Les terrasses fluviatiles quaternaires de la Somme à Abbeville : 150 ans de recherches mêlant géologie et archéologie,de Boucher de Perthes à aujourd’hui

Paleolithic localities associated with fossil fluvial formations of the Somme River at Abbeville have played a significant role in the recognition of human antiquity. From the end of the 18th century, the work carried out in particular by the “Société d’émulation d’Abbeville” led to the emergence of studies in the Somme valley on both Quaternary Geology and Prehistory. It was within this scholarly society that Jacques Boucher de Perthes initiated his research on the famous localities of Menchecourt, Hospital and Moulin Quignon in 1837. These observations were quickly confirmed by geologists such as Albert Gaudry, Charles-Joseph Buteux or Joseph Prestwich. It was the latter who introduced the Abbeville deposits into the United-Kingdom scientist community and his observations were used as early as 1859 by Charles Lyell to demonstrate the antiquity of the archaeological remains collected from these localities. However, the controversy surrounding by the discovery of the human jaw of Moulin Quignon (1863–1864), which undermined the credibility of this locality for a long time, led to the decrease of the archaeological researches in the Somme valley, and in particular in Abbeville, but it then resumed quite quickly. However, if the work carried out by Geoffroy d’Ault du Mesnil from 1875 to 1898 had the merit to reaffirm the archaeological importance of the Abbeville deposits and to reveal the complexity of their Quaternary records, the geographical and stratigraphic imprecision resulting from these rare publications, coupled with the gradual abandonment of the various gravel pits whose exploitation he followed, cast a significant discredit on these discoveries and on the stratigraphical observations he made there. The researches carried out then by Victor Commont at Carpentier Quarry between 1904 and 1918 were characterized by the excellence of the stratigraphic and iconographic data he produced for each of the studied sites. However the absence of archaeological discoveries in the fluvial deposits he studied lead the prehistorians who succeeded him increasingly doubted of the veracity of Ault's observations, despite the efforts made by Abbé Henri Breuil to reconcile these with Commont's observations. Between the two world wars, Breuil and Léon Aufrère resumed the study of the Abbeville deposits then still accessible, tried to ensure their preservation, through the purchase and protection of the Carpentier and Léon quarries, a decisive intervention in the archaeological rehabilitation of these deposits. These works were at the origin of the chronostratigraphic studies carried out on these major localities of the ancient European prehistory until today. Over the past ten years, some of these localities have been studied again, thus largely confirming the observations made by the elders (Boucher de Perthes, d’Ault du Mesnil, Commont) and validating the major place of Abbeville in the understanding of ancient prehistory settlement of northwestern Europe.     

Retourner à Moulin Quignon

In March 1863, Jacques Boucher de Perthes found in the alluvia of Moulin Quignon (Abbeville, France) a human jaw in association with lithic artifacts and fossil bones of mammals. Quickly challenged, this discovery will appear as a fraud whose prehistorian was the victim. Since that time, the collection and the Moulin Quignon site are synonymous with an archaeological forgery. Forgotten for almost 150 years, these collections became in 2012 a research topic in the context of a program devoted to oldest settlements and Acheulean. The objective was to identify and study the whole 1863–1864 documentation set (human remains, artifacts, sediments, fauna bones, malacological remains, archives, papers, site) to allow further analysis of the whole subject.     

La redécouverte du site de Moulin Quignon (Somme) et les premiers Acheuléens du Nord-Ouest de l’Europe (650 ka)

