Karavanic, Smith

8/13/2019 Karavanic, Smith

1/26

Ivor KaravanicDepartment of A rchaeology,Faculty of Philosophy,University of Zagreb, I .L uc ica 3, 10000 Zagreb,Croatia

Fred H. SmithDepartment of Anthropology,N orthern I llinois University,DeK alb, IL 60115, U .S.A .

Received 7 May 1997Revision received11 September 1997 andaccepted 22 September1997

K eywords:Mousterian,

Aurignacian, Neanderthals,early modern humans,Central Europe, UpperPaleolithic, Vindija, VelikaPecina.

The Middle/Upper Paleolithic interfaceand the relationship of Neanderthals andearly modern humans in the HrvatskoZagorje, Croatia

This paper presents the first detailed analysis of the artefacts from the

Mousterian level G3 at Vindija Cave and a revision of the artefactanalysis for the early Upper Paleolithic levels (j, i) at Velika Pecina,both in Croatia. Combined with an assessment of the artefacts fromthe crucial G1 level at Vindija, results of these analyses are used toargue that the combination of Middle and Upper Paleolithic elementsin the upper G complex at Vindija is not necessarily the result ofgeological mixing but may well represent a natural cultural assem-blage. Some Upper Paleolithic elements are possibly derived from thelocal Mousterian, while others result from extraneous cultural influ-ences into this region. Interestingly, currently available radiocarbondates indicate that Neanderthals (Vindija level G1) and early modernhumans (Velika Pecina ) were penecontemporaneous in this region atca. 33 ka, or perhaps somewhat earlier if the radiocarbon dates aretaken as minimum age estimates. Therefore some Upper Paleolithictools associated with the Vindija G1 Neandertals, such as bonepoints, may result from imitation of or trade with early modernpeople. While there is external influence on the development of theearly Upper Paleolithic in this region, it exhibits a unique characterwhich does not conform to that of classic Western or CentralEuropean Aurignacian. 1998 Academic Press Limited

Journal of Human Evolution(1998) 34,223248

Introduction

The sites of the Hrvatsko Zagorje (north-western Croatia) are well known in paleo-anthropology because of the importantfinds of fossil humans and their materialcultures in this region. The most significantMiddle Paleolithic sites in this area areKrapina, Vindija, Velika Pecina andVeternica near Zagreb (Figure 1). The

Neanderthal remains from Krapina(Gorjanovic-Kramberger, 1906; Malez,1971; Radovcicet al., 1988;Smith, 1976a;Wolpoff, 1996) and Vindija (Malez,1975;Malezet al. , 1980;Malez & Ullrich,1982; Smithet al. , 1985; Smith & Ahern,1994; Wolpoffet al., 1981) have beenstudied from various perspectives. Alsothe Mousterian artefacts from Krapina(Gorjanovic-Kramberger, 1913; Malez,

1970,1978;Simek, 1991;Simek & Smith,1997)and, to a lesser extent, Velika Pecina(Malez, 1967, 1974) have been describedand analyzed.

Unlike Krapina, Vindija and VelikaPecina also contain Upper Paleolithic strati-graphic units with modern human fossilremains (Smith, 1976b, 1982). The UpperPaleolithic bone tools and lithics from thesesites have been analyzed (Karavanic, 1994,

1995, 1996; Malez, 1967, 1974, 1978,1988), while the Middle Paleolithic materialfrom Vindija has been presented only in part(Malez, 1978; Karavanic, 1996). Further-more, since Vindija and Velika Pecina haveMiddle and Upper Paleolithic componentsin association with remains of fossil humans,these two sites have the potential to make asignificant contribution to both the unsolvedissue of the relationship between Middle

00472484/98/030223+26 $25.00/0/hu970192 1998 Academic Press Limited


2/26

and Upper Paleolithic in Central Europe(seeAllsworth-Jones, 1986) and also to thedebate concerning the appearance of earlymodern Europeans (see Stringer, 1989;

Wolpoff, 1989;Smith, 1991).Vindija Cave is one of the rare European

sites with the possibility of an associationbetween Upper Paleolithic tools andNeanderthal skeletal remains. Other suchassociations involve the Chatelperronian inFrance at the sites of Arcy-Sur-Cure and St.Csaire, but at Vindija the Upper Paleolithicassemblage is clearly not Chatelperronianand has been suggested to represent theAurignacian (Malez et al., 1980;Smith, 1982;Smith & Ahern, 1994;Karavanic, 1995).

We present the first full arachaeologicaldescription of the Mousterian level G3 atVindija, as well as further information on thearchaeological remains from level G1 and

the early Upper Paleolithic levels at VelikaPecina. We also present a revisiting ofthe hypothesis that Neanderthals and theUpper Paleolithic are associated in Vindija,based on analysis of both the artefactsand the fossils human remains, as well aschronometric dates from the pertinent levelsat Vindija and Velika Pecina. Finally, wereassess evidence pertaining to the originof both modern human morphology and

Upper Paleolithic technology in theHrvatsko Zagorje.

Figure 1. Important Middle and Upper Paleolithic sites of the Hrvatsko Zagorje (North-Western Croatia):1. Vindija (Middle and Upper Paleolithic), 2. Velika Pecina (Middle and Upper Paleolithic), 3. Krapina(Middle Paleolithic), 4. Veternica (Middle Paleolithic).

224 . . .


3/26

The material from Vindija and VelikaPecina that is analyzed here was exacavatedunder the direction of M. Malez and ishoused in the Institute for Paleontology and

Quaternary Geology of the CroatianAcademy of Sciences and Arts in Zagreb.Archaeological material from Vindija exca-vated by S. Vukovic is not included in thisanalysis because of the incompatibility of hisstratigraphy and that ofMalez & Rukavina(1979). The stone tools were describedaccording to the terminology of Bordes(1961)for the Middle Paleolithic tool types,even if these derive from Upper Paleolithiclevels. Similarly, the terminology ofSoneville-Bordes & Perrot (1953, 1954,1955, 1956a, 1956b) is used for UpperPaleolithic tool types, even if these areassociated with primarily Mousterian assem-blages. This approach is taken because itprovides greater precision in the descriptionof certain tool categories. Since statisticalcomparison cannot be applied to these data,the benefits of greater descriptive precisionare not overridden by possible problems of

statistical comparison. The bone toolswere classified according to Albrechtet al.(1972).

Background

VindijaVindija is a large limestone cave located2 km west of the village of Donja Voca. Theentrance overlooks the narrow gorge on

the southwestern slope of Krinjakov Vrh.The cave is more than 50 m long, and itsmaximum width and height measure 28 mand 10 m, respectively. More than 9 m ofdeposits were present, and the stratigraphi-cal profile comprises about twenty stratathat Malez & Rukavina (1979) interpretedas extending from the Riss glaciation (oxy-gen isotope stage 6 or earlier) through theHolocene. A UTh date on bone from level

k of 114 ka suggests a last interglacial age forthese deposits (Wildet al. , 1987/88). Level k

is stratigraphically older than the G complexand basically equivalent to Mousteriandeposits at Krapina, recently dated toca. 130 ka (Rinket al., 1995). Unfortu-

nately, UTh dates for older levels at Vindija(L & M) are inconsistent and seeminglyunreliable.

Excavations were conducted in Vindijafor over 30 years, beginning in 1928, byS. Vukovic (1950). Vukovic excavated bothinside and in front of the cave, mostly in theupper levels. Subsequently, M. Malezdirected excavations from 1974 until 1986.It was during this period that most of thePaleolithic archaeological and Pleistocenefaunal material, as well as all of the fossilhuman remains, pertinent to our studywere recovered. Additionally, the cavesPleistocene stratigraphic sequence wasestablished at this time (Malez & Rukavina,1979).

Veli ka PecinaVelika Pecina is situated near the village ofGoranec on Ravna Gora, between the sites

of Krapina and Vindija. The entrance is onthe western slope of Plat hill (brdo Plat),and the cave itself is 25 m long. In someparts of the cave the sediments are over10 m deep. There are 16 defined strati-graphic levels, ranging from the end of theRiss glaciation (oxygen isotope stage 6)through the Holocene (Malez, 1974,1979).

