Micro-regions of the Lepenski Vir culture, Padina in the Upper Gorge and Hajducka Vodenica in the Lower Gorge of the Danube

8/13/2019 Micro-regions of the Lepenski Vir culture, Padina in the Upper Gorge and Hajducka Vodenica in the Lower Gorge of

1/36

1

UDK 903'12\'15(4\5)''633\634''>575.17< 903'12\'15(4\5)''633\634''>574.91

Documenta Praehistorica XXXV (2008)

Micro-regions of the Lepenski Vir culture>Padina in the Upper Gorge and Hajdu;ka Vodenicain the Lower Gorge of the Danube

Borislav JovanovicSerbian Academy of Arts and Sciences, Belgrade, Serbia

[email protected]

Introduction

Viewed from a distance of nearly forty years (19702007), the excavations at Padina and Hajduka Vo-denica were done in the same way as those that areusually conducted at prehistoric lowland and hill-fort sites. To date, the results have been publishedfragmentarily, which has contributed to conflictingviews on the origins and chronologies of these twosites. Such discussions were mostly concerned withunique features of the Lepenski Vir culture, such asstone sculpted boulders, specific architecture andtypical settlement organization situated at specially

selected locations (Fig. 1). Evaluations of the archa-eological content of such sites based on stratigraphicand typological grounds alone has proven inade-

quate among other reasons, due to the lack of rele-vant analogies in the prehistory of the Balkans andsouth-eastern Europe (Banffy 2004.36769). Thestagnation in the study of the Lepenski Vir culture,however, was avoided largely thanks to the activi-ties of the Neolithic Seminars that have been conti-nuously held at the Department of Archaeology, Ljub-ljana University, and publications of theDocumentaPraehistorica series.

The immediate surroundings of the Late Mesolithic

and Early Neolithic in the Upper Gorge of the Da-nube was the winding coastal zone, i.e. the space be-tween the steep cliffs of the hinterland and the Da-

ABSTRACT This paper was compelled into existence when confronting the fact that the most com-monly discussed transitional period connecting the Late Mesolithic and the Early Neolithic in theerdap or the Iron Gates Gorges (present-day Serbia and Romania) is determined by the often em-phasized lack of adequately published evidence, thus leaving this period largely unknown. We exam-ine previous conclusions concerning the Lepenski Vir culture, or at the very least, try to put such dis-cussions back in their archaeological context. We discuss the features that are essential for the organ-isation of the settlements, stratigraphy, and adaptability of architecture caused by geomorphologyand climate, as well as the remarkable loss of settlement space due to different types of erosion.

IZVLEEK lanek je nastal kot odgovor na dejstvo, da pogoste diskusije o prehodu med mezolitikomin neolitikom v soteski erdap (v dananji Srbiji in Romuniji) ne temeljijo na ustrezno objavljenihrezultatih izkopavanj in katalogih najdb. V lanku reanaliziramo interpretacije kulture Lepenski virin jih s pomojo rezultatov naih izkopavanj umestimo v arheoloke kontekste. Predstavljamo nam-re kljune znailnosti prostora, ki so vplivale na organiziranost naselij in arhitekturo, prilagojenogeomorfolokim in klimatskim pogojem. Predstavljamo tudi krenje njihovih obsegov in spremenje-nih stratifikacij kot posledic velikega tevila erozij.

KEY WORDS Late Mesolithic; Neolithic; Iron Gates Gorges; Lepenski Vir culture; settlement struc-tures; stratigraphy


2/36

Borislav Jovanovic

2

nube basin. Such a position, narrowed by the rockycliffs and the course of the river was the site of var-ious types of erosion (Fig. 3). The geomorphology ofthe Danube Gorges must be emphasized, with slight

differences visible everywhere in this landscape which restricted the settlement space for both theLate Mesolithic populations and the Lepenski Vir cul-ture. After the abandonment of these settlements,the impact of erosion significantly altered the origi-nal configuration of the terrain, and to a large extentreduced the original size of settlements. Hence, forthe stratigraphic determination of architectural fea-tures and portable material culture, it is necessary toestimate the extent of the preservation of differentsettlements at Sectors IIV at Padina (Jovanovi2004.6374).

Longitudinal and transversal erosions are related tothe dynamics of the Danube River, as well as smal-ler rivers, streams and seasonal torrents in the im-mediate vicinity. Their channels, full of cascadescarved in the rocky base, crossed sidewise settle-ment spaces, occasionally producing floods of de-structive power. The simplified classification of ero-sion dynamics we recognized in the stratigraphic se-quence at Sectors IIII of Padina is as follows:

1. Erosion caused by the Danube (erosion activity

visible on river banks; changing river courses).2. Erosion caused by seasonal torrents (outbursts

that caused large quantities of sediment to bewashed down the steep cliffs).

3. Erosion caused by short watercourses (water inestablished channels; from the spring to theirconfluence with the Danube, altering the direc-tion of the river course).

1. Eolic and atmospheric erosion in the depositionof anthropogenic layers (mini erosions within a

settlement of upper and lower terraces, be-tween buildings).2. Stratigraphic significance of archaeological finds

that were impacted by the Danube River erosion(accumulated or redeposited artefacts next tothe cleaned river banks).

3. Stratigraphic significance of archaeological findsthat were impacted by seasonal torrents (destru-ctive power that such torrents had within semi-subterranean buildings and crofts, accumulatedarchaeological finds).

Stratigraphic significance of archaeological findsat palaeo-surfaces and occupation levels (settle-ment surfaces placed on different terraces).

Relative chronological significance of 14C results incomparison to the stratigraphy (inconsistencies be-tween 14C results and their stratigraphic position).

Padina. Gospoin Vir (Upper Gorge)

The site of Padina consists of several settlements lo-cated at horseshoe shaped coves downstream fromthe Gospoin Whirlpool (Sectors IIII). The earliestphase here belongs to the Late Mesolithic of the IronGates, followed by the entire development of the Le-penski Vir culture. After a long hiatus, the Late Neo-lithic Kostolac culture, which at the time spread ac-ross the Danube and the Sava river basins, is thenext cultural manifestation at this site, finally fol-

lowed by the Early Iron Age Kalakaa culture. Spo-radic finds have been assigned to the Roman andMedieval periods.

Sectors I to III were connected, while the signifi-cantly larger Sector IV looked like a spacious pla-teau, bordered with steep cliffs. Various geomorpho-logical characteristics affected formation processes:Sector Ia, which was almost at the same level as theDanube River, was significantly affected by high wa-ters (Fig. 2).

Padina A. Sector I, Trench 1, Profile 1

The Late Mesolithic period. Sector I consists of twostepped stony terraces: the lower is bounded by thehigh river bank profile, at the base of which is a pa-laeosoil sporadically covered by river sand. Partlypreserved remains of architectural features at thissector provided a reliable vertical stratigraphy. De-pressions in the bumpy rocky base surfaces of bothterraces, as well as of segments 1 and 2 in front ofthe river bank profile, are largely filled with this

palaeosoil of intense black colour. It contains LateMesolithic flint and bone artefacts. No Early Neoli-thic Starevo pottery is found at this level (Fig. 4).

Padina A B. Base of a building structure.Phase of contacts. Iron Gates Late Mesolithic,the Early Neolithic Starevo culture

The layer of river-deposited sand in segments 1 and3 extends almost to the high river bank profile. Insegments 23, this deposit covers the flattened baseof an architectural feature from the final stage of the

Late Mesolithic the phase of contact. A long rectan-gular hearth, made from vertically placed stoneblocks, was found in the central part of this base. In


3/36

Micro-regions of the Lepenski Vir culture> Padina in the Upper Gorge and Hajdu;ka Vodenica in the Lower Gorge of the Danube

3

the vicinity, there is a scatter of split and slab-likestones left behind on the levelled palaeosoil. Only afew stone and bone artefacts were found here, whilethe Early Neolithic Starevo pottery is absent (Jova-

novi 1971.Pl. III, Figs. 1, 5, 6).Padina B1. House 1. The Lepenski Vir culture

On a narrow ridge of one of the cliffs that slopes to-ward the river (segment 3), a rectangular hearth ofHouse 1 was found, directly above the previouslydescribed feature dated to the contact phase. Thehearth platform was composed of two parallel stoneslabs, while between them was a partly damagedboulderaltar in situ occupying a standard placewhen it comes to architectural features of all phases

of the Lepenski Vir culture. Here, in the washed-outlayer around the platform of House 1, only a fewfragments of coarse and monochrome ware werefound (Fig. 6).

Padina B2. House 2. The Lepenski Vir culture

Houses 2 and 3 mark the next two building horizons.These were semi-subterranean structures dug intothe culture layers deposited on two rocky terracesup to a depth of 2m. The terraces slope toward theDanube. In other words, segment 1 represents a se-

ries of settlements located in a small arc-shaped cove.These settlements were built on river terraces, likehill-top sites, where their downtowns are placed ontheir steep slopes all the way to the base of the hill.

The riverbank profile of Sector I, which was the re-sult of river erosion, in segments 1 and 2 exposesthe trapezoidal base of the semi-subterranean House2, assigned to the Lepenski Vir culture. The hard-beaten clay floor and a rectangular hearth with ananiconic boulder were found next to the back of the

platform. The floor of House 2 levelled the palaeo-soil, i.e. the Late Mesolithic layer found on terrace 1.This layer is only sporadically preserved next to therocky base, and contains bone tools, and no pottery.On the floor of House 2, similar bone artefacts wereassociated with Starevo-Cris pottery (Jovanovi1971.T. V.1) (Fig. 7).

