Boric&Miracle2004-AMS-Padina i H Vodenica

7/26/2019 Boric&Miracle2004-AMS-Padina i H Vodenica

1/32

DU AN BORI AND PRESTON MIRACLE

MESOLITHIC AND NEOLITHIC (DIS)CONTINUITIES IN THEDANUBE GORGES: NEW AMS DATES FROM PADINA ANDHAJDUCKA VODENICA (SERBIA)

Summary. In this paper we present 17 new AMS dates from the

MesolithicEarly Neolithic sites of Padina and Hajdu ka Vodenica and discussthe continuity and nature of occupation at them in the context of the

MesolithicNeolithic transformations in the Danube Gorges region (north-central Balkans), c.100005500 Cal BC. The dates indicate long occupationsequences and help refine the stratigraphies of the two sites. They, also enableus to date architectural features, burial positions and bone/antler tools, and to

further our understanding of the impact of the noted aquatic reservoir effecton radiocarbon dating of human and dog remains from this region. Finally,these dates suggest continuity of occupation at sites other than Lepenski Vir inthe Danube Gorges at the time of the MesolithicNeolithic transition,

c.63005950 Cal BC.

introduction

Since the excavation of MesolithicEarly Neolithic sites in the Danube Gorges (IronGates) region of the north-central Balkans (Fig. 1) in the 1960s and 1970s it has not been easyto estimate the actual depths of the occupational sequences or to phase numerous human burialsand architectural features to particular chronological periods. This situation is due to the complexstratigraphic superposition of features and the nature of occupation at these locations. New AMSdates from two sites Padina and Hajdu ka Vodenica in the Danube Gorges,1 which we present

here, allow us for the first time i) to distinguish clearly Mesolithic and Early Neolithic contextsat the site of Padina, ii) to obtain the first radiometric dates for the site of Hajdu ka Vodenica,iii) to provide further information about the noted freshwater reservoir effect in this region withregard to differences in dating human burials and associated bone tools, iv) to date specific burialpositions found in this culture complex, and v) to date specific typological categories ofantler/bone tools. We further examine whether the distribution of the current radiometric

c

c

c

CS

1 Six AMS dates (OxA-9052, OxA-9053, OxA-9054, OxA-9055, OxA-9056, OxA-9034) were obtained in the

course of the NERC-funded project directed by Alasdair Whittle (Whittle et al. 2002) and the remaining sevenAMS dates from Padina and four dates from Hajdu ka Vodenica came through the NERC-funded ORADS facility

in the course of Du an Bori s Ph.D. project, supervised by Preston Miracle, and in collaboration with theexcavator of Padina and Hajdu ka Vodenica, Borislav Jovanovi .cc

cs

c


2/32

evidence in this region indicates continuity of occupation during the Mesolithic and Neolithicperiods.

conventional and bone collagen radiocarbon dates from padina

Padina (4430 N 2225 E) is situated in the Upper Gorge of the Danube in the vicinityof the sites of Lepenski Vir and Vlasac (Fig. 1). There are three excavated sectors of thissettlement along the Danubes bank with a complex stratification of the remains of severalprehistoric and later periods (Jovanovi 1969a, 1987). A superposition of Mesolithic and Early

Neolithic features and finds was noted in all three sectors. Mesolithic architectural features weremainly represented by elongated stone constructions, primarily in Sectors II (Fig. 2) and III (Fig.3; see Fig. 7), while an Early Neolithic date for buildings with trapezoidal floors has beensuggested on the basis of the associated Early Neolithic Star evo-Krs-Cri pottery found insitu on their floors (Jovanovi 1969a, 1987; Bori 1999, 2002b, fig. 2). The trapezoidal buildingswere found only in Sectors I (Fig. 2) and III.

There are nine previous conventional radiocarbon dates from Padina. Six dates on bonecollagen samples from all three sectors were made at the British Museum (Table 1), five onhuman bones (BM-1143, BM-1144, BM-1146, BM-1147 and BM-1404) and one on a bear bone(BM-1403) (Burleigh and ivanovi 1980; Radiocarbon 24, 1689). These dates indicated a

mostly Early Mesolithic occupation at the site. On the other hand, three charcoal dates (GrN-8229, GrN-8230 and GrN-7981) came from trapezoidal structures in Sector III (Table 2), i.e.

c

ccsc

c

MESOLITHIC AND NEOLITHIC (DIS)CONTINUITIES IN THE DANUBE GORGES

Alibeg

Padina

Stubica

Lepenski Vir

Vlasac

RazvrataIcoana

Veterani

KulaOstrovul Mare

km 875 & 873

Ostrovul

Corbului

Ostrovul Banului

Schela CladoveiHajducka

Vodenica

SERBIA

ROMANIA

DANUBE

Privod

Vodneac

>1000

700-1000 m

500-700 m

300-500 m

100-300 m

< 100 m

0 25

km

Climente I & II

Baile Herculane

Cuina

Turcului

Velesnica

cave sites

open-air sites

Ilisova

Izlaz

SvinitaVirtop

4430'

2225'

Figure 1

Map of the Danube Gorges with Mesolithic and Early Neolithic sites.


3/32

OXFORD

JOURNALOFARCHAEOLOGY

BlackwellPublishingL

td.2004

343

Figure 2

Architectural features and burials, Sectors III, Padina; provenance of dated samples marked


4/32

OXFORD

JOURNA

LOFARCHAEOLOGY

344

B

lackwellPublishingLtd.2004

Figure 3

Architectural features and burials, Sector III, Padina; provenience of dated samples marked


5/32

DU AN BORI AND PRESTON MIRACLECS

from hearths and charred remains of timber beams of roofs and posts that framed the upperconstruction of buildings (Groningen Database, obtained from Centrum voor IsotopenOnderzoek, Groningen; see Bori 1999, fig. 6). At the moment we lack more precise informationabout the exact provenience of the latter dates at the level of building associations. However,

these dates indicate that the occupation of trapezoidal structures coincides with the duration ofthe Early Neolithic in the wider region (cf. Whittle et al. 2002).

ams dates and correcting the aquatic reservoir ages in the danube gorges

Absolute dating by using Accelerator Mass Spectrometry (AMS) on human anddomesticated dog bone samples in the Danube Gorges is affected by the freshwater reservoireffect (cf. Bonsall et al. 1997, 2000; Cooket al. 2002; Whittle et al. 2002). This situation is dueto the substantial intake of both freshwater and migratory fish, and possibly other food sources

c

table 1

Bone collagen dates from Padina (after Burleigh and ivanovi 1980, table 1). All dates calibrated with OxCal v. 3.9

(Bronk Ramsey 1995, 2001) using INTCAL98 calibration curve (Stuiver et al. 1998)

lab ID context burial position species 14C Age Cal BC at Cal BC at

(bp) 1 s.d. 2 s.d.

BM-1146 Burial 12 flexed/right, Homo sapiens 9331 58 87208470 87508330(Sector III) perpendicular to the

Danube headupslope/away fromthe Danube

BM-1404 Burial 39 (?)2 (?) Homo sapiens 9292 148 87308290 91508200

BM-1147 Burial 14 (?) Homo sapiens 9198 103 85408280 86908230(Sector III)

BM-1144 Burial 7 extended, perpendicular Homo sapiens 8797 83 82007650 82507600(Sector II) to the Danube

head upslope/awayfrom the Danube

BM-1403 (?) Sector III n/a Ursus arctos 8138 121 75406800 75506700

BM-1143 Burial 2 extended, parallel to Homo sapiens 7738 51 66406480 66506460(Sector I) the Danube head

downstream

cZ

table 2

Charcoal dates from Padina (source: Groningen Database). All dates calibrated with OxCal v. 3.9 (Bronk Ramsey

1995, 2001) using INTCAL98 calibration curve (Stuiver et al. 1998)

lab ID context material 14C Age (bp) Cal BC at 1 s.d. Cal BC at 2 s.d.

GrN-8229 trapezoidal bldg(?) hearth charcoal 6570 55 56105470 56305380GrN-8230 occupation layer(?) charcoal 7100 80 60605840 61605780GrN-7981 trapezoidal bldg(?) charcoal 7075 50 60005840 60305800

2 It is not clear to what burial this date relates. Burleigh and ivanovi 1980 mention it as Infans I. However,there is no burial marked as Infans I at Padina (see Roksandi 1999, Appendix I).c

cZ


6/32

rich in protein, depending on the complexity of the food web (Grupe et al. 2003; Bori et al.in press). This situation makes the measurements older due to the deposition of old carbon(e.g. Lanting and van der Plicht 1998) and requires a correction of uncalibrated measurements.These corrections are critical for the dating of specific contexts and for the question regardingthe continuity of occupation at Padina.

