The “abri du Bison”

1The site sits in a locality called Ségriès, in the municipality of Moustiers-Sainte-Marie, in the south of the Alpes-de-Haute-Provence department and in the heart of the Verdon regional natural park. The shelter opens within a rocky ridge, visible from afar in the landscape (fig. 1). It is 2 m deep for a height of 3 m and a span of 3 m. A concrete construction protects the wall (fig. 2). The shelter is 770 m NGF, and has a South-East exposure.

The bison in its archaeological context

2The Ligurian-Provençal arc probably constituted an important route of settlement during the Paleolithic, bordered to the north by the important glaciers of the Alps and to the south by the Mediterranean Sea. For about thirty years, the discoveries of settlement sites and the study of population dynamics, for example from raw materials (Binder 1994 and 1997, Tomasso 2015), has enabled our knowledge of regional prehistory to evolve. In addition, work carried out since 2012 about fifty km west of Ségriès, in the Jabron valley (site of Prés-de-Laure in Combs-sur-Artuby, Var) revealed an occupation from the Upper Paleolithic, attributable to the Final Gravettian (Porraz et al. 2014, Tomasso et al. 2018).

History of an unusual discovery

3The bison was identified on August 27, 1963 by a Parisian architect, Paul Moscovino. An expertise led by Bernard Bottet at the request of Sylvian Gagnière concluded that the figure was authentic. A molding was made at the same time. In 1966, H. de Lumley published his study on the occasion of the Simposio internacional de Arte Rupestre in Barcelona (Spain).

An original study method

4The current state of conservation of the wall (fig. 3) is making a detailed study of the representation almost impossible. We have therefore studied both the wall and the original molding of the figure (fig. 4).

“This is a bison”

5The parietal system is limited to a single representation: a bison. Other elements engraved in the immediate vicinity have also been identified, but they offer a more modern character both technically and thematically. They correspond to another phase of occupation, or at least of passage, in the shelter.

6The animal is engraved in its left profile (fig. 5). It is complete. The head is small and triangular in shape with a large ear and a pair of horns seen from the front. The forequarters are massive, the withers prominent. There is a single Y-shaped fore member. Two hind legs are visible.

7The jars of the chest are represented by hatching.

8The style of production conforms to what is known in Paleolithic art, beyond the cultural divisions made for the Upper Paleolithic period. H. de Lumley proposed parallels both with the Magdalenian representations from Languedoc (belonging, for the most part, to the old / middle Magdalenian, around 17 000 years BP) and with those from the Pyrenees, such as Niaux cave (middle Magdalenian, around 15 000 years ago cal. BP). It must be said that at the time of its study in the 1960s, there was a lack of sources of comparison, which is no longer the case today.

9Stylistic similarities with some figures from Cosquer cave should be emphasized. We find the same Y-shaped representation of the limbs, and “craddle”-like treatment of the horns (Clottes et al. 2005). However, this type of formal treatment occurs in other chrono-cultural periods, including very recent ones such as the Laborien of Aquitaine (Paillet & Man-Estier 2014).

Times of an unloved engraving

10The first issue with Segriès, as identified in de Lumley’s study (de Lumley 1968a), is the age of the representation. An attribution to the Upper Paleolithic, defended by this author, meant a notable evolution in the landscape of Paleolithic art in South-East France at the time (fig. 7). The geographical isolation of the Bison shelter was mentioned in early publications (Anonyme 1964; de Lumley 1968a), as well as in texts alerting on possible deliberate damages made on the engraving (Clappier 1986). The attribution to the Upper Paleolithic is possible on the basis of several criteria. The site constitutes a milestone in terms of artistic expression between sites in the Rhone valley and, further south, sites on the Mediterranean coast.

11The second issue, which must be considered regardless of the age of the engraving, is conservation. Since its discovery, the motif has been altered by a series of treatments intended to make it more legible (ironing of lines with chalk, posibly charcoal, and casting), but poorly suited to its conservation. The engraving is now extremely degraded. On a macroscopic scale, the wall surface shows detachments of millimeter to centimeter-size flakes. In fact, it is not so much the wall that crumbles, but the superficial layer of concretions that we have studied in detail. We will see that they have played an important role in the preservation of the rock until today (fig. 6).

12While it is exceptional, in the strictest sense of the word, the conservation of an engraved rockshelter like Ségriès is no less plausible, even over millennia. In addition, H. de Lumley pointed out as early as 1966 the existence of what he called a “recovery cullet”, which seemed to him a valid argument to attest to the age of the engraving. In turn, we suggested in 2013 (Man-Estier 2013), that this was indeed an excellent argument. The detailed study carried out in 2014 allows us to go much further on this subject.

13The rock face is locally covered by very thin concretions with the appearance of whitish veils of calcite, millimeter thick, themselves affected by numerous desquamations and exfoliations. Several overlapping films have been identified, some of them separated by thin layers of grayish clay.

14The microstratigraphic study of the support at the level of the bison and of the engraved cross (fig. 8), has shown an interesting chronological sequence which makes it possible in particular to demonstrate that a significant lapse of time has elapsed between the engraving of the bison and that of the cross. Moreover at Ségriès, the execution technique and the identifiable backdrops on both the in situ representation and the molding absolutely exclude the use of a metallic point.

15The presence of the engraved crosses in the immediate area of the bison raises questions. It was carried out by a very deep engraving which largely cut into the rock. In our view, its position is no accident. It testifies to the legibility of the animal engraving at the time of its own realization. One probably wanted to Christianize a place considered as pagan from the presence of the bison representation.

16While prospecting along the rocky ridge, we discovered other historical engravings of Christian inspiration. More precisely, they are two Jesuit monograms (fig. 9a and fig. 9b).

Conservation, a major issue

17The current and future conservation of the engraving, but also of the shelter as a whole, must be considered with the greatest care. Thanks to the analysis of photographic archives, we can offer a dynamic reading of the evolution of the surface condition of the wall. It shows how fast changes are taking place, but it also shows that the representation has not been entirely damaged. In addition, a condition report was carried out, which could be used in the future for monitoring the state of conservation. It seems to us that this little bison, which we thought had been deliberately destroyed, can become an example of the joint work of all the heritage partners in the broad sense.

18Acknowledgements. This work was authorised by the prefecture and funded by the DRAC PACA, Service régional de l’Archéologie. We would like to thank all the scientific and administrative partners we met in 2013 and 2014, and especially Isabelle Dubset, Mathias Dupuis and Xavier Margarit. We would like to thank Émilie Lesvignes for her permission to publish the photographs she took in 2014 as part of the study, as well as Vincent Buccio and Yann Dedonder (Alpes-de-Haute-Provence departmental archaeology service) for their permission to use the topographic survey carried out at the same occasion.
We would like to thank Michel Clappier and Jean-Paul Moscovino for their interest in our study.
The archival research carried out in the context of the survey by J.-F. Devos (Devos 2011) was supplemented by numerous documents provided by the Service archéologique départemental (M. Dupuis), by the team of the Musée de Préhistoire des Gorges du Verdon (I. Dubset) and by the Clappier family (M. Clappier).