Padina B2. House 3. The Lepenski Vir culture

Although two buildings on these terraces were notbuilt at the same time House 3 is definitely youn-

ger their end was simultaneous. They were de-stroyed by a torrent wave; and marked in the sec-tion of this Sector as layer 3b. It had prominent con-

tents of torrent-like washed-out materials: depositedmud and a light yellow deposit of gravelly sand. Thetorrents thrust crossed both terraces diagonally, hit-ting with full force across the area of the current ri-

ver bank to the old channel of the Danube, split bothterraces diagonally, and continued with full strengthdown to the present river bank area into the Da-nube. Since House 2 was built at the highest spot ofthe settlement, it was the first to be hit by the tor-rent: its walls of piled stones were cut through, whileits simple ridged wooden roof was destroyed. Therectangular hearth of stone blocks was moved fromits foundations and transported down the slope be-tween the terraces and stopped by some kind ofbarrier, but in the inverse position from the one inthe base of House 3. It is unusual that the hearth

construction was preserved, as well as both A-sup-ports, one on each of the longer, lateral sides (Fig. 8).On the floor of House 3, a semi-subterranean con-struction next to the rocky edge of terrace 2, frag-ments of coarse pottery were found, as well as acomplete model of an oven (25 x 35cm) decoratedby impresso technique (Fig. 30 A).

After this catastrophic event, these buildings on ter-races 1 and 2 were not renewed. However, it did notmean that life had stopped down at the coastal partof the settlement, which was at this time covered by

river sand or washed out down to the rocky base (aswas the case at Sector Ia).

Padina B2. Pottery workshop between Houses2 and 3

One more feature found between Houses 2 and 3 re-lates to settlement horizon B2. This was probably alarge, open-air hearth, ellipsoid in shape, paved withsmall flat stones, and encircled by a wider ring ofpiled stones (1.50m by 1.10m). The hearth and the

area around it were carefully levelled. Pieces of car-bonized wood and ash were found, along with clus-ters of larger fragments of coarse pottery. One suchcluster allowed a reconstruction of a typical pot-am-phora, ornamented with impresso technique, andwith symmetrical placed handles, similar to those ofthe Starevo-Krs-Cris cultural complex. This fea-ture is most likely a pottery kiln, or an open-air (?)hearth with or without a dome. This might havebeen a workshop with a simple hearth, where pre-pared pots were covered by wooden brunches, andfired in a pyre. It is the first indication of the exis-

tence of local pottery workshops in the Lepenski Virculture.


4/36

Borislav Jovanovic

4

Padina B occupation level

As previously shown, the architectural features ofthe upper and lower terraces fairly precisely define

the occupation level in the immediate vicinity ofHouses 2 and 3. The croft of House 2 was also clear-ly outlined in segment 2 of the riverbank profile,and can also be seen in the longitudinal sections ofTrench 1 (quadrants 1 and 2). The occupation levelsof the croft and of the floor of House 2 are identi-cal, which is also confirmed by a round base madeof split stones found at the same level in quadrant1a. The surface level of the occupation zone outsidethe semi-subterranean croft is identical to the levelfrom which the trapezoidal base of House 2 wasdug, and this is clearly visible on segment 1 of the

riverbank section (Fig. 7).

In this way, there is a more secure stratigraphic de-termination of surface activities for the period whenHouses 2 and 3 were in use. This horizon yields alarger concentration of finds in comparison withthe culture layer above and below the given level.Leaving aside the division of the Late Mesolithic andthe Lepenski Vir culture settlements on the erosivezones of both terraces and the costal zone, it is pos-sible to comprehend the connections between theseoccupation levels of the terrain, which sloped even

more steeply toward the Danube in those days. Ar-chaeological material was in this way washed awayfrom the occupation levels upslope and deposited atthe base of the terrace, in the costal zone. This pro-cess caused inverted and complex stratigraphy inthe costal zone, characterized by inverted 14C dates(see below). This inverted stratigraphy of absolutedates was noticed at Padina immediately after thefirst 14C results in 1978 (Clason 1980). The inversestratigraphy was also noticed in the costal zone ofSector III, after it was realized that every higher ter-

race with semi-subterranean buildings of the Lepen-ski Vir culture had been gradually depositing itswaste on the lower horizon of abandoned buildings.Thus, some finds from Sector I, such as a cult table,were refitted from pieces that originated from twodifferent culture levels (segments 3 and quadrant 1a)(Fig. 30 B).

The dynamic of formation processes at Padina ena-bled definitions of different zones of the settlementon the basis of an approximate order by which thearchaeological material was deposited. The stratig-

raphy and relative chronology of this sector can bereduced to the order of occupation levels of succes-sive living horizons. The absolute heights of the liv-

ing horizon start from architectural features in or-der to connect a particular cultural layer with thebase of a feature, i.e. ideally its floor.

Padina. Sector I. Relative chronology of thesettlements

It is obvious that occupation levels at settlementsbased on slopes are not characterized by signifi-cant horizontal deposits, since the cultural layersare formed around horizontally placed buildings.The occupation layer should be considered as theensemble of single surfaces inserted and placed onthe steep sides of the gorge, or secondarily connec-ted to the costal zone. The cultural layer of Sector Irests on a natural, rocky base, which prevented the

pit-digging that characterizes settlements located onloess or clay soils.

Stratigraphy at Sector I begins with costal palaeo-soil, which contains Late Mesolithic finds. On thisthin layer of palaeosoil, with an average thickness of0.30m, the first Mesolithic dwellings were placed inpockets of the cliffs and on the sloping platforms ofthe rocky base. Such dwellings were more promi-nently present at Sector II.

Another important aspect related to Sector I was de-

termining the precise stratigraphic position of therectangular hearth connected to phase Padina AB.This architectural feature enables us to connect theheights of the stone-lined hearth and the occupationhorizon around it. The entire feature covers the le-velled zone of the palaeosoil (Mesolithic cultural la-yer), and does not reach down to the rocky base ofsegment 2 of the costal zone (Fig. 5). This leveling,made here before the feature was built, confirms thatthe hearth (or oven) covers undisturbed palaeosoilwith Late Mesolithic occupation residues. It is un-

clear how much time elapsed between the Late Me-solithic occupation in this zone and the period whenthis later feature was built, although there is perhapssome continuity of occupation. There is no evidencethat this feature was built by digging into the ground.It is probable that the period that followed the LateMesolithic occupation was characterized by an archi-tectural form different from the preceding period asmuch as from the architecture built during the en-suing phase of the Lepenski Vir culture.

The anthropogenic layer formed over this atypical

architectural feature upon its abandonment was for-med after a considerable temporal hiatus, since atthis location one finds a building with recognizable


5/36


5

elements of early the Lepenski Vir culture. This buil-ding (No. 1) had only the rear platform of the hearth,with two stone slabs and an ornamented boulder be-tween them. Between this surprisingly well-preser-

ved architectural detail of the building (

1.50m)and the feature from the contact phase (AB 2.80m) is a cultural layer 1.28m thick (Fig. 56).

When it comes to the architecture of the LepenskiVir culture at Padina, the level of the hearth andfloor mark the average depth of digging for a parti-cular building around 0.85m (cut of House 2 inTrench I, quadrant 16). From this, one can indirect-ly calculate the accumulation of cultural layers: thedifference between the floors of the contact phasefeature and the Lepenski Vir culture building (House

1) is 1.28m. One can then add the average depth forthe latter feature of 1.50m. These two values indi-cate the build-up of the occupation layer at this pointin the settlement (2.75m), which can then be chec-ked against the section of Sector I, Profile I, seg-ments 23. One notices an accelerated developmentof the settlement during the contact phase (AB)(Fig. 7).

Three settlement horizons the Late Mesolithic, con-tact phase and the early Lepenski Vir culture phase were preserved only in the coastal zone (segments

13). This also means that Sector I represents theperiphery of these settlements, which spread up tothe steep cliffs of the gorge. Unfortunately, the cen-tral area of these settlements used to be in the pre-sent-day Danube channel. Architectural features(Houses 2 and 3) at both terraces were well-elabo-rated, containing elements from sedentary Balkan-Danubian Early Neolithic communities, such as pot-tery and polished stone tools. However, it is lessclear what caused the abandonment of the youn-gest settlement horizon of the Lepenski Vir culture.

It seems this might have been caused by local clima-tic changes (some isolated catastrophe in the case ofHouse 18), but even more probably by a more signi-ficant series of events, which also might have cau-sed the abandonment of other settlements of theLepenski Vir culture that were concentrated in sucha narrow zone of the Upper Gorge of the Danube.

The pottery from Sector I is important for two rea-sons: its first appearance at the site dates to the be-ginnings of the Lepenski Vir culture, and, its typolo-gical classification to a special (Iron Gates) variant of

Early Neolithic culture known as the Starevo andCris culture complex.