Excavations in Velika Pecina were con-ducted initially by M. Malez in 1948, with

the intention of establishing the stratigraphyand preparing for future systematic excava-tions. Subsequent excavations began in1957 and, with several interruptions, lasteduntil 1970.

Stratigraphy and chronology

VindijaThe critical levels for assessing the relation-

ship between the Middle and the UpperPaleolithic at Vindija are G3, G2 and G1

225 /


4/26

(Figure 2). The description of these levelsused here is based on the chronostratigraphyemployed by Malez & Rukavina (1979).The stratigraphically higher F complex,

which is important to issues concerningthe early Upper Paleolithic at Vindija, isdescribed elsewhere (Malez & Rukavina,1979;Karavanic, 1995).

G3 level. This level comprises a distinctivesandy green sediment, with relatively fewstone fragments. Thickness of the levelvaries from 10 cm to 30 cm. The only cur-rently available date for this level is an

amino-acid date of 42,4004300 BP doneby R. Protsch (Smithet al. , 1985). Unfortu-nately, dates by this technique must beviewed cautiously due to the well-knownproblems with amino-acid dating of bone.

G2 level. Made up of grayish clay sedimentwith abundant small stone rubble, this levelis present only in some parts of the cave.Thus there are portions of the cave whereG1 and G3 are in direct stratigraphiccontact. Thickness of the G2 level rangedbetween 1 and 30 cm.

G1 level. This is a distinctive stratum,formed by a red-brown clay sediment,820 cm in thickness. It is easily dis-tinguished from other levels of the G andthe overlying F complexes. Occasionally itcontains carbonaceous particles. Artefactsand bone from G1 are often distinguished

by adhering grains of this matrix, which isunique in the cave. ChronostratigraphicallyG1 correlates to the Wrm 2/3 interstadialin the French version of the Alpine termino-logical scheme (Rukavina, 1983). Recentradiocarbon anaysis of a cave bear bonesample from this level produced an ageestimate of 33,000400 BP (Karavanic,1995).

There are still many unresolved issues

relating to the stratigraphic interpretation ofVindija. The most significant of these is the

occurrence of cryoturbational phenomena inthe cave (Malez & Rukavina, 1975), and itsrole in potential mixing of elements betweenlayers (Stringer, 1982a, 1982b; Kozowski,

1996;Montet-White, 1996). This issue willbe discussed in the context of describing thearchaeological and paleontological remainsfrom the site.

Velika PecinaThe critical levels for assessing the relation-ship between the Middle and UpperPaleolithic at Velika Pecina are k, j and i(Figure 2). The description of these levels

used here is based on the chronostratigraphyemployed byMalez (1979).It is importantto note that there is no evidence of cryotur-bational activity documented for VelikaPecina (Malez, 1974; Radovcic, personalcommunication).

k level. This is a light yellow, sandy claystratum, with stones comprising ca. 95%of its volume. Level k measures between160180 cm in thickness.

j level. This layer is formed of light brown,compact clay approximately 40 cm thick.

i level. This stratum is a pale, light brownclay with a large number of stones. It rangesbetween 8085 cm in thickness. Radio-carbon dating yielded an age estimate of33,850520 BP (Malez & Vogel, 1970).

Vindija industries from levels G3, G2and G1

Level G3Level G3 yielded 357 pieces of knappedstone, of which 50 (14%) represent typo-logically defined tools. Debitage consistsof 242 flakes (53 primary decortication,60 secondary decortication), 23 cores,24 chunks and 16 broken pebbles. One

hammerstone is also present. Most of thedebitage (779%), as well as many of the

226 . . .


5/26

Middle Paleolithic-type tools, are made ofwhite quartz. Other Middle Paleolithic and

all Upper Paleolithic-type tools are madeof other raw materials (chert, tuff, etc.).Debitage is also present in G3 that matchesthe raw material for three of the endscrapersrecovered from this level. This might besupport for the production of these UpperPaleolithic type tools in level G3. Unfortu-nately, this is not conclusive, because someMiddle Paleolithic-type tools are also madefrom the same raw material.

The tools from level G3 are: nine notchedpieces (Figure 3, nos. 2, 5), nine denticu-lated pieces (Figure 3, no. 6) plus oneprobable pseudo-tool (Figure 3,no. 7), foursingle convex sidescrapers (Figure 3, no. 8),five single straight sidescrapers (one ofwhich is in combination with an end-scraper), a single concave sidescraper, fournaturally backed knives (one being partiallybacked by retouchFigure 4, no. 6), a

double straight convex sidescraper (Figure4, no. 5), a double concave convex side-

scraper, an alternate retouched sidescraper(Figure 3,no. 3), an unfinished leaf-shaped

bifacial piece (Figure 3, no. 4), three end-scrapers on flakes (Figure 4, nos. 1, 2, 4),one endscraper on a core, two limaces, oneraclette, a burin, a rabot, a chopping tool(Figure 4, no. 7), a blade with two continu-ously retouched edges (Figure 3, no. 9) anda tool in the shape of a small point. Also,there are 13 additional pieces that cannot beformally classified. They are made of whitequartz and were probably used as tools. The

minimum frequency of sidescrapers andlimaces in the tool assemblage is 30%.

Flake technology is dominant in level G3as would be expected in a Mousterianassemblage, but there is also evidence ofblade technology (Figure 3,nos. 1, 2, 8, 9;Figure 4, no. 3) and bifacial technology(Figure 3, no. 4). Some tools were made onlarge blades, and one unfinished leaf-shapedbifacial piece was probably made by two

unifacial treatments (A. Marks, personalcommunication). No use of the Levallois

Figure 2. Potential correlation of the Vindija and Velika Pecina levels discussed in the text (after Malez,1974;Rukavina, 1983).

227 /


6/26

Figure 3. Selected artefacts from Level G3, Vindija Cave: 1. blade (with edge damage); 2 & 5. notchedpieces; 3. alternate retouched sidescraper; 4. leaf-shaped bifacial piece (unfinished); 6. denticulated piece;7. pseudo-tool; 8. single convex sidescraper; 9. blade with two continuously retouched edges. Scale is

in cm.

228 . . .


7/26

Figure 4. Selected artefacts from Level G3, Vindija Cave: 1, 2 & 4. endscrapers on flakes; 3. endscraperon (broken) blade; 5. double straight convex sidescraper; 6. naturally backed knife (partially backed byretouch); 7. chopping tool. Scale is in cm.

229 /


8/26

technique is evident in the lithics from thislevel. The presence of Upper Paleolithictypes (such as endscrapers) is readily observ-able, and I. Turk (personal communication)

has noted that numerous endscrapersof similar types are present in the UpperPaleolithic of Bacho Kiro in Bulgaria (seeKozowskiet al. , 1982).

The presence of these Upper Paleolithictools in this Mousterian level might possiblyhave been caused by mechanical mixing ofelements from different strata. However,none of these Upper Paleolithic tools exhibitrounded edges which would clearly demon-

strate abrasion due to cryoturbation (seeLavilleet al., 1980: Fig. 3.12). Further-more, raw material similarities between thefinished tools and the debitage lend somesupport to the assertion that the majority ofthe tools, including the Upper Paleolithictypes, were produced during the time spanrepresented by this level. Bearing all thesefacts in mind, an argument can be madethat level G3 represents a late Mousterianindustry with the co-presence of flake, bladeand bifacial technology, as well as toolsdisplaying Upper Paleolithic traits.

G2/3Eight pieces of the lithic industry carry theG2/3 designation. These are: one large flake(probably used), one nosed endscraper(Figure 5, no. 1), two blades with twocontinuously retouched edges (Figure 5,nos. 2, 3), a single straight sidescraper, a

single convex sidescraper, an alternateretouched sidescraper, and a scraper-drill.These tools also represent a mixture ofMiddle and Upper Paleolithic elements,including one Aurignacian tool (the nosedendscraper). For these tools it is impossibleto establish whether they belonged to levelG2 or G3.

Level G2

This level yielded nine pieces of debitage(one chunk and eight flakestwo primary

decortication, four secondary decortication)and four tools. The tools are: one notchedpiece, a transverse convex sidescraper(Figure 5, no. 5), a sidescraper with bifacial

retouch (Figure 5, no. 4), and a naturallybacked knife. Level G2 is present only insome parts of the cave (Malez & Rukavina,1979: 190), which means that the levels G1and G3 are in contact in portions of the cave.All debitage is white quartz, but not all toolsare produced from this raw material. Alltools are basically Mousterian in character.

