Navigation – Plan du site

AccueilNouvelle série10Rethinking Orgnac 3 in Time and S...

Rethinking Orgnac 3 in Time and Space: Implications for Reconstructing Neandertal Origins

Jason E. Lewis, Jean Combier† et Ludovic Slimak
p. 77-111


L’un des objectifs importants de la recherche paléoanthropologique et archéologique est de clarifier les données qui restent encore mal cernées du Pléistocène moyen (PM). Combler les lacunes du PM eurasiatique nous permet de mieux comprendre l'évolution comportementale et biologique des humains archaïques, des Néandertaliens, des Denisoviens, et de leurs ancêtres. Orgnac 3, situé en France méditerranéenne, est particulièrement adapté pour aborder ces questions. Cet article présente le site à de nouveaux lecteurs, résume les travaux antérieurs réalisés sur le site et son matériel, et propose différents points de vue fondés sur une nouvelle analyse de son matériel faunique. Sur la base des éléments de démonstration actuellement disponibles, il faut envisager que la séquence d'Orgnac 3 soit probablement beaucoup plus ancienne que ce qui est généralement accepté. Cette vaste séquence pourrait ainsi dater des OIS 12-8, et non seulement des OIS 9-8 comme cela a été traditionnellement proposé. Les dents d'homininés, qui n'ont pas été décrites en détail ni assignées à un taxon, dateraient d'environ 375 à 400 000 ans, période centrale du débat sur la divergence entre les populations néandertalienne, dénisovienne et humaine moderne. Des analyses plus exactes et plus précises du matériel paléontologique, archéologique et géologique d'Orgnac 3 sont nécessaires pour clarifier notre compréhension de l'évolution des homininés pendant le PM.

Haut de page

Notes de la rédaction

Received 22.03.2018 – Received in revised form 11.11.2022 – Accepted 03.01.2023.

Author contributions: J. Combier performed original excavations, lithic analyses, allowed access to the site's material, and helped write sections of the paper. J. Lewis and L. Slimak reviewed existing literature, performed renewed analyses on some of the site's material, and co-wrote the overall paper.

Texte intégral

Acknowledgements: We would like to thank Dr. Richard Klein for his support and guidance during our work on Orgnac 3. Thanks also go to Dr. Henry de Lumley and Anne-Marie Moigne for facilitating access to the Orgnac 3 material and associated publications at the CERPT in Tautavel, France, and their hospitality during research visits. Dr. Évelyne Crégut-Bonnoure provided helpful comments on a draft of the manuscript. This research was funded by a NSF Doctoral Dissertation Improvement Grant (#0925571) to JEL, the Leakey Foundation, Stanford University’s France-Stanford Center for Interdisciplinary Studies, Department of Anthropology, and Stanford Archaeology Center. This article is dedicated to our dear colleague and friend Jean Combier who passed away in 2020, and whose life was dedicated to understanding the history of human origins in the Rhône Valley.

1. Introduction

1An important aim of paleoanthropological research over the last two decades is to further our understanding of the biological and behavioral evolution of Middle Pleistocene hominin populations in Eurasia that included the ancestors of Neanderthals, Denisovans, and potentially other archaic Homo populations (e.g. Mellars et al. 2007, Reich et al. 2010, Abi-Rached et al. 2013, Fu et al. 2014, Higham et al. 2014, Meyer et al. 2014, 2016, Prüfer et al. 2014,). A critical component of this endeavor is to clarify “the muddle in the middle” (as it was termed by Isaac in 1975), that is, the poorly resolved record of the Middle Pleistocene. This time period saw the transition from Lower to Middle Paleolithic stone tool technologies, the evolutionary differentiation of Homo heidelbergensis into Neanderthals and Denisovans in Eurasia and Homo sapiens in Africa (Arsuaga et al. 1997, 2014, Rightmire 1998, 2008, Hublin 2007, 2009, Weaver et al. 2008), as well as the purported development of consistent control of fire (Roebroeks & Villa 2011, MacDonald et al. 2021), possibly of dwelling structures (but see Villa 1977, Kolen 1999), and the development of more advanced hunting techniques (Klein 1987). Given the evolutionary import of these events, every potential source of data bearing on these issues is precious.

2Orgnac 3 is one of the rare long European sequences of the Middle Pleistocene site in Mediterranean France that, although very suited to address the aforementioned issues in hominid morphological and behavioral evolution, and having a fair amount of ink spilled about it (see bibliography), has not been treated in a synthetic manner nor included in the majority of discussions on these important evolutionary questions. Exceptions are the site’s inclusion in reviews on the development of consistent control of fire (Roebroeks & Villa 2011), the appearance of the Levallois flaking method and the origins the Middle Paleolithic in Southern Europe (Moncel et al. 2012 and references therein), new radiometric dating analyses on the site itself (Michel et al. 2011, 2013), and how to detect heat-induced changes of fossil bone mineral using Fourier transform infrared spectrometry (Lebon et al. 2010). However, misunderstanding of, or inattention to, details in previous research on the site’s stratigraphy, chronology, and archaeology could call into question the validity of these interpretations.

3This paper aims to introduce the site to new readers, summarize the previous work that has been performed on the site and its archaeological material, offer some different views based on that work, and propose directions for future research, with hope that a more accurate picture of Orgnac 3 will help clarify the “muddle” in our understanding of hominin evolution during the Middle Pleistocene.

1.1 The Middle Pleistocene Record in Europe and the Importance of Orgnac 3

4In Europe, there are a handful of sites that preserve well-dated deposits of Middle Pleistocene faunal remains, stone tools, and in some cases, hominin remains themselves (fig. 1). In Spain, sites in the Sierra de Atapuerca, such as the Sima de los Huesos (Arsuaga et al. 1993) and the Gran Dolina (Carbonell Roura & Rodriguez Alvarez 1994), contain deposits that span ~800 ka to ~300 ka and contain the “best evidence to date for the existence of human habitation in the Iberian Peninsula from the early Middle Pleistocene” (Carbonell Roura & Rodriguez Alvarez 199: 291). Other sites, such as La Caune de l’Arago (France), Torralba and Ambrona and Bolomor (Spain), Isernia la Pinetta and Visogliano (Italy), Boxgrove and Hoxne (England), Schöningen (Germany), Vértesszöllös (Hungary), are some of the largest, relatively well dated, best known, and most studied sites that help fill in the Middle Pleistocene sequence from Western Europe. Figure 2 summarizes the presence or absence of four major types of archaeological and fossil evidence from twelve of the best-studied Middle Pleistocene sites in Europe. All of the sites shown contain stone tools and faunal remains, and some also contain hominin fossils or convincing evidence for the use of fire, but only Arago, Vértesszöllös, Bolomor, and Orgnac 3 contain all of these in their well-stratified deposits. Precise dating of these deposits remains a major issue as they usually don’t contain volcanic material for 40K/40Ar or 40Ar/39Ar dating, leaving mainly U-series, luminescence, electron spin resonance (ESR), and biochronological dating methods.

1. The approximate locations of major European middle Middle Pleistocene sites with early human fossils or artifacts.

1. The approximate locations of major European middle Middle Pleistocene sites with early human fossils or artifacts.

2. Major European Middle Pleistocene sites showing proposed dates and presence of hominin remains, fire, lithics, and faunal remains (using the chronology for Orgnac 3 proposed herein). See Supplement for references.

2. Major European Middle Pleistocene sites showing proposed dates and presence of hominin remains, fire, lithics, and faunal remains (using the chronology for Orgnac 3 proposed herein). See Supplement for references.

5Given that even the best-studied European Middle Pleistocene sites still leave a lot to learn about hominin morphological and behavioral changes over a half-million year time-period, the Orgnac 3 sequence and its material have important potential for answering major questions, such as: a) the timing of the origins of the use of fire and its behavioral and morphological consequences on H. heidelbergensis in Europe; b) the timing of and reasons for the transition from the Lower Paleolithic to the Middle Paleolithic and how the Levallois concept may be considered as involved in that process?; c) were there changes in subsistence behavior over that transition? Were they hunting, and what was the relationship between H. heidelbergensis and other carnivores, etc.?

2. History of Excavation, Geographical and Geological Setting

6The site of Orgnac 3 (Figure 3; 44°17'39.75"N, 4°25'12.65"E) was discovered in 1956 by A. and M. Héritier (two schoolteachers who were visiting the area) because local villagers were uncovering lithic and faunal remains among the rocks they were using to make a ‘cabane de charbonnier’, or charcoal-maker’s hut (Combier 1967), in the middle of the vast bushy ‘garrigue’ vegetation characteristic of the Mediterranean region (Combier 1989). Test pits into the site were performed in 1959 by J. Combier, J.-P. Thévenot, and J. Thiant, making it the first stratified Acheulean site identified in the Rhône Valley of France (Combier 1963). Another larger test excavation was performed in 1962, and the site was systematically excavated between 1964 and 1972, removing several hundred cubic meters of sediment and recording the vertical and horizontal placement of all recovered lithic and faunal material (Combier 1967). A second, smaller and targeted excavation was undertaken in 1988, with the participation of J. Da Silva, B. Gély and J.-L. Porte, in Level 1 of the south-western portion of the site (squares H, I & J 7 – 9), in order to completely recover what has been interpreted as an in-situ single knapping event (Combier 1996; fig. S1). [Please note that all of the referenced work on Orgnac 3, whether published journal articles, book chapters, or theses, can be made available upon request to the authors.]

7Orgnac 3 is located halfway between the Ardèche and the Cèze rivers in the Ardèche department of the Rhône-Alpes region of southeastern France (fig. 1). It is situated at 325 m elevation on the Bourg-Saint-Andéol plateau, which is a large Lower Cretaceous (Urgonian) karstic limestone block. This limestone is compact and white in color, and the block is tilted slightly downward towards the Rhône River to the east. This formation is prone to subterranean erosion, forming well-known and extensive karstic systems such as the Aven d’Orgnac, of which Orgnac 3 (also known as Mattecarlinque) is a part (Jaillet et al. 2007, Delannoy et al. 2009). Orgnac 1, also known as Baume Ronze, is a very large collapsed aven with Upper Paleolithic through end-Neolithic archeological deposits (Combier 1989); Orgnac 2, also known as Baume Flandin, is a smaller cave containing Middle Paleolithic archeological deposits attributed to the Eemian, around the end of OIS 5e (Moncel et al. 2008).

8The site of Orgnac 3, from the surface is 40 m long and about 15 m wide (fig. 3). To delimit the extent of the site, many core soundings were taken in 1968 and again in 1984. In excavation squares C11, D11, & E11, the floor of the substratum (or at least very large boulders where the excavations stopped) is 7 m deep, whereas in square O11 (to the west), it is 2.5 m deep. The northern part of the excavation reached a depth of 6.4 m, which diminishes towards the northeast to 3 m in depth. In the eastern part of the excavation, the depth varies from 4 m to 7 m. In the southern part, the depth of the deposits has not been established (Aouraghe 1992).

3. Aerial plan of the site showing areas of excavation and core soundings (adapted from Aouraghe 1992: 20).

3. Aerial plan of the site showing areas of excavation and core soundings (adapted from Aouraghe 1992: 20).

2.1 Site Formation

9Taking the core soundings, excavation results, and other geological evidence into account, it was concluded that Orgnac 3 is more like a collapsed cave than a true sinkhole (Combier 1967). The western part of the site is hypothesized to be the entrance to the cave (Aouraghe 1992).

10J. Combier (Combier 1967) and E. Debard (Debard 1988) both proposed several stages in the development of the site (fig. 4), progressing from a limestone block to completely filled in sinkhole. In Stage A, the infiltration of groundwater in the mass of limestone in the direction of the dip of the plateau created an underground opening. In Stage B, the cavity grows larger and larger by the processes of erosion and structural shifts. In Stage C, the cavity is the form of a large cave, with rock debris falling from the ceiling and the formation of stalagmites and flowstone slabs. In stage D, the roof partially collapses, leaving a rock shelter. During the final stage, E, the last of the overhangs crumble, and the rest of the basin fills with sediment.

4. Proposed phases of site formation and development for Orgnac 3: left from Combier 1967: 35; right from Debard 1988: 91.

4. Proposed phases of site formation and development for Orgnac 3: left from Combier 1967: 35; right from Debard 1988: 91.