The stratigraphic significance of this pottery is deter-mined on the basis of its context in Sector I, anddivided into the following zones:

Pottery of the coastal zone (Profile I, segments13), which was deposited secondarily by hill-washerosion from terraces 1 and 2, as well as by the ad-ditional accumulation of pottery finds from erodedblocks of Profile I undercut by the Danube. The pot-tery that was washed down from the terraces be-longs to various levels of the cultural layer, or fromdamaged and abandoned buildings; the pottery thatcame from the eroded profile I could be connectedto older levels of the cultural layer, characterized bya mixture of archaeological material from both ter-races. The chronological classification of these find-

ings is primarily based on typology and is thus notcompletely reliable.

Pottery (in much smaller amounts) from architec-tural features, found both on and beneath the floors,around open-air hearths or ovens, as well as in situpots these can be considered as closed finds; inaddition, those finds connected with the occupationhorizon can be said to have a slightly less prominentstratigraphic significance.

Early Neolithic pottery from Padina leaves the im-

pression of a coherent typological and ornamentalwhole. The narrow range of ornamental motifs andtechniques is noticeable when compared with theStarevo culture settlements outside of the confinesof the Danube Gorges. The pottery assemblage atSector I is dominated by coarse ware, weakly tempe-red, with chuff and larger sand grain inclusions. Thedominant form is the large open dish with a rectan-gular hollow pedestal with openings, as well as chro-nologically younger specimens with flat bases andshallower receptacle. Spherical pots and deeper sphe-

rical bowls and amphorae are less numerous, butwere of better quality and temper.

Most of the ornamentation of these vessels involvedtwo techniques: coarsening or engraving surfaces,i.e. polishing. The first technique relies on impres-sions made by fingertips or semi-circular nail/im-pression (impresso), producing various combina-tions of horizontal and vertical impressions, oftenshowing the recognizable motif of a wheat ear. So-metimes, these imprints were made with the tip ofa wooden or bone tool. Thin lines, cut by a sharp

blade, form asymmetrical sheaves or simple geome-trical motifs, occasionally coupled with a series oftriangles. The second technique, used on mono-


6/36

Borislav Jovanovic

6

chrome vessels by polishing their surfaces, was con-fined to bowls and amphora-like vessels.

Ornamentation by painting or barbotine (the surface

of the vessel is channelled with irregular plastic ap-pliqus imitating tree bark), widespread across thearea of the Starevo culture, were not found on pot-tery from the Lepenski Vir culture settlements at Pa-dina Sectors IIII.

Horse-shoe polished stone axes, which characterizethe Starevo culture, were found in pottery horizonsat all 3 Sectors of Padina, mostly outside closed con-texts (Antonovi 2004.Fig. 46). Cult objects fromthe Lepenski Vir culture settlement at Sector I werefew and had different roles. The aniconic boulder

from House 2 and damaged boulder-altar (ornamen-ted) from House 1 were the only finds of this kindin buildings. Admittedly, the hearth of House 3 wasmoved from its original location, and the presenceof the boulder-altar (on the rear platform) is notcompletely certain. However, an ornamented mo-del of an oven (Fig. 30) with close typological linksto a similar model from Ajmana, the settlement ofthe Starevo culture in Klju (lower Danube region)(Stalio 1986.Fig. 13), comes from the floor of thisbuilding.

Padina. Sector I. Relative-chronological division

Settlement of the Iron Gates Late Mesolithic

As previously explained, the occupation layer of theLate Mesolithic settlement is primarily related to pa-laeosoil full of organic material and discarded arte-facts. The most important among these finds wereflint blades and flakes, as well as fragmented bonetools found beneath the rectangular hearth. This la-

yer had lenses of charcoal and burning, distingui-shed from the strip of burned soil under the fire-box. The hearth was placed on the palaeosoil. In thelayer under the rectangular hearth, no pottery wasfound. A similar context was found beneath the part-ly damaged floor of House 2; the layer contained flintand bone artefacts found in the palaeosoil depositedover the rocky base.

Architectural feature with elongated rectan-gular hearth. Profile I. Segments 23

We have already explained the stratigraphic positionof this feature found between House 1 of the Lepen-ski Vir culture (segment 3) and the occupation ho-

rizon of the Mesolithic settlement (palaeosoil) (seg-ments 23). The functions as well as constructiondetails of this feature remain unclear. The interiorof the feature has no partitions, unless the piles of

stone slabs were not the remains of partitioningwalls. There were no traces of a floor or any kind ofsubsidiary stone construction. The elongated hearthhas analogies only in the architecture of Vlasac andHajduka Vodenica (the latter in the Lower Gorge ofthe Danube, see below), although in a different con-text (Figs. 5 and 6).

Architectural elements cult place of House 1.The Lepenski Vir culture. Profile 1

Leaning against the steep cliff, which limits the

downstream edge of the cove, the base of House 1was partially preserved, with the rear platform ofthe hearth and ornamented boulder-altar. On theone hand, this architectural detail dates House 1;and on the other, it testifies the long duration of theearly phase of the Lepenski Vir culture settlementto which this building belongs. The orientation ofHouse 1 did not deviate from the generally accept-ed rule found in the Upper Gorge: in all settlementsof this culture, the buildings faced the Danube withtheir (wider) frontages most commonly at a right-angle or slightly deviating, depending on the topo-

graphy of a particular terrain. It has been mentio-ned previously that the approximate dimensions ofthe base of House 1 cover the whole gauge of the ol-der feature with the rectangular hearth. In this way,a clear stratigrafic relationship was established be-tween two architectural features, or two differentphases in the development of the Lepenski Vir cul-ture (Fig. 5).

House 2. The Lepenski Vir culture. Trench 1,quadrant 1b

The floor of this building was situated on a levelledsurface of the palaeosoil. The floor, made of fineclay, was not preserved entirely. The trapezoidalbase, which was fairly accurately constructed, hada rectangular fireplace with slab-free firebox embed-ded on its front. The ash-place with massive slabs ar-ranged fan-like, extended to the front (entrance) areaof the base. On the cross-section of the frontal ofHouse 2 and segment 1 (Profile 1), one can clearlynotice the cut into the cultural layer that was formedbefore the construction of the trapezoidal base. This

section of the cultural layer confirms that the settle-ment spread farther on the sloping, wide terrace 3,where the riverbed is found today. The accumula-


7/36


7

tion of this layer confirms that the primary phaseof the settlement (B1), which precedes the construc-tion of House 2, was not only related to terrace 3,but that it lasted considerably longer, before the

construction of Houses 2 and 3 (Fig. 7).House 3. The Lepenski Vir culture. Trench 2,quadrant 2b

This building was located at the far end of the areaoccupied by the settlements of the Lepenski Vir cul-ture, and bordered by the semi-circular cove of Sec-tor I, above which there are steep, uninhabited cliffs.We have previously mentioned the destruction ofHouse 3 in a catastrophe which left only part of thefloor, while the fireplace slid down along the steep

edge of the ridge. The context of finds from this basehas already been considered. The settlement of theLepenski Vir culture expanded to this natural borderonly during its youngest phase (Houses 2 and 3; B2)(Figs. 5 and 8).

Stratigraphy and chronology

The first 14C date for Sector I at Padina was ob-tained through the collagen analysis of Burial 2, dugnext to the edge of House 2 (Trench 1, quadrant 1b)belonging to phase B2 of the Lepenski Vir culture

settlement: BM1143, 6220 (95.4%) 5740 cal BC. Inthe absence of a larger series of 14C dates, it is use-ful to connect the archaeological evidence with avai-lable 14C dates in order to estimate their correspon-dence. On the basis of this date, Burial 2 could havebeen interred between the last quarter of the 7thmillennium and the first quarter of the 6th millen-nium BC (Bori and Miracle 2004.315). A similardate probably refers to the collective burial interredin the vicinity of the house 3 (Fig. 9).

It is not possible to make a reconstruction of the Le-penski Vir culture settlements at Sector I, and evenless so of the Late Mesolithic settlement. We can onlyestimate the area of the settlement space. Previouslydiscussed evidence indicates a significant loss of thecultural layer, including portable finds and archi-tectural features, to the Danube. On the basis of thedimensions of segments of Profile 1, which was re-searched at the low water levels of the Danube, theeroded cultural layer was found in a range of 6.0mfrom Profile 1. This layer was coverd by sand andgravels, thinning gradually down the slope of ter-

race 3, which was in part already under the Danubewaters.

The settlement of the Late Mesolithic phase left notrace of building features at this sector and consis-ted only of the cultural layer that is the same as thepalaeosoil. Between the palaeosoil and layer of the

Lepenski Vir culture, two architectural features werefound with a particular building element and a spe-cial stratigraphic position: the elongated rectangularhearths in Trench 1, block 1. Both features were pla-ced on the palaeosoil, making them stratigraphicallyyounger than the Mesolithic settlement.

The position of the building with the rectangularhearth indicates the periphery of the settled areaat this time, which also included terrace 3, the low-est and widest of which is today under the Danube.It is certain that this entire terrace was settled dur-

ing the phase of contact between the autochthonouspopulation and the Early Neolithic Starevo cultureincomers (Padina phase AB).

If the architectural pattern of the base with the rec-tangular hearth is compared with the rear platformwith boulder-altar of House 1, one notices a progressin the development of architecture, for which onelacks transitional examples. The evidence of such atransitional example must have disappeared withthe erosion of the cultural layer of terrace 3 and thecoastal zone. On the other hand, a certain level of

development can be determined by comparing Hou-ses 13 of the Lepenski Vir culture (Profile 1, seg-ment 2; Trench 1, quadrants 1b, 2b). The last twofeatures repeat the same form, with likely differen-ces in details and dimensions. The exclusive use ofa dry wall technique was undoubtedly conditionedby the rocky terrain of terraces 1 and 2, unsuitablefor a wooden dwelling construction (Fig. 5).