G1/G3

One hundred and thirty-five pieces ofknapped stone are labelled G1/G3. Of these,25 (185%) are tools. Debitage consisted of102 flakes (including 20 primary decorti-cation, 28 secondary decortication, tworetouched), four cores and four chunks.Typical tool types are: a sidescraperretouched on the ventral surface (Figure 5,no. 7), an abrupt retouched sidescraper, analternate retouched sidescraper (Figure 5,no. 6), a single concave sidescraper, anendscraper on a flake (Figure 5, no. 9),an endscraper on a broken blade (Figure 5,no. 10), a straight dihedral burin (Figure 5,no. 11), a drill (Figure 5,no. 12), a limace,a drill-endscraper, notches, denticulatedpieces and a notched bladelet (Figure 6,no. 6). For these tools it is impossible toestablish whether they belonged to level G1,G3 or, in some cases, G2. One flat pebblewas also found. It had edges sharpened from

use (scraping-polishing). It should be men-tioned that a single artefact appears to bemarked as G1/G4, but we are unaware ofany contact between these layers. This isa notched piece on a fragment of blade

(Figure 5, no. 8). Thus, G1/G3 marked tools

are fundamentally Mousterian in character

with only a few Upper Paleolithic elements.

G/g

The designation G/g (meaning G/up) refersto upper levels of the G complex. It was

230 . . .


9/26

Figure 5. Selected artefacts from Vindija Cave. G2/3: 1. nosed endscraper, 2 & 3. blades with twocontinuously retouched edges; Level G2: 4. sidescraper with bifacial retouch, 5. transverse convex

sidescraper; G1/G3: 6. alternate retouched sidescraper, 7. sidescraper on the ventral surface, 9. endscraperon flake, 10. endscraper on (broken) blade, 11. straight dihedral burin, 12. drill; G1/G4: 8. notched piece.Scale is in cm.

231 /


10/26

Figure 6. Selected artefacts from Vindija Cave. G/g: 1. Leaf-shaped bifacial piece, 2. blade with twocontinuously retouched edges, 3. endscraper on flake, 4. nosed endscraper, 5. sidescraper-endscraper,

7. nosed endscraper, 8 & 11. single straight sidescrapers, 9 & 12. offset sidescrapers, 10. denticulatedpiece; G1/G3: 6. notched bladelet. Scale is in cm.

232 . . .


11/26

probably used at the beginning of theexcavation when it was not yet possible todistinguish the individual levels of the com-plex clearly. The G/g label is found on 43

pieces of debitage and on 34 tools. Amongthe debitage items, we can identify 39 flakes(including eight primary decortication, eightsecondary decortication), three chunks andone blade. The most typical tool types are aleaf-shaped bifacial piece (Figure 6,no. 1),a blade with two continuously retouchededges (Figure 6, no. 2), an endscraper on aflake (Figure 6, no. 3), nosed endscrapers(Figure 6, nos. 4, 7), a sidescraper-endscraper (Figure 6,no. 5), single straightsidescrapers (Figure 6, nos. 8, 11), singleconvex sidescrapers, a convex transversesidescraper, a straight transverse side-scraper, a concave transverse sidescraper,offset sidescrapers (Figure 6, nos. 9, 12),naturally backed knives, denticulated pieces(Figure 6,no. 10) and notches. These toolsalso represent Mousterian types togetherwith some Upper Paleolithic types. Thetwo nosed endscrapers are specifically

Aurignacian type tools. In addition to thesepieces, it is possible that some of the flakesincluded in the debitage representde factotools (expedient sidescrapers, notches,denticulated pieces).

Level G1This level contains 62 chipped stone items,of which 15 (242%) are typologically recog-nizable tools. Among the debitage, 28 flakes

(five primary decortation, ten secondarydecortation), one core, ten chunks, and twobroken pebbles can be identified. A ham-merstone and complete pebble were alsofound. The Upper Paleolithic tool types are:an endscraper on a flake (Figure 7, no. 3),and endscraper on an Aurignacian blade(Figure 7, no. 4), a straight dihedral burin(Figure 7, no. 2) and a blade with twocontinuously retouched edges (Figure 7,no.

11). Five sidescrapers (Figure 7, nos. 5, 6,7) and four denticulated pieces (Figure 7,

nos. 8, 9, 10) are more typical of theMousterian tradition. However, it is possiblethat some pieces classified as denticulatepieces are in fact pseudo-tools. One of

these is illustrated in Figure 7 (no. 8). Aleaf-shaped bifacial point, nicely retouchedon both sides (Figure 7, no. 1), and a rabotalso originate from this level. As in levelG3, the Levallois technique was not usedin level G1.

Associated with this stone industry arebone tools of types generally characteristic ofthe Upper Paleolithic. Especially interestingis a split-base bone point (Figure 8,no. 1)and three massive-base (Mladec) bonepoints (Figure 8, nos. 2, 8, 9). There arealso five other tool fragments, four of whichare presented inFigure 8(nos. 3, 5, 6, 10).Also, a bear baculum with engraved circum-ferential markings (Figure 8, no. 7), and aso-called bone button (Figure 8, no. 4)are designated as deriving from this level.The latter was probably produced by cavebear activity or another natural process andnot by humans (see Turk, 1988). On the

other hand, the bear baculum is clearly theresult of human modification. Although ithas been attributed to level G1 (Malez,1988), a note associated with this specimensuggests that it may in reality have comefrom the upper part of G3, which wouldmake it even older.

While the bone tools suggest an attri-bution of this level to the Aurignacian, thelithic industry is more equivocal. It contains

typological elements typical of both theMiddle and Upper Paleolithic, but onlyone endscraper on an Aurignacian bladeis unequivocally characteristic of theAurignacian. However, Kozowski (1996)ascribes the G1 lithics to a Moustero-Levalloisian assemblage and states thatboth the dated cave bear bone and thetwo Aurignacian bone points are in all like-lihood an admixture in layer G1 from the

materials resting on the interface of layersG/F, the layers which represent thetypical

233 /


12/26

Figure 7. Selected artefacts from Level G1, Vindija Cave: 1. leaf-shaped bifacial piece; 2. straight dihedral

burin; 3. endscraper on flake; 4. endscraper on an Aurignacian blade; 5. single concave sidescraper;6. double convex sidescraper; 7. single straight sidescraper; 8, 9 & 10. denticulated pieces; 11. blade withtwo continuously retouched edges. Scale is in cm.

234 . . .


13/26

Figure 8. Bone assemblage from Level G1, Vindija Cave: 1. split-base point; 2, 8 & 9. massive-base points;3, 5, 6, & 10. fragments; 4. bone-button; 7. bear baculum bone with carved circular markings (possiblyfrom G3). Scale is in cm.

235 /


14/26

Aurignacian with blade tools and bonepoints (p. 211: emphasis ours). The pro-posed mixture solution (Stringer, 1982a,b; Kozowski, 1996; Montet-White, 1996)

could be possible, because in the originaldocumentation of the excavations we foundthat a few stone tools from level G1 camefrom the areas in the cave disturbed bycryoturbation. However, as was the case inlevel G3, we didnotnote any modification ofthe Upper Paleolithic type lithic tools (e.g.,nibbled, rounded edges) which would sug-gest mixing by this process. Furthermore,there are actually three virtually complete

bone tools and several fragments from levelG1 all of which would have to be attributedto mixing. As was the case with the lithicitems, the bone tools lack the abrasion andbattering expected if they were disturbedsignificantly by cryoturbation.

Unfortunately, excavation records do notidentify the precise grid location for most ofthe bone tools. However, the split-base bonepoint and virtually all of the human fossilremains, including the Vindija 207 mandible(Figure 11), were excavated north of profileII as presented inMalez & Rukavina (1975).The strata from this portion of the cave donot exhibit any obvious evidence of cryotur-bational activity, as the photographs, pub-lished profiles, and descriptions from thispart of the cave demonstrate. The part ofthe cave showing extensive cryoturbation islocated approximately 2 m south of Malezand Rukavinas profile II. Thus, while some

mixing may have occurred, it is difficult toattribute all of the archaeological character-istics of the Vindija G1 level to the effects ofcryoturbation.