2.2 Stratigraphy

11J. Combier originally divided the Orgnac 3 deposit into 21 sedimentological layers and 10 archaeological levels (Combier 1967, 1979; fig. 5, table 1). E. Debard (Debard 1988), after thorough sedimentological and granulometric analyses, refined this stratigraphy by dividing and subdividing some of these initial layers (fig. 5). S. Khatib (Khatib 1989), based on separate sedimentological and geochemical analyses, proposed grouping the layers into four different sedimentological packages: Package I lies below Level 8; Package II comprises archaeological Levels 8 & 7; Package III comprises archaeological Levels 6, 5a and b, 4a and b, and 3; Package 4 comprises archaeological Levels 2 & 1. S. Laafar (Laafar 1995) proposed yet another grouping scheme based on microfacies analyses (table 1).

5. Stratigraphic sections from Orgnac 3. Left, from Combier 1967: 37), and Right, from Debard 1988: 57). See figure 3 for locations of profiles.

5. Stratigraphic sections from Orgnac 3. Left, from Combier 1967: 37), and Right, from Debard 1988: 57). See figure 3 for locations of profiles.

Table 1. Geological levels, archaeological levels, lithic industries, faunal assemblages, paleoecological and paleoclimatological reconstructions, and chronology for Orgnac 3 (adapted from Aouraghe 1999: 182, and Laafar 1995: 61).

Table 1. Geological levels, archaeological levels, lithic industries, faunal assemblages, paleoecological and paleoclimatological reconstructions, and chronology for Orgnac 3 (adapted from Aouraghe 1999: 182, and Laafar 1995: 61).

12Level 8 is the lower-most archaeological level, situated between 4.25 - 4.0 m depth at the top of S. Khatib’s ‘Ensemble I’ and E. Debard’s ‘layer k’, designated based on the presence of 17 lithic artifacts. It was only excavated over an area of 21 m2 (according to current plans), in squares C-G 8-11. The sediments below this archaeological level were excavated a further ~2.5 m in zones C-G 10-11, but few lithic artifacts were found (although fossil fauna, predominantly carnivores, were present).

13Before discussing Level 8 itself, it is important to understand the history of the opening of the Orgnac 3 karstic cavity and the deposition of the deposits within. E. Debard describes these initial steps:

esting on the substrate, the first deposits were laid down under the ceiling of a cavity that still wasn’t too open to the exterior. The general sedimentation is of a coarse type; it is interspersed with rock-fall material, staying however reduced. The coarse material comes essentially from the walls; the fine material comes essentially from the outside. The deposits follow the general slope of the substrate, which dips towards the southeast.

The presence of the ceiling is attested to by the concretion of sediment at several places in the site; an opening to the exterior is evidenced by gelifracted stones and pieces detached from the walls. This opening, however, was not large enough for predatory birds to roost and reject their pellets. Only some carnivores were able to slip and became trapped inside the aven.

This first phase of sedimentation ended with the loss of a part of the uppermost deposits which no doubt by a down-drawing phenomenon, was washed to the lowest zone towards the east-south-east (Debard 1988: 70).

14At the time of deposition of this layer, the site was still more of a cave with a vertical/slanting opening than a sinkhole; after the initial carbonate dissolution and opening of the cavity (which most likely took place during a warm wet period,) a cool dry period occurred which brought in aeolian sediment (that forms the ~2.5 m of non-archaeological layers below Level 8 (Debard 1988: 70, Khatib 1989: 174). The slope of the bottom of the cavity and these layers suggest that the site was not likely to have been preferred by hominins as an occupation site, and the faunal remains in these layers might well represent animals that fell into a shaft-like opening to the cavity, or came in on their own volition to hibernate, and died there (Khatib 1989: 174). E. Debard noting the rolled appearance of the lithic artifacts from Level 8 and the sedimentological facies of this layer, suggests that “essentially, the material was brought in from outside by running water, which flowed from West to East. This resulted in a beveled deposit [laterally thinning]. The running water caused the reworking of a previous Acheulean habitat, the industry which, being heavily rolled, forms archaeological level 8” (Debard 1988: 90).

15Archaeological Level 7 is situated between 4.0 - 3.5 m depth in the upper portion of S. Khatib’s ‘Ensemble II’ and E. Debard’s ‘layer j’. It is separated from Level 8 by nearly a half meter of archaeologically sterile material. E. Debard describes the geological context:

This is the principle phase of the opening of the cavity in the southern part of the site. Coarse sedimentation is predominant. The distribution of blocks and stones and the evolution of their size allow the following of the collapse of the ceiling, from west to east, from south to north, going along through the deposits. Thus, just until the last collapse, at the origin of layer j1, a ceiling existed above the infill in the zone of squares D and C. This is attested to by the concretizing of sediment, no doubt located directly below ancient gutters, and the stalagmite in place, formed on top of layer j3 in square C8. The progressive opening of the cavity allowed the installation of predators and humans (Level 7) (Debard 1988: 90).

16Allochthonous material was also brought in by running water and colluvial action, such as silts and course sands from the plateau, and corroded stones from the karstic system itself (Khatib 1989: 166). S. Khatib and E. Debard both signal the importance of carbonate precipitation during this period, with the deposition of the stalagmite and the flowstone. Specifically, S. Khatib stated “The two stalagmitic slabs of Orgnac 3 are located in the same stratigraphic level. They were formed at the same time, in an environment rich in calcium carbonates and thanks to the presence of microorganisms, under a warm climate” (Khatib 1989: 147).

17Level 6 is located between 3.5 – 3 m depth, at the base of S. Khatib’s ‘Ensemble III’ and in E. Debard’s ‘layer i2’. According to E. Debard, starting with layer i2 and continuing to d1 (archaeological Level 2):

This phase is characterized by instability of the walls, which is at the origin of a constant course sedimentation. The northern parts of the ceiling continue and eventually finish collapsing. The cavity was largely open; a part of the coarse material was probably coming from the exterior at the same time as the fine material. The latter have a marked aeolian origin. Permanent percolation of water, tied to the proximity of the walls, deposited carbonates beneath the stones.

Humans were present in several episodes. They installed themselves in a sort of depression, largely open and which progressively continued to crumble during the sedimentation. The walls, no doubt higher than the current level of the limestone, gave them shelter against predators and the elements (Debard 1988: 90).

18Level 5b is located between 3- 2.5 m depth, at the middle of S. Khatib’s ‘Ensemble III’ (sub-Ensemble IIIa) and in E. Debard’s ‘layer hb’. It is separated from Level 6 below by archaeologically sterile layers i1 and hc (table 1; Debard 1988) but directly contacts Level 5a above. The geological context is broadly the same as described by E. Debard for Level 6. S. Khatib describes this layer as containing an abundance of large stones and blocks, with large amounts of pyroxene and quartz. The autochthonous material comes mainly in the form of stones and large blocks from the walls and remaining ceiling of the cavity, whereas the allochthonous material is mainly silts, coarse sands and quartz colluvially entering the cavity from the surrounding plateau.

19Level 5a is located between 2.5 - 2.0 m depth, in the middle of S. Khatib’s ‘Ensemble III’ (interface between sub-Ensembles IIIa and IIIb) and in E. Debard’s ‘layer ha’. Level 5a lies directly on tip of Level 5b, but appears to be separated from Level 4b by a thin layer of sterile sediment.

20Level 4b is located between 2 - 1.8 m depth, in the middle of S. Khatib’s ‘Ensemble III’ (sub-Ensemble IIIb) and in E. Debard’s ‘layer g’. Level 4b is separated from Level 4a by archeologically sterile ‘layer fb’ (composed of large fallen blocks). S. Khatib noted a larger proportion of clays, sphenes, potassium feldspaths and plagioclases, a higher grain-size of the sands and a diminution of quartz (Khatib 1989: 167). Level 4a is located between 1.8 - 1.5 m depth, in the middle of S. Khatib’s ‘Ensemble III’ (top of sub-Ensemble IIIb) and in E. Debard’s ‘layer fa’. Level 4a directly underlies Level 3.

21Archaeological Level 3 is situated between 1.5 - 1 m depth at the top of S. Khatib’s ‘Ensemble III’ (sub-Ensemble IIIc) and is E. Debard’s ‘layer e’. It is directly between Levels 4a and 2, with no discernible sedimentary hiatus or erosion between layers. S. Khatib describes how this layer has the highest content of silts and quartz, and the lowest content of fine sands in the entire site (Khatib 1989: 167). The autochthonous material comes mainly in the form of stones and large blocks from the walls and remaining ceiling of the cavity, whereas the allochthonous material is mainly silts, coarse sands and quartz colluvially entering the cavity from the surrounding plateau and, starting in Level 3, a strong aeolian component with an increase in zircons and micas.

22Archaeological Level 2 is situated between 1 - 0.25 m depth at the base of S. Khatib’s ‘Ensemble IV’ (sub-Ensemble IVa) and E. Debard’s ‘layer d2’. It is directly above Level 3, but separated from Level 1 by E. Debard’s ‘layer d1’ which, as described below, is a pedogenic horizon that was partly removed by erosion, before the deposition of ‘layer c’ that contains Level 1. E. Debard describes the geological context of this layer:

This is a very complex phase, of which the importance and the duration are largely unknown. Several phenomena occurred, and it is likely that traces of certain ones were not preserved.

A first group of events comprise the formation of a certain number of deposits after layer d1 and their pedological transformation. Layers d1 and d2, consolidated in breccia, constitute the pedogenic carbonate accumulation of a paleosol horizon.

A second group of phenomena, of an erosive type, caused the disappearance of the upper layers of the paleosol and a part of layer d1. The upper part of the latter is dug into by small basins and craters that sometimes almost completely dislocate the layer (Debard 1988: 90).

23S. Khatib concurs, describing how the cavity was almost full, with few rocks being eroded from the walls and most of the sediment, predominantly clays, is of aeolian origin (Khatib 1989: 169).

24Archaeological Level 1 is situated between 0.25 - 0 m depth at the top of S. Khatib’s ‘Ensemble IV’ (sub-Ensemble IVb) and is E. Debard’s ‘layer c’. Level 1 is separated chronologically from Level 2 by a potentially considerable amount of time. The geological context and origin of sediments are broadly the same as described by E. Debard and S. Khatib for Level 2, with S. Khatib noting that the roof had totally collapsed and the presence of an increased proportion of aeolian transport of plagioclases, pyroxenes, and zircons (Khatib 1989: 170).

25S. Laafar (Lafaar 1995) conducted micromorphological analyses on the entire Orgnac 3 sequence, providing detailed color plates of thin sections magnified under natural and polarized light. Laafar’s results are generally in agreement with E. Debard’s and S. Khatib’s interpretations, and bring further evidence about the site’s post-depositional history deserving of special attention. In geological layers i and j, corresponding to Levels 6 and 7, he notes “Carbonate features are represented by recarbonation in the basal mass and microsparite particles. Dissolution phenomena, especially isolated gaps and voids, affect the calcareous elements (wall fragments or stalagmitic encrustations) and the basal mass. [In layer i], this subset contains very abundant pebbles (limestone and fragments of stalagmitic slab) with an open fabric. They present polarized carbonate crusts (on the lower surface), also affected by dissolution voids” (Lafaar 1995: 44). About the fossils from these layers, he states “The bones are fragmented with fissures filled with secondary carbonates... The phosphate deposits are restricted to fragments of coprolites with detrital inclusions, bone, and carbonate epigenizing deposits (phosphate encrustation)” (Lafaar 1995: 46).

26About layers h to f (Levels 5b to 4a), Laafar reports “The basal mass and especially the fine fraction, show an enrichment in sparitic grains, of reddish-brown color. Some voids are filled by sparitic carbonates, in the form of root pseudomorphosis. Up sequence, the carbonates impregnate all the basal mass as well as the biological structures” (Lafaar 1995: 48), and “Bones are abundant and fractured... and sometimes affected by secondary sparitic carbonates. Some bones are embedded in rounded aggregates” (Lafaar 1995: 48). For layers e and d (Levels 3 and 2), he states “Carbonate features are present in the form of an intense impregnation of layer d. This impregnation is constructed by clear sparites and microsparites partially impregnating the basal mass and invading the sand grains. Towards the top of this layer and to the east of the site, we have associated with the sparitic facies, the acicular facies visible in the voids” (Lafaar 1995: 52), and “includes an abundance of bone-like elements, calcined and ferruginized by oxides whose voids are filled with microsparite” (Lafaar 1995: 52). Please see figure S2 for images taking during our work of how this post-depositional alteration, brecciation, and expansion have affected macrofaunal fossils in the site.

27A. Djerrab (Djerrab 2001) conducted a magnetic susceptibility study of the Orgnac 3 layers, and S. Laafar, S. Abdessadok and A. Djerrab (Laafar et al. 2008) published a combined micromorphological and magnetic susceptibility analysis of the hearths in the sequence.