Yet another characteristic of the stratigraphy ofphase B1B2 at Sector I is the existence of the invi-

sible dwelling horizon, which is represented by theremains of stone constructions of abandoned ordestroyed buildings. The damaged trapezoidal basebetween Houses 1 and 2, found along the edge ofsegments 2 and 3 of profile 1, belongs to this hori-zon. This trapezoidal base was noticed on the basisof horizontal groups of split stones and stone slabsthat were at first marked as stone constructions.Further, the remains of the rear hearth platformwere preserved in the profile of segment 1, nearHouse 2. Judging from their stratigraphic position,these construction elements point to the existence

of an interpolated horizon with 2 building featuresbetween House 1 and the end of segment 1 (Profile1, point A).


8/36

Borislav Jovanovic

8

One can thus speak of settlements rather than of asettlement of the Lepenski Vir culture at Sector I. Inthis way one recognizes, at least in basic outlines,the evolution of the architecture typical of this cul-

ture, formed during phases B1B2 at Sector I. Thesame relative-chronological framework can be ap-plied to the presence of the Early Neolithic potteryat Sector I. Depending on the stratigraphic positionof such pottery finds, it is possible to follow the evo-lution of this local variant of the Starevo-Cris cul-ture.

Padina A. Sector II. The settlement of the IronGates Late Mesolithic

A narrow ridge that diagonally slopes down thestepped cliffs into the Danube, divided Sectors I andII as two separately populated spaces. Less underthe impact of river erosion, Sector II is characteri-zed by more favourable conditions for the Late Me-solithic settling of this area than Sectors I and III. Itwas exactly at this place that the central settlementof the Late Mesolithic period at Padina was located.This layer, with lenses of burning and rich in decom-posed organic materials, was characterized by a darkcolour and the compactness of the deposit (Fig. 10).Similar to Sector I, flint artefacts and modified bo-

nes were found in this layer. Here, such finds wereconcentrated around the remains of dwelling struc-tures: oval or ellipsoid surfaces built of split or slabstones.

This central part of the sector (Trench II, quadrants2a3a; 3b4b) was not under the erosion impact ofthe Danube, which does not apply to the parts of thesettlement closer to the river bed. In comparison tothe river bank of Sector I, it is likely that approxima-tely only half of the original Late Mesolithic settle-

ment was eroded away. The riverbank profile of Sec-tor II extends for 52m, with an average height be-tween 2.20 and 2.80m. This profile was firstly ver-tically cut and then marked similarly to the longitu-dinal section of the whole sector.

The pattern of the distribution of the Late Mesoli-thic dwelling features at this sector indicates somedegree of organisation, based on the chosen posi-tion, and indicates a semicircular arrangement onthe left shore of a small tributary of the Danube.Coming down the steep slopes of the Gorge, this

stream cut a narrow bed in order to flash out ontothe open and sloping terrain of Sector II.

For the base of dwellings, people chose shallow orwider depressions, as well as channels between par-ticular rocky blocks. The oval base of the biggestdwelling, placed approximately in the centre of the

settled space, had two floor levels made of piledstone slabs and smaller pebbles. Dwellings were pla-ced along the left bank of the tributary, where thecentral dwelling was found.

According to the bases with piled stones, there wereeight dwellings, the size of which varied between3.60 by 1.90m and 0.80 by 0.54m. The total area ap-proximated 30m2, but it might have been conside-rably larger, since the paving of the bases of dwel-lings was done in particular due to the special func-tions of these surfaces: hearths, workshops, sleeping

areas. On these surfaces, along with the remains ofcharred wood, there were residues of stone andbone tools, next to flat, massive boulders with shal-low depressions in the centre (anvils and pounders).

The structure of particular paved areas in blocks 3a4a of Trench 2 provides enough examples of thiskind: Base 4 pavement consisted of sparsely distri-buted stones, among which was a larger slab stone(dimensions 0.46m x 0.20m) indicating a workingsurface like a table; the construction of Base 5 pave-ment was an exception, since it was built entirely of

massive stone slabs from 0.50m by 0.21m to 0.24mby 0.20m; Base 7 pavement was the biggest, ovalsurface (around 6.50m2) and was constructed fromstone slabs of the following dimensions: 0.40m by0.18m and 0.10m by 0.08m; Base 8 pavement hadan approximately square shape, with a massive table-like stone slab that was 0.50m by 0.22m.

The gully of the brook was fan-shaped across thesector close to the riverbank, providing an outlet forperiodic outbursts of seasonal torrents down the

steep slopes of the Gorge. The diagonal course ofthis gully and its gradual shifting towards the down-stream rocky ridge was caused by the higher terrainin the north-western part of the sector. The south-eastern part of the sector, according to this evidence,was composed of the following zones of Mesolithicsettlement: a functional zone next to the right bankof the brook, without dwellings; a central, settlementarea along the left bank of the brook, with a concen-tration of dwellings; a periphery, towards the south-eastern border of the excavation area, with no dwel-lings.

As visible on the profiles of the costal zone, the Me-solithic cultural layer (palaeosoil) is present in seg-


9/36


9

ments 13, Profile 2 due to hill wash processes thataccumulated Mesolithic occupation residues in thiszone. As a consequence, in this part of the profilethe Mesolithic layer has a thickness of up to 0.75m,

covering the rocky ground evenly.These differences in the occupation of the upper andlower zones/terraces at Sector II depended on thecourse of the torrential brook and topography of therocky base. Two preconditions for settling duringthe Mesolithic period were fulfilled at this sector: asource of running water and a convenient place forthe construction of dwellings. Such differences inthe use of space at Sector II affected the stratigraphyhere. It remains unclear how large was the area thatthe Mesolithic settlements covered in this cove, and

it probably depended on the width and depth ofbordering rocky ridges at Sector II.

The most frequent category of Mesolithic materialculture at Padina was chipped stone tools, which canbe separated by form and function into 14 basictypes. Most of these finds were found at Sector II,i.e. in the preserved part of the settlement. It waspossible to separate these finds stratigraphically: themajority were found in the central dwelling that hadtwo floor levels (the intermediate layer betweenthem was 0.16m thick), while a similar pattern was

seen in other dwellings. One should also mentionworkshop areas on the left side of the brook, closerto the downstream ridge of the sector (Radovano-vi 1981).

Besides the secure stratigraphic position of flint findsand retouched boulders, the irregularity in their de-position was caused by a specific terrain of the sec-tor. We should also remember that a part of this set-tlement was destroyed by the Danube. The most nu-merous chipped stone artefacts were scrapers and

sidescrapers (32.83%). This is followed by denticu-lates and notched tools (17.16%). Retouched flakesconstituted 23.19%. These percentages together makeup 73.18% of all 332 retouched specimens. There isa notable increase in burins between the older A1(1.28%) and younger A2 (5.42%) phases, which mayindicate more frequent wood-working activities.Other tools must have been used for hide-working,and were possibly hafted in wooden or bone han-dles (Radovanovi 1996.Fig. 5.4; Fig.5.5; Mihailo-vi 2004.6786; Kozowski 2001.61, Fig. 9).

A special category of knapped stone tools at Padinawere trapezoidal choppers/axes (ranging in size be-tween 3.5cm6.5cm). They also occur in the Upper

Palaeolithic and Mesolithic of the Lower Danubearea, the Ukrainian steppes, and in the north andnorth-east of Europe. This common trait may indi-cate that this type of tool originates in the early Me-

solithic of the Lower Danube region (Fig. 12).That these artefacts are related on typologicalgrounds is based on the following common featu-res: straight and trapezoidal head and slightly ban-ded, edge as well as massive shafting, part rectangu-lar in section. This tool must have been used in awooden or bone handle, which enabled its multipur-pose use. Interestingly, this type of tool is very rarein the collection of chipped stone tools from the dis-cussed area: the duration of this type of tools is re-stricted between the use of larger Late Mesolithic

chipped stone tools and polished stone tools, andwere similar in shape to shoe-last axes that charac-terize the Early Neolithic of the Danubian-Carpathianzone (Figs. 12 and 29). The largest number of thesetools were found at Padina. Moreover, it seems thatthere is evidence for their relatively longer use atthis site: two specimens can be assigned to phase A1,and the other four specimens to phase A2 of theMesolithic settlement at Sector II. There is also aslight typological difference between them duringphase A1, their typical form was not completely de-veloped, while they are perfected during phase A2.

The stratigraphy of stone constructions of the cen-tral dwelling in square 2a from where these findsoriginate is securely established, and this also holdsfor artefacts found in this area. It is clear that be-tween the two phases there is a typological progressin the symmetry of these specific tools, as well as inthe shape and retouching of the cutting edge, andalso in variations in size. All these elements indicatethat such improvements were directly proportionalto the use of such a new tool for wood-working,which resulted in the gradual development of a more

secure type of dwelling structure. In this way, thepopulation minimized its dependence on the local(rocky) relief of Padina (Fig. 12). On the basis of thespread of this type of tool across the plains of north-ern Europe and southern Danubian areas, such inno-vations could be understood as signs of the processthat conjoined the early experience of plant cultiva-tion with a long tradition of hunting and gathering.

