The Neanderthal legacy: an archaeological perspective from western Europe 9781400843602, 9780691034935, 9780691167985

Citation preview

CY

EANDERTHAL | | An Archaeological Perspective from Western Europe

NEANDERTHAL A An Archaeological Perspective

from Western Europe Paul Mellars Department of Archaeology Cambridge University, UK

PRINCETON UNIVERSITY PRESS, PRINCETON, NEW JERSEY

Copyright © 1996 by Princeton University Press Published by Princeton University Press, 41 William Street, Princeton, New Jersey 08540 In the United Kingdom: Princeton University Press, Chichester, West Sussex All Rights Reserved

Library of Congress Cataloging-in-Publication Data Mellars, Paul. The Neanderthal legacy : an archaeological perspective from western Europe / Paul Mellars.

p. cm.

, Includes bibliographical references and index. ISBN 0-691-03493-1 (cloth : acid-free paper) 1. Neanderthals — Europe. 2. Paleolithic period — Europe. 3. Human evolution — Europe —- Philosophy. 4. Behavior evolution — Europe. 5. Europe — Antiquities. IJ. Title.

GN285.M45 1995 | 936 — dc20 95-4300

This book has been composed in Palatino Princeton University Press books are printed on acid-free paper and meet the guidelines for permanence and durability of the Committee on Production Guidelines for Book Longevity of the Council on Library Resources Printed in the United States of America by Princeton Academic Press

10987654321

ontents

List of Tables vii

List of Illustrations 1x

Preface xvii

Chapter 1 Introduction 1 Chapter 2 The Environmental Background to Middle Palaeolithic Occupation 9

Chapter 3 Stone Tool Technology 56 Chapter 4 Tool Morphology, Function and Typology 95 Chapter 5 The Procurement and Distribution of Raw Materials 141

Chapter 6 Industrial Taxonomy and Chronology 169

Chapter 7 Middle Palaeolithic Subsistence 193

Chapter 8 Sites in the Landscape 245

Chapter 9 The Spatial Organization of Middle Palaeolithic Sites 269 | Chapter 10 The Significance of Industrial Variability 315

Chapter 11 Neanderthal Society 356 Chapter 12 The Neanderthal Mind 366

Chapter 13 The Big Transition 392 References 420

Index of Sites 461

General Index 465 v

List of Tables

Table 3.1 Principal stages of flake and tool production, use and discard in the reduction sequence 58 scheme of Geneste.

Table 3.2 Principal products recognized by Geneste in his reduction sequence scheme for the 59 production of Levallois flakes.

Table 3.3 Average numbers of dorsal flake scars on samples of Levallois flakes in Middle 93 Palaeolithic assemblages from France and the Middle East.

Table 5.1 Principal Middle Palaeolithic sites analysed for raw material sources in southwestern 146 France.

Table 5.2 Quantities and extent of utilization of raw materials transported over varying distances 147 in Middle Palaeolithic sites.

Table 6.1 Standard type-list of Lower and Middle Palaeolithic flake tools proposed by Francois 170 Bordes.

Table 6.2 Sites with levels of Mousterian of Acheulian tradition overlying levels of Quina or 184 Ferrassie Mousterian in western France.

Table 6.3 Sites with levels of Mousterian of Acheulian tradition underlying Upper Palaeolithic 187 levels.

Table 10.1 Mean lengths of racloirs and unretouched flakes in levels of Ferrassie/Quina and 328 Denticulate Mousterian at Combe Grenal.

Table 12.1 ‘Neocortex ratios’ for different hominid taxa over the past two million years. 368

vil

List @ 5é of Illustrations Figure 2.1 General pattern of oxygen-isotope fluctuations in deep-sea sediments over the past 10 300,000 years.

Figure 2.2 Temperature variations over the past 140,000 years estimated from oxygen-isotope and 10 , deuterium ratios in the Voztok ice core (eastern Antarctica).

Figure 2.3 Typical vegetational succession in pollen sequences spanning the last interglacial 12 period in northern Europe.

Figure 2.4 Estimated mid-summer temperatures across Europe at the peak of the last interglacial. 13

Figure 2.5 Estimated mid-summer temperatures in Europe at the time of the Amersfoort 13 interstadial.

Figure 2.6 Vegetational sequence in the long pollen core from La Grande Pile (northeastern 15 France).

Figure 2.7 Vegetational sequence covering the last glacial/interglacial cycle at Les Echets (eastern 16 France).

Figure 2.8 Estimates of mean annual temperatures and precipitation over the past 140,000 years 18 , derived from climatic analyses of the vegetational patterns at Les Echets and La Grande Pile.

Figure 2.9 Principal stadial and interstadial periods in northern and western Europe during the 19 earlier part of the last glaciation, with associated summer temperature estimates.

Figure 2.10 Reconstructed vegetational sequence through the last interglacial/glacial cycle in 20 Holland and northern Germany.

Figure 2.11 Estimated mid-summer sea-surface temperatures in the north Atlantic region over the 21 past 130,000 years recorded in deep-sea core V23-82.

Figure 2.12 Reconstructed fluctuations of cold and warm sea-water currents in the north Atlantic 22. region, based on faunal evidence from deep-sea cores.

Figure 2.13 Estimated mid-summer and mid-winter sea-surface temperatures in the Atlantic 24 region at the time of the last-glacial maximum, ca 18,000 BP.

Figure 2.14 Fluctuating oxygen-isotope ratios recorded through the ‘GRIP Summit’ ice core 26 (Greenland).

Figure 2.15 Oxygen-isotope ratios over the past 140,000 years recorded in core KET-8004, from the 27 northern Mediterranean.

Figure 2.16 Comparison of age estimates for various interstadial episodes during isotope stage 3, 28 derived from various climatic sequences.

Figure 2.17 Sea level fluctuations over the past 140,000 years derived from studies of oxygen- 29 isotope ratios in surface and deep-water ocean sediments and coral terraces in New Guinea.

Figure 2.18 Distribution of loess and related wind-blown deposits in western Europe. 31

Figure 2.19 Chronology of major periods of loess deposition in northern France during the last 32 glacial period.

Figure 2.20 Stratigraphy of the rock-shelter deposits at Combe Grenal (Dordogne, southwestern 34 France).

Figure 2.21 Pollen succession through the Combe Grenal sequence. 36 1x

x LIST OF ILLUSTRATIONS Figure 2.22 Frequencies of principal faunal taxa recorded in the Combe Grenal sequence. 37 Figure 2.23 Proposed correlation between the climatic sequence at Combe Grenal and the sequence 39 of oxygen-isotope stages in deep-sea cores.

Figure 2.24 Pollen sequence through the earlier last-glacial deposits at Pech de l’ Azé II. 43 Figure 2.25 Proposed correlation of the faunal and vegetational sequence at Pech de l’Azé II. 44

Figure 2.26 Electron-spin-resonance dates for the sequence at Pech de |’ Azé II. 44

Figure 2.27 Stratigraphic section in the lower shelter at Le Moustier. 45 Figure 2.28 Stratigraphic and archaeological sequence in the lower shelter at Le Moustier. 46 Figure 2.29 Pollen sequence through the upper levels at Le Moustier. 47

Figure 2.30 Pollen sequence through the later Mousterian and Chdatelperronian levels at Arcy-sur- 49 Cure (central France).

Figure 2.31 Reconstruction of major vegetation zones, ice sheets, and coastlines in Europe at the 51 time of the last glacial maximum.

Figure 2.32 Estimated extent of glaciers in the Massif Central region of central France at the time 54 of the last glacial maximum.

Figure 3.1 Analysis of retouched and unretouched pieces in the assemblage from layer VII of 60 Grotte Vaufrey, according.to the ‘chaine opératoire’ scheme of Geneste.

Figure 3.2 Refitted block of flakes from level VIII of the Grotte Vaufrey. 60

Figure 3.3 Stages of production of a ‘classic’ Levallois core. 62 Figure 3.4 ‘Classic’ Levallois core from the site of La Borde. 63

Figure 3.5 Schematic representation of the basic ‘Levallots concept’ as defined by Eric Boéda. 63 Figure 3.6 Examples of cores oriented towards the production of a single major Levallois flake 64

from the site of Bagorre (northern France).

Figure 3.7 Levallois points and similar forms from the later Mousterian levels of the Kebara cave 65 (Israel).

Figure 3.8 Levallois points and associated core forms. 66

Figure 3.9 Schematic representation of the ‘recurrent unipolar’ Levallois technique. 68

Figure 3.10 Illustrations of the ‘recurrent bipolar’ and ‘recurrent unipolar’ Levallois techniques as 69 represented at Biache-Saint-Vaast (northern France).

Figure 3.11 Examples of flakes produced by the ‘recurrent unipolar’ and ‘recurrent bipolar’ 70 techniques at Biache-Saint-Vaast.

Figure 3.12 Illustration of the ‘recurrent centripetal’ Levallois technique as represented at 71

Figure 3.13 Examples of ‘disc’ cores. 72 Figure 3.14 Diagram to illustrate the potential of flakes with varying cross-sections for repeated 74 Corbehem (northern France).

edge-resharpening.

Figure 3.15 ‘Salami slice’ technique of flake production. 75

Figure 3.16 More complex patterns of core reduction employed in Quina-Mousterian assemblages. 76 Figure 3.17 Map of Middle Palaeolithic sites with blade-dominated industries in northern Europe. 78 Figure 3.18 Examples of cores used to produce ‘Levallois blades’ from sites in northern France. 79

Figure 3.19 Schematic representation of Boéda's ‘specialized Levallois’ technique of blade 80 production.

Figure 3.20 Examples of blades from the site of Seclin (northern France). 81

Figure 3.21 Blade cores from Seclin. 82 Figure 3.22 Blades and blade cores from Riencourt-lés-Bapaume (northern France). 83 Figure 3.23 Refitted group of blades and flakes from St-Valéry-sur-Somme. 85 Figure 3.24 Two groups of refitted blades and flakes from St-Valéry-sur-Somme. 86 Figure 3.25 Differing ‘volumetric’ use of raw materials achieved by typical Levallois techniques 87 compared with two forms of blade techniques.

Figure 3.26 Relationship between the thickness and surface area of Levallois flakes in French 89 Middle Palaeolithic assemblages.

Figure 3.27 Metrical parameters of a succession of flakes produced during the experimental 91 replication of Levallots flaking techniques.

Figure 3.28 Metrical parameters of 41 refitted Levallois and other flakes from the site of 91 Maastricht-Belvédere (Holland).

Figure 4.1 Various side-scraper forms, classified according to the categories of Francois Bordes. 97

LIST OF ILLUSTRATIONS x1

Figure 4.2 Transverse racloir forms. 98 Figure 4.3 Resharpening flakes from the Cotte de St Brelade (Jersey). 100 Figure 4.4 Resharpening spall refitted to the edge of the original tool, from the Cotte de St Brelade 100 (Jersey).

Figure 4.5 Mean lengths of racloirs in levels of Ferrassie, Quina and Denticulate Mousterian at 102 Combe Grenal.

Figure 4.6 Schematic illustration to show how the application of retouch can increase the effective 103 length of the working edge in the production of a typical ‘racloir’ form.

Figure 4.7 Relationship between the relative frequencies of transverse versus lateral racloirs and 105 the variable utilization of Levallots flaking techniques in Ferrassie and Quina Mousterian industries.

Figure 4.8 Dibble’s hypothetical reconstruction of the transformation from lateral to transverse 106 racloir forms.

Figure 4.9 Variable intensity of edge retouch on different forms of racloirs in the Mousterian 107 assemblages from Bisitun, La Quina and Combe Grenal.

Figure 4.10 Maximum dimensions of lateral and transverse racloir forms in different levels of 108 OQuina Mousterian at Combe Grenal.

Figure 4.11 Relative frequencies of transverse and convergent racloir forms, and the ratios of 109 denticulates to notches, in different Mousterian variants.

Figure 4.12 Dibble’s hypothetical reconstruction of the transformation from single to convergent 110 racloir forms.

Figure 4.13 Examples of pointed forms from French Middle Palaeolithic sites. 111

Figure 4.14 Pointed forms with basal trimming of the bulbar surfaces. 114

Figure 4.15 Convergent racloir forms showing micro-wear traces of hafting. 115

Figure 4.16 Two models for the production of pointed forms. 116

Figure 4.17 Apparent impact fractures on Mousterian points from the Cotte de St Brelade 117 (Jersey).

Figure 4.18 Notched and denticulated tools from French Middle Palaeolithic sites. 118

Figure 4.19 Typical backed knives from MTA sites in southwestern France. 121 Figure 4.20 ‘End scrapers’ and ‘burins’ recorded in French Middle Palaeolithic industries. 123

Figure 4.21 ‘Micoquian’ type hand axes from the sites of La Micoque (France) and 125 Bocksteinschmiede (Germany).

Figure 4.22 Typical cordiform and related biface forms. 126

Figure 4.23 Length-over-breadth ratios of bifaces. 127

Figure 4.24 Breadth-over-thickness ratios of bifaces. 127 , Figure 4.25 Sharply triangular hand-axe forms, from surface contexts in western France. 128 Figure 4.26 ‘Bout-coupé’ hand-axe forms, from sites in southern England. 129 Figure 4.27 Bifacial leaf points from the later Mousterian levels of the Mauern cave (southern 130 Germany).

Figure 4.28 Flake-cleaver forms from the ‘Vasconian’ Mousterian levels of El Castillo, Cantabria. 131

Figure 4.29 Chopper/chopping-tool forms. 132 Figure 4.30 Comparison of backed-knife forms in early Upper Palaeolithic and Mousterian 134 industries.

Figure 4.31 Relative frequencies of racloirs versus notched and denticulated tools manufactured 137 from different raw materials.

Figure 4.32 Cumulative graphs of the tool assemblages manufactured from flint and quartzite in 139 the Quina-Mousterian assemblage from Mas-Viel (Lot).

Figure 5.1 Map of the major geological outcrops in the Perigord and adjacent areas of 142 southwestern France.

Figure 5.2 Cross-section of geological outcrops through the central Perigord area. 142 Figure 5.3 Distribution of geological sources sampled for raw material supplies in recent raw- 143

material provenancing studies in southwestern France. *

Figure 5.4 Distribution of distinctive varieties of flint in the Perigord and adjacent areas. 144 Figure 5.5 Integrated map of raw material sources exploited from Middle Palaeolithic sites in the 145 Perigord and adjacent areas.

Figure 5.6 Raw material sources exploited from the sites of La Plane and La Borde. 148

xil LIST OF ILLUSTRATIONS Figure 5.7 Raw material sources exploited from the Quina-Mousterian sites of La Chapelle-aux- 149 Saints and Mas-Viel.

Figure 5.8 Raw material sources exploited in different levels of the Grotte Vaufrey. 150 Figure 5.9 Frequencies of raw materials deriving from varying distances in the assemblages from 151 Grotte Vaufrey and Fonseigner.

Figure 5.10 Frequencies of raw materials deriving from different sources in the assemblages from 152 Grotte Vaufrey layer VII and Fonseigner layer D.

Figure 5.11 Patterns of utilization of raw materials in the assemblages from Grotte Vaufrey layer 154 VII and Fonseigner layer D.

Figure 5.12 ‘Intensity of utilization’ of raw materials in the assemblages from Fonseigner layer D 155 and Grotte Vaufrey layer VIII.

Figure 5.13 Map of Middle Palaeolithic sites studied by Geneste in the Euche valley. 156

Figure 5.14 Variable frequencies of cores recorded in different Middle Palaeolithic sites in the 156 Euche valley.

Figure 5.15 Relative frequencies of cortical flakes, non-cortical flakes and retouched tools 157 manufactured from two different raw materials at Marillac (Charente).

Figure 5.16 Frequencies of different categories of flaking debitage and cores in three assemblages 158 from southwestern France.

Figure 5.17 Three hypothetical models of raw material transport by Middle Palaeolithic groups. 162 Figure 5.18 Comparison of the distances over which raw materials were transported in Middle 164 Palaeolithic sites in southwestern France and central Europe.

Figure 5.19 Integrated map of the raw material sources exploited from Upper Palaeolithic sites in 166 the Perigord and adjacent areas.

Figure 5.20 Frequencies of raw materials deriving from varying distances recorded in Middle 167 Palaeolithic and early Upper Palaeolithic sites in southwestern France.

Figure 6.1 Overall distribution of side-scraper frequencies in southwestern French Mousterian 171 assemblages.

Figure 6.2 Racloir frequencies recorded in different industrial variants of the southwestern 172 French Mousterian.

Figure 6.3 Cumulative graphs of tool-type frequencies for three Mousterian variants. 173 Figure 6.4 Frequencies of backed knives recorded in MTA and other industrial variants. 173 Figure 6.5 Bifacially-worked ‘tranchotrs’ from the sites of Montgaudier and La Quina. 174 Figure 6.6 Bifacial ‘tranchoirs’ from the Quina-Mousterian assemblages of La Quina and 175 Hauteroche.

Figure 6.7 Double-pointed ‘limace’ forms. 176 Figure 6.8 ‘Micoquian’ type bifaces from the site of Klausenische (Germany). 177 Figure 6.9 Bifacial leaf points from the later Mousterian levels of the Mauern cave (southern 178 Germany).

Figure 6.10 Plan and edge views of racloirs with ‘Quina-type’ retouch. 178

Figure 6.11 Percentages of tools shaped by ‘Quina-type’ retouch in different industrial variants of — 179 the southwestern French Mousterian.

Figure 6.12 Relative frequencies of transverse racloir forms in Quina and other Mousterian 179 industries in southwestern France

Figure 6.13 Multi-dimensional-scaling analysis of 33 Mousterian assemblages. 180

Figure 6.14 Canonical variates analysis of 96 Mousterian assemblages. 182

Figure 6.15 Discriminant function analyses of paired groups of Mousterian industries. 182

Figure 6.16 Stratigraphic distribution of Ferrassie, Quina and MTA industries at Combe Grenal. 184 Figure 6.17 Levallois indices recorded in stratified sequences of Ferrassie and Quina Mousterian 185 assemblages.

Figure 6.18 Technological and typological features of stratified sequence of Ferrassie and Quina 186 Mousterian assemblages at Combe Grenal.

Figure 6.19 Thermoluminescence dates for the lower shelter at Le Moustier. 187 Figure 6.20 Comparison of chronologies proposed by Laville and Mellars for Mousterian 188 successions in southwestern France.

Figure 6.21 Frequencies of reindeer remains in association with Quina and MTA industries in 189 southwestern France.

LIST OF ILLUSTRATIONS xii Figure 6.22 Stratigraphic distribution of Denticulate and Typical Mousterian industries at Combe 190 Grenal.

Figure 7.1 Frequencies of the principal faunal taxa recorded in Mousterian faunal assemblages in 195 southwestern France.

Figure 7.2 Frequencies of the four principal faunal taxa recorded throughout the Mousterian 197 succession at Combe Grenal.

Figure 7.3 Frequencies of the dominant faunal taxa recorded in Mousterian faunal assemblages 200 from southwestern France.

Figure 7.4 Percentages of reindeer remains recorded in Mousterian and Upper Palaeolithic levels 201 in southwestern France.

Figure 7.5 Relative frequencies of skeletal elements of red deer and reindeer at Combe Grenal. 205 Figure 7.6 Kelative frequencies of the main skeletal elements of horses and bovids at Combe 207 Grenal.

Figure 7.7 Relative proportions of postcranial remains to teeth recorded for different species in 209 Mousterian faunal assemblages from southwestern France.

Figure 7.8 Frequencies of different skeletal elements of red deer in levels 50A-54 at Combe Grenal, 213 compared with the relative amounts of ‘soft tissue’ and marrow associated with the different bones.

Figure 7.9 Frequencies of different skeletal elements of reindeer in layers 23-25 and layers 20-22 214 at Combe Grenal, compared with the amounts of soft tissue and marrow volumes associated with the different bones.

Figure 7.10 Frequencies of different skeletal elements of reindeer in layers 23-25 and 20-22 at 215 Combe Grenal, compared with the amounts of soft tissue and percentages of oleic acid associated with the different bones.

Figure 7.11 Butchery marks on different parts of the skeleton of horse and bovids in Mousterian 218 faunal assemblages at Combe Grenal.

Figure 7.12 Age distribution of horses at Combe Grenal. 219

Figure 7.13 Contrasting age-mortality profiles for exploited animal populations. 223 Figure 7.14 Relative frequencies of skeletal elements in faunal assemblages accumulated by 224 , different types of carnivores.

Figure 7.15 Relative frequencies of horn and head parts to limb bones in faunal assemblages 225 accumulated by scavengers and hunters.

Figure 7.16 Mortality patterns of animals expressed as triangular scattergram plots. 226 Figure 7.17 Points of wooden spears from Clacton on Sea, England, and Lehringen, northern 227 Germany.

Figure 7.18 Schematic section of the Mousterian deposits at La Quina (Charente). 228 Figure 7.19 Accumulations of mammoth and rhinoceros bones in levels 3 and 6 at the Cotte de St 231 Brelade (Jersey).