2.3 Radiometric Dating

2.3.1 Volcanics

28Volcanic sediments were found in very small amounts all along the Orgnac 3 sequence. Green clinopyroxene was found in Level 2 of the site. E. Debard and J.-F. Pastre correlated these sediments to those deriving from an eruption of Mont-Dore (located a few hundred kilometers to the northwest), with the authors stating, “It is very probably correlated with a trachyandesite emission from the final eruptions of the Sancy [volcano] appearing around 0.30 Ma” (Debard & Pastre 1988: 1517). Additionally, fission-track dating of the same sediment provided an age of 298 ± 55 ka for Level 2 (Khatib 1989, 1994). Aeolian sphenes were found in the sequence as well, with E. Debard and J.-F. Pastre stating “We note however some fresh [sphene] in layers j4 and i2 [archaeological Levels 7 & 6 respectively] which could come from acidic trachyandesite emissions which mark the beginning of the terminal phase of activity at Sancy around 0.4 Ma” (Debard & Pastre 1988: 1517).

29V. Michel et al. (Michel 2011, 2013) performed new 40Ar/39Ar total-fusion isotopic age estimates on 57 sanidine grain populations from the same volcanic source in Level 2. Eleven were excluded from the results for producing ages that were either too old (>550 ka) or too young (277.8 ka). The 2σ weighted mean of the remaining 46 samples is 302.9 ± 2.9 ka, which is agreement with previous estimates (table 2).

Table 2. Every radiometric date ever published for the Orgnac 3 sequence. Please see main text for background, discussion, and references.

Table 2. Every radiometric date ever published for the Orgnac 3 sequence. Please see main text for background, discussion, and references.

2.3.2 Uranium-series and Electron Spin Resonance

30Most of the radiometric dates on the Orgnac 3 sequence derive from uranium-series and ESR analyses of samples taken from both the flowstone slab and 47 cm high stalagmite located in the lower levels of the site (fig. 6). G. Shen (Shen 1985) stated that the flowstone slab formed on top of Level 6, and that the stalagmite formed on top of Level 8 and reached to Level 6. Subsequent sedimentological, geomorphological and geochemical studies indicate that both the flowstone slab and stalagmite are in Level 7 (Debard 1988, Khatib 1989: 145). As seen in figure 6 (and others in Michel et al. 2013), the stalagmite and flowstone slab do not show the pure calcite structure of dripstone-like speleothem, and instead display a poorly bedded, porous, and chalky texture. Whether this appearance is due to how the carbonate was precipitated (biomediated or not; Pedley 1990, Gandin & Capezzuoli 2008), or the post-depositional alteration described above and below, this material appears quite susceptible to contamination by remobilized uranium, thorium, or other materials, which would especially affect uranium-series and ESR dates (Srdoc et al. 1994). As carbonate precipitation requires water to form, E. Debard and S. Khatib place the 'stalagmite' and 'flowstone slab' in Level 7 where there is sedimentological evidence for water action. There is no such evidence in Levels 6 - 5a.

6. Comparison of the (A) flowstone (left) and stalagmite (middle) from Orgnac 3, with a closeup of the stalagmite (right) from Michel et al. 2013, to cross sections of (B) travertine speleothem (Luray Caverns, Virginia, USA, photo credit: James St. John), and (C) typical dripstone speleothem calcite (from figure 2 of Schwarcz 2007).

6. Comparison of the (A) flowstone (left) and stalagmite (middle) from Orgnac 3, with a closeup of the stalagmite (right) from Michel et al. 2013, to cross sections of (B) travertine speleothem (Luray Caverns, Virginia, USA, photo credit: James St. John), and (C) typical dripstone speleothem calcite (from figure 2 of Schwarcz 2007).

31G. Shen performed uranium-series dates on two samples from the flowstone slab and two from the stalagmite, one from the base and one from the top. When describing these samples, G. Shen stated:

Samples ORG82-1 and ORG82-2 are from the slab. The light-yellow calcite is quite pure and well-crystallized. Slices of bones and pieces of small stalactites fallen from the ceiling are visible in the chemically precipitated calcite, attesting to the in situ formation of this slab” (Shen 1985: 135), and “Samples ORG82-3 and ORG82-4 represent respectively the top and the base of the stalagmite. This stalagmite does not seem to be very well preserved; indeed, some layers are detached and some alveolar zones are filled with clay, which implies partial dissolution due to groundwater. We have proceeded to a very careful selection of the samples to be analyzed, and small pure and compact pieces are still available (Shen 1985: 135).

32With all these qualifiers, G. Shen produced ages of 374 +165/-94 ka and 356 +266/-76 ka for the slab, 364 +204/-69 ka for the top of the stalagmite and 288 +82/-45 ka for the base (table 2). This is a wide range, and in reverse stratigraphic order. G. Shen continued:

The calcite from this site is characterized by its low uranium content and low uranium isotope activity ratio. The 230Th/232Th ratios vary between 15.7 and 18.5. Due to the very old age of these samples, correction of the initial 230Th is not necessary. The four 230Th/234U ratios are identical within the statistical error of 1 sigma. This indicates the good reproducibility of our measurements. The fact that all these ratios are close to secular equilibrium gives a slightly different aspect to what happens in the dating of the Caune de l'Arago or Petralona cave where this ratio sometimes exceeds secular equilibrium. The difference between equilibrium and the average of these four ratios, 0.953 ± 0.016, is statistically significant, and it is therefore likely that the age derived from this last value, 339,000 years, better reflects the actual age of these stalagmitic formations (Shen 1985).

33It should be noted that the 230Th/232Th activity ratios reported for these samples are all less than 20, which is considered indicating contamination by detrital materials (Michel et al. 2013, Martínez‐Aguirre et al. 2019). Shen concluded that given the poor state of preservation of the samples, and that in 1985 this antiquity was at the upper limit of the U/Th technique, it was difficult to distinguish between the ages of the different samples and to rule out the possibility that the real age of that level was older than 350 ka (Shen 1985: 135-137). C. Falguères (Falguères 1986) calculated an Electron Spin Resonance (ESR) date for a sample from the base of the flowstone slab, giving an age of 309 ± 34 ka. C. Falguères et al. (Falguères et al. 1988) re-reported the results from G. Shen (Shen 1985) and C. Falguères (Falguères 1986).

34H. Masaoudi (Masaoudi 1995) performed U/Th dating on bones, enamel, and dentine, as well as ESR dates on those materials and the stalagmite, flowstone slab, and burnt quartz. The U/Th dates for bones, enamel and dentine are incoherent and out of chronological order; when explaining this result, H. Masaoudi stated “These results show that the site of Orgnac 3 has known very important geochemical alteration. The opening of the site to the migration of uranium has been made evident. The lower levels are the most affected by this alteration. It proves difficult to apply the U/Th method to the bones, dentine and enamel of this site” (Masaoudi 1995: 69). The ESR dates on bones, enamel and dentine are also very incoherent and out of chronological order, often getting younger going down the sequence (table 2); when explaining this result, H. Masaoudi stated “No matter which incorporation model is used, the ages obtained by ESR on enamel are too young in relation to the geochronological and biostratigraphic data from the site, except for Level 1 which gives coherent ages” (Masaoudi 1995: 79). For the ESR dates on the stalagmite, flowstone slab, and burnt quartz, H. Masaoudi presents sets of wide dates, but does not make any mention of how the geochemical alteration or other problems he previously described might affect those dates. It is important to note that H. Masaoudi’s ESR dates for burnt quartz, the one dating method that is less vulnerable to geochemical alteration because its ‘clock’ is based on a high temperature event rather than the deposition of carbonates, are coherent and are around 100 ka older than the stalagmite/flowstone slab dates (table 2), yet come from the 3 levels above the stalagmite/flowstone slab (Masaoudi 1995: 84).

35V. Michel et al. (Michel et al. 2011) performed new MC ICPMS U/Th analyses on four carbonate samples from throughout the stalagmite and flowstone slab. These authors place the stalagmite and flowstone slab generally in Levels 7, 6, and 5b, without more precision or explanation of how they arrived at this new stratigraphic interpretation. They do not report any resulting data, only that their age estimates range from 261 ± 6 to 305 ± 14 ka (table 2).

36V. Michel et al. (Michel et al. 2013) performed 14 additional ICP-MS U/Th analyses on carbonate samples from throughout the stalagmite and flowstone slab. These authors now place the stalagmite in Levels 6 & 7, and the flowstone slab in Levels 6 & 5b. The upper 3 of the 8 samples from the flowstone slab were excluded from age calculation due to low 230Th/232Th activity ratios caused by “contamination by detrital materials” (Michel et al. 2013). One of the 6 samples from the stalagmite was excluded from the results as it was much too young compared with the adjacent samples, the authors stating that this sample was “probably biased by post-depositional diagenesis” (Michel et al. 2013). The two-third of samples that were retained produced ages that are also incoherent and out of chronological order (table 2) with dates ranging from 265 ± 4 ka to 312 ± 14 ka, but not in a logical sequence from top to bottom stratigraphically. When trying to account for the differences between this date range and the 302.9 ± 2.9 ka 40Ar/39Ar date from the top of Level 2 higher up in the sequence discussed earlier, they do not question the validity of the U/Th dates and instead argue that either there was “a slight contamination of sanidine grains or by a minor excess of 40Ar” or “the volcanic minerals were transported to the site tens of thousand years [sic] after the Sancy eruption” (Michel et al. 2013: 5). They make the latter argument despite showing very well preserved sanidine grains under SEM magnification in their figure 7 and stating in the legend that “The minerals are sharp-edged and unweathered” (Michel et al. 2013: 8).

7. Two separate age models that can be constructed from published data for the Orgnac 3 deposits (background adapted from Khatib 1989: 10).

7. Two separate age models that can be constructed from published data for the Orgnac 3 deposits (background adapted from Khatib 1989: 10).

37Most recently, J.-J. Bahain et al. (Bahain et al. 2022) published new thermoluminescence (TL) ages on burnt flints from Levels 5 & 6 (though it is not explained why Levels 5a & 5b were combined), and recalculations of H. Masaoudi’s (Masaoudi 1995) ESR/U-series dates on mammal teeth. These ESR/U-series recalculations were informed by in situ dosimetry data collected via dosimeters installed into Levels 4b, 5a, 5b, and 6 (though no information is given about when or under what auspices this dosimetric fieldwork was conducted). Six TL dates from Level 5 range from 210 ± 20 ka to 332 ± 35 ka, with an average age of 262 ± 36 ka; five dates from Level 6 range from 244 ± 21 ka to 337 ± 29 ka, with an average age of 286 ± 30 ka (table 2). The seven recalculated ESR/U-series dates from Levels 1 through 7 range from 236 ± 2 ka to 311 ± 30 ka, but are not always in chronological order down the sequence (table 2). Based on these dates, they propose that Level 8 and below were deposited during OIS 10 (ca. 375 - 340 ka); that Levels 7 & 6 were deposited during OIS 9 and the OIS 9 - 8 transition (up to Amargiers interstadial; ca. 340 - 270 ka); that Levels 5 through 1 were deposited during OIS 8 (post Amargiers interstadial; ca. 270 - 240 ka). When trying to account for the differences between this date range and the 302.9 ± 2.9 ka 40Ar/39Ar date from the top of Level 2, they argue that while it “is analytically accurate, it likely corresponds to reworked minerals in the sediments associated with an older volcanic event disconnected from the archaeological record, and it should be considered with caution as the dated minerals are probably reworked into the Orgnac 3 sequence” (Bahain et al. 2022: 9).

3. Archaeology and Paleoanthropology of Orgnac 3

3.1 Artifacts

38The site of Orgnac 3 has also yielded just over 50,000 stone artifacts (table S1, modified from Moncel 1999), over half of which come from Level 1 alone. The vast majority are knapped on local flints which outcrop a few kilometers from the site in the form of long plates. A small number of pieces are knapped on flint from the Rhône Valley, which is 15  km away, and also on limestone, basalt, and quartzite. The proposed presence of complete chaînes opératoires suggests that intense flaking activities took place at the site, although details of the spatial associations of these chaînes opératoires are not described (Aouraghe 1992, Moncel et al. 2005). Work by M.-H. Moncel et al. (Moncel et al. 2005) have shown that at least in Levels 2 & 6, dense concentrations of stone tools occur in association with dense concentration of faunal remains, and that these are often spatially associated with ash and charcoal lenses that could be interpreted as hearths. No fluvial sorting or preferential orientation has been observed for the material, and refits generally occur between pieces that were located close to one another. The lithic material presents a white patina on most flints. Rare physical alteration (notches and breakage) suggests little if any post-depositional movement or slumping of the sequence.

