Muge 150th : The 150th Anniversary of the Discovery of Mesolithic Shellmiddens—Volume 1 [1 ed.] 9781443884167, 9781443880077

Muge 150th: The 150th Anniversary of the Discovery of Mesolithic Shellmiddens is organized into two volumes. This first

178 106 7MB

English Pages 408 Year 2015

Report DMCA / Copyright

DOWNLOAD PDF FILE

Recommend Papers

Muge 150th : The 150th Anniversary of the Discovery of Mesolithic Shellmiddens—Volume 1 [1 ed.]
 9781443884167, 9781443880077

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Muge 150th

Muge 150th: The 150th Anniversary of the Discovery of Mesolithic Shellmiddens—Volume 1 Edited by

Nuno Bicho, Cleia Detry, T. Douglas Price and Eugénia Cunha

Muge 150th: The 150th Anniversary of the Discovery of Mesolithic Shellmiddens— Volume 1 Edited by Nuno Bicho, Cleia Detry, T. Douglas Price and Eugénia Cunha This book first published 2015 Cambridge Scholars Publishing Lady Stephenson Library, Newcastle upon Tyne, NE6 2PA, UK British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Copyright © 2015 by Nuno Bicho, Cleia Detry, T. Douglas Price and Eugénia Cunha and contributors All rights for this book reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the copyright owner. ISBN (10): 1-4438-8007-8 ISBN (13): 978-1-4438-8007-7

TABLE OF CONTENTS

Foreword .................................................................................................... ix List of Contributors .................................................................................... xi Chapter One ................................................................................................ 1 Carlos Ribeiro and Francisco António Pereira da Costa: Dawn of the Mesolithic Shell Middens of Muge (Salvaterra de Magos) João Luís Cardoso Chapter Two ............................................................................................. 19 The Archaeological Excavations at Muge Shell Middens in the 1930’s: A New Contribution to the History of its Investigation Ana Abrunhosa Chapter Three ............................................................................................ 33 Sources for the Reconstruction of Cabeço da Arruda Mary Jackes, David Lubell, Pedro Alvim and Maria José Cunha Chapter Four ............................................................................................. 45 Cabeço da Arruda in the 1860s Mary Jackes, David Lubell, H. Cardoso, José Anacleto and Chris Meiklejohn Chapter Five .............................................................................................. 59 Antler Debitage in Muge Shell Middens: The Collections of the Geological Museum Marina Almeida Évora Chapter Six ................................................................................................ 77 Reading the Lithics in Flint of Cabeço dos Morros Shell Midden Anabela Joaquinito

vi

Table of Contents

Chapter Seven............................................................................................ 89 Marine Invertebrates and Models of Economic Organization of the Coastal Zone during the Mesolithic: French and Portuguese Examples Catherine Dupont and Nuno Bicho Chapter Eight ........................................................................................... 105 What’s New? The Remains of Vertebrates from Cabeço da Amoreira— 2008-2012 Campaigns: Preliminary Data Rita Dias, Cleia Detry and Alexandra Pereira Chapter Nine............................................................................................ 119 Preliminary Techno-Typological Analysis of the Lithic Materials from the Trench Area of Cabeço da Amoreira (Muge, Central Portugal) João Cascalheira, Eduardo Paixão, João Marreiros, Telmo Pereira and Nuno Bicho Chapter Ten ............................................................................................. 135 New Functional Evidence for Human Settlement Organization from the Mesolithic Site of Cabeço da Amoreira (Muge): Preliminary Lithic Use-Wear Analysis João Marreiros, Juan Gibaja, Eduardo Paixão, Telmo Pereira, João Cascalheira and Nuno Bicho Chapter Eleven ........................................................................................ 147 Raw Material Procurement in Cabeço da Amoreira, Muge, Portugal Telmo Pereira, Nuno Bicho, João Cascalheira, Célia Gonçalves, João Marreiros and Eduardo Paixão Chapter Twelve ....................................................................................... 161 The Midden Is On Fire! Charcoal Analyses from Cabeço da Amoreira (Muge Shell Middens) Patrícia Diogo Monteiro, Nuno Bicho and Lydia Zapata Chapter Thirteen ...................................................................................... 177 Fire and Death: Charcoal Analyses from Two Burials in Cabeço da Amoreira (Muge Shell Middens) Patrícia Diogo Monteiro and Olívia Figueiredo

Muge 150th Volume 1

vii

Chapter Fourteen ..................................................................................... 185 The Importance of New Methodologies for the Study of Funerary Practices: The Case of Cabeço da Amoreira, a Mesolithic Shell Midden (Muge, Portugal) Olívia Figueiredo and Célia Gonçalves .

Chapter Fifteen ........................................................................................ 199 The Mesolithic Skeletons from Muge: The 21st Century Excavations Maria Teresa Ferreira, Cláudia Umbelino and Eugénia Cunha Chapter Sixteen ....................................................................................... 209 Life in the Muge Shell Middens: Inferences from the New Skeletons Recovered from Cabeço da Amoreira Cláudia Umbelino, Célia Gonçalves, Olívia Figueiredo, Telmo Pereira, João Cascalheira, João Marreiros, Marina Évora, Eugénia Cunha and Nuno Bicho Chapter Seventeen ................................................................................... 225 Tracing Past Human Movement: An Example from the Muge Middens T. Douglas Price Chapter Eighteen ..................................................................................... 239 Cranial Morphology and Population Relationships in Portugal and Southwest Europe in the Mesolithic and Terminal Upper Palaeolithic Christopher Meiklejohn and Jeff Babb Chapter Nineteen ..................................................................................... 255 The Late Mesolithic of Southwest Portugal: A Zooarchaeological Approach to Resource Exploitation and Settlement Patterns Peter Rowly Conwy Chapter Twenty ....................................................................................... 273 Living On the Edge of the World: The Mesolithic Communities of the Atlantic Coast in France and Portugal Grégor Marchand Chapter Twenty One................................................................................ 287 Settlement and Landscape: Poças de São Bento and the Local Environment Lars Larsson

viii

Table of Contents

Chapter Twenty Two ............................................................................... 301 At the Edge of the Marshes: New Approaches to the Sado Valley Mesolithic (Southern Portugal) Pablo Arias, Mariana Diniz, Ana Cristina Araújo, Ángel Armendariz and Luis Teira Chapter Twenty Three ............................................................................. 321 Lithic Materials in the Sado River’s Shell Middens: Geological Provenance and Impact on Site Location Nuno Pimentel, Diana Nukushina, Mariana Diniz and Pablo Arias Chapter Twenty Four ............................................................................... 333 High Resolution XRF Chemostratigraphy of the Poças de São Bento Shell Midden (Sado Valley, Portugal) Carlos Simões, Eneko Iriarte, Mariana Diniz And Pablo Arias Chapter Twenty Five ............................................................................... 347 New Preliminary Data on the Exploitation of Plants in Mesolithic Shell Middens: The Evidence from Plant Macroremains from the Sado Valley (Poças De S. Bento And Cabeço Do Pez) Inés L. López-Dóriga, Mariana Diniz and Pablo Arias Chapter Twenty Six ................................................................................. 361 Lithics in a Mesolithic Shell Midden: New Data from the Poças de São Bento (Portugal) Ana Cristina Araújo, Pablo Arias and Mariana Diniz Chapter Twenty Seven............................................................................. 375 Pots for Thought: Neolithic Pottery in Sado Mesolithic Shell Middens Mariana Diniz and Miriam Cubas

FOREWORD

In 1863, Carlos Ribeiro and his team from the Portuguese National Geologic Services discovered the first Mesolithic shell midden in the Muge area (central Portugal) in the Tagus Valley. In the following years and decades, a dozen or so of other similar sites were found in the region. In the following century the same pattern of discovery took place in the Sado Valley. The two Mesolithic complexes composed the main bulk of knowledge on the Iberian Mesolithic up to the end of the 20th century. Those discoveries are well known internationally, not only due to the considerable effort that the Portuguese academics put forth for the international recognition of the importance of those sites, including the 1880 World Congress of Archaeology and Ethnology, but also due to the very high number of human burials (over 300) excavated during the last 150 years. Despite a century and half of research, until recently those sites were poorly known and very little data were analyzed and published, at least at the international level. In 2013, celebrating 150 years of the discovery of the Muge Shellmiddens, we were able to organize an international conference, MUGE150th - The 150th Anniversary of the discovery of the Mesolithic Shell Middens, held in Salvaterra de Magos in March, 2013. There, recent results from the Mesolithic of both the Muge and Sado regions were presented, together with important and new results and data from Mesolithic research around Europe. The proceedings of the MUGE150th - The 150th Anniversary of the discovery of the Mesolithic Shell Middens is organized in two volumes and includes most papers presented in the congress covering the Mesolithic sites from many regions in Europe, as well as a set of general syntheses on the state of the art on specific topics such as the use of Isotopes in diet determination and migration, plant use, and burial practices. In addition to these papers, there is also a set of papers on the general topic of shell middens. Volume 1, now published by Cambridge Scholars Publishing, focuses only on the Mesolithic of the Muge and Sado Valleys, with a total of 27 chapters. The second volume, to be published soon, will integrate all the remaining papers, covering an array of topics and regions.

x

Foreword

The organization of the conference and the publication of the Proceedings was only possible due to the help of Casa Cadaval, where many of the Muge Shell Middens are located, the Câmara Municipal de Salvaterra de Magos, and the CIAS - Research Centre for Anthropology and Health, from the University of Coimbra.

LIST OF CONTRIBUTORS

Abrunhosa, Ana ICArEHB Faculdade das Ciências Humanas e Sociais Universidade do Algarve Campus Gambelas. 8005-139 Faro, Portugal [email protected] Alvim, Pedro Department of Archaeology, Durham University, Durham, UK, DH1 3LE [email protected] Anacleto, J.A. 3Museu Geológico, Rua Academia das Ciências, Nº. 19 – 2º. 1249-280, Lisboa, Portugal. [email protected] Armendariz, Ángel Instituto Internacional de Investigaciones Prehistóricas (IIIPC), Universidad de Cantabria, Edificio Interfacultativo, Avda. de los Castros s/n, Santander 39005, Cantabria, Spain Araújo, Ana Cristina Direcção Geral do Património Cultural/LARC; EnvArch/CIBIO/InBIO [email protected] Arias, Pablo Instituto Internacional de Investigaciones Prehistóricas (IIIPC), Universidad de Cantabria, Edificio Interfacultativo, Avda. de los Castros s/n, Santander 39005, Cantabria, Spain [email protected]

xii

List of Contributors

Babb, Jeff Department of Mathematics and Statistics, University of Winnipeg, Winnipeg, MB, R3B 2E9, Canada [email protected] Bicho, Nuno ICArEHB Faculdade das Ciências Humanas e Sociais Universidade do Algarve Campus Gambelas. 8005-139 Faro, Portugal. [email protected] Cardoso, H.F.V. Department of Archaeology, Simon Fraser University, Burnaby, BC, V5A 1S6, Canada. [email protected] Cardoso, João Luís Universidade Aberta Centro de Estudos Arqueológicos do Concelho de Oeiras (Câmara Municipal de Oeiras) Centro de Arqueologia da Universidade de Lisboa [email protected] Cascalheira, João ICArEHB Faculdade das Ciências Humanas e Sociais Universidade do Algarve Campus Gambelas. 8005-139 Faro, Portugal Cubas, Miriam Instituto Internacional de Investigaciones Prehistóricas de CantabriaSociedad de Ciencias Aranzadi. Edificio Interfacultativo. Avd de los Castros s/n. E- 39005. Spain. [email protected]