Bone artefacts. Trench II, square 2a. Phase A1A2

Bone artefacts were made for specific purposes andtheir form is more or less constant, hence they areusually a chronologically less sensitive class of ma-


10/36

Borislav Jovanovic

10

terial culture. One of the larger closed features ofthis class of tools from Padina relates to the two pre-viously described horizons of the Mesolithic settle-ment (Trench 2, block 2a A1A2) (Fig 11).

Phase A1

The layer under the construction of the older floorof the central dwelling in the settlement (block 2a),lying directly on the rocky base, contained the follo-wing forms of bone tool:

1. Projectile, ellipsoidal cross-section, length 4cm.2. Awl tip prod, end broken (5.2cm x 1.2cm)3. Awl pin, round cross-section (6.8cm x 2cm)4. Awl prod (10.8cm x 2.1cm)

5. Scraper polishing tool, rounded tip (9.2cm x2.2cm)

On the stone construction of the floor of the centraldwelling bone artefacts were few:6. Pin prod, tip of red deer antler (12.5cm x 2.0cm)7. Fragmented polishing tool of good quality (9.5cm

x 2.3cm)8. Boulder/anvil, found with bone artefacts (Fig. 13)

Phase A2

The following finds come from the younger floor ofthe central dwelling and directly above it:

1. Projectile, round cross-section (Length 5cm)2. Arrow tip (pin with spike) (5.3cm x 1.3cm)3. Perforator, lower part in the form of a chisel

(3.9cm x 1.3cm)4. Knife (blade?) made of boar tusk (5.4cm x 1.5cm)5. Dagger (top of spear?), hollow bone (8.0cm x

2.2cm)6. Double-sided awl, round cross-section (length

10.8cm)7. Needle, lower portion broken (6.7cm x 1.1cm)8. Polishing tool scraper (6.2cm x 1.8cm)9. Head of scraper, oval blade (6.5cm x 5.0cm)10. Chisel, damaged, flat blade (6.5 cm x 3.5 cm)

From the renewed floor of the central dwelling ofthe Mesolithic settlement, there is a complete list ofbone artefacts belonging to the younger phase ofthe settlement, made of red deer antler, wild boartusk, hollow and cylindrical animal bone (Fig.14).

From the upper level of the palaeosoil above thefloor of the central dwelling (block 2a).

1. Awl, oval cross-section (Length 6.2cm)2. Perforator, red deer antler (Length 5.5cm)3. Perforator, red deer antler (Length 5.6cm)

Padina B2. The Lepenski Vir culture

Although larger than Sector I, Sector II was not in-habited so extensively during the Lepenski Vir cul-ture phase. The advantage of having running freshwater in this zone was of less importance than thefear of the destructive power of the occasional tor-rential flood. However, the terrace of Sector II hadother purposes during this period. The south-easterncorner of the terrace that was outside of the floodzone seems to have served as a workshop zone forHouses 13 from Sector I, or even of a part of the

settlement of Sector III. Separated by a diagonalridge, the terrace might have served as a corral forstock-breeding, or as a workshop for fishing activi-ties, with some additional stone constructions thatcan be recognized here. A larger amount of splitstones, the remains of fireplaces and smaller stoneconstructions, as well as a significant amount of pot-tery fragments of Starevo-Cris (the Iron Gates),along with other finds, indicate an everyday rangeof activities.

Burials in the course of the Lepenski Vir culture

phase are few at this sector, without particular buri-al constructions or offerings. The deceased werelying in extended positions on their backs, with theirarms along their bodies. The dead were mostly bu-ried in the costal zone (Trench 2, block 1b).

One should also mention another burial where noburial cut was recognized. The burial contained onlya skull, and was found at the periphery of the ter-race in the foothill of the steep sides of the gorge.This partial burial, with some symbolic significance,

was found in block 5d, Trench 2, cut into the Meso-lithic palaeosoil that was later covered by a massivelayer of hill wash; such a stratigraphic position coulddate this cut to the contact phase (AB). The skullwas covered by several stone plaques which wereencircled with split stones, partly damaged by ero-sion. The remains of burning were noticeable in thevicinity of this construction to a diameter of 1.20m.Among broken bone finds there was a haft of a reddeer antler. Apart from skull fragments, no post-cra-nial fragments were found here (Fig. 15). There isone 14C date on a skeletal inhumation (Burial 7)

from this sector (analyzed bone collagen) with theresult BM1144: 82507600 calBC (Bori and Mi-racle 2004.Tab. 1).


11/36


11

Padina B1 3. Sector III

The chronology of prehistoric cultures at Padina in-dicates how particular locations were used during

these periods, which depended on the micro-charac-teristics of the terrain, as well as on the technical ca-pabilities of the culture to adapt to a particular cho-sen location. The settled area of Sector III is muchlarger than the two previous sectors, and is alsomore complex with regard to the organization of thesettlement, including the separate area of a necro-polis. Although the erosion of the costal area is alsoprominent at Sector III, due to the higher absolutelevel of the terrain, it was less destructive than attwo other sectors (Fig. 16). The following functionalzones can be separated on the basis of their relative

chronological order:

The costal area with high profiles is the limit ofthe river erosion.

The lower level of the costal zone, with partly pre-served traces of buildings (B1) the second rowis in a better preserved part of the settlement,above the costal profile (B2). The upper horizonof the terrace at the periphery of the settled area,i.e. the third row of buildings (B3).

Necropolis: a group of burials around elongatedstone constructions of the necropolis (A2); a groupof burials beneath dome-like and other stone con-structions (AB); a group of burials buried aroundbuildings of the third row (B3).

All burial groups were found in the foothills of thesteep sides of the gorge at the very periphery ofthe terrace and were southnorth orientated.

The necropolis of Sector III of the Lepenski Vir

culture settlement can be divided into two groups:The first was contemporaneous with those settle-ments (B3), while the other one significantlyprecedes it (AB; A2). In both instances, one mustaccept the possibility of burials of people whowere not part of the settlement at Sector III (Fig.16).

Padina A2. Trench 7. Sector III. The necropoliswith elongated stone constructions

The oldest building feature at Sector III was not pla-

ced next to the costal zone as one would expect. Theopposite was the case: a relatively long construction(1012m), some type of a shallow base, built in four

levels of piled stones, extended along the foothillsof the vertical massive gorge. Built and piled one ontop of the other, separated by a levelled layer of soil,the stone constructions were a sacred feature a

site place for only few, single burials. However, thedead were not placed inside constructions, but be-side them or in their vicinity (Figs. 17 and 18).

Another important (technical) characteristic of thesestone piles was a layer of loessic soil with some gra-vels that separated the layers of stones. In this way,the whole area of the necropolis was gradually ele-vated, reaching a height of 0.50m0.65m and awidth of 0.80m2.60m. Inside the construction, orin the levelled zones, there were broken bone arte-facts, small pieces of chipped stone, fragments of

animal bone and, in places, notable remains of bur-ning and charred wood.

In total, in all three levels of the necropolis, therewere four burials: in the older levels (12) therewere three burials Burials 19, 21, 22; in the youn-gest level (4) there was only one Burial 23. Burialswere of two types: with the deceased lying supine inan extended position oriented towards the Danube(Burials 19 and 22); in a sitting position, which de-manded more complex stone constructions. Usually,the latter stone constructions and the burial remains

in them were less preserved. An older burial of thiskind comes from level 2 (Burial 21), and a youngerone from the final level 4 (Burial 3).

Burial concentration with domed stone con-structions. Trench 6, block 2

Immediately beside the necropolis of elongated stoneconstruction, there was a burial zone with domedstone constructions. They were constituted of a stone

ring, wider at its base, narrowing toward its top. In-side these constructions, there were one or two bu-rials in sitting positions with crossed legs Burials15 and 16 (two burials in the same stone construc-tion); it is not absolutely clear about Burials 17 and18, since the stone construction was damaged, al-though the individual in Burial 17 was most likelyplaced in a sitting position (Fig. 19). However, with-in this group of burials there are also intermentsplaced in flexed positions and with their hands be-low their heads: Burials 14 and 26. Domed-like bu-rials were less frequent on the left bank of the Da-

nube; a good example is found in the mortuary re-cord of the Late Mesolithic and Early Neolithic ofOstrovul Corbului; a burial of this type was discov-


12/36


13/36


13

dal shape (example of House 12, middle row, phaseB2) (Figs. 21, 22, 23). Rows of such buildings placedparallel to the Danube were the main stratigraphicfeatures with a contemporaneous existence. One pos-

sibility is that each row was inhabited during thesame period as a single settlement horizon, whilethe older horizon was abandoned and littered withthe residue of the settlement (Fig. 20). In the follo-wing, we discuss the function of Houses 13, 15, 17,and 18 from the younger and middle rows of theLepenski Vir culture settlement at Sector III, chosendue to their position, size, architectural details andartefacts found in situ.