Here, for the correction of the freshwater reservoir effect, we adopt one of three methodssuggested by Cooket al. (2002). According to that study, one should apply the correction of200500 years to those dated human burials with high d15N values, depending on individualvalues of d15N and d13C and their ratios. In order to estimate the correction factor for the reservoireffect, these authors take the highest d15N value of +17 measured for an adult individual fromLepenski Vir (Burial 89a: Bonsall et al. 1997, table 5) as an end-point for a 100 per cent aquaticdiet while an end-point of +8 is equated with a 100 per cent terrestrial diet (Cooket al. 2002,81). Cook et al. suggested three methods of correcting the freshwater reservoir effect in theDanube Gorges on the basis of d15N values of particular human burials and associated ungulate

bone tools/weapons that were either found as grave offerings with dated burials or wereembedded in a skeletal part of deceased humans (the latter due to violent encounters) from thesite of Schela Cladovei (ibid., table 3). They approximate a weighted mean age offset for a 100per cent diet to 540 70 radiocarbon years. According to Method 1, the applied correction factordepends on the percentage of estimated aquatic diet and it is calculated with regard to d15Nvalues (e.g. the value of +15.1 indicates 79 per cent of aquatic diet). Method 2 relies on allthe data from table 3 in Cooket al. (2002) and applies the correction factor of 440 45 yearsoffset (100 per cent aquatic diet) for those measurements that show d15N values are >+13 and220 23 years offset (50 per cent aquatic diet) when d15N values are between +10 and +13.Finally, Method 3 relies on 19 dated bone artefacts from Schela Cladovei, which relate to the

Mesolithic period, and ten dated human burials from the same period, calculating a weightedmean of this series of measurements for an average human diet with d15N values of 15.4

c


table 3

Corrected ages for human and dog AMS dates with d15N values >+10 (affected by the aquatic reservoir effect) from

Padina and Hajdu ka Vodenica using all three methods suggested by Cook et al. (2002). The d15N values used toestimate percentage of aquatic diet

lab ID burial/ species 14C age d15N protein corrected corrected correctedspecimen (bp) () (aquatic?) 14C age (bp) 14C age (bp) 14C age (bp)no. signal (%)* Method 1 Method 2 Method 3

PadinaOxA-11106 21 Homo 10095 55 14.1 67 9733 72 9655 71a 9668 72OxA-11104 11 Homo 10000 60 13.0 55 9703 72 9780 64b 9649 72OxA-11105 15 Homo 9480 55 13.7 63 9140 70 9040 71a 9078 70OxA-11107 1a Homo 7975 50 15.5 84 7521 77 7535 67a 7439 77OxA-9053 8.70/358/1 Canis 7685 60 11.3 37 7485 65 7465 64b 7449 65OxA-9056 8.70/347/1 Canis 7625 55 12.5 50 7355 65 7405 60b 7306 65

Hajdu ka VodenicaOxA-11128 8 Homo 8645 65 16.0 89 8146 90 8205 79a 8077 90OxA-11127 12 Homo 7990 55 15.8 86 7526 81 7550 71a 7441 81OxA-11126 15-younger Homo 7980 50 15.6 84 7526 77 7540 67a 7444 77OxA-11109 20 Homo 7875 55 16.1 90 7389 84 7435 71a 7301 84

* used for Methods 1 and 3; a= 100 per cent reservoir correction applied (440 45 years); b = 50 per cent reservoir correctionapplied (220 23 years).

c

c


7/32


table 4

AMS dates on animal bones and antler from Padina (OxA-9055, OxA-9034, OxA-9053, OxA-9056, OxA-9052 and

OxA-9054 after Whittle et al. 2002). AMS dates with d15N values >+10 corrected for the freshwater reservoir effectaccording to Method 2 as described by Cook et al. (2002, 82). All dates calibrated with OxCal v. 3.9 (Bronk

Ramsey 1995, 2001) using the INTCAL98 calibration curve (Stuiver et al. 1998)a = 100 per cent reservoir correction applied (440 45 years); b = 50 per cent reservoir correction applied (220 23

years)

lab ID sample species skeletal S # d15N d13C 14C age corrected Cal BC atcontext element & () () (bp) 14C age 2 s.d.

inv. no. (bp)OxA-11102 underneath Cervus astragalus 3 4.6 -20.0 9990 55 n/a 99659275

House 14 elaphus 8.70/377/5

OxA-9055 midden, Cervus mandible P5 4.7 -20.8 8445 60 n/a 76007340Profile 3 elaphus 7.70/191/1segm. 1,exc. level 3

OxA-9034 above House Canis tibia P1 8.6 -17.7 7755 65 n/a n/a12 under familiaris 8.70/289/15the levelwith animalbones

OxA-11108 offering in Cervus antler tool 12 6.2 -20.9 7750 50 n/a 66506460Burial 1a elaphus

OxA-9053 underneath Canis ulna P3 11.3 -17.7 7685 60 7465 64b 64406210House 18 familiaris 8.70/358/1

OxA-9056 floor of Canis tibia P6 12.5 -18.1 7625 55 7405 60b 64106090House 9 familiaris 8.70/347/1

OxA-11103 hearth of Mammalia bone tool 7 7.2 -22.4 7315 55 n/a 62506025House 17 IB 395

OxA-9052 floor of Cervus antler P2 6.6 -22.2 6965 60 n/a 59905720House 18 elaphus 7.70/169/1

OxA-9054 underneath Mammalia worked bone P4 5.7 -21.5 6790 55 n/a 57805560House 15 8.70/ku a 15/1c

approximated to 638 70 years. We find Method 1 inadequate, pretending to provideunattainable accuracy, which is at the current state of research unrealistic. Moreover, it ismisleading to suggest that +17 indicates a 100 per cent aquatic diet in humans, and one shouldbe reminded that d15N values signal the contribution of protein in general, which can be a

combination of several protein-rich food sources and not only fish (cf. Bori et al. in press). Onthe other hand, Method 3 relies on a limited number of artefacts from Schela Cladovei that werenot associated with particular burials, only assuming the contemporaneity of the two. In fact,all three methods are to some degree speculative and at best tentative at present, relying on arather limited number of dated burials and associated artefacts from Schela Cladovei alone.However, in order to approximate the real age of samples affected by the freshwater reservoireffect we have corrected new AMS dates from Padina and Hajdu ka Vodenica by all threemethods (Table 2).

In the following discussion all new dates with d15N values >+10 from these two sitesare corrected according to Method 2 (see Tables 46). Although Method 2 provides only a rough

c

c


8/32

estimate of the correction factor, we find it still the most realistic among the three offeredmethods. Method 2 closely relies on the relationship between dated burials affected by thefreshwater reservoir effect and unaffected associated artefacts. It also takes into accountdifferences in the d15N values in a realistic way with regard to the quality and type of datepresently at our disposal. One should not forget, however, that these associated ungulate artefactsare only relatively contemporaneous with the individuals in whose burials they were found, andit would be reasonable to expect that the obtained date for these artefacts is always at least

marginally older than the true age of the deceased.

ams radiocarbon dating at padina

Recently, 13 new AMS dates were obtained for Padina. Nine dates were made on animal(Table 4) and four on human bones (Table 5).

In the case of Burial 1a (OxA-11107, uncorrected 7975 50bp), both the skeleton andthe associated antler mattock were dated in Sector I (see Fig. 2). Due to the correction for thefreshwater reservoir effect for the deceased, the actual date could only be in the same range oryounger than the associated antler mattock, which was dated in the range 6650 to 6460 at 95

per cent probability (OxA-11108, 7750 50bp). The age offset between the dated burial andthe dated tool is 225 71. On the other hand, by applying the correction for the freshwater


table 5

AMS dates on human bones from Padina. All dates corrected for the freshwater reservoir effect according to Method

2 as described by Cook et al. (2002, 82) and dates calibrated with OxCal v. 3.9 (Bronk Ramsey 1995,2001) using the INTCAL98 calibration curve (Stuiver et al. 1998)

a = 100 per cent reservoir correction applied (440 45 years); b = 50 per cent reservoir correction applied

(220 23 years)

lab ID burial no. burial position skeletal S # d15N d13C 14C age corrected Cal BC at& context element () () (bp) 14C age 2 s.d.