Muge 150th Volume 1

xiii

Cunha, Eugénia University of Coimbra, Department of Life Sciences Coimbra, Portugal [email protected] Cunha, Maria José Museu de Antropologia e Pré-História Mendes Correia, Museu de História Natural, Praça Gomes Teixeira 4099 - 002 PORTO, Portugal. [email protected] Detry, Cleia Centro de Arqueologia da Universidade de Lisboa (UNIARQ), Faculdade de Letras, Alameda da Universidade, 1600-214 Lisboa, Portugal. Dias, Rita ICArEHB Faculdade das Ciências Humanas e Sociais Universidade do Algarve Campus Gambelas. 8005-139 Faro, Portugal Diniz, Mariana Centro de Arqueologia da Universidade de Lisboa (UNIARQ), Faculdade de Letras, Alameda da Universidade, 1600-214 Lisboa, Portugal. [email protected] Duarte, Carlos Instituto Internacional de Investigaciones Prehistoricas de Cantabria (IIIPC), Universidad de Cantabria, Edificio Interfacultativo, Avda. de los Castros, s/n., 39005 Santander , Cantabria, SPAIN Dupont, Catherine CNRS UMR 6566 CReAAH Université de Rennes 1, Campus Beaulieu, bât. 24-25, 35042 Rennes, France [email protected]

xiv

List of Contributors

Évora, Marina Almeida ICArEHB FCHS - Universidade do Algarve Campus Gambelas. 8005-139 Faro, Portugal [email protected] Ferreira, Maria Teresa Forensic Sciences Centre, Department of Life Sciences, University of Coimbra, 3001-401 Coimbra, Portugal, [email protected] Figueiredo, Olívia NAP Faculdade das Ciências Humanas e Sociais Universidade do Algarve Campus Gambelas. 8005-139 Faro, Portugal [email protected] Gibaja, Juan Department of Archaeology and Anthropology. Milá y Fontanals Institution. Spanish National Research Council (CSIC). Egipciacas 15, 08001, Barcelona, Spain. [email protected] Gonçalves, Célia ICArEHB Faculdade das Ciências Humanas e Sociais Universidade do Algarve Campus Gambelas. 8005-139 Faro, Portugal [email protected] Iriarte, Eneko Laboratorio de Evolución Humana, Dpto. Ciencias Históricas y Geografía, Universidad de Burgos, Plaza de Misael Bañuelos s/n, Edificio I+D+I, 09001 Burgos, SPAIN Jackes, Mary Department of Anthropology, University of Waterloo, Waterloo, ON, N2L 3G1, Canada. [email protected]

Muge 150th Volume 1

Joaquinito, Anabela Portuguese Association for Archaeological Investigation (APIA) Portugal [email protected] Larsson, Lars Institute of Archaeology and Ancient History, Lund University, Box 117, SE-221 00 LUND, Sweden. [email protected] López-Dóriga, Inés Instituto Internacional de Investigaciones Prehistóricas (IIIPC), Universidad de Cantabria, Edificio Interfacultativo, Avda. de los Castros s/n, Santander 39005, Cantabria, Spain. [email protected] Lubell, David Department of Anthropology, University of Waterloo, Waterloo, ON, N2L 3G1, Canada Marchand, Grégor CNRS Université de Rennes 1– UMR 6566 CReAAH, France [email protected] Marreiros, João ICArEHB Faculdade das Ciências Humanas e Sociais Universidade do Algarve Campus Gambelas. 8005-139 Faro, Portugal. [email protected] Meiklejohn, Christopher Department of Anthropology, University of Winnipeg, Winnipeg, MB, R3B 2E9, Canada [email protected]

xv

xvi

List of Contributors

Monteiro, Patrícia ICArEHB Faculdade das Ciências Humanas e Sociais Universidade do Algarve Campus Gambelas. 8005-139 Faro, Portugal [email protected] Nukushina, Diana UNIARQ, Faculdade de Letras Universidade de Lisboa [email protected] Paixão, Eduardo ICArEHB Faculdade das Ciências Humanas e Sociais Universidade do Algarve Campus Gambelas. 8005-139 Faro, Portugal Pereira, Alexandra NAP Faculdade das Ciências Humanas e Sociais Universidade do Algarve Campus Gambelas. 8005-139 Faro, Portugal Pereira, Telmo ICArEHB Faculdade das Ciências Humanas e Sociais Universidade do Algarve Campus Gambelas. 8005-139 Faro, Portugal [email protected] Pimentel, Nuno CeGUL, Faculdade de Ciências Univesidade de Lisboa [email protected]

Muge 150th Volume 1

Price, T. Douglas Archaeology Programme Department of Culture and Society University of Aarhus [email protected] Rowley-Conwy, Peter Department of Archaeology Durham University P.A.Rowley-Conwy@ durham.ac.uk Teira, Luis Instituto Internacional de Investigaciones Prehistóricas (IIIPC), Universidad de Cantabria, Edificio Interfacultativo, Avda. de los Castros s/n, Santander 39005, Cantabria, Spain Umbelino, Cláudia University of Coimbra, Department of Life Sciences, CIAS, [email protected] Zapata, Lydia Universidad del País Vasco/Euskal Herriko Unibertsitatea (UPV/EHU)

xvii

CHAPTER ONE CARLOS RIBEIRO AND FRANCISCO ANTÓNIO PEREIRA DA COSTA: DAWN OF THE MESOLITHIC SHELL MIDDENS OF MUGE (SALVATERRA DE MAGOS) JOÃO LUÍS CARDOSO Universidade Aberta Centro de Estudos Arqueológicos do Concelho de Oeiras (Câmara Municipal de Oeiras) Centro de Arqueologia da Universidade de Lisboa [email protected]

Abstract This paper presents the original unpublished documentation on the identification by Carlos Ribeiro, of the first two Mesolithic shell middens in the region of Muge: Arneiro do Roquete (or Quinta da Sardinha), in the Magos creek valley, and Cabeço da Arruda, in the Muge creek valley, also known as Paul do Duque (Duke’s Marsh), on the 13 and 14 April 1863 and on the 24th of the same month, respectively. It is interesting to point out that Carlos Ribeiro did not realize the archaeological importance of the piles of shells and bones he came upon in both locations, admitting that they were the result of torrential transports. Such discoveries were framed in the geological studies developed at the time in Portugal by the two pioneers who were directly connected to them: Carlos Ribeiro and Francisco António Pereira da Costa, Director members of the Comissão Geológica de Portugal (the Portuguese Geological Commission), whose personal and institutional relations are also characterized in this paper, for being determinant in the evolution of the geological and archaeological investigations then made in the country.

2

Chapter One

The Comissão dos Trabalhos Geológicos was created by the Decree of 31 December 1852. However, the Commission was only regulated by the Decree of 8 August 1857 and by two other royal Decrees made on the same day when people were appointed to it as members: Carlos Ribeiro (1813-1882), who was at the time Captain of the Artillery, and General Filipe Folque, “Lente”, that is, Professor of Mineralogy and Geology of the Escola Politécnica. Both members of the Commission were dependents of the Director of the Trabalhos Geodésicos do Reino, functioning within the scope of the Ministério das Obras Públicas, Comércio e Indústria (Boletim 1857). Pedro de Aguiar (1941), who was married to a granddaughter of Carlos Ribeiro, wrote that the Minister had wanted Carlos Ribeiro to occupy the position of Director, due to the fact that he had held the position of the Chief of the Mines Section since the Ministry was created in 1852, by invitation of the Minister at the time, António Maria Fontes Pereira de Melo. However, it had been recognized that there was a clear advantage for the said Commission to include someone particularly devoted to paleontological studies which was well considered as essential for such purposes; Carlos Ribeiro was, apparently willingly, co-opted for the direction of the newly-formed Service along with Francisco António Pereira da Costa (1809-1889), and their assistant was the young army officer Joaquim Filipe Nery da Encarnação Delgado (1835-1908). Pereira da Costa had previously worked, fruitfully, with Carlos Ribeiro: both had written the Lei de Minas, published on 31 December 1852, which remained in print for a long period of time. Therefore, mutual confidence and esteem existed between them. It was an alliance between two complementary personalities: Carlos Ribeiro (Fig. 1.1), with his great experience in fieldwork was, alongside Nery Delgado, in charge of assuring the coordination of a remarkable team of collectors, whose relevance in the fast progress of the geological survey of the country has only very recently been the object of study (Carneira 2005); Pereira da Costa, an academic and Professor of Mineralogy and Geology at the Escola Politécnica, was in charge of the cabinet work, involved in the studies of the paleontological collections, and with the objective of stratigraphic support for the geological surveys, as well as the organization of the collections. This complementarity, however, originated due to the serious dissent that ended in the extinction of the Second Geological Commission in February 1868. Nery Delgado, who knew very well the causes of the divergence, elegantly silenced them with the remarkable historical compliment paid to Carlos Ribeiro in 1905, without, however, failing to

Carlos Ribeiro and Francisco António Pereira da Costa

3

register the scientific merits of Pereira da Costa, inasmuch as the first headquarters of the Geological Commission were located in the private residence of Pereira da Costa in Lisbon, São Roque St. (Delgado 1905) from November 1857 until April 1859 (Almeida & Carvalhosa 1974). In fact, from that date the Commission occupied the upper floor of the extinct Convento de Jesus, which was then already the office of the Academia Real das Ciências de Lisboa. As the last testimony of this state of affairs, the Geological Museum is still located in this same place today, though it is now integrated into the Laboratório Nacional de Energia e Geologia (LNEG).

Fig. 1.1. Carlos Ribeiro (1813-1882).

The LNEG is the successor of its two prior institutions, the Serviços Geológicos de Portugal, from 1918, and, more recently, the Instituto Geológico e Mineiro. The LNEG keeps the collections of the Muge shell

4

Chapter One

middens, recovered after the extinction of the Second Geological Commission, which occurred, as already mentioned, in 1858. Carlos Ribeiro was a pioneer: he established the survey of the mineral and hydrogeological resources of the country, knowing that the development of the country depended, in part, on its geological knowledge, and his early studies of geological nature in its various aspects: without forgetting the so-called “Pure Geology”, with the integration of geological cartography, for which the knowledge of Stratigraphy and Palaeontology was essential. Nery Delgado accompanied him many times on his long stays in the field, resulting in the publication of the first geological maps as early as 1866 and 1867 at the 1/100 000 scale, which constitute the first serious attempt at the geological cartography of the country and reveal the scientific and graphic quality of the work associated with its execution (Zbysewski 1949). Unfortunately, they were never distributed. This was due to purely scientific motives (Cardoso, 2013) since there could be no other reasons after the reconstitution of the Geological Commission (although with another name), in December 1869. Indeed, in 1866 Carlos Ribeiro had already charted a good part of the detrital deposits of the Tagus Basin, thought to be Quaternary. These deposits attained the thickness of 400 m and had been heavily subjected to severe tectonic processes, some in the vertical position. However, when confronted with the odd mapping of the deposits transmitted to him by de Verneuil, Carlos Ribeiro changed his opinion and correctly remitted those formations to the Miocene and Pliocene (Ribeiro 1871), thus rendering useless those two geological maps. Carlos Ribeiro was, indeed, a man of the field and a man of action, even though he did not forget the scientific publication of his field results. Nery Delgado reported (Delgado 1905) that Ribeiro was born in 1813, the son of a silver caster of the Casa da Moeda (the institution in charge of printing and coining currency in Portugal), and that he had to start working very early in his life, as a clerk apprentice: at the age of ten he worked in a grocery shop in São João da Mata St., in Lisbon. In 1905, his main biographer, Nery Delgado, who was by then Division General, tells us that such an establishment was attended by boys coming from good families who used to make that shop their meeting point. One of the most assiduous attendees of those meetings was Filipe Folque, his future hierarchic superior but at that time just a student-cadet of the Academia Real de Fortificação, Artilharia e Desenho, who assumed for himself the role of protector of the smart clerk apprentice. Carlos Ribeiro acquired a knowledge of French grammar and was soon able to take advantage of the contents of the books that Folque lent him. This allowed him, at the age of

Carlos Ribeiro and Francisco António Pereira da Costa

5

19 on 4 August 1833, to enlist in the Artillery, thus embracing the Constitutional cause, against the wishes of his father, who was an absolutist. Because of that decision, he could no longer expect the slightest aid from his father: he was all by himself and dependent only on himself, based on his qualities and his behaviour. As a matter of fact, the busy personal and professional life of Carlos Ribeiro was the object of a short yet touching novel written by Camilo Castelo Branco, his former classmate at the Academia Politécnica do Porto. This novel was published in 1884, two years after the passing of Camilo’s colleague (Castelo Branco 1884) and it very well portrays the generous and romantic soul of the future geologist.