Figure 7.20 Location of the bison-butchery site of Mauran (Haute-Garonne). 232

Figure 7.21 Detail of the bison-bone accumulation at Mauran. 233 Figure 7.22 Location of the La Borde site (Lot). 234 Figure 7.23 Cross-section through the deposits at La Borde. 235 Figure 7.24 Estimated age distribution of the remains of aurochs from La Borde and remains of 236 bison from Mauran.

Figure 7.25 Relative frequencies of different skeletal elements of bison at Mauran. 239 Figure 7.26 Relative frequencies of different skeletal elements of aurochs at La Borde. 239 Figure 7.27 Comparison of frequencies of different skeletal elements of bovids in the faunal 240 assemblages from Mauran, La Borde, and Combe Grenal.

Figure 7.28 Notched and denticulated tools associated with the bison-bone assemblage from 241 Mauran.

Figure 7.29 Flake tools associated with the aurochs-bone assemblage from La Borde. 242

Figure 7.30 Chopping tools of quartzite from Mauran. 243 Figure 7.31 Chopping tools of quartz from La Borde. 244

Figure 8.1 Distribution of Middle Palaeolithic cave and rock-shelter sites in the Perigord and 24.6

, adjacent areas of southwestern France.

Figure 8.2 Distribution of Upper Palaeolithic sites in the Perigord and adjacent areas. 248

XIV LIST OF ILLUSTRATIONS Figure 8.3 Solar orientation of Middle and Upper Palaeolithic cave and rock-shelter sites in the 249 northern Perigord region.

Figure 8.4 View of the Dordogne valley, close to the site of Combe Grenal. 251

Figure 8.5 Distribution of Middle Palaeolithic open-air and cave/rock-shelter sites in the northern 255 Perigord region.

Figure 8.6 Comparison of the distribution of Middle and Upper Palaeolithic sites in the Dronne 256 — valley region.

Figure 8.7 Occurrence of Middle Palaeolithic sites in relation to sources of lithic raw materials. 257 Figure 8.8 Principal geological outcrops in the areas between the Dordogne and Lot valleys. 258 Figure 8.9 Geological distribution of flint sources in the areas between the Dordogne and Lot 259 valleys.

Figure 8.10 Overall distribution of Middle Palaeolithic sites adjacent to the Dordogne and Lot 260 valleys.

Figure 8.11 Distribution of finds of cordiform and related hand-axe forms in southwestern 261 France.

Figure 8.12 Distribution of MTA industries between the Dordogne and Lot valleys. 262 Figure 8.13 Distribution of Quina Mousterian industries between the Dordogne and Lot valleys. 263 Figure 8.14 Distribution of surface finds of Acheulian, Mousterian, Upper Palaeolithic and 265

Neolithic artefacts at the site of La Croix-Guémard (Deux-Sedvres). | Figure 9.1 Plan and section of the Grotte Vaufrey. 270 Figure 9.2 Overall distribution of lithic artefacts in layer VIII of Grotte Vaufrey. 272

Figure 9.3 Distribution of refitted artefacts in layer VIII of Grotte Vaufrey. 273 Figure 9.4 Distribution of Levallois flakes in layer VIII of Grotte Vaufrey. 274

Figure 9.5 Distribution of unworked river cobbles in layer VIII of Grotte Vaufrey. 275

Figure 9.6 Distribution of racloirs in layer VIII of Grotte Vaufrey. 276

Figure 9.7 Distribution of notches and denticulates in layer VIII of Grotte Vaufrey. 277 Figure 9.8 Overall distribution of faunal remains in layer VIL of Grotte Vaufrey. 278

Figure 9.9 Distribution of small bone splinters in layer VIII of Grotte Vaufrey. 279 Figure 9.10 Distribution of traces of burning in layer VIII of Grotte Vaufrey. 280

Figure 9.11 Plan of the Grotte du Lazaret. 281 Figure 9.12 Distribution of lithic artefacts in the Grotte du Lazaret. 282

Figure 9.13 Distribution of burning traces in the Grotte du Lazaret. 283 Figure 9.14 Distribution of small bone splinters in the Grotte du Lazaret. 284

Figure 9.15 Distribution of small sea shells and foot bones of fur-bearing animals in the Grotte du 285 Lazaret.

Figure 9.16 Reconstruction of the probable organization of space in the Lazaret cave. 286

Figure 9.17 Reconstruction of the hypothetical hut structure in the Lazaret cave. 287

Figure 9.18 Plan of the excavated area in the Les Canalettes rock shelter. 288 Figure 9.19 Distribution of flint flakes and quartz artefacts in layer 2 of Les Canalettes. 289

Figure 9.20 Distribution of retouched tools and cores in layer 2 of Les Canalettes. 290

Figure 9.21 Distribution of faunal remains in layer 2 of Les Canalettes. 291

Figure 9.22 Three forms of hearths in the Acheulian levels at Pech de Il’ Azé II (Dordogne). 297 Figure 9.23 Possible hearth in the Denticulate Mousterian level of the Hauteroche rock shelter. 298 Figure 9.24 Circular hearth, associated with burned stones, recorded in the later Mousterian levels 300 of the Grotte du Bison at Arcy-sur-Cure (Yonne).

Figure 9.25 Sequence of stratified hearths at the early last glacial site of Saint-Vaast-la-Hougue on 301 the Normandy coast.

Figure 9.26 Area of cobble paving recorded in the ‘Rissian’ levels of the Baume-Bonne cave (Basses- 302 Alpes).

Figure 9.27 Claimed remnants of stone walling recorded in the later Mousterian levels of Cueva 304 Morin (Cantabria).

Figure 9.28 Cross-section and plan of the pit in the earlier Mousterian levels at Combe Grenal. 306

Figure 9.29 Two adjacent pit features in the uppermost Mousterian level at Le Moustier. 307

Figure 9.30 Cast of post-hole in layer 14 at Combe Grenal. 308

Figure 9.31 Plan of hut structures in the Chatelperronian levels of the Grotte du Renne Cave at 312 Arcy-sur-Cure (central France).

LIST OF [ILLUSTRATIONS XV

Brune (eastern France). ,

Figure 9.32 Circular hut structure, with a centrally located hearth, at the Gravettian site of Vigne- 314 Figure 10.1 Frequencies of identifiable animal remains per 100 retouched stone tools in association 330 with different Mousterian variants at Combe Grenal.

Figure 10.2 Frequencies of main faunal species associated with levels of Denticulate and Quina 330 Mousterian in levels 22—11 at Combe Grenal.

Figure 10.3 Summary of Dibble’s model for the progressive reduction of unretouched flakes into 333 various racloir forms.

Figure 10.4 Dibble & Rolland’s hypothetical model for the transformation of a denticulate- 334 dominated industry into one dominated by racloir forms.

Figure 10.5 Rolland’s model of how various environmental and other factors can influence the 335 degree of tool reduction in Middle Palaeolithic sites.

Figure 10.6 Associations between Mousterian variants and climatic regimes in western 339 Europe.

Figure 12.1 Increase in hominid cranial capacities over the past three million years. 367 Figure 12.2 Fragments of manganese dioxide apparently used as pigments, from the MTA levels of — 370 Pech de l’Azé I.

Figure 12.3 Fossils of sea shells and other marine organisms from the Mousterian levels of Chez- 371 Pourrez and the Grotte de l’Hyene.

Figure 12.4 Incised bones from Lower and Middle Palaeolithic contexts in Europe. 372.

Figure 12.5 Perforated bones and teeth from Middle Palaeolithic sites. 373

Figure 12.6 Fossil nummulite from the Mousterian levels of Tata (Hungary), with incised cross. 374 Figure 12.7 Section and plan of the Neanderthal grave at La Chapelle-aux-Saints (Corréze). 376 Figure 12.8 Upper: Plan of the six Neanderthal burials in the Mousterian levels at La Ferrassie. 377 Lower: Plan and section of burial no. 6 from La Ferrassie.

Figure 12.9 Photograph of the skeleton of Neanderthal burial no. 2 at La Ferrassie. 378 Figure 12.10 Supposedly ‘ritual’ placement of the Neanderthal skull from the Grotta Guattari 379 (Monte Circeo, western Italy).

Figure 12.11 Burial of one of the anatomically modern hominids from Djebel Qafzeh (Israel) 380 associated with a large pair of fallow deer antlers.

Figure 13.1 Aurignacian stone and bone artefacts from sites in western France. 394 Figure 13.2 Bone and ivory artefacts from early Aurignacian levels in Europe. 395

Figure 13.3 Perforated deer tooth from the early Aurignacian levels of La Souquette (southwestern 396 France),

Figure 13.4 Animal figurines of mammoth ivory from the early Aurignacian levels at Vogelherd 397 | (southern Germany).

Figure 13.5 Lion-headed human figure of mammoth tvory from the early Aurignacian levels in the 398 Hohlenstein-Stadel cave (southern Germany).

Figure 13.6 Upper: Perforated sea shells, from the early Aurignacian levels of La Souquette, 399

southern France. | Figure 13.7 Skull of anatomically modern form from Djebel Qafzeh (Israel). 402 southwestern France. Lower: sources of sea shells found in Aurignacian levels in

Figure 13.8 Left: Late Neanderthal skull from the Chatelperronian levels at Saint-Césaire 403 (southwestern France). Right: anatomically modern skull from the early Aurignacian levels at Vogelherd (Stetten, southern Germany).

Figure 13.9 Suggested chronology of the principal Neanderthal remains from southwestern 404 France.

Figure 13.10 Geographical distribution of Aurignacian and other early Upper Palaeolithic 406 industries in Europe and the Middle East.

Figure 13.11 Aurignacian stone tools from the Hayonim cave (Israel). 407 Figure 13.12 Bone artefacts and animal-tooth pendants from the Aurignacian levels of the Hayonim 408 cave.

Figure 13.13 Absolute age measurements for early Aurignacian and Chatelperronian industries in 409 Europe.

Figure 13.14 Chatelperronian stone tools from the Grotte du Renne (Arcy-sur-Cure, central 413 France).

xvi LIST OF ILLUSTRATIONS Figure 13.15 Bone artefacts and animal-tooth pendants from the Chatelperronian levels at Arcy-sur- 415 Cure.

Figure 13.16 Bifacial leaf points and associated tools from the Szeletian site of Vedrovice V 417 (Czechoslovakia).

Preface My colleague Chris Stringer warned me that and (to a lesser extent) faunal assemblages. writing a book about Neanderthals would be My central belief in embarking on this book rather like ordering menus for College High was that it was only by taking a very broadTable dinners — there would be no way of ranging look at all aspects of the behaviour pleasing everyone. For the past eighty years and organization of Neanderthal communior so we seem to have been caught between ties — their technology, subsistence patterns, two opposing camps — between those who spatial and demographic organization and

see the Neanderthals as our immediate patterns of communication and cognition — ancestors, only mildly different in at least that we could hope to form any overall their basic biological and behavioural pat- impression of their true behavioural capaciterns from fully ‘modern’ populations; and _ ties and, in particular, how far these may those who prefer to adhere to Marcellin have differed from those of the ensuing anaBoule’s original vision of the Neanderthals as tomically modern populations. As far as lam

representing an extinct side line of human aware, this is the first book which has evolution, with both behavioural patterns attempted to bring all these issues together in

and most probably innate capacities for a single study, focussed on a specific and behaviour which were radically different exceptionally rich and well documented from those of later populations. The recent body of archaeological evidence. controversies surrounding the origins and The extent to which this book depends on dispersal of anatomically modern popula- the research of my French colleagues will be tions have put the spotlight firmly onto the immediately apparent from dipping into any Neanderthals as arguably the last surviving of the individual chapters. Following the representatives of the original ‘archaic’ line- pioneering research of Francois Bordes, ages which populated the world prior to the André Leroi-Gourhan and others shortly dramatic demographic and behavioural dis- after the Second World War, there has been

persal of genetically modern populations. an extraordinary upsurge in studies of the ,

While there have been many studies of the Middle Palaeolithic over the past twenty biological and anatomical aspects of Nean- years, extending far beyond the traditional derthals, there have been surprisingly few obsession with typological and taxonomic systematic attempts to review their behav- issues which dominated the field for so long. ioural patterns. There has of course been no In southwestern France the work pioneered shortage of excavations of Middle Palaeo- by Jean-Philippe Rigaud on the spatial dislithic sites, combined with a deluge of analy- tribution and organization of Mousterian ses of the associated stone-tool industries sites has been systematically extended by a Xvil

xVill PREFACE number of younger colleagues— most notably behavioural patterns in the three decades or by Jean-Michel Geneste, Allain Turq and _ so since I wrote the original thesis.

Christine Duchadeau-Kervazo in the Peri- It is hard to exaggerate the debt [owe to my gord region, and further south by Jacques French colleagues in sharing their ideas with

Jaubert, Jean-Marie Le Tensorer, Liliane me, providing information on the latest Meignen and others. Over the same period results of their fieldwork and laboratory anathe research of Francoise Delpech, Jean-Luc lyses and allowing me to reproduce data and Guadelli, Guy Laquay, Stephane Madelaine, illustrations from their own publications. Philip Chase and others has taken the analy- Jean-Philippe Rigaud in particular has been a sis of Middle Palaeolithic faunal assemblages tower of support, as he has been to many far beyond the traditional concern with the other British and American workers involved purely climatic and chronological aspects of in the study of French prehistory. For similar

the faunas, into a determined attempt to help and cooperation in a variety of ways I reconstruct the economic and carcase-pro- am equally indebted to Eric Boéda, Jeancessing strategies which lay behind the Pierre Chadelle, Jean-Jacques Cleyet-Merle, : archaelogical bone assemblages. Above all, André Debénath, Francoise Delpech, Pierre' perhaps, the study of lithic technology has Yves Demars, Christine Duchadeau-Kervazo, been revolutionized by the application of the Catherine Farizy, Jean-Michel Geneste, Jean‘chaine opératoire’ approach to technological Luc Guadelli, Jacques Jaubert, Guy Laquay, analysis (involving the use of extensive refit- Henri Laville, Michel Lenoir, Francois Lévéting studies, experimental replication tech- que, Stephane Madelaine, Liliane Meignen,

niques and microscopic use-wear analyses), Jacques Pelegrin, Denise de Sonnevilleas well as by systematic studies of the sources Bordes, Allain Tuffreau, Allain Turg and Berand distribution patterns of the raw materials nard Vandermeersch. Without the coopera-

employed for tool production. tion of these and other French colleagues the

What is remarkable about this recent work book could never have been written. The is not merely its scientific quality, but the same applies, from a slightly different perextent to which much of the critical dataand spective, to the help I have received from a analysis still remains embedded inarangeof number of north American colleagues specialist site reports, conference volumes involved in closely related studies of the and (above all perhaps) privately circulated European Middle Palaeolithic — above all to

and only partially published doctoral dis- Harold Dibble, Nicholas Rolland, Philip sertations, which are largely inaccessible to Chase, Art Jelinek and Lewis Binford. Some the majority of non-French workers. One of of them may feel that their interpretations the primary aims of this book has been sim- have received rather rough justice in some ply to bring together this extraordinary store sections of the book, but I hope they will see of recently acquired information and to make _ the close attention I have paid to their work it more readily available to a wider audience. above all as a genuine mark of respect for

Beyond this I have of course added my own _ their contributions. Lewis Binford in partic- , material and analyses, gleaned in over thirty ular, I believe, has been the most positive and years of research into Middle Palaeolithic inspiring influence on European Palaeolithic

problems, and extending back to the PhD studies over the past 30 years, and I am thesis I originally wrote at Cambridge (under greatly indebted for his generosity not only the supervision of Charles McBurney) in the in discussing his ideas with me at length

early 1960s. In one sense, this book is a (most notably during a memorable visit to personal odyssey to discover exactly what 1 Albuquerque in the summer of 1992) but also

believe has been learnt about Neanderthal in allowing me to quote from his extremely

PREFACE x1X important unpublished research at Combe Finally, I owe a special debt of gratitude to

Grenal. my wife. Anyone who embarks on writing a The preparation of the book has benefited lengthy book knows that this demands end-

greatly from the encouragement and guid- less tolerance for lost weekends, late nights, , ance of Bill Woodcock and Emily Wilkinson bouts of exasperation and no doubt other , of Princeton University Press, and the artistic strains — backed up by endless promises that skills of John Rodford, who drew or redrew _ the end is just around the corner. As ever, she

well over half the illustrations in the book. bore this with tolerance and good humour, My colleagues in the Archaeology Depart- and contributed greatly and in many ways to ment at Cambridge have provided various _ the final product.

forms of intellectual stimulation ~- and Financial support towards the preparation restraint — over the past 15 years, while a of the book was provided mainly by the succession of exceptionally bright and enthu-_ British Academy. In addition to travel grants , siastic undergraduate and graduate students for visits to France in 1990, 1991, 1992, and have done even more to stimulate and chan- 1993, the Academy provided the initial stim-

nel my thinking along new lines. The con- ulus and support for the work which lay tributions of Nathan Schlanger, Paul Pettitt behind the book, through the award of a twoand Gilliane Monnier in particular will be year Research Readership from 1989 to 1991. apparent from the discussions in Chapters 3 For similar support ~ and for providing the and 4. The great inspiration throughout my ideal geographical and intellectual environown studies at Cambridge was provided by ment in which to write a book —Iam indebted Professor (now Sir) Grahame Clark, anditisa to the Master and Fellows of Corpus Christi

pleasure to acknowledge the debt | owe to College. him. Two other colleagues at Cambridge —

Nick Shackleton and Tjeerd Van Andel -

were particularly helpful in reading through Paul Mellars

drafts of the environmental chapter and Cambridge

offering valuable comments. July 1995

e

To my wife and parents

Introduction = CHAPTER 1

The Neanderthals have always been some- logically modern populations in Eurasia, and thing of an enigma. Since their initial discov- indeed that the Neanderthals as a whole ery in the middle of the last century opinions might well represent a separate biological

have tended to polarize between two species (Cann et al. 1987, 1994; Stoneking & extremes: between those who saw the Nean- Cann 1989; Stoneking et al. 1992; Stringer & derthals as standing directly astride the main Gamble 1993 etc.). Similarly, recent dating of

course of human evolution, only slightly dif- a range of essentially modern anatomical ferent in either their physical or mental capa-~ remains at the sites of Skhul and Qafzeh in bilities from modern populations; and those Israel, and at a number of sites in southern who saw them, by contrast, as much more’ Africa, has shown that forms closely similar

primitive figures, with behavioural and_ to ourselves had already emerged in several physical capacities radically different from parts of the world long before their appearthose of later populations and almost cer- ance, in a remarkably sudden and abrupt tainly representing an extinct side branch of form, in the more western zones of Eurasia human evolution. According to one view- (Stringer & Andrews 1988; Stringer 1990, point the Neanderthals were our direct 1992, 1994; Stringer & Gamble 1993; Brauer ancestors, while according to the other they 1989; Vandermeersch 1989). These concluwere rather distant, and not very respectable, sions have been contested by proponents of

cousins. A spate of characterizations in the ‘regional continuity’ or ‘multi-regional media cartoons, as well as more thoughtful evolution’ school, who argue that the entire presentations in popular novels (such as Wil- framework of both genetic and anatomical liam Golding’s The Inheritors, and Jean Auel’s evidence which has been used to support the

The Clan of the Cave Bear) have served to demographic extinction of the Neanderthals enhance the mystique and uncertainty sur- is based on rank misrepresentations of the rounding the true role of the Neanderthals in biological evidence, or at best on serious

our own evolution. ambiguities in the interpretation of this evi-

Research and discoveries over the past ten dence (Wolpoff 1989, 1992; Wolpoff ef al. years have tended to heighten rather than 1994; Thorne & Wolpoff 1992; Smith 1991, reduce these long-standing controversies 1994). According to them European readers over the place of the Neanderthals inhuman of this book are far more likely to have a evolution. Recent research in molecular strong component of Neanderthal genes in genetics has been interpreted to suggest that their direct ancestry than genes of a hypothe Neanderthals may have made no direct thetical intrusive modern population, from contribution to the genetic ancestry of bio- some exotic African or Asian source. 1

2 THE NEANDERTHAL LEGACY Similar debates have plagued recent inter- the Neanderthal populations of Europe with pretations of the archaeological records of the ensuing populations of anatomically and Neanderthal behaviour. To many prehistor- behaviourally modern humans, a transition ians the archaeological records of the Nean- which seems to have taken place in most

derthals suggest a pattern of behaviour regions of Europe between ca 40,000 and which is not only radically and fundamen- 35,000 years ago. Specifically, the major tally different from that of the ensuing bio- issues in this context can be reduced to three

logically modern populations but which critical questions: indicates a fundamentally dit ferent structure 1 To what extent, if at all, did the Nean-

of. derthals mind. Recent characterizations in this ;; contribute to the genetic ancestry

vein have suggested that the Neanderthals of later populations in Eurove?

may have been incapable of hunting most of PoP Pe

the larger species of animals; that they 2. How far, and in what ways, did the behavformed social groupings which were more iour of Neanderthal populations contrast akin to the sexually segregated foraging units with that of the ensuing anatomically and of most primate communities than the fam- behaviourally modern populations?

ily-based structure of modern human pop. , —_ 3. If we can document major contrasts ulations; that theyaelacked the forpatterns of Neanbetween thecapacity behavioural long-range planning or organization of their how . ve derthal andactivities; modern populations, economic and social and that ; Do . should these contrasts be they explained?

