Abstract:
The Châtelperronian dates to around 40.5-45 ka cal BP and thus lies at the transition between the Middle and Upper Palaeolithic. So far, mostly sites in France are known, with only a few existing in northern Spain (Soressi & Roussel 2014). The highest density of Châtelperronian sites is located west of the Massif Central, while only a handful of sites are known in eastern France. These include the eponyme site of the Châtelperronian, the Grotte des Fées in Châtelperron; the easternmost known site, Grotte de la Verpillière I in Germolles; and the important Grotte du Renne site in Arcy-sur-Cure (Floss 2003; 2005; Soressi & Roussel 2014). As the Châtelperronian has so far been mainly associated with Neanderthal remains (Bailey & Hublin 2006; Welker et al. 2016), but the technology contains elements normally associated with H. sapiens (d'Errico et al. 2003; Caron et al. 2011; Vanhaeren et al. 2019), the Châtelperronian is an important element of the debate on the so-called ‘cultural modernity’ in Neanderthals. For this reason, the present work begins with the definition of Châtelperronian and its technology, the palaeoanthropological debate and the definitions of ‘cultural modernity’ and the various acculturation hypotheses. This is followed by a presentation of the Châtelperronian of eastern France. In the course of the work, the sites already known there were re-evaluated and further indications of possible Châtelperronian sites were investigated. For this purpose, collections from the Musée Denon in Chalon-sur-Saône, the Musée d'Archéologie Nationale in Saint-Germain-en-Laye, the Musée des Beaux-Arts in Dole, the Musée National de Préhistoire in Les Eyzies-de-Tayac, the Musée Anne de Beaujeu in Moulins, the Integrative Prähistorische und Naturwissenschaftliche Archäologie in Basel and the British Museum in London were newly recorded and analysed. Further sites were identified through literature research and evaluated on the basis of the known sources. The sites included in the work are La Tour Fondue, Theillat, Grotte des Fées and Tilly for the Département Allier, the sites Neuzy-Paray, Grotte de la Folatière, Rue Cataux and Champ de Fourches in Chenôves, La Petite Grotte, Saint-Désert, La Roche in Saint-Martin-sous-Montaigu, several districts in Dracy-le-Fort and Grottes de la Verpillière I and II in the department of Saône-et-Loire. In the Côte d'Or department, the Abri Virely site was analysed, while the Trou de la Mère Clo-chette and Frettes sites and the Montbleuse and Vantoux open-air sites were considered for the Jura department. In the department of Yonne, the Roche-au-Loup, Grotte du Bison and Grotte du Renne sites were analysed for this study, and a final site that became relevant is located in Zeiningen in Switzerland. Due to the large number of sites, different methods of recording and analysing were used. In the case of large collections such as for Grotte du Renne, the focus was placed on individual artefact categories, with particular emphasis on the core technology. In the case of sites with only a few artefacts, all finds were recorded. In these cases, working step analyses were carried out where possible. Some sites could only be assessed on the basis of the literature. During the evaluation of the various sites from the east of France and the north of Switzerland, different degrees of reliability were established, ranging from clear Châtelperronian sites to sites for which the presence of a Châtelperronian is rejected. For the following summary, the sites were divided into four groups: Group 1: Secured sites In the author's opinion, the site has clear evidence for the presence of a Châtelperronian. This includes Châtelperronian points, cores that fall within the technological range of variation of the Châtelperronian, a suitable accompanying industry with a focus on backed elements and end-scrapers on massive blanks, or a reliable stratigraphy or dating. Group 2: Possible sites In the author's opinion, the site can be plausibly assigned to Châtelperronian, but it lacks characteristic features that would make this categorisation incontestable. Sites in this group either have clear Châtelperronian points, which cannot be easily confused with other backed elements, or fall into the Châtelperronian spectrum due to their technology. Group 3: Uncertain sites In the author's opinion, the site contains elements that could fit into the châtelperronian, but there is a lack of reliable evidence. This group includes in particular sites for which there exist only a few possible Châtelperronian points, sites that have no accompanying industry or sites that could not be examined in person for this study and do not have good sources. Group 4: Rejected sites In the author's opinion, the site cannot be considered as Châtelperronian based on the current data. This group includes sites with backed artefactsthat have been incorrectly identified as Châtelperronian points, sites for which only individual pieces exist that have a Châtelperronian affinities but could also be categorised as belonging to other periods, or sites for which the find context is too uncertain. In the author's opinion, the Grotte de la Verpillière I site (Floss et al. 2016), which has at least 48 Châtelperronian points from different zones of the site, is the most secure site of its region. Despite the long and complex excavation history, a remnant of the formerly intact Châtelperronian layer was found during recent excavations, which contained a Châtelperronian point. There exist no cores and accompanying Châtelperronian tools from this layer due to its small size, but from the inventory of the mixed layers and the old excavations, cores could be identified that are similar, at least in their technology, to the Châtelperronian from other sites. For