Fig. 1.2. Francisco António Pereira da Costa (1809-1889).

6

Chapter One

In contrast, Francisco António Pereira da Costa (Fig. 1.2), born in Lisbon on 11 October 1809 was, as mentioned above, the prototype of a cabinet researcher, with sedentary and regular habits. Like Carlos Ribeiro, he fought for Liberalism in the Academic Batallion, and, right after the triumph of the Constitutional cause, he graduated in Medicine at the Universidade de Coimbra (University of Coimbra). At the time of the foundation of the Escola Politécnica in 1837 Pereira da Costa was much more attracted by the studies of the Natural Sciences than by Medicine, and he applied for the position of Professor of Mineralogy and Geology, which he only took up in 1840. Gifted with a meticulous spirit, Pereira da Costa took an interest in the cataloguing of the collections of the Museu de História Natural, located at the time in the Academia das Ciências de Lisboa, where he temporarily conducted the classes (Gomes, 1903/1904) during the reconstruction of the building of the Escola Politécnica after the great fire of 1843. In the collections of that Institution, the most important core consisted of the antique Colecção Real da Ajuda, which had, in the meanwhile, been transferred to the Academy. This collection, although very depleted, had lately been incorporated into the Escola Politécnica, certainly as the result of the intervention of Pereira da Costa, where a major part of it was destroyed in 1978 by the great fire that took place there. So, Pereira da Costa’s lack of enthusiasm for the daily growth of the collections of the Comissão Geológica is understandable, in unfair competition with those of the Polytechnic (Antunes 1986): this was so, in spite of the fact that he was the author of an important proposition presented after the reform of the Academia das Ciências de Lisboa in December 1851, during the session of 10 March 1852 (Costa 1949), transcribed below (in Costa, 1949: 8): The capacities of the Reformative Decree of the Academia Real das Ciências de Lisboa point to the satisfaction of the purposes of this institution, the following two are mentioned: First–the special study of Portuguese soil, both on the mainland and in the overseas provinces, considered in itself and with regard to the industrial arts. Second–the enlargement of the collections of the natural products in the Academy. The duty of taking care of the execution of these two legal precepts belongs to the section for Historic-Natural Sciences and the duty of indicating what seems to me to be the more convenient thing to do in that respect belongs to me, who has until now been responsible for classification in the Museum,…. (Our translation)

Carlos Ribeiro and Francisco António Pereira da Costa

7

This stand had a precise goal: Pereira da Costa insistently claimed that the First Geological Commission of the Kingdom was responsible for the collections reunited by Charles Bonnet, saying that: he thought it indispensable to gather in a single location all the elements of study of the Historic-Natural Sciences that were scattered all over Lisbon, and to minister the teachings in the establishment where those elements were reunited, giving them a ‘definitely practical’ character. For that, he suggested that not only the collection existent in the Casa da Moeda– started by Eschwege (…)–should be transferred to the Museum of the Academy, but also the collections of the Escola Politécnica (Polytechnic School) and those, as he said, of the ‘Commission of the Geological Map of the Kingdom’ (in Costa 1949: 8). (Our translation)

Due to the fact that he had been unable to achieve his goals at the time, and thanks to the re-establishment of the teaching in the Escola Politécnica after the reconstruction of the building following the fire of 1843, it is perfectly understandable that Pereira da Costa had reservations regarding the flourishing assets of the Museum of the Second Geological Commission, which had furthermore been enriched with the materials of study and comparison gathered during the travels made by Carlos Ribeiro from 4 July to 14 December 1858 (Delgado 1905). When the importance of the Academia das Ciências de Lisboa to the geological survey of the country had been diminished as a result of the creation of the Second Geological Commission by the Ministério das Obras Públicas, only a few months after Pereira da Costa’s proposal transcripted above, the only thing he could do was to achieve an understanding with Carlos Ribeiro, with whom he had always maintained good relations. Such working relations, developed within the principle of complementarity mentioned above, would be maintained until 1866/1867. But, as expected, Pereira da Costa’s growing grudge increased further. Several reasons for this can be pointed out: the enlargement of the assets of the Geological Commission and the use of funds made available to the Institution where teaching was not practised; all this was done to the detriment of the Museu da Escola Politécnica where Pereira da Costa still belonged to the teaching board; naturally, the School had a limited budget–and only by serving the teaching would the Museu would make any sense. These were, amongst other reasons, at the origin of unrecoverable disagreements. Once again, Nery Delgado discreetly mentioned those reasons: With the authorization of a special fund for the geological survey of the country (…), the survey work began on 18 May 1866, with Carlos Ribeiro

8

Chapter One leaving for Alentejo accompanied by the Commission assistant [Nery Delgado himself]. They both returned to Lisbon in November, and in April 1867 they left for the north of the kingdom, with the work of this first survey finishing in September of that year. One can date from this period the misunderstandings between the two directors of the Geological Commission which had such dire results. For some reason, which will not be easy to understand, Pereira da Costa fiercely opposed him over the execution of the survey, creating for Carlos Ribeiro the greatest difficulties and openly denying him his collaboration in the work and in the supervision of its expenses, unlike his usual actions on the regular funding of the Commission (…). Having been witness to this fight since it began, I have never been able to understand its true motive. In the private correspondence of Carlos Ribeiro, an immediate cause for this quarrel with Pereira da Costa appears to be the disagreement over an issue of administration; but I believe that this dissent could have been easily fixed given the intense friendship that bonded them, as long as another hidden cause did not contribute to that result. Even in their correspondence it appears that, beyond the presented reasons, there was a more intimate cause for this breaking up of relations. Maybe a misunderstanding or mistrust that later turned into a serious offence originated that disgraceful fight that injured one of the contenders, and for which the other has won nothing with its triumph. (Delgado 1905: 23 and 24). (Our translation)

Pedro de Aguiar was the only one who, until today, has pointed out concrete leads to the causes for the quarrel between the two colleagues. He says: My dear friend and uncle by affinity, (…) José Vitorino Damásio Ribeiro [son of Carlos Ribeiro], told me, yet very secretly, that it was a misunderstanding or a refined susceptibility by Dr. Pereira da Costa. Even after a certain age, Dr. Pereira da Costa had some hidden love affairs, which infatuated him. Carlos Ribeiro, always sensitive and polite, may have said something to him that Dr. Pereira da Costa did not appreciate. Maybe that is the origin of the possible resentment and then divergence, because the love offences are unforgivable. (Aguiar, 1941: 25) (Our translation)

This was also the version taken by M. Telles Antunes as the immediate cause for the disastrous known outcome, even though he does not quote Pedro de Aguiar (Antunes 1986: 795, 1989: 144) and also by other researchers before him, who were apparently unaware of that work (Costa 1949:13). However, beyond these mundane affairs invoked without concluding evidence, other deeper and more imposing realities which were certainly much more important, will better explain what happened.

Carlos Ribeiro and Francisco António Pereira da Costa

9

Without speculating on the private life of the apparently austere Professor of the Polytechnic, he had permanent motives that would justify the attempt to neutralize the Second Geological Commission, as Nery Delgado mentions in his statement. Carlos Ribeiro’s priority was the geological survey of the country from which resulted the already mentioned first maps at 1/100 000 scale, but for Pereira da Costa, fieldwork was only interesting if it provided him with the raw material for scientific publications. These publications involved very high typographic expenses. In fact, Pereira da Costa wrote and published a remarkable work (Costa 1866/1867), having left an important set of lithographs which were printed and published after his death (Dollfus et al. 1903/1904). In other words, for Pereira da Costa, the fieldwork was not an end in itself, as opposed to Carlos Ribeiro’s perspective. Thus, it is in this light that the words of Nery Delgado, transcripted above, gain such major importance: one can understand that for Pereira da Costa it would have been preferable for the financial resources that Ribeiro was planning to channel into the fieldwork to be mainly spent on publication where his merit was undoubtedly distinguished. This was, then, the origin of the poisonous atmosphere that was brooding in the Geological Commission in 1866 and 1867 (Carneiro 2005: 155). The extinction of the Geological Commission took place soon afterwards, by the direct intervention of Pereira da Costa, making use of his relations with the Minister of Public Works, and his former colleague in the Escola Politécnica, who gave him the lead in the geological studies of the country. The Decree of 23 December 1868 ensured the transfer to the Escola Politécnica of all the study materials gathered there, including books, collections and furniture (Delgado 1905: 24, 25). In this way, Pereira da Costa achieved, even if only for a moment, the goal for which he had been fighting for so long: the gathering in a single public institution in Lisbon of all the collections and documentation, including the library of the extinct Comissão Geológica, which would serve teaching and geological investigation. However, this was a wasted effort because the Geological Commission was again restored one year later when the political conditions allowed, but without Pereira da Costa. Until the end of his long life of about 80 years (he passed away on 3 May 1889), he continued to dedicate his efforts towards the maintenance and enlargement by acquisition of the geological collections (petrography), paleontology and mineralogy of the Museu da Politécnica, without publishing anything remarkable, probably thanks to the lack of means, which had not previously been a problem. It must be noted that the majority of the collections of the old Geological Commission were never

10

Chapter One

returned when this Institution was restored, as we had the opportunity to personally verify in 1977, at the Museu Mineralógico e Geológico da Faculdade de Ciências da Universidade de Lisboa. This also happened with the editions of the Commission before its extinction in 1868, especially those by Pereira da Costa: some of them were still there for sale at derisory prices in the 1980s, survivors of the terrible fire that almost completely destroyed the interior of the building in 1978. *** *** *** These two men with such disparate personalities and also with very distinct visions about the development of geological studies in Portugal and the methodologies, purposes and priorities, are both inextricably connected to the investigations into the Mesolithic shell middens of Muge. The conditions in which the first discoveries took place, in the valleys of the creeks of Magos and Muge, were reported by Carlos Ribeiro himself: It was at the expense of hard work and many adversities that, in 1863 and in the following years, we discovered, and collected under our immediate and exclusive supervision, the remains of human skeletons, vertebrate animals and molluscs, in Cabeço da Arruda, in Salvaterra and in other places in the Tagus valley, as well as the biggest part of the numerous objects of pre-historic human art that, in 1868, could have been seen in the Museum of the Geological Commission. The origin of these investigations and discoveries was the necessity for surveying, by the examination of the geological facts, which were the most important movements that had occurred in our land after the opening of the first order valleys that nowadays cut it, and to indicate which of those movements were contemporary with the human species; these were questions of the highest importance that are interesting not only to the geology of the Iberian Peninsula, but also to the hydrography, and to the difficult hydraulic problems that engineering has to solve in order to allow the restoration and conservation of our ports and harbours. (Ribeiro 1871: 1, footnote) (Our translation)

Notice that the author uses the past tense when he refers to the collections exhibited in 1868 in the Museum of the Portuguese Geological Commission, given the fact that in 1871 the said collections where no longer there, for they had been transferred to the Escola Politécnica, as it was called at the time, after the extinction of the mentioned Commission.

Carlos Ribeiro and Francisco António Pereira da Costa

11

The field records of Carlos Ribeiro, published here for the first time, prove that it was on 13 and 14 April 1863 that the first discoveries were made (Fig. 1.3):

Fig. 1.3. Description of the identification of the shellmidden of Arneiro do Roquete (April, 13, 1863). Contemporary handwritten copy of the field notebooks of Carlos Ribeiro, written by himself, in sheets of lined paper. April 13 and 14 Tour to Paul dos Magos (Magos Marsh) and the setting of boundaries. The left bank of the Tagus Valley presents itself as a very smooth plain from the Cazal or Quinta do Sargento till Salvaterra. In the valley of Paul de Magos, the right bank is flattened and the left one is abrupt. The relative heights of the bank from the top of the plain to the field are about 10 to 15 metres.