New surveys were undertaken in Abbeville in 2016 and 2017 on the site of the Moulin Quignon discovered before 1850 by Boucher de Perthes, relocated thanks to the archival work carried out by MNHN researchers. This research led 150 years later to the rediscovery of this emblematic Paleolithic site. Test pits and extensive excavation have led to the discovery of alluvial levels in situ and a lithic series composed of 254 fragments, 15 cores and 5 bifaces. The cores and flakes attest to in situ debitage on flint nodules according mainly to unipolar and centripetal unifacial technology, except for one core, which indicates a management on several debitage surfaces. The bifaces are diversified morphologically and technically (shaping modes). They attest to a general management of the bifacial volume and geometry of the tool, to the care taken at the tip of the tool and to the regularisation of the often peripheral edges. They resemble those discovered by Boucher de Perthes in 1837–1868 and described in the past as “Abbevillian”. Stratigraphic and ESR age studies on quartz date human occupation between 670 and 650 ka. These artifacts belong to the earliest Acheulean occupation ever dated in the Somme basin and more generally north of the 50th parallel north. The lithic series discovered at Moulin Quignon is indisputable proof of the presence of human occupations dated at more than 650 ka in the Somme valley, a natural pathway to northern sectors, present-day Great Britain. This discovery thus ages (1) the age of the oldest occupation in the North of France (previously 550 ka) and (2) the age of the oldest bifaces for Northwest Europe (100 ka older than the bifaces in Great Britain).     

La séquence de Moulin Quignon est-elle une séquence archéologique ?

The sequence of the Moulin Quignon quarry yielded archaeological findings (faunal remains, artefacts) and some of them are well located in the stratigraphy thanks to Jacques Boucher de Perthes and colleagues. These remains are often used to attest the early age of the site and the human remains. Some bifaces have been found in close relationship to the human remains and, in particular, the mandible found in 1863. Thanks to the handwritten label put on the pieces and their deposit at the National Museum of Natural History (Paris), it is possible to study them again in relation with the sequence. The study concerns the rare faunal remains and the bifaces given, among others, by Boucher de Perthes to the Museum. The paleontological and technological analyses allow us replacing the lithic material among the north-western European data. The material is presented according to the historical context of his discovery.     

Moulin Quignon : la redécouverte d’un site

The collection from Moulin Quignon is a coherent one, despite the diversity of its pieces, in the certainty of its origin, in its attachment to the researches led on the site in 1863–1864. Considered as a historic heritage, its study has nevertheless delivered significant scientific information, valid for the present, and, beyond that, to reintroduce the site of Moulin Quignon in the oldest Palaeolithic panel sites of the Somme valley.     

La stratigraphie du Pléistocène supérieur et le Paléolithique en Corée

In Korea, Quaternary deposits are characterized by marine and fluvial terrace deposits, commonly found in the coastal areas and at the sides of river valleys. The major terraces are classified into 5 different levels, i.e., alluvial plain (3-5 m asl), lower terrace (10-25 m asl), middle terrace (33-50 m asl), high terrace (55-85 m asl), and highest terrace (90-160 m asl). They were formed by the repetition of sedimentation and erosion due to sea level fluctuation and uplifting along the coastal areas during Quaternary period. The formation age of the lower marine terraces at the level of 10-20 m is interpreted to be the last interglacial (75-125 Ka, MIS = 5a-5e). The middle coastal terrace gravels distributed at the level of about 33-45 m above mean sea level and are to be formed before the last interglacial. Above coastal terrace deposits, slope deposits and paleosol layers are overlain and freezing and thawing crackings were developed in them. Major prehistory archaeological sites in the coastal areas are found in the slope deposits formed during last glaciations. The upper parts of the slope deposits are typified by paleosol layers with typical frost cracks (soil wedges) and frangipanes. As geochronological bases for subdividing Paleolithic cultures, two typical paleosol layers are useful because they are mostly found and imprinted with several horizons of frost cracks or soil wedges. The uppermost crack structures were formed under the boreal-periglacial climatic regime during the last glacial period of extremely cold-dry climate in Korea. The upper pedogenetic sediments containing frost cracks or soil wedges were formed between 24 and 17 Ka in many Paleolithic profiles. The lowermost paleosol layer with frost cracks (soil wedge), strongly attacked into hillslope sediments or last interglacial saprolite layers on the weathered basement rocks, were detected at least as old as 59-74,000 Ka. Lastly upper Paleolithic cultural layers were found in the young fluvial deposits, which had been formed since the Last Glacial Maximum (LGM). Some of them were particularly continued as late as the Pleistocene-Holocene Transition Period.     