(bp)

OxA-11106 21 (stone seated (?) rib 10 14.1 -18.9 10095 55 9655 71a 92508790construct.)

OxA-11104 11 extended, rib 8 13.0 -18.9 10000 60 9780 64b 93608920(beneath perpendicularHouse 15) to the

Danube head upslope/away fromthe Danube

OxA-11105 15 seated with trapezium 9 13.7 -19.0 9480 55 9040 71a 84507960(stone crossed legsconstruct.)

OxA-11107 1a extended, long bone 11 15.5 -17.9 7975 50 7535 67a 64706230(Sector I) perpendicular frag.

to theDanube head upslope/away fromthe Danube


9/32


reservoir effect according to Method 2 (Cooket al. 2002) the burial is dated in the range 6470to 6230 Cal BC at 95 per cent probability (corrected 7535 67bp). In this case the deceasedseems to be at least 100 years younger than the tool placed in the burial. This date indicates avery late (terminal) Mesolithic3 occupation of this portion of the site. That this burial belongsto the local Mesolithic tradition in this region can further be suggested on the basis of its burialposition an extended inhumation placed perpendicular to the Danube with its head pointingupslope/away from the Danube (cf. Radovanovi 1996, 160224; see below for the sameposition and date characterizing Burial 12 at Hajdu ka Vodenica). Burial 1a was placed directlyon the bedrock (Fig. 4). The physical anthropologist who first examined the skeleton noted thatthe postcranial skeleton is gracile and that the measurement indices are characteristic ofNeolithic skeletons ( ivanovi 197374, 141). While ivanovi (197374) identifies thedeceased as adult male, in a more recent analysis of this skeleton, Roksandi sexed the individualas old adult male? (1999, Appendix I). The associated antler mattock (OxA-11108, Fig. 5) fallsunder the typological category of tools that are frequently found in MesolithicEarly Neolithiccontexts at sites in the Danube Gorges (e.g. Srejovi and Letica 1978; Boronean 2001). Yet,tc

cccZ

cc

table 6

AMS dates on human bones from Hajdu ka Vodenica. All dates corrected for the freshwater reservoir effect according

to Method 2 as described by Cook et al. (2002, 82) and calibrated with OxCal v. 3.9 (Bronk Ramsey1995, 2001) using the INTCAL98 calibration curve (Stuiver et al. 1998)

a = 100 per cent reservoir correction applied (440 45 years); b = 50 per cent reservoir correction applied (220

23 years)

lab ID burial no. burial position skeletal S # d15N d13C 14C age corrected Cal BC at& context element () () (bp) 14C age 2 s.d.

(bp)

OxA-11128 8 (older extended, cranial 20 16.0 -18.6 8645 65 8205 79a 74807050hearth) parallel to the frag.

Danube headupstream

OxA-11127 12 extended, cranial 19 15.8 -17.3 7990 55 7550 71a 65006230(flooring perpendicular frag.of the to the Danube sacrificial head upslope/hearth away fromarea) the Danube

OxA-11126 15 extended, rib 18 15.6 -18.6 7980 50 7540 67a 64706230younger parallel to the(grave Danube headchamber) downstream

OxA-11109 20 (grave extended, cranial 17 16.1 -18.0 7875 55 7435 71a 64406090chamber) parallel to the frag.

Danube headdownstream

c

3 The designation Mesolithic is used only as a heuristic tool here, and primarily to describe an archaeological

context with no Early Neolithic pottery (for a discussion with regard to theoretical, conceptual and empiricalproblems of using Mesolithic/Neolithic labels in the Danube Gorges see Bori in press a, in press b).c


10/32

currently, there are a limited number of dates for bone and antler tools only from the site ofSchela Cladovei (Cooket al. 2002), and one would need more dated bone/antler tools to providea firmer chronology for certain typological groups.

The correction for the freshwater reservoir effect should also be applied to three dateddog bones, which were found immediately on or below floors of trapezoidal buildings in SectorIII (Fig. 3). The isotopic values of the dogs are most probably indicative of a similar diet to thatof humans, where a larger proportion of protein was from freshwater fish, particularlyanadromous fish, reflecting the higher d15N values in aquatic ecosystems. Humans might have

fed the dogs with fish remains and/or dogs might have scavenged on the residues of humanconsumption. This explanation is largely based on their d13C values (Table 4). The d15N valuesof the two dogs (OxA-9053 and OxA-9056) are higher, too, while one dog (OxA-9034) has aconspicuously low d15N value for this explanation, and would perhaps require a newmeasurement. We have corrected two dated dog bones from Padina that have d15N values>+10 with the same method applied to the dated human samples since dogs show roughlythe same general trend with regard to their d13C and d15N values as that in humans, and accordingto a general assumption that dog isotope values in certain instances can be used as proxies forhuman diets (Schulting and Richards 2002). However, in the future, one should try to developa specific method for the correction of dates for dog bones since humans in the Danube Gorges

are clearly at the top of the food chain and are more enriched in15

N than dogs (see Bori et al.in press). With regard to OxA-9034 (House 12) with its relatively low d15N value of +8.6, we

c


Figure 4

Burial 1a (OxA-11107), Sector I, Padina (photo: after ivanovi 197374, T. 1.1).cZ


11/32


refrain from applying any correction factor.4 OxA-9056 (uncorrected 7625 55bp) provides thedate for the occupation and/or abandonment of House 9 in the range 6410 to 6090 Cal BC at95 per cent probability (corrected 7405 60bp). On the other hand, the dated dog bone fromHouse 18 (OxA-9053, uncorrected 7685 60bp) originates from the area underneath the floorof this building and would represent a terminus ante quem for the construction and/or occupationof this later building, dated in the range 6440 to 6210 Cal BC at 95 per cent probability (corrected7465 64bp).

Similarly, three new AMS dates made on ungulate bones, which do not require thecorrection for the freshwater reservoir effect, further reaffirm the Early Neolithic context of

Figure 5

AMS-dated antler mattock (OxA-11108) found in Burial 1a, Sector I, Padina.

4 Due to its d13C value, which is almost identical to two other dog measurements, we may assume that the actualdate for this individual has to be 200500 years younger.


12/32

trapezoidal buildings. Thus, OxA-11103 (7315 55bp) dates a bone projectile point (Fig. 6)from the hearth of House 17 in Sector III, indicating the occupation/abandonment of thisbuilding in the range 6250 to 6025 Cal BC at 95 per cent probability. A sample from House 18(OxA-9052, 6965 60bp) and another from House 15 (OxA-9054, 6790 55bp) indicate a

later phase of the Early Neolithic occupation at the site after 6000 Cal BC, i.e. they fall in therange 5990 to 5720 and 5780 to 5560 Cal BC at 95 per cent probability. Significantly, theselatter dates come from Houses 18 and 15, which on the basis of their archaeological contextcould well be among the latest buildings to be occupied at the site (cf. Jovanovi 1987). Thereis also an expected chronological difference between OxA-9053 (sample beneath House 18) andOxA-9052 (sample on the floor of House 18), which are vertically stratified within the samebuilding. In sum, these samples derive from bones found in association with Early Neolithicpottery, and they date the earliest phase of the Early Neolithic period in this region. This periodoverlaps with the beginnings of Early Neolithic settlements in the surrounding areas of the north-central Balkans with the starting date at around 63006200 Cal BC (cf. Whittle et al. 2002).

Early indications about a significant chronological depth to the occupational sequenceat Padina on the basis of the first collagen dates made on human bones (Table 1), found inassociation with the area defined by the excavator as the stone construction of the necropolisin Sector III (Fig. 7), are now confirmed with new AMS dates. Burials 21 (OxA-11106,uncorrected 10095 55bp) and 15 (OxA-11105, uncorrected 9480 55bp, Fig. 8), also found

c


Figure 6AMS-dated bone projectile point (OxA-11103) found inside the rectangular hearth of House 17, Sector III, Padina.