Chapter One

12

[From there] to the valley of the Q[uin]ta da Sardinha is a distance of 1 kilometre to the west. In the place of Córte grande and in the properties of the Arneiro do Roquete, one finds a very large quantity of recent shell fragments, mixed with loose rough sand. The fragments are in such quantity that they whiten the soil in spots, according to the larger or smaller accumulations. Their aspect is exactly like the sand of the beaches that have many rock fragments. In this sand or on the surface one finds many pieces of this sand conglutinated with the fragments of the shells, looking like many conglutinations found on the shoreline like, for example, at the place of Sete Bicas between Foz d’Albufeira and Foz da Fonte. The extension that it occupies is small, under 400 metres. And it is covered by the thin Pliocene sands, but those that the wind has set in motion. Its height is about 14 metres. Starting from the field or from a small ditch, it forms a ramp that precedes a plain where the animal remains finish. The remains found there are:–A phalanx that appears to be from a digitigrade [.] A vertebra fragment of a land mammal [.] A skull fragment (thick bone)–Idem. Bone fragments of the limbs–Idem. Rib fragments– Idem. Small thin bones and fragments of others– Crab claws– Buccinum–Tellina–Nucula–Cardium edule–Oyster (???) in the Tagus estuary–Chitton (sic)–Idem–the one that is edible–Pecten–Solen– Some of these shells have colours, the pearly nacre, and the others have the freshness of new shells. They are evidence of the old Tagus estuary. (…)

Some days after these observations, Carlos Ribeiro observed similar evidence in another tributary of the same margin of the Tagus located upstream (Fig. 1.4): April 24–Tour by Raposa, Vale de Postigo and Machadinhos. Above the layers of reddish tawny Pliocene sandstone that forms the cliff that people call Fonte da Burra next to Pontes do Ralão right outside Mugem there are conglutinations with abundant shell remains (Cardium, Telina, etc.) at a level of 12 metres above the fields of Mugem inside Paul. I followed the Motta da Valla until I got to the road that leads to Raposa and next to it, a quarter of an hour after leaving the Motta, there is a small hill called Cabeço d’Arruda. It is about 6 metres above the Paul. It presents, to the side of the Paul, an abrupt cliff formed by a deposit1 of marine shells and the bones of land animals with a visible thickness of 3 metres. There I found Cardium edule, a single piece of a Solen valve,and a small Cypraea, with the deposit being constituted mainly of Fragiliae and broken shells and all bivalves. I also found a jaw piece of a land mammal 1

N.T.: In the original, lumachela.

Carlos Ribeiro and Francisco António Pereira da Costa

13

with teeth; a portion of a rodent jaw, a phalanx and different bones of rodents and other animals. In the same deposit at Cabeço d’Arruda, pebbles of quartzite can be found up to the size of turkey eggs, scattered in an irregular pattern inside the mass and with its bigger axis slightly leaning in relation to the horizontal plain in which the deposit stands. Everything proves that it is a transport deposit originated by a violent torrent that appears to have followed side by side with the axis of the Tagus and in a time in which the Pliocene deposits were already formed. In this deposit many fragments of carbonized wood are found. (…). (Our translation)

Fig. 1.4. Description of the identification of the shellmidden of Cabeço da Arruda (April, 13, 1863). Contemporary handwritten copy of the field notebooks of Carlos Ribeiro, written by himself, in sheets of lined paper.

An extract of Carlos Ribeiro’s original field notebook had been published, belonging to the Historic Archive of the LNEG, in which the

14

Chapter One

discovery of the shell midden of Arneiro do Roquete or Quinta da Sardinha was described, reportedly on 13 April 1863 (Cardoso & Rolão 1999/2000: Document 3). The present transcript was based on the handwritten copy of the field notebooks, made at the same time by Carlos Ribeiro on sheets of lined paper, but this was more readable than the one used earlier and with slightly different contents, which occurred because the appointments recorded on the field were copied. They were recovered in 1975/1976 from the piles of rubbish accumulated at the door of the Serviços Geológicos de Portugal (Portuguese Geological Survey), which came from the attic of the building that had been transformed into working offices. These are the first references to the famous Muge shell middens. In this way, it can be concluded that the shell midden from Arneiro do Roquete was the first to be discovered, in the valley of the creek of Magos, confirming the information published by Carlos Ribeiro years later (1884: 280). This shell midden was also designated as Quinta da Sardinha, the name that Afonso do Paço (1938) would in the future relate to two other shell middens, the Cova da Onça and the Monte dos Ossos, both of them located on the right bank of the creek of Magos. Therefore, it can be concluded that the three place names correspond to a single archaeological location (Cardoso & Rolão 1999/2000: 85). Carlos Ribeiro, in his only publication dedicated to the shell middens of the Tagus valley and presented to the International Congress of 1880, declared that it was not possible to undertake archaeological excavations at the location because of the opposition of the owner. On the other hand, in the Muge creek, designated Paul do Duque, the first shell midden to be discovered by Carlos Ribeiro was the shell midden of Moita do Sebastião, then called Fonte da Burra, immediately before the identification of the shell midden of Cabeço da Arruda; it was here that the most important excavations took place. Such excavations were visited by the participants of the organized tour in 1880, during the International Congress that was then held in Lisbon. After this site, other shell middens followed even before 1880, along with the identification of Fonte do Padre Pedro and Cabeço da Amoreira (Ribeiro 1884: 280), and all of them were still the object of excavations or mere surveys in the nineteenth century. Therefore, it can be concluded that after the completion of the survey of Vale de Magos, the survey of Paul do Duque ensued, which was located further north. During this later survey, the important and well-preserved shell midden of Cabeço da Arruda was referenced. In 1864 and 1880, this would be the object of major extended excavations and publications, both conducted by Carlos Ribeiro (Fig. 1.5).

Carlos Ribeiro and Francisco António Pereira da Costa

15

It is also important to note that, by the time of its identification, Carlos Ribeiro considered its formation to be due to natural causes and not to anthropic activity. This was only properly re-evaluated at the time of the excavations conducted in the year following its discovery, due to the presence of numerous human skeletons, which reached the number of 45 individuals (Costa 1865:13), a number that was much increased by the excavations conducted in 1880, with some of them being photographed in situ (Ribeiro 1884: P1. I, II), in what was the first record of this kind published in Portugal.

Fig. 1.5. Cabeço da Arruda. Partial view of the excavations of 1880 (in Ribeiro, 1884:Pl. 1).

The conditions observed at the time of the excavations conducted in 1864 at Cabeço da Arruda, which corresponded to a cut through all layers of the shell midden on its southwest side facing the Paul do Duque, were thoroughly described and interpreted by Carlos Ribeiro, as we can see in the monograph signed only by Pereira da Costa: the first scientific publication in Portugal of a pre-historic site (Costa 1865). In that publication (Fig. 1.6), Pereira da Costa used the observation elements obtained by Carlos Ribeiro without even mentioning his name. So, it is not surprising that Carlos Ribeiro, when the relationship between them was already destroyed, felt it was his right to underline in a concise and objective way the real authorship of the identification and the

16

Chapter One

archaeological works conducted on those sites, which are among the most remarkable of Mesolithic European sites.

Fig. 1.6. Front cover of the first monograph published about the Muge shellmiddens (Costa, 1865).

Carlos Ribeiro and Francisco António Pereira da Costa

17

In a manuscript by Carlos Ribeiro, recently published by Cardoso (2013), written shortly after the publication of 1865, it is possible to read: “Among the vertebrates found in Cabeço d'Arruda figured a lot of human skeletons. Mr. F.A.P. da Costa, also Member of the Geological Commission, we gave for the sudy of these human remains, giving all the information necessary that was collectedduring the observation of the facts in the location where they were found.” (Cardoso, 2013:86).

References Aguiar, P., 1941. Os Generais José Vitorino Damásio e Carlos Ribeiro. Porto: Tip. Empresa Guedes, Lda. Almeida, F. Moitinho de & Carvalhosa, A. Barros e (1974. Breve história dos Serviços Geológicos de Portugal. Comunicações dos Serviços Geológicos de Portugal. 58, 239-265. Antunes, M. T., 1986. Sobre a História da Paleontologia em Portugal. In História e desenvolvimento da Ciência em Portugal. Lisboa: Academia das Ciências de Lisboa, 2, p. 773-814. —. 1989. Sobre a História do Ensino da Geologia emPortugal. Comunicações dos Serviços Geológicos de Portugal. 75,127-160. Boletim, 1857. Boletim do Ministério das Obras Públicas, Commercio e Industria. 8. Lisboa: Imprensa Nacional, p. 305-307. Castelo Branco, C., 1884. O General Carlos Ribeiro (recordações da mocidade). Porto: Livraria Civilisação. Cardoso, J. L., 2013. Carlos Ribeiro, a "Breve noticia acerca do terreno quaternario de Portugal", e a questão do homem terciário em Portugal. Estudos Arqueológicos de oeiras, 20, 27-88. Cardoso, J. L. & Rolão, J. M., 1999/2000. Prospecções e escavações nos concheiros mesolíticos de Muge e de Magos (Salvaterra de Magos): contribuição para a história dos trabalhos arqueológicos efectuados. Estudos Arqueológicos de Oeiras. Oeiras. 8, 83-240. Carneiro, A., 2005. Outside government science, “Not a single tiny boné to cheer us up!” The Geological Survey of Portugal (1857-1908, the involvement of common men, and the reaction of civil society to geological research. Annals of Science, 62 (2), 141-204. Costa, F. A. Pereira da, 1865. Da existencia do Homem em epochas remotas no valle do Tejo. Primeiro opúsculo. Noticia sobre os esqueletos humanos descobertos no Cabeço da Arruda. Lisboa: Commissão geológica de Portugal.

18

Chapter One

Costa, F. A. Pereira da, 1866/1867. Molluscos fosseis. Gasteropodes dos depósitos terciários de Portugal. Lisboa: Commissão Geologica de Portugal. Costa, J. Carrington da, 1949. Aspectos da evolução dos serviços encarregados do estudo geológico de Portugal. Memórias da Academia das Ciências de Lisboa. Classe de Ciências. Lisboa. 5, 14 p. (separata). Delgado, J. F. Nery, 1905. Elogio histórico doGeneral Carlos Ribeiro. Lisboa: Imprensa Nacional. Dollfus, G. F. et al., 1903/1904). Molusques tertiaires du Portugal. Planches de Gastéropodes et Pélécipodes laissées par F. A. Pereira da Costa accompagnées d´une explication sommaire et d´une esquisse géologique. Lisboa: Comissão do Serviço Geológico de Portugal, p. VIII-IX. Gomes, J. P., 1903/1904. Notice biographique sur Francisco Antonio Pereira da Costa. In DOLLFUS, G. F. et al., (1903/1904), Molusques tertiaires du Portugal. Planches de Gastéropodes et Pélécipodes laissées par F. A. Pereira da Costa accompagnées d´une explication sommaire et d´une esquisse géologique. Lisboa: Comissão do Serviço Geológico de Portugal, p. VIII-IX. Paço, A. do, 1938. Novos concheiros do vale do Tejo. Brotéria, 27 (1), 6675. Ribeiro, C., 1884. Les kioekkenmoeddings de la valée du Tage. IX Sessão do Congresso Internacional de Antropologia e de Arqueologia PréHistóricas (Lisboa,1880). Actas. Lisboa: Tipografia da Academia Real das Ciências de Lisboa, p. 279-289. Zbyszewski, G., 1949. Deux cartes géologiques inédites, oeuvres de Carlos Ribeiro et J. F. Nery Delgado. Lisboa: Serviços Geológicos de Portugal.