almost certainly lacked complex, highly ;; - they reflect simply a gradual, progressive

structured language (Binford 1989; Lieber- increase in the overall complexity of difman 1989; Chase &; Dibble 1987; Soffer 1994; , PL. . ferent behavioural systems over the course

Stringer & Gamble 1993). Opposing this view of time? Or do they represent a much more

are those who see the general behaviour and ie,

vs sudden andNeanderthals radical shift in behavioural cultural capacities of the as patterns, which reflects proonly marginally different from those .an equally foundof shiftlater in theLey associated mental and populations, with the. cognitive exceptioncapacities of a few postve: vs ; of the for behaviour Neanderthal embellishments in the form of ; ; populations involved? representational art, more complex forms of

bone and antler technology and a general These questions, addressed primarily to predilection for manufacturing stone tools the archaeological records from western from more elongated and economical blade Europe, form the central focus and subject forms in preference to larger and heavier matter of the present book. flakes (Clark & Lindly 1989; Lindly & Clark 1990; Clark 1992; Hayden 1993). The latter

developments, it is argued, are more likely to Who were the Neanderthals?

represent a gradual, cumulative increase in

the overall complexity of behavioural pat- I shall make no attempt to discuss in any terns over the course of later human evolu- detail here the biological and anatomical fea-

tion than a radical transformation in the tures of the Neanderthals since this has underlying intellectual and cultural capabil- already been dealt with comprehensively in

ities of the populations involved. two recent books ~ In Search of the Nean-

So what exactly are the central issues in derthals by Chris Stringer and Clive Gamble current studies of the Neanderthals? The (1993), and The Neanderthals: Changing the question which lies at the heart of the present Image of Mankind by Erik Trinkaus and Pat debate centres on the precise relationships of Shipman (1993). Both studies seem to agree

INTRODUCTION 3 that, in at least their major anatomical fea- the habitual use of jaws and teeth for various tures, the Neanderthals form a reasonably ‘paramasticatory’ activities (i.e. using the distinctive and fairly well defined taxonomic jaws as tools) than to any cold-climatic adapgrouping, even if the precise geographical tation (Smith & Paquette 1989). and chronological limits of the grouping are To put exact limits on the Neanderthals ina more difficult to define. As in other fossil time and space framework is more difficult. hominids, the most distinctive features of However, there seems to be reasonable agreeNeanderthal morphology are reflected inthe ment that most of the distinctive features of skull and facial regions. In the case of the Neanderthal anatomy can be traced across a

Neanderthals these include heavily enlarged broad arc of Europe and western Asia, supraorbital brow ridges, a generally low extending from the Atlantic coasts of France and flattened cranial vault with a strongly and the Iberian peninsula to the western developed occipital ‘bun’, the heavily built parts of the Middle East and central Asia — for

structure of the jaws and teeth, with little example at Tabun, Amud and Kebara in trace of a chin, and a surprisingly large cra- Israel, at Shanidar in Iraq, and as far eastnial capacity of around 1400-1600 cc imply- wards as Teshik Tash in Uzbekistan (Smith ing an overall brain volume at least as large 1991; Stringer & Gamble 1993; Trinkaus & as that of modern populations. There seems Shipman 1993). Whether anything distincto be equal agreement that at least some of tively Neanderthal can be identified to the these distinctive features of Neanderthal south of this zone (for example, some of the morphology can be seen as an adaptation to North African fossils such as those from Jebel

the specific environmental conditions of the Irhoud and _ Dar-es-Soltan in Morocco) more northern zones of Eurasia during the remains more controversial. The main geocolder, glacial and sub-glacial episodes of the graphical range corresponds, in other words,

later Pleistocene. Thus the large noses and to the more western zones of Eurasia, and the generally inflated form of the facial predominantly to those areas which experiregion as a whole are often seen as anadapta- enced recurrent episodes of sharply colder tion to accommodate the very large nasal climate during the middle and later stages of channels that were essential to warm the _ the Pleistocene. cold, dry air of these exceptionally harsh In a chronological sense, most of the well climates (Howell 1957; Coon 1962; Wolpoff dated and ‘classic’ Neanderthal forms belong 1980 but see Trinkaus 1989b for a different to the earlier stages of the last glacial period, view). Similarly, the generally short, heavy between ca 110,000 and 35,000 BP. Both Trinbody structure typical of the Neanderthals is kaus & Shipman (1993) and Stringer & Gamusually seen not only as an adaptation to a_ ble (1993), however, have argued that many

very active and strenuous life style which distinctive Neanderthal traits can be traced demanded considerable physical strength, back into the period of the penultimate glabut also as an adaptation to conserve body cial and perhaps, as for example in the heat in severe, seasonally fluctuating cli- remains from Biache-Saint-Vaast in northern mates (Trinkaus 1983, 1989b; Trinkaus & France, to the period of isotope stage 7, Shipman 1993; Smith 1991). Whether all the around 200—250,000 BP. Earlier forms such as

distinctive features of Neanderthal morphol- the hominids from Swanscombe in England, ogy can be explained in these terms is more Steinheim in Germany, Petralona in Greece controversial. Smith and others, for example, and Tautavel in France tend to be regarded as have suggested that the large and heavily ‘pre’ or ‘proto’ Neanderthal forms, anatombuilt form of the Neanderthal face may have ically transitional between the late Homo been related more to the stresses involved in erectus/Homo heidelbergensis populations of

4 THE NEANDERTHAL LEGACY Europe and the succeeding Neanderthal in several parts of the Old World by the time

forms (Stringer et al. 1984; Stringer & of the penultimate interglacial period Andrews 1988; Stringer & Gamble 1993). between ca 200,000 and 250,000 BP (for examClearly, if one accepts that the Neanderthals ple at Biache-Saint-Vaast and the Grotte Vauare indeed the direct descendants of the ear- frey in France and Maastricht-Belvédére in

, lier erectus populations in Europe it would be Holland) and probably at roughly the same unrealistic to expect to recognize a sharp line time at a range of sites in western Asia and of demarcation between these taxa in ana- Africa (see Chapter 4, and Klein 1989a; Bar-

tomical or evolutionary terms. Yosef 1992). Whether it is entirely coinciSimilar problems are encountered in dental that these complex stone-working

attempts to define the precise chronological techniques appeared in Europe at roughly limits of what conventionally has been’ the same time as the rapid increase in cranial defined as the ‘Middle Palaeolithic’ phase of capacities which is one of the most distinctive

technological development. It should be features of the Neanderthals, is an interesting stressed that it is not the aim of this book to point for speculation. look at the problem of the technological ori- The chronological scope of this book theregins or emergence of Middle Palaeolithic fore coincides essentially with the period technology from the earlier patterns of Lower from the final disappearance of the NeanPalaeolithic technology. Any serious study of derthals in western Europe around 30-35,000

this question would not only require a book years ago to the period of oxygen-isotope in its own right but would be seriously handi- stage 7 around 250,000 BP. The surviving capped by the extremely patchy, coarse- archaeological records of human behaviour

grained and above all very poorly dated within this time range are very unevenly records of human technological development distributed and are far more abundant, more prior to the last 250,000 years. As a working fully documented and more chronologically definition lam happy to conform to what has _ fine-grained during the later stages of the

now become the conventional practice of Middle Palaeolithic than during its earlier regarding the prime hallmark of Middle Pal- stages. The book should therefore be seen aeolithic technology as being the emergence primarily as a study of the archaeological of more complex and sophisticated patterns evidence of human behaviour during the earof prepared-core flaking, classically illus- ler stages of the last glacial period, between trated by the various Levallois and allied ca 115,0000 and 35,000 years ago. It is this techniques discussed in Chapter 2. Asseveral period for which I will reserve the term

workers have recently stressed, the emer- ‘Mousterian’ throughout the book. In my gence of these techniques could be seen as a__ view it is only for this period that we have a

major turning point not only in a purely sufficient quantity of well documented evitechnological sense but also as a potential dence, and sufficient control over the quality watershed in the whole conceptual and cog- and resolution of this evidence, to present nitive basis of lithic technology, implying a_ any really secure and convincing reconstruc-

much greater degree of forward planning, tions of human behavioural patterns within time depth, and capacity for strategic prob- the Neanderthal time range. lem-solving in the working of stone resources

(Roebroeks et al. 1988; Rolland 1990; Mellars Th h I cal ti 1991; Klein 1989a). Whether or not this € ATCHAECOLOSICAL Perspeci1ve viewpoint is accepted, it is now clear that It soon became clear in planning this book remarkably complex and varied forms of pre- that to attempt a general survey of the pared-core techniques were being practised archaeological evidence for Neanderthal

INTRODUCTION 5 behaviour across the whole geographical correlations between the archaeological and range discussed above — i.e. extending from skeletal records. While we can never be sure

the Atlantic coast of Europe to the Middle that every Middle Palaeolithic industry in East — would not only be a daunting task in western Europe was produced by a Neanderterms of the amount of material and data to thal, we can be far more confident in making

be considered, but could become a rather this correlation in this region than in any questionable exercise. There are two princi- other part of Eurasia. pal reasons for this. First, there is the problem My second reason for choosing a more of knowing exactly where in the archaeo- restricted geographical focus for this book is logical records of western Eurasia we see the more pragmatic and relates to the way in

products of Neanderthal populations. One of which we approach the analysis and inthe most significant facts to emerge during terpretation of the archaeological evidence. the last decade is that we can no longer The importance of adopting a specifically assume that all archaeological assemblages regional approach in this context is now conventionally classified as ‘Middle Palaeo- widely recognized (e.g. Gamble 1984, 1986).

lithic’ in a purely technological sense were It would make little sense, for example, to indeed the products of Neanderthals, or make direct comparisons between the pat-

indeed other archaic forms of hominids. The terns of animal exploitation or the relative , discoveries at both Skhul and Qafzehin Israel frequencies of different species of animals and possibly at Staroselje in the Crimea have exploited, even in two areas as geographdemonstrated that in certain contexts techno- ically close as southwestern France and logically Middle Palaeolithic industries were northern Spain, where the overall patterns of

produced by hominids who in most anatom- climate, vegetation, topography etc. are ical respects were much closer to biologically likely to have been significantly different. modern populations than to Neanderthals - The same applies with equal force to comparand at the remarkably early date of around isons between, say, the relative use of cave

100,000 BP (Vandermeersch 1989; Bar-Yosef versus open-air sites in different areas, the | 1992; Stringer & Gamble 1993), Even in parts overall patterns of settlement and mobility

of central and eastern Europe there is still of the human groups, or the patterns of debate as to whether some of the skeletal procurement and distribution of different remains recovered from Middle Palaeolithic raw materials (Féblot-Augustins 1993). Only

contexts (such as Krapina in Croatia, or by adopting a regional approach to these Kulna in Moravia) can be confidently attrib- issues can we hope to provide a coherent uted to Neanderthals as opposed to anatom- reconstruction of the detailed patterns of ically modern populations (Smith 1984, behavioural adaptation of Neanderthal com1991). These observations raise critical ques- munities to the rapidly oscillating climatic tions concerning the ultimate relationships and ecological conditions of the later Pleistobetween Neanderthal and anatomically mod- cene. The choice in the present context was ern populations in these areas, which willbe therefore fairly stark: either to focus the prespursued further in the final chapter of this ent study primarily on the evidence from one book. The fact remains, however, that it is specific and well documented region; or to

, only in the extreme western zones of Europe range more widely, and inevitably more that we have a sufficiently large, well docu- superficially, over the evidence from many mented and consistent association between different regions and run the risk of failing to technologically “Middle Palaeolithic’ indus- deal adequately with the situation in any one

tries and taxonomically ‘Neanderthal’ region.

remains to make any reasonably confident When viewed in these terms the rationale

6 THE NEANDERTHAL LEGACY for focusing primarily on the evidence from rich and well studied regions of western western Europe, and largely on the evidence Europe cover an impressive span of time and

from the so-called ‘classic’ region of south- reflect the behaviour and adaptations of western France, is self-evident. This is not Neanderthal populations in sharply contrastmerely the classic region for Middle Palaeo- ing climatic and environmental conditions. If lithic studies in a purely historical sense (e.g. we can discern any general patterns or reg-

Lartet & Christy 1864; de Mortillet 1869, ularities in behavioural patterns over this 1883) but it has also produced a wealth and_ time range, these should not be dismissed concentration of hard archaeological evi- lightly. Wherever possible and appropriate dence for Middle Palaeolithic behavioural in the following chapters I have attempted to patterns which, by any criteria, is much make specific comparisons with the evidence richer, more detailed and better documented from other regions of Europe, especially than that from any other area of comparable where these seem to hint at a pattern of size in either Europe or Asia. From south- behaviour significantly different from that western France alone we now have a total of reflected in the western European data. Overover fifty cave and rock shelter sites with all, however, the arguments for focusing any relatively substantial and well documented study of this kind primarily on the evidence evidence for Middle Palaeolithic occupation, from one specific and well studied region are

in many cases in the form of long and _ largely self evident and I make no apology detailed occupation sequences. The majority for choosing the region. where the available

of these sites have produced rich and well archaeological records of Neanderthal preserved faunal assemblages, and from behaviour are exceptionally abundant and many of them we now have detailed informa- well documented. Given the specific theorettion not only on the technological features of ical orientation of this study, there is also

the stone-tool assemblages but also on the something to be said for choosing an area character and geological sources of the raw _ where archaeological records for the behavmaterials from which they were made (Gen-_ iour of the succeeding anatomically modern este 1985, 1989a; Turg 1989a; 1992b). All this populations are equally rich and clearly

research has been combined with a metic- defined. , ulous concern with the geological and climatic associations of the human occupations,

reflected ina wealth of published data on the Theoretical perspectives sedimentological, palaeobotanical and fau-

nal associations of the different sites. My aim in this book is not to adopt any The point of these observations, needless to specific a priori theoretical stance towards the

say, isnot toimply that the patterns of Nean- analysis and interpretation of the archaeoderthal behaviour in the extreme western logical record beyond what I would describe zones of Europe are inherently more impor- asasimple ‘rationalist’ one —i.e. to focus ona tant or more interesting than those in other number of specific, clearly defined questions regions of Eurasia, still less to suggest that relating to particular aspects of Neanderthal

any patterns or regularities observed in this behaviour and to evaluate a range of alterregion can be extrapolated automatically and native answers to these questions from uncritically to other regions. Whether such whichever aspects of the archaeological extrapolations can be made must remain one record seem most directly relevant. There of the prime targets of future research. The are, nevertheless, a number of themes which point to be emphasized is that the archaeo- run through many of the chapters and which logical and behavioural records from these should be recognized at the outset.

INTRODUCTION 7 First, it is clear that much of the folowing Palaeolithic communities but that it virtually discussion will be seen in many quarters as_ ignores the entire dimension of socially con-

an explicit reaction against some of the more structed behaviour and learning processes extreme ‘hyperfunctionalist’ approaches which must inevitably have shaped the which have dominated much of the literature behavioural patterns of all communities in on the Middle Palaeolithic over the past two the past, extending back to our primate oridecades - as I would see it largely as an_ gins (e.g. Mithen 1994; Mellars 1992a, 1994; overzealous extrapolation of the more mod- Bar-Yosef & Meignen 1992; McGrew 1993). est forms of functionalism which character- These and related issues are discussed more ized the ‘processualist’ archaeology of the fully in Chapter 10.

1960s and 1970s. As a product and great The second point is to some extent interadmirer of many aspects of the New Archae- related with the first and concerns the imology of the mid-1960s, I regard myself, in portance of viewing the organization and many respects, as a committed processualist, behaviour of past human communities in the sense that I believe that a close analysis within a demographic perspective, that is, in of the potential functional interrelationships terms of networks of communities linked by between the different aspects of behavioural varying degrees of social interraction. This

patterns represents the obvious starting approach assumes in other words _ that point for any productive analysis of hu- human populations do not form uniform, man behaviour in the past. I also admire the homogeneous networks extending indefi-

fact that functionalist or processualist ap- nitely across the landscape, but that they are proaches place the primary emphasis on divided into smaller and more restricted those aspects of human behaviour and adap- social and breeding units which interact in tation which leave the most direct traces in specific and often closely prescribed ways the archaeological record — such as technol- (Wobst 1974, 1976; Dunbar 1987; Foley & Lee

ogy, subsistence, settlement patterns, 1989; Rodseth et al. 1991). This approach to human-environment relationships etc. In _ the analysis of demographic and behavioural these respects I believe that a broadly ‘func- patterns is commonplace in studies of other tionalist’ viewpoint forms a natural and logi- animal and primate communities but has so

cal starting point for any constructive study far received remarkably little attention in of human behavioural patterns and adapta- studies of the behaviour and organization of

tions in the past. Palaeolithic communities. As discussed fur-

What disturbs me is the lengths to which ther in Chapter 10, this question has direct functionalist interpretations have been car- implications not only for understanding how ried in some of the studies of Neanderthal Neanderthal communities may have been behaviour over the past twenty years. detect organized in purely social terms but also for an indication that the pursuit of functional issues such as the viability of these commuinterpretations has become not merely area- nities in conditions of rapid climatic and

sonable point of departure for the analysis ecological change, as well as for any associ- , of behavioural patterns, but something ated population shifts or displacements approaching an a priori credo to be pursued during these periods of environmental or

and defended against almost any alternative demographic change. All these questions interpretation. As several authors have have direct and important implications for recently pointed out, the limitation of this studies of Neanderthal technological patapproach is not merely that it largely terns, as the discussions in Chapter 10 will excludes the possibility of any explicitly cul- attempt to make clear.

tural component in the behaviour of earlier The third dimension which I believe has

8 THE NEANDERTHAL LEGACY been surprisingly neglected in earlier studies led to some of the more serious errors and of the Middle Palaeolithic is the question of confusions in studies of the Middle Palaeochanges in behavioural and adaptive pat- lithic archaeological record during the last

terns over the course of time. Even if we fifty years. define the Middle Palaeolithic in its most The final theme of the book is that the true restricted sense (i.e. as coinciding with the picture of Neanderthal behaviour, and earlier stages of the last glaciation) itis now behavioural capacities, almost inevitably lies clear that this covers a span of at least 80,000 somewhere between the two _ positions years — that is almost three times as long as_ described at the start of this introduction. the entire time-span of the Upper Palaeolithic There seems to be an irresistible urge to succession in Europe. If we extend the Mid- polarize scientific debate into extreme posidle Palaeolithic to include all the manifesta- tions. The truth is rarely that simple. I believe tions of Levallois and related prepared core that there is ample evidence that many of the technologies reaching back to the time of most basic behavioural patterns of the Nean-

, isotope stage 7, the time range expands to. derthals were significantly different from over 200,000 years. It is now clear that this those of the ensuing anatomically modern period witnessed some of the most rapid and populations, that the cognitive capacities of extreme shifts in climatic and ecological pat- the Neanderthals for certain behavioural patterns during the whole of the Pleistocene terns may have been different from our own period, with conditions in western Europe (though this remains difficult to prove) and shifting repeatedly from dense deciduous that, at least in the western zones of Europe,

forest to almost treeless tundra or steppe (e.g. the Neanderthals probably did become Zagwijn 1990; Behre 1990; Jones & Keen 1993; extinct (Mellars 1989a,b, 1991, 1992a). But

Dansgaard et al. 1993). Itisinconceivable that this is a far cry from some of the more this period could have passed without radi- extreme scenarios which see the Neanderthcal shifts or adaptations in both human tech- als as having virtually no capacity for stratenological patterns and in the specific geo- gic or symbolic thought, no language and graphical and ecological ranges occupied by social patterns which were more akin to those different human groups. Admittedly, in the of the great apes than to behaviourally and

past, studies of these patterns of techno- anatomically modern populations. I prefer logical and ecological adaptation have been not to see my own position as reflecting ‘a seriously hampered by the long-running typically British spirit of compromise’ (Haddebates over the relative and absolute chron- denham 1980: 48), but as adopting a more ology of many of the most important Middle realistic approach to the behaviour and cog-

Palaeolithic sequences in western Europe. nitive capacity of populations who had Nevertheless, the importance of understand- brains just as large as ours and who were the

ing and controlling the chronological dimen- product of at least five million years of sion of human behavioural patterns is no less intense socially and ecologically competitive self-evident in the Middle Palaeolithic than evolution from their closest primate ancesin any of the later periods of prehistory. The tors. If this is seen as a characteristically attempt to ignore or side-step this dimension Anglo-Saxon compromise, so be it! of human behavioural patterns has, I believe,

CHAPTER 2

>» L. 4A f | .