13/32


in association with the stone construction of the necropolis (Fig. 7) like previously dated Burial12 (Table 1), despite the necessary correction for the reservoir effect are significantly early,indicating an Early Mesolithic use of this location as a burial ground in the range 9250 to 8790and 8450 to 7960 Cal BC at 95 per cent probability (corrected 9655 71 and 9040 71bp)(Table 5). Burial 21 is identified as old adult female? and Burial 15 as old adult female(Roksandi 1999, Appendix I). Still earlier, Burial 11 (OxA-11104, uncorrected 10000 60bp),found partly in an area beneath the floor of House 15 (Fig. 9), falls in the range 9360 to 8920

Cal BC at 95 per cent probability (corrected 9780 64bp). With regard to Burial 11, one hasto emphasize the difference of at least c.3000 years between OxA-11104 dating Burial 11 andOxA-9054 dating the construction/occupation of this building (see above). Burial 11 was foundat the level of the stone construction that was found underneath House 15 (Padina field diaryfor 13/8/1970), placed in the loess soil, without visible traces of a burial pit ( ivanovi 197374,145). The deceased is identified as an 11-year-old child on the basis of its dental eruption stage(ibid.; also Roksandi 1999, Appendix I). This individual is characterized by the lowest d15Nvalue of +13 if compared to other burials dated to the Mesolithic in the Danube Gorges (cf.Tables 3 and 5; Bori et al. in press). This difference may significantly relate to the young ageof this individual, which if further confirmed, may significantly differentiate dietary practices

of adults and children during the Mesolithic in this region. In addition, one should note thatthere was a damaged area in the floor in the part of the building approximately above Burial 11

c

c

cZ

c

Figure 7

Stone construction of the necropolis, Sector III, Padina (plan: courtesy B. Jovanovi ).c


14/32

(Figs. 3 and 9). Despite such conspicuous damage to the floor, it is unreasonable to expect thatBurial 11 was interred from the level of the building (hence the damaged floor according to thislogic), which would have been an extremely odd outlier to the overall chronology of thetrapezoidal buildings in the Danube Gorges. This particular situation may rather indicatethat there might have been a later awareness about the existence of this burial, and that thismuch older burial and other features at the site might have been imbued with some significancein the course of the Early Neolithic occupation, as traces of previous occupation at the

locale, maintaining some elements of habitual continuity of practices and social memory (forthe same argument with regard to architectural and mortuary evidence at Lepenski Vir see Bori2003).

Two new AMS dates also indicate that Sector III at Padina was not used only as a burialground during this early phase but most likely also as a seasonal or more permanent occupationbase (for evidence about the seasonality of occupation at the site see Bori 2002c). OxA-11102(9990 55bp) dates the remains of occupation (a red deer bone) beneath (i.e. sealed with thefloor of) House 14 in the range 9965 to 9275 Cal BC at 95 per cent probability, while OxA-9055 (8445 60bp) dates a midden deposit in the area around House 6 in the range 7600 to7340 Cal BC at 95 per cent probability. Although there is a significant chronological gap of at

least around 1500 calendric years between these two dates, there is a possibility that the locationmight have been used continuously over several millennia during the Mesolithic period in the

c

c


Figure 8

Burials 15 (OxA-11105) and 16 (sitting positions with crossed legs), stone construction of the necropolis, Sector

III, Padina (photo: courtesy B. Jovanovi ).c


15/32


region. A large series of new dates from single features would be necessary to show clearlywhether the site was continuously or discontinuously occupied during this period.

In terms of dating specific burial positions in the Danube Gorges, Burial 15 isparticularly interesting. This is one of two burials at Padina that were placed in a sitting positionwith crossed legs next to each other (Fig. 8; Jovanovi 1969a, T. XIII.1). Padina is the only sitein the Danube Gorges where two individuals were placed in this way since at all other sites withseated burials only one sitting individual was found per site. This burial position was noted atthe sites of Lepenski Vir (Burial 69: Srejovi 1972, fig. 52), Vlasac (Burial 17: Srejovi andLetica 1978, 72), Kula (Burial 5: Miki and Sladi 1994), Velesnica (Burial 2-skeleton G: Vasi

1986) and Ostrovul Corbului (Burial 25: P unescu 1996). In an attempt at relative (stratigraphic)dating of specific burial positions at the sites in the Danube Gorges, Radovanovi (1996: forc

a

ccccc

c

Figure 9

House 15 (OxA-9054), Sector III, Padina; Burial 11 (OxA-11104) was found below the building floor (photo: courtesy

B. Jovanovi ).c


16/32

Padina pp. 171174; for Vlasac pp. 20910, 218) assumes that this burial position characterizesthe later phase of the Mesolithic occupation in the region. However, OxA-11105 for Burial 15at Padina indicates a much earlier Mesolithic date between 8450 and 7960 Cal BC for theappearance of this burial position, and it will be of some interest to date other burials in the

region placed in the sitting position with crossed legs. The oldest among currently dated burialsfrom Padina is child Burial 11, an extended inhumation with the orientation westeast( ivanovi 197374, 145), i.e. perpendicular to the Danube. The similar arrangement of Burial1a in Sector I (OxA-11107 and OxA-11108) (Fig. 4) indicates that this position of extendedinhumations placed perpendicular to the Danube course with their heads pointing upslope/awayfrom the river relates both to this very late Mesolithic occupation of the site and to the oldestdated burial at Padina (see below for Burial 12 from Hajdu ka Vodenica placed in the sameposition).

Thirteen new AMS dates from Padina (Fig. 10; OxA is not included in this figure) havesignificantly clarified our understanding of the chronological sequence at the site. The dates

indicate that people used and returned to the site over a very long time span, at least from the

c

cZ


Atmospheric data from Stuiver et al. (1998); OxCal v3.9 Bronk Ramsey (2003); cub r:4 sd:12 prob usp[chron]

12000CalBC 10000CalBC 8000CalBC 6000CalBC

Calibrated date

OxA-11102 999055BP

OxA-11104 978064BP

OxA-11106 965571BP

OxA-11105 904071BP

OxA-9055 844560BP

OxA-11108 775050BP

OxA-11107 753567BP

OxA-9053 746564BP

OxA-9056 740560BP

OxA-11103 731550BP

OxA-9052 696560BP

OxA-9054 679055BP

Figure 10

Plot of calibrated AMS dates on animal bones and human burials from Padina; AMS dates with d15N values >+10

corrected for the freshwater reservoir effect according to Method 2 as described by Cook et al. (2002, 82). All dates

calibrated with OxCal v. 3.9 (Bronk Ramsey 1995, 2001) using the INTCAL98 calibration curve (Stuiver et al.1998).


17/32

mid-tenth to mid-sixth millennia Cal BC. This occupation span might have been discontinuous,with periods of abandonment, i.e. dispersion and renewed aggregation of occupation, up toaround 6300 Cal BC. At this time, most likely, the construction of the first structures withtrapezoidal floors starts at Padina and at other sites in the Upper Gorge of the Danube coinciding

with the appearance of Early Neolithic pottery. This Early Neolithic phase of occupation at thesite can largely be characterized as continuous, most likely from around 6300 to 5500 Cal BC,when the location was finally abandoned by this regional tradition.

ams radiocarbon dating at hajducka vodenica

Hajdu ka Vodenica (4428 N 2235 E) is situated at the downstream entrance to theLower Gorge of the Danube on the right bank of the Danube (Fig. 1). At Hajdu ka Vodenica,there are two distinctive areas of the site (Jovanovi 1966, 1967, 1969b, 1984). In the first arearectangular hearths were covered by stone constructions, which were piled up in four distinct

levels. These elongated stone constructions follow the same outline. One could assume aMesolithic date for the lowest levels of this construction with abundant bone tools and no pottery.Early Neolithic pottery of Star evo-Krs-Cri type was found in the two upper levels of thestone construction. This first area of the site remains without radiometric dates. The second areaof the site consists of a grave chamber where a number of burials were placed in extendedpositions parallel to the Danube, with their heads pointing downstream (Figs. 1112). The burialswere found behind a rectangular hearth in an area named the sacrificial hearth area, which wassurrounded by a packed flooring of red burned earth (Jovanovi 1984, 307). Beneath the hearth,on the lower level, remains of an older hearth and structure were recognized (ibid., 307). Twolarge ornamented boulders were found in the front part of the hearth area at a lower level. All

these features were interred into virgin soil (loess-sandy deposit) while the accumulation ofdeposits above the features (e.g. Fig. 12) in certain areas reached up to 4 m in height (Jovanovi1967, 182).