39The Orgnac 3 sequence is important to understanding the evolution of lithic technology, as it contains the gradual appearance of Levallois flaking methods. Levallois flaking is not currently attested in the deeper Levels 7 through 4b, then gradually appears in Level 4a and reaches a frequency of 20% in Level 2 (Combier 1967). Moving vertically up the sequence, the number of bifaces diminishes, while the number of choppers and chopping tools increases. Scrapers are present throughout the sequence, but the flaking techniques used to make them are considered as more standardized, with thinly backed scrapers with bifacial retouch present in Level 3 (Combier 1967). Level 1 is the richest in lithics, with over 27,000 artifacts, amongst which 80% are flakes and cores. The lithics of this level are characteristically Levallois, with a predominance of scrapers, and also the appearance of denticulates, but bifaces are absent (Combier 1963).

40The archaeological levels were initially grouped into four units based on lithic technology: Level 1 as Pre-Mousterian, Levels 2 - 3 as Upper Acheulean III, Level 4 as Upper Acheulean II, and Levels 5, 6, & 7 as Upper Acheulean I (Combier 1967). The Pre-Mousterian is characterized by containing a large amount of Levallois flaking methodologies. The Upper Acheulean III is characterized by “a clear tendency for ‘mousterianisation’, despite having very pure [stylistically and technically real] Acheulean bifaces... The scrapers, whose representation by types barely changes, are better than in the former levels; the percentage of thick forms of Clactonian removals lessens, as do the scrapers, burins, and perçoirs” (Combier 1967: 91). The Upper Acheulean II is different from the industries that surround it. “It is less rich, in total, in scrapers, while correspondingly the richest in notches and denticulates; they are everywhere in Level 4b” (Combier 1967: 91). These Upper Acheulean sub-designations have not been widely discussed or used subsequently in the Paleolithic literature.

41Anglophone authors have included data from the Orgnac 3 lithic material in their analyses. In 1985, H.L. Dibble included length, width, and thickness measurements from the Levallois flakes from Levels 2 and 3 of Orgnac in his analysis of the effect of variation in lithic raw material on flake manufacture (Dibble 1985). G.F. Monnier (Monnier 2006) included length, width, and thickness measurements, as well as data on the type and location of retouch and the degree of symmetry of retouched tools in her analysis of whether tool types become more standardized over time. His results allow him to “reject M.-H. Moncel and J. Combier’s (Moncel & Combier 1992a, 1992b) claims that retouched tools at Orgnac 3, particularly scrapers and ‘convergent tools’, become more standardized from the oldest to the youngest levels” (Monnier 2006: 76). J. Shea (Shea 2009) included measurements of maximum width and thickness on points (taken from M.-H. Moncel [Moncel 2005] and unpublished data) in his analysis of the development of projectile weapon technology. His results suggest that the lithics from Orgnac (indeed, all European MP lithics) were not used as projectiles, and that, for him, projectile weapon technology was first developed “among African Homo sapiens populations between 50 and 100 ka” (Shea 2009: 195).

42More recent work has continued to examine the evidence at Orgnac 3 for the emergence of Levallois core technology (Moncel et al. 2020), and technological and traceological analysis of percussion activities on both lithic and faunal material (Mathias et al. 2021).

3.2 Hominid Remains

43Seven hominid teeth were discovered from Levels 5a, 5b, and 6. An adult upper right canine was discovered in 1962, and two deciduous incisors and 4 deciduous molars were discovered between 1968 and 1971. M.A. de Lumley (de Lumley 1976) reports that two of the deciduous molars come from a juvenile around 9 years old, whereas two incisors come from a juvenile around 5 years old. Examination of one lower left deciduous 2nd molar (specimen Homo 2) shows that the crown is more mesio-distally elongated and generally bigger than in modern human and Neanderthal juveniles. M.-A. de Lumley claims that this specimen shares many characteristics with the lower left deciduous 2nd molars from la Caune de l’Arago, such as a more mesio-distally elongated crown, a large and deep anterior fovea, and a developed cervical tuberosity on the buccal side. Based on these comparisons, M.-A. de Lumley attributed the hominid remains from Orgnac 3 to ‘pre-neanderthal’ (de Lumley 1976). No information is given about the two remaining molars, and no photographs or drawings of any of the specimens have ever been provided, except a photograph of one of the deciduous molars in situ during excavation in 1968 (fig. S3; Combier 1968, 1971).

4. Paleontology, Zooarchaeology, and Biochronology of Orgnac 3

4.1 Paleontology

44Approximately 10,000 identifiable faunal remains were uncovered at Orgnac 3; ~7,000 large mammal specimens are listed by level by H. Aouraghe (Aouraghe 1992). Different categories of the fauna have been analyzed by different researchers at different times (table S2). The carnivores, equids, cervids, and bovids were described by H. Aouraghe (Aouraghe 1990, 1992, 1999). Morphological description, the number of identified specimens, the minimum number of individuals, and sometimes measurements for skeletal elements were reported (table S3).

45The equids were further analyzed by A. Forsten and A.-M. Moigne (Forsten & Moigne 1998). This work provides an age profile for the equid assemblage based on dental eruption and wear, and grapples with the species determination between steinheimensis and mosbachensis. They divide age into 7 general categories and report the number of individuals in each age category by stratigraphic level (table S4; note that the MNIs in this analysis differ from those of H. Aouraghe [Aouraghe 1992]). Based on morphological comparisons of the teeth and metapodials, they attributed the Orgnac horse to E. steinheimensis, though H. Aouraghe attributed them to E. mosbachensis. The Rhinocerotidae material was examined by C. Guérin (Guérin 1980). The Macaca material was examined by A.-M. Tillier and B. Vandermeersch (Tillier & Vandermeersch 1976). Lagomorphs were identified and analyzed by K. El Guennouni (El Guennouni 2001). Micromammals from the site were systematically identified and analyzed by M. Jeannet (Jeannet 1974, 1981, 2000) and further by N. El Hazzazi (El Hazzazi 1991, 1998a, 1998b) and A.-C. Paunesco (Paunesco 2007).

46Birds from the site were systematically identified and analyzed by C. Mourer-Chauviré (Mourer-Chauviré 1975a, b). The carnivores were initially identified by R. Ballesio (unpublished), and the felidae and hyaenidae were further studied by A. Testu (Testu 2006). The Ursus thibetanus and Hemitragus materials were revised by E. Crégut-Bonnoure (Crégut-Bonnoure 1996, 2002).

4.2 Zooarchaeology

47The focus of H. Aouraghe’s (Aouraghe 1992) work was to identify the large mammal remains to element, side, and species, and to use these to inform paleoecological and biostratigraphic reconstructions for the site. Only 13 pages (out of almost 500) were dedicated to any type of zooarchaeological analysis, which was admittedly preliminary in scope. Skeletal part representations for the equids, cervids and bovids were provided (these values do not include shaft fragments). Specimens were aged to young, adult, and old based on levels of dental wear and epiphyseal fusion, but these are only reported colloquially throughout the text and no analysis of mortality profiles was performed. The presence of cut marks and carnivore and rodent tooth marks on the faunal remains are qualitatively described throughout the text and provides 5 photographs thereof (3, 1, and 1, respectively) are provided (Aouraghe 1992). While this work represents an important first step, it does not contain the analyses required to support the zooarchaeological conclusions that were presented: that hominids were the main, almost sole accumulator of the materials in the site, and that they occupied the site for a very long time.

48Further discussion on zooarchaeological aspects of the Orgnac 3 fauna were presented by A.-M. Moigne and D.R. Barsky (Moigne & Barsky 1999). Only two paragraphs are spent, most of which is citation of H. Aouraghe’s work, many of which are contradictory. For example, it is stated that “it is important to note that burnt bones are very numerous in all archaeological levels, especially in Level 6” (Moigne & Barsky 1999: 229), whereas H. Aouraghe stated “however, the quantity of burnt remains is slightly more important in the upper levels: Level 2 is very rich in burnt bones, particularly in bone splinters” (Aouraghe 1992: 433). It is stated that bone preservation is excellent, with teeth making up less than 50% of the material, and that “all skeletal elements are well represented” (Aouraghe 1992: 228). This directly contradicts H. Aourgahe’s skeletal part representation data (fig. S4). The conclusion that “at Orgnac 3, prehistoric hunters, successively occupying the changing floors in the shelter, brought complete animal carcasses back to the site and processed them there” (Aouraghe 1992: 229) is not demonstrated by the data presented.

49Y. Sam’s (Sam 2009) zooarchaeological study did not include identifiable elements from over 30 drawers of material (from which we identified ~1400 additional elements, see Supplement for details), nor any mention of this material or their taphonomic implications. The faunal data used were derived from the CERPT online inventory database, which contains 15 and 20% mis-identified elements, either to element, side, species, or a combination thereof (see Supplement for details). The majority of the zooarchaeological calculations in the analyses were done on faunal remains from the entire site, not for each individual level. For example, his figure 95 (fig. S5) shows anthropogenic damage locations on red deer [page 200], or calculations of skeletal element profiles [pages 125-137], etc.). This means that, according to the chronology used by Y. Sam (which places the site’s layers between 300 - 360 ka), 60,000 years of faunal accumulation by multiple factors are being time-averaging into one single analytical unit.

50Interestingly, K. El Guennouni (El Guennouni 2001) mentions two lagomorph metatarsals with cutmarks on their plantar surface indicative of skinning (El Guennouni 2001: 200).

51There are over 30 drawers of material from Orgnac 3 in the CERPT in Tautavel for which square and level provenience are known but had never been studied because they lacked specific 3-D XYZ coordinates. As an example of what these drawers contain, figure S6 shows the contents of bag chosen at random and sorted, which included: natural rocks, lithic pieces, identifiable faunal specimens, faunal specimens with surface modifications (e.g. carnivore tooth marks, cut marks), burnt faunal specimens, faunal specimens that were too covered with concretions to be analyzed, etc. The lithic and faunal samples used in the spatial analyses presented by M.-H. Moncel (Moncel 1996) and M.-H. Moncel et al. (Moncel et al. 2005) are therefore potentially biased to include only the specimens large/important/identifiable enough to have been given specific 3-D XYZ coordinates during excavation 50 years ago (see OSM for further discussion on these spatial analyses).

52Certain aspects of the aforementioned work suggest that carnivores may have been involved in accumulating bones in the lower levels of the site. In addition to determining that carnivores in general were involved in accumulating bones at a site, it is often possible to investigate which genera/species of carnivore in particular was involved. For example, K. Cruz-Uribe (Cruz-Uribe 1991) identified and described six indicators that a bone assemblage was accumulated by hyenas:

  1. Carnivore/ungulate ratio. In hyena accumulations the ratio to carnivore to carnivore + ungulate remains is at least 20%; in hominid accumulated assemblages it is usually less than 10% and always less than 13%.
  2. Damage to bone surfaces. Distinctive hyena damage includes striations, pitting, grooves, scooping, and acid-etching. Bone surface preservation in fossil assemblages may not be so good, therefore low percentages of carnivore damaged bones do not necessarily exclude hyenas as a bone accumulator.
  3. Bone breakage. Hyena accumulations are generally characterized by ‘bone cylinders’ in which the epiphyses have been chewed off, whereas hominid accumulations tend to have many broken shafts and intact epiphyses. It has also been generalized that carnivore accumulations are less fragmented than human accumulations. Post-depositional bone fragmentation and destruction must always be accounted for.
  4. Cranial/postcranial ratio. In hyena accumulations, the cranial/postcranial ratio tends to increase with ungulate size. Large ungulate postcranial bones tend to come from adults, whereas the large ungulate dentitions tend to come from juveniles. No such patterning is visible in hominid accumulations.
  5. Representation of small, hard bones. These tend to be rare in hyena accumulations, regardless of the preservation level of bone in the site. Conversely, small hard bones tend to be common in hominid accumulations, and superabundant in very fragmented samples.
  6. Age profiles. Age, or mortality, profiles, tend to be attritional in hyena accumulations, whereas they can be either attritional or catastrophic in hominid accumulations, depending on how the animals were obtained.