Boucher de Perthes (1788–1868). Des rêves métaphysiques à l’invention de la préhistoire

The director of customs in Abbeville (Somme), Jacques Boucher de Perthes (1788-1868) had the project of renewing Christianity with metempsychosis. He had no scientific training but, as president of the Société d’émulation d’Abbeville, he was aware of what was being debated in the sciences. He integrated the scientific problem of fossil man into his religious faith. He thus discovered the Palaeolithic. His flagship book, the three volumes of Celtic and Antediluvian Antiquities (1849, 1857, 1864), one of the founder works of Prehistory, succeeded in combining science and religion and this is what makes his study so interesting.     

Moulin Quignon 1863–1864 : détours inédits et bilan historiographique

The historical file of Moulin Quignon (Abbeville, France), around the human jaw unearthed in this site, is a form of black hole that swallowed a prehistoric archaeological site and Lower Palaeolithic collection. This article, based on a corpus of unpublished archives, illuminates the matter in a different light. It focuses on archaeological and epistemological issues, the roles of the different actors, interaction games and solidarity in a global scientific community previously welded by the recognition in 1859 of the paradigm of high antiquity of Man.     

Moulin-Quignon et Homo heidelbergensis. Contextes épistémologiques et enjeux taxinomiques

In 1863, when a human mandible was discovered in the quarry of Moulin-Quignon (Abbeville, Somme), it completed the demonstration of Jacques Boucher de Perthes (1788–1868) proving the existence of an ante-diluvian man. However, this discovery was quickly questioned and would eventually disqualify even the site itself and the industries it had delivered. The recent re-examination of these fossils brought to light in 1863 (one mandible) and 1864 (28 bones and teeth), now housed at the Muséum national d’Histoire naturelle (musée de l’Homme), has confirmed their attribution to Homo sapiens. In the light of current knowledge, this attribution is incompatible with the stratigraphic provenance of these anthropological remains, whose age is estimated between 670–650,000 years. In fact, direct radiocarbon dating of these bones dates them back to a historical period, between the 13th and 18th centuries, which confirms their intrusive nature in the deposit. In this paper, after describing the discovery and rediscovery of the site, we will question who is the man who could have been discovered at Moulin-Quignon. This will be the occasion for a review of human evolution, in the context of the taxonomic upheavals of the 1950s and 1960s, in Western Europe in the Middle Pleistocene, where recent discoveries attest to a significant variability among Homo heidelbergensis, some of which are already strongly involved in the Neanderthal lineage. This retrospective will voluntarily follow the course of paleo-antropological research throughout the 20th century in order to highlight, as discoveries are made, changes in paradigms and practices.     