House 13. Middle row. Profile III, segment 2

It is unnecessary to repeat the main construction de-tails of this and other buildings described here, sincethese are common to the whole architectural tradi-tion. It is more useful to discuss specific, i.e. localconstruction details. House 13, for instance, was theonly building at Padina with its rear zone limited byblocks of rock placed vertically, similar to a massivestone wall. Bringing such large rocks (1.50m0.80m), which probably originated from the cliffsthat overhang the costal zone, must have dependedon developed engineering skills. The configurationof the terrain at this site excludes any ritual func-

tion for such a dry wall in the back of the building House 13 had, in fact, a stone dam as protectionagainst the torrential brook that ran down the hillinto the Danube, and that also impacted House 18,which will be mentioned later. Hence the protectiverole of this rocky dam: preventing flooding, washingout or destruction of the building, similar to theproblem that caused the destruction of House 3 atSector I (Figs. 16 and 21).

House 18. Trench 5, block 1 Trench 6, block 2

With regard to this building, it was most importantto establish the time of its abandonment, whichmight have been caused by the torrential brook orpreceded this event. The base of this building wascompletely preserved, its complete inventory wasfound in situ, which was rare for the settlements ofthe Lepenski Vir culture at Sector III. House 18 wasalso one of the youngest buildings in this row (B3)and certainly indicates the end of the use of this set-tlement (Fig. 27); on the floor of the building a groupof pots was found which were typologically differ-

ent from the typical Starevo-Cris pottery found inother buildings (with the exception of House 15 inthe same row), in the occupation layer or middensof the costal zone.

It seems that the torrential brook took the inhabi-tants by surprise, but was not the only reason forthe abandonment of this settlement. It seems thatthe abandonment of this building marks much larg-

er changes in the traditional economic system, suchas a decline in demand for fish from the Danube, es-pecially of anadromous fish, such as beluga andother species ofAcipenseridae, or other large fishspecies that were fished by collective efforts. Thus,the abandonment of House 18, i.e. its youngest set-tling horizon, certainly was not sparked by a suddendisappearance of these species, since up until thehydroelectric dams erdap I and II were built, thewhole region was famous for fishing on these spe-cies.

In stratigraphic terms, House 18 cut the front ofHouse 17, built on higher terrain (0.30m0.40m),hence its lower position is an exception to the ruleof building the Lepenski Vir culture settlements younger buildings always being on higher terrain.The exception to this rule must have been related tothe lack of adequate space in the given area; shiftingHouse 18 to the lower position exposed it to the tor-rential brook. Being positioned on loessic sand,House 18 could not have been supported by a wallof massive stone blocks like House 13.

In front of the front, wider side of House 18, i.e. infront of its entrance, there was a preserved part ofthe floor, partly levelled by gravels and small sizesplit stone, like a porch of some kind. The buildingsof the second row, and to some extent those of thefirst horizon, or at least some of them, could alsohave had a similar paved porch, thus giving horizon-tal access to the building, as some kind of narrowplatform that was cut into the layer of loessic sand(Figs. 16 and 17).

House 15. Trench 5, block 2

The entrance area with the patio is clearly visiblein the case of this building, similar in its construc-tion to House 18. What distinguishes this buildingfrom other buildings is a completely preserved cutfor the building, with clearly dug in sides that werealmost vertical, and with a carefully levelled, trape-zoidal floor, along with the massive rectangular built-in hearth. There is an exception in this buildingwhen it comes to the construction of the hearth. Thelonger massive stone slab was found slanting on

the side of the hearth, as thrown into the hearth; itmight have served various functions. There is also alarge block of rock in this building with a slantingflat surface in the back of the building. The stone


14/36

Borislav Jovanovic

14

might have served as a firm construction element forthe longitudinal beam of a gabled roof with the ne-cessary inclination. Pots found on the floor or House15 were few, and were typologically similar to those

found in House 18 (Figs. 16, 26, 28).House 17. Trench 5, block 1

The floor construction of this building, in the youn-gest settlement horizon at Sector III, was carefullyfurnished, while the building can be connected to aspecial sacred/sacrificial function. At the bottom of afunnel-shaped pit (depth from the floor level 0.50m),an oval stone slab was placed as a lid of some sort,while underneath it, a carefully manufactured ellip-soidal stone altar with a shallow receptacle and a

draining lip (0.35m by 0.20m) was found. Similar toother cult places in the settlement, at this place therewere neither visible traces of ritual offerings noranimal bones. Remains of burning and charcoal werenot very prominent here, which is clear by the ab-sence of ash and soil burning. A ritual rite dedicatedto a particular divinity might have been symbolic.House 17 was also abandoned without any visibletraces of violent or hasty abandonment. Yet the in-ventory of finds discovered on the floor of this buil-ding is quite modest (Figs. 16 and 24).

Padina B1 3. Sector III. Settlement organiza-tion

The settlement of the Lepenski Vir culture at SectorIII is not found on top of a previous settlement aswas the case with contemporaneous trapezoidalbuildings at Sector I. However, this could be relatedto the loss of settlement space due to the erosioncaused by the Danube. This was the reason test tren-ches were dug in the costal zone in order to checkthe possibility of the existence of the oldest horizon

in the settlement. In the immediate vicinity of theexcavation area (19681970), using a sandy ridgeformed during the low water table, Trench 3, Sec-tor 3 was excavated to a depth of 2.50m, when un-derground water was reached. This showed that theslope of the terrace was significantly steeper at thetime of the duration of this settlement than at thetime of the excavation. This conclusion was also con-firmed by a limited number of Early Neolithic findsin this zone, putting the boundary of the prehistoriccoastal zone 1520m toward the Danube from theriverbank profile encountered in the 1960s.

It is clear that the width of this zone was irregular,depending on the topography, although it is possi-

ble to assume that approximately one third of thesettlement of Sector III was lost. Such an estimatecould be supported on the basis of the current cos-tal profile, with a cross-section of the bases of trape-

zoidal buildings in the middle row (B2), while thebases of the first, and the oldest, row of buildingswere covered by layers of sand which accumulatedwhen the water table of the Danube was high, i.e.during inundations.

The organization of the settlement space at SectorIII during the Lepenski Vir culture phase was not de-pendent on the complex of necropolises, whichwere significantly older than this settlement, or onthe preceding phase of contacts with the Early Neo-lithic groups, since the space where we might have

had such traces is now covered by the Danube. Whatsurvived was a settlement from the time of the de-veloped Lepenski Vir culture, with three rows of ty-pical buildings. This settlement was gradually builtup. It is realistic to suppose that the main settlementhorizon (B2) was connected with the buildings ofthe older phase (B1), during which the preparationsfor the building of a new row of buildings was anti-cipated (B3) (Fig. 28).

The middle row of buildings on the settlement plancontained Houses 1115/1 (in total 5 floors); the

older, lower horizon contained Houses 510 (in to-tal 6 floors) (Fig. 25). One can notice that the mid-dle row is reduced by one building (6:5), althoughit is advanced on the basis of a new architectural de-tail that was introduced at the time: A-frames placedasymmetrically over the hearth. This reduction inthe number of buildings continues, if one comparesthe middle row of buildings (B2) with the latest,higher row (B3), which was also reduced by onebuilding (5:4), since the space for building was nar-rowed. The conclusion that the settlement space pre-

viously used went out of use at this time, and evenstopped being visible during the latest horizon (B3),is based on the existence of an exceptionally largemidden zone (c. 28m by 6.5m), covering the middleand older parts of the settlement, reaching up to thevery platforms of Houses 15 and 18. The thicknessof the midden was uneven, and it was formed by agradual accumulation of residues from each higherlevel of buildings (Fig. 16). This explains why norubbish pits were found at the sectors of Padina.This role was taken up by the costal zone (with thezone of the abandoned settlement horizon) and high

waters of the Danube. The most important palaeo-zoologial and ichthiological finds from Padina (forthe settlement of the Lepenski Vir culture) were


15/36


15

found in this particular stratigraphic feature. Thescarcity of finds from the contact phase is understan-dable, since these were probably washed out by theDanube at Padina, in this way deepening the enig-

ma of the appearance/disappearance of the Lepen-ski Vir culture.

Padina. Sectors IIII. Pottery and the chronol-ogy of the Lepenski Vir culture

Pottery is not often found within closed contexts inthe Lepenski Vir phase buildings at Padina, despitethe small settlement area and density of finds with-in the buildings. Therefore, we focus on the disco-very of circumstances and contexts of those findsand not on the typological or chronological schemes

based on pottery. There are only two, quantitative-ly different groups of pottery finds. The first typeof pottery comes from the cultural layer, includingmiddens, and the second, much smaller, from buil-dings, hearths and burials. Traditionally, most atten-tion has been paid to the Starevo-Cris pottery fromthe Lepenski Vir settlements, mostly representativevessels, which are even more scarce than the secondgroup of finds (Budja 1999.13435; 2007.50, Figs.31, 32, 33, 34, 35).

However, the first group of pottery finds is surpri-

singly large in number, even if it is compared tothe assemblages from the Early Neolithic sites fromthe middle Danube area. But formal analogies donot imply that the pottery is the same, as potteryfrom the Upper Gorges sites represents a distinctivevariant of the Starevo-Cris complex. This pottery isrobust and poorly made. Vessels of large dimensionsare common; monochrome pottery of better qualityis rare, while the painted pottery is almost absent.At Padina, this last technique of ornamentation iscompletely missing. On the other hand, decoration

of the outer surface of the vessel by pinching andimpressing is the dominant method of decorationfrom the beginning to the end, when it becomes lesscommon. Typical Starevo ornamentation techniquessuch as combing, relief furrowing and the applica-tion of wet clay (barbotine) is completely absent(Bori 1990.4953).