53The skeletal remains from the lower levels of Orgnac 3 are very interesting in this regard. According to figure S4, in Level 7-8, carnivores make up ~40% of the faunal remains (compare this to figure S7, in which this number is ~20%), and spotted hyenas make up ~27% of the carnivores in Level 7-8 (both of these values are the highest of all levels) and 19.3% of carnivores in the site overall (fig. S8). No indications of hyena stomach acid etching on the bones were previously reported, but also that the bone surfaces are often hard to see because of sediment concretions, which is inconclusive. To address the level of bone fragmentation, for species where NISP and MNI values were reported by H. Aouraghe (Aouraghe1992) by level, the NISP was divided by the MNI to calculate a ‘fragmentation index’. For Ursus deningeri, Equus caballus, Dicerorhinus hemitoechus, Sus scrofa, and Cervus elaphus, Level 7-8 is not the most highly fragmented level (usually in the 50 – 80 percentile), whereas for Canis lupus, it is the highest (table S3). For the equids in the site, A. Forsten and A.-M. Moigne (Forsten & Moigne 1998) provide a mortality profile for Level 7 that contains 1 infant, 1 juvenile, 1 sub-adult, and 1 very aged individual (table S4). This is the most attritional equid mortality profile of any level in their report. It is not possible to determine the cranial/postcranial ratio, small-hard bone representation, or age profiles for other taxa, as the necessary data has not previously been reported for Orgnac 3. We know hyenas were present in the site, that in Pleistocene Europe spotted hyenas were important bone accumulators (e.g. Campbell 1977, Stuart 1982, Sutcliffe 1985, Gamble 1986, Fourvel 2012), and that their assemblages are reported to contain a large number of carnivores, including members of their own species (Stuart 1982, Gamble 1986). The three indicators of hyena accumulation for which data was available in the 1990’s suggest that hyenas may have been involved in accumulating the faunal remains in Levels 7-8.

8. Abundances of Emys and Testudo carapace remains in Orgnac 3’s levels.

8. Abundances of Emys and Testudo carapace remains in Orgnac 3’s levels.

4.3 Biochronology

54The first to attempt a biochronological date for the site was C. Guérin (Guérin 1980: 615-618). He placed geological layers from Orgnac 3 into his mammal chronological ‘Zones’ as follows:

  • layers o through k into his Zone 22, equivalent with Mindel, Elster, Lovestoft (generally correlated to OIS  12, 478–424 ka).
  • layer i into his Zone 23, equivalent with Mindel-Riss, Holsteinian, Needian, Hoxnian (generally correlated to the early part of OIS 11, 424 ~ 390 ka).
  • layer g into his Zone 24, equivalent with Riss, Saale, Gipping (generally correlated to the early part of OIS 10, 364 ~ 340 ka).

55According to H. Aouraghe (Aouraghe 1992), the biochronological aspects of the fauna from the site, such as the presence of Canis lupus lunellensis, Apodemus sylvaticus, Microtus brecciensis, Eliomys quercinus, and an archaic form of Crocuta spelaea indicate that the lowest archaeological levels derive from the end of the early Middle Pleistocene, whereas Level 2 can be attributed to just before the late Middle Pleistocene based on the presence of Rupicapra rupicapra, Hemitragus bonali, Ursus deningeri, and Equus caballus mosbachensis. This is generally congruent with the radiometrically-derived dates, but spans a wider time range. H. Aouraghe concluded that the depositional history suggested by the fauna would “require a longer amount of time than has been proposed [by the radiometric dates]” (Aouraghe 1992: 462). A. Testu (Testu 2006) described aspects of the Felidae that also suggest a mismatch between the radiometric dates and the biochronology: “The cat [Felis silvestris] from Orgnac 3, although more recent, preserves some archaic characters, in particular the shortening of the protoconid on the P4, which recalls more ancient forms” (Testu 2006: 283).

56Based on the microfauna, M. Jeannet (Jeannet 1974, 1981: 65), concluded:

All of this encourages us to think that Orgnac 3 is situated in the transitional Mindel-Riss. I [Jeannet] was tempted to correlate the site with some other sites that I was able to work on in the following manner:

Orgnac 3 (g, h) - phase IV

= La Fage (en partie)

= Riss I

Orgnac 3 (i, j) - phase III

= Lunel-Viel 1, Coudoulous 1 niveau 8

= Mindel-Riss

Orgnac 3 (k, l, m) - phase II

= Saint-Estève-Janson H

= Mindel final

Orgnac 3 (n, o, p) - phase I

= Montoussé 3

= Mindel superior

Due to the presence of some particular species, the chronostratigraphic relations can be established between Orgnac 3 and the other sites where they can be collected.

57This conclusion was mirrored by J. Chaline (Chaline 1975: 114): “A sure stratigraphy of these periods will be possible when we have at our disposal very long sequences that overlap in time, like at Saint-Estève-Janson, Orgnac 3, and Aldène.” She showed her proposed correlation between those sequences, and others, in her figure 36 (fig. S9), which matches M. Jeannet’s chronological placement of the Orgnac 3 layers. Based on this same microfauna, N. El Hazzazi (El Hazzazi 1998a: 129) concluded:

This study allows us to sketch a regional chronostratigraphy based on the evolution of the genus Arvicola in this part of France. Thus, in the site of Orgnac 3, there is the presence of a homogeneous archaic population, as indicated by their morphology as well as the degree of differentiation of the enamel bands; this population alongside Pliomys episcopalis and Pliomys chalinei (Jeannet 1974), and Erinaceus davidi (El Hazzazi 1991), belongs to the species Arvicola cantiana. The SDQp indices obtained for this site allow us to give a Holsteinien [OIS 11] age to levels 6 and 7 and a Saalien [OIS 10] age to levels 4 and 5.

58In A.-C. Paunesco’s (Paunesco 2007) more recent and detailed work on the microfauna, most morphological and biometric analyses indicated a very strong correspondence between the community at Orgnac 3 and those of La Caune de l’Arago Ensembles  I to III (dated to 690 - 400 ka, OISs 14 - 9; Falguères et al. 2015) and Visogliano (dated to 500 - 350 ka, OISs 13 - 10; Falguères et al. 2008), and appears slightly more archaic than Ensembles II-III of Baume Bonne (OISs 10 - 8; Gagnepain & Gaillard 2005).

5. Chronological Issues with the Orgnac 3 Sequence

59As described above and detailed in table 2, there is a mismatch between the different radiometric dates for the Orgnac 3 sequence and between these mismatched measurements and the biochronological timeline of the sequence. S. Khatib (Khatib 1989: 181), described sedimentological indicators of the site’s chronology and how they do not match with the radiometric dates:

The two absolute dates made on the site allow us to situate the sequence in the isotopic stratigraphy of the Quaternary.

- Ensemble II dated to -339,000 years on average corresponds to the lower limit of OIS 10, but the sedimentological and geochemical data indicate that it represents a climatic warming period, which allows us to situate Ensemble II in OIS 11. This supposes that the date is too recent to be attributed to this stage.

- Ensemble IV dated to -298,000 ± 55,000 years, corresponds to OIS 9 or the equivalent of a climatic warming period. The sedimentological data is going in the direction of a climatic cooling period; it is very likely that here too, the date is lightly underestimated. This ensemble could therefore still be in OIS 10. The superficial alteration of this ensemble (level IVb) corresponds to warm stage 9.

- According to the isotopic stratigraphy of the Quaternary, we can situate Ensemble I, for which the sedimentological analyses indicate a very cold and dry climate, in OIS 12.

The same goes for Ensemble III which can be situated in cold OIS 10, as the sedimentological and geochemical data correspond to a cold and humid climate.

60As seen in table 2, in much of the dating work the younger dates come from further down in the sequence. This is the opposite of what would be expected from a closed system environment, and what would be expected if uranium was mobilized and reprecipitated further down the stratigraphy and/or the inverse for thorium. The poor precision, accuracy, and/or internal consistency of the radiometric dates argue for a longer period of deposition of the Orgnac 3 sediments than mainly occurring in OISs 9 and 8, as proposed in most publications (e.g. Moncel et al. 2005, 2011, 2012, Michel et al. 2011, 2013, Bahain et al. 2022 and references therein). Based on the available data, two distinct chronological reconstructions can be made for the Orgnac 3 deposits (fig. 7):

  • Age Model 1, based exclusively on published radiometric dates, would suppose that the non-archaeological layers below Level 8 are from OIS 10, and that the archaeological layers are from OISs 9 and 8.
  • Age Model 2, based on the geological and paleontological data, would suppose that the non-archaeological layers below Level 8 are from OIS 12, and that the lower half of the archaeological layers are from OIS 11, the upper half are from OIS 10, and that Level 1 is from the beginning of OIS 8, while OIS 9 is mainly represented by Levels 3, 2, and the erosional non-conformity between Levels 2 and 1. During OIS 9, the OIS 10 deposits were post-depositionally altered and the circulation of carbonate-charged water formed the dense concretions covering the bones and lithics from those levels.

61In order for Age Model 1 to be accepted, one must disregard nearly all of the conclusions stated in the analyses of the site’s geology (e.g. Debard 1988, Khatib 1989, Lafaar 1995) and faunal assemblages (especially the microfauna; El Hazzazi 1998a, Paunesco 2007), and allow for the volcanic minerals in layer d1 to have been blowing and washing around on the surface for over 50 thousand years before ending up there, but still look “sharp-edged and unweathered”. Meanwhile, the reasons to reject this age model have already been given by many previous researchers: after the collapse of the karstic cavity and the opening of the layers to the surface, rain and ground water movement through the layers caused remobilization of uranium and/or thorium in the sequence (an open system), causing any U-series or ESR dates to be underestimated (e.g. Claes et al. 2020).

62If Age Model 2 is accepted, instead of archaeological levels 8 - 2 (corresponding with geological layers k - d) dating from 350 - 250 ka (centering around 300 ka), they would instead date from OIS 11 & 10, 427 - 334 ka. The non-archaeological levels below, geological layers l - t, would therefore date to OIS 12, 474 - 427 ka. Level 1, from geological layer c, is separated from Level 2 by an erosional event and the clinopyroxene and its associated date of ~300 ka is situated on top of the altered and ‘dug into’ layer d1; the ~300 ka date is therefore a maximum age of Level 1, but these capping sediments, fossils and artifacts could in fact be much younger. This revised chronology is substantially older and of a longer duration than the radiometric dates suggested and appear to be more in line with the depth of deposition and both macro- and micro-faunal changes recorded in the site. Additionally, figure 8 shows the abundance of two reptiles in the sequence, the European pond turtle (Emys orbicularis) and a terrestrial tortoise Testudo sp. (Lewis 2011). Both are markers of warm and wet environments, especially the latter, which is characteristic of interglacials in the European Middle Pleistocene record (Morales Pérez & Sanchis Serra 2009, Sommer et al. 2007). Their abundances match what would be expected if there were a transition from interglacial, to glacial, and then another interglacial at the top of the sequence.

63Overall, the history of work presented here suggest that Age Model 2 is more likely, with the non-archaeological layers below Level 8 being from OIS 12, and that the lower half of the archaeological layers are from OIS 11, the upper half are from OIS 10, and that Level 1 is from OIS 8, while OIS 9 is represented by an erosional non-conformity between Levels 2 and 1 and an increase in warm signals (like red deer, turtles, carnivores, etc.) in Levels 2 and 3 due to mixing. The sedimentological signals of cold periods are reflected in changes in species abundances throughout the sequence, arguing against the deposits having been laid down in one single OIS stage. With the 40Ar/39Ar date of 302 ± 2.9 ka (a date right at the transition between OIS 9 and 8) between Levels 1 and 2, it would be questionable to place all of Levels 3 though 8 in the warm OIS 9 period. Such a model would require the 7 m of stratigraphy at Orgnac 3 to have been deposited in less than 100 ka, which is rarely seen in major Mediterranean Middle Pleistocene sites, the vast majority of which required several hundred thousand years to form such depth (e.g. Gran Dolina, Arago, Vallparadís, Visogliano, Petralona, etc.; Cook et al. 1982, Carbonell Roura & Rodriguez Alvarez 1994, Grün 1996, Falguères et al. 2008, 2015, Martínez et al. 2013). Most of the above-cited sedimentological and dating researchers describe how post-depositional geochemical alteration has made Uranium-series and most ESR analyses dubious. Meanwhile, the fission track and ESR dates on burnt quartz match up with what would be predicted in Age Model 2. Previous authors’ reconstructions of the lower levels dating from OIS 9 and the upper levels from OIS 8 (e.g. Moncel et al. 2005, 2011, 2012, Michel et al. 2011, 2013, Bahain et al. 2022 and references therein) therefore are not supported by the majority of the available evidence. The history of the KBS Tuff Controversy from the Pleistocene of Koobi Fora, Kenya (Lewin 1997) serves as a cautionary tale of what happens when questionable radiometric dates are prioritized over the weight of stratigraphic and biochronological evidence. If Age Model 2 is shown over time to be correct, Orgnac 3 would record one of the oldest and longest paleoanthropological sequences in Western Europe.