Le site de Longgupo dans son environnement géologique et géomorphologique

On a regional scale, Longgupo is found within a karstic landscape of fenglin-ouvala developed in the Triassic limestones of the Yangtze platform. After several seasons, it is now possible to define and propose a scenario for the formation of the site. During one of the intense karstification phases during the Cretaceous or early Tertiary, karstic conduits were created deep in the substratum. In concert with these deep flows and by karstic collapse, a cave was created. With the start of Himalayan uplift during the Oligocene and Miocene, the beginning of fluvial cutting was reflected by the partial erosion of the upper sandstone formations and the creation of the Miaoyu Valley. This action opened the cave, which would later be occupied by the hominid groups that interest us here. Probably at the end of the Tertiary, the cave, open at the bottom of the valley slope and facing upstream, began to be filled with alluvial deposits from the valley. The basal deposits (C III 10 of the current stratigraphic sequence, levels 20-13 of the first sequence) begin with black clays; at the Plio-Pleistocene transition, deposits become clearly alluvial, increasingly marked by reworking of surface and karstic clays and by systematically coarser contributions. During the deposition period of archaeological complexes III/III’ (previously levels 12-6), karstic collapse appears to have still been active. During the Lower Pleistocene sensu lato, the coarse nature of the deposits becomes accentuated. The material of complex II (previously levels 5-2) is a deposit that is clearly heterometric and coarse, composed at the end mainly of gravels, pebbles and more or less rounded blocks in a clayey matrix. At the conclusion of this new stage, the cave was entirely filled and, purely and simply, disappeared from the landscape. As the valley deepened during the Middle and Upper Pleistocene, the Longgupo slope underwent profound changes, the last of the limestone blocks disappeared from the slope limit and the archaeological site was protected by the coarse breccias that carpeted the slope. The main cause for the sedimentation processes at the site of Longgupo is the hydrosystem of the Miaoyu he, although injections of karstic type sensu stricto cannot be excluded. All of the arguments effectively support alluvial deposition: pebbles, an alternation of coarse and clayey beds, systematic and repeated granulometric decrease. But the stratigraphic disposition was disturbed by deformations due to gravity and karstic origins; the alluvial sediments, rendered plastic by the high proportion of clayey beds, were deformed by gravity (deformation of beds, faults, slopes) following the direction of the slope and by karstic collapse along the two limestone walls north and south of the sites (in sub-parallel faults to the south and by circular aspiration to the north). To this is added a calcitic consolidation of the upper levels that made it difficult to excavate complex C II. The stratigraphic disposition of the archaeological levels at Longgupo is now well understood. It has thus been possible to collect samples for absolute dating, to be published soon.     

Late Quaternary paleoenvironmental records from the western Lena Delta,Arctic Siberia

The three main Lena Delta terraces were formed during different stages of the late Quaternary. While only the first floodplain terrace is connected with active deltaic processes, the second and third terraces, which dominate the western part of the delta, are erosional remnants of arctic paleolandscapes affected by periglacial processes. The landscape dynamics of the second and the third terraces, and their relationship to each other, are of particular importance in any effort to elucidate the late Quaternary paleoenvironment of western Beringia.Multidisciplinary studies of permafrost deposits on the second terrace were carried out at several sites of the Arga Complex, named after the largest delta island, Arga–Muora–Sise. The frozen sediments predominantly consist of fluvial sands several tens of meters thick, radiocarbon-dated from > 52 to 16 kyr BP. These sands were deposited under changing fluvial conditions in a dynamic system of shifting river channels, and have been additionally modified by synsedimentary and postsedimentary cryogenesis. Later thermokarst processes affected this late Pleistocene fluvial landscape during the Lateglacial and the Holocene. In addition, eolian activity reworked the fluvial sands on exposed surfaces at least since the Lateglacial, resulting in dune formation in some areas. Contrary to the Arga Complex, the third terrace is mainly composed of polygenetic alluvial and proluvial ice-rich permafrost sequences (Ice Complex deposits) radiocarbon-dated from 50 to 17 kyr BP which cover older fluvial sand units luminescence-dated to about 100–50 kyr BP. Paleoecological records reflect tundra-steppe conditions that varied locally, depending on landscape dynamics, during the Marine Isotope Stage (MIS) 4 and 3 periods, and a persistent change to shrub and arctic tundra during Lateglacial and Holocene periods.The study results indicate a continuous fluvial sedimentation environment for the Laptev Sea shelf in the region of the second Lena Delta terrace during the late Pleistocene, and confirm the presence of a dynamic channel system of the paleo-Lena River that flowed at the same time as the nearby subaerial Ice Complex deposits were being formed.     

État des connaissances sur la préhistoire et l’histoire des recherches en vallée de la Somme (XIXe–XXIe siècles)

During the middle of the 19th century, the Somme valley played a major role in the recognition of prehistory as a scientific discipline. It is still at the heart of issues and research on the first human settlements in northwestern Europe. The multidisciplinary colloquium “Toute une (pré)histoire en Somme”, which was held on 22–24 November 2018 in Abbeville, set itself the primary objective of establishing a state of knowledge and current issues in the fields of prehistoric studies and the history of prehistory.     