Four new AMS dates on human burials from the described area at Hajdu ka Vodenica(Table 6) are the first radiometric dates from this site. Our previous discussion about thefreshwater reservoir effect correction applies here (see above).

Burial 8 (OxA-11128, uncorrected 8645 65bp) is currently the oldest dated burial asis also suggested by its stratigraphic position. It was placed before the sacrificial hearthconstruction was built on a lower level than the rest of the burials. In contrast to most otherburials at this site, which were placed parallel to the Danube with their heads pointing

downstream, Burial 8 was placed also parallel to the Danube with its head pointing upstream.This burial provides a terminus post quem for the use of the rectangular hearth and flooring inthe range 7480 to 7050 Cal BC (corrected 8205 79bp). The deceased is identified as old adultmale? (Roksandi 1999, Appendix I).

Burial 12 (OxA-11127, uncorrected 7990 55bp) was placed over a flooring of burnedearth, reddish in colour, that surrounded the area of the sacrificial rectangular hearth. It providesa terminus ante quem for the use of this structure, and once corrected for the freshwater reservoireffect falls in the range 6500 to 6230 Cal BC at 95 per cent probability (corrected 7550 71bp). The deceased has not been sexed; it is identified as an adult individual.

Burials 20, unsexed old adult (OxA-11109, uncorrected 7875 55bp) and 15-

younger, unsexed juvenile (OxA-11126, uncorrected 7980 50bp) were placed inside theburial chamber, which contained 20 other articulated and partial human skeletons (for various

c

c

c

c

sc

cc

c



18/32


Figure 11

Architectural features and burials, Hajdu ka Vodenica; dated burials marked (plan: courtesy B. Jovanovi ).cc


19/32


remarks concerning the physical anthropology of these burials see Roksandi 1999, 2000). Afterthe necessary correction, Burial 20 falls in the range 6440 to 6090 and Burial 15-younger inthe range 6470 to 6230 Cal BC at 95 per cent probability (corrected 7435 71 and 7540 67bp). Burial 20 was accompanied by a stone axe (?) placed beside the deceased, while Burial 15-younger was accompanied by a herbivore mandible (Radovanovi 1996, 221).

These three dates (OxA-11127, OxA-11126 and OxA-11109) form a relatively tightcluster that, at face value, seems to indicate a relatively short period of time during the Late

Mesolithic occupation of the site (Fig. 13). After correction for the reservoir effect the dates fallin the range 6500 to 6090 Cal BC, which straddles the appearance of the Early Neolithic in the

c

c

Figure 12

Isometric view of the sacrificial hearth area, front part of the grave chamber, and Section 3-1, Hajdu kaVodenica (adopted after Jovanovi 1969c, 68 and 1967, fig. 2).c

c


20/32

wider region. This chronological bracketing can perhaps be applied to other articulatedinhumations found in the grave chamber at Hajdu ka Vodenica. On the other hand, a group ofdisarticulated bones labelled as Burials 2628 found in the rear of the grave chamber (Fig. 11)could be of older date. Both hypotheses remain to be tested by future radiocarbon measurements.

In terms of their burial position, the placing of extended inhumations parallel to theDanube with their heads pointing downstream confirms Radovanovi s suggestion that thisspecific position characterizes the Late Mesolithic phase of burial practices in the Danube

Gorges (1996, 224), what she calls later type of formal disposal areas (1996, 222). However,it seems that on the basis of our radiometric evidence for Burial 12, which was placed in anextended position perpendicular to the Danube with its head upslope/away from the Danube,and is similar to Burial 1a at Padina, both dated to almost exactly the time span between 6500and 6230 Cal BC, one cannot support Radovanovi s conclusion that the latter burial positioncharacterizes exclusively the earlier Mesolithic phase in this region. Her conclusion wasprimarily based on her stratigraphic analyses of burials from Vlasac (ibid., 218). Also, it seemsthat rather than accepting the claim that frequent orientation of graves parallel to the course ofthe river [. . .] became an absolute standard during Hajdu ka Vodenica Ib (ibid., 222), we arenow able to assert that extended inhumations both oriented parallel to the Danube (heads

pointing downstream) and perpendicular to the Danube characterize Late (terminal) Mesolithicburial practices at Hajdu ka Vodenica. The chronological difference is noted in the orientationof older Burial 8, which was placed parallel to the Danube but with its head pointing upstream.

At present, Hajdu ka Vodenica appears as a site used as a burial location during theMesolithic period from around 7500/7000 Cal BC, and possibly earlier. Again, it is possible thatthe site might have been temporarily abandoned for a part of the period between around7500/7000 to 6500/6400 Cal BC. However, new dates in the future may cover this chronologicalgap. Our radiometric evidence shows that the location was certainly used again after 6500/6400Cal BC for burial purposes. The spatial position of the later phase burials in relation to the earlyphase Burial 8 indicates likely recognition of the same location for the same purpose and

possibly by people that belonged to the same culture/regional tradition as those earlier occupantsof the site. This situation, similar to Padina, exemplifies the depths of social memory, since the

c

c

c

c

c

c


Atmospheric data from Stuiver et al. (1998); OxCal v3.9 Bronk Ramsey (2003); cub r:4 sd:12 prob usp[chron]

8000CalBC 7500CalBC 7000CalBC 6500CalBC 6000CalBC

Calibrated date

OxA-11128 820579BP

OxA-11127 755071BP

OxA-11126 754067BP

OxA-11109 743571BP

Figure 13

Plot of calibrated AMS dates on human burials from Hajdu ka Vodenica; AMS dates with d15N values >+10 corrected

for the freshwater reservoir effect according to Method 2 as described by Cook et al. (2002, 82). All

dates calibrated with OxCal v. 3.9 (Bronk Ramsey 1995, 2001) using the INTCAL98 calibration curve (Stuiveret al. 1998).

c


21/32


current chronological gap between these two radiometrically dated phases of occupation at thesite is at least 500 years. The later phase of occupation most likely lasted up until around 6200Cal BC but perhaps even longer. The presence of Early Neolithic pottery in the presently undatedarea of the site (two upper levels of the elongated stone constructions, cf. Jovanovi 1984,

30910) could tentatively suggest an unbroken continuity of occupation at this location to theEarly Neolithic.

gaps or continuities in the meso-neolithic transformations?

In a recent paper, Bonsall et al. (2002) suggest that there is a chronological gap betweenc.82507900BP (calibrated), i.e. 6300 and 5950 Cal BC in the occupation of all sites in theDanube Gorges except at the type-site of Lepenski Vir (Srejovi 1969, 1972).

5 In their view,this discontinuity of occupation on the banks of the Danube in the region was caused by the8200BP event (Alley et al. 1997), a global phase of climatic cooling that might have caused

the rise in water-levels of the Danube and frequent flooding of great magnitude and duration ofthe sites along the Danubes shores. The onset of this climatic event would significantlycorrespond with the appearance of the first Early Neolithic settlements in the wider region ofthe Balkans and significant changes in the material culture of the sites in the Danube Gorges(cf. Whittle et al. 2002; Radovanovi 1996; Bori 1999, 2002b). The postulated flooding phasemight have been an important stimulus for culture change (Bonsall et al. 2002, 11).