6. Conclusion

64Orgnac 3 is a key site in the quest to understand the paleoecological, behavioral, and morphological changes that were taking place during the Middle Pleistocene that included the separation of the modern human, Neanderthal, and Denisovan lineages. However, a review of the previous work performed on the site and the conclusions drawn from those studies show that Orgnac 3 remains misunderstood. The Orgnac 3 deposits most likely date to OISs 12, 11, 10, and some of 9 and 8, not only OIS 9 and 8 (making the site ~100 ka older than proposed by other authors based solely on U-series and ESR dates from post-depositionally altered samples). Meanwhile, the dating of the upper levels has important implications for understanding the arrival of the Levallois method into the existing Acheulean and the transition to the Middle Paleolithic with the Mousterian culture of the Rhône Valley, as the Orgnac 3 sequence has been used as a key argument on the timing of Levallois emergence in Europe overall. Level 1 can only confidently be assigned a maximum date of 300 ka due to the erosional event between it and Level 2, and there is no way of providing a minimum date. The fossil fauna most likely represents a combination of both carnivore and hominin-acquired carcasses, and not only hominin hunting and butchering as proposed by other authors. The hominin remains have not been described in detail or assigned to a taxon. Orgnac 3 has the potential to be one of the oldest, largest, richest, and most important Middle Pleistocene sites in Europe. Building upon the pioneering work of the 1960’s and research since then, in-depth synthetic analysis of Orgnac 3’s geological, archaeological and fossil material, and possible future excavations, have the potential and help answer the important questions about hominin morphological and behavioral evolution during the Middle Pleistocene in Europe.

Haut de page


Abi-Rached et al. 2011, ABI-RACHED L., JOBIN M., KULKARNI S., MCWHINNIE A., DALVA K., GRAGERT L., BABRZADEH F., GHARIZADEH B., LUO M., PLUMMER F.A., KIMANI J., CARRINGTON M., MIDDLETON D., RAJALINGAM R., BEKSAC M., MARSH STEVEN G.E., MAIERS M., GUETHLEIN LISBETH A., TAVOULARIS S., LITTLE A.M., GREEN R.E., NORMAN P.J., PARHAM P., The Shaping of Modern Human Immune Systems by Multiregional Admixture with Archaic Humans, Science, 2011, 334, 6052, p. 89-94.

Aouraghe 1990, AOURAGHE H., Les Cervidés du site Pléistocène Moyen d'Orgnac 3 (Ardèche, France), Quaternaire, 1990, 3-4, p. 231-245.

Aouraghe 1992, AOURAGHE H., Les Faunes de grands mammifères du site Pléistocène Moyen d'Orgnac 3 (Ardèche, France), Paris, Museum national d'histoire naturelle - Institut de Paléontologie humaine, 1992, Thèse de Doctorat : Quaternaire, paléontologie humaine, préhistoire, 495 p.

Aouraghe 1999, AOURAGHE H., Reconstitution du paléoenvironnement par les grands mammifères : les faunes du Pléistocène moyen d'Orgnac 3 (Ardèche, France), L'Anthropologie (Paris), 1999, 103, 1, p. 177-184.

Arsuaga Ferreras et al. 1993, ARSUAGA FERRERAS J.L., MARTÍNEZ MENDIZÁBAL I., GRACIA TÉLLEZ A., CARRETERO DÍAZ J.M., CARBONELL ROURA E., Three new human skulls from the Sima de los Huesos Middle Pleistocene site in Sierra de Atapuerca, Spain, Nature, 1993, 362, p. 534-537.

Arsuaga Ferreras et al. 1997, ARSUAGA FERRERAS J.L., MARTÍNEZ MENDIZÁBAL I., GRACIA TÉLLEZ A., LORENZO C., The Sima de los Huesos crania (Sierra de Atapuerca, Spain). A comparative study, Journal of human Evolution, 1997, 33, 2-3, p. 219-281.


Bahain et al. 2022, BAHAIN J.-J., MERCIER N., VALLADAS H., FALGUÈRES C., MASAOUDI H., JORON J.-L., FROGET L., MOIGNE A.-M., COMBIER J., MONCEL M.-H., Reappraisal of the chronology of Orgnac 3 Lower-to-Middle Paleolithic site (Ardèche, France), a regional key sequence for the Middle Pleistocene of southern France, Journal of Human Evolution, 2022, 162, p. 103092.

Campbell 1977, CAMPBELL J.B., The Upper Palaeolithic of Britain: a study of man and nature in the late Ice Age, Oxford, Clarendon Press, 1977, 264 + 376 p.

Carbonell Roura & Rodríguez Alvarez 1994, CARBONELL ROURA E., RODRÍGUEZ ALVAREZ X.P., Early Middle Pleistocene Deposits and Artifacts in the Gran-Dolina Site (Td4) of the Sierra de Atapuerca (Burgos, Spain), Journal of human Evolution, 1994, é-, p. 291-311.

Chaline 1975, CHALINE J., Les Rongeurs, l'âge et la chronologie climatique du remplissage de l'Aven I de la Fage (Corrèze), Nouvelles Archives du Museum d'Histoire naturelle de Lyon, 1975, 13, p. 113-117.

Claes et al. 2020, CLAES H., TÖRÖK Á., SOETE J., MOHAMMADI Z., VASSILIEVA E., HAMAEKERS H., ERTHAL M.M., ARATMAN C., CHENG H., EDWARDS R.L., SHEN C., ÖZKUL M., KELE S., MINDSZENTY A., SWENNEN R., U/Th dating and open system behavior: implications for travertines based on the study of Süttő (Hungary) and Ballik (Turkey) sites, Quaternaire, 2020, 31, 2, p. 117-132.

Combier 1963, COMBIER J., Découverte d'un gisement acheuléen dans le bassin du Rhône, Comptes rendus hebdomadaires des séances de l'Académie des sciences, 1963, 257, p. 201-203, 1 fig.

Combier 1967, COMBIER J., Le Paléolithique de l'Ardèche dans son cadre paléoclimatique, Bordeaux, Imprimerie Delmas, 1967, 463 p. (Mémoire de l'Institut de Préhistoire de l'Université de Bordeaux ; 4).

Combier 1968, COMBIER J., Compte-rendu des recherches préhistoriques effectuées en 1968 dans la circonscription Rhône-Alpes, Compte-rendu d'activités annuelles - Association régionale pour le développement des recherches de paléontologie et de préhistoire et des amis du Muséum de Lyon, 1968, p. 28-38.

Combier 1971, COMBIER J., Le gisement prémoustérien et acheuléen d’Orgnac, Études préhistoriques, 1971, 1, p. 24-26.

Combier 1979, COMBIER J., Unpublished Excavation Report, s.l., s.n., 1979, n. p.

Combier 1989a, COMBIER J., La "Garrigue" du Plateau d'Orgnac (Ardèche), Terre Vive, 1989a, 76, p. 14-22.

Combier 1989b, COMBIER J., Autour d'Orgnac... La Préhistoire, Terre Vive, 1989b, 76, p. 23-26.

Combier 1996, COMBIER J., Mattecarlinque ("Orgnac 3"), Gallia Informations, 1996, p. 46-48.

Cook et al. 1982, COOK J., STRINGER C.B., CURRANT A.P., SCHWARCZ H.P., WINTLE A.G., A Review of the Chronology of the European Middle Pleistocene Hominid Record, Yearbook of Physical Anthropology, 1982, 25, p. 19-65.

Crégut-Bonnoure 1996, CRÉGUT-BONNOURE É., A review of small Middle Pleistocene bears from France, Acta Zoologica Cracoviensia, 1996, 39, 1, p. 89-101.

Crégut-Bonnoure 2002, CRÉGUT-BONNOURE É., Les Ovibovini, Caprini et Ovini (Mammalia, Artiodactyla, Bovidae, Caprinae) du Plio-Pléistocène d’Europe : systématique, évolution et biochronologie, Université de Lyon I, 2002, Thèse de Doctorat d'État : Sciences, 2 vol., 429 p., 686 fig., 33 pl., 212 tabl.

Cruz-Uribe 1991, CRUZ-URIBE K., Distinguishing Hyena from Hominid Bone Accumulations, Journal of Field Archaeology, 1991, 18, p. 467-486.

de Lumley 1976, DE LUMLEY M.-A., Les Anténéandertaliens dans le sud, in La Préhistoire française. Tome 1 : Les Civilisations paléolithiques et mésolithiques de la France, Lumley (de) H. (dir.), Paris, Éditions du CNRS, 1976, p. 546-560.

Debard 1988, DEBARD É., Le Quaternaire du Bas-Vivarais d'après l'étude des remplissages d'avens, de porches de grottes et d'abris sous roche : dynamique sédimentaire, paléoclimatologie et chronologie, Lyon, Université Claude Bernard - Lyon I, 1988, 315 p. (Documents des Laboratoires de Géologie Lyon ; 103).

Debard & Pastre 1988, DEBARD É., PASTRE J.-F., Un marqueur chronostratigraphique du Pléistocène moyen à la périphérie du Massif Central : la retombée à clinopyroxène vert du Sancy dans le site acheuléen d'Orgnac III (Bas-Vivarais, SE France), Comptes rendus de l'Académie des sciences (2), 1988, 306, p. 1515-1520.

Delannoy et al. 2009, DELANNOY J.-J., GAUCHON C., HOBLÉA F., JAILLET S., MAIRE R., PERRETTE Y., PERROUX A.-S., PLOYON E., VANARA N., Karst: from palaeogeographic archives to environmental indicators, Géomorphologie. Relief, processus, environnement, 2009, 2, p. 83-94.

Dibble 1985, DIBBLE H.L., Raw-Material Variation in Levallois Flake Manufacture, Current Anthropology, 1985, 26, p. 391-393.

Djerrab 2001, DJERRAB A., Étude des propriétés magnétiques du remplissage de sites préhistoriques pléistocènes : contribution à l'étude stratigraphique, archéozoologique de leurs dépôts et à leur évolution géochimique post-dépositionnelle, Université de Perpignan, 2001, Thèse de doctorat en Géologie du Quaternaire, 439 p.

El Guennouni 2001, EL GUENNOUNI K., Les lapins du Pléistocène moyen et supérieur de quelques sites préhistoriques de l'Europe méditerranéenne : Terra-Amata, Orgnac 3, Baume Bonne, grotte du Lazaret, grotte du Boquete de Zafarraya, Arma delle Manie : étude paléontologique, taphonomique et archéozoologique, Paris, Muséum national d'Histoire naturelle, 2001, Thèse de doctorat en Sciences naturelles, 403 p.

El Hazzazi 1991, EL HAZZAZI N., Contribution à l'étude de la microfaune d'Orgnac 3 : données paléontologiques, paléoclimatiques et palethnographiques, Museum national d'histoire naturelle, 1991, Diplôme d'Études approfondies : Quaternaire, géologie, paléontologie humaine, préhistoire, 99 p.

El Hazzazi 1998a, EL HAZZAZI N., Apport du genre Arvicola (Rodentia, Mammalia) dans la connaissance de la chronostratigraphie des sites du Pléistocène moyen et supérieur : Orgnac 3, Payre 2 et l'Abri des Pêcheurs (Ardèche, France), in Theoretical and methodological problems - Archaeometry - Paleoecology, Arias C., Bietti A., Castelletti L. et al. (dir.), Forlí, A.B.A.C.O., 1998a, p. 125-133 (Actes du 13ème Congrès de l'Union internationale des sciences préhistoriques et protohistoriques, Forlí 1996).

El Hazzazi 1998b, EL HAZZAZI N., Paléoenvironnement et chronologie des sites du pléistocène moyen et supérieur : Orgnac 3, Payre et abri des pêcheurs (Ardèche, France) d'après l'étude des rongeurs, Paris, Muséum national d'Histoire naturelle, 1998b, Thèse de doctorat en Géologie, paléontologie humaine et préhistoire, 246 p.

Falguères 1986, FALGUÈRES C., Datations de sites acheuléens et moustériens du Midi méditerranéen par la méthode résonance de spin électronique (ESR), Museum national d'histoire naturelle, 1986, Thèse de Doctorat, 173 p.

Falguères et al. 1988, FALGUÈRES C., SHEN G., YOKOYAMA Y., Datation de l'Aven d'Orgnac III : Comparaison par les méthodes de la résonance de spin électronique (ESR) et du déséquilibre des familles de l'uranium, L'Anthropologie (Paris), 1988, 92, 2, p. 727-730.