CHAPTER TWO THE ARCHAEOLOGICAL EXCAVATIONS AT MUGE SHELL MIDDENS IN THE 1930’S: A NEW CONTRIBUTION TO THE HISTORY OF ITS INVESTIGATION ANA ABRUNHOSA Ph.D. Student University of Algarve, [email protected]

Abstract The old researchers must be remembered in the celebration of 150 years since the discovery of the Mesolithic shell middens of Muge (18632013). Their hard work contributed to the development of archaeology as a scientific field and to the study of Prehistory in Portugal and Europe. The Mesolithic shell middens of Muge turned out to be one of the major Portuguese and European Prehistoric archaeological sites and the many teams that based their research on these sites throughout the last 150 years reflect different stages in the development of Portuguese archaeology. In the 1930s Mendes Corrêa (1888-1959) was responsible for leading a team of researchers from the University of Porto in the study of the origins of the Muge Men and the chronology of the shell middens. A new set of unpublished documentation discovered in the 1980s and only now studied and brought to public notice sheds light on the objectives of the research and the fieldwork methodologies applied. This area is usually poorly described, even though it is considered to be of the utmost importance for the knowledge of the development of archaeological practice in Portugal and Europe.

20

Chapter Two

Introduction The Mesolithic shell middens of Muge are scattered along the banks of the Ribeira de Muge and Ribeira de Magos, both tributaries of the Tagus River. They were discovered in 1863 by geologist Carlos Ribeiro (18131882) while on a survey in order to register the quaternary terraces of the region and to complete the geological map of Portugal (Ribeiro 1866). After a hiatus in the excavations between 1880 and 1930, A. A. Mendes Corrêa (1888-1959), Professor of the Faculty of Sciences of the University of Porto, directed a new phase of archaeological excavations in the Muge shell middens located in the margins of the Ribeira de Muge (Corrêa 1933; Monteiro 1959, VV. AA. 2012). He was aided both in the field and with further research back in Porto by a group of students and assistants of the Anthropological Institute that he had founded at the University of Porto (Fig. 2.1). The three who were the most important and visible throughout his work were Rui de Serpa Pinto, Santos Júnior and Alfredo Athayde (Pinto 1931; 1932; 1933).

Fig. 2.1. Mendes Corrêa (middle) with Santos Júnior (left) and Rui de Serpa Pinto (right) at Cabeço da Amoreira shellmidden in 1930.

Thanks to the many letters and postcards exchanged by the named researchers and to the field notes that they all while in Muge, there is enough data to reconstruct the fieldwork methodology with unusual detail.

The Archaeological Excavations at Muge Shell Middens in the 1930’s

21

Some of the surviving documents which were mainly field notes on pocket agendas belonging to Mendes Corrêa were published earlier and their content exposed and discussed by João Luís Cardoso (Cardoso 1999; Cardoso and Rolão 2002/2003) and by A. Huet B. Gonçalves (Gonçalves 1983; 1983/1984; 1986’ Pinto 1986). The latter was curator at the Dr. Mendes Corrêa Anthropology Museum and was responsible for finding, saving and starting to organize the new set of unpublished documents studied for a Master’s thesis at the Faculty of Arts of the University of Porto (Abrunhosa 2012a).

Objectives and Methodology The aim of Mendes Corrêa was the study of the singular osteoanthropological characteristics of the human skeletons found in the shell middens and the origins of Homo mugem, as it was initially named, and later changed by him to Homo afer taganus (Abrunhosa 2012b; Corrêa 1936). For that purpose he needed more specimens and so, a new excavation cycle began in 1930 and ended in 1937 (Corrêa 1917). The shell middens studied in that decade were the ones found by Carlos Ribeiro–Cabeço da Arruda, Moita do Sebastião, Cabeço da Amoreira and Fonte do Padre Pedro. The best preserved were targeted by the new excavations – Cabeço da Amoreira and Cabeço da Arruda in the years 1930-1931 (Cabeço da Amoreira), 1933 (Cabeço da Amoreira and Cabeço da Arruda) and 1937 (Cabeço da Arruda). The main objectives of this research were i) to continue the study started by A. Huet B. Gonçalves; ii) to organize the new documentation about the Muge shell middens’ archaeological excavations during the 1930s; and iii) to analyse these archaeological excavations on the shell middens directed by Professor Mendes Corrêa through more than two hundred unpublished documents. The new abundant documentation consists of a total of 227 documents from the years 1921, 1930 to 1934, and 1936 to 1939 (Fig. 2.2) revealing the motivations, strategies, programming, development and fieldwork techniques of that time (Abrunhosa 2012a). Each document was organized by the year of its production and its typology as defined by the author (Fig. 2.2) – correspondence (49 documents), drawings (12), topographic maps (4), expenses (26), reports (6), notes (18), lists (7), texts to be read on oral communications at congresses (3) and photographs (102).

22

Chapter Two

Fig. 2.2. Num mber of documeents by typology y per year.

Each doccument was digitally d scann ned in high reesolution imag ge format for detailed study and arcchiving. Then n they were inndividually an nalysed in graphical annd iconographhic perspectiv ves (Abrunhoosa 2013). Th he overall set was analysed with the objectivee of finding which major themes embodied in this docum mentation cou uld bring new w informatio on to the understandinng of the devvelopment of Portuguese arrchaeology in n the first half of the twentieth ceentury. In paarticular, we wanted to know k the methodologies employed in the study and research of the Mesoliithic shell middens of M Muge by the team t directed by Mendes C Corrêa.

N Contrributions New Given thhe objectives proposed abo ove, it was poossible to sorrt out the information obtained from m the study off the new set oof documents. The new documentatiion allowed the reconstru uction of parrt of the exccavations’ methodologyy and the ressearchers’ callendar duringg the 1930s. The T main contributionns are related to four subjects: s i) nnew insightss on the preparation and methodollogy of fieldw work; ii) labouur divisions; iii) travels, expenses annd activities reeports; and iv) the support oof Casa de Cad daval.

The Arrchaeological Excavations E at Muge M Shell Midddens in the 193 30’s

23

Preparation n and Metho odology of F Fieldwork The studdy of researchh methodology y in the field, i.e., the archaaeological excavations,, was basedd mainly on the analyysis of pho otographs, correspondeence and nottes taken in the field. Thhrough each of these elements it w was possible to t understand the methodollogies of fieldwork and the instrumeents that were employed forr the excavatiion process. The T letters and postcardds helped the analysis a of thee adaptation oof strategies th hroughout the work annd the discusssion of fieldw work prior to tthe scientific questions that structurred the project. In this paarticular projeect the main objective prior to the excavation was w known. Mendes M Corrêaa needed to ob btain new specimens too continue hiss study of the origins of thee Muge Man–tthe Homo afer taganuss. All questionns or concernss were commuunicated to the director of the excavvation even whhen he was ab bsent.

Fig. 2.3. Wom men working att Cabeço da Am moreira shellmiddden in 1933.

The insttruments usedd in the archaaeological exxcavation as visible v in most photoggraphs are hoees, shovels, wo ooden wheelbbarrows, wood den frame metal meshh sieves, trow wels, brushess and consollidants (parafffin in a solvent) forr fine work regarding r thee exhumationn of human skeletons. s Some instruuments are of agricultural origin. The soiil excavated in n the area

24

Chapter Two

that was considered to have archaeological significance was transported in wheelbarrows to the sieving area and potential archaeological objects were hand-picked. There were two kinds of sieves. Small circular sieves could be operated by only one person. The larger rectangular sieves on a swing structure were operated by two women or men. The sieve aperture is unknown (Fig. 2.3). Several letters and postcards mention excavation strategies, discussions on which areas to excavate, decisions to be taken regarding which areas to prioritize, and the definition of stratigraphy and the layers’ characteristics, among other issues that arised during the fieldwork. The shell middens were “attacked”, closely following the nomenclature of that time, from the defined limits towards the centre. When a layer of shells was detected, the excavation would start from the top layer until reaching what was thought to be the bottom of the midden. With this we have a horizontal progression and a vertical progression. The terrain being studied was divided in squares 2.5 metres large, marked on site with wooden stakes, and limited by code letters. The squares were each named with the letter that limited them in one direction. Perpendicularly the rows of squares were named by a number (Fig. 2.4). For example square A-B row 1, B-C row 3, and so on. There were three layers–a superior layer, the middle layer and a lower layer–and they were defined by their depth in relation to the highest point of the shell midden and the typology of artefacts found in each one of them. With a horizontal and vertical progression at the same time and a delimitation of a squared mesh the archaeologists were able to record a relative spatial position for each of the most important finds for Mendes Corrêa–the human skeletons. Every time a skeleton was found, it was isolated, carefully excavated and recorded by means of photography throughout the process of excavation. Sometimes drawings or sketches were also made but the number of photographs of the skeletons reveals the importance that was given to this new method of recording (Fig. 2.5).

Labour Divisions Mendes Corrêa was a professor at the University of Porto and active on the political scene (he was the Mayor of Porto municipality from 1937 to 1940, years in which further excavations took place in Muge), and he rarely went out into the field. He would only do so when there was a need and when his duties in Porto allowed him to complete the journey. His disciples directed the excavation and this was held only in the presence of at least one of them.

The Arrchaeological Excavations E at Muge M Shell Midddens in the 193 30’s

25

Fig. 2.4. Draw wing of excavatted areas of Cab beço da Amoreeira from 1930 to t 1933.

Mendes Corrêa carefuully hired pro ofessionals whho specialized d in areas important too the recordiing of the miiddens prior to any archaaeological interventionn. Before the excavation e he hired a profeessional topog grapher to survey eachh one of the foour middens to t be studied at the time–C Cabeço da Arruda, Mooita do Sebasstião, Cabeço da Amoreiraa and Fonte do Padre Pedro (Fig. 2.6). A proofessional pho otographer alsso recorded the t same middens in pristine connditions. It iss the first tim me in the history h of Portuguese archaeology that t we see th his kind of syystematic atten ntion and accuracy abbout the detaills of the recorrding of an arrchaeological site prior to the archaaeological fieldwork. And we w are glad thhat Mendes Corrêa C did

26

Chapter Two

so because iit is thanks to him that we now have a ggood idea of how h these shell moundds stood out inn the landscape.

o da Amoreira shellmidden in n the 1933 Fig. 2.5. Skeeleton number 8 from Cabeço campaign.

Moita do Sebasttião, 2- Cabeço da Amoreira, 33- Cabeço da Arruda A and Fig. 2.6. 1- M 4- Fonte do P Padre Pedro. Phhotos 1, 2, and 3 were taken iin 1930 by a prrofessional photographerr. 4 was taken by Mendes Corrrêa in 1921.

The Arrchaeological Excavations E at Muge M Shell Midddens in the 193 30’s

27

Agricultuural workers,, mainly wom men, from thee Casa de Cadaval, a major landoowner where the t shell midd dens were situuated, were contracted c to do the ffieldwork witth a supervissor from the University of o Porto. Workers weere divided intto boys, wom men and men aand tasks werre defined by sex and age (Fig. 2.7). Women were responssible for siev ving even though theyy could be heelped by men n. The latter, with the help of the younger boyys, were respponsible for excavating e annd carrying th he dirt in wheelbarrow ws from its orriginal place to the area w where the wom men were sieving. Onn the many photographs p we w cannot seee women pusshing the wheelbarrow ws but somettimes we do see them exccavating with h hoes or shovels.

Fig. 2.7. Grouup of workers at a Cabeço da Arrruda shellmiddden in 1937.