Velika Pecina industries from levels k,j and i

Level k

Level k is divided into seven zones (k1k7).The lower part of the level (zones k7, k6

and the lower part of k5) produced onlyfive tools, one flake and a piece of a quartzpebble (Malez, 1967: 2225). Recent re-analyses of these specimens revealed tools

belonging to the following types: oneLevallois point (Figure 9,no. 1), two alter-nate retouched sidescrapers (Figure 9,nos. 3, 5), and two denticulated pieces(Figure 9, nos. 2, 4). In the upper part ofzone k5, and zones k4 and k3, therewere no artefacts; and only three tools havebeen found in zones k2 and k1 (Malez,1967: 26). These are: one probable notchedpiece (Figure 9,no. 7) and two pseudo-tools

(Figure 9, nos. 8, 9). One so calledbone button (Figure 9,no. 6) was foundin zone k1, but, as previously noted, theseare not human-made artefacts (see Turk,1988).

Malez (1967: 28) attributed the artefactsfrom the lower part of level k to theMousterian and those from the upperpart tentatively to the proto-Aurignacian,because they were not typical of any definedindustry. He suggested that they could alsobelong to the Mousterian. Considering thesmall quantity and typological characteris-tics of the tools from level k, there is noconvincing reason to recognize two differentindustries. It is likely that the lower part oflevel k belongs to the Mousterian, and theupper part may contain only pseudo-tools(geofacts).

Level jOnly one stone tool originates from thislevel, a blade with two continuouslyretouched edges with a notch on the left side(Figure 10,no. 1). Because it was imposs-ible to determine a distinct industry affili-ation, and because the next higher level wasthought to represent the Aurignacian,Malez(1967: 28) used the term proto-Aurignacianfor the single tool in level j. A human frontal

fragment (Smith, 1976b) derives from thislevel.

236 . . .


15/26

Figure 9. Selected artefacts from Velika Pecina. Level k (zones k7, k6, k5): 1. Levallois point, 2 & 4.denticulated pieces, 3 & 5. alternate retouched sidescrapers; Level k (zones k2 and k1): 6. bone-button,7. notched piece, 8 & 9. pseudo-tools. Scale is in cm.

237 /


16/26

Figure 10. Selected artefacts from Velika Pecina. Level j: 1. blade with two continuously retouched edges

(with notch on left side); Level i: 2. drill, 3. endscraper on (broken) blade with a notch, 5. canted dihedralburin, 4, 6 & 7. bone points (probably split-base), 8. massive-base bone point. Scale is in cm.

238 . . .


17/26

Level iSeven stone and three bone artefacts werefound in level i. The stone tools are: onecanted dihedral burin (Figure 10, no. 5), a

drill (Figure 10, no. 2), double convexsidescraper, convergent convex sidescraper,alternate retouch sidescraper and end-scraper on (broken) blade with a notch(Figure 10,no. 3). There is also one bladeletcore. Bone tools consist of three points thatprobably had split bases (Figure 10, nos. 4,6, 7) and one massive-base (Mladec)point (Figure 10, no. 8). Originally Malez(1967) attributed the base of this Mladecpoint to level h, but it fits with a point tipfrom level i.

While the stone artefacts include someUpper Paleolithic types, Middle Paleolithictypes (three sidescrapers) are also repre-sented. However, the bone points stronglysuggest an Upper Paleolithic affiliation.The precise determination of which UpperPaleolithic industry is represented is moredifficult because of the lack of typicalstone tools. Mixture of Upper and Middle

Paleolithic type tools at Velika Pecina can-not result from the effects of cryoturbation,because there is no evidence of this processreported in this cave (Malez, 1967, 1974;Radovcic, personal communication).

Human remains

Fossil human remains from the HrvatskoZagorje are known from the sites of Krapina,

Vindija, and Velika Pecina and have beendiscussed in several recent assessments oflate Pleistocene human evolution (Malez,1978; Smith, 1982, 1984, 1991, 1994;Wolpoff, 1980, 1996). These remains aredivisible into three groups: Neanderthals(Krapina and Vindija levels G3 and G1),early modern Europeans (Velika Pecina andVindija levels Fd and Fd/d), and later UpperPaleolithic-associated specimens (Vindija

level D). The most pertinent of these to theissues under consideration here are Velika

Pecina and the Vindija G1 and F complexsamples.

The human remains from Vindija levelG3 unquestionably represent Neanderthals

(Wolpoffet al. , 1981; Smithet al. , 1985),albeit with a distinct pattern of changes in

facial morphology compared to earlierNeanderthals. Among such changes are ver-tical mandibular symphyses, mandibles withan incipientmentum osseum andincurvatiomandibulae, maxillae with narrower nasalapertures and shorter alveolar processes,and supraorbital tori with a shape some-what intermediate between the KrapinaNeanderthals and early modern Europeans(Smith, 1994;Wolpoff, 1996).

The human remains from Vindija levelG1, are fragmentary and not extensive(Table 1), but they clearly representNeanderthals. This assessment is based onthe presence of a true supraorbital torus onVi 308, the morphology of the Vi 307 zygo-matic, the large size and shoveling patternon the Vi 290 incisor, and the retromolarspace and horizontal-oval mandibular

foramen on Vi 207 (Figure 11) (Smith,1984;Smith & Ahern, 1994;Wolpoffet al. ,1981).

Many of these features are not uniqueto Neanderthals. Retromolar spaces(Franciscus & Trinkaus, 1995) and colum-nar frontal processes on the zygomatic(Smith, 1976a; Smith & Ahern, 1994)occur, but are rare, in post-NeanderthalEuropeans. Features like multiple zygomati-

cofacial foramina, well-developed Breschetssulci, horizontal-oval mandibular foramina,and maxillary incisor shoveling, all of whichcharacterize the Vindija G1 remains,occur in varying frequencies in UpperPaleolithic-associated remains, but arestill more common in Neanderthals (e.g.Smith, 1978;Frayer, 1992;Wolpoff, 1996).Despite some uncertainty about individualfeatures, the complex of features exhibited

by the six G1 specimens from Vindija,taken collectively, would be extremely

239 /


18/26

Table1 Vindija human skeletal remainsfrom levels G1, Fd, andFd/d (original descriptionsin Wolpoffet al ., 1981,Smith et al ., 1985)

Specimen Level Description Salient features

Vi 207 G1 Right mandibular ramus withedentulous posterior corpus

(1) retromolar spaceb(2) horizontal-oval mandibular foramenb

Vi 208 G1 Anterior, superior fragment of left parietal (1) Breschets sulcus well-developedb

Vi 287 G1 Right upper canine

Vi 290 G1 Right upper central incisor (1) strongly shovel-shapedb

(2) large sizeb

Vi 307 G1 Left zygomatic bone (1) columnar frontal processb

(2) multiple zygomaticofacial foraminab

Vi 308 G1 Left frontal fragment with medialsupraorbital torus

(1) true supraorbital torus(2) large frontal sinus, restricted to torus

Vi 204 Fd/da

Right posterior parietal; articulates withVi 302 (left posterior parietal) (1) gabled coronal contour(2) slight lambdoidal flattening

Vi 286 Fd Lower right lateral incisor

Vi 289 Fd Upper right lateral incisor (1) strongly shovel-shapedb

Note: aLocated at the interface of Fd and Fd/d. bGenerally associated with, but not unique to, Neandertalmorphological pattern.

Figure 11. The Vindija 207 mandible viewed from above (top) and medially. Note the retromolar space

between the alveolus for M3 (B) and the anterior border of the ramus, the horizontal-oval mandibularforamen and the rather medial position of the intersection between the mandibular notch (incisura) andthe condyle (A).

240 . . .


19/26

hard to accommodate in anything but aNeanderthal sample.

Only four specimens of early modern

humans are known from the HrvatskoZagorje. These are the partial frontal fromlevel j at Velika Pecina (Smith, 1976b), thearticulated posterior parietals (Vi 204 and302) from Vindija complex F (Smith &Ahern, 1994), and two isolated teeth fromthe F complex (Wolpoffet al. , 1981). TheVelika Pecina frontal (Figure 12) is unques-tionably modern in morphology, with abrow ridge clearly divided into a distinct

superciliary arch and supraorbital trigone(Smith, 1976b). Similarly, the F complexparietals from Vindija exhibit a gabled coro-nal contour that conforms to the conditionseen in other early modern Europeans. Theincisors from the F complex could beaccommodated in either a Neanderthal orearly modern sample.