The Environmental Background to

Middle Palaeolithic Occupation The Middle Palaeolithic spans a period of The period which is of most concern to this

dramatic climatic and ecological change, study is that of the last glacial and intercoinciding broadly with the later stages of the glacial cycle which spans the past 130,000 Pleistocene period. The Pleistocene period as years. Most of the evidence for climatic and

a whole was characterized by a long succes- environmental conditions over this time sion of climatic oscillations in which condi- range comes from two main sources: first, tions shifted repeatedly between periods of from studies of the changing oxygen-isotope very cold, ‘glacial’ climate, leading to major ('°O/'°O) ratios in deep-sea sediments (Fig. expansions of the ice sheets in the northern 2.1), which provide a fairly direct record of and southern hemispheres, and intervening the total quantity of sea water contained in ‘interglacial’ episodes in which the climate the world’s oceans and therefore of the total

returned to conditions broadly similar to amount of water locked up on land in the those of the present day. In all, at least ten of form of ice sheets (see Shackleton 1977, 1987; these glacial/interglacial cycles can now be_ Bradley 1985); and second, from some of the

documented during the past one million remarkably long and detailed pollen sequenyears. Exactly what triggered these major ces recently documented from a number of changes in world-wide climate is still the sites in western and southern Europe, such as subject of lively debate amongst climatic spe- La Grande Pile and Les Echets in eastern cialists. Most climatologists, however, now France (Figs 2.6, 2.7), Tenaghi Phillipon in subscribe to a version of the Milankovitch Greece, Valle di Castiglione in Italy and ‘astronomical hypothesis’ in which subtle but Padul in southern Spain (Woillard & Mook

significant variations in the pattern of the 1982; de Beaulieu & Reille 1984 etc.). Other , earth’s rotation around the sun, with regular important sources of information have come

and predictable cycles of around 23,000, from studies of some of the shorter pollen 41,000 and 100,000 years, are assumed to sequences recorded from sites in northern have provided the initial stimulus for the Europe (Brerup, Amersfoort, Hengelo etc.) majority of documented climatic shifts (see (Fig. 2.9), from studies of the changing surImbrie et al. 1989, 1992; Bradley 1985). Aclear face temperatures of the world’s oceans understanding of these climatic changes and _ based on studies of the varying frequencies of

their effects on vegetation patterns, animal particular planktonic species of marine

populations, annual temperature regimes organisms preserved in sea-bed sediments , etc. is an essential prerequisite to any study (Fig. 2.11) and from more general records of

of the patterns of human occupation by climatic and temperature changes derived

Neanderthals in Europe. from studies of Arctic and Antarctic ice cores 9

10 THE NEANDERTHAL LEGACY

1} 2 3 4 5 6 7 8 “TY al lc Mie |!e =) 1 || so loot o0|| Isotope stages ——_—>

Ih! | I} ot |

ee 2om |

rr)

| 0 100 200 300 Age BP (x 1000) ——>

Figure 2.1 Pattern of oxygen-isotope (°O/'°O) fluctuations in deep-sea sediments over the past 300,000 years, based on studies of five separate ocean cores, with the divisions into major climatic stages indicated. The high points in the curve reflect warm periods with reduced glaciation, while the troughs represent glacial episodes, when the lighter '°O isotope was selectively removed from the oceans (in the form of water vapour) to form continental ice sheets. Stage 5e represents the peak of the last interglacial. The time scale is derived from the ‘orbital forcing’ calculations of Martinson et al. 1987.

Isotope stages —-——-———->

3

oc Oo

oo 1

Ze)

@ -4 . Oo -5 | | 3© 6 | | | Q 7 | -Q-10 | | < Of ®4f

OQ.

& -2

S «3 @

—_ -8 p 5

|

0 20 40 60 80Age 100 BP 120 140 160 (x 1000)

Figure 2.2 Temperature variations over the past 140,000 years estimated from studies of oxygen-isotope and deuterium ratios in the Voztok ice core from eastern Antarctica. The inferred correlation with deep-sea core isotropic stages is shown at the top of the graph. After Jouzel et al. 1987.

,

THE ENVIRONMENTAL BACKGROUND TO MIDDLE PALAEOLITHIC OCCUPATION 11

(Figs 2.2, 2.14) (Zagwijn 1990; Behre 1990; cial there was a period of exceptionally cold Behre & Plicht 1992; Bond et al. 1993; Dans- climate (represented by isotope stage 6 in the

gaard et al. 1993; Jouzel et al. 1987). Througha deep-sea core records) during which tem- , combination of these different lines of evi- perature conditions probably fell to levels dence we can now presenta picture of chang- similar to those attained at the peak of the last

ing climatic and environmental conditions glaciation (around 18,000 BP) and the conduring the later stages of the Pleistocene tinental ice sheets expanded to almost their period with a level of detail and clarity which maximal positions in northern Europe. The would have been unimaginable twenty, or abrupt termination of this glaciation seems to

even ten, years ago. have been caused by a sharp increase in solar In the following sections attention will be radiation at around 130,000 BP. Within a

focused first on some of the more general period of ca 5000 years year-round tempera- , patterns of climatic and environmental tures rose by at least 10—15°C, the ice sheets change which can be documented in western retreated to near to their present positions Europe over the course of the later Pleisto- and world-wide sea levels rose to 5-6 metres cene period, and second, onsome of the more higher than those of today (Fig. 2.17) (Martin-

specific evidence from sites in the region son et al. 1987; Chappell & Shackleton 1986; which will provide the main case study Imbrie et al. 1989; LIGA 1991a). throughout the remainder of this book, Le. The most detailed records of climatic and the Perigord region and adjacent areas of ecological conditions during the last inter-

southwestern France. glacial are provided by the numerous pollen sequences which have been recorded from

- well over a hundred separate locations in

The last interglacial period different parts of northern, western and eastThe last interglacial period extending fromca ern Europe (see for example Watts 1988; Zag-

126,000 to 118,000 BP is one of the most wijn 1990; Woillard 1978; de Beaulieu & intensively studied and best documented Reille 1984; Bowen 1978; West 1977; LIGA periods of the Pleistocene with relatively 1991a). Most of these show a similar pattern

detailed records available from deep-sea of forest development, with only minor

cores, ice cores, high sea-level stands and_ variations in different parts of the continent many detailed pollen sequences (Watts 1988; (see Figs 2.3, 2.7). Typically, the vegetational Bradley 1985; Bowen 1978; West 1977; Shack- sequences commence with a brief period of

leton 1969, 1977; LIGA 1991a; Sejrup & birch and pine dominated forest during the Larsen 1991). The period is known by dif- opening stages of the interglacial (the soferent names in different parts of Europe — called ‘pre-temperate’ or ‘protocratic’ phase) the ‘Eemian’ in northern Europe, the ‘Ips- which is then followed by a consistent suc-

Wichian’ in Britain, the ‘Riss-Wtirm’ in cession of warmth-demanding deciduous France, and the ‘Mikulino’ in Russia. In the species ranging successively from elm and

isotope records of deep-sea cores (Fig. 2.1) it oak, through to alder, hazel, yew and is now generally agreed that the last inter- hornbeam. During the later stages of the

glacial in the strict sense should be restricted interglacial there was a reversion to more to the period of isotope stage 5e (Shackleton coniferous species such as spruce, silver fir 1987; Chappell & Shackleton 1986; Watts and pine. Only during the very earliest and 1988; LIGA 1991a; Sejrup & Larsen 1991). latest stages of the interglacial would there Perhaps the most striking feature is the have been any significant areas of open vegeabruptness with which the last interglacial tation to break the monolithic dominance of started. Immediately preceding the intergla- these dense forests which characterized the

12 THE NEANDERTHAL LEGACY

. Tilia = Quercus Picea Abies

ap eS , aj gE | a; A] ft

ee v YY v _ vy. .Y

Af

Eemian Betula Pinus Ulmus ¥Fraxinus¥ Alnus Corylus Taxus YCarpinus¥

E37}

at CE : mm an.-rrté‘ ie rises during the major interstadial episodes

Pf} 108,000 (U-Th) of isotope stage 3 (see Figs 2.2, 2.8, 2.9, 2.11). _ id _|_ _ ff — == 115,000 (0TH) Even if climatic conditions during stage 3

ve _ > 131,000(U-Th) ereofaera re seem to have attained the degree climatic P< 120,000 (U-Th) highly oscillatory, therefore, they never

HH - -- - Hee amelioration which characterized the more 6 140 + * major interstadials during the earlier stages of the last-glaciation.

a The exact causes of these rapid and sudden Figure 2.15 Changing oxygen-isotope ratios over climatic oscillations during isotope stage 3 the past 140,000 years recorded in core KET-8004, have recently generated lively controversy from the north Mediterranean. In this area, variations among Quaternary climatologists (e.g. Dansin the oxygen-isotope ratios are likely to be much more gaard et al. 1993; Bond et al. 1993; Kerr 1993).

directly controlled by temperature variations than in The most significant discovery is that the stadial and interstadial phases obtained by Potassitum- most pronounced periods of rapid cooling Argon and Uranium-Thorium methods are indicated seem to have coincided with periods when on the right. After Labeyrie 1984 (see also Paterneet | large numbers of icebergs broke away from

the major ocean basins. Absolute dates for the major ; yg:

al. 1986). the Laurentide ice sheet in the North Atlantic

, region (leaving clearly rital defined layers of detmaterial in the contemporaneous deepsea sediments) which in turn would have

were much less marked than those which reduced the salinity of the surface waters in , characterized the earlier, more pronounced the North Atlantic, and triggered off move-

interstadials of isotope 5a and 5c. In the ments in ocean currents which would have pollen records from northern Europe, for caused an influx of much warmer waters — a example, the Denekamp and Hengelo inter- pattern which has recently been described as stadials were marked simply by a shift from ‘Bond cycles’ and related “Heinrich events’ open tundra to shrub tundra communities, (Bond et al. 1993; Kerr 1993). Other climatolowhereas the earlier, much longer, Amers- gists, however, are inclined to see most of foort/Brorup and Odderade interstadials these climatic oscillations as a possible direct had been characterized in the same areas by result of various astronomical orbital-forcing

28 THE NEANDERTHAL LEGACY

_ 20. 30 40 | 50 60 kyrs BP

23-282 SST . ,of

= ma _

Oxygen-isotope = = = = - Oxygen-isotope = = _ _— KET 8004

sea levels

terraces = = = = Voztok = = = = New Guinea

ice core

North Eur ope Sheng suupernenemaae Sake ieee

Int erst adi al S (C-1 4) Papen > Ses > EERE TER > See >

| 20 30 a 40 | 60 Denekamp Hengelo Glinde Oerel Age BP x 1000

Figure 2.16 Comparison of age estimates for various interstadial episodes during the course of isotope stage 3, derived from the various climatic sequences shown in Figs 2.2, 2.11, 2.15, & 2.17. Note that the dates for the major interstadials recognized in northern Europe (at the base of the diagram) are based mainly on radiocarbon measurements, and may therefore understimate the true age of these intervals by up to 3,000 years (cf. Bard et al. 1990a).

mechanisms of the kind discussed earlier Sea-level fluctuations (Kerr 1993).

Leaving these debates aside, the pattern of As noted earlier, fluctuations in world-wide highly variable, rapidly fluctuating climatic sea levels during the course of the later Pleisand ecological conditions during the period tocene period can now be reconstructed with

of isotope stage 3 is of particular interest a remarkable degree of accuracy. The evifrom an archaeological and behavioural dence comes from two major sources (Fig. standpoint, since it coincides with the final 2.17): from the oxygen—isotope records in stages of the Middle Palaeolithic occupation deep-sea cores, which can be used to provide of Europe and the ensuing cultural and anestimate of the total volume of sea water in demographic transition to the Upper Palaeo- the world’s oceans (Shackleton 1977, 1987); lithic. As we shall see in a later chapter, the and from the more direct evidence of varying character of these ecological oscillations may sea-level positions preserved in the form of well have a critical bearing on the specific coral terraces and other forms of raised beach

demographic and other mechanisms by deposits in various areas of the world

which anatomically modern populations (Aharon & Chappell 1986; Chappell & Shackwere able to compete with and apparently leton 1986). Estimates of sea levels from oxyeventually replace the local Neanderthal gen-—isotope records need to be handled with populations within the different regions of caution, since some specific allowance must

Europe. be made as to the particular component in

fe Ee a oT % y 7 ,

THE ENVIRONMENTAL BACKGROUND TO MIDDLE PALAEOLITHIC OCCUPATION 29

Isotope stages ——>

— | Sea level SL from New Guineaestimated

An. terraces ae.

o2 \50h - = PANN N P\e“\IS \ Va R/S on “ \\

8 100 \ | \ Q. / SY

a A| \A sea level'”O estimated = using deep-sea data

W

= 150}

(ENE SOE OR OOO Bn __—__L ___ 1 it 0ee RE 20EE40 60 80 100 120 140 Age BP (x 1000)

Figure 2.17 Estimates of world-wide sea level fluctuations over the past 140,000 years derived from studies of oxygen-isotope ratios in surface and deep-water ocean sediments (continuous line) and dated coral terraces in New Guinea (dashed line). After Shackleton 1987.

these changing isotope ratios which is attrib- stages of the Pleistocene period is summautable directly to temperature changes, rized in Fig. 2.17. The period of maximum sea rather than to the removal of isotopically level can now be shown to have coincided light sea water from the oceans to form the closely with the period of oxygen—isotope continental ice sheets (Shackleton 1987). In stage 5e (i.e. the peak of the last interglacial), principle the direct records of coral reefsand and to have reached a level around 6-10 terraces are simpler to interpret, although in’ metres higher than that of the present day. these cases some allowance must be made for There is clear evidence for this in many parts the effects of local tectonic displacements in of the world not only in the coral terraces the recorded altitudes of the individual ter- preserved in New Guinea and elsewhere but

races. The most detailed and extensively also in the form of beach deposits located at studied traces of coral terraces are those pre- altitudes of ca 6-10 metres in many parts of

served in the Huon Peninsula of New the world (for example in northern France, Guinea, from which large numbers of precise southern Britain and the Mediterranean) and internally consistent radiocarbon and (Bowen 1978, 1990). Following this period of uranium-series dates have been obtained interglacial high sea level, there was a series (Aharon 1983; Aharon & Chappell 1986; of complex fluctuations during the earlier staChappell & Shackleton 1986). Similar terrace ges of the last glacial period. During the initial

sequences are documented from Barbados, cold stages of isotope stages 5d and 5b it

Timor and the New Hebrides, all of which seems that world-wide sea level fell to around | have yielded patterns comparable with those 50-60 metres below present levels (Shackledocumented from the New Guinea terraces ton 1987). During the much warmer intersta-

(Aharon & Chappell 1986). dial episodes of stages 5c and 5a, by contrast,

The general picture of these ‘eustatic’ sea~ sea levels rose sharply to levels only perhaps level fluctuations which emerges for the later 10-20 metres below those of the present day.

30 THE NEANDERTHAL LEGACY Clearly defined episodes of relatively high J oess deposition sea-level stands during these major intersta-

dial episodes are well documented in several One of the most conspicuous geological parts of the world (especially New Guinea effects of periods of glacial climate in the and Barbados) and have been dated by direct northern zones of Europe was the formation _

uranium-series measurements to around of thick deposits of wind-blown loess. The 105-110,000 and 80-85,000 BP (Aharon & precise origins and mode of formation of Chappell 1986). Sea levels fell to their early these deposits are still open to some debate glacial minimum during the period ofisotope (Butzer 1972; Bowen 1978; Wintle 1990). The stage 4, to around 60-80 metres below present deposits were evidently laid down by heavy

levels. A succession of rapid fluctuationsthen wind action and were presumably derived occurred during the period of isotope stage 3. from areas where extensive spreads of fine(between ca 60,000 and 30,000 BP) before the grained sand or silt deposits were exposed to

massive reduction in sea levels, down to the effects of wind erosion without the proaround —130 metres below modern levels, tection of a continuous vegetation cover. But

occurred during the overall maximum of the how far they were derived from river or glaciation at around 18,000 BP (Shackleton beach deposits and how far from various

1987). forms of glacial outwash deposits is more The effects of these sea-level fluctuations debatable, and no doubt varied between dif-

on the coastal geography of Europe are rea- ferent locations. sonably easy to reconstruct from the patterns In western Europe, thick deposits of loess. of submarine contours around various parts can be traced ina broad arc from the western, of the coastline. The major reduction of sea coasts of Brittany and Normandy through level during isotope stage 4 had the effect of | Picardy and the Paris basin into Belgium and exposing large areas of the current NorthSea northern Germany. Southwards, occasional basin and English Channel as dry land, effec- deposits of loess can be traced into the Loire tively integrating southern Britain, northern valley and into at least the northern parts of France, the Low Countries and southern Burgundy (Fig. 2.18). The thickness attained Scandinavia as part of a greatly extended canbeimpressive;in parts of northern France, North European plain. At the same time the for example, it is estimated that up to 10-15 coastal plain along the Atlantic coast of metres of loess may have built up during the France was extended by at least 30-40km_ various stages of the last glaciation. towards the west. The much higher sea levels The major problem has always been precise which occurred during the warmer inter- dating of loess deposits. Wintle et al. (1984) stadial episodes brought the coastlines much have suggested that the formation of loess closer to modern patterns. Even so, the high- may be a rare event during the Pleistocene est sea levels recorded during the major period asa whole, perhaps occupying only 10 interstadials of the last glaciation would have | percent or less of Pleistocene time. The only had a major impact on the geography of the form of absolute dating which can be applied areas bordering the North Sea basin and_ directly to the loess deposits is that of therwould have left Britain an integral part of the moluminescence (TL) dating — based on the

main European land mass throughout the presumed bleaching of the TL signal in Whole of the last glacial period. It was only quartz grains by the action of sunlight as during the period of the last interglacial that particles were transported through the air | Britain was fully isolated from the continent, (Wintle 1990). By applying this technique to with coastlines broadly similar to those of the loess profiles in Normandy, Wintle et al.

present day. (1984) have suggested that the principal

a THE ENVIRONMENTAL BACKGROUND TO MIDDLE PALAEOLITHIC OCCUPATION 31

ey SB es ALES s PE ty A eed

ae oo) —”:”*~*=“‘“‘“‘“C:*S*CURUERUE UaESStaSESSIONS ena nnn a Loe eeOES SERN CRSA SRE . Pm sitet RE SHSAR RSTO SHAHN SONOS SESE NENTS SESE OUD SON ERE ESS SES SARIS ENO ESR wa Sse |S SRE ES SS Sr hehe a anata

: oe be Se ee ee ee ee oe

eenolin IAs . ee NR 2K ebenhayn Se ee ~~ rss c rrCC Yen Ce

a eee ee a ar

nfCRA Statham, (Siar Sen a;ce eyaCEaoe cf SES RNS RRERS Tie ene odRG CE aaa aie etches tatera9 ae ae heater Ser SUN Sr ‘ . ES EE ~ ie all CR eR ER . os Want ee MICO aN MCC OC RR OURS rey ADAH EAT eel opeccasartotetet batenatsbebebetpneeceregensteenetecntenerstesassteneneeegeseienerecasenegegeeccaenenmgegetereenepeceenseetraeecentyy , poe ey eleietetees “ea eee

Bh TERR ARA EAT NAR EIEN AS SOC Te ; ASSESS SETS ORE Oe tadctrcnegecnnniatoteter muusotageltgetitete celeeter ote teuaseteences crests grater OY fay RUD oe NOIRE NRE ERIN

Be aR en eHRR EE a} RR COURS Se aoy ettHq SOND CR EEN sae AE ay raat]vee pan - t Seo Cee inva MbUpame SXS/ FSR aaa ARS NE tLe Seen NC eR nora) aa , Re ; ‘xCatt SOBMCARS ECeeey BEWhe FS oy RS ty an Ree ~~ NE SO EK kuti Ca fo DS

ee ee ee MS NR SU di aPe lc DCU ee IN i eee Eee in ena Coe an2. ce ee AS EDIE legen wermeernern C5 fe ae ORS a en raiDigi eNOS 2 ES SO PELE get aasrs ee Foy_& 7 ESPLOID teenacundicad RR ES Pry p CLO Ge BREE IE Ay gy ONE EP ru, pag GE 0D aa EEO asseeertentttanl

OR, etre LE OF i! Choa S20 SESE Bess Crdg Xi eS gmeecd d Ber]ja Neeet

ee es ee fo not ‘ Lf Gers Pal g pote eee) OR a iN ie fe Ies ON 6AaeOSC arene ond Fo Ogee Gree © SRR es 9 9? Bee QUT Ce CosYS os TO ~~" ae

cS eee soa = & 2) 0 KE Potion EG) em g One MEO g CLOW a yh Vj pe °D, 4 PO Sl a Wp IO ee ~~ Tou SEE ~ Ba \er®%, fS> 3| |eee SSE Vi AP? | pe SOK AA ey 6° Ly Se DO ld, 8 get \ uy. EE Af f gh PMin cheno. kin» fe aad aneneeare ete AGe/ Boe Ses tit x iP WSS Gb t eA RS EEac ieasa OR NEcenOS , RPe SFYS sat

peter aia AOeso a Qawp oC ®a 4aea \:fereeraat anu Shoe 0 Hfa iv p15, DUAL & O° fEY°©Zoos (J? . oe,a €MV Cae e Nh fc, Naty) +° Sees

ee Lo ern 6 : ¢ eS 4 | fj —, CB CC Lyon Xe RP Cg ESS NS Se ey x SD © , fA. SUN | | oY ySas LD4 iQ §49 ) \ iB Veneer )Bilp 70 Ti othe serine, aN G 2UloUgs. gr. By , - oN LAVASG) a

en Ss [CS __ — ee fs *Sille paz ees 19 “=ape TAT Sd 7 ae ed -reerenn OY, ee reRRS ae - qEADY y ‘y PIP]