Falguères et al. 2008, FALGUÈRES C., BAHAIN J.-J., TOZZI C., BOSCHIAN G., DOLO J.-M., MERCIER N., VALLADAS H., YOKOYAMA Y., ESR/U-series chronology of the Lower Palaeolithic palaeoanthropological site of Visogliano, Trieste, Italy, Quaternary Geochronology, 2008, 3, 4, p. 390-398.

Falguères et al. 2015, FALGUÈRES C., SHAO Q., HAN F., BAHAIN J.-J., RICHARD M., PERRENOUD C., MOIGNE A.-M., DE LUMLEY H., New ESR and U-series dating at Caune de l'Arago, France: A key-site for European Middle Pleistocene, Quaternary Geochronology, 2015, 30B "LED14 Proceedings / Rainer Grün & Frank Preusser", p. 547-553.

Forsten & Moigne 1998, FORSTEN A., MOIGNE A.-M., The horse from the middle Pleistocene of Orgnac-3 (Ardèche, France), Quaternaire, 1998, 9, 4, p. 315-322.

Fourvel 2012, FOURVEL J.-B., Hyénidés modernes et fossiles d'Europe et d'Afrique : taphonomie comparée de leurs assemblages osseux, Université Toulouse le Mirail – Toulouse II, 2012, Thèse de Doctorat Archéologie et Préhistoire, 597 p.

Fu et al. 2014, FU Q., LI H., MOORJANI P., JAY F., SLEPCHENKO S.M., BONDAREV A.A., JOHNSON P.L.F., AXIMU-PETRI A., PRUFER K., DE FILIPPO C., MEYER M., ZWYNS N., SALAZAR GARCÍA D.C., KUZMIN Y.V., KEATES S.G., KOSINTSEV P.A., RAZHEV D.I., RICHARDS M.P., PERISTOV N.V., LACHMANN M., DOUKA K., HIGHAM T.F.G., SLATKIN M., HUBLIN J.-J., REICH D., KELSO J., VIOLA B., PAABO S., Genome sequence of a 45,000-year-old modern human from western Siberia, Nature, 2014, 514, 7523, p. 445-449.

Gagnepain & Gaillard 2005, GAGNEPAIN J., GAILLARD C., La grotte de la Baume Bonne (Quinson, Alpes-de-Haute-Provence) : synthèse chronostratigraphique et séquence culturelle d'après les fouilles récentes (1988-1997), in Les premiers peuplements en Europe : colloque international [...], Rennes, 22-25 septembre 2003, Molines N., Moncel M.-H., Monnier J.-L. (dir.), Oxford, John and Erica Hedges Ltd, 2005, p. 73-85 (British archaeological Reports - International Series; 1364).

Gamble 1986, GAMBLE C.S., The Palaeolithic settlement of Europe, Cambridge, University Press, 1986, 471 p., couv. ill.

Gandin & Capezzuoli 2008, GANDIN A., CAPEZZUOLI E., Travertine Versus Calcareous Tufa: Distinctive Petrologic Features and Stable Isotopes Signatures, Il Quaternario, 2008, 21, 1B, p. 125-136.

Grün 1996, GRÜN R., A re-analysis of electron spin resonance dating results associated with the Petralona hominid, Journal of human Evolution, 1996, 30, 3, p. 227-242.

Guérin 1980, GUÉRIN C., Les rhinocéros (Mammalia, Perissodactyla) du Miocène terminal au Pléistocène supérieur en Europe occidentale : comparaison avec les espèces actuelles, Lyon, Université Claude Bernard - Lyon I, 1980, 1188 p. (Documents des Laboratoires de Géologie Lyon ; 79).


Hublin 2007, HUBLIN J.-J., What Can Neanderthals Tell Us About Modern Origins?, in Rethinking the Human Revolution, Mellars P.A., Boyle K.V., Bar-Yosef O. et al. (dir.), Cambridge, McDonald Institute for Archaeological Research, 2007, p. 235-249.

Hublin 2009, HUBLIN J.-J., The origin of Neandertals, Proceedings of the National Academy of Sciences of the United States of America, 2009, 106, 38, p. 16022-16027.

Isaac 1975, ISAAC G.L., Sorting out the muddle in the middle: an anthropologist's post-conference appraisal, in: After the australopithecines, Butzer K.W., Isaac G.L., Wenner-Gren foundation for anthropological research (dir.), The Hague, Mouton 1975, p. 875-888.

Jaillet et al. 2007, JAILLET S., DELANNOY J.-J., BERSIHAND J.-L., NOURY M., SADIER B., TOCINO S., L'aven d'Orgnac : un grand réseau paragénétique, étude spéléogénétique des grands volumes karstifiés, in L'aven d'Orgnac : valorisation touristique, apports scientifiques, Delannoy J.-J., Gauchon C., Jaillet S. (dir.), Le Bourget-du-Lac, UMR5204 Environnements, Dynamiques et Territoires de la Montagne (EDYTEM), 2007, p. 57-78 (Collection EDYTEM - Cahiers de Géographie ; 5).

Jeannet 1974, JEANNET M., Les Rongeurs d’Orgnac 3 (Ardèche), Université de Lyon, 1974, Diplôme d'Études Supérieures Spécialisées, 113 p.

Jeannet 1981, JEANNET M., Les rongeurs du gisement acheuléen d'Orgnac 3 (Ardèche) : essai de paléoécologie et de chronostratigraphie, Bulletin mensuel de la Société linnéenne de Lyon, 1981, 50, 2, p. 49-71.

Jeannet 2000, JEANNET M., Biométrie et écologie de Microtus brecciensis (Mammalia, Rodentia), Revue de Paléobiologie, 2000, 19, 2, p. 339-357.

Khatib 1989, KHATIB S., Le site d'Orgnac 3 (Ardèche, France) : étude sédimentologique et géochimique - cadre chronologique et évolution paléoclimatique, Paris, Museum national d'Histoire naturelle, 1989, Thèse de Doctorat du Muséum national d’Histoire naturelle : Quaternaire, 209 p., 45 fig. & tabl.

Khatib 1994, KHATIB S., Datation des cendres volcaniques et analyses géochimiques du remplissage d'Orgnac 3 (Ardèche, France), Quaternaire, 1994, 5, p. 13-22.

Klein 1987, KLEIN R.G., Reconstructing how early people exploited animals: Problems and prospects, in The evolution of human hunting, Nitecki M.H., Nitecki D.V. (dir.), New York, Plenum Press, 1987, p. 11-45.

Kolen 1999, KOLEN J., Hominids without homes: on the nature of Middle Palaeolithic settlement in Europe, in The Middle Palaeolithic Occupation of Europe, Roebroeks W., Gamble C.S. (dir.), Leiden, University, 1999, p. 139-176.

Laafar 1995, LAAFAR S., Étude de microfaciès sédimentaires, pédologiques et archéologiques aux remplissages de grottes en milieu méditerranéen (Orgnac III et Lazaret, France), Paris, Muséum national d'Histoire naturelle, 1995, Thèse de Doctorat : Terre, océan, espace, 193 p.

Laafar et al. 2008, LAAFAR S., ABDESSADOK S., DJERRAB A., Caractérisation des foyers préhistoriques : apport de la susceptibilité magnétique et de la micromorphologie des sols (Orgnac 3, Lazaret et abri Pataud, France), in Le Quaternaire marocain dans son contexte méditerranéen : actes de la quatrième rencontre des quaternaristes marocains (RQM4), Oujda, 15-17 novembre 2007. Vol. 2, paléontologie, taphonomie & préhistoire Aouraghe H., Haddoumi H., El Hammouti K. (dir.), Oujda, Université Mohamed 1er, 2008, p. 447-460.

Lebon et al. 2010, LEBON M., REICHE I., BAHAIN J.-J., CHADEFAUX C., MOIGNE A.-M., FRÖHLICH F., SÉMAH F., SCHWARCZ H.P., FALGUÈRES C., New parameters for the characterization of diagenetic alterations and heat-induced changes of fossil bone mineral using Fourier transform infrared spectrometry, Journal of Archaeological Science, 2010, 37, 9, p. 2265-2276.

Lewin 1997, LEWIN R., Bones of contention: controversies in the search for human origins 2nd edition, Chicago, University of Chicago Press, 1997, 366 p.

Lewis 2011, LEWIS J.E., Taphonomy And Zooarchaeology of The Middle Pleistocene Site of Orgnac 3 (Ardèche Valley, Southeastern France), Department of Anthropological Sciences - Stanford University, 2011, PhD Dissertation, 440 p.

MacDonald et al. 2021, MACDONALD K., SCHERJON F., VAN VEEN E., VAESEN K., ROEBROEKS W., Middle Pleistocene fire use: The first signal of widespread cultural diffusion in human evolution, Proceedings of the National Academy of Sciences of the United States of America, 2021, 118, 31, p. e2101108118.

Martínez et al. 2013, MARTÍNEZ K., GARCIA J., CARBONELL ROURA E., Hominin multiple occupations in the Early and Middle Pleistocene sequence of Vallparadís (Barcelona, Spain), Quaternary International, 2013, 316 "Middle to Upper Palaeolithic biological and cultural shift in Eurasia II", p. 115-122.

Martínez Aguirre et al. 2019, MARTÍNEZ AGUIRRE A., ALCARAZ PELEGRINA J.M., RODRÍGUEZ VIDAL J., U/Th dating of impure carbonates: 230Th/232Th activity ratios in detrital material, Journal of Radioanalytical and Nuclear Chemistry, 2019, 321, p. 71-81.

Masaoudi 1995, MASAOUDI H., Application des méthodes du déséquilibre des familles de l'uranium (230Th/234U) et de la résonance de spin électronique (ESR) à la datation des sites d'Orgnac 3, de Payre, et de l'Abri des Pêcheurs (Ardèche), Paris, Museum national d'histoire naturelle, 1995, Thèse de Doctorat : Quaternaire, 155 p.

Mathias et al. 2021, MATHIAS C., VIALLET C., MOIGNE A.-M., Diversité des outils et des activités de percussion à la fin du MIS 9 dans le niveau 6 d’Orgnac 3 (Ardèche, France), Comptes Rendus Palevol, 2021, 20, 11, p. 175-198.

Mellars et al. 2007, MELLARS P., BOYLE K.V., BAR-YOSEF O., STRINGER C.B., Rethinking the human revolution: new behavioural and biological perspectives on the origin and dispersal of modern humans, Cambridge, McDonald Institute for Archaeological Research - University of Cambridge, 2007, 436 p. (McDonald Institute Monographs).

Meyer et al. 2014, MEYER M., FU Q., AXIMU-PETRI A., GLOCKE I., NICKEL B., ARSUAGA FERRERAS J.L., MARTINEZ I., GRACIA TÉLLEZ A., BERMUDEZ DE CASTRO J.M., CARBONELL ROURA E., PAABO S., A mitochondrial genome sequence of a hominin from Sima de los Huesos, Nature, 2014, 505, 7483, p. 403-406.

Meyer et al. 2016, MEYER M., ARSUAGA FERRERAS J.L., DE FILIPPO C., NAGEL S., AXIMU-PETRI A., NICKEL B., MARTÍNEZ MENDIZÁBAL I., GRACIA TÉLLEZ A., BERMUDEZ DE CASTRO J.M., CARBONELL ROURA E., VIOLA B., KELSO J., PRÜFER K., PÄÄBO S., Nuclear DNA sequences from the Middle Pleistocene Sima de los Huesos hominins, Nature, 2016, 531, 7595, p. 504-507.

Michel et al. 2011, MICHEL V., SHEN G., SHEN C., FORNARI M., VÉRATI C., GALLET S., SABATIER D., Les derniers Homo heidelbergensis et leurs descendants les néandertaliens : datation des sites d’Orgnac 3, du Lazaret et de Zafarraya, Comptes Rendus Palevol, 2011, 10, 7, p. 577-587.

Michel et al. 2013, MICHEL V., SHEN G., SHEN C., WU C., VÉRATI C., GALLET S., MONCEL M.-H., COMBIER J., KHATIB S., MANETTI M., Application of U/Th and 40Ar/39Ar Dating to Orgnac 3, a Late Acheulean and Early Middle Palaeolithic Site in Ardèche, France PLoS ONE, 2013, 8, 12, p. e82394.

Moigne & Barsky 1999, MOIGNE A.-M., BARSKY D.R., Large mammal assemblages from Lower Palaeolithic sites in France: La Caune de l'Arago, Terra Amata, Orgnac 3 and Cagny l'Epinette, in The role of early humans in the accumulation of European lower and middle Palaeolithic bone assemblages, Gaudzinski S., Turner E. (dir.), Mainz, Römisch-germanisches Zentralmuseum - Forschungsinstitut für Vor- und Frühgeschichte, 1999, p. 219-236 (Monographien; 42).