Although the Vindija Neanderthalsexhibit systematic morphological changes

toward the modern European condition(when compared to earlier Neanderthals),

there is a distinct gap in cranial, thoughperhaps not in dental, morphology betweenthem and the early modern remains from

Velika Pecina and the Vindija F complex.Both the Velika Pecina frontal and theVindija F parietals would not fit morpho-logically into the Vindija G Neanderthalsamples, although in their brow ridge andlambdoidal flattening patterns they mayexhibit Neanderthal-reminiscent features(Smith, 1984; Smith & Ranyard, 1980;Smithet al. , 1985).

Assuming the accuracy of currently avail-

able dates, both early modern humans andlate Neanderthals appear to inhabit theHrvatsko Zagorje at about 33 ka, or perhapsearlier if the radiocarbon dates are taken asminimal age estimates. If these populationsare indeed penecontemporaneous, it seemsunlikely that early modern people couldhave appeared in this region without asubstantial increase in gene flow, includingpopulation movements, into Europe at this

time (Trinkaus & Smith, 1985; Smith &Trinkaus, 1991).

Figure 12. The human frontal bone from level j at Velika Pecina. Note the distinct separation of the browridges into a superciliary arch and supraorbital trigone.

241 /


20/26

Overview of the MiddleUpperPaleolithic interface in north-western

Croatia

Typological analysis of the stone tools fromVindija level G3 indicates the presence ofcharacteristic Upper Paleolithic type tools ina fundamentally Mousterian assemblage. Inaddition to typical Mousterian tools (e.g.,sidescrapers) and flake technology, level G3lithics also include evidence of bifacial tech-nology (seeFigure 3,no. 4) as well as bladetechnology (see Figure 3, nos. 1, 2, 8, 9).Blade technology and some Upper Paleo-

lithic tool types can be present in theMousterian (see Bordes, 1961); however,the Vindija G3 endscrapers (see Figure 4,nos. 1, 2, 3, 4) are smaller and less crudelymade than is common in the Mousterian.The raw material comprising the debitagefrom level G3 suggests, but does not prove,that almost all toolscouldhave been madein situ by the Neanderthals. Furthermore,the types of stone (white quartz, chert, tuff,sandstone) used for tool production inVindija level G3 derive from environs nearthe cave, and no item suggests any othersource for raw material (see Kurtanjek &Marci, 1990).

As in level G3, mixture of Middle andUpper Paleolithic typological and techno-logical characteristics is also present in thetools from the level G1 (see Figure 7).Furthermore, the connection between G3and G1 is also suggested by the presence of

leaf-shaped bifacial pieces, which may rep-resent a typological and technological con-nection between the Middle and the UpperPaleolithic in Central Europe (see Valoch,1968; Allsworth-Jones, 1986). These toolsat Vindija are found in levels G3 (Figure 3,no. 4), and G1 (Figure 7, no. 1). Anadditional specimen is marked as G/g(Figure 6, no. 1), meaning that it comesfrom some level in the upper G complex;

and yet another one carries the markingG/d, indicating an origin in the lower G

complex (probably level G4 or G5). Malez(1979: Fig. 31, no. 6) described a fifthleaf-shaped bifacial piece from the upperstrata of the complex G. Unfortunately this

specimen cannot be located.One of these specimens, from level G1, isa particularly well-made piece, fashioned ofred radiolarite. It differs from the otherleaf-shaped bifacial pieces from the lowerlevels, which are more crudely produced onblack chert. The G1 point (Figure 7, no. 1)shows great similarity, both in terms oftypology and color, to finds from JankovichCave (Hungary). There is no debitage

(small bifacial thinning flakes) from Vindijalevel G1 reflecting production of this piece,and it is possible that this point wasimported (Montet-White, 1996). However,because systematic dry and water screeningwere applied on only a very limited part ofthe sediment, it is possible that small piecesof debitage from such relatively rare rawmaterial would not have been recovered.

Except for the possibility the bear bacu-lum derives from G3 rather than G1 and aworked rib fragment of unknown proven-ience within this complex, the bone industryfrom the G complex at Vindija was recov-ered entirely from level G1. Massive-basebone points are found both in G1 and the Fcomplex, raising the possibility that they areintrusive into G1 (e.g., Kozowski, 1996).The single split-base bone point from G1 isthe only such point recovered from the site.Absence of mechanical alteration on the

tools usually associated with the effects ofcryoturbation is a logical argument againstextensive mixing of artefacts by geologicalprocesses at Vindija.

If the radiocarbon dates are correct,Vindija level G1 and Velika Pecina level i arepenecontemporaneous at around 33 ka.Archaeologically, these two levels are similarin that both contain bone points. The lithicindustry in the Velika Pecina level is meager

(six tools), but does include a burin, drilland endscraper on a broken blade that are

242 . . .


21/26

typically associated with Upper Paleolithicindustries. The Vindija G1 lithic industry ismore extensive and also contains distinctlyUpper Paleolithic type tools. But in neither

site do the industries from these levels, northe F complex levels at Vindija, exhibit theartefact profile typical of Aurignacian in theclassic western European sense. Thus, it canbe argued that we find both Neanderthals(Vindija level G1) and early modern humansassociated with similar archaeologicalmanifestations at roughly the same time inthe Hrvatsko Zagorje. However, it is notprudent to consider these manifestationsto represent the Aurignacian in anypan-European sense (Miracle, 1997).

Because this situation is unusual, if notunique, in Europe, the general tendency hasbeen to consider these associations atVindija to be the result of mixing throughcryoturbation. We have noted that there arefactors that do not support this interpret-ation. Chief among these are the absence ofevidence for mechanical alteration on thepertinent stone and bone tools and the fact

that a significant amount of material comesfrom portions of the cave not obviouslyaltered by this process. Furthermore, severalG1 human specimens and artefacts exhibittraces of the distinctive reddish matrix char-acteristic of this level rather than the matrixof the lower G complex or the stratigraphi-cally higher F complex. We certainly do notclaim that these factors prove absence ofmixing. On the other hand, we feel that this

evidence is at least as strong as the evidencefavoring the mixing explanation. With this inmind, alternative explanations to mixingneed to be explored.

If we exclude the mixing by cryoturbationinterpretation, there are four possibilities toexplain the Vindija G1 bone industry: (1)bone points were produced by early modernhumans and were obtained in some way by(or found in association with) Neanderthals;

(2) Neanderthals adopted the technology ofmaking Upper Paleolithic bone points

from contemporary early modern humans;(3) early modern humans adopted thetechnology of making bone points fromNeanderthals; and (4) both groups devel-

oped the technology of making bone points.The third and fourth possibilities areperhaps the least likely, because mostNeanderthal sites yield no evidence of bonetechnology. However, there are a few otherMousterian sites that exhibit some evidenceof bone point production. These are fromthe late Mousterian at the German sitesof Weinberghlen, Grosse Grotte, andVogelherd (Hahn, 1988). Since the originsof Upper Paleolithic bone technology arenot clearly known, the possibility thatthis technology developed during theMousterian should not be categoricallyexcluded. Indeed,Montet-White (1996)hassuggested that massive-base and split-basebone points represent an adaptation to high-land hunting in the regions around the Alpsand Carpathians, beginning not with theUpper Paleolithic, but with the lateMousterian. Furthermore, the recently

described wooden spears from Schningen(Thieme, 1997) along with a lance fromLehringen (Thieme & Veil, 1985), both inGermany, demonstrate that pre-modernEuropeans were capable of producingsophisticated tools from raw materials otherthan stone.

It is difficult to distinguish between thefirst two possibilities in an archaeologicalcontext. At Arcy-sur-Cure (France),Hublin

et al. (1996: 226) explain the possession ofChatelperronian industries by Neanderthalsas a . . . high degree of acculturation, butattribute items of personal adornment totrade rather than imitation of technology byNeanderthals. The functional usefulnessof the Vindija split-base point might bequestionable, because the basal flangesare fragile and the overall dimensions(31156 mm) suggest a structurally

weak point. This could be support for theimitation explanation (C. Roubet, personal

243 /


22/26

communication). Either possibility, trade orimitation, indicates a rather complex patternof interaction between late Neanderthal andearly modern human populations in the

Hrvatsko Zagorje.