Foi Ea erm ETN 1 ih oa one AtA IC KR Aseee per Fed EL CN ee ee A Le Js Be SNE NE VERN 6 a era — a? LJ oe 32ea@ ty af X 3653 a PREM GIS ta eon eee ra cgstaeniaee peg ee Se ae a Ris whee: v Oem e o * Fad BIE mE pra TI ieee TE aN: tite Bees Nea A Date UTES in — Cy PAG: ry Sed Gee Zeeay Yop ,

Coe 29 \\ 27 2532\ 26 me weal | |

lana the its ere: Ni V+ : NLS oe Len GUC ae : SIRT DT A th $00

W B S>NA: I

|

E ' (,’ Pi: at > j YP \ Pe

ana Soe coe aQO )D aA a an 38 7 Leu cal) Ope! O mpm as Fen ta ire 4 Lomi erio tela 1S es 0 a Gp?

meqanced GS Rigi Meds WG RAEae Penn sh EEE say \ : (at feof aaal oie AES Wey {3 39 bis Eerrrrcupe VAR

gna eneeet et wicte ecco Taree

Gis ty ' EF ARNE. Ae ae

4 ee 42.41 40\ 39 atsy ¢iceeteaaia te Leeann it ryeT Wie ge Ns BO NANG On sae eS Ce Sd Nee \

rs JOSE aEERNe TRINOR No) O77os O Tseres BEC CERpeer I der =24 TN ne? oe waa, FV re i Serf uy Upeiale wird 30 37

oeeeeOe eree1LL f Yo ACUmt Wer EP Arai a EO Re 759 mo" ¥ "2490 i ti wale vey 38

OV Seay apna soars,Ee Boe pin RUN EERE Ose rnea On ee erWeA meer BOE Pret Lite A

Sepeinsnen Srusoommanrnypnine manana 11M “a L Nea NR

Dra Re RRR 7 LKR NCN Vee) HLL Dae i A ee 998 57 56 Gee ‘ ITTdo Ue\44 gee iu < “a 43 | es 48 ee De Taine > \ho\ Beep veceoressscassagquueesemug teal, aU AT a. (eens ewe Psa iii ag ee Yow 48 Lt td

eee 60 Td

||J,

i ae eee Ds OD OOS Uns re

S62

Figure 2.20 Stratigraphy of the rock-shelter deposits at Combe Grenal (Dordogne, southwestern France), showing the upper (above) and lower (below) parts of the sequence. (Note that the section shown here is folded, and therefore includes both lateral and transverse components of the stratigraphy, as indicated in the central diagram.) Note how the deposits lie on a series of erosional platforms in the underlying limestone, which become rogressively younger in age as the rock shelter recedes towards the north. The deposits span a total of over 13 metres, and cover the period from the end of the penultimate (‘Riss’) glaciation (layers 56-65) to around the middle of the last glaciation (see Fig. 2.23). After Bordes 1972.

o ry . f . vA . «

THE ENVIRONMENTAL BACKGROUND TO MIDDLE PALAEOLITHIC OCCUPATION 35

Briefly, the main features of the sequence soil formation which affected virtually all can be summarized as follows: deposits formed during the preceding ‘Ris-

ay: sian’ period to a depth of at least 1.0 metres. |

1. The basal part of the filling (layers 65-56) According to Twill (1975; Laville ef al consists of a series of stony ‘eboulis’ deposits 1980), it is unlikely that any entirely new which extend over a depth of ca 2.9 metres deposits were laid down during this interval and lie on the lowermost erosiona’ step a the and it would seem that the upper levels of the bedrock forma tion (Fig. 2.20). All available associated soil horizon were truncated and evidence indicates that these levels were eroded by the effects of heavy humidity durformed under extremely vee climatic ron ing the later stages of this period. As a result ditions — possibly the coldest conditions in there are no direct records of either pollen, tne Whole o me Some renal sequence. ai faunal remains or associated archaeological

P 5 ; pos ; assemblages for this interval. Nevertheless, it these levels (Fig. 2.21 a consistently is clear that this period represents a pro-

below 15 percent and in the me) ority of levels longed episode of very warm, humid climate

(Bor dec me ny arn ecles represented which must have persisted for several thoureflected in the character of the ‘ssociated sand years. All authors are now agreed that sediments. which also point to a period of this can be correlated with the peak of the last

P JP interglacial period (i.e. isotope stage 5e of the ,

extremely cold, dry climate with intensive deep-sea core sequence) and accordingly frost action and little evidence for any con- dated to around 120-125,000 BP (Fig. 2.23)

temporaneous chemical weathering of the (Laville et al. 1983, 1986) ‘ 5: *

deposits (Bordes et al. 1966; Laville 1975; ’

Laville et al. 1980). The composition of the 3. The third major phase in the Combe Greassociated faunal assemblages provides fur- nal sequence is represented by the block of ther support for this conclusion (Fig. 2.22); deposits between levels 55 and 38 spanning throughout all levels remains of reindeer approximately 2.5 metres in depth and lying,

account for between 72 and 97 percent of total once again, on a separate erosional step in the , ungulate PEDarns with only sporadic traces bedrock formation. The combination of the ‘Bordes «Prat 1 os), as red deer and horse sedimentological and pollen evidence shows

. . . that these deposits formed during a succes-

All the available evidence therefore points sion of four major climatic oscillations com-

to exceptionally rigorous conditions during prising two major cold episodes and two the formation of the lower levels, which are prolonged episodes of much warmer climate

assumed to correlate with the period of The cold episodes, represented in layers extremely cold, full glacial conditions repre- 55-53 and 49-44, were characterized by

sented by ' he later ae of a yeas om - severe climatic conditions leading to inten-

stage 6 (Fig. 2.23) (Laville ef al. 1983, 1986, sive frost action in the formation of the | Guadelli & Laville 1990; Mellars 1986a, 1988). associated sediments and reducing tree-pol-

Archaeologically, these levels produced “len frequencies in the local habitat to around series of late Acheulian industries charac- 12-15 percent (Fig. 2.21). Nevertheless, it is terized by rather crude ry made hand axes significant that arboreal pollen frequencies in and low frequencies of Levallois technique these levels are consistently higher than

(Bordes 1971a), those recorded in the majority of underlying

2. The second major climatic episode in the ‘late Rissian’ levels and show a continuous

Combe Grenal sequence is represented by a_ record of species such as juniper alder and |

- period of heavy weathering and associated hazel in addition to the hardier species of

36 THE NEANDERTHAL LEGACY

2 8 “ 46 4 SE ~~Se & Re 9S 8 ; Ss 4 2® ESS EFssSues 5 PQ t a Ss

2 gb el 3 2 & = S bee ZS E © S & ay

XY—SGESRPTEEESS S DiS x =At = felt Ig 4 ARBRE utab Hn mbH

? MEEEE i rinmiinimemii SM GVEseaaeeeee ~—e po S0RR ERE At _ cS 4 : aiming duu munineunignganias

poy ep =a Se a SG + | JESS SEES aiT=HidBeNONE ;206010 ; _ eeunre ee et Aa , Eoin cx ep he a one GE ees 5) , | L | a A steerer

9 : _—— ; Tan ornemneamee ener cneraene

a5y}-+23 , , }, Hebej Gracocaarieren RecOE RIUTY sae rememenen 76 — : I Garacemmemnmnehesnennanedinntins

35 Re ue bininsnnnngncguinas

am S82 Saas i ee RE

ope a

= 8 ) . Sapa cepthaenn =e leiceapeterm eee RteTentEs

ale S00 een) §6¢ Ge) OSS

|= a56 55 a -: . |> EEE ,“~— ::°-.

| Ge SGGGnidan “ai . Pep ee

= TEE TT D—D_NO"— Fa sc REet Renen RNR _ TCT 57 ‘ TT \ : SURRRORRR RR RENCE

jet OO Seeceseceeeeet “Sas . “tes ON ON OS A , ma : , Sct Figure 2.21 Pollen succession recorded through the Combe Grenal sequence. After Bordes et al. 1966.

birch and pine (Bordes et al. 1966), which layers 52-50A and 43-41) reflect a dramatic suggest that climatic conditions never change from these conditions. In these levels approached the severity of those recorded overall arboreal pollen ratios rise to between during the preceding cold phase of isotope 60 and 70 percent and now include sub-

stage 6. stantial frequencies of a wide range of The two warm episodes in the sequence (in warmth-demanding deciduous species such

470 [oe 114.42] yoy > 14 ( }_(iene, re f od ; 15 SA

THE ENVIRONMENTAL BACKGROUND TO MIDDLE PALAEOLITHIC OCCUPATION 37

Reindeer ————— Bovids —————

Layers = Red deer == == = = = Horse === =

9,10 —_> _—. = 13 f f ae — ee ame

17 ? NS ( oo 18.19 & Pa , = 20 S i ( 21 f q. \ “7 ~ S 22 ? te j ? a 23 4 > i Pa wurm Il 24 Se a y 25 sy f + | 27 \ ‘ V7 28 oe _ 29.30 a } ee, 32.34 Pa ‘. 35 f “7 a 36 | ( ‘. 2 —-——- ,

40 | o x | Att fee (> 51 52 \.r oa CN a |

37.39 |f re Py

42.45 §mm oo oo! vem | 46.49 | | f Wurm 50A | “Ss 1 f

50 f - VSS 6A ne ae oo J_&by LB A + .. —_ N

Zz | 1:i>¥2a~OU a g _ ? : D Y — _— foey SS =_MIPA 4—_ — Tr :;z _ :' :ay +.+ ++. ge 6=& —O

:i:Sy & A ||2|y:| -:| LL S : Ar ;23|H

:| _26 .7= = .—_ & es x _ Or L . | | T ,. =| ,72>7-. :q:+7/—|° —— Y f. |._— - =- =aZ=N4 ___ r 2 —~a+- i=

; 4| +] ,|A°; U cc ,Va G ~| a,SS| a7, =es aaeP ' :n ~ . . : = ; . 7 _. + __ —— P

7 nl . | : of = Pn ~ | > : 7 | | ee | 2

;am aoe|aeo +)Bae _Ox: IIe-— — ,,4o| :7.4|* = ~:=— ”= =+aS Ooe saa !|,4'| — Be 6| Ly ’. Ly a5-AS: a-Zes

|.+._f =; :+.= +| 4~ = 41 ey ee Bla Ae _| — : :.-== : + = tian ai ee = ; , ~ n + + : i — a =: -+.+:ad =st+HSS . _Oo. . | .=_# oe x2oh > ,=:“ —_ + Dae + . ce =o: ae hs 55

h, .||= b= ++ . Rea ns|°es 4S ce >a" mi,ee megs fi aca tiE “he *oe : eae 4 |: : ~ = 3] 5 '| — |_J;a:-.:=5--’‘4 : ne tir }aes=~onne ==: ere os |3|7:yeat :¥z:E 7 —_, soos 5a 53:| -< z |O. yfGun “2 eeathoe

,6é| . : ; : : “i 5 = :--= am :5 :|ae :.|j= _— ee ae ee ae OF 6f 0 Fr ia 7 _ : s 5 os Z : + | : = es or ae ore 4 2 | ru j . —_ nee Too nites ee * 6 5 Pasien _ + ' | F y — ae ce ee 8 : _ L i = _ . ; ; i: . ; : = a 9 4 ea 4Th : 1 a | 4 ) ce ae o | ee 4 HA , Ex =e ae . t.9|o¢9 ? ; al a e mot . = el Eyes =" a 3: :opr _— i | a a V : le ; * = rt Be ay — " 3 — ; ; ae : — wo? _-— 7———— ————— :FE . 54 :| ::: pr — ry | ———— = U "ue : as :sed se — [—— = < ) | ——— : on ne = = os a i oe 0 0efo _— an oe = =; ae aoeae “4J:8 | 5 JO = : - ad . < Aree C D bas ° + oe th - : | 7=_="— aN— Pea aayaefyesneae = J. Axe

= —~ eat ioe 2 Y. ip cs) Bae i ee ie 6 f aeE_—h— x ax ay: t :iia ae oe a7, — "oad. Ce re =cava! ee = B85 aS{. 8~=—_ ty=ony) Hee iSoe aes ates 44 359H;:; wy ste aNE ey 4ae Sat ni24— He ce aie He Se —

= -fe) Ree ae61S 6; hiro&gh — ==aa, - iy — ==624 rTer -— 55 -re i—ae— oO 5 2 ’=— a - at tone Aytea LY ie CA oFane AY % by te ee Ni 5 6C

) : 50 | | isleposi _ a 3 8 , , a se | ic = — a _ 5 oe S : 5 af8eit|rie 5Ca a = 7 _ ay | | a—4=92oe5|| 1g ”:ecleae

SetCa fs aae »5:5565::: -ige10 1oe = aieae = ie ceis tsa e- id1dbad |gl oe .™ fe)ae =nes eae sere) Pp ~te! i98

de |J |af.5:J| .— : A:a4

e€ S 0 — ~ ens ¢ 2ee ate Te “f 2 sett - :

T] —— ;

ze

toy

44 THE NEANDERTHAL LEGACY | stage Figure 2.2 correlation Layer (N) Reindeer % Arboreal Isotope pollen % correlation 5 c osed S Prop C ation ffofe th faunal al and a

3 ——FE - ) vegetational sequence at Pech de l’Azé II with isotope

8 tm f stages 5a and 4 in deep-sea cores. As shown in Fig.

—| | 2.26, ESR dates for layer 4 centre on ca 75-80,000

lif| on st tage oo 4

ff. R 55-72,000 BP. From Griin et al. 1991.

| | _ — 5a

48 | 315

[ _ __. oe _ oxygen isotope stages

Nc Tp ic lol |: ayer 6330 a ae :A

0 10 20 30 40 5060% 10 20 30 40 5060% 12 34 5 6 2 |

:c“oot : 3 6370 an ! ! ean : 998% a 7 :i E 2

i783¢ae : : F 78 ie 5 | 78a = : ‘ wat | : G c eo ! 1 1 ‘ 5340 = ; 63st i : : ° b = 2 : 627¢ a | 628 eg | : b —i : 62 Ge —“~T ._‘: : 4 eo 1 ! c a ae : = : : C d _s : ; 63 fa / af 6260 | 4 6 ! 7930 co eS b ot SS 5 7940 Do 4 b ; +s — ‘e) : Cc \, !—+— + — fe) . 6234 re Cc : ! + 8 6249 : ag ‘e) 6 625a ! ——_« Oo bc; : + —_4—— iO fe) 6200 a | 4+ Io b BO he 0 62 fo ; aa er aa| °° 7 6220

6i b ronan ; ; ;

yZ ; Oe .:.'t;‘°

Figure 2.26 Electron-spin-resonance (ESR) dates b ot AT 8G obtained on animal teeth for early-uptake the sequence(EU) at Pech de 6172 l’Azé II. Dates based on the model 61a ipo eea 6 3 of uranium uptake in the teeth are indicated by 6 Iho 7 a 0 !

. . . ; ’ b ’ : ——_4.—. ie) ‘

triangles, the slightly earlier dates based on the linear- ¢ po! a — oO! uptake (LU) model are indicated by circles. The lower 615¢ hy —— oc 8: 9 levels in the sequence (layers 5—9) contain Acheulian 6165 a —o o |

. . -: a‘ se‘ —_—4--_— .: we re i en ee A eee ee° ae J

industries and evidently date from the period of c Pt =‘ o | isotope stage 6. After Griin et al. 1991. 0 50 100 150 200 age (ka)

. yALTIATIN arty

THE ENVIRONMENTAL BACKGROUND TO MIDDLE PALAEOLITHIC OCCUPATION 45

BO Sig ee ir Eee Se ae 5 en oo | a 7 a NEE age a oe a aeee adewesiisee rs wotsEOIN “ptt ae se os, _ Se i ee; aan S e ©Pe a 7 ae a. : oo Tog Dibba : eSa; ~My OE ee a a eo) ar r “eZ ee rg oS oc egetapage | a~~Ee aeee Ae eee. ae , 5RS age fhe ieee gf RUBNM AE htOE tom, aee SA. ee .ait reFeegE see :Ree Fe oaey -. *we ger pe LG OO fact Pee Saga 8 [ ee ae aeons Went te “s CP Ee : > # Pik ee ee _ sae : , *pe_es ee ; Pes a eeameyer gs EE "aS Woes a ;Tee ee aei aeeagSe Aaiguacien FR veg . a ee Par a Serie @ oS : eg ._._ —s ¥S oe - Sar elBley : . 4 : idRT — aROBES ee: ye ee osABASSRS oe . . ts, BSS OM Ae a4 ~ 4 Eg 4 . ee oor oa .. fe ice ONE a BR . . ; Ye CL) ~e — aa a.| a.Sa . me aI: “iVeo BO: dad eee AS ~ iWee | Boris ce: aoo hl oe rea aA eeoS ve Mouton eee BS -oe . ; Ge Pa ae ~~ en ee - leita itl =ll 2S aa oe ae PS pee PSE F ee 2. — oy eee se "es 4 ae MS ewer me Ee er a ag Se me nn ee aera al

a a ee eS SD Moa Sn Ne PARR eG = Bee a a ee, Se ERE, Wey as: Se SFoa a | ee i reer ee ee Ce ee a ae ere ey a ee ie oo ia oo ee ae a i Oe eS ON ee es Sy eoH ee OT ee eee eee aeaaOe re eres ee eeOe eeee ek a ee 2ee re et ee ee ne Ne eS Mt a ee ee Eg Pe es SG er as {0 cod de Wallon Tress a. eei we ceAeeBeg x: ar -~os. eo. .-co. < SS : 28 Sea fs oe So LotRO poe ietea a .eefa é-ae~ES .:% .s Fe Feeee Res . ne6. ~ _Oe ht eet A ee ae Res “4° ee ee ee a ee eae Cee a eee yk, Ae, a ae ee ak i ee ee es ES efaSanil aei a cal he Pe eS fk PE oo a ee iva lig TAs NO ae acta lle ee ee a a eas aces A ae Ne ia. ee ee ee ee ee ee Oe NS oe a ee ee ee ae es Ce a ae ane Lees EG RO REET 8S LSS ee ae - aS ee ee

ae &SRGae Toe 2 of eee SSSe FN2Sehe ee APG “k ee28aSke.oa veSLae i oages . ae eo a+,ae SERS Se pe niet SERRE NEE SN SRS SES ogeee RS a 5 gages BN ri care he WEEN. oe TAR gia he gs eas Bos BBs Ege ek ABrmSets pee oe (8 Gees ue PERE GR SEfs RE Nee ece TR O8 BS Gsecee SE es

SABRE ET Sa can : : 2s ae 2eagRRE SAS 3 1 pet SESE gy SAR, SE & oe SRE es ae a ee ee ere oa Mowsbivien ee a re

Oo eeMsae pareehAa Me RES eeWohee Re ee : ic _Sebo ARES ON Os oeee BE “OUESSSSS . PURPA eSoe GR 2 ce Ba SNoe SSS RET o's RC Oe SURWR ite ee SeBS RCC RRS eo Se NS EES gE EEet See5St ae KO SOUS ERE a Oe Ce OE oe oe Be ee ageee SEIS ams Set eee ae . ee

ae aa Se eer eV, 8 eNOS re re CRO OR ee SeOe.Skeeee . a OS eeAEa gy —— sO a : Bott, Sil ie ok = Bo. oe ee « 7 mye NTR SRE S PARR SN SSIES OSE RES ESR SE Tigh SESE oO ESSE oo ee Se ig Be SRE Se 8 LOSES UP LOTOS NAG sh Seg 2, Ses lite CS Se Se hae i eo A Be giles ee

ee ERR a eeSNES ee Cee ee ere | RSs 2g) Ee? ekgog ke ores nee aog eS RC AeSRR Me deS ocLS RC - SS RSa BR oeGaRR by) ReN Ce Bb So tg beg?rk. Beir7oaER gS USSR Reg Be Rec asf URVRES GRR RRO cn NP Bing GEOi to. 2ee ee ae .< eeeG 2 aefee Re See Se 7 OT Be =

Mop gas PEAY o.A ee ee re Ce Ake se aek Bs 8ee ae oo aac hyeeee Sree eeee. et— ee oS ae aoi. eeaRR eeBae SS, ae 2. Se S 2: re ee eat ee BS oeeeol. eee eran. eet

i ie 3. a ee eee ee ae > + =e a ee en a Ga ee Seg + ae” ee Oe ee i tt ee Lesa a SS See. be ok ee ss - es ad = ies oe Sys Oop e Sau R aig g- SOOT ec 2 ee See ge ty SL, SS ote Ga a - a. =

a : : oes HS oS SS eS ee oe 8 Bie. OSes ben TR) 9 Se Sere GBB EEE REPS egag 0 ME ecre ARAE: a So te. Ree = oo Sa Oe a ee ee =¥ un SS eee

- Beco Sh SigeB Beene S85: _geo etcoye TP -eee i. ce. ce RS orsES aSiyaue> eSies a eC oeSe hs Se Ses & BRAG cee a f : _og, . PRS OU aofigreer sd 2ees OE pf SSae ee =e ed Sy ee St See Reeee SEE vera SS SP ee SABE GSB- & SOS ger SRB a a aBa osSAE HESTE ee

ye eeae BB eeayo: le ee Ce Goeseee yg Ceti—“‘“‘CiésSS ee SE Sg es ape ee ’ 5 : j - ; ; ’ iv , Ly Vv 4 i }- ) - 1: a - 2

Figure 2.27 Photograph of the stratigraphic section in the lower shelter at Le Moustier (see also Fig, 2.28).