The stratigraphic position of concentrations of pot-tery within the cultural layer is uncertain, due tothe formation of large middens upslope. Like on hill-forts, large fragments tend to move down slope,

where they cover abandoned parts of the settlement.An important issue when discussing the emergenceand development of the Iron Gates variant of Early

Neolithic pottery is to locate its production. There isnow evidence for local production at Padina (at Sec-tor I, block Ib), but only after the contact phase. Thesame can be said, at least in principle, for chipped

and polished stone tools, especially at Sector III (Fig.29)

Padina. Sectors IIII. 14C dates and their corre-lations with stratigraphy

A series of 14C dates from several laboratories havebeen obtained for samples from Padina over theyears, and due the publication of new AMS datesfrom Padina and Hajduka Vodenica, we present thefull list here. We stress the context of samples andtheir correlation with the stratigraphy of the settle-

ment.

The first contradictions between the stratigraphyof building horizons at Padina and corresponding14C dates (from 1978 and without calibration whichwas performed subsequently) were caused by theinverse relative chronology, where younger horizonswere defined as the oldest:

GRN 8229; 6570 55 BP; 56305380 calBC at 95per cent confidence (charred wood, the hearth fromHouse 8, Sector III, Profile 3)

GRN 7981: 710080 BP; 61605780 calBC at 95per cent confidence (cultural layer, Sector III, Trench6, block 1 [phase B3])

GRN 7981: 7075 50 BP; 60305800 calBC at 95per cent confidence (charred wood, House 15, Sec-tor III, Trench 5, block 2 [phase B3])

This contradiction was later attributed to the conta-mination of samples due to the middens. A large vo-

lume of the midden material affected the samples,while middens were also post-depositionaly distur-bed by flash floods.

Obtained dates can be arranged in two intervals the first half of the 6th millennium for phase B3, andthe second half and end of the 7th millennium forphase B2 (Bori and Miracle 2004.345).

A part of the same sample of charred wood from thefloor of House 8 (B1) was dated to 7065 110 BP;61055725 calBC at 95 per cent confidence, which

is approximately contemporary with the date GrN8229. This unpublished analysis was carried out bythe University of Minnesota (USA) in 1974. Absolute


16/36

Borislav Jovanovic

16

dates for House 9 (B1), situated immediately adja-cent to House 8, is according to the date of the sam-ple from the floor calibrated in the range 64106090 calBC at 95 per cent confidence (Canis famil-

iaris, OxA9056), which puts the date for phase B1(according to its stratigraphical priority) in the ex-pected place. House 9 is therefore dated to the mid-dle and second half of the 7th millennium BC (Boriand Miracle 2004.346).

Temporal interval between Houses 15, 17 and 18of the youngest dwelling horizon, Sector III, Trench56 (B3).

Houses 18 and 15

OxA9052, floor of House 18 (deer antler, Cervuselaphus): 6965 60 BP; 59905720 calBC at 95 percent confidence.

OxA9054, under House 15 (Mammalia, workedbone): 6790 55 BP; 57805560 calBC at 95 percent confidence.

The absolute temporal gap between both houses,which is between 210 and 160 calendar years, sug-gests that they are more or less contemporary, whichconfirms the stratigraphic relationships recognized

during the excavations (Bori and Miracle 2004.347).

Houses 17 and 18

OxA11103, from the hearth of House 17 (Mam-malia, bone tool): 7315 55 BP; 62506025 calBCat 95 per cent confidence.

Absolute temporal gap between both houses from thesame horizon, B3, which is between 305 and 260 ca-

lendar years. This confirms the stratigraphical rela-tion between houses, where the floor of House 17was cut during the construction of House 18. The que-stion remains as to whether they are really separatedby three centuries (Bori and Miracle 2004. 347).

Two chronological terminus, related to thebase of House 18

OxA9052. House 18 (red deer antler, Cervus ela-phus): 59905720 calBC at 95 per cent confidence.

OxA9053. House 18 (under the house floor/Canisfamiliaris): 64406210 calBC at 95 per cent confi-dence.

The age difference of both samples corresponds totheir stratigraphical position. However, the intervalbetween dates seems to be too large, almost half amillennium (450490 calendar years). It is possible

that the sample of Canis familiaris belongs to theLate Mesolithic phase, which in turn leads us to dis-cuss the age of dog domestication at Padina (Boriand Miracle 2004.346).

Age difference between Burial 11, under thefloor of House 15 and the base of this building

OxA11104. Burial 11, under the floor of House 15(Sector III, Trench 5, block 2): 93608920 calBC at95 per cent confidence.

OxA9054 (Mammalia, worked bone): 57805560calBC at 95 per cent confidence.

These dates are separated by three millennia. Igno-ring the issue of the acceptability of this difference,we can observe that the pit for Burial 11 could nothave been cut during the existence of House 15. Bu-rial 11 is located 0.6m below the building floor,which together with the depth for the cut made forthe building floor (0.65m) amounts to 1.25m. This issignificantly deeper than the depth of burials nextto the stone structure of the Late Mesolithic necropo-

lis in test Trenches 6 and 7 of the same sector. Thisis similar to the case of Burial 13 under House 18,which was interred before the building was con-structed. As the radiocarbon dates indicate, Burial11 is older than House 18. The large age differencecan be provisionally explained by the early settlingof the local Late Mesolithic population of the DanubeGorges (Bori and Miracle 2004.Tab. 3 and 5; Bon-sall at al. 2002/3. Figs. 34).

Burials 1a and 2. Sector 1, Trench 1, blocks 1

and 1b

Burial 1a is greatly disturbed, only the bones of lo-wer extremities were preserved. Their position indi-cates an unusual position of the body, which was po-sitioned supine (test Trench 1, block 1). However, inthe photograph published by Bori and Miracle(2004.Fig. 4) instead of Burial 1a, Burial 2 (testTrench 1, block 1b) is shown, located next to the Le-penski Vir culture House 2. This is due to the misla-belling of the photograph of this burial in ivanovi(19731974) that Bori and Miracle used in their

paper. However, the stratigraphic position of bothburials puts them in the contact phase or the earli-est stage of the Lepenski Vir culture (B1) (Fig. 5).


17/36


17

The comparisons of 14C dates with the stratigraph-ic evidence from Padina does not end with theseexamples, and it remains important to provide anadequate correspondence between these sets of evi-

dence as one of the key issues in the study of theLate Mesolithic and Early Neolithic settlements at Pa-dina and the Upper Gorges.

Padina IIII. The Lepenski Vir culture

The study of the stratigraphic relationships of theMesolithic settlements of the Lepenski Vir culture isfocused on those parts of the cultural layer, occupa-tion surfaces and features that were spared post-de-positional disturbances caused by erosion or the mo-

vement of material from middens in areas outsidethe living settlement zone. This approach hasbrought new ideas about the appearance of settle-ments, based on the excavated area with the hori-zontal and vertical arrangement of trapezoidal buil-dings with stone constructions on the inner sides oftheir cuts. These entities are composed of past occu-pation surfaces, positioned on terraced surfaces out-side the areas exposed to flood hazards caused bythe Danube and away from torrent streams comingdown the slopes of the Gorges. Building horizons ex-cavated at the Sectors I and II provide a good exam-

ple of their primary function on which the organiza-tion of the settlement is based, by adapting to thespecific environment. Thus, there is a symbiosis be-tween the natural environment and the technologyfor extracting particular resources.

People were making very selective choices aboutwhere to settle in the Upper Gorges in the Late Meso-lithic, sometimes even inconsistently in the choiceof suitable locations, as some less favorable locationswere selected. However, as a general rule, settle-

ments were located in the vicinity of whirlpools (Le-penski Vir, Gospoin Vir), or in the vicinity of rapidsand cataracts (Greben near Vlasac), usually locateddownstream from a particular settlement. The hos-pitable riverbank at Sector IV at Padina, from thefossil Gospoin Vir at Sector III to the outlet of theezava River was not settled, nor was the river bankbetween Stubica and Padina. However, it should bestressed that the shallow coves of Padinas sectorswere located along the last rapids of the GospoinVir whirlpool.

These locations were suitable for collective fishingfocused on large fish. The presence of large Euro-pean sturgeon, which weighed around 200 kg and

was up to 5m in length, was attested by ichthyologi-cal analyses of fish bones from Padina (Brinkhui-zen 1986.18). However, no fish hooks were foundat Padina. Bone fish hooks present in riverine set-

tlements can often be the evidence of organized fish-ing. Therefore, we must assume the use of fish netsand traps at these locations, where similar fishingtechniques were used until the construction of theIron Gates dam.

Fishing in the course of the Lepenski Vir culture isan interesting issue. The cultural development ofcommunities was based on the exploitation of cer-tain natural resources with a highly developed spe-cialized technology, which resulted in unique cultu-ral expressions. We will only briefly touch upon

these phenomena. One of the more general pointsabout the essence of the Lepenski Vir culture is thatits uniqueness can be connected directly with thetransformation of the Late Mesolithic populationsinto the communities of the first producers of foodin the area.

