oder Neolithikum?"
Prangsgaard,K. Museum Ur-und Friihgeschichte
1992 "Introduktionaf keramiki den yngre Ertebolle-Kulturi Veroffuntlichung Potsdam
Sydskandinavien,LAG 3: 29-52. 14/15: 129-142.
Price, T. D. Skaarup,Jorgen
1973 Hesselo-Solager. Jagdstationen der Sudskandinavischen
1985 "AffluentForagersof Mesolithic SouthernScandinavia," in Trichterbeckerkultur.
T. D. Priceand J. Brown, eds., Prehistoric Copenhagen:AkademiskForlag.
Hunter-Gatherers.
Theemergence Sorenen, Soren
of CulturalComplexity. New York:Academic
1996 Kongemosekulturen i Sydskandinavien.
Press, 341-363. Fargegarden, Den-
mark:Egnsmuseet.
2000 "The Introductionof Farmingin Northern Europe,"in T.
D. Price, ed., Europe'sFirst Farmers.Cambridge: Cam- Stafford,Micheal
1999 From Foragerto Farmer in Flint. Aarhus: University of
bridgeUniversityPress,260-300. AarhusPress.
Price,T. D.. and ErikB. Petersen
1987 "PrehistoricCoastal Settlement in Mesolithic Denmark," Stuiver,Minge, P. J. Reimer,E. Bard, J. W. Beck, G. S. Burr,K. A.
ScientificAmerican March: 112-121. Hughen, B. Kromer,F. G. McCormac,J. v. d. Plicht, and
M. Spurk
Rackham,Oliver 1998 "INTCAL98 Radiocarbon age calibration 24,000-0 cal
1980 Ancient Woodlands.Cambridge: Cambridge University
BP"Radiocarbon 40: 1041-1083.
Press.
Tauber,Henrik
Randsborg,Klaus 1981 "'13C Evidence for Dietary Habits of PrehistoricMan in
1975 "SocialDimensions of EarlyNeolithic Denmark,"Proceed-
Denmark,"Nature 292: 332-333.
ingsof thePrehistoricSociety41: 105-118.
Trolle-Lassen,Tine
Rasmussen, Peter 1987 "Human Exploitationof Fur Animals in Mesolithic Den-
1990 "LeafFodderingin the EarliestNeolithic Agriculture:Evi- 12: 85-102.
dence from Switzerlandand Denmark"ActaArchaeologica mark,"Archaeozoologia
60: 71-86. 1992 "Butcheringof Red Deer (CervuselaphusL.)-a CaseStudy
from the Late Mesolithic Site of TybrindVig, Denmark,"
Rasmussen,L. W., and JaneRichter JournalofDanishArchaeology 9: 7-37.
1991 Kainsbakke.Grena:DjurslandMuseum.
Vang Petersen,Peter
Regnell, Mats, M.-J. Gaillard,T. S. Bartholin,and Per Karsten 1984 "Chronological and Regional Variation in the Late
1995 "Reconstructionof Environmentand History of PlantUse Mesolithic of EasternDenmark,"Journal ofDanishArchae-
During the Late Mesolithic (ErtebolleCulture) at the In- ology3: 7-18.
land Settlementof BlokebergIII, Souther Sweden,"Veg-
etationHistoryandArchaeobotany Zapata,L., A. Cava,M. J. Iriarte,J. B. Baraybar,and C. De La Rua
4(2): 67-91. in press "MesolithicPlant Use in the Western Pyrenees:Implica-
Richards,Micheal,T. D. Price, and E. Koch tions for VegetationChange,Use of Wood andHuman Di-
in press "Mesolithicand Neolithic Subsistencein Denmark:New et,"in S. L. R. Mason and J. G. Hather,eds., Hunter-gath-
StableIsotope Data,"CurrentAnthropology. ererArchaeobotany. London: Institute of Archaeology,Oc-
Richter,Jane,and Nanna Noe-Nygaard casionalPublications.
in press "ALateMesolithicHunting Station at Agernas, Fyn, Den- Zvelebil,Marek
mark:Differentiationand Specializationin the Late Erte- 1995 "PlantUse in the Mesolithic and Its Role in the Transition
bolle CultureHeralding the Introduction of Agriculture," to Farming,"Proceedings of thePrehistoric
Society60: 35-74.
ActaArchaeologica.
Robinson, David E., and J. A. Harild
in press "Archaeobotanyof an Early Ertebolle (Late Mesolithic)
Site at Halskov,Zealand,Denmark,"in S. L. R. Mason and
J. G. Hather,eds., Hunter-gathererArchaeobotany. London:
Institute of ArchaeologyOccasionalPublications.
Rowley-Conwy,Peter
1995a "Wildor Domestic? On the Evidencefor the EarliestDo-
mesticCattleandPigs in South ScandinaviaandIberia"In-
ternationalJournalof Osteoarchaeology 5: 115-126.
1995b "MakingFirst FarmersYounger:The West EuropeanEvi-
dence,"CurrentAnthropology 36: 346-353.