Travels, Expensees and Rep orts Reports and receipts are a very impo ortant for undeerstanding thee logistics and manageement of a reesearch projecct in the firstt half of the twentieth century. The excaavations weree funded by Junta J de Educcação Nacionaal (JEN)– the Nationaal Education Board–the government g aagency that regulated matters perttaining to eduucation and reesearch in varrious degrees. In 1930 this institutioon awarded Mendes M Corrêaa a grant to fuund the researcch project into the shelll middens of Muge. At thee end of each yyear the direcctor of the

28

Chapter Two

excavation had to send JEN a final report of activities showing the goals set and met throughout the year, and how the money assigned had been implemented. The funds were spent mainly on train travel between Muge and Porto, in the purchase of recording materials (film packs, notebooks and the development of photographs), consumables (e.g. sieve nets), and on the payment of weekly wages and prizes to workers. In 1932 Rui de Serpa Pinto also received a grant from JEN to study abroad in England and France for a few months where he collected information about the investigations developing in Europe, made important contacts and exchanged bibliographies with researchers and libraries. Fieldworkers were paid by labour-day. On a weekly payroll the supervising archaeologists would register who had worked that week, and on which days, and how much each one should receive per day of what they called the “molhadura”, a quantity of wine. Some expense reports have shown us that women earned less than men (and they also had less wine than men) – which was current at that time. But because women were the ones sieving and collecting the small artefacts and men were the ones excavating – it was the latter who received prizes for finding human skeletons.

Casa Cadaval Casa Cadaval is a vast rural real estate that belongs to the Cadaval family which has a long history with royal roots, nobility titles and culture. The estate’s logistical support to the archaeological fieldwork at Muge has been important from before the 1930s until the present day. That is clearly shown in the correspondence exchanged between the family representatives and the archaeologists (Fig. 2.8). Casa de Cadaval was very important (and still is) in helping to protect the shell middens and giving the archaeologists a home during their stays in Muge. Contact was made between the Administrator of Casa de Cadaval– Armindo de Jesus–and the archaeologists from Porto, who maintained a cordial friendly relationship that has grown over the years as noted by the decrease in formality and the increased references to personal matters in correspondence. Armindo de Jesus was responsible for lodging the archaeologists, taking care of their welfare and hiring the necessary workers. He also provided agricultural tools to be used in the excavations, wheelbarrows for earthmoving, and timber to build the sieving structures. Some archaeological artefacts were deposited in the manor for display as evidenced by correspondence.

The Archaeological Excavations at Muge Shell Middens in the 1930’s

29

Without the help of Armindo de Jesus on behalf of Casa Cadaval it would have been a lot more difficult for the archaeologists to manage the site, their needs and the workers because of the geographical distance and poor means of communication at that time.

Conclusion The excavations by Mendes Corrêa stand out by the intensity of their fieldwork and the density of materials found. The importance of his research was stated by a number of publications and oral presentations at national and international levels. Also, the articulation between researchers is outstanding for that time taking into account the means of communication available and the big cuts in research projects as the country was drowning in an economic crisis and a dictatorship regime. His work reveals a clear definition of objectives–the search for evidence for the antiquity of the presence of Man in Portugal–and the application of correct methods which were scientifically accepted at the time but seldom applied in Portuguese archaeology and which resulted in good practice for archaeological excavations and the hiring of professionals to perform specific tasks. In 1932 his most beloved disciple–Rui de Serpa Pinto–died at a very young age and Mendes Corrêa became more involved in political life. After his work only a few resisted the decrease in interest for the study of Prehistory, choosing to look for the origins of a Portuguese nationality in more recent times (pre-Roman or Roman times). During the 40-year dictatorship in Portugal (1933-1974) and contrary to other similar contemporaneous regimes in Europe, archaeology was not used as an ethno-political argument and because of that there were very few projects, most of them with private financial support. Thanks to Mendes Corrêa’s academic and political status and his interest in archaeology and anthropology he was able to extend his connections to his own and other reputable archaeologists’ investigations, as was the case of Jean Roche in the Mesolithic shell middens of Muge. His political profile was of the utmost importance to enable him to gather both public and private financial support for his research and to support the others that followed.

30

Chapter Two

Fig. 2.8. Leetter written in 1931 by Armindo A de Jeesus (Casa dee Cadaval Administratorr) to Mendes Corrêa

The Archaeological Excavations at Muge Shell Middens in the 1930’s

31

Acknowledgements I would like to thank António Huet B. Gonçalves for saving the documents from being lost and for allowing me to study them.

References Abrunhosa, A., 2012a. As escavações arqueológicas nos concheiros mesolíticos de Muge: contributo para o estudo da historiografia das investigações da década de 1930. Unpublished Master Thesis in Archaeology, University of Porto. —. 2012b. Mendes Corrêa e o Homo afer taganus. Paper presented at the V Jornadas de Jóvenes en Investigación Arqueológica. Santiago de Compostela. 9-11May. 2012. —. 2013. Archaeology through photography: the case of Muge shellmiddens (Portugal) during the 1930’s. Paper presented at the VI Jornadas de Jóvenes en Investigación Arqueológica. Barcelona. 911May. 2013.in press Cardoso, J. L., 1999. O Professor Mendes Corrêa e a Arqueologia Portuguesa. al-madan IIª série(8): 138 - 156. Cardoso, J. L. & Rolão, J., 2002/2003. Prospecções e Escavações nos Concheiros Mesolíticos de Muge e de Magos (Salvaterra de Magos): Contribuição para a História dos Trabalhos Arqueológicos Efectuados, Estudos Arqueológicos de Muge, 1. Correa, A. A. M., 1917. Novos subsidios para a Antropologia portuguesa, in Asociacíon Español para el Progreso de las Ciencias, Congresso de Sevilla, Madrid: 142-150. —. 1933. Les nouvelles fouiles à Muge (Portugal), Paris, Institut International D'anthropologie: 1 - 16. —. 1934. Novos elementos para a cronologia dos concheiros de Muge. Anais da Faculdade de Ciências do Porto, XVIII. —. 1936. A propósito do Homo Taganus. Africanos em Portugal, Boletim da Junta Geral do Distrito de Santarém. Gonçalves, A. H. B., 1983. O Eng.º Dr. Rui de Serpa Pinto – Estudo Biobibliográfico, Arqueologia, 76:1-7. —. 1983/1984. Rui de Serpa Pinto - o Homem e a Obra, Actas do Colóquio Inter-Universitário de Arqueologia do Noroeste, Homenagem a Rui de Serpa Pinto, Portugália, Nova série, Vol. IV-V, Instituto de Arqueologia, Faculdade de Letras da Universidade do Porto, Porto: 9-12.

32

Chapter Two

—. 1986. Inéditos de Rui de Serpa Pinto sobre as escavações arqueológicas de Muge. TAE, nº 26 (1-4): 211 - 220. Monteiro, E., 1959. Professor A. A. Mendes Corrêa, Trabalhos de Antropologia e Etnologia, Volume de Homenagem ao Prof. Mendes Corrêa, Porto, 17(1-4): 5-8. Pinto, R. d. S., 1931. Nouvelles Recherches sur le Miolithique en Portugal. Association Française pour l'Avancement des Sciences, LV sessions Extr. des C. rendus du Congrès de Nancy: 327 - 329. —. 1932. Notas sôbre a indústria microlítica do Cabêço da Amoreira (Muge). Asociación Española para el Progreso de las Ciencias. —. 1933. Sur la taille du silex, à Muge (Portugal). C. P. d. France. NimesAvignon, Bureaux de la Société Préhistorique Française, Paris: 219222. —. 1986. Les fouilles du Kjoekkenmoedding de Cabeço da Amoreira, à Muge (Portugal). Trabalhos de Antropologia e Etnologia, 26(1-4):221229. Ribeiro, C., Ed. 1866. Description du terrain quaternaire des bassins du Tage et du Sado. Lisboa, Typographie de L'Académie Royale des Sciences. VV. AA., 2012. Mendes Corrêa (1888-1960) entre a ciência, a docência e a política, Coord. MARTINS, A. C., ACD Editores, Lisboa.

CHAPTER THREE SOURCES FOR THE RECONSTRUCTION OF CABEÇO DA ARRUDA MARY JACKES1, DAVID LUBELL1, PEDRO ALVIM2 AND MARIA JOSÉ CUNHA3 1

Department of Anthropology, University of Waterloo, Waterloo, ON, N2L 3G1, Canada. [email protected] 2 Department of Archaeology, Durham University, Durham, UK, DH1 3LE [email protected] 3 Museu de Antropologia e Pré-História Mendes Correia, Museu de História Natural, Praça Gomes Teixeira 4099 - 002 PORTO, Portugal. [email protected]

Abstract Archival resources provide us with some limited information on the excavation of Cabeço da Arruda in the spring and early summer of 1880. Further excavation in the 1880s is not well recorded, but we can make some comments. We will locate the excavations within the site and discuss the evidence of burials up to the present century. A greater understanding of the site and its mortuary archaeology allows us to present updated information on the human skeletal sample. Cabeço da Arruda is a late Mesolithic site in central Portugal which lies on the northern bank of the Muge River, a tributary of the Tagus. Here, midden deposits lie on a terrace remnant which rises above the surrounding land and overlooks the broad marshy valley of the river. This site has a long history, beginning with its discovery in 1863 and the excavation in 1864 by Carlos Ribeiro (1867:715). The significance of the

34

Chapter Three

first description by Pereira da Costa (1865) of materials from the site is outlined by Jackes et al. (this volume). Jackes and Meiklejohn (2004) made a first attempt at assessing the minimum number of individuals found at the site over the period from 1880 to the 2000s. The study did not include the material excavated in the 1930s. Because the excavation of Arruda has a more complex history than other Muge sites, it is more difficult to determine the details needed for a mortuary or demographic study. We need to know what has been lost and how much more might be found. The basic question is the relationships between the Arruda skeletons: where were they all found between 1864 and 2000?

Fig. 3.1. Cabeço da Arruda, 1956 aerial © IGP, 2012 overlain by a tracing after a 1930 contour map. White areas indicate previous excavations and vegetation to the right was replaced by the rice fields evident on the aerial photograph. The inset shows the estimated locations of skeletons excavated in April-May 1880 (grey circles to the right) and in the 20th century (dark circles, 1937, and rectangles, 1964-5). The image is oriented to accord with approximate cartographic north (up).

Sources for the Reconstruction of Cabeço da Arruda

35

Fig. 3.1 shows a 1956 aerial photograph. Rice was being grown in the mid-twentieth century and flooding led to the development of a peninsula. The peninsula was not present on old maps and is no longer shown, but on a 1947 aerial photograph it is even more prominent and is also emphasized in a contour map drawn in 1964 by Veiga Ferreira who collaborated with Jean Roche in the 1960s (Jackes and Meiklejohn 2004: Fig 9 ). In Fig. 3.1 contours overlain on the aerial photograph come from a map found at the Geological Museum, Lisbon in 1989. We see here, in white, the areas generally according with nineteenth-century archaeological work. It appears that the 1930s excavations by Mendes Correia2 have not yet been undertaken and we can confirm the map is from 1930 (Abrunhosa, 2012: 122). The inset shows, on the left, our reconstruction of the location of the skeletons found by Mendes Correia in 1937 (dark circles), while the rectangles indicate the locations of the Roche skeletons. Upper level skeletons were reported for the first time in 1964, shown by the uppermost rectangles. The four upper level skeletons lay more than four metres above the nine lower level skeletons (rectangles grouped to the right), which were stated to be directly on the sand. In 2000, Rolão found a child’s skeleton in or on the sand, close by the 1964 deep skeletons (Roksandic 2006, pers. comm. 2013; Rolão 2004, pers. comm.). But just to the east, there was another skeleton only 30 cm below the modern surface which was fragmentary, as were the upper skeletons from the 1960s (Roche 1974: 30). The exact positions of the skeletons excavated in 1864 cannot be specified, but our reconstruction of their locations (Jackes et al., this volume) suggests that they were found in the area of the southern (bottom) lobe of the central white area shown in Fig. 3.1. That would accord with all the clues provided by Pereira da Costa (1865). The 1864 excavation would have extended far to the west of that area (all along the southwest face of the mound), but we know that skeletons were found only in a restricted area in the eastern part of the trench. Clear signs of the trench would have been removed by flooding which occurs frequently (Azevedo et al. 2004). Roche (1967:80) recorded that in 1966 the flood waters reached half-way up the profile, that is, perhaps five metres above the average level of the Muge. A photograph from 1880 (Jackes and Meiklejohn 2004: Fig 13) clearly indicates the situation at the time of a sketch of the site dated 25 April 1880: part of this sketch is reproduced here as Fig. 3.2. Both the 2

The spelling accords with the birth certificate of A. A. Mendes Correia (cf. Corrêa).