If we observe the whole process of transformationoutside the existing classificatory schemes, we cannote that Mesolithic populations were already in-volved in a system of production, fishing the largeseasonal migratoryAcipenseridae, which was very

similar to food production. A careful considerationof the material and religious superstructure of theseearly communities before the adoption of farmingmay indicate constant efforts at achieving the stageof a food-producing economy. The moment of aban-doning the millennia-long tradition of hunting andgathering at Lepenski Vir and Padina was preservedby chance.

Hajduka Vodenica, Mali Kazan (Lower Gorge)

The prehistoric site of Hajduka Vodenica is locatedin the Lower Gorge of the Danube (Mali Kazan) andconsists of two locations: the settlement in the cen-tre of a large semicircular depression below the Malitrbac mountain massif, while the necropolis was lo-cated beside the Danube to the west. However, nei-ther location has been extensively researched or stu-died, despite large-scale systematic excavations be-tween 1965 and 1969. Hypothetically, the settlementmight have been situated in the middle of the cove,near a large spring. This water source was strong

enough to run a wooden mill, hence the name ofHajduka Vodenica (meaning Hajduks mill or out-laws mill).


18/36


19/36


19

of the Lepenski Vir culture. Downstream from thestone constructions, builders set apart a special area,clearly planned in advance. In the stratigraphicsense, this is a secondary shift of the sacrificial al-

tar, which would be in the tradition of the LepenskiVir culture, although the feature in itself is slightlydifferent from this tradition (Fig. 37).

The first construction of a narrow and long hearth,with elongated lateral sides, was carried out on thelevel of loessic sand, based on preserved foundationditches. Similar to the construction technique in thehorizon with rectangular hearths in the zone of stoneconstructions, long and large stone slabs were in-serted in the foundation trenches here. The immedi-ate area around this hearth remains unknown, al-

though it is certain that on both sides of this firsthearth, burial rites were performed, connected withthe burials found in the cultural layer (Figs. 38, 43).

The hearth was accompanied by an older boulder ofgreen volcanic stone, the surface of which was chan-nelled by narrow flutings, which form an ornamen-tal motif distantly related to patterns found on theboulders of Lepenski Vir (Srejovi and Letica 1983.Fig. 27) (Fig. 55).

After some time, the space around the sacrificial

hearth was filled in and the whole construction wasmoved back, away from the river bank. A burialchamber, cut deep into the layer of flowstone be-tween two ridges which ran into the Danube, wasextended further (dimensions 7.5 x 4.0m, Fig. 41).

A new hearth was built, no longer rectangular, butin a funnel-like shape, with a narrow channel alongone of its lateral sides. There can be no doubt aboutits function it is a conduit for liquid, which doesnot come immediately from the sacrificial altar, but

accumulates at the end of the channel. The design ofthis type is unknown among the hearths of the Le-penski Vir culture (Figs. 38, 39, 42).

The time lapse between the older and younger sac-rificial hearths is evident in the basis of the changein style of decorating the boulder associated withthe younger hearth. It is a sandstone boulder, withsome anthropomorphic details (Srejovi and Letica1983; Antonovi 2004.7374) (Fig. 56).

The floor, built on both sides of the hearth (6.0 x

2.5m), does not respect the symmetry of the con-struction with the hearth on the main axis whichis one of the main construction rules of the Lepenski

Vir culture. The part of the floor that was upstreamthe Danube from the location of the hearth was twiceas large in comparison to the one on the oppositeside; the floor even partly covered the hearth and

was on the same level with two stone slabs. Thefirst, a massive rectangular slab, functioned as thethreshold, while the other one, much narrower,leads to the hearth. There is no ash place, as bothstone slabs were located where the ash place isusually found. The hearth might not have been con-structed for keeping a continuing fire, but for special(cult), periodic events (Fig. 39).

Some burials (Burial 11 and partial remains of otherburials, Figs. 38, 39, 40) were located below the ho-rizon with hearths, in a layer of dark earth with con-

siderable quantities of charcoal.

The burial chamber extends directly from the sacri-ficial hearth. Here, a large number of bodies wereburied in single, double or multiple burials, in theextended position, parallel to the Danube, with theheads pointing downstream (Roksandi 2000.3738). In the latest horizon, burials were covered bystone plaques, and the whole feature was abando-ned. It was preserved until the excavation, with bothsculpted boulders accordingly in association withthe hearths to which they belonged.

We have no answer to the question about the popu-lation which maintained this as well as other ritualconstructions for such a long period. Preliminarycomparisons can be made with Hajduka Vodenica,Vlasac (Bori, French, Dimitrijevi this volume) andPadina AB during the contact phase. For example,the long hearth from this phase at Padina (approxi-mately 1.50m x 0.50m in Sector I, segment 23) isthe only example of the elongation of lateral sidesof the hearth.

These similarities are related to the earliest phase inthe formation of the Lepenski Vir culture, namelywith the first material evidence of the contact phase.They demonstrate the influence of local traditions inthe process of accepting the sedentary life-style andfood production. The strong presence of the LowerDanube variant of Starevo culture, with importantsettlements found distributed down from Kladovoto Prahova and to the mouth of the Timok River,should be taken into consideration when discus-sing the process of transformations at Hajduka Vo-

denica. On the other hand, there are no reliable dataon the fate of the destroyed settlement with Stare-vo pottery found in the middle part of the Hajduka


20/36

Borislav Jovanovic

20

Vodenica cove. Such pottery was also found in hori-zons 3 and 4, with elongated stone constructions ofthe necropolis.

Hajduka Vodenica necropolis, radiocarbondates correlations with the stratigraphic po-sition of burials

A series of 4 AMS dates from Hajduka Vodenica date4 burials from the necropolis with the sacrificialconstruction (Bori and Miracle 2004.346, Tab. 3;for OxA13613 this replaces previously publishedOxA11128 seeBori et al. 2004.Tab. 3). They be-long to different levels of this construction:

OxA13613. Burial 8: 70766699 calBC at 95 per

cent confidence. One of the oldest double burialscut into loess, at the south-east corner of the founda-tion trench of the first sacrificial hearth. The bodywas placed supine with arms crossed, parallel to theDanube, elevation 46.86m.

OxA11127. Burial 12: 65006230 calBC at 95per cent confidence. The burial was placed directlyon the floor next to the sacrificial hearth, after thesacrificial hearth had been abandoned. Here, oneagain finds a double Burial (12 and 9): both skele-tons were fully articulated, elevation 47.74m.

OxA11109. Burial 20: 64406090 calBC at 95per cent confidence. The burial was placed in the ex-tension of the burial chamber and belonged to agroup burial (Burials 17, 17a, 18, 19), elevation 48.14m.

OxA11126. Burial 15: 64706230 calBC at 95per cent confidence. This burial was covered withstone constructions and belonged to the burials ofthe second, younger horizon, elevation 48.54m.

The elevation of the excavated area of the site lo-cated on the riverbank is 48.86m, which makes aheight difference of 3.32m, exactly the height of thesection above the construction of the necropolis.The deposition of the scree hill-wash can be attribu-ted to the time interval between the abandonmentof the site and our excavations. This thick depositscontributed to the preservation of the cultural layerand stone constructions.

AMS dates suggest two chronological horizons of the

necropolis. The older belongs to the beginning ofthe 7th millennium BC (Burial 8), while the youngercan be dated to the middle and second half of the

7th millennium BC. In other words, there may be atemporal gap of several hundred years in the use ofthe necropolis. Both sacrificial structures are clearlyrelated, and the renovation of the younger necropo-

lis continues from the use of the older.Discussion of the date of the construction of the firstsacrificial structure is hindered by the lack of analo-gous features from other areas. Certain architectu-ral analogies with the contact phases of Vlasac andPadina point to the middle and the second half ofthe 7th millennium BC (Bori, French, Dimitrijevithis volume). We must also consider the amount ofwork put into the construction of massive stonestructures layered one on top of the other. In thisway, a monumental stone structure was built, the

dimensions of which can only be guessed at. The ap-pearance of Early Neolithic pottery in the third andthe fourth horizons of the stone structure places thewhole (sacral?) settlement in the formation phase ofthe Lepenski Vir culture.

Conclusions

The results of the systematic excavations of Padinaand Hajduka Vodenica have been concisely presen-ted here. However, since these views are very dis-tant from the time of the initial excavations, they

are viewed through the lenses of several decades ofresearch and interpretation of the the Lepenski Virculture. The issues presented here are the result ofheterogeneous strands of evidence in the presentknowledge of the Lepenski Vir culture, which has fora long time served as a model for the Neolithisationof the frontier area on the Danube. This discussion,based on facts and finds gathered from Padina andHajduka Vodenica, actually announces their immi-nent monographic publication.

If Padina is taken as a point of departure, then theissue was the difficulty of establishing the relativechronological position of the material from Sectors Iand III in relation to other settlements of the Lepen-ski Vir culture.

Sector I was first thought to be chronologically older,as its closed assemblages are typical of the bloomingof the traditional elements of this culture, while theconsiderably larger settlement at Sector III, organi-cally incorporated into the known scheme of rowsof buildings with trapezoidal floors, is distinguished

by a unique find a closed assemblage of potteryfrom the floor of the latest House, 18. This assem-blage has no stylistic or formal similarities with Star-


21/36


21

evo-Cris pottery; it is more similar to an unex-pected phenomenon of the Proto-Vina pottery,with black burnish

Documents

Micro-regions of the Lepenski Vir culture, Padina in the Upper Gorge and Hajducka Vodenica in the Lower Gorge of the Danube