Stratigraphie et chronologie du site de Gaolingpo dans le bassin de Bose,sud de la Chine

The Bose Paleolithic industry in the Bose Basin, Guangxi, South China is well known in the international scientific community for its bearing handaxes dated back to 803 ka. Seven terraces were developed in the basin and terrace 4 is the most important because handaxes and tektites were recovered from this terrace. Since the first site was discovered in 1973, additional sites of the Bose Paleolithic industry have been investigated and excavated, most of which are located on terrace 4. The previous studies concluded that there is only one phase within this industry and its age is 803 ka. However, before 2013, no attempt had been made to dig down to the basal gravels from the top of terrace 4 in archaeological excavations. Therefore, the stratigraphy and the layers from which the stone artifacts derive of terrace 4 are not clear. In 2013–2014, we conducted an excavation of 200 m² at the Gaolingpo site, which is one of the most important Paleolithic sites in the Bose Basin. We excavated to the basal gravel deposits from the top of this terrace, and a section of more than 7 meters thick deposit was exposed and clearly shows the stratigraphic layers of Pleistocene and Holocene deposits. Furthermore, features and about 800 stone artifacts including choppers, picks, scrapers, etc. were excavated from several stratigraphic layers, which are different in age and other aspects. Based on stratigraphy and the stone artifacts, three stages of Paleolithic cultural remains from the Gaolingpo site can be established. The first stage is in the age of at leat 803 ka. The second stage can be dated to 15 ka and the third stage to about 10 ka.     

Galets aménagés en provenance des terrasses supérieures du Rhône au sud de Lyon (Sud-Est de la France)

Between 1990 and 1993, we found 175 pebble-tools and two flake-tools on the western high-terrace south of Lyon (Southeast, France) at two different levels (290 m and 235 m NN). Most of them show the classical spectrum of choppers and chopping-tools, joined by six protobifaces. They all belong to the developed Olduvan (Olduvan II). On the lower terrace, corresponding with its younger age, we met a more progressive pebble-tool culture. Both stocks are secondary sites, brought from the East by the most extreme tongue of the Rhône glacier and, therefore, cannot be dated exactly.     

Les industries acheuléennes de la France septentrionale

The excavations of several open air sites in Northern France and, especially in the Somme basin, have given numerous data concerning the age and the environment of the human occupations for the period from MIS 12 to MIS 8. The quality of the raw material and the function of the different sites have a great effect on the composition of the lithic assemblages as it has been seen for the different sites located near the confluence of the Somme and Avre (Cagny-la Garenne, Cagny-l’Épinette, Ferme de l’Épinette, Gentelles). The characteristics of the lithic industries of Gouzeaucourt show the complexity of the transition from Lower to Middle Palaeolithic.     

Les nappes alluviales de la Basse Somme revisitées

The new approach of the fluviatile sequence preserved downstream of Abbeville from Port-le-Grand to Saint-Valery, brings together two sets of complementary data in a same canvas. The first set combines the field records of V. Commont and A. Briquet slightly above the basal cravels. The second set encompasses the data gained for the high and very high terraces mapped in the eighties. They point to a succession of ten main terraces and two second order ones, occurring in continuity with the Abbeville terrace system. The upper terrace being related to Jaramillo, we are dealing with one million year strong regional record with a high-resolution palaeoenvironmental signature fitting in with the ODP 677 marine isotopic record and with the Chinese loess record. In this way, the laminar tool-kit of Croix-l’Abbé could be positioned in the middle part of MIS 8, around 250 ka.     

Charophytes, indicateurs de paléobathymétrie du lac Tigalmamine (Moyen Atlas, Maroc)