This scenario is based on evidence of an apparent gap in the distribution of radiocarbonmeasurements from the sites of Padina, Vlasac, Schela Cladovei, Icoana, Ostrovul Banului andOstrovul Corbului (Bonsall et al. 2002). Only at Lepenski Vir, as emphasized by Bonsall et al.,is there continuity of occupation during the period between 63005950 Cal BC confirmed by a

continuous sequence of radiocarbon measurements (see also Bori 1999, fig. 7; 2002b, fig. 5).In their view, the continuity of occupation at Lepenski Vir was due to its conceptual importanceand sacred character for Mesolithic communities in the region; hence the site could not havebeen abandoned or displaced to another location, like other less important sites, despite theflooding. The postulated flooding of the Danube in the period after 6300, according to theseauthors, necessitated the construction of strong, durable floors at Lepenski Vir, which only atthis site were made of concrete-strong limestone, designed to resist flooding (Bonsall et al.2002, 5), while large stone boulders with carved representations of hybrid fish-human beingshad an apotropaic6 function of appeasing mysterious powers responsible for flooding. Theargument with regard to the durability of trapezoidal floors and stone boulders in the Danube

Gorges in resisting floods has already been previously advanced by Chapman (1993, 107). Theauthors rely on the view that the mentioned features could have resisted the incoming floods,whereas smaller artefacts might have been easily washed away (Chapman 1993, 100, 2000,195).

c

cc

c

c

5 Bonsall et al. (2002) use dates expressed in BP (calibrated). We use dates expressed in Cal BC years (see Figs.1415), calibrated with OxCal v. 3.9 (Bronk Ramsey 1995, 2001) using the INTCAL98 calibration curve (Stuiver

et al. 1998). Thus 8200BP (calibrated) is the same as 6250 Cal BC.6 It is not quite clear what kind of meanings Bonsall et al. (2002) evoke by the term apotropaic. The concept of

apotropaic power usually relates to protective qualities (mana) of material objects, visual images and performativepractices against evil eye and other similar magical concepts. For the use of the idea with regard to the possible

instances and significance of apotropaic practices and materialities in the Danube Gorges see Bori 2002a, 2003;Bori and Stefanovi 2004; Stefanovi and Bori in press.cccc

c


22/32


This catastrophic scenario concerning the appearance of well-known features atLepenski Vir is an innovative and exciting venue of research, and it strongly suggests adiscontinuity of occupational histories between Mesolithic and Early Neolithic periods at allother sites in the Danube Gorges but Lepenski Vir. However, there are a number of possible

problems with such a scenario with regard to different lines of evidence that we would now liketo examine more closely.

(a) In previous publications concerning the geological history in the formation ofarchaeological sites in the Danube Gorges, we find no mention of any evidence about the fluvialdeposits that would indicate that any of the examined sites were flooded either seasonally orover the longer term (Markovi -Marjanovi 1969, 1978; Brnnacker 1971). Indeed, Bonsall etal. indicate that the evidence of flooding might have been overlooked and would have left littletrace in sedimentological records (2002, 4). We should mention that small pockets of thinflooding deposits (yellow sterile soil and river pebbles) indeed do appear at Padinas Sectors Iand II (Padina field diaries for 19681970). However, these lenses of fluvial deposition are

scattered across the settlement and appear to relate to the seasonal outburst of mountain creeksthat run into the Danube over the two sectors at Padina (Fig. 2). They do not reflect majorflooding by the Danube. Moreover, the date for the deposition of these fluvial sediments is mostlikely Early Holocene, significantly earlier than the flooding phase postulated by Bonsall et al.(2002). Thus, it does not appear convincing that major floods during the period of c.300 yearswould leave no detectable trace.

Furthermore, a reconstruction of the alterations in water-levels of the Danube based onthe recorded data over the last 150 years indicates that even at the time of the greatest floodsin this region architectural features of the MesolithicNeolithic sites were out of reach of therising waters of the Danube (Bori 2002c, fig. 4.8; contra Chapman 1993, 100, 2000, 195).

Undercutting, i.e. riverbank erosion, of certain features at the lowermost portions ofarchaeological sites is most likely the consequence of more recent changes in the water-levelregimes of the Danube, due to the melioration works in the Great Hungarian Plain over the last300 years draining massive amounts of underground water from the Plain into the Danuberiver system. This process might have caused the rise of the average water-level of theDanube in the region of the Danube Gorges over the last several centuries, and, as a morerecent phenomenon, might have been responsible for the evident riverbank erosion ofarchaeological sites, making visible sections of stratified deposits in the river profile (cf. Bori1999, n. 1, fig. 25).

In sum, we see no direct evidence for major flooding in the region in the period

63005950 Cal BC.(b) Radiometric dates that Bonsall et al. (2002) use to suggest the continuity of

occupation at Lepenski Vir come from the dated charcoal found on the floors of trapezoidalbuildings. These charcoal samples originate from timber beams of post-framed upperconstructions or from the remains of scattered charcoal found in rectangular stone hearths andover building floors (Quitta 1975, 283; Bori 1999, fig. 7, 2002c, Appendix 1). What remainsunclear in the scenario proposed by Bonsall et al. for the construction of durable building floorsat this site to resist floods can be summarized as follows: if the site was periodically floodedduring the period 63005950 Cal BC, including all buildings that were built during this earlyphase, how is it possible that the charcoal used for dating (deposited on the floors of these

buildings), and which provided measurements for the abandonment of these structures with datescovering the period 6300 to 5950 Cal BC, was not washed away by the floods? The same

c

c

c

cc


23/32


question is relevant with regard to the presence of other smaller artefacts and disarticulatedhuman bones found on the floors of these buildings. In addition to these dates on material(charcoal) that would have been washed away in a flood, there are two recent AMS dates (OxA-8725 and OxA-8618, cf. Whittle et al. 2002; Bori 2002b, table 1; see Fig. 14) associated with

trapezoidal floors that cover exactly the period 63005950 Cal BC.(c) In the presentation of old charcoal dates and new AMS dates to suggest the

chronological gap in the sequence, Bonsall et al. (2002, figs. 23) use radiocarbon mean agesdepicted as individual circles/dots. This procedure of visual representation seemsmethodologically inappropriate due to the magnitude of the encountered errors at 1 and 2standard deviations. Thus, if one represents those measurements from Vlasac that are on themargins of either side of the postulated gap in the range at 95 per cent probability as we havedone in Figure 14, the sharp limits of the suggested chronological hiatus between 6300 to 5950Cal BC blur. Furthermore, Bonsall et al., for no apparent reason, omit at least two dates fromVlasac that clearly fall in the range between 6300 and 5950 Cal BC (ibid., fig. 2). These are

Bln-1168 (7475 60bp) and Bln-1954 (7440 60bp) (Srejovi and Letica 1978, 129),indicating dates in the range 6440 to 6220 and 6430 to 6100 Cal BC at 95 per cent probability(Bori 2002c, table 4.7; see Fig. 147).

(d) Most importantly, our new AMS dates from Padina and Hajdu ka Vodenica closethe 63005950 Cal BC gap (Fig. 16). New AMS dates from Padina practically cover thesuggested gap in its entirety, indicating both a final Mesolithic (Burial 1a: OxA-11107 and OxA-11108) and an earliest Neolithic use of the site in the period 6300 to 5950 Cal BC with evidenceof occupation of the following trapezoidal buildings in Sector III: House 12 (OxA-9034), House18 (OxA-9053), House 9 (OxA-9052) and House 17 (OxA-11103). New AMS dates from thissite rule out the possibility of a significant chronological gap between the Mesolithic and

Neolithic periods.As previously indicated for Hajdu ka Vodenica, after the correction of dates made on

human burials for the freshwater reservoir effect, it is possible to suggest that the site wasoccupied during the final Mesolithic phase, but perhaps significantly overlapping with thedevelopment of the Early Neolithic in the wider region of the Balkans, with the dated occupationup to c.6200/6100 Cal BC (Figs. 13 and 1516). Hajdu ka Vodenica, with only four dates, alsocloses the gap. Four dates, however, cannot be viewed as a representative sample from a singlesite. This limitation leads us to question the reliability of five, four and two dates from,respectively, Icoana, Ostrovul Corbului and Ostrovul Banului to demonstrate the discontinuityof occupation (contra Bonsall et al. 2002, fig. 3). Furthermore, new AMS dates are necessary

to date the upper levels of the stone construction at Hajdu ka Vodenica, where Early Neolithicpottery appears stratified above Mesolithic remains, and these new dates would provide firmevidence for a (dis)continuity from the Mesolithic to the Neolithic at this site.

On the basis of the (lack of) flooding evidence, problems with Bonsall et al.sinterpretation of charcoal dates from Lepenski Vir, their omission of some existing dates, aswell as our new AMS dates from Padina and Hajdu ka Vodenica, we reject their scenario ofsite abandonment caused by a catastrophic flooding associated with the 8200BP event.

c

c

c

c

cc

c

c

7 We have not included human bone collagen dates from Padina (Table 1) in Figures 14 and 16. It is reasonable

to assume that these dates on human bones are also affected by the freshwater reservoir effect, and since thereis no information regarding their d15N values we have not been able to apply the correction factor to these dates.