°°.«Ld . ry. °e °ee. .-

ymbGrenal dl in and, parti r, rei is reinforcing Pe) 1930; Laville 1975; Laville etal.al Combe in particular, (Peyrony ; Laville 1975; Laville et

_@ e _ . Pi ° . « ra j f . _ — — 1 =

the specific correlations proposed with the 1980; Paquereau 1975; Valladas et al 1986; general climatic succession in deep-sea Mellars & Griin 1991). As noted earlier, these

_ i° _. L ri i -a?=_ ry L°...

final ste in thesequence M lar ice % cores. final stages in the Mousterian are The stratigraphic and climatic sequence not well represented at Combe Grenal and

e -— F = e e _ _ _ - s e e _ ~ -

recorded in the lower shelter at Le Moustier cannot be dated with any confidence or preci(Figs 2.27-2.29) is important because it com- sion in absolute terms. By contrast, we now

_I _the*| fLthe¢ M4:hel _ L7 ier, « _b e

plements the evidence recorded at Combe have what would seem to be an internally Pech deyl’Azé II by addiandhapparently h d 3 |secure chronology Grenal and and Pech de l’Azé adding much coherent more detail to the pattern of climatic events for the stratigraphic sequence in the lower during the later stages of the Mousterian shelter at Le Moustier, based on long and

. - - 1 ~4 £ L

sequence between ca 55,000 and 40,000 BP detailed sequences of both TL measurements

46 THE NEANDERTHAL LEGACY

Industry TL Dates ESR Dates aLayer (x 1000) (x 1000) L Aurignacian BORO TS BG OR | rs EU LU

| ee a

TS Le reee ge eS) er, ; , GB ee Oe. 42.643.7 K Chatelperronian Pe ee LEGGE C SOBER See e,

GS: —< onsBe5, ane Tn ON Pa or

ee ec ee —_— = Dine aBo a EG ——_—— Mousierian Poe Bg ae OePP eseJ sta an URIS | Denticulate =) ~~ ge spe 40.94.5.0

— Pi OS ES eke Se yy

Se ee

cee eR rer Op ieee OOO BES Hy

SOPOT MO Gree RNR ES eto e, He Gisgpon sg SRR Bio ES oe itetacmreon i He 42.542.0 39.7424 41.0426 PAARL Bob Rt Ola vs Oe ep one a ON ORs

Type B ee a a ee

H MTA ots an Og OL Sp oA ee 00) OT el RB He ON CaN ES SORES" GOS IO RECO OSSn nO aero Ss HS HOE Ch Sr Gs OC SEQ VE AGAR S He

STUER, SES Rs GAR ME OE ees Hd Ree NT KEN TENCES On tren aE MORI BS SO VA EORSUT NTT PERL CoLT USES EOI RCE toy RACER RU SCR TIENT

2 See eS’ 46.3180

| Teal OCTLEMIMRcArELATTECSEPDSETEE OSboeaccnisesteay areas

(Weathering) am i iateaires eas aay | ‘ | sapseeea, Sapeeh Ye AGe \ G350.345.5 MTA i ‘ut eee SEMBLE! fears Peer Asta nee AA YaMen SI aRcaatedacesestesbcone/eG tama Nbacetcg] CHCHiNeenaTt

G T ype A i alae CelTe i: i iee : ron aa ts aH ee G2 + -43.0423 47.0425 Dest lS oN OEE HEEL a PU ELP pee Socotra Eek tre HUME Tire REE ESE G 55.8 +5.0

Figure 2.28 Stratigraphic and archaeological sequence recorded in the lower shelter at Le Moustier, showing the results of TL dating of burnt flint samples carried out by Valladas et al. (1986) and ESR dating of animal teeth by Mellars & Griin (1991). Note that the ESR dates are calculated according to two different models of uranium uptake in the teeth (the ‘early-uptake’ and ‘linear-uptake’ models), which yield slightly different results; on both models the ESK dates suggest a slightly younger age for layer G (43—47,000 BP) than that suggested by the TL dating (50-55,000 BP). After Mellars & Griin 1991 (see also Fig. 6.19).

of burnt flint samples (Valladas et al. 1986) Mousterian succession in southwestern and by a parallel series of ESR measurements France.

on animal teeth from the same levels (Figs The general climatic pattern recorded in 2.28, 6.19) (Mellars & Griin 1991; see also the dated part of the stratigraphic sequence Mellars 1986a). At present therefore the at Le Moustier, (i.e. between layers G and J) is sequence recorded in the lower shelter of Le summarized in Fig. 2.29. As will be seen, the Moustier must be regarded as by far the best pollen record throughout the greater part of and most securely dated sequence of strati- the sequence indicates a succession of pregraphic and archaeological levels so far dominantly cold climatic conditions characdocumented from the later stages of the terized by generally low frequencies of arbo-

THE ENVIRONMENTAL BACKGROUND TO MIDDLE PALAEOLITHIC OCCUPATION 47

z » $ se 8g 3 $

ole r 2 o28 S38 & 38g 254 & Egs2 Se. e3 8 $S— 538 5 esse sgesle oc S2eeas

iish ene | _ ao — , AP NAP ity TF] ry Se eaaeSS a _ . ie | . PAE AN SRO AE Sack ee 7 we +1) | + [+ HERE Sat 54 5M SHEo @ Pw at £ “Peveeaaz = 2Se=a5R80 SO} Sst-=i 36{> 8DMPO Cekt2o&f& AL OF ownso£59 dtoc rab) eos © OonwoonoxBS =z Ww ftBESTS oOoanaudHO IW

c| | i | | IR ee iene ooo Saas a Hei aa ont im | g eats PSTN Pee OeT : 7: > Te BCUS EN EFS RS 3 : : - : - : | Pe Pee Be it ae SRR: min | Bt q | aSS ae : 8RRR Sea MIA ge S:. . 7 :- |leEAA as S|, aan } ; + Se | me | | Sat are £§ -_ Be RAI te ye SEAR Eepee RS J| a| |@ aa | Bs aneneeee a Genk beeen aaa Geaees

| 1mt| |eee } ee HP OoU eee Faa Us] weein) ae one ae La. HeeRun ees ieeiN ee ee 7 7 = a. f . : —— Fe Fe Ne ae SEG oe MN FLY

|| . Senet a Se ea :

4 4 RO EPS SCONE : Dip raten eea 0 ; . STK AER CARE ATG 7 ae ae g | Peace ~7Liq+ — S| Haig ae GE eke a |_§ | oe ER ae . ; 7 SCUe UR Rese we See PATS ey eae Saha aeee es Oe - +5. __§ _ » —{ [| jt _H Pena aS Ss Paes ace ans

e tH fem fof a -#-8 § - - FF -— ff fh PA Ae Sy he TN sali al gga

it ._ Wee oe Jptto tt bb bd. ERE ES Ey ee Se

05050505050 2 40.6 10 $0 05050 10 0505 0 2% 40 0505 0580505055505050 10 0505 0505050505050 50

Figure 2.29 Pollen sequence recorded by Paquereau (1975) through the upper levels (layers H-]) at Le Moustier. For details of the stratigraphy and dating of the sequence, see Fig. 2.28.

real pollen, mostly between 2 and 5 percent, ance of more warmth-demanding deciduous and with pine as the only tree species repre- species such as birch, alder, hazel and wil-

sented in the majority of the levels (Paquer- low. In the sedimentary record there is eau 1975). At three separate points in the evidence for an additional and more major _ sequence, however, there are indications of climatic amelioration coinciding with the phases of much milder climate. In the pollen interface between levels G and H (Laville records, traces of apparent interstadials can 1975; Laville et al. 1980). In this case there is be recognized in level G2 and in levels 11-13, no direct pollen evidence for the climatic in each case marked by a sharp rise in the amelioration — since the phase was evidently overall frequency of tree pollen from 2-5 to marked more by weathering and erosion of around 15-20 percent and with the appear- the existing deposits than by the accumula-

48 THE NEANDERTHAL LEGACY tion of new sediments — but the presence ofa _ graphic and erosional hiatus between the

major warm interval is strongly indicated by formation of these levels and that of the the presence of a heavily weathered soil hori- overlying layers G—-J which, in the absence of

zon which penetrates for a depth of almost any direct dating evidence, could relate to 60cm into the underlying deposits of layerG almost any point within the last glacial (Fig. 2.28). In the earlier literature this level sequence. The pollen evidence from these was generally taken to represent the position levels reveals two major warm episodes of the main ‘Wiirm I/II’ interstadial within which might well equate with those recorded the Le Moustier succession, before the recent in levels 50-52 and 41-43 at Combe Grenal redating of the site by Valladas et al. (see (i.e. coinciding with isotope stages 5c and 5a) Laville 1975, 1988; Laville et al. 1980: 174-181; (Paquereau 1975; Laville et al. 1980: 181; Lav-

1986; Mellars 1986a, 1988). In all we can ille 1975: 186-7). In the absence of any indeprobably recognize a succession of at least pendent chronological control for this part of three separate interstadial episodes within the Le Moustier sequence, however, it might the later stages of the Le Moustier sequence be unwise to press these correlations any separated by phases of much colder climate further.

and all lying within the time range of Although not within the Middle Palaeo55-40,000 BP (Figs 2.28, 2.29). In archaeo- lithic time range, it should be noted that there logical terms, these levels were associated is evidence for a further, well defined inter-

with typical Mousterian of Acheulian tradi- stadial during the opening stages of the tion (MTA) industries (of both Type A and Upper Palaeolithic sequence in southwestern Type B forms: Bourgon 1957) and withasuc- France, coinciding broadly with the earlier cession of faunal assemblages heavily domi- stages of the Chatelperronian. This can be nated by remains of large bovids (i.e. either seen in the detailed pollen sequences from Bos or Bison) (Peyrony 1930; 5. Madelaine, both Saint-Césaire and Quingay and _ is personal communication). All these climatic strongly hinted at in the contemporaneous oscillations fall within the time range of iso- pollen and sedimentological sequences from tope stage 3 of the deep-sea core sequences, Les Cottés, Les Tambourets, La Ferrassie, which can be seen again as a period of rela- Trou de la Chévre and, further to the northtively complex, sharply fluctuating climate. east, at Arcy-sur-Cure (Fig. 2.30) (Leroi-

To attempt a more specific correlation of Gourhan 1984; Leroyer 1986, 1988, 1990; these interstadials with those recorded in Paquereau 1984; Leroi-Gourhan & Renaultsome of the other climatic sequences summa- Miskovsky 1977; Renault-Miskovsky & rized in Fig. 2.16, however, might be pre- Leroi-Gourhan 1981; Laville et al. 1980). The mature for the reasons already discussed. existence of a major interstadial at this point

The precise dating and correlation of the in the early Upper Palaeolithic is now so various climatic fluctuations recorded in the widely recognized that it has been formally basal part of the Le Moustier sequence designated by Leroi-Gourhan and Renault(between levels A and F) remain for the Miskovsky (1977; also Renault-Miskovsky & present much more debatable. As Laville et Leroi-Gourhan 1981) as the ‘Les Cottés’ inter-

al. (1980: 174-7) have emphasized, these stadial. Indeed, in the latest stratigraphic deposits represent an entirely separate epi- schemes of Laville, this has been elevated to sode in the geological history of the Le the most significant interstadial in the whole Moustier site, dominated by a series of flu- of the last-glacial succession in southwestern viatile deposits laid down during successive France and accordingly designated the ‘interperiods of flooding by the adjacent Vézére stade Wiirmien’ (Laville et al. 1986; Laville river. There is almost certainly a major strati- 1988; Guadelli & Laville 1990). The exact

THE ENVIRONMENTAL BACKGROUND TO MIDDLE PALAEOLITHIC OCCUPATION 49

oe O Birch 6 5 6 ; ge oO S Ss RY © O Grasses > | AP napP| S € © @ #8 + Composites© g 2 @ Pine

x ae ae— | CO FT YY W wf | 0

Ss BS fos st 3 \ So 4! = a) * * baa wee A oO sce| e' * * BSS tO see ccb e [xx| & es O+ =: N sS ._ S~— xEN * xese xBS ie,or Te&oe = oF = L Vk po 3isAA ef SfBFSo 1% 5% 1% | 1% | 1% | 1% | 1% | 1% BSS] 10 % SD % 34 2 %

BS p eA oun tenn nen een ne BS o% see

a a “OL aN a 4 1 x x x 335 LR Sod

Sh ec die bk fee x) 2 8 yf

| pees | ss | ee SS Ses \ soe O & x | x Sos + Q ese | re 1x | x S33 + 0 se = | ) = [ ae S se = ON 2 \ S/S a ”3 & * sos] + 2) Sox -} ES / ! oe 2 i oe ! a = e }* | % * S331 2 Beg / x Sos} +t; on con id © osey ‘ SS cab) LS oso ‘ \4A ote. a + rj} ¢ 1% o3e) > 2 see _ & > $3) RO On, a = |@ASa O ss cQO | Soa @S254 oT O San S33 ‘\ 9" ssf a © * * So Qt Se

| . sy of ce jp LS} SO} BS of O BS

Figure 2.30 Pollen sequence recorded through the later Mousterian and Chatelperronian levels at Arcy-surCure (central France), showing evidence for a major climatic amelioration during the formation of the Chatelperronian levels. This interval is generally referred to as the ‘Les Cottés’ interstadial in France, and assumed (more tenuously) to correlate with the Hengelo interstadial in the Netherlands. After Leroi-Gourhan 1988.

dating of the interstadial remains ambig- come to an end by ca 34,000 BP when there is

uous, mainly owing to the imprecision of clear evidence for the re-establishment of most radiocarbon dates in this time range, extremely cold full-glacial conditions coinbut is generally estimated around 36,000 BP ciding with the appearance of the earliest (Leroi-Gourhan & Renault-Miskovsky 1977; Aurignacian industries in the Perigord Renault-Miskovsky & Leroi-Gourhan 1981). region (Movius 1975; Farrand 1975, 1988, Certainly, the amelioration is known to have Renault-Miskovsky & Leroi-Gourhan 1981).

50 THE NEANDERTHAL LEGACY The obvious correlation, therefore, would be trated record of human occupation during to regard this interval as the equivalent of the the various stages of the last glacial sequence so-called ‘Hengelo’ interstadial, now docu- so far documented in Europe (Sackett 1968; mented widely from many other parts of Laville et al. 1980; David 1985).

northwestern Europe and similarly dated No doubt part of the explanation for this by radiocarbon to around 36-38,000 BP wealth and intensity of Palaeolithic occupa(Figs 2.10, 2.16: Van der Hammen et al. 1967; tion lies in some of the basic topographic and

Leroi-Gourhan & Renault-Miskovsky 1977; geological features of the Perigord region Zagwijn 1990). In absolute terms we should including the abundance of naturally proprobably regard this interval as lying closer tected cave and rock shelter sites, the obvious to ca 39-41,000 BP, if we take into account the protection offered by these sheltered valley accumulating evidence for an approximately habitats against harsh glacial climates and 3,000-year displacement in the radiocarbon the relative abundance and widespread distime-scale over this period (Bard et al.1990a). tribution of high quality flint supplies for In any event, this well defined interstadial tool manufacture (see Chapter 5). It is arguduring the earliest stages of the Upper Pal- able, however, that the most critical features aeolithic sequence must be seen as a further which attracted and supported a high level of major component within the overall sharply human occupation in this region throughout oscillating climatic succession of oxygen-iso- the last glacial period were specific local

tope stage 3. climatic and ecological conditions. As I have discussed more fully elsewhere (Mellars 1985) the main features of these conditions

Southwestern Franceasahuman — 2% as follows:

habitat 1. All attempts to map the distribution of One of the most striking and widely recog- ecological zones throughout Europe under nized features of the archaeological record of the various ‘glacial’ regimes of the Pleistothe Perigord and adjacent areas of southwest- cene period indicate that the southwestern ern France is the sheer wealth and abundance zone of France must have supported the most of the evidence for Palaeolithic occupation. southern areas of essentially open tundra or This pattern is well documented for both the steppe-like vegetation within the European Middle and Upper Palaeolithic periods andis continent (e.g. Butzer 1972: Fig. 51; Iversen

reflected not only in the overall totals of 1973: 16-17). The explanation for this is occupied sites — running into several hun-_ related simply to the highly maritime charac-

dred for both the Middle and Upper Palaeo- ter of the climate in this extreme western lithic (see Chapter 8) —- but also in the density zone of Europe, which keeps summer tem-

and concentration of the archaeological peratures to a much lower level than those remains recovered from the sites (Mellars experienced in more continental zones of 1985). Making allowance for various possible central and eastern Europe and provides corforms of bias in the survival and documenta- respondingly less favourable conditions for tion of the archaeological evidence (such as_ tree growth. As shown in Fig. 2.31, the lat-

the long history of Palaeolithic research in itudinal margin of forest growth tends to the region and the favourable conditions for shift progressively towards the north as one preservation of occupation deposits in cave moves from west to east across Europe into and rock shelter sites) no one would seriously regions of increasingly continental climate question that the southwestern French region and correspondingly warmer summers. The has yielded the richest and most concen- important implication is that these extreme

THE ENVIRONMENTAL BACKGROUND TO MIDDLE PALAEOLITHIC OCCUPATION 51

(ee — 7 ee —— a5 DinTae — = = —— Ban ‘ i OPS v Sset a

7 IER EEE CIEE a NIE 2 | ORAAO ENR , NES PY fe Yi. - ce n on v7 Ever as

rr ae Be TS SS ETSEE EN ES |

0| a200 400| .~—600 miles EGRESS RILSMS TS )i ee CO oN TINT SS RNG RPO = RS OLIN

0 500 1000 iOS GNI A km ORQOAAI INE

fo RBs YK ER LL OSE INES oes eed |

esti aS Oe. xr | SOR aneBRR Fie ern ee Crea rr EIA NEES SIS ee ERE Ip ES SS Pe AD IA SEES Stir cnet

CE Eis RE ON GS EER IS SF CC

BN RE AN EN OIE RON IN ES . . sr

Ne OI IT INN PO EV NI GAS ee AS oo EEE CAR SSAA RR it aORO SE oft CEPR PAV PRS AMES MESGo SEecee

| ~) . eee © “

es PRE RIS S71 ,| en ORR oe BE Rhy NY EAS ooVee ee (SB PeNLA ca Cid ICe SS ON UNEEES ~2>IN AaTES aC SON afASe 6 eee doa” Pk See NeeN

! / ENN ET EAA aARSS EE ne (f -cones , SNR See AML IRIN : A SS oe, Oe a \\ nen Cs EES Se ae Go Fs 6Ck|CllURe, ee ae

(ee se AACR Coe Se)

Le a abate ane eo - cS ~ 8 OS on So Parktundra

ESSE EEE ESS SC lU IS) : Y ot Steppe

Figure 2.31 Reconstruction of the major vegetation zones, ice sheets, and coastlines in Europe at the time of the last glacial maximum, ca 18—20,000 BP, according to Iversen 1973. Note how the zone of open tundra and steppe vegetation extends much further to the south in western Europe than in central and eastern Europe.

southern tundra and steppe landscapes 2. A second factor which would almost cerreceived exceptionally high levels of solar tainly have had a major influence on the radiation (due to their latitude) and conse- overall carrying capacity of southwestern quently would have supported some of the France for animal populations was the relahighest levels of plant productivity within tive oceanicity of the climate and the correthe European continent (Butzer 1972: 463). As sponding mildness, by glacial standards, of a habitat for many open-living herbivorous the winters. Accurate temperature estimates species, including reindeer, horse, bison etc., are notoriously difficult to derive from pal-

these tundra and steppe zones of southwest- aeoecological evidence, but one estimate ern France were probably unique within the based on botanical data has suggested an

last-glacial landscapes of Europe. annual temperature range in the Perigord

52 THE NEANDERTHAL LEGACY region around the time of the last glacial these areas are characterized by significantly maximum from ca 12—15°C insummer to 0°C_ different micro-habitats - with generally in winter (Wilson 1975: 185). Temperatures warmer climatic conditions in the valleys would have risen a few degrees higher than associated with greater protection from wind this during the major interstadial phases of exposure, and richer, deeper and more fertile

the last glaciation. soils than on the higher, more exposed plaThe comparative mildness of the winters teaux (Laville 1975; Duchadeau-Kervazo

would have benefited animal as well as 1982).