A paleolimnological study can be particularly accurate in situations where the species found in the sediments have persisted in the same lake and can provide a direct modern analogue for interpretation. Lake Tigalmamine, thus, represents an exceptionally favourable context, compared to most paleolakes in North Africa, as this site allows comparing the fossil remains of the Characeae to those produced by the living plants. Tigalmamine is a hydrosystem composed of three karstic lakes located at 1626 m above sea level in the Middle Atlas of Morocco. The site had been the subject of a 16 m long core taken in the center of the central lake under 16 m of water (core C 86 in Lamb and Kaars, 1995 [The Holocene 5 (1995) 400-408]). The section represents the total of the Holocene from 10,200 BP to Present (El Hamouti et al., 1991 [Comptes Rendus de l’Académie des Sciences de Paris 2 (1991) 259-265]; Lamb et al., 1995 [Nature 373 (1995) 134-137]). The latter published a synthesis of the pluridisciplinary studies conducted on this core. Also, the numerous gyrogonites isolated from that core had been determined as Chara hispida (Soulié-Märsche in Benkaddour, 1993 [Changements hydrologiques et climatiques dans le Moyen-Atlas marocain : chronologie, minéralogie, géochimie isotopique et élémentaire des sédiments lacustres de Tigalmamine. PhD thesis, Université Paris-Sud, Orsay (inédit)]). Further fieldwork at Tigalmamine by the authors of the present paper in year 2000 showed that Chara hispida is still present in the lake where it forms a belt of submerged vegetation restricted to the range of 2 to 9 m water depth. This paper describes the morphological and ecological characteristics of the extant Chara hispida as a reference for the interpretation of the fossil remains. Based on these data, the frequency changes of the gyrogonites in the core sediments become significant in terms of bathymetry. The detailed analysis of the charophytes from the core, as a complement to the previous studies, confirms the existence of four Holocene phases with lake level low stand at Tigalmamine. The aim of the present study is to enhance the value of the charophytes as a useful complementary tool for paleolimnological reconstruction.     

Évolution des caractères technologiques et typologiques des industries lithiques dans la stratigraphie de la Caune de l'Arago

The thick Quaternary deposits of the Caune de l’Arago (Pyrénées-Orientales, France) are dated to between 690 000 and 90 000 years old. At least fifteen different archeostratigraphical units have been identified within these deposits, each corresponding to distinct prehistoric occupations. Numerous stone tools made from several different rock types, have been discovered in each unit. The tools present specific characteristics concerning the choice of raw materials, the typology, and the technology used to produce them. Morpho-technological study of the different components of the assemblage contributes to a better understanding of the debitage methods used for their production. Each raw material is considered individually in order to ascertain its origin in the environment, its typological role and the technology applied during its exploitation. Defining production systems leads to the characterisation of the assemblages from each unit. When compared, they reveal common elements, as well as differences, suggesting evolutionary trends. Some observations are also made concerning the extent to which changing uses of the site may have influenced the general morphology of each assemblage, therefore taking into account exterior factors. Analysis of this rich stone tool assemblage helps to situate the Caune de l’Arago industry within the larger evolutionary context of the Lower Paleolithic in Mediterranean Europe.     

Alluvial sequence in the north piedmont of the Chinese Tian Shan over the past 550 kyr and its relationship to climate change

A new division of Middle and Late Pleistocene alluvial sequence in the north piedmont of the Chinese Tian Shan based on geomorphologic, stratigraphic, and chronologic criteria provides a framework for examining their relationship to climate change during glacial–interglacial transitions. Over the past 550 kyr at least four major alluviation episodes occurred within the piedmont. Along the major river valleys in this region, each episode of alluvial fan deposition morphologically correlates with a major river terrace. These correlations create a regionally applicable framework for subdivision of the Quaternary alluvial sequence in the study area, where seven stepped river terraces are defined. Our new chronology of this fluvial sequence suggests that, following intervals of aggradation, three highest river terraces and equivalent alluvial fans were abandoned at ～ 530 ka, ～ 300 ka and ～ 10 ka, respectively. Paleosols at the base of the loess sequences that directly overlie the older terraces and fans suggest that episodes of aggradation occurred late in the glacial cycles. The subsequent incision that caused abandonment/stabilization of these terraces and fans occurred near to glacial–interglacial transitions. A relatively high degree of synchrony in major river incision events across the piedmont, despite disruption by several discrete structure zones with asychronous tectonic activities, supports the dominant control exerted by climatic conditions on alluvial deposition and terrace creation during the Quaternary across the north piedmont of the Chinese Tian Shan.