the Grotte de la Verpillière II (Frick 2016b), only a single Châtelperronian point exists, which also stems from a disturbed context. However, since it is an unmistakable specimen and the distance to the VP I site is barely 50 metres, the author considers it plausible to regard VP II as an extension of the Châtelperronian of VP I. As the eponymous site of the Châtelperronian, Grotte des Fées (Angevin et al. 2021) is beyond all doubt. More than 70 Châtelperronian points, at least five Châtelperronian end-scrapers, 77 blade or bladelet cores and an associated Châtelperronian industry were found here. An exact temporal classification of the individual artefacts is not always possible due to the complex research history and the mixed stratigraphy, but in the author's opinion, Châtelperronian tool forms make up the majority of the inventory. The Grotte du Renne (Connet 2002; 2019b) with its three Châtelperronian layers, an inventory of over 80,000 lithic artefacts, including over 1000 cores and 380 Châtelperronian points, is a confirmed Châtelperronian site. Doubts about the stratigraphic integrity are justified, but must be considered on a case-by-case basis. Grotte du Bison (David et al. 2006) is less clear-cut than the neighbouring site of Grotte du Renne. Above all, the relatively recent dating and the lack of images of clear Châtelperronian points make it difficult to categorise the site. In the author's opinion, however, it remains one of the confirmed sites due to its direct proximity to the Grotte du Renne, to which it also has a physical connection. However, further analyses or more detailed publications would be desirable. Due to the confirmation of the presence of Châtelperronian points, Roche-au-Loup (Bodu et al. 2014; Horard-Herbin 1990) can still be considered a confirmed Châtelperronian site. Here too, however, more detailed analyses would be necessary in order to obtain a better picture of this site. Among the sites which, in the author's opinion, can be labelled as possible Châtelperronian sites is La Roche, near Saint-Martin-sous-Montaigu (Herkert 2020). There exist at least nine possible Châtelperronian points from this site as well as an overall high proportion of backed elements. Both Aurignacian and Moustérian occur locally, and the châtelperronian site Germolles lies only about 3 km away. A further investigation of this locality would definitely be of great interest. The data on the finds from Rue Cataux near Chenôves (Gros & Gros 2005) is somewhat more limited, but two of the artefacts are clearly arch-backed points, one of which even appears to be a typological Châtelperronian point. Possible Châtelperronian end-scrapers also exist, and the approximate location of the find spot could be determined. However, it is uncertain whether there is potential for further work. La Tour Fondue near Chauriat (Pasty et al. 2012) shows a stratigraphy from Middle to Upper Palaeolithic, in which possible Châtelperronian artefacts were discovered. Unfortunately, no clear Châtelperronian points have been published and the dates provided are somewhat too recent. However, according to the authors, the accompanying inventory seems to fit into the Châtelperronian, as do the cores. A more detailed publication of the finds would be advantageous in order to allow a clear categorisation. La Petite Grotte in Saint-Vallerin (Floss et al. 2022) is a special case. Clear Châtelperronian points are said to exist, but their discoverers are keeping them away from science, as well as the exact location of the site. Research by the Floss working group has at least made it possible to locate the cave, but without the discovery of Châtelperronian points, Saint-Vallerin must still only be regarded as a possible site. The site of Frettes (Lamotte et al. 2014) is particularly interesting. Here there exist at least three possible Châtelperronian points as well as cores that fall within the range of variations of the Châtelperronian. Since a Moustérien is primarily described at the site and no other Upper Palaeolithic ensemble is mentioned, the finds could be dated to the Châtelperronian. However, as the pieces have not yet been personally examined by the author, no clear conclusion can be drawn at this point. While in the previously mentioned sites Châtelperronian points were the decisive factor for an attribution to the Châtelperronian, Uf Wigg is classified only as a possible Châtelperronian site due to their absence. In the author's opinion, Uf Wigg (Brogli 1975) is an Early Upper Palaeolithic inventory, the technology of which could fall within the range of variation of the Châtelperronian. Without the Châtelperronian type fossil as a guide, however, it is not possible to make a clear classification. Three sites were categorised as uncertain Châtelperronian sites. The first is ‘La Foussotte’ near Dracy-le-Fort, where a possible Châtelperronian point was discovered (Gros & Gros 2005). As this is a surface collection and the context of the individual piece is uncertain, this artefact should be viewed critically. The Neuzy-Paray site (Peyrouse et al. 2007), the exact location of which was unfortunately not possible to find due to a lack of sources, has a clear Châtelperronian point, which is said to stem from an excavation. It is categorised as a possible site, as the context is not clear, the location is unknown and further research is made difficult due to the lack of data. Investigating whether there are other Châtelperronian sites in the region between Germolles and Châtelperron could, however, yield interesting results. Although the pieces from Abri Virely, Saint-Aubin (Joly 1959) have been lost, the two Châtelperronian points are clear enough, at least from the illustration, to be able to speak of a possible Châtelperronian. Several caves