24/32

OXFORD

JOURNA

LOFARCHAEOLOGY

364

B

lackwellPublishingLtd.2004

Figure 14Distribution of available radiometric evidence for the sites situated in the Upper Gorge of the

corrected for the freshwater reservoir effect according to Method 2 as described by Cooket a3.9 (Bronk Ramsey 1995, 2001) using the INTCAL98 calibration curve (Stuiver et al. 1998)

fill: charcoal/conventional dates; white fill: AMS dates; graded sh

Sources:Alibeg: Boronean 2001; Padina: Whittle et al. 2002; Bori 2002b; Bori and Miracle t

Bonsall et al. 1997, table 5; Whittle et al. 2002; Bori 2002b; Vlasac: Sretable 3.

Note: Bonsall et al. (1996, table 1, 1997, table 3, 5, 2000, table 3) report the same Lab ID OxAfrom Lepenski Vir.

ccct


25/32


Figure 15

Distribution of available radiometric evidence for the sites situated in the Lower Gorge and the Klju areas of the

Danube. AMS dates with d15N values >+10 corrected for the freshwater reservoir effect according to Method 2 as

described by Cooket al. (2002, 82). All dates calibrated with OxCal v. 3.9 (Bronk Ramsey 1995, 2001) using theINTCAL98 calibration curve (Stuiver et al. 1998). Solid bars show 1 s.d.; lines show 2 s.d. Black fill:

charcoal/conventional dates; white fill: AMS dates.

Sources:

Cuina Turcului: P unescu 1978;Icoana: Boronean 2001, 2034;Hajdu ka Vodenica: Bori and Miracle this paper,Table 1; Razvrata: Boronean 2001; Ostrovul Banului: Boronean 2001; Schela Cladovei: Bonsall et al.

1997; Ostrovul Corbului: P unescu 1996.att

ccta

c


26/32


Figure 16

Close-up distribution of new AMS dates for Padina and Hajdu ka Vodenica in the period between 75005500 Cal

BC. AMS dates with d15N values >+10 corrected for the freshwater reservoir effect according to Method 2 as

described by Cooket al. (2002, 82). All dates calibrated with OxCal v. 3.9 (Bronk Ramsey 1995, 2001) using theINTCAL98 calibration curve (Stuiver et al. 1998). Solid bars show 1 s.d.; lines show 2 s.d. Black fill:

charcoal/conventional dates; white fill: AMS dates; graded shade: bone collagen.

c


27/32


Furthermore, it could well be that larger series of dates from individual sites, which have beenavailable for Lepenski Vir and are now also available for Padina, would bridge the postulatedchronological gap. Nonetheless, there is a stark discrepancy in the number of radiometric datesthat cover particular periods in the long development of the Danube Gorges Epipalaeolithic,Mesolithic and Early Neolithic sequences (Fig. 17). Individual sites (and also particular areasof respective sites) may exhibit differences in the chronological coverage, i.e. in the intensityof occupation, as well as in the nature of transformation from the Mesolithic to the Neolithic inthis region during the period from 6300 to 5950 Cal BC. However, the newly available seriesof radiometric dates from Padina and Hajdu ka Vodenica strongly suggest a continuous

occupation during this period without major chronological gaps at the two sites.

conclusion

New AMS dates from Padina indicate a significant chronological depth for the use ofthis site (this also applies to some other sites in the Danube Gorges, such as Vlasac, cf. Fig. 14)as both a burial ground and a seasonal or semi-permanent settlement, at least from c.9500 CalBC. There are significant chronological differences in dating individual human burials, scatteredresidual remains and remains found within architectural features, such as trapezoidal buildings.For the moment, it is difficult to claim unbroken continuity for almost 4000 years, i.e. during

the MesolithicNeolithic development that the available radiometric dates cover at this site, andwe suspect that there were discontinuous phases and (seasonal) episodes of occupation.

c

Danube Gorges Mesolithic-Neolithic (101 dates)

0

2

4

6

8

10

12

14

16

18

5500

6000

6500

7000

7500

8000

8500

9000

9500

10000

10500

11000

11500

12000

12500

13000

Cal BC

No

14C

Dates

First Pottery

"8200 BP Event"

Epipalaeolithic

Early Mesolithic:

Upper Gorges

Late Mesolithic:

Upper+Lower

Gorges

Figure 17

Quantitative distribution of radiometric evidence in the Danube Gorges over the chronological span 13000 to 5500

Cal BC (for sources see Figs. 1415).


28/32

However, it seems that populations which visited this site over several millennia of theMesolithic development belonged to the same material culture/regional tradition throughout theperiod, with a good knowledge of the significance of the locale. Furthermore, it seems that thereis no major chronological break in the sequence of radiometric dates at the time when new

material paraphernalia, such as pottery, Balkan yellow-spotted flint and polished stone tools,appear in this region as general traits of Early Neolithic material culture tradition. We see noevidence to support a recently suggested scenario of flooding that would have caused theabandonment of this site in the period 63005950 Cal BC (Bonsall et al. 2002) since a numberof new AMS dates cover exactly this period.

New dates from Hajdu ka Vodenica suggest the Mesolithic use of the site as a burialground from at least c.7500 Cal BC. There is a possible gap of around 500 years between theearlier use and a later use of the site as a burial ground around 6500 to 6100 Cal BC. However,there might have been a clear recognition of the previous use of the site on the basis ofstratigraphic and spatial relations of earlier and later dated burials and architectural features.

Material culture associations at the site related to the lower two portions of stone constructions(without pottery and with abundant bone tools) also suggest that the site might have been usedas an occupation settlement/base during the Mesolithic, which should be determined by futuredating of bone tools and animal bones from this area.

This revised chronological framework on the basis of new AMS dates from Padina andHajdu ka Vodenica also significantly helps us understand the nature of the Meso-Neolithictransformations in this region, with abundant indications regarding the chronological overlappingand cultural coexistence of Mesolithic and Neolithic realms (Bori in press a; in press b).

Further research and new AMS dates from both sites are necessary to achieve a fine-grained resolution of the chronological sequences and to discern particularities of very long and

complex occupational histories at these and other locales in the Danube Gorges.

Acknowledgements

We would like to thank the National Environmental Research Council (UK) for funding newAMS measurements from Padina and Hajdu ka Vodenica. We are grateful to Borislav Jovanovi and theArchaeological Institute in Belgrade for their kind permission to sample osteological material from thetwo sites. We also thank Vesna Dimitrijevi and Borislav Jovanovi for comments on earlier versions ofthis paper.

(DB) Department of Anthropology

Columbia University452 Schermerhorn Ext.

1190 Amsterdam Avenue, MC 5538New York, NY 10027

USAE-mail: [email protected]

(PM) Department of ArchaeologyUniversity of Cambridge

Downing Street

Cambridge CB2 3DZE-mail: [email protected]

cc

cc

c

c

c



29/32

references

alley, r.b., mayewski, p.a., sowers, t., stuiver, m., taylor, k.c. and clark, p.u. 1997: Holoceneclimate instability: a prominent widespread event 8200yr ago. Geology 25, 4836.

bonsall, c., boronean , v. and srejovi , d. 1996: AMS Radiocarbon Determinations on Human Bonefrom Lepenski Vir, Vlasac and Schela Cladovei. Mesolithic Miscellany 17(2), 610.

bonsall, c., lennon, r., mcsweeney, k., stewart, c., harkness, d., boronean , v., bartosiewicz,l., payton, r. and chapman, j. 1997: Mesolithic and early Neolithic in the Iron Gates: a palaeodietaryperspective.Journal of European Archaeology 5(1), 5092.

bonsall, c., cook, g., lennon, r., harkness, d., scott, m., bartosiewicz, l. and mcsweeney, k. 2000:Stable isotopes, Radiocarbon and the MesolithicNeolithic Transition in the Iron Gates. DocumentaPraehistorica 27, 11932.

bonsall, c., macklin, m., payton, r. and boronean , a. 2002: Climate, floods and river gods:environmental change and the Meso-Neolithic transition in southeast Europe. Before Farming 34(2),115.

bori , d. 1999: Places that created time in the Danube Gorges and beyond, c.90005500 BC.DocumentaPraehistorica 26, 4170.

bori , d. 2002a: Apotropaism and the temporality of colours: colourful MesolithicNeolithic seasons inthe Danube Gorges. In Jones, A. and MacGregor, G. (eds.), Colouring the Past: The Significance of Colourin Archaeological Research (Oxford), 2343.

bori , d. 2002b: The Lepenski Vir conundrum: reinterpretation of the Mesolithic and Neolithic sequencesin the Danube Gorges.Antiquity 76, 102639.