human populations in a variety of ways, most This complex interdigitation of plateau notably by reducing the depth and duration and valley habitats helps to explain the ecoof snow cover and by ameliorating some of logical diversity reflected within both the the worst effects of the punishing weather palaeobotanical and faunal evidence from conditions which must have posed a major the southwestern French sites throughout the obstacle to human and animal survival in last glacial period. As Paquereau (1979) has

some of the central and eastern zones of pointed out, the major river valleys in the Europe (e.g. Gamble 1983). Possibly the most Perigord and adjacent areas almost certainly important consequence of these milder win- supported some tree growth throughout the ters, however, would have been to extend the last glacial succession. During the coldest duration of the growing season for most spe- periods this would have been reduced to cies of plants and therefore to increase sub- the hardier species such as pine and birch, stantially the quantities of forage which were perhaps confined mainly to the deeper and available to animal populations throughout more sheltered valleys. During the milder the winter season. This in turn would have interstadial periods, however, this range of

had a major impact on the overall carrying species expanded to include a variety of capacity of the region for animal herds, since specifically warmth-loving species such as it is now generally recognized that it is the oak, elm, lime, hazel and alder. As Paquereau

availability of winter forage, especially dur- (1979) and others have emphasized, we

ing the late winter and earliest spring should visualize vegetational conditions in months, that represents the most critical fac- the Perigord and adjacent areas as a mosaic of

tor in determining the overall density and contrasting communities with the zones of biomass of animal populations which can be woodland extending mainly along the more supported in any region ona long-term basis sheltered valleys and predominantly open,

(Moen 1973: 404-13). tundra or steppe-like vegetation flourishing

over the higher and more exposed plateau 3. Whilst the abundance and concentration areas. of local animal populations was no doubt one This element of ecological diversity is of the critical factors in supporting high equally reflected in the composition of the human population densities, it is arguable faunal assemblages from Middle and Upper that the most important single factor was the Palaeolithic sites (Jochim 1983; Mellars 1985; sheer variety and diversity of local ecological Delpech 1983). As discussed in Chapter 7, conditions within this region (Jochim 1983; many of the sites show a predominance of Mellars 1985). The Perigord region is charac- one particular species (normally either reinterized by two contrasting types of habitat: deer, red deer, horse or large bovids) which the very open, exposed environments on the could indicate either a deliberate selection for extensive limestone plateaux; and the shel- the exploitation of these particular species by tered habitats within the major river valleys the human groups themselves, or alternawhich dissect these plateaux. At present tively, the specific character of the local eco-

THE ENVIRONMENTAL BACKGROUND TO MIDDLE PALAEOLITHIC OCCUPATION 53

logical conditions within the immediate Central (Fig. 2.32) and by sharply reduced environment of each site. However, the most temperatures within these upland areas. As remarkable aspect of the faunal evidence is Raynal and Guadelli (1990) have pointed out, the exceptional range of different animalspe- there appears to be evidence for this in some cles represented in most sites and the waysin of the recent pollen records from the Massif which their remains normally occur, side by and adjacent areas which point to a rapid

side, within the same occupation levels. change in vegetational and associated cliIn economic terms the crucial importance matic patterns between the Massif Central of ecological diversity lies in the degree of and the Atlantic coast. security which this provides against periods The direct effect of this steepening of ecoof occasional failure of particular economic logical gradients from east to west was to add resources (Drury 1975). Even if the popula- a further dimension of potential ecological tions of some animal species, such as rein- diversity to the overall range of environdeer, may occasionally have suffered rapid ments available to the human communities. declines as a result of short-term ecological The patterns of animal migration within this fluctuations — or even overexploitation by the region are still subject to some controversy

human groups themselves (Mithen 1989, (e.g. Bouchud 1966; Spiess 1979; Gordon 1993) — it is likely that the human commu- 1988; Boyle 1990; Pike-Tay 1991, 1993; Burke

nities would still have had access to many 1993) but there can be little doubt, on purely other, complementary species of animals to ecological grounds, that some of the major tide them over these periods of temporary migratory movements of species such as reinresource failure. It is this aspect of diversity deer, and perhaps horse and red deer, were in economic resources which helps to explain along an essentially west-east trajectory the capacity of southwestern France to sup- from the more low-lying and sheltered valley port dense and concentrated human popula- areas of the Perigord and adjacent areas dur-

tions and, apparently, to provide a high ing the winter months towards the higher degree of economic security for them over elevations of the Massif Central and _ its

long spans of time. immediate foothills during the summer. Bouchud (1966) and others, however, have

4. The fourth aspect of ecological conditions argued that any migrations along this axis are

in southwestern France which has been likely to have been on a limited scale, and emphasized in several studies (e.g. Drury unlikely to have extended more than perhaps 1975; Mellars 1985; Raynal & Guadelli 1990) 80-100km between winter and summer ranis the marked compression or steepening of ges. With the sharp compression of climatic ecological zones along an east-to-west axis and ecological gradients along this east-west which characterized this region throughout axis during the coldest glacial episodes, it the last glacial period. The general topog- could be argued that any migrations along raphy and relief of the Perigord and adjacent this axis were even further reduced. The areas is such that there is a natural and fairly implication is that these migratory animal rapid succession of topographic and environ- populations were probably never very far mental zones as one moves progressively from the Perigord region during any part of westwards from the higher elevations of the the annual cycle, and presumably were accesMassif Central towards the coastal Atlantic sible for exploitation by the local human Plain. Under glacial conditions, however, groups within at most a few days of travel these ecological gradients were made even either towards the middle or upper foothill steeper by the presence of local glaciers in zones of the Massif Central, or perhaps into some of the higher elevations of the Massif the coastal Atlantic Plain (Mellars 1985).

54 THE NEANDERTHAL LEGACY

MOULINS |

7.~f.APa | 30

i;

a

o

Y)

25

wa \ GRE ERRAND

CLERMONT

o> Plateau { MO BN) ais. LYON

Ki A aM fw QAP ae Ko Oe Mo EN ae , ; co ORS x un % We fo A) © ere as eS fe @ to we AS Pr SY

> preenes a Se es aD7 a “ed 2. aSlax VCANTAL © rr ay @

oS ee Ae RN po Pe een Ge As : ety Po Sw OI S Gee , me ae AQ a. Vy ff ‘ane’, er oeBae G4orcas ao Oe de = ee,SO OK OU es )|

Sy ~~ ky NSS XN :

“ee

Xarn ~Causse oy , Larzac . | ) du

| : | | WS

pitiee 1 20km 6 . 2Z IF a = NOS Se

7 RU 4 OLA

MONTEELLIER ore

Figure 2.32 Estimated extent of glaciers in the Massif Central region of central France at the time of the last glacial maximum, ca 18—20,000 BP (after Daugas & Raynal 1989). The presence of these glaciers would inevitably have created much sharper east-west gradients in climatic and ecological conditions across the southwestern French region than at the present day.

THE ENVIRONMENTAL BACKGROUND TO MIDDLE PALAEOLITHIC OCCUPATION 95

5. The final point concerns the probable role this is probably the most important single of the major river valleys which traverse the factor. Perigord region as the major, habitual migra-

tion trails of species such as reindeer. This has been discussed in several earlier studies The factors discussed above are very gen(e.g. Bouchud 1966; Spiess 1979) and may be eral features of the ecological and environcritical to understanding many aspects of the mental conditions within southwestern

detailed distribution of both Middle and France, and their precise character would Upper Palaeolithic sites in this region. The have fluctuated sharply and repeatedly durpoint is that by locating settlements or hunt- ing the different chronological and climatic ing locations directly astride these major phases of the Upper Pleistocene. We are still,

migration trails it was possible for human unfortunately, very poorly equipped with groups to intercept animal populations information on some of the more specific deriving from relatively large territories aspects of environmental patterns, most within southwestern France — i.e. the com- notably reliable estimates of temperature bined summer and winter ranges —atasingle conditions, varying snow-fall regimes and location (Mellars 1985: 280). As an explana- the extent of seasonal contrasts in climate. It tion for the remarkable concentration of is against this background that we will examMiddle and Upper Palaeolithic sites at par- ine the archaeological records of the behavticular locations in the Perigord region (for iour and organization of Neanderthal com-

example in certain sections of the Vézere munities within these western fringes of and Dordogne valleys: see Figs 8.1, 8.2) Europe.

CHAPTER 3

Stone Tool Technology

Studies of stone-tool technology have always little doubt that these and other approaches occupied a central position in approaches to to the analysis of stone-tool technology will

the Middle Palaeolithic. The reasons for this continue to provide a central focus of are evident. Here, as in the rest of the Palaeo- research into the behaviour and organization

lithic, stone-tool assemblages provide by far of Palaeolithic communities well into the

the most durable and complete record of future. human development with a degree of conti- The analysis of stone-tool technology can be nuity and fine-scale resolution whichis much carried out at many different levels, each pos-

better than that of the associated faunal ing its own particular problems of methodassemblages and far more complete than that ology and interpretation and each providing of the skeletal remains of the populations rather different insights into the structure and involved. Not surprisingly, the intricacies of organization of the activities which lay behind changes in stone tool flaking techniques and the production of stone-tool residues. The the fine, chronologically patterned changes most basic aspect, which forms the focus of in the forms of stone tools have always pro- this chapter, is the strategies by which the vided the principal framework both for con- available sources of flint or other raw materistructing regional chronological sequences als were systematically reduced into various and for documenting divergences in patterns flakes, blades or other blank forms either for of technological and cultural developmentin immediate use or for subsequent modification

different regions. into a range of retouched tool forms. Second, As a result of more sophisticated approa-_ there is the question of how these initial blank ches to analysis developed over the past two forms were reduced into more regular, exten-

decades, however, it is now clear that sym- sively retouched implement types and the pathetic approaches to the study of lithic significance in functional, stylistic or other technology can go much further. As the terms which can be attached to these morphosubsequent chapters will show, a correct logically or typologically distinct tool cateunderstanding of the nature and structure of gories (discussed in Chapter 4). Equally sigstone-tool technology can shed important nificant are the strategies and procedures by light on the patterns of movement of human which the original sources of raw materials groups over the landscape; the particular were exploited by Middle Palaeolithic groups economic and technological activities carried and subsequently distributed across the land-

out in different sites, and potentially on the scape for use or for further reworking on mental and cognitive processes which lay eventual occupation or special-activity sites behind the production of the tools. There is (Chapter 5). Finally, and currently most con56

STONE TOOL TECHNOLOGY 57 troversial, there is the question of precisely undertaken by Jaques Tixier, for example, what significance should be attached to the and more recently by a number of his stubewildering technical and typological varia- dents (most notably Eric Boéda and Jacques tions which can now be documented over the Pelegrin) working in the Laboratory for Pre-

wide space and time range of the Middle history and Technology at Meudon (Tixier Palaeolithic. Since the latter issues have effec- 1978; Tixier ef al. 1980; Boéda 1982, 1986; tively monopolized much of the literature on Pelegrin 1986, 1990; Boéda & Pelegrin 1983). the Middle Palaeolithic over the past thirty Similar approaches have been developed by years, they will be raised and discussed in Bruce Bradley (1977), Harold Dibble and oth-

some detail in Chapter 10. ers in the United States, by Mark Newcomer

(1971) and Peter Jones (1981) in Britain, and

. by Jean-Michel Geneste (1985), Alain Turq

Primary flaking technology (1989b, 1992a, 1992b), Liliane Meignen and studies of the primary flaking techniques by several other workers in France. Equally which available supplies of flint or other raw important advances over the same period materials were systematically reduced into have come from systematic studies of the suitable blanks or pre-forms for tool produc- detailed spatial distribution of lithic artefacts tion are fundamental to any study of Palaeo- and flaking debris over occupation surfaces,

lithic technology. This is an area in which combined with painstaking reconstructions extensive research has been carried out over of the products of knapping debris (Fig. 3.2), the past ten years and in which some of the to allow insights into the overall sequence most impressive advances have been made. and underlying strategies of the flaking proCredit for these developments belongs cedures involved (Tuffreau & Sommé 1988; mainly to the French school of Palaeolithic Rigaud 1988; Geneste 1988; Révillion 1989; de

studies, stimulated initially by the pioneer- Heinzelin & Haessaerts 1983; Roebroeks ing research of Francois Bordes (Bordes 1947, 1988; Schlanger 1994). All these approaches 1950a,b, 1953a,b, 1954a, 1961a, 1980, 1984; are directly relevant not only to understandBordes & Bourgon 1951). It was Bordes who _ ing the various stages of production of lithic introduced the notion of controlled quantita- artefacts, but to investigating the basic con-

tive approaches to the study of Lower and ceptual and mental processes which lay Middle Palaeolithic technology and thereby behind the sequence of planning and manulaid the foundations for all later systematic facture of stone tools. studies of technological variation and devel- The central concept which underlies all of opment (Bordes 1950a, 1961a). It was Bordes, these recent studies of lithic technology is too, who was responsible for some of the generally referred to in the French literature earliest experimental approaches to lithic as the chaine opératoire approach, and in the technology, involving not only a personal English literature as ‘lithic reduction’ studies mastery of different Palaeolithic techniques (e.g. Tixier 1978; Geneste 1985, 1991; Boéda but also controlled experiments to investi- 1986; Boéda et al. 1990; Bradley 1977). All gate the effects of different flaking strategies these approaches are based on the recognion both the forms of finished tools and the _ tion that the entire process of production and kinds of flaking debris resulting from their shaping of stone artefacts is essentially a

manufacture (Bordes 1947). reductive procedure which passed through More recent studies have developed these several discrete and separate stages (see approaches in a variety of ways. Further Tables 3.1, 3.2; Fig. 3.1). The process started experiments in the systematic replication of with selection of suitable nodules or other

Palaeolithic flaking techniques have been blocks of material which were assessed,

58 THE NEANDERTHAL LEGACY Table 3.1 Principal stages of flake and tool production, use and discard recognized in the reduction sequence scheme (chaine opératoire) of Geneste (1985: 179).

Acquisition stage

Stage 0: Extraction and testing of nodule Production stages

Stage 1: Decortication of nodule Initial shaping of core Preparation of striking platforms

Stage 2: Production of primary flake blanks (flakes, blades etc.) Shaping/retouching stage

Stage 3: Retouching of tools Utilization stage

Stage 4: Use of retouched and/or unretouched pieces Resharpening /reworking of tools Discard stage

Stage 5: Breakage Terminal edge-wear/damage Discard The major products generated during the different reduction stages are listed in Table 3.2.

usually by one or two trial blows, as suitable either in the unretouched state or for systemfor the flaking procedures envisaged.In most atic shaping by retouch into regular imple-

cases it then proceeded through a stage of ment forms. The process could be extended systematic ‘decortication’ of the nodule to by subsequent reworking or resharpening of remove the outer covering of irregular cortex the edges of the tools as they became blunted or skin. Once this procedure was completed, or damaged in the course of use. Finally the there was usually a phase of shaping the core __use-life of the tool was effectively exhausted

into a more or less regular pre-form from and the piece was discarded to form part of which flakes of a preselected shape, size or the accumulating refuse on the site. This is an regularity could be removed. Subsequent sta- idealized sequence and in particular contexts ges in the sequence included successive certain stages of the reduction sequence were removals of these preferred flake forms, either attenuated or even omitted — for exam-

usually accompanied by intermittent epi- ple in some of the simpler, non-Levallois sodes of reshaping or correcting the core to techniques where the initial phase of systemallow further, controlled flake removals. The atic decortication of the nodules was often

final stages in the reduction sequence by-passed. Nevertheless, as a basic concepinvolved selecting specific flakes for use tual framework for analysing the different

STONE TOOL TECHNOLOGY 59 Table 3.2 Principal types of products recognized by Geneste in his reduction sequence (chaine opératoire) scheme for the production of Levallois flakes

Technological Product Type of product

phase no.

0 0 Unworked/tested nodule 1 Cortical flake (>50%)

1 2 Partial cortical flake (30 mm

25 Flaking debris a eee LS SIS

Cs —~ ES aN 1} —_ re Ea ~ eae Ty a = , = \S == € >|] |) BE ee, \y Z Of EEE j):= === A f= =~ ;a7ZZ ! SS |ath , =Ee ) SAI EE \ LEE - == ae LEE - + AM EEE “see Zee zoe IZ = EAA =a —= : —— ee LLE_-=TO = Ee z | SS Z Lion Fo L\

ge 2 oS BE LE gE : LY ZZ NSE iif, BS GE. Z. D 2 Whi Be —— Wy = / a ma SS i 1 I 2 Vif 73 SS BRED 0 KROQ 2 3 \\s SSo SS = Bs ié Be &> ee 3 Be" A : See: Ze AR Z} ZL, Ra ane celles SS \: fs ee anaie = ne AS) SSS Boe4 da a -j. ee oeSING oo > SN SS Ss ~ — ig : SS as CXS ; ee Ga YS , i rehae SS tAN> Ss Ee roa er al Ne See we 7- A ries 32. te

oa= es SbRRASS Yi XG CEM x SN lo ZL Sage

Figure 3.6 Examples of cores oriented towards the production of a single major Levallois flake (illustrating Boéda’s ‘lineal Levallois’ strategy) from the site of Bagorre (northern France). By different patterns of core preparation, flakes ranging in shape from rectangular (no. 1) to triangular (nos 2, 3) could be produced. After Boéda 1988a.

|A

STONE TOOL TECHNOLOGY 65 dé»

ay f a i” > ° a “rn, Fa ~ ie, OS x SS ire aan SON PE /f= io iA — ~ oe, ) > -Z —-~ Zo. — — = = ae = Fa “ ue > i ae ot, —— kr \ feats. ES. SS a SL a, : =. C27. . ae ~ “7 SS Fe ~ZUe rn FFfig SgSS Wi\res

ee ioy se 7 Tl , Toe LEY AN ae ; ae

Zaz oo : BR ae Sa a so a z OO ULB Jy.)=.LT OE " C= aw, er | ae Ne

\ | ays ad fc. by | — ‘\ Me 4 __

y \ IE’

E alff7A ja-\~f>\!i E is fA Vy3 Se

4 | (\ C ~ KV

VES B he = JT, 3 IE he / v5 UE - | | EER 7 Ts an, ‘co 7 EX ES eo oA \ \\mn 4 ~~P j=_ re=—— ~ WZ ie A ap ‘

FE = 2-™

“ES Z wwe ~ | NZS 0a ae | Sem < LH { L =a Br» SS y W YS

Figure 3.20 Examples of blades from the site of Seclin (northern France) dated by thermoluminescence to ca 90,000 BP. The two pieces on the bottom right-hand-side are examples of typical crested blades (‘lames-a-créte’) used to initiate the sequence of blade detachments from the core, and closely resembling techniques used on Upper Palaeolithic sites. After Révillion 1989.

82 THE NEANDERTHAL LEGACY

‘ 18 ET carn

a Ae AD . | ea = : . | i= , Ly sas 45 SS | 4 RES EI 4 \ E f

, = Te Z : Z ae ) NESE B yy — 2 al \) SESE

WY —y 7 |iy— ‘\ NES Sal

SxS —_ O ascma a

Q (AG - 5 oN

=| : HI Th zy |

of) |

; is Ki TG Ra 7 3 ia TP] Hl SY XQ E AEE (tN 5: f : LAH ae “EP OAR OEE CHA AY eH EBf)Fo LEN a wee 1 | fe i BN ES

A A Ey J f= Ro J BRE Y RES Kee RE SV 4 5 eRe ¢ F)

S gy E PAF a ewal i FF EX =e)a¢EX +

att 6 é 7

Figure 3.21 Blade cores from the site of Seclin. Nos 1-5 are examples of Boéda’s ‘specialized Levallois’ blade cores (see Fig. 3.19), while nos 6-7 are examples of fully prismatic cores, closely similar to Upper Palaeolithic forms. After Révillion 1989.

STONE TOOL TECHNOLOGY 83

aN yj, meet’ Yee UTE SS] Se RP,aefoEN ON KE>.ae‘ = SS ~ i: oo ney ft} : AOA Saat hs ~. [oe.- ¥ SRA) Th ERS RE Gly ley

: ; p— = /KNS 7) EE we 3 AAW)~/ YEE BSEen Vr A

a fo Gib II = Ey ‘if

= XY RON — ;2 A aa S - ANNE:

EN EZ a Ye GG Zo ="See, ea mS | rg SS - Sy =" Yo. C1) B ' BSS SR ae Se —— = NB 6 EA Fit Aa RSs 72 -W (fae S\ 2a ee oe WSS Ye SS Ba =e - ee “he | PS BS rm 2 POs \A\\A WEESBS fapes Alf | We ae cae ries Op a leg | OIN MNRAS BT Eh Pee gf Ne A 7 hams _ on CR ee = Se te oo Lf ll be = : Gyr {; °.4 RNY Se Se BE. =S== ae es 2: ‘ oe FD stir, Y 7

=n2an BEERS. SiGea . AN®SS —SSS , — 2. cs EE... So . =Seay KF z=——— peices eo, NN =—. S\ Zu SS BS SS... Fee) SS fesse: ae ae Se,he \ =e SX, SS. SQ OS pie 4RRS fi: MEM errr cL EY RI N\QE 2 INS. if ey PEDERI ah aya age

ey VRE 0) ee ae NWR 4 aa. Ye, Se ewe \|\\Sep careAi Y's, ee Bike ees BBs EBS) OS AIRS ee ey

@ Se\\ay /\ oe|PESOS BR CN yeWy \i =e“cae mee, °° ewe Yh WE ci eR CREE, ( NSS Paes yas) Speers SN xy eee, ANS SA AN | WN Wi [} Lj } \

\DE dy 7|SSIS ! ‘es 7 \\Dances N _ . eer i, exe ares ons 6 =vee) Geshe

Figure 4.2 Typical transverse racloir forms — characterized by the presence of a retouched edge located opposite the bulb of percussion and striking platform of the parent flake. After Bordes 1961a.