The question of industry

Another issue of importance concerns theclassification of these industries at Vindijaand Velika Pecina. For Kozowski (1996:211), the stone tools from Vindija level G1represent a Moustero-Levalloisian assem-blage and the bone tools have affinities

with the Aurignacian. Thus for him, G1must be mixed. Recently, Miracle (1997;Miracle & Crummett, 1995) has demon-strated that the lithic tool assemblage fromVindija levels G1 to Fd is significantlydifferent from the typical pattern ofthe French, and even from the CentralEuropean Aurignacian and is more similarto Szeletian. Furthermore, he thinks that thesplit-base point cannot be used as a typefossil for Aurignacian in Central Europe,because such points also occur with theSzeletian (Szeleta Cave Level 4, DzeravSkla Level 5-11) and in non-Aurignacianlithic assemblages in Istllsk layers 9 to 7.

At Vindija, several Upper Paleolithic typetools derive from level G1, but only one ofthese (and three others from the upper Gcomplex) might be considered definitivelyAurignacian. There are also a few othersfrom the lower F complex and from its

interface with level G1. The F complexcontains remains of early modern humans,so perhaps this represents an intrusive cul-tural entity into this region which has someaffinity with the Aurignacian in a broadsense. In this case perhaps the lack of moretypical Aurignacian stone tools at Vindijais the result of relatively poor-quality rawmaterials, the different technological experi-ences of the Vindija population, or some

type of functional specialization (cf. Hahn,1977).

There is another possibility that warrantsconsideration. At Velika Pecina, as well asthe Slovenian sites of Mokrika Jama andDivje Babe I (Turk & Kavur, 1997), bone

tools like those in Vindija are found with asmall number of stone artefacts that are notespecially typical for any particular industry,including the Aurignacian. In PotockaZijalka (Brodar & Brodar, 1983), anotherSlovenian site, many stone tools show char-acteristics of the Mousterian, but bladesand Aurignacian retouch are also frequent(Brodar & Osole, 1979). This lithic assem-blage is associated with several bone tools

that are very similar to those from theCroatian sites. The industry from PotockaZijalka was at first called Olschewian byBayer (1929). After this designation wasabandoned, the industry was included inAurignacian (Brodar, 1971). More recently,Brodar and Osole (1979) have proposedthat the old name should be restored, not inthe exact sense that Bayer used it, but ratheras a regional cultural term for what hasgenerally been called Central EuropeanAurignacian. Recently this term also wasused byMontet-White (1996)to designatecultural manifestations around the Alps andCarpathians characterized by the presenceof bone points like those from the sitesmentioned here.

Some researchers believe that PotockaZijalka and Mokrika Jama can be clearlyassociated with the Aurignacian (seeBrodar& Osole, 1979; Brodar & Brodar, 1983;

Allsworth-Jones, 1990). However, thearchaeological assemblages of PotockaZijalka display both some differences from,and also some similarities to, other men-tioned sites. Besides many bone points,the archaeological assemblage of PotockaZijalka also contains many stone tools,which is not true for Mokrika Jama, DivjeBabe, Vindija and Velika Pecina. Further-more, the stone tools include many more

distinctively Aurignacian types, particularlycarinated and nosed endscrapers (see

244 . . .


23/26

Brodar & Brodar, 1983; Allsworth-Jones,1990) than at these other sites. The distinctsimilarities to Mokrika Jama and theCroatian sites lie in the presence of charac-

teristic bone tools and a continuation ofMousterian lithic tradition into the UpperPaleolithic. Several years ago,Hahn (1977)suggested that components comprising rela-tively small sample sizes of lithic tools withbone implements likely represented activity-specific (specifically hunting) foci of theAurignacian. While this is a possible expla-nation for the Slovenian and Croatian sites(see alsoMontet-White, 1996), it should benoted that there is no obvious unique hunt-ing activity evident for these sites, whichmight explain their distinctive artefactualpattern.

It may be that the industries from Vindija,Velika Pecina, Divje Babe, and MokrikaJama are closely related and that all repre-sent an early Upper Paleolithic industry thatrepresents an indigenous cultural develop-ment in this region, at least in part. Webelieve that Potocka Zijalka is also closely

related to these sites, but this last siteprovides the best evidence of intrusiveinfluences into this region during the earlyUpper Paleolithic. This influence may wellbe the influx of some type of Aurignacianinto this region, but once present, this intru-sive influence is assimilated into a regionalcultural expression that may have someroots in the local Mousterian.

Conclusion

Level G3 at Vindija contains an essentiallyMousterian lithic assemblage, in whichsome distinctly Upper Paleolithic type itemsare included. Additionally, one rather crudeleaf-shaped bifacial piece is present in thislevel and another one is derived from thelower G complex. These may be forerunnersof the more typical bifacial leaf points of

G1 and the upper G complex. Thus, thereare indications of some technological and

typological continuity between the G3 andG1 levels. We believe that the entire assem-blage from level G1, including the bonepoints, was produced by the Neanderthal

population represented by the human skel-etal remains recovered from this level. Webase our interpretation on the fact that cryo-turbational mixing is unlikely to have causedthe intrusion of neither the single split-basebone point (or all of the other bone tools)from the site nor the Upper Paleolithic typelithic elements into level G1.

Furthermore, we note that an associationof Neanderthals with lithic and other itemsconsidered typical of the Upper Paleolithichas also been demonstrated at the Frenchsites of Saint Csaire (Lvqueet al. , 1993)and Arcy-sur-Cure (Hublinet al., 1996).Thus, such associations at Vindija are notunique and may be also the result of sometype of interaction between Neanderthalsand early modern Europeans, at least inpart. This possibility is supported by thechronological and paleontological indi-cations that there may have been overlap

between these populations in the HrvatskoZagorje around 33 ka. While arteficial mix-ing cannot be completely rejected as a poss-ible explanation for this association, it wouldseem no more reasonable an explanation,given all of the pertinent evidence, than oneinvolving complex interaction between twodifferent Pleistocene human populations.

A similar bone industry is present inVindija, Velika Pecina, Mokrika Jama,

Divje Babe, and Potocka Zijalka associatedwith stone artefacts exhibiting both Middleand Upper Paleolithic features. The stoneassemblages of Vindija level G1, VelikaPecina, Divje Babe and Mokrika Jamaare clearly not Aurignaciansensu str icto,although one Aurignacian tool type ispresent in Vindija G1. In Potocka Zijalkathe number of Aurignacian types present isconsiderably greater. While the situation in

Potocka Zijalka might represent what maybe Central European Aurignacian, the lithic

245 /


24/26

assemblages from the other sites reflecteither a continuation of the Mousteriantechnological tradition (Vindija G1) withthe inclusion of Upper Paleolithic elements

or a regional Upper Paleolithic variant withdistinct connections to the Mousterian.In any case, we conclude that the termAurignacian should not be used for the earlyUpper Paleolithic in these Croatian andSlovenian sites, at least not in any pan-European sense. This is because the earlyUpper Paleolithic in these sites is unique inmany ways and does not conform to theclassic Aurignacian pattern. Furthermore, in

some areas, like the Hrvatsko Zagorje, theearly Upper Paleolithic potentially has rootsin the Middle Paleolithic, from which itdeveloped, combined with what was likelysignificant external influences. It may bethat some of these influences correspond tothe appearance of early modern people inthis region, but it is important to rememberthat some aspects of the early UpperPaleolithic were already developing beforethese populations arrived.

Acknowledgements

We are grateful to the following individualsfor various forms of assistance relative to theinformation presented in this paper: Henryde Lumley, Anthony Marks, PrestonMiracle, Anta Montet-White, MajaPaunovic, Vida Pohar, Jakov Radovcic,Jean-Philippe Rigaud, Colette Roubet,

Darko Rukavina, Jan Simek, LawrenceStraus, Ivan Turk, and Robert Whallon.Miljenko Gregl drew Figures 110 and KimReed drew Figure 11. The paper alsobenefited greatly from the comments ofTerry Harrison and two anonymous review-ers. Financial support was received from:Ministry of Science and Technology of theRepublic of Croatia, University of ZagrebDepartment of Archaeology, U.S. National

Academy of Sciences and Northern IllinoisUniversity. IK has also been supported by

an Institute of International Education(Fulbright) Fellowship. To all of these indi-viduals and agencies we extend our grati-tude. Finally, we owe a great debt to the late

Mirko Malez, who excavated the Croatiansites we discuss and gave each of us theopportunity to study this material.

References

Albrecht, G., Hahn, J. & Torke, W. G. (1972). Merk-malanalyse von Geschossspitzen des mittleren Jung-pleistozns in Mittel- und Osteuropa.A rchaeologicaVenatoria. Stuttgart: Kohlhammer.

Allsworth-Jones, P. (1986).The Szeletian and the tran-

siti on from M iddle to Upper Paleolithi c i n CentralEurope. Oxford: Clarendon Press.Allsworth-Jones, P. (1990). The Szeletian and the

stratigraphic succession in Central Europe and adja-cent areas: main trends, recent results, and problemsfor resolution. In (P. Mellars, Ed.)The Emergenceof M odern H umans, pp. 160242. Ithaca: CornellUniversity Press.

Bayer, J. (1929). Die Olschewa Kultur.Eiszeit undUrgesch. 6,83100.

Bordes, F. (1961).T ypologie du Palolithique Ancien etM oyen. Bordeaux: Delmar.

Brodar, M. (1971). Olschewien. Die Anfangsstufen desJungpalolitnikuns in Mitteleuropa. In (G. Novak,

Ed.) Actes du VIIIme Congr. Intern. UISPP (1),pp. 4352. Beograd: UISPP.

Brodar, M. & Osole, F. (1979). Paleolitske i mezolitskeregije i kulture u Sloveniji. In (A. Benac, Ed.)Praistorija jugoslavenski h zemalja (1), pp. 159194.Sarajevo: Svjetlost.

Brodar, S. & Brodar, M. (1983).Potoc ka Zijalk a.Ljubljana: Slov. akad. znan. umjet.

Franciscus, R. & Trinkaus, E. (1995). Determinants ofretromolar space presence in PleistoceneH omomandibles.J. hum. Evol. 28,577595.

Frayer, D. W. (1992). Evolution at the European Edge:Neanderthal and Upper Paleolithic Relationships.

Prehist. Europ. 2,969.Gorjanovic-Kramberger, D. (1906). Der DiluvialeM ensch von K rapina in K roatien: Ein Beitrag zurPaloanthropologie. Wiesbaden: Kreidel.

Gorjanovic-Kramberger, D. (1913). ivot i kultura dilu-vi jalnog c ovjeka iz K rapine u H rvatskoj(with Germansummary). Djela Jugosl. akad. znan. umjet. 23.Zagreb: Jugosl. akad. znan. umjet.

Hahn, J. (1977). Aurignacien. Das ltere Jungpalo-lithikum in Mittel- und Osteuropa.Fundamenta, Ser.A., Vol. 9. Cologne: Bhlau.

Hahn, J. (1988). Fiche sagaie a base simple de traditionaurignacienne. In (H. Camps-Fabrer, Ed.)Fiches

typologiques de l industri e osseuse prehistorique. Cahier I,Sagaies, pp. 117. Aix-en-Provence: Universit deProvence.

246 . . .


25/26

Hublin, J.-J., Spoor, F., Braun, M., Zonneveld, F. &Condemi, S. (1996). A late Neanderthal associatedwith Upper Paleolithic artifacts.N ature 381, 224226.

Karavanic, I. (1994). Gornjopaleoliticke kamene ikotane rukotvorine iz pilje Vindije (with French

summary).Opvsc. A rchael. 17 (1993), 53163.Karavanic, I. (1995). Upper Paleolithic occupation

levels and late-occurring Neandertal at Vindija Cave(Croatia) in the Context of Central Europe and theBalkans.J. A nthropol. Res.51(1), 935.

Karavanic, I. (1996). The problem of association ofNeanderthals and the Upper Paleolithic in VindijaCave (NW Croatia). Paper presented at the38th Meeting of the Hugo Obermaier Society,Regensburg, Germany.

Kozowski, J. K., Dagnan-Ginter, A., Gatsov, I. &Sirakova, S. (1982). Upper Palaeolithic assemblages.In (J. K. Kozowski, Ed.)Excavation in the Bacho

Kiro Cave (Bulgaria), Final Report, pp. 119167.Warszawa: Panstwowe Wydawnictwo Naukowe.Kozowski, J. K. (1996). Cultural context of the last

Neanderthals and early Modern Humans in Central-Eastern Europe. In (O. Bar-Yosef, L. L. Cavalli-Sforza, R. J. March & M. Piperno, Eds)T he L owerand the M iddle Palaeolithic(5), pp. 205218. Forli:Int. Union Prehist. Protohist. Sci.

Kurtanjek, D. & Marci, V. (1990). Petrografskaistraivanja paleolitskih artefakata spilje Vindije (withEnglish summary).Rad Jugosl. adad. znan. umjet.449(24), 227238.

Laville, H., Rigaud, J.-Ph. & Sackett, J. (1980).Rock

Shelters of the Perigord. New York: Academic Press.Lvque, F., Bocker, A. M. & Guilbaud, M. (1993).Context of a L ate Neandertal. Madison: PrehistoryPress.

Malez, M. (1967). Paleolit Velike pecine na Ravnoj goriu sjeverozapadnoj Hrvatskoj (with German sum-mary).Arheoloki radovi i rasprave Jugosl. akad. znan.umjet. 45,764.

Malez, M. (1970). Paleolitska kultura Krapine u svjetlunovih istraivanja (with English and German summa-ries). In (M. Malez, Ed.) Krapina 18991969,pp. 57129. Zagreb: Jugosl. akad. znan. umjet.

Malez, M. (1971). Yugoslavia. In (K. P. Oakley, B. G.Campbell & T. I. Molleson, Eds)

Cata logue of FossilH ominids, Part I I : Europe, pp. 337343. London:British Museum (Natural History).

Malez, M. (1974). Noviji rezultati istraivanja paleo-litika u Velikoj pecini, Veternici I S{andalji (withGermany summary).Arheoloski radovi I raspraveJugosl. akad. znan. umjet.7,744.

Malez, M. (1975). Die Hhle Vindija: Eine neueFundstelle fossiler Hominiden in Kroatien.Bull. Sci.Conceil A cad. Sci. A rts RSF Yougoslav ie, Section A20(56), 139141.

Malez, M. (1978). Novija istraivanja paleolitika uHrvatskom zagorju. In (. Rapanic, Ed.)A rheolokaistraivanja u sjeverozapadnoj H rvatskoj, pp. 669.I zdanja H rv. arheol. drutva 2. Zagreb: Hrv. Arheol.Drutvo.

Malez, M. (1979). Nalaista paleolitskog i mezolitskogdoba u Hrvatskoj. In (A. Benac, Ed.)Praistorija

jugoslavenskih zemalja ( 1), pp. 227276. Sarajevo:Svjetlost.

Malez, M. (1988). Prehistorijske kotane rukotvorine izspilje Vindije (Hrvatska, Yugoslavija) (with German

summary).Radovi zavoda za znan . rad Jugosl. akad.znan . umjet. 2,217252.

Malez, M. & Rukavina, D. (1975). Krioturbacijskepojave u gornjopleistocenskim naslagama pecineVindije kod Donje Voce u sjeverozapadnoj Hrvatskoj(with German summary).Rad Jugosl. akad. znan.umjet.371(17), 245265.

Malez, M., & Rukavina, D. (1979). Poloaj naslagaspilje Vindije u sustavu clanjena kvartara ireg pod-rucje Alpi (with German summary).Rad Jugosl. akadznan . umjet.383(18), 187218.

Malez, M., Smith, F. H., Radovcic, J. & Rukavina,D. (1980). Upper Pleistocene hominids from

Vindija, Croatia, Yugoslavia.Curr. Anthropol. 21(3),365367.Malez, M. & Ullrich, H. (1982). Neuere palanthro-

pologische Untersuchungen am Material aus derHhle Vindija (Kroatien, Jugoslawien).Palaeont.Jugoslavica29,144.