36

Chapter Three

photograph and the sketch show skeletons mostly lying within four metres of the back wall of the trench. On the other hand, the plot published by Jackes and Meiklejohn (2004: Fig 12), showing an overall rough plan of the site which must derive from May 1880, accords with the transept published here as Figure 3.3, in which skeletons lie around 12.5 metres from the trench wall. The locations of the skeletons sketched on 25 April are indicated by no more than wavy lines on this wider plot, while the transept proves the skeletons had been lifted and that part of their former location had been excavated down to the sterile terrace sands. On the Figure 3.1 inset, we can see the two sets of skeletons, their positions indicated by the light circles: one group further in to the centre of the mound which was excavated in late April, and the other (less accurately recorded) to the south closer to the marsh, excavated in May and probably retained under shelter through the dry summer as an exhibit for participants at the 1880 Congrès International d'Anthropologie et d'Archéologie Préhistoriques. As well as resources like these plots which were found in the Geological Museum, Lisbon in 1989 and copied by Lubell, giving an idea of the excavation at two points in time in the first half of 1880, there are also written reports which were copied by Alvim ten years ago when they were in the Instituto Geológico e Mineiro archives at Alfragide. These reports provide more answers and help date the transept and broader sketch plot. We have a series of short reports to Carlos Ribeiro in Lisbon, written by the man who was actually excavating at Muge–Manuel Roque d’Oliveira (see Jackes et al., this volume for details). The record begins on 19 March 1880 when the hired local men were moving earth that had previously been dug; by 21 March they had found a skeleton even though they were still moving back dirt. By the 23rd they were already four or five metres down and had dug a 15 x 4 metre trench. Obviously they were following the bottom of the (partly erosional, no doubt) scarp shown in the sketches of the site. In a letter which perhaps dates to 25 March, Roque reported on another skeleton, but he was getting frustrated and planning to move east and south, that is, further from the centre of the mound. He noted that local men had seen skeletons in that area. By 27 March he reported that skeletons had been found together with animal bones, four metres below the surface. Within a month they had dug a great deal of deposit to the east, judging from the 25 April sketch shown in Fig. 3.2. On 3 May, Ribeiro visited and wrote in his notebook that they had found 13 skeletons and explored 20 to 30 metres (this accords with the

Sources for the Reconstruction of Cabeço da Arruda

37

distance of 25 metres that we estimate as the extent of the trench shown in Fig. 3.2). Ribeiro was apparently noting the skeletons to the east on the Fig. 3.2 plot: they numbered exactly 13 on the 25 April.

Fig. 3.2. Part of a sketch dated 25th April 1880 held in the Geological Museum, Lisbon in 1989, showing the location of 15 skeletons (the dots marking 13 of the skulls have been enlarged). The sketch is shown with estimated magnetic north for April 1880 oriented to accord with Fig. 3.1.

On 14 May, Roque stated that he had made transects across the mound, with one going over the skeletons to make the situation clear to Ribeiro–no doubt the transept reproduced here as Fig. 3.3. Next, on 19 May, Roque reported two more skeletons and said that as soon as all the work was finished near the skeletons they would move near to the big cork oak; this is shown in the plot published by Jackes and Meiklejohn (2004: Fig. 12) as the large black shape to the right–that is, they expanded the trench further to the south and east. Roque, by that time, also had men digging down to the sands at the deepest part, 5.5 metres below the surface. That work would have been carried on to the northwest, closer to the centre of the mound. However, now the focus switched to Moita, where–as Roque reported on the 1 June–the excavators had uncovered the apparent mass burial of a group of perhaps 16 skeletons described in Jackes and Alvim (2006).

38

Chapter Three

A man called Scolla was left in charge at Arruda and more work was done in the southern area. But, on 11 June, Roque reported that the excavations were also being extended to the northwest. There is, in addition, the record of nine skeletons found in a 1 x 2 metre area, all piled together like those found at Moita. By 21 June, three more skeletons had appeared. And that is almost all the information we have on the Arruda burials from 1880.

Fig. 3.3. A transect from May 1880. This transect runs from the high point of the mound, across the excavation edge, where a skeleton is shown on a pedestal, across an area from which the skeletons shown in Figure 2 have been removed, down to the level of the terrace sands at 2 m. The location of skeletons left in situ under a shelter for the 1880 Congress is indicated inside a pile of back dirt and the slope down to the edge of the mound. The trajectory of this transect (a – a’) is not clearly specified on any sketch, but can be estimated to lie at ~42 degrees west of cartographic north.

What else is there for 1880? There are two plates from the Ribeiro (1884) publication which are in fact parts of the same image. The correct reversed orientation was published in Cartailhac (1886) as an engraving which was stated to come from a photo in the Geological Museum, Lisbon, showing a wider view than Ribeiro’s plates. Perhaps the burials shown were those protected by a shelter until displayed to the participants in the Congrès International d'Anthropologie et d'Archéologie Préhistoriques, held at Lisbon in late September 1880. First visiting Moita, the visitors

Sources for the Reconstruction of Cabeço da Arruda

39

were given a sumptuous lunch and then a hardy few crossed the Muge valley to see the more extensive excavations at Arruda. While the skeletons shown in the 1880 Congress publication plates appear similar to the 13 described by Ribeiro on 3 May–in parallel lines, and not dispersed–there are fewer than 13. That suggests the plates were indeed illustrating the burials retained under the shelter. However, there is one burial grouping not yet mentioned: on 3 May, Ribeiro discussed another group of burials–nine skeletons found at a depth of four metres in which the layout was apparently different, that is, not aligned. He mentioned that the slight differences in height made it seem as though the skeletons were distributed around a broad basin. The plan drawn by Ribeiro of this significant grouping has not been published and this notebook was apparently retained by Veiga Ferreira. He had had many of the resources discussed here and marked and wrote on them before returning them to the Geological Museum sometime between 1986 and 1989. Unfortunately, since this particular notebook was retained by Veiga Ferreira, it is not available to researchers although it has been partly published (Cardoso and Rolão 1999/2000: 121-122). Sadly, Ribeiro was already ill in 1880, and he and Manuel Roque d’Oliveira both died in 1882. Further excavation was undertaken in 1884 (Paula e Oliveira, 1889: 59). Thirteen burials were excavated that year. Very few, if any, of those 13 skeletons came from Moita: Arruda is the more likely source (Jackes and Alvim 2006). Paula e Oliveira noted that some skeletons were in very poor condition, so there could have been upper level skeletons among these, especially as Arruda N in the Geological Museum is 1000 years younger than the oldest Arruda burial excavated in 1937 (Jackes et al., 2014). There was one more image in the Geological Museum – an unpublished photograph. Based on the burial mode, it is clearly from Arruda, perhaps looking towards the eastern wall of the trench near the big cork oak. This would be at the edge of Veiga Ferreira’s record of previous excavations, where the sand and the burials were very close to the surface slope of the mound. Perhaps this image came from late May or early June 1880. We have no record for the 13 burials from 1884 and the work in 1885 at Arruda was not productive (Paula e Oliveira 1889: 59; Jackes and Alvim 2006: 96 ftn 11). So our study moves to January 1928, when we know from one photograph in the Museu de História Natural, Porto that a preliminary visit was made to Arruda. In 1933, while digging at Amoreira, Mendes Correia began some work at Arruda laying out east-west and north-south test trenches in 2.5 metre

40

Chapter Three

squares. Then in 1937, a large scale excavation was undertaken. Apart from fragments, many mixed in with faunal remains perhaps as a result of the extensive erosion which can be seen to have occurred before June 1932 (Abrunhosa 2012: Document nº 32-07: 228), most of the in situ burials came from a very restricted area (marked by dark circles on Fig. 3.1 inset), except for a fragmentary small child found in 1933. All were found just above the terrace sands. All subsequent finds have been in the same area as the 1937 burials.

Fig. 3.4. From Porto positive xxvi.25.B, neg 1937/5. The profile of part of the Mendes Correia excavation, showing one full 2.5 meter square (for comparison with the 1864 profile and Roche 1974 Plate 3 reversed). The profile shown reaches down only to the beginning of the deep layer.

Fig. 3.4 is part of one of two photographs in the archives of the Museu de História Natural, Porto which gives us a very clear idea that the deposits exposed in 1937 were similar to those illustrated by Roche (1967: Fig. 2; 1974 Pl. III printed reversed). Cartailhac (1886: 55) described equivalent deposits that had been shown to him on a visit to Arruda, as we can determine from his statement that the 1865 illustration of the Arruda

Sources for the Reconstruction of Cabeço da Arruda

41

profile was similar (the angle of the deposits appears to deepen slightly, further into the mound). It appears that a broad swathe along the mound produced skeletons from 1864, 1880, 1884, and 1937, 1964 and 2000 and the deposits below these also remained consistent, with the skeletons described as being just above the sands except for the burials furthest into the mound, below the highest point, which were excavated by Roche and Veiga Ferreira. Those skeletons are said to lie on the sand (Roche 1974: 27) while the 1937 burials are noted as between 25 and 140 cm above the sand (derived from field books, Cardoso and Rolão 1999/2000: 174-179). Are there other areas of the mound that may contain skeletons? We know that the 1864 excavations found skeletons only in the eastern portion of the trench and that Mendes Correia considered that the 1880s work was the most extensive and deepest in the eastern part of the mound (Abrunhosa 2012: 363). No doubt skeletons to the south, and possibly to the west, could have been lost to flooding, and there may well be deep burials below the present scarp, as well as some upper level skeletons still below the top of the mound. However, there are indications that some areas have no further skeletons. We know there was an excavation at Arruda in 1885 which was so unproductive that the campaign for that year was switched to Moita. On 4 June 1885, Paula e Oliveira wrote a letter to Nery Delgado in which he said that the work at Arruda had been “fruitless”. Fig. 3.1 shows a flat area to the east of the mound and an indication of an old excavation. A cadastral map (IGP 1960) indicates that the NE quadrant was disrupted, and partially removed from the mound, and this is echoed in Veiga Ferreira’s 1964 map at exactly that point. In each case, these indications of old excavations were to the northeast and it is very possible that this was the location of the 1885 excavation. Paula e Oliveira would not have been able to specify that there were burials only to the south if he did not dig in the north. He contrasted Moita, where the burials were in the northeast part of the mound, with Arruda where the burials were all located in a southern quadrant (Paula e Oliveira 1889: 74). It is unlikely then, that the burial of the dog found in 1880 at four metres would have come from that area (cp. Detry and Cardoso 2010: Fig. 5). Our 2004 demographic study of the 1880s’ Arruda materials counted 97 individuals. Subtracting the 1884 maximum of 13, it seems that 84 burials were excavated in 1880. However, only about 50 are mentioned in the 1880 records, indicating that those records are indeed incomplete. Because of this, it is likely that further 1880 work pulled back the excavation deeper into the mound. Judging from the difference between the early 1880 scarp and the scarp surveyed in the 1950s (IGP 1960), it seems that there must have been an excavation in that area and, indeed,