(variously applied to wood, meat or skin) to. clearest reflection of this can be seen in the more heavy-duty scraping of hide or bone. complex system of typological divisions for Clearly, the functions of these tools were various racloir forms proposed in the standmuch more flexible than their conventional ard typology of Francois Bordes (1961a).

English and French names would imply. Leaving aside the more complex forms of When defined in these terms thereis scope convergent and déeté racloirs, Bordes advofor relatively wide variation in both the pre- cated 17 separate divisions for the simpler cise forms of side scrapers and in the charac- racloir forms (see Table 6.1). The major dister and treatment of the retouched edges. The _ tinctions in Bordes’ system are based on the

TOOL MORPHOLOGY, FUNCTION AND TYPOLOGY 99 position of the retouched edges in relation to mented variations in the forms of the tools the main flaking axis of the tool and on the — either in terms of function or specific overall shape of the retouched edge itself. design norms or ‘mental templates’ that Thus a basic distinction is made between lay behind the conception and production pieces where the retouched edge is aligned of the tools (Dibble 1987a,b,c, 1989; Dibble parallel to the main axis of the flake (defined & Rolland 1992; Kuhn 1992a)?

as ‘lateral racloirs’) and those where the retouched edge is oriented transversely As discussed earlier, the question of tool across this axis (‘transverse racloirs’). Further reduction and resharpening has emerged as a subdivisions of these types are based on the central issue in Middle Palaeolithic technol-

forms of the retouched edges - whether ogy with radical implications not only for the straight, convex or concave. In addition to morphology of individual tool forms but also these basic forms Bordes recognized separate for the more general issue of inter-assemcategories of tools retouched either exclu- blage variation in the Mousterian (see Chapsively or partially on the ventral surface of ter 10). In a series of publications over the the flakes (racloirs sur face plane or a retouche past 15 years Nicholas Rolland and Harold

alterne) and other forms with extensive Dibble have presented a stark alternative to retouch on both the dorsal and ventral sur- conventional perspectives on side-scraper faces (racloirs a retouche biface and racloirsados typology by suggesting that these tools may

aminci). Tools with characteristic retouch on never have been planned as retouched tool two separate edges are grouped together col- forms but may simply have acquired their lectively as double racloirs, and further sub- characteristic retouch in the course of prodivided into six categories based on various’ gressive resharpening of their edges during potential permutations in the curvature of use (Rolland 1977, 1981, 1988a, 1990; Dibble the two retouched edges (see Table 6.1; Figs 1984a,b, 1987a,b,c, 1988a,b, 1989, 1991a,b;

4.1, 4.2). Rolland & Dibble 1990; Dibble & Rolland From these distinctions it is clear that the 1992). They envisage that the great majority manufacture of even relatively simple racloir of conventional racloir forms started their

forms left Middle Palaeolithic artisans with use-lives simply as unretouched flakes, much flexibility in their choice of both overall which were only systematically retouched as

design (i.e. the number and location of the their originally sharp edges became progresretouched edges) and the precise patterns of sively worn and damaged through heavy retouch applied to the worked edges. Recent use. The logical extension of this is that effecdebates on the significance of these formal tively all the documented variation in the

variations in racloir morphology have relative frequencies of racloirs in different ,

focused on two basic issues: Middle Palaeolithic assemblages can be 1. How far was the retouch applied to attributed largely to variations in the degree

to awhich thesefeature tool-resharpening processes racloirs deliberate applied in the . , ; ;Thus ainitial , or were carried out in the different sites. stages of tool production and how wae taxonomically Denticulate Mousterian far was it simply a result of ad hocdustries, resharpLo the overall frequencyinin which of ening and reworking theseen edges they aereduced | . . racloirs is low,ofare asaslightly

became progressively. wornand bluntedin . ol were ;| industries in which few raw flakes

the course of use (Rolland 1977, 1981; Dib- transformed into retouched racloir forms ble 1984a, Rolland & Dibble 1990; , ’ ; while the various Ferrassie andor Quina-type Dibble &1987a; Rolland 1992)? .; industries (in which overall racloir frequency

2. What significance can be attached todocu- is high) are seen as heavily reduced indus-

100 THE NEANDERTHAL LEGACY

a7) Ze) pe | Oe)) A 1: y} Z ILe, “ Lyon He Wad Win — - [;| ni a Uf ee re

-F\

are: cat Qe=a Ame ] we {- =a _ P-7, ™ a~Ar a aeTe saree - in

— th — -H] 2wo (5) — Gn _Wy wy =an} Uh " 8 Seat \ FA cays ; Vy

Ny LY,ae , . SP ZY SY aY/ re Tir BaéLp TSS US

Cc > \_—

~ Oy “4 ty ‘i : & Was By- A - {" =| Ss = Bile) Qk)- EE NS_@i RE

|Ow” = SS a A aa~ va) TFI \ |

Bo By me vars pi 2 n 7 beat f We = df \ = Names me= fo A 7, (BY SPT TLR N WR: % [ISS \ SS (tt Eh.|

Wi Z a mY ex yt : ye ve \| NS u \ ~~, Su zip

: = ee aa a | }

Figure 4.4 Resharpening spall refitted to the edge of the original tool, from the Cotte de St Brelade (Jersey). After Cornford 1986.

TOOL MORPHOLOGY, FUNCTION AND TYPOLOGY 101 tries in which a large proportion of the cases have been refitted directly to the parent available flakes were transformed into sys-_ tools (e.g. Fig. 4.4). These discoveries leave tematically resharpened racloir forms (Roll- no doubt that systematic resharpening of and 1977, 1981, 1988a; Rolland & Dibble 1990; blunted or damaged edges of side scrapers

Dibble & Rolland 1992). was carried out commonly on a range of These patterns in turn are thought to be Middle Palaeolithic sites. closely related either to environmental fac- The central issue here is the relative scale on tors (reflecting the relative abundance and’ which resharpening was carried out in Mid-

accessibility of local raw material supplies) dle Palaeolithic sites. Can we use these or to varying degrees of duration and inten- resharpening models to argue that typical sity of occupation in the different sites (Dib- side-scraper forms were rarely if ever manuble & Rolland 1992; Dibble 1991a,b). These factured as deliberate, a priori tool forms (as issues will be discussed in more detail in Rolland and Dibble imply) or was resharpenChapter 10 in the context of more general ing applied purely as a secondary technomodels of inter-assemblage variation in the logical device to extend the use-life of tools Mousterian. The assumptions of these mod- conceived and produced from the outset as els, however, are directly relevant tothe pres- retouched tool forms? The main points are as

ent issue of how far tool-reduction models follows:

can De invoked to account for the patterns of 1. The evidence from many sites does not ° There is no doubt that Rolland and Dibble SUPPO™ easily the nouion that systematic have identified an important potential aspect reeuction of raw Hake me cnaracteristi¢

of variation in Middle Palaeolithic technol- rac OM FORMS Was & PrOGUes OF EM ACT Inten-

ogy which is in line with the findings of much on patterns of site occupalion or the scarcity recent research. Recent results of microscopic of local raw materials. There ae now several

use-wear studies, for example, leave no examples of sites where high frequencies of doubt that many specimens of unretouched typical racloirs coincide with locations where flakes in Middle Palaeolithic industries were local raw materials we _ both ot high tlaking

subjected to heavy use, apparently for a quality and apparently available in almost range of different functions. Amongst others, am wes SUPP ae vemos! pvious CXAMNP le

the studies carried out by Sylvie Beyries on is the si seh , 1: - Cy Winer a meyer the assemblage from Corbehem in northern outcrop of Nigh quality wnt occurs directly

France (Beyries & Boéda 1983) and similar on NOS). th site Meyrony 1s pour studies by Patricia Anderson-Gerfaud (1990) Oo am lost - Be, ach ‘(Farizy 1985) wg on assemblages from several southwestern Biach a Srint-Vaast in Srthorn Frames (Tuf-

French sites leave no doubt about this (see ,

also Beyries 1986, 1987, 1988a,b, 1990; Keeley {"eav & Somme 1988), there would have been

1980). There is also now clear evidence from oy amp < capacity Lor producing rreshy seed several sites that the retouched edges of typ- con h BOAES BOE echare, ne ee “ak nee

ical racloir forms were systematically i? y intensive Tes va gem © qd i‘ “S h resharpened by the removal of deliberate © Whe it should he heen mrougs

resharpening spalls (Fig. 4.3). Typical speci- meee NY it shou aye een MECES Say NN mens of these resharpening flakes have been these contexts to apply highly oO LORMAN recovered, for example, from Combe Grenal strategies in the use of available flint supplies (Lenoir 1986), Marillac (Meignen 1988), La is by no means clear. Cotte de Saint-Brelade (Cornford 1986) and 2. Similarly, there are now many sites where La Micoque (Schlanger 1989), and in some a high overall frequency of racloirs (in rela-

102 THE NEANDERTHAL LEGACY tion to other tool forms) can be seen to assemblages comprising the lowest overall coincide with an equally high frequency of percentages of racloirs (i.e. those of the Denunretouched flakes. At Biache-Saint-Vaast, ticulate variant) the average lengths of the for instance, an industry comprising over 60 side scrapers present in the assemblages are percent of retouched racloir forms occurs in’ shorter than those recorded in some of the the context of an assemblage which is domi- most heavily reduced industries, such as nated by unmodified Levallois flakes (Tuf- those of the Ferrassie and Quina Mousterian freau and Sommé 1988). A similar pattern is variants (Fig. 4.5; Table 10.1). This has been reflected at Champvoisy (Marne: Tuffreau clearly documented, for example, by Rolland 1989b) and again at Champlost in Burgundy himself for the various levels of Denticulate, (Farizy 1985: 406). In these and other cases it Quina and Ferrassie Mousterian at Combe is difficult to see why intensive retouching Grenal (Rolland 1988b: 173, Table 9.4B) and is and resharpening of tools should have been equally apparent in similar assemblages from needed when there were large numbers of the Abri Chadourne, La Quina, Hauteroche, unretouched flakes immediately accessible Arcy-sur-Cure and elsewhere. Exactly how and readily available for use on the occupa-_ these observation can be reconciled with the

tion sites. hypothesis that the extent of systematic

resharpening and reduction of tools is actu3. Even more problematic in terms of the ally much greater in the Quina and Ferrassie racloir-reduction models is the available data industries than in those of the Denticulate on the relative sizes of side scrapers in dif- variant (i.e. how systematic reduction can ferent types of Mousterian assemblages. For somehow make the tools larger) has yet to be example, several studies have shown that in explained.

} | Layer:17 "

21 / 23 | ff 25 7.

Quina/Ferrassie

27 ! z= 35 ! a

Figure 4.5 Mean lengths of racloirs Denticulate | | recorded in various levels of Ferrassie, Layer:13

Quina and Denticulate Mousterian at 14

Combe Grenal, as documented by Rolland

(1988b: Table 9.4b). The larger sizes the 16Ferrassie and | 20 a racloirs recorded inofthe Quina Mousterian levels clearly conflicts

with the notion that these are heavily , : — ‘reduced’ versons of the tools encountered 20 30 40 80 60 70 80 in Denticulate industries. Mean length of racloirs (mm)

TOOL MORPHOLOGY, FUNCTION AND TYPOLOGY 103 4, Finally, it should be emphasized that there sites. But in most of the simpler flaking stratare strong a priori reasons for assuming that egies (such as those in many Quina or discin many contexts racloirs would have been core strategies) it was only by systematically produced as an essential and quite deliberate modifying the edges of the original flakes policy in general tool-making strategies. All that these two objectives could be effectively studies of racloir morphology have recog- achieved (Turq 1988b, 1989b, 1992a). Direct nized that two major objectives lie behind the illustrations of these strategies can be seen in

production of typical racloir forms: first, to the detailed patterns of retouch applied in secure the maximum possible length of work- the shaping of racloirs in several sites. At ing edge from the available flake blanks, and Biache-Saint-Vaast it can be seen that the second, to impose a regular, smooth form on retouch on the edges of many of the racloirs this worked edge (Bordes 1961a; Mellars was applied ina very sparing, discontinuous 1964; Dibble 1987a). A moment’s reflection way apparently intended either to remove will show that the only way to achieve these some obvious irregularity on the original objectives simultaneously on a high propor- edge of the flake or to extend the effective tion of flake blanks is by applying systematic length of the working edge down the maxretouch to the edges of the tools (Fig. 4.6). In| imum possible length of the tool (Tuffreau cases where primary flaking strategies pro- and Somme 1988). In these and many other duced elongated, regular flakes (forexample cases it seems that retouch was applied not in some of the Levallois strategies discussed merely to rejuvenate heavily worn and damearlier) then substantial numbers of flakes aged edges but asa deliberate policy to maxiwith naturally long, regular edges would mize the inherent potential of the available have been readily available for use on the flakes for the specific functions envisaged.

__ ~~ ql —(

Wa

A

ave) application of retouch can substantially increase NG Figure 4.6 Schematic illustration to show howthe the (7 effective length of the working edge in the production

ws of a typical ‘racloir’ form.

104 THE NEANDERTHAL LEGACY None of this is meant to deny either the 1984) and others have pointed out, the relaextensive use of totally unretouched flakesin tive frequencies of these two forms show many Middle Palaeolithic industries nor the some striking variations between different reality of systematic resharpening of certain industrial variants of the Mousterian. In partools as a way of extending their natural use- ticular, very high frequencies of transverse lives. As noted above, both features can now _ racloirs seem to be especially characteristic of be documented in the archaeological material the classic Quina-type industries in western from several sites. But to suggest that allthe Europe (Fig. 6.12), and of the Yabrudian abundant and highly varied racloir forms industries in the Middle East (Bordes 1955b,

encountered in Lower and Middle Palaeo- 1984). In most other industries typically lithic industries can be dismissed as opportu- transverse forms are relatively rare and nornistic end-products of these resharpening mally account for only about 5-10 percent of

sequences would be not only in conflict with racloir forms in general. , several specific features of the archaeological Most workers in the past (including evidence but contrary to most reasonable Bordes) have assumed that the production of expectations of tool production strategies. transverse as opposed to lateral racloirs represented a deliberate decision on the part of

Lae . | the flint workers controlled largely by the

Variations in side-scraper forms forms of the original flake blanks selected.

Where available flakes were relatively long Even if we accept that most side-scraper in relation to their breadth (as for example in

forms were produced from the outset as many Levallois industries) the longest workretouched tools, this still leaves much varia- ing edges on the tool could usually be tion to be explained within the overall side- obtained along the main, longitudinal flaking scraper range — i.e. the contrast between axis of the original flake (Bordes 1961b: 806, lateral and transverse types, single versus 1968a: 101, 1977: 38, 1981: 78-9, 1984: 164; double-edged forms, variations inthe shapes Bordes & de Sonneville-Bordes 1970: 61; Turq and treatment of the retouched edges (Figs 1989b; Mellars 1967, 1992a). By contrast, 4.1, 4.2). What significance can be attached to. where the original flake blanks were rela-

this variation, either in terms of deliberate tively broad in relation to their length (for design norms in the initial production of the example in most of the Quina-type flaking tools or in terms of subsequent use and re-_ strategies discussed in the preceding chap-

sharpening? ter) then it was more often possible to obtain

Debates on the significance of varying fre- the maximum length of working edge by quencies of transverse versus lateral racloir retouching the flakes on the transverse marforms in different Middle Palaeolithic indus- gin, directly opposite the striking platform. tries provide a classic illustration of the kind The distinction between lateral and transof issues generated not only by studies of verse racloirs represented a simple technoside-scraper forms but by studies of Middle logical decision dictated by the form of flakes Palaeolithic tool morphology in general. The immediately available for tool manufacture distinction between these two forms rests (see Fig. 4.7). strictly on the location of the retouched edge A sharply conflicting interpretation of the in relation to the main flaking axis of the tools transverse/lateral racloir distinction has — i.e. essentially parallel to the flaking axisin been put forward by Dibble in the context of the case of lateral racloirs and at right angles his general tool-reduction models in the Mouto this axis in the case of transverse forms sterian (Dibble 1984a,b, 1987a,b,c; Dibble & (Bordes 1961a). As Bordes (1961b, 1968a, Rolland 1992 etc.). His hypothesis is that

i

BO pe , _

TOOL MORPHOLOGY, FUNCTION AND TYPOLOGY 105

C1 Ferrassie

40 + %

= ef | ——e

7 oe © 8

ie a oP Figure 4.7 Relationship between the S e° relative frequencies of transverse versus Qo e@ lateral racloirs, and the variable 3 “Or 9@ utilization of Levallois flaking | G O techniques (as reflected in the Levallois - mM Index) recorded in Ferrassie and Quina O 4 Mousterian industries in southwestern 10 f France. The data suggest strongly that I IE Nog O the production of transverse as opposed

to lateral racloir forms was dictated mainly by the shapes of the original flake

pL _ 30 blanks by theMellars different flaking 0 10| |20 40 50produced strategies. From 1992a: Fig. Levallois index 2.2, with additions.

most, if not all, transverse racloirs started life thickness to total tool-area and generally initially as more elongated lateral racloir heavier retouch along the worked edges (Fig. forms (either with or without deliberate 4.9) (Dibble 1987a: Fig. 1, Table 1). He argues retouching of the utilized edges) and were’ that these features are consistent with the

only transformed gradually into typical idea that transverse racloirs were initially transverse forms as a result of repeated much larger in their original form than in resharpening of the worked edges as they their final, discarded, form and appear to became progressively damaged or blunted show direct evidence for this heavy reducby use. According to this model, the orienta- tion of the tools in the invasive character of tion of the worked edges shifted progres- the retouch along the worked edges. sively from a lateral to a transverse orienta- As in the more general racloir-reduction tion during successive phases of reworking arguments discussed above there may well and reduction. Dibble’s own representation beanelement of truth in Dibble’s arguments. of this transformation is illustrated in Fig.4.8. There is no doubt that many typical speciIn support of this hypothesis Dibble advan- mens of transverse racloirs do show evidence ces a number of observations on the metrical of apparently heavy and perhaps repeated features and retouch characteristics of lateral resharpening of the worked edges (e.g. and transverse racloir forms drawn from Lenoir 1986; Meignen 1988) and it may well sites in both southwestern France (La Quina, be that in progressive resharpening some Combe Grenal, Pech de l’Azé) and the Middle racloirs were occasionally transformed from East (Bisitun) (Dibble 1987a,b,c). He argues lateral to transverse forms. The question, that transverse racloirs tend to show larger again, is the scale on which this technological striking platforms than those on most lateral transition occurred. Does this transformation racloirs together with higher ratios of flake- account for almost all documented trans-

106 THE NEANDERTHAL LEGACY

eo ms. Oe YRS ce 3oseee - ' fo N

a ee Fees \(;gel nTAA, yee , i , ae MO he * ke ” : °} \ 4 sy “ ‘ 2 . \

ae , moe ey my 4.) ; ; | dca Gi

AB|

; ; a a Sy 7, & . Pe ners aia:

ee - oat te, ee nae KU

SPC saree FED, os QQ

:=- m 2 :S. aSS : “ (\ ahd,IN ~ eee -fee i. 7og X 8ft RAE - Seer .— ooae ny he PR,a2. “ Ay a. Og A 4o

Le BoE AG, SL \

CD

Figure 4.8 Dibble’s hypothetical reconstruction of the transformation from lateral to transverse racloir forms, in the course of successive episodes of resharpening the edges of the tools. Note how the overall length of the tool reduces progressively in the course of this transformation. From Dibble 1987a.

verse racloir forms, as Dibble apparently verse racloir forms in most Middle Palaeoimplies, or was it a relatively rare and atypi- lithic assemblages? The relevant issues are as

cal occurrence which accounts for only a_ follows: small percentage of the documented trans-

-ff{foe o | / / Jf |: Yo A | ffs //f/f. | [lowe

TOOL MORPHOLOGY, FUNCTION AND TYPOLOGY 107

100 5 100 7% ; 75 4 ! Ye 754 li:

504 [lf « 504 fod

of - are

25 4 / / a 25 4 } f foo

Awe MATa __ om ho m h BISITUN LA QUINA


\ A t! o oO fe) 00 0 ° AD fol a re) e * a? 7st 7 e

Se : ese .

. on pitsJr Fesee o 2*. °e. stots ees ae

OO co. eeey rm sf° °° ”°oo oO o °e? SEE .| ee materi oyee °?fitienee oue rar °° °°? oo ?°°°oO ao ®Ps ‘ at sassotss sti | spe Raia ° 7s ao!fO Kee SESE iar eH -ak |. -oor! i BAO o0Oo °o2%” ooo)r* e*

Saeco a] eet te | pases states SSS tetas ma) v7 “I ib 2 °° oo 2 rs) o ? ° sisiese 2388 see aot state rete! reret cote tage eatatatate petatas sretgete feral fret sores: aceatese Ris eats a islet sete’ Se Pee cates eens Pie cata foe wt 47 ° fe] o