Moncel 1996, MONCEL M.-H., Les niveaux profonds du site pléistocène moyen d'Orgnac : habitat, repaire, aven-piège ? L'exemple du niveau 6, Bulletin de la Société préhistorique française, 1996, 93, 4, p. 470-481.

Moncel 1999, MONCEL M.-H., Les assemblages lithiques du site pléistocène moyen d'Orgnac 3 (Ardèche, moyenne vallée du Rhône, France). Contribution à la connaissance du Paléolithique moyen ancien et du comportement technique différentiel des hommes au Paléolithique inférieur et au Paléolithique moyen, Liège, Service de Préhistoire - Université, 1999, 446 p. (Études et Recherches archéologiques de l'Université de Liège (ERAUL) ; 89).

Moncel et al. 2005, MONCEL M.-H., MOIGNE A.-M., COMBIER J., Pre-Neandertal behaviour during isotopic stage 9 and the beginning of stage 8: new data concerning fauna and lithics in the different occupation levels of Orgnac 3 (Ardeche, South-East France): occupation types, Journal of archaeological Science, 2005, 32, 9, p. 1283-1301.

Moncel et al. 2008, MONCEL M.-H., CRÉGUT É., DAUJEARD C., LARTIGOT A.-S., LEBON M., PUAUD S., BOULBES N., CROIZET S., Le site de la baume Flandin (commune d'Orgnac-l'Aven) : nouvelles données sur ce gisement du Paléolithique moyen = The site of the Baume Flandin (Orgnac-l’Aven): New data on this Middle Palaeolithic site, Comptes Rendus Palevol, 2008, 7, 5, p. 315-325.

Moncel et al. 2011, MONCEL M.-H., MOIGNE A.-M., SAM Y., COMBIER J., The Emergence of Neanderthal Technical Behavior: New Evidence from Orgnac 3 (Level 1, MIS 8), Southeastern France, Current Anthropology, 2011, 52, 1, p. 37-61.

Moncel et al. 2012, MONCEL M.-H., MOIGNE A.-M., COMBIER J., Towards the Middle Palaeolithic in Western Europe: The case of Orgnac 3 (southeastern France), Journal of Human Evolution, 2012, 63, 5, p. 653-666.

Moncel et al. 2020, MONCEL M.-H., ASHTON N.M., ARZARELLO M., FONTANA F., LAMOTTE A., SCOTT B., MUTTILLO B., BERRUTI G.L.F., NENZIONI G., TUFFREAU A., PERETTO C., Early Levallois core technology between Marine Isotope Stage 12 and 9 in Western Europe, Journal of Human Evolution, 2020, 139, p. 102735.

Monnier et al. 2006, MONNIER G.F., [WITH CA COMMENT BY], BISSON M.S., KLEIN R.G., MCNABB J., FLUCK H., MONCEL M.-H., STRAUS L.G., TUFFREAU A., The Lower/Middle Paleolithic Periodization in Western Europe: an Evaluation Current Anthropology, 2006, 47, 5, p. 709-744.

Moncel & Combier 1992a, Moncel, M.-H., Combier, J., Industrie lithique du site pléistocène moyen d’Orgnac 3 (Ardèche), Gallia Préhistoire, 1992, 34, p. 1-55.

Moncel & Combier 1992b, Moncel, M.-H., Combier, J., L’outillage sur éclat dans l’industrie lithique du site pléistocène moyen d’Orgnac 3 (Ardèche, France), L’Anthropologie, 1992, 96, p. 5-48.

Morales Pérez & Sanchis Serra 2009, MORALES PÉREZ J.V., SANCHIS SERRA A., The Quaternary fossil record of the genus Testudo in the Iberian Peninsula: archaeological implications and diachronic distribution in the western Mediterranean, Journal of archaeological Science, 2009, 36, 5, p. 1152-1162.

Mourer-Chauviré 1975a, MOURER-CHAUVIRÉ C., Faunes d'oiseaux du pléistocène de France : systématique, évolution et adaptation, interprétation paléoclimatique, Geobios, 1975a, 8, 5, p. 333-352, 6 fig., 1 pl.

Mourer-Chauviré 1975b, MOURER-CHAUVIRÉ C., Les oiseaux du Pléistocène moyen et supérieur de France, Université Claude-Bernard - Lyon I, 1975b, Thèse d'État : Sciences, 2 vol., 624 p.

Paunesco 2007, PAUNESCO A.-C., Les rongeurs (Rodentia, Mammalia) d’Orgnac-3 (Ardèche, France). Étude paléontologique et paléoécologique, Buletinul Muzeului "Teohari Antonescu" Giurgiu, 2007, 9, 9-12, p. 11-94.

Pedley 1990, PEDLEY H.M., Classification and environmental models of cool freshwater tufas, Sedimentary Geology, 1990, 68, 1-2, p. 143-154.


Reich et al. 2010, REICH D., GREEN R.E., KIRCHER M., KRAUSE J., PATTERSON N., DURAND E., VIOLA B., BRIGGS A.W., STENZEL U., JOHNSON P.L.F., MARICIC T., GOOD J.M., MARQUES BONET T., ALKAN C., FU Q., MALLICK S., LI H., MEYER M., EICHLER E.E., STONEKING M., RICHARDS M.P., TALAMO S., SHUNKOV M.V., DEREVIANKO A.P., HUBLIN J.-J., KELSO J., SLATKIN M., PAABO S., Genetic history of an archaic hominin group from Denisova Cave in Siberia, Nature, 2010, 468, 7327, p. 1053-1060.

Rightmire 1998, RIGHTMIRE G.P., Human evolution in the middle Pleistocene: the role of Homo heidelbergensis, Evolutionary Anthropology, 1998, 6, 6, p. 218-227.

Rightmire 2008, RIGHTMIRE G.P., Homo in the Middle Pleistocene: Hypodigms, variation, and species recognition, Evolutionary Anthropology, 2008, 17, 1, p. 8-21.

Roebroeks & Villa 2011, ROEBROEKS W., VILLA P., On the earliest evidence for habitual use of fire in Europe, Proceedings of the National Academy of Sciences of the United States of America, 2011, 108, 13, p. 5209-5214.

Sam 2009, SAM Y., Étude paléontologique, archéozoologique et taphonomique des grands mammifères du site Pléistocène moyen d'Orgnac 3 (Ardèche, France), Université de Perpignan, 2009, Thèse de doctorat en Préhistoire. Archéozoologie.

Shea 2009, SHEA J.J., The Impact of Projectile Weaponry on Late Pleistocene Hominin Evolution, in: The evolution of hominin diets: integrating approaches to the study of palaeolithic subsistence, Hublin J.-J., Richards M.P. (dir.), Amsterdam, Springer, 2009, p. 189-199.

Shen 1985, SHEN G., Datation des planchers stalagmitiques de sites acheuléens en Europe par les méthodes des déséquilibres des familles de l'uranium et contribution méthodologique, Université Pierre et Marie Curie (Paris VI) / Museum national d'Histoire naturelle, 1985, Thèse de doctorat d'État, 162 p.

Sommer et al. 2007, SOMMER R.S., PERSSON A., WIESEKE N., FRITZ U., Holocene recolonization and extinction of the pond turtle, Emys orbicularis (L., 1758), in Europe, Quaternary Science Reviews, 2007, 26, 25-28, p. 3099-3107.

Srdoč et al. 1994, SRDOČ D., OSMOND J.K., HORVATINČIĆ N., DABOUS A.A., OBELIĆ B., Radiocarbon and Uranium-Series Dating of the Plitvice Lakes Travertines, Radiocarbon, 1994, 36, 2, p. 203-219.

Stuart 1982, STUART A.J., Pleistocene vertebrates in the British Isles London, Longman, 1982, 212 p.

Sutcliffe 1985, SUTCLIFFE A.J., On the Track of Ice Age Mammals, London, British Museum of Natural History, 1985, 224 p.

Testu 2006, TESTU A., Étude paléontologique et biostratigraphique des Felidae et Hyaenidae pléistocènes d'Europe méditerranéenne. (sites de la Caune de l'Arago, Orgnac 3, le Portel-Ouest, Bize-Tournal, l'Hortus, la Crouzade en France, la Covade de l'Arbreda en Espagne, Karaïn E en Turquie, Université de Perpignan, 2006, Thèse de doctorat en Préhistoire, 320 p.

Tillier & Vandermeersch 1976, TILLIER A.-M., VANDERMEERSCH B., Les Cynomorphes, in La Préhistoire française. Tome 1 : Les Civilisations paléolithiques et mésolithiques de la France, de Lumley H. (dir.), Paris, Éditions du CNRS, 1976, p. 367-370.

Villa 1977, VILLA P., Sols et niveaux d'habitat du Paléolithique inférieur en Europe et au Proche-Orient, Quaternaria, 1977, 19, p. 107-134.

Weaver et al. 2008, WEAVER T.D., ROSEMAN C.C., STRINGER C.B., Close correspondence between quantitative- and molecular-genetic divergence times for Neandertals and modern humans, Proceedings of the National Academy of Sciences of the United States of America, 2008, 105, 12, p. 4645-4649.

Haut de page



The documentary appendices are available in the Mediterranean Prehistories collection in Nakala’s repository (Huma-Num).

Haut de page

Table des illustrations

Titre 1. The approximate locations of major European middle Middle Pleistocene sites with early human fossils or artifacts.
Fichier image/jpeg, 1,3M
Titre 2. Major European Middle Pleistocene sites showing proposed dates and presence of hominin remains, fire, lithics, and faunal remains (using the chronology for Orgnac 3 proposed herein). See Supplement for references.
Fichier image/jpeg, 887k
Titre 3. Aerial plan of the site showing areas of excavation and core soundings (adapted from Aouraghe 1992: 20).
Fichier image/jpeg, 758k
Titre 4. Proposed phases of site formation and development for Orgnac 3: left from Combier 1967: 35; right from Debard 1988: 91.
Fichier image/jpeg, 268k
Titre 5. Stratigraphic sections from Orgnac 3. Left, from Combier 1967: 37), and Right, from Debard 1988: 57). See figure 3 for locations of profiles.
Fichier image/jpeg, 1,4M
Titre Table 1. Geological levels, archaeological levels, lithic industries, faunal assemblages, paleoecological and paleoclimatological reconstructions, and chronology for Orgnac 3 (adapted from Aouraghe 1999: 182, and Laafar 1995: 61).
Fichier image/jpeg, 523k
Titre Table 2. Every radiometric date ever published for the Orgnac 3 sequence. Please see main text for background, discussion, and references.
Fichier image/jpeg, 566k
Titre 6. Comparison of the (A) flowstone (left) and stalagmite (middle) from Orgnac 3, with a closeup of the stalagmite (right) from Michel et al. 2013, to cross sections of (B) travertine speleothem (Luray Caverns, Virginia, USA, photo credit: James St. John), and (C) typical dripstone speleothem calcite (from figure 2 of Schwarcz 2007).
Fichier image/jpeg, 896k
Titre 7. Two separate age models that can be constructed from published data for the Orgnac 3 deposits (background adapted from Khatib 1989: 10).
Fichier image/jpeg, 302k
Titre 8. Abundances of Emys and Testudo carapace remains in Orgnac 3’s levels.
Fichier image/jpeg, 266k
Haut de page

Pour citer cet article

Référence papier

Jason E. Lewis, Jean Combier† et Ludovic Slimak, « Rethinking Orgnac 3 in Time and Space: Implications for Reconstructing Neandertal Origins »Préhistoires Méditerranéennes, 10 | -1, 77-111.

Référence électronique

Jason E. Lewis, Jean Combier† et Ludovic Slimak, « Rethinking Orgnac 3 in Time and Space: Implications for Reconstructing Neandertal Origins »Préhistoires Méditerranéennes [En ligne], 10 | 2022, mis en ligne le 13 novembre 2023, consulté le 21 février 2024. URL : ; DOI :

Haut de page


Jason E. Lewis

Department of Anthropology and Turkana Basin Institute, Stony Brook University, Stony Brook, New York 11794-4364, USA. *Corresponding Author.

Articles du même auteur

Jean Combier†

Articles du même auteur

Ludovic Slimak

Laboratoire CAGT, Faculté de Médecine Purpan, Bat A, 37 Allées Jules Guesde, 31000 Toulouse, France.

Articles du même auteur

Haut de page

Droits d’auteur


Le texte seul est utilisable sous licence CC BY 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search