42

Chapter Three

further along to the west. There was erosion, no doubt, but given the fast pace of the 1880 excavation (once they learned to keep areas of potential interest free of spoil heaps), the digging of deposit is likely. The area of around 48 square metres to the southeast alone had the potential to add perhaps 30 skeletons during the continuing 1880 campaign. In comparison, between 1937 and 2000 at least parts of 21 deep level skeletons were found within an area of about 13 square metres. Our suggestion is, then, that we should be able to undertake a demographic study of the surviving Arruda skeletons, given that no excavator has described an area in which there was the preferential burial of adults or children, and that a reasonable sample representing a broad area of the mound has been retained and is available to us. On that basis, we can add individuals from the 1964 basal sands’ excavation (Jackes and Meiklejohn 2004: 102), and now materials from the deeper levels of the 1937 Arruda excavation (in the Museu de História Natural, Porto), to the 1880s material for a new total fertility rate (TFR) estimation. Previous estimates (Jackes and Meiklejohn 2008: 232) of higher fertility at Arruda (c. 6.5) than at Moita (c. 4.5) were based on a minimum of 105 individuals (MNI, derived from mandibles). Now we add the upper level 1964-5 individuals and attempt to arrive at a better understanding of how many burials were actually encountered in the 1930s. The materials in Porto are difficult to study because of possible mixing and loss, but, after studying the Porto material in 2010, we can confirm that five juvenile mandibles need to be added to the ten individuals numbered by Mendes Correia, for a new total of 124 individuals, which may still exclude one or two Porto adult mandibles no longer firmly assigned to a Muge site. Our method of arriving at a TFR allows us to compare the estimates using a 95% CI from two estimators: a wide range would arise from a discrepancy between the two estimators indicating a defective sample, in this case an underrepresentation of adults. The estimates of around seven to eight children live-born to the average woman of reproductive age is too high a rate for a group that is no more than semi-sedentary. As a result, it is important that we acknowledge the loss of the 1864 skeletons. The underrepresentation of adults could well result from that loss, suggesting a TFR that is probably too high for a late Mesolithic group. The addition of 45 adults to the Arruda sample would give us an estimate of around five live-born children for the average woman during her reproductive years (the total range is still too wide since not all of the 1864 skeletons were fully adult). It is clear that we can make no more than a guess, but that all indications are still that the Arruda TFR was slightly higher than that of the Moita sample, at least 5 to 5.5.

Sources for the Reconstruction of Cabeço da Arruda

43

Acknowledgements We are grateful to Dr. M. Magalhães Ramalho, Coordenador, Museu Geológico, Laboratório Nacional de Energia e Geologia for permission to examine Muge materials. MJ acknowledges the bequest made to her by Esther Palmer which funded the scanning and recording of the negatives and prints, which are used or referred to with the permission of Prof. Doutor José Luís Santos, vice-director of the Faculdade de Ciências do Porto.

References Abrunhosa, A., 2012. As escavações arqueológicas nos concheiros mesolíticos de Muge: contributo para o estudo da historiografia das investigações da década de 1930. Master’s thesis, Faculdade de Letras Universidade do Porto. Azevêdo, T.M., Nunes, E., & Ramos, C., 2004. Some morphological aspects and hydrological characterization of the Tagus floods in the Santarém region, Portugal. Nat. Hazards 31, 587–601. Cardoso J.L. & Rolão J.M., 1999/2000. Prospeçcões e escavações nos concheiros mesolíticos de Muge e de Magos (Salvaterra de Magos): contribuição para a história dos trabalhos arqueológicos efectuados, Estud. Arqueol. Oeiras 8 Cartailhac, E., 1886. Les Ages Préhistoriques de L’Espagne et Portugal, Reinwald Librarie, Paris. Detry, C. & Cardoso, J.L. 2010. On some remains of dog (Canis familiaris) from the Mesolithic shell-middens of Muge, Portugal. J. Archaeol. Sci. 37, 2762-2774. IGP, 1960. Secção Cadastral N6, Freguesia de Muge, Concelho de Salvaterra de Magos, Distrito de Santarém 1:2000 Instituto Geográfico Português www.dgterritorio.pt used under copyright 22 August 2012. Jackes, M., & Alvim, P., 2006. Reconstructing Moita do Sebastião, the first step. In: Bicho, N.F., Veríssimo, H. (Eds.), Do Epipaleolítico ao Calcolítico na Península Ibérica: Actas do IV Congresso de Arqueologia Peninsular, Faro, Universidade do Algarve, pp,13–25. Jackes, M., Lubell, D., Cardoso, H.F.V., Anacleto, J.A., & Meiklejohn, C., 2015. Cabeço da Arruda in the 1860s. This volume. Jackes, M.K., Lubell D., & Cunha M.J., 2014.. Redating a Mesolithic skeleton from Cabeço da Arruda, Muge, Portugal. Mesolith. Misc.22(2), 40-44.

44

Chapter Three

Jackes, M., & Meiklejohn, C., 2004. Building a method for the study of the Mesolithic-Neolithic transition in Portugal. In: Budja, M. (Ed.), Neolithic Studies 11, Doc. Praehist. 31, 89–111. Jackes, M., & Meiklejohn, C., 2008. The paleodemography of central Portugal and the Mesolithic-Neolithic transition. In: Bocquet-Appel, J.P. (ed.) Recent Advances in Paleodemography: Data, Techniques, Patterns. Springer, Dordrecht, pp. 209–258. Paula e Oliveira M. F., 1889 Nouvelles fouilles faites dans les kioekkenmoeddings de la vallée du Tage (posthumous publication). Comun. Com. Trab. Geol. II (i), 57-81. Pereira da Costa, F.A., 1865. Da existencia do homem em epochas remotas no valle do Tejo. Primeiro opusculo. Noticia sobre os esqueletos humanos descobertos no Cabeço da Arruda. Imprensa Nacional, Lisbon. Ribeiro, M. C., 1867. Note sur le terrain quaternaire du Portugal. Bull. Soc. Geol. Fr. Séance du 17 Juin 1867, Series 2, 48, 692–717. Ribeiro, C. 1884 Les Kioekkenmoeddings de la Vallée du Tage Congrès international d'anthropologie et d'archéologie préhistoriques. Compte rendu de la neuvième session a Lisbonne 1880. Typographie de l’Academie Royale des Sciences, Lisbonne, pp. 279-290. Roche, J., 1967. Note sur la stratigraphie de l’amas coquillier mésolithique de Cabeço da Arruda (Muge). Comun. Serv. Geol. Port.. LII, 79–94. —. 1974. Sépultures de l’amas coquillier mésolithique de Cabeço da Arruda (Muge). Actas do Congresso Nacional de Arqueologia, 25–36. Porto Roksandic, M., 2006. Analysis of burials from the new excavations of the sites Cabeço da Amoreira and Cabeço da Arruda (Muge, Portugal). In: Bicho, N.F., Veríssimo, H. (Eds.), Do Epipaleolítico ao Calcolítico na Península Ibérica: Actas do IV Congresso de Arqueologia Peninsular, Faro, Universidade do Algarve, pp. 43–54.

CHAPTER FOUR CABEÇO DA ARRUDA IN THE 1860S MARY JACKES1, DAVID LUBELL1, H.F.V. CARDOSO2, J. A. ANACLETO3 AND C. MEIKLEJOHN4 1

Department of Anthropology, University of Waterloo, Waterloo, ON, N2L 3G1, Canada. [email protected] 2 Department of Archaeology, Simon Fraser University, Burnaby, BC, V5A 1S6, Canada. [email protected] 3 Museu Geológico, Rua Academia das Ciências, Nº. 19 – 2º. 1249-280, Lisboa, Portugal. [email protected] 4 Department of Anthropology, University of Winnipeg, Winnipeg, MB, R3B 2E9. Canada. [email protected]

Abstract The identification of osteological evidence from the earliest excavation of Cabeço da Arruda in 1864, still preserved in two Lisbon museums, provides the occasion for a reminder of the importance of the burials in the history of European anthropology. The material is briefly summarized, together with a discussion on the location of the finds. An AMS analysis of a charcoal sample contained within a skull provides a date consonant with other evidence.

46

Chapter Four

Introduction: the First Publication Cabeço da Arruda is a Portuguese Mesolithic site on the northern bank of the Muge River consisting of midden deposits formed on a remnant of terrace sands overlooking the marshy valley. Following its discovery in 1863 and its excavation in 1864 by Carlos Ribeiro (1867: 715), da Costa (1865) published the first description, noting the excavation of at least 45 human skeletons (Pereira da Costa 1865: 13), mostly in one level and all in one area. In order to situate these burials, the published cross-section and profile must be scaled (Pereira da Costa 1865: 6, Figs. 1 & 2). Our scaling (Fig. 4.1 and 4.2) derived from statements that the mound was 40 x 95 m, rising 5 m above the sand (da Costa 1865: 4), the burials mostly occurring in a 70 cm thick level. The scaling is only approximate because the mound cross-section was idealized. Excavations could not have reached the high point of the mound. Ribeiro (1884: 282) said the mound rose 7 m above the sands, based on the 1880 excavations where 5 m of midden deposits were exposed closer to the high point: the 1880 excavation face was still ~2 m below that point (Jackes et al., this volume). Furthermore, the upper part of his Fig. 2 image (Fig. 4.2 here) appears to represent not the actual excavation profile, but an artistic rendering of the topography of the mound above that cut. Since the angled cross-section was rendered as a profile image at 90 degrees, ~10% change in height might be expected.

Fig. 4.1. Reconstructed cross-section of the 1864 excavation from da Costa (1865, Fig.1). The italic letters correspond to those used on the original.

Cabeço da Arruda in the 1860s

47

Pereira da Costa’s section (1865: 6, Fig. 1) imagines the original, preerosion, mound edge at point A’. Rather than envisioning the mound prior to erosion, our Figure 1 begins at A”, showing a reconstruction of the excavation based on the 1865 section through the site, from roughly SW to NE, corresponding to da Costa’s points A” and B. The section begins at an erosional edge stated to be 10 m in from a drainage ditch: our scaling is such that da Costa’s A’ to A” equals 10 m. Because the excavation floor and profile were both illustrated as angled, we estimate that burials were exposed in a limited portion of the excavation (Fig. 4.1), even though the trench cut into the mound ~18 m. All burials were in a restricted layer which overlay 2 m of the sterile terrace sand that was exposed at the edge, covered with eroded archaeological materials.

Fig. 4.2. Scale suggested for the “profile” of the 1864 excavation, facing along the marsh, from da Costa (1865, Fig. 2).

The reported 95 m length of the mound accords with Veiga Ferreira (Cardoso and Rolão 1999/2000: 225, Fig. 56). We have scaled Pereira da Costa’s profile at 25 m length (Fig. 4.2) on the basis that the general horizontal of Layer D is specified as 70 cm high, leading to the profile having a height of 5 m, with the actual midden deposits just over 4 m high. This agrees with evidence that the excavation did not extend far into the mound with the profile not extending down to the sand. Pereira da Costa (1865: 4) stated that the excavation extended along the southwest face of the mound. Our 25 m length and the location of the profile fit well with changes in the orientation of the eroded scarp edge, shown in maps from 1880 and the 1960s (Jackes and Cunha, 2015). There is also concordance with an area of previous excavation noted on a 1930 map (Jackes et al., this volume; see also Jackes and Cunha, 2015). Since Pereira da Costa stated that burials were found in Layer D, with very few in F’, the skeletons were within a restricted area of ~10 m in length on the basis of our scaling. The description of the burials as above the sands, and underlying a layer (C) containing fire-cracked pebbles,

48

Chapter Four

broken mammal bones and charcoal, concurs with the 1880 sources. Skulls were crushed, with knees folded on the thorax in the one fully described burial, typical again of skeletons excavated in 1880. Individuals were of diverse ages. We can never be sure how many individuals were found at Arruda in 1864, but it seems likely that at least 45 were excavated. A clustering of burials was not unusual. While “mass burials” were found both at Moita and Arruda in 1880 (Jackes and Alvim 2006; Jackes et al. this volume), these included only ~16 and nine individuals, found in areas of only ~2 m x 4 m and 1 m x 2 m, respectively. In fact, Pereira da Costa (1865: 13) quotes the excavators as having said that the skