Burials are an integral component of modern human settlement in Europe. Especially from Central Europe, we have a sizable record of well-preserved Upper Palaeolithic burials many of them accompanied by numerous ornaments. Of these more notable are finds from Dolní Věstonice II (triple burial), Předmosti (multiple burial), Brno II, Mladeč Cave and child burials recently discovered at Krems Wachtberg (e.g. Valoch, 1959; Teschler-Nicola, 2006; Svoboda, 2007, 2008; Händel et al., 2009). UP human burials are very rare and unique, and every new discovery furnishes new data on past rituals, demographic trends, cultural traditions or local social relations (Vanhaeren & d'Errico, 2002, 2006; Zilhao & Trinkaus, 2002; Trinkaus & Svoboda, 2006; Henry-Gambier, 2008).
The recently discovered assemblage from Borsuka Cave (southern Poland) comprises 112 pendants made from teeth of large herbivores, which may be interpreted as grave goods of the child burial destroyed by natural post-depositional processes. This find represents new data on mortuary practices of Mid Upper Palaeolithic (MUP) societies in Central Europe—in terms of both the age of the buried person and the type of ornaments used as grave goods.
Materials and methods
The Borsuka Cave was discovered during a survey of caves made in 2007 in the southern area of the Kraków-Częstochowa Upland (southern Poland) by Jakub (Nowak, 2007). Excavations in the cave were initiated after the discovery of several bones of an adult woolly mammoth, radiocarbon dated to 24 850 ± 200 (Poz-26124) (Nadachowski et al., 2011). The excavations were carried out 2008–2010. A 12 m2 (4 × 3 m) trench was cut next to the visible cave entrance (Figure 1). A deposit of seven layers was explored and wet sieved (Figure 2). In general, the layers may be divided into two different sequences: an older one, of Pleistocene age (layers V–VII), and a younger one, dated to the Holocene (layers I–IV). Archaeological material recovered from all of the layers except those marked V and VII was dated to the Upper Palaeolithic through to the Middle Ages (Wilczyński et al., 2012b).
Mammal bones other than microfauna were identified by comparison with modern specimens held by the Institute of Systematics and Evolution of Animals, PAS, and with reference to available manuals (e.g. Gramova, 1950; Pales & Garcia, 1981a, 1981b).
The human teeth were described on the basis of their morphological traits (Hillson, 1996). The measurements of the teeth were made using the method of Hillson & Trinkaus (2002). Stages of dental development were determined according to Moorrees et al. (1963) and the later revisions by Liversidge & Molleson (2004) and AlQahtani et al. (2010).
Skeletal elements of different mammalian taxa from Borsuka Cave were quantified in terms of the number of identified specimens (NISP), minimum number of elements (MNE) and minimum number of individuals (MNI). NISP is defined as the number of identified specimens in a collection, where identified means ascribed to taxon. For each taxon, MNI was estimated by sorting—but not matching—left and right elements. The age of the specimen was not considered. MNE is defined as the minimum number of skeletal elements necessary to account for the specimens observed. It is an estimation of the number of skeletal elements represented by specimens in the assemblage, based on the most common portion of the epiphysis considered (Klein & Cruz-Uribe, 1984; Lyman, 1994).
All bones were examined carefully to document possible modifications, namely the traces of processes caused by humans, animal activity and the abiotic environment. Cut marks were identified using criteria established by several authors (e.g. Olsen & Shipman, 1988; Lyman, 1994; Domínguez-Rodrigo et al., 2009). Carnivore modifications and rodent gnawing were identified using the criteria described in Binford (1981, 1983), Lyman, (1994), Haynes (1980, 1983) and Sutcliffe (1970) and other sources.
Traces of manufacture and use-wear on pendants were identified using the following microscopes: Olympus SZX9 stereomicroscope (Wrocław, Poland) with range of magnification ×6.3–57 and metallographic microscope Nikon Eclipse LV100 with range of magnification ×50–100 (Wrocław, Poland).
Layer VI—faunal remains and Upper Palaeolithic assemblage
The most interesting assemblage was obtained from layer VI. Discovered at a depth of 150 cm to the bedrock, this deposit contained fresh and sharp-edged, unweathered limestone rubble within a matrix of greyish-yellow loam. Finds recovered from layer VI included a few skeletal remains of large mammals, human deciduous teeth, a distal fragment of a flint blade and numerous tooth pendants. The pendants and the teeth of a child were discovered in a disturbed arrangement, evidently dragged out (both the human teeth and the pendants) NW in quite a unidirectional–linear rearrangement down the slope (Figures 2 and 3). No grave pit was detected, and this could explain the spatial distribution of the assemblage. Two fragments of pendants made of European elk (Alces alces) and/or steppe wisent/aurochs (Bison priscus/Bos primigenius) incisors discovered in layer VI yielded uncalibrated accelerator mass spectrometry radiocarbon dates of 25 150 ± 160 (Poz-38236) and 27 350 ± 450 bp (Poz-32394). Taken together with the single date from layer VI obtained from a reindeer metatarsus (26 430 ± 180 bp (Poz-38237)—Wilczyński et al., 2012a), all of these dates fall towards the end of the interpleniglacial (MIS 3).
Layer VI yielded a faunal assemblage of over 1931 bone fragments—additionally to 112 pendants (Table 1) belonging to fish, amphibians, birds and mammals. The identified mammals are species associated with a steppe-tundra and taiga environment. Also found were the remains of a few species known to occur in warmer conditions of mixed forest, such as European pine marten (Martes martes), European badger (Meles meles), Eurasian lynx (Lynx lynx) and Artiodactyla (European elk A. alces and aurochs B. primigenius) (Wilczyński et al., 2012a). Most of the animal remains, ~80% (NISP = 1589), belong to small species (amphibians, insectivores, rodents). They represent the prey of raptors and smaller carnivores, and animals that did not survive the winter (amphibians). Smaller mammals (hare with NISP = 82 and Arctic fox with NISP = 47) are represented by relatively numerous remains. The remains of medium-sized mammals (e.g. wolverine, badger and reindeer) and larger mammals (e.g. woolly rhinoceros, horse and bovids) are less numerous. With the exception of the tooth pendants, very likely, from necklaces, only 125 bones and teeth were found from medium and larger mammals. Reindeer (NISP = 44), horse (NISP = 14) and bovids (NISP = 12) are the most numerous among the remains of medium and large mammals. There are represented by bones from different parts of the skeleton but appear to belong to single individuals. Other medium-sized and large mammals are represented by isolated bones or teeth (Table 1).
Table 1. Borsuka Cave, faunal assemblage from layer VI: number of identified specimens (NISP) and minimum number of individuals (MNI)
Despite the small number of remains of medium and large mammals, signs of carnivore activity were identified. Four indeterminate bone fragments have traces of digestive corrosion, and six bone fragments have marks from carnivore gnawing. They were observed on a horse tibia and mandible, and on the epiphysis of an undetermined fragment of a long bone. The human element in layer VI is represented by the small number of human teeth and the pierced teeth from a necklace(s) (pendants made of ungulate teeth). There was no evidence of other human activity, for example, cut marks or percussion marks on the bones.
Personal ornaments—the necklace
Discovered at a depth of 160–220 cm, the animal teeth pendants (112 pieces) included 78 specimens made from incisors and canines of steppe wisent or aurochs (B. priscus/B. primigenius), 34 from incisors and canines of European elk (A. alces). Deciduous and permanent teeth of both taxa were used to make the pendants. Their spatial distribution clearly shows that their original concentration was disturbed by a flow towards the NW (Figure 3). Some of the perforated teeth have fine cut marks on the roots and necks, made presumably when the teeth were harvested from the killed animals. They are similar to the marks observed on finds from Aven des Iboussieres, in south-eastern France (d'Errico & Vanhaeren, 2002). All the roots had been heavily scraped on mesial and distal aspect, which visibly reduced their thicknesses (Figure 4). The perforations were made very close to the root apex by drilling from both sides. On three pendants, traces of polishing of the crown are visible. It is hard to say for what purpose this activity was undertaken. Some of the incisors have traces of a red colourant, mainly in the cut marks. All pendants are probably from a necklace or may have been sewn onto the clothing. It is notable that no identifiable traces of a burial pit were found during the excavation. The original position of the artefacts presumably was in grid square B 5/6 (Figure 3), possibly the site of their original deposition.
Human remains from layer VI
Six isolated deciduous human teeth were found, three in situ during the excavation the other three recovered during sorting of the washed sediments in the laboratory. They were identified as follows: right first upper incisor (udi1), a left second upper incisor (udi2), right and left maxillary first molars (udm1), a left mandibular first molar (ldm1) and a fragment of a cusp of a deciduous molar (Table 2, Figure 5).
According to AlQahtani et al. (2010); midpoint of 1 year.
Crown complete, root partially damaged
R3/4 to Rc
On the lingual surface, the marginal ridges are easily seen, and there is a slight extension from the tuberculum dentale
Crown complete, root damaged
On the lingual surface, the marginal ridges are easily seen at its distal margin, and an extension from the tuberculum dentale is noticeable
Crown complete, root damaged
In the occlusal aspect are trapezoidal, they have three cusps, two mesial (protocone and paracone) and a moderately developed metacone; on the buccal side, the crown bears the molar tubercle of Zuckerkandl
R¼ to R½
Crown complete, root damaged
On the buccal side, the crown bears the molar tubercle of Zuckerkandl
Within the trench, the teeth rested to the west of the concentration of pendants, in grid square B5, C7 and E7. They were the only human remains found at the site. The teeth are all similar in colour: greyish-white with yellow spots and are with a fully developed crown. Only the right udi1 and left udm1 retained their roots, helping to establish their stage of development. The roots of the other teeth survived incomplete, damaged by post-depositional processes or during wet sieving. At the same time, none of them had marks of scratching, perforation or other human modification. Five teeth were sufficiently well preserved to be measured (Table 3). Dental crown metrics were then compared with measurements available from other Upper Palaeolithic sites and also with teeth of medieval and modern human populations (Table 4). The bucco-lingual and mesio-distal measurements of teeth under analysis here occupy an average position among the contemporary Upper Palaeolithic finds.
Table 3. Borsuka Cave, human deciduous teeth, dental metrics, in mm
The teeth appear to be from a 12- to 18-month-old child (infans I) of undetermined sex. All of the roots were in the process of formation rather than resorption, so they were not being shed at the time of dental replacement. Their diagnostic features and the fact that they all belong to the same dental development phase suggest that the teeth may belong to a single individual (Table 2). Dental development patterns have continued to evolve from the Upper Pleistocene to the present age, and human populations today present a higher variability in dental development (Tompkins, 1996, Liversidge & Molleson, 2004); it is important to keep in mind that the infant from Borsuka Cave was a Palaeolithic individual, and its teeth do not necessarily reflect the ontogenetic age of more recent infants. The reasons for the differences in tooth eruption in different populations are currently unknown but may be related to nutrition, disease or genetic factors (Holman & Jones, 1998).
The most important finds from Borsuka Cave are the animal tooth pendants and the human remains, all of them discovered in layer VI. Two radiocarbon dates (about 25 000 and 27 000 bp) secured for the pendants lead us associate these artefacts, and the human remains too, with the Gravettian technocomplex—more precisely, with the Pavlovian culture. At the same time, we cannot rule out the possibility that the pendants and human remains may be linked to a late Aurignacian occupation, as suggested by finds from Central Europe, for example, from Albendorf, Sirgenstein, Bockstein-Törle and Mitoc-Malu Galben (Bachner et al., 1996; Haesaerts, 2007; Conard & Bolus, 2008; Jöris & Moreau, 2010) or from an Epiaurignacian site in Moravia (Oliva, 1996; Svoboda, 2002; Trnka, 2005). The lack of diagnostic stone tool types makes it very difficult to confirm which Upper Palaeolithic culture this material belongs to. That is why we need to concentrate on a stylistic analysis of the pendants from Borsuka Cave.
In Central Europe, human remains from Mladeč Cave in the Czech Republic have been dated reliably to the Aurignacian age, similarly as the remains from Peştera cu Oase, Peştera Muierii and Peştera Cioclovina Uscată in Southern Carpathians (Trinkaus et al., 2003; Wild et al., 2005; Soficaru et al., 2006, 2007; Teschler-Nicola, 2006). A larger number of human remains is known from Pavlovian sites in Central Europe, such as Pavlov I, Dolní Věstonice II and Předmosti, where a few complete human skeletons and isolated bones and teeth were found (Svoboda, 1997, 2008; Trinkaus et al., 2000b, 2010; Trinkaus & Svoboda, 2006). From Poland, we have only some isolated remains of Neanderthals (Stajnia Cave) and anatomically modern humans (Obłazowa, Borsuka and Maszycka caves, and the open-air site Wilczyce) (Kozłowski, 1996; Valde-Nowak et al., 2003; Irish et al., 2008; Urbanowski et al., 2010). The finds from Borsuka Cave are doubly important because deciduous teeth from Upper Palaeolithic sites are rarely found (Legoux, 1975; Frayer, 1978; Sládek et al., 2000; Hillson & Trinkaus, 2002; Teschler-Nicola et al., 2004; Benazzi et al., 2011). The measurements of the teeth from Borsuka Cave show that Palaeolithic and recent human populations are not markedly different. Differentiation in crown metrics of teeth may be connected to sexual dimorphism with male deciduous teeth possibly being bigger than female teeth to some extent (Black, 1978; Żądzińska et al., 2008). However, primary crown dimensions are less sexually dimorphic than those of permanent teeth, and the degree of dimorphism varies within and between populations (Harris & Lease, 2005). Therefore, it is difficult to study the biological relationships between population samples for this period. Every discovery of infant remains brings important information about past human groups. As evidenced by other finds of this type, even infants were considered important and rightful members of Upper Palaeolithic hunter–gatherer communities (Teschler-Nicola et al., 2004, Einwögerer et al., 2006).
The Borsuka Cave assemblage is unusual because of the presence of a larger number of pendants made from teeth of large herbivores in contrast to other Central European MUP burials, which mostly yielded many small ornaments made from pierced carnivore teeth, ivory beads or mollusc shells, most notably the child burials from Krems Wachtberg, Sunghir and Lagar Velho (Bader, 1998; Vanhaeren & d'Errico, 2002; Händel et al., 2009). The size of the pendants discovered in Borsuka Cave—some of the teeth are up to 4 cm long—and their number indicate that they were worn by adults and may be interpreted as a grave offering or, if they were sewn on clothing, could all fit on clothing of an infans child size. Because this assemblage had been redeposited, we cannot say anything about the ornaments of young children.
Aurignacian people made pendants from the teeth of various species of mammals: including the teeth of carnivore species—bear, wolf, fox (Hahn, 1972; Conard & Bolus, 2006; Vanhaeren & d'Errico, 2006). However, it is notable that pendants were also made from the teeth of large ungulates (e.g. European elks and wisent/aurochs), which are known from quite a few Central European sites (Vanhaeren & d'Errico, 2006). For example, pendants made from herbivore teeth (European elk and horse incisors) were discovered in Mladeč Cave (Czech Republic) (Oliva, 2006). The production methods used in making the pendants from Borsuka Cave and from Mladeč Cave are very similar and involved root scraping and drilling.
Pendants made from mammal teeth are found very often at Pavlovian sites. However, the specimens discovered at Pavlov I and Dolní Věstonice I and II (Czech Republic) are mainly made from canines or third incisors of fox (Arctic and red fox) and wolf (Klíma, 1963; 1987; 1994; 1997). Pendants made from the teeth of other animal species (e.g. reindeer or bear) are very rare. Of more than 350 pendants excavated at Pavlov I, only three were made from herbivore teeth (reindeer and Bison/Bos) (Wojtal and Wilczyński, unpublished data). At other Gravettian sites of Central Europe, such as Dolní Věstonice I or Krems-Wachtberg, pendants were also mainly made from carnivore teeth (Klima, 1987; Händel et al., 2009). The method of production of the Pavlovian artefacts (the scraping of the tooth root without bilateral drilling, sometimes punching of the root; Wojtal and Wilczyński, unpublished data), is also different from methods of production observed on the pendants from Borsuka Cave. Only from Trenčanske Bohuslavice site, dated around 25–22 000 bp, has a pendant similar to those from Borsuka Cave (Vlačiky 2012).
The modest faunal assemblage discovered at Borsuka Cave could be interpreted as results of the action of carnivores. It is possible that the herbivore remains belong to the prey of large carnivores (cave hyena or wolf) killed in the vicinity of Borsuka Cave. It is possible that the predators used the cave during short-term visits. The lack of typical features observed in cave hyena dens, such as hyena bones, their coprolites and hyena milk teeth, confirms this suggestion. It is confirmed further by the small size of the cave, which rarely has a width greater than 1 m.
We can think of a number of alternative explanations for the presence of infant teeth and pendants in the Late Pleistocene deposit in Borsuka Cave, as follows:
1.Human activity may explain the finds. In this scenario, the steppe wisent/aurochs and European elk teeth would be a part of a necklace of mixed animal and human teeth, lost by accident in the vicinity of the cave.
This suggestion also has some weaknesses. For example, the roots of the human deciduous teeth are in the stage of formation, some with evidence of secondary damage, but none are with evident and typical marks of attempted perforation or other human activity, such as cutting, drilling or covering in ochre, such as seen on other teeth used as pendants. The conclusion would be that human teeth were not part of the necklace. There is no evidence of modification on them. The discovery of the human teeth in association with so many animal teeth pendants, which differ significantly in their appearance, would confirm that these finds do not belong to a single, lost necklace (Figure 6).
2.The head or body of the child was brought to the site by cave hyenas or other large carnivores. Modern spotted hyenas are able to carry to their dens even very heavy portions of animal bodies such as the head of a giraffe (Pickering, ). A characteristic feature of hyena dens is the presence of a large number of hyena remains (both adult and young individuals), partially digested bones and coprolites. In Borsuka Cave, no digested bones or cave hyena remains were found, and only isolated bones with carnivore gnawing marks, suggesting that the site was not a cave hyena den. If the presence of the child's remains is a result of cave hyena activity, this would have been only an incidental event, and the visit of cave hyena to the cave was casual.
Our view is that carnivore activity may be excluded as an explanation for the finds from Borsuka Cave; this is supported by the spatial distribution of the human remains and pendants (Figure 3). It is possible to identify the exact original spot where the teeth and artefacts were deposited and the direction (north-west) of their later redeposition. If the materials had been the result of carnivore activity, their arrangement presumably would have been less regular. Moreover, carnivore activity does not fully explain the presence of only the teeth of a child and the pendants.
3.A third possibility is that Borsuka Cave is the site of a pendant-making workshop, and the human teeth are not related to this activity. However, the fact that all the animal tooth pendants are finished specimens, and there are no semi-finished or incomplete/broken specimens undermines this explanation. The entire deposit was wet sieved, and no lithic artefacts (except for a single fragment of a flint blade) needed in the production process were discovered.
4.In our view, the best explanation for the presence in the cave of human infant teeth, a large number of animal tooth pendants and no lithic or bone artefacts, and any other signs of human activity, is that the assemblage is what survives of an intentional burial of the child. This is our most grounded hypothesis based on various lines of evidence and known properties of bone material. During the fieldwork, no grave pit was discovered. It is possible that the body was deposited on the surface of the ground, perhaps covered with loose stones. The human remains and pendants were deposited in a small area (ca. 0.5 m2) near the entrance of the cave, below an overhang (Figure ). We believe that this was the original location of the child's body and the necklace of animal tooth pendants. The human remains and the artefacts were intentionally placed on the surface of the ground and next, as a result of natural post-depositional processes were shifted in a NW direction in a unidirectional–linear rearrangement, highly characteristic for abiogenic movement, either of the entire deposits or by running water. The whole deposit containing the human teeth and pendants was wet sieved, but no other human remains were found. Because cave water, rich in CaCO3, is corrosive even to bedrock, if the deposit was soaked often, this may have produced bone mineral dissolution. It could be the reason that only teeth remained.
The necklace of rather large number of animal tooth pendants (at least 112, some 4 cm long) is possible but unlikely to have been an element of such a small child's garment. In our view originally, the pendants were strung together in a necklace (Figure 6), possibly a burial offering. The lack of signs of human activity (lithic and bone artefacts, hearths, remains of killed animals, etc.) supports the idea that the burial place was not related to the occupation of the cave or its vicinity.
The site recently discovered in Borsuka Cave has provided us with material of extraordinary value for the study of MUP mortuary practice and human symbolic behaviour in Central Europe. The human remains, the large number of ornaments, presumably from a necklace, the lack archaeological material associated with domestic activities (such as cores, debitage, flint and bone tools) and the lack of remains of hunter–gatherer mammalian prey all indicate an intentional burial of the child, placed on the surface of the ground, with the necklace as a grave offering. At Borsuka Cave, it is possible to identify two separate events. First, the burial of the child and, second, occasional use of the cave by carnivores. Noticeable is the unusual endowment of the child. It is actually seen in the fact that tooth pendants were collected from at least 20 elk and wisent or aurochs individuals—far more individuals than of any other mammals deposited in entire level VI. Either sewn or worn as necklace, it is assumed that the adults cared to amass a considerable amount of them, renouncing of this sign of status/personhood, or involved in hunting to acquire them especially for this purpose.
The radiocarbon dating and the manufacturing technique of the pendants place the assemblage of interest in the Upper Palaeolithic. Unfortunately, the lack of diagnostic lithic tool types makes it very difficult to attribute the finds from Borsuka Cave to a specific UP culture—Aurignacian or Pavlovian/Early Gravettian.
The faunal assemblage from Borsuka Cave is important as well. For we have here the first (and relatively rich) palaeontological record intimating the presence of forest areas in southern Poland, during what probably were brief episodes at the end of the Upper Plenivistulian (about 27 000–25 000 bp). Layer VI has yielded otherwise rare remains of mammals associated with a forest environment (European elk, aurochs, lynx and beaver). Unfortunately, the small number of bones and teeth of medium- and large-sized mammal excavated in Borsuka Cave makes a more detailed zooarchaeological analysis unfeasible. The lack of butchery traces, few gnawing marks and deposition of animal remains could be interpreted as results of sporadic carnivore activity and winter hibernations.
We are indebted to the four anonymous reviewers for their comments that helped us to significantly improve our paper. Our thanks go to Erik Trinkaus for letting us access his human skeletal material database and Gary Haynes as well as two anonymous reviewer for their valuable comments and suggestions.