bori , d. 2002c: Seasons, Life Cycles and Memory in the Danube Gorges, c.100005500 BC Cal(Unpublished Ph.D. Dissertation, University of Cambridge, Cambridge).

bori , d. 2003: Deep time metaphor: mnemonic and apotropaic practices at Lepenski Vir.Journal ofSocial Archaeology 3(1), 4674.

bori , d. in press a: Fuzzy horizons of change: Orientalism and the frontier model in the Meso-Neolithictransition. In Milner, N. and Woodman, P.C. (eds.), The Mesolithic: new directions and alternativenarratives (Oxford).

bori , d. in press b: Deconstructing essentialisms: unsettling frontiers of the Meso-Neolithic Balkans. InBailey, D. and Whittle, A. (eds.), Unsettling the Neolithic (Oxford).

bori , d. and stefanovi , s. 2004: Birth and death: infant burials from Vlasac and Lepenski Vir.Antiquity78, 582601.

bori , d., grupe, g., peters, j. and miki , . in press: Is the MesolithicNeolithic subsistence dichotomyreal? New stable isotope evidence from the Danube Gorges. Antiquity (submitted).

boronean , v. 2001: Paleolithique superieur et epipaleolithique dans la zone des Portes de Fer(Bucure ti).

bronk ramsey, c. 1995: Radiocarbon calibration and analysis of stratigraphy: the OxCal Program.Radiocarbon 37(2), 42530.

bronk ramsey, c. 2001: Development of the radiocarbon program OxCal.Radiocarbon 43(2A), 35563.

brnnacker, k. 1971: Geologisch-pedologische Untersuchungen in Lepenski Vir am Eiserner Tor. InSchwabedissen, H. (ed.), Die Anfnge des Neolithikums vom Orient bis Nordeuropa, Teil 2: stlichesMitteleuropa (Kln/Wien, Fundamenta, Monographien zur Urgeschichte, Reihe A, Band 3), 2032.

burleigh, r. and ivanovi , s. 1980: Radiocarbon dating of a Cro-Magnon population from Padina,Yugoslavia, with some general recommendations for dating human skeletons.Zeitschrift fr Morphologieund Anthropologie 70(3), 26974.

chapman, j.c. 1993: Social Power in the Iron Gates Mesolithic. In Chapman, J. and Dolukhanov, P. (eds.),Cultural Transformations and Interactions in Eastern Europe (Aldershot), 71121.

cz

st

zcc

cc

c

c

c

c

c

c

c

t

t

ct



30/32

chapman, j.c. 2000: Fragmentation in archaeology. People, places and broken objects in the prehistoryof south-eastern Europe (London and New York).

cook, g., bonsall, c., hedges, r.e.m., mcsweeney, k., boronean , v., bartosiewicz, l. and pettitt,p.b. 2002: Problems of dating human bones from the Iron Gates. Antiquity 76, 7785.

grupe, g., manhart, h., miki , . and peters, j. 2003: Vertebrate food webs and subsistence strategiesof Meso- and Neolithic populations of central Europe. In Grupe, G. and Peters, J. (eds.), DocumentaArchaeobiologiae 1. Yearbook of the State Collection of Anthropology and Palaeoanatomy, Mnchen,Germany (Rahden/Westf.), 193213.

jovanovi , b. 1966: Sculptures de la ncropole de lage du fer ancien Hajdu ka Vodenica.ArchaeologiaIugoslavica 7, 314.

jovanovi , b. 1967: Nekropola starijeg gvozdenog doba u Hajdu koj Vodenici. StarinarXVIII, 18192.

jovanovi , b. 1969a: Chronological Frames of the Iron Gate Group of the Early Neolithic Period.Archaeologica Iugoslavica 10, 2338.

jovanovi , b. 1969b: Hajdu ka Vodenica naselje i nekropola starijeg gvozdenog doba. Arheolo ki

pregled11, 927.jovanovi , b. 1969c: Hajdu ka Vodenica. In Bo kovi , Dj. (ed.), Stare kulture u Djerdapu (Beograd,SANU).

jovanovi , b. 1984: Hajdu ka Vodenica, praistorijska nekropola. StarinarXXXIIIXXXIV (19821983),30513.

jovanovi , b. 1987: Die Architektur und Keramik der Siedlung Padina B am Eisernen Tor, Jugoslawien.Germania 65(1), 116.

lanting, j.n. and van der plicht, j. 1998: Reservoir effects and apparent ages. The Journal of IrishArchaeology 9, 15165.

markovi -marjanovi , j. 1969: Geologija i stratigrafija. In Srejovi , D., Lepenski Vir Nova

praistorijska kultura u Podunavlju (Beograd, Srpska knji evna zadruga), 18592.

markovi -marjanovi , j. 1978: Geologija i stratigrafija. In Gara anin, M. (ed.), Vlasac mezolitskonaselje u Djerdapu Vol. II (Beograd, Srpska akademija nauka i umetnosti), 1127.

miki , . and sladi , m. 1994: Ostaci mezolitskog oveka iz Kule u -Derdapu.Zbornik Filozofskog fakul-teta 18, 3741.

p unescu, al. 1978: Cercetarile archeologice de la Cuina Turcului-Dubova (Jud. Mehedinti). Tibiscusistorie. Volum inchinat celei de-a 60 a aniversari a Unirii, 1156.

p unescu, al. 1996: Ostrovul Corbului. Die mesolithische Siedlung der Schela Cladovei-Kultur 1.b. InHauptmann, H. and Roman, P. (eds.), Ostrovul Corbului (Bukarest),155214.

quitta, h. 1975: Die Radiocarbondaten und ihre historische Interpretation. In Srejovi , D.,Lepenski Vir.Eine vorgeschichtliche Geburtssttte europischer Kultur(Bergisch Gladbach), 27285.

radovanovi , i. 1996: The Iron Gates Mesolithic (Ann Arbor, International Monographs in Prehistory,Archaeological Series 11).

roksandi , m. 1999: Transition from Mesolithic to Neolithic in the Iron Gates gorge: Physicalanthropology perspective (Unpublished Ph.D. dissertation, Department of Anthropology, Simon FraserUniversity, Vancouver).

roksandi , m. 2000: Between Foragers and Farmers in the Iron Gates Gorge: Physical AnthropologyPerspective. Djerdap Population in Transition from Mesolithic to Neolithic.Documenta Praehistorica 27,1100.

schulting, r.j. and richards, m.p. 2002: Dogs, ducks, deer and diet: New stable isotope evidence on

early Mesolithic dogs from the Vale of Pickering, North-east England.Journal of Archaeological Science29, 32733.

c

c

c

c

a

a

cczc

scc

z

ccc

c

cc

cscc

scc

c

cc

cc

zc

t



31/32

srejovi , d. 1969: Lepenski Vir Nova praistorijska kultura u Podunavlju (Beograd, Srpska knji evnazadruga).

srejovi , d. 1972:Europes First Monumental Sculpture: New Discoveries at Lepenski Vir(London).

srejovi , d. and letica, z. 1978: Vlasac. Mezolitsko naselje u Djerdapu (I arheologija) (Beograd, Srpskaakademija nauka i umetnosti).

stefanovi , s. and bori , d. in press: The new born infant burials from Lepenski Vir: in pursuit of

contextual meanings. In Bonsall, C., Boronean , V. and Radovanovi , I. (eds.), The Iron Gates Prehistory:New Perspectives (Oxford, BAR Int. Ser.).

stuiver, m., reimer, p.j., bard, e., beck, j.w., burr, g.s., hughen, k.a., kramer, b., mccormac, g.,van der plicht, j. and spurk, m. 1998: INTCAL98 Radiocarbon Age Calibration, 24000-0 cal BP.Radiocarbon 40(3), 104183.

vasi , r. 1986: Compte-rendu des fouilles du site prehistorique Velesnica. 19811982. -Derdapske sveske3, 26480.

whittle, a., bartosiewicz, l., bori , d., pettitt, p. and richards, m. 2002: In the beginning: new

radiocarbon dates for the Early Neolithic in northern Serbia and south-east Hungary.Antaeus 25, 63117.ivanovi , s. 197374: Prvo saop tenje o rezultatima antropolo kog prou avanja skeletnih ostataka sa

Padine u Djerdapu. Starinar2425, 13953.csscz

c

c

ct

cc

c

c

zc



32/32

Boric&Miracle2004-AMS-Padina i H Vodenica

Documents