Malez, M. & Vogel, J. C. (1970). Die Ergebnisse derRadiokarbonanalysen der quartren Schichten derVelika Pecina in Nordwest-Kroatien.Bull . Sci. Con-ceil Acad. Sci. Arts RSF Yougoslavie, Section A 15(1112), 390391.

Miracle, P. T. (1997). The spread of modernity inEurope. In (K. Omoto & P. Tobias, Eds)The Originsand Past of M odern H umans. Toward Reconcilia tion,

Singapore: World Scientific.Miracle, P. T. & Crummett, T. (1995). Just who were

those Aurignacians? Vindijas place in the CentralEuropean Early Upper Paleolithic. Paper presentedat the Sixtieth Annual Meeting of the Society forAmerican Archaeology, Minneapolis.

Montet-White, A. (1996).L e Paloli th ique en ancienneYougoslavie. Grenoble: Jrme Millon.

Radovcic, J., Smith, F. H., Trinkaus, E. & Wolpoff,M. H. (1988).T he K rapina H ominids: A n I llustratedCatalog of Skeletal Collection. Zagreb: Mladost &Croatian Natural History Museum.

Rink, W. J., Schwarcz, H. P., Smith, F. H. & Radovcic,

J. (1995). ESR ages for Krapina hominids.N ature378,24.Rukavina, D. (1983). O stratigrafiji gornjeg pleistocena

s osvrtom na topla razdoblja i njihov odraz unaslagama na podrucju Jugoslavije (with Englishsummary).Rad Jugosl. akad. znan. umjet. 404(19),199221.

Simek, J. F. (1991). Stone tool assemblages fromKrapina (Croatia, Yugoslavia). In (A. Montet-White & S. Holen, Eds)Raw M aterial Economiesamong Prehistoric H unter-Gatherers, pp. 5871.Publications in A nthropology 19. Lawrence: Universityof Kansas.

Simek, J. F. & Smith, F. H. (1997). Chronological

changes in stone tool assemblages from Krapina(Croatia).J. hum. Evol.32,561575.

247 /


26/26

Smith, F. H. (1976a).The Neanderta l Remains fromK rapina: A Descriptive and Comparative Study.Reports of Investigation 15. Knoxville: University ofTennessee, Department of Anthropology.

Smith, F. H. (1976b). A Fossil hominid frontal fromVelika Pecina (Croatia) and a consideration of Upper

Pleistocene hominids from Yugoslavia.Am. J. phys.Anthrop. 44,127134.

Smith, F. H. (1978). Evolutionary significance of themandibular foramen area in Neandertals.Am. J.phys. Anthrop. 48,523531.

Smith, F. H. (1982). Upper Pleistocene hominid evo-lution in South-Central Europe: A review of theevidence and analysis of trends.Curr. A nthrop.23(6),667703.

Smith, F. H. (1984). Fossil hominids from the UpperPleistocene of Central Europe and the origin ofmodern Europeans. In (F. H. Smith & F. Spencer,Eds)The Or igins of M odern H umans: A World Survey

of the Fossil Ev idence, pp. 137209. New York: AlanR. Liss.Smith, F. H. (1991). The Neanderthals: evolutionary

dead ends or ancestors of modern people? J.Anthropol. Res. 47 (2), 219238.

Smith, F. H. (1994). Samples, species, and specula-tions in the study of modern human origins. In(M. H. Nitecki & D. V. Nitecki, Eds)O rigins ofAnatomically M odern H umans, pp. 227249. NewYork: Plenum Press.

Smith, F. H. & Ranyard, G. C. (1980). Evolution ofthe supraorbital region in Upper Pleistocene fossilhominids from South-Central Europe.Am. J. phys.Anthrop. 53,589610.

Smith, F. H. & Ahern, J. C. (1994). Brief Communi-cation: additional cranial remains from Vindija Cave,Croatia.Am. J. phys. A nthrop. 93,275285.

Smith, F. H. & Trinkaus, E. (1991). Les origines delhomme moderne en Europe centrale: un cas decontinuit. In (J.-J. Hublin & A.-M. Tillier, Eds)Auxorigines dH omo Sapiens. Paris: Presses Universitairesde France.

Smith, F. H., Boyd, D. C. & Malez, M. (1985).Additional Upper Pleistocene human remains fromVindija Cave, Croatia, Yugoslavia.Am. J. phys.Anthrop. 68,375388.

Sonneville-Bordes, D. de & Perrot, J. (1953). Essai

dadaptation des mthodes statistiques au Palo-lithique suprieur: Prmiers rsultas.Bull. Soc.Prhist. Fran. 50,323333.

Sonneville-Bordes, D. de & Perrot, J. (1954). Lexiquetypologique du Palolithique suprieur.Bull. Soc.Prhist. Fran. 51(7), 327335.

Sonneville-Bordes, D. de & Perrot, J. (1955). Lexiquetypologique du Palolithique suprieur.Bull. Soc.Prhist. Fran. 52(12), 7679.

Sonneville-Bordes, D. de & Perrot, J. (1956a). Lexiquetypologique du Palolithique suprieur.Bull. Soc.Prhist. Fran. 53(78), 408412.

Sonneville-Bordes, D. de & Perrot, J. (1956b). Lexiquetypologique du Palolithique suprieur.Bull. Soc.Prhist. Fran. 53(9), 547559.

Stringer, C. B. (1982a). Comment on Upper Pleisto-cene evolution in South-Central Europe: A review ofthe evidence and analysis of trends.Curr. A nthropol.23(6), 690691.

Stringer, C. B. (1982b). Towards a solution to theNeanderthal problem.J. hum. Evol.11,431438.

Stringer, C. B. (1989). The origin of early modernhumans: a comparison of the European and non-European evidence. In (P. Mellars & C. Stringer,Eds) The H uman Revolution, pp. 232244.Edinburgh: Edinburgh University Press.

Thieme, H. (1997). Lower Paleolithic hunting spearsfrom Germany.N ature385,807810.

Thieme, H. & Viel, S. (1985). Neue Untersuchungenzum eemzeitlichen Elefanten-Jagdplatz Lehringen,Ldkr. Verden.Die KundeN. F. 36,1158.

Trinkaus, E. & Smith, F. H. (1985). The fate of theNeanderthals. In (E. Delson, Ed.)Ancestors: TheH ard Evi dence, pp. 325333. New York: Alan R.Liss.

Trinkaus, E. & Smith, F. H. (1995). Body size of theVindija Neandertals.J. Hum. Evol. 28,201208.

Turk, I. (1988). Kosceni gumbi iz Divjih Bab I: Pris-pevek k razlagi njihovega nastanka (with Germanysummary).Porocilo o raziskov. paleol. neol. eneol. vSloveniji16,5964.

Turk, I. & Kavur, B. (1997). Survey and description ofPaleolithic tools, fireplaces, and hearths. In (I. Turk,Ed.)M ousterian Bone Flute and other finds from

D ivje Babe I Cave Site in Slovenia, pp. 119156.Operni inst. archaeol. Sloven. 2. Ljubljana:Znanstvenoraziskovalni Centar SAZU.

Valoch, K. (1968). Evolution of the Palaeolithic inCentral and Eastern Europe.Curr. A nthropol. 9(5),351390.

Vukovic, S. (1950). Paleolitska kamena industrija spiljeVindije (with French summary).H istorijski zbornik14,241256.

Wild, E., Steffan, I. & Rabeder, G. (1987/88).UraniumSeries dating of fossil bones, progressreport.Insti tut fr Radiumforschung und K ernphysikWien53,5356.

Wolpoff, M. H. (1980).Paleoanthropology. New York:A. A. Knopf.

Wolpoff, M. H. (1989). Multiregional evolution: Thefossil alternative to eden. In (P. Mellars & C.Stringer, Eds)The H uman Revolution, pp. 62108.Edinburgh: Edinburgh University Press.

Wolpoff, M. H. (1996).H uman Evolution. New York:McGraw-Hill.

Wolpoff, M. H., Smith, F. H., Malez, M., Radovcic, J.& Rukavina, D. (1981). Upper Pleistocene humanremains from Vindija Cave, Croatia, Yugoslavia.Am.J. phys. Anthrop. 54,499545.

248 . . .

Karavanic, Smith

Documents