along the cliffside of Saint-Aubin were found during the author’s own inspections, and artefacts were also discovered in a badger burrow, although they could not be identified as Châtelperronian. However, it is unclear whether a possible Châtelperronian layer could still be present and the context of the finds is not described. Eight sites that were of interest during the research on the Châtelperronian of eastern France must be rejected as Châtelperronian sites. These include the three artefacts found at the surface sites ‘Champ au Geny’ near Vantoux and ‘Sur la Prairie’ near Mon-tbleuse. Although at first glance they appear to be very similar to Châtelperronian points, the method of retouching is too atypical to speak of a Châtelperronian. If further finds are made in this area that place the three artefacts in a more precise context, this assessment could perhaps change, but at the present time we must refrain from classifying them as Châtelperronian. The Theillat site near Sanssat (Raynal et al. 1989) must also be rejected as a Châtelperronian site, as only one of the four lithic artefacts discovered has a lateral retouch, which is not sufficient for a definition as a Châtelperronian point. While ‘Champ de Fourches’ near Chenôves (Gros & Gros 2005; Herkert 2020) yielded small arch-backed elements, the simultaneous presence of Azil points makes it more likely that the artefacts date to the Azilien. Although the Grotte de la Folatière (Herkert 2020) has a history of occupation from the Middle Palaeolithic through the Aurignacian, Gravettian, possibly Solutrean and into post-Palaeolithic times, a clear Châtelperronian could not be distinguished, and the presence of the Gravettian means that the occasional arch-backed point might well date to this period. The site ‘Les Varennes’ at Dracy-le-Fort (Gros & Gros 2005) has no clear Châtelperronian points, nor does Saint-Désert. No accompanying inventory typical of Châtelperronian could be identified either. Both sites are therefore not categorised as Châtelperronian. Finally, at the site Trou de la Mère Clochette (Feuvrier 1907; Brou 2009c), previous analyses by other authors agree that the arch-backed elements are part of the Gravettian. Some show a stronger affinity to the Châtelperronian, but without a more precise stratigraphic resolution and an accompanying Châtelperronian inventory an assignment to the Gravettian is the most obvious explanation. The site of Tilly was discarded as a possible Châtelperron site, as the evidence for this was due to a mislabelling of Châtelperronian artefacts from other sites. In conclusion, it was shown that eastern France has more evidence for the presence of the Châtelperronian than previously believed. Starting points for further research are new excavations and surface prospections at promising loci in the Département Saône-et-Loire and a targeted search in the ‘gaps’ of the distribution could also lead to new findings. Furthermore, it could be shown that the core technology of the Châtelperronian is similar in many respects to the core technology of the Aurignacian on the Swabian Alb, while the tool inventory in particular differs. Among the similarities are the blade production, which is orientated towards the sequential reduction of different surfaces and is initiated by the creation or exploitation of lateral edges (Hahn 1988; Bataille & Conard 2018; Roussel 2011). During the production of blades, there is an integrated production of technological flakes, which are retouched into Châtelperronian end-scrapers in case of the Châtelperronian and used as carinated burins in the Aurignacien. However, a specific bladelet industry is only documented at some sites in the Châtelperronian (Roussel et al. 2016), while at others a progressive reduction from blades to bladelets takes place. The author defines bladelet cores on massive flakes, which were reduced with low intensity and over an edge, as a typical artefact of the Châtelperronian. They show certain similarities to carinated pieces, but without their systematic bladelet reduction. Differences between the Aurignacian and the Châtelperronian include the regular presence of specialised osseous industry in the Aurignacian (Kitagawa & Conard 2020), the absence of pieces with typical Aurignacian retouch and Aurignacian pointed blades in the Châtelperronian, and the systematic production of bladelets on carinated pieces. Dufour bladelets occasionally occur in the Châtelperronian (Roussel 2014), but there is not yet enough data to define it as a regular tool group of the Châtelperronian. In the author's opinion, the Châtelperronian is therefore much more similar to the early Upper Palaeolithic than to the preceding local technologies, at least in terms of its lithic technology. One possible explanation for this is transculturation (Le Brun-Ricalens 2019), in which the local population - regardless of species - came into contact with other technologies either directly or indirectly and adopted them in a modified form. In the author's opinion, the existence of overlaps in the core technology speaks in favour of indirect contact between the populations, also known as contact along the pathways (Tostevin 2007). However, it will also be necessary to collect further data to answer this question. In particular, a clear identification of the human species that was the carrier of the Châtelperronian is necessary, but also more comparisons between the different technologies that existed at the time of this transition in and outside Europe is needed. For the purpose of such comparisons, the appendix contains illustrations of all the cores included in this thesis. This is intended to circumvent the problem of different terminologies and create the possibility of making far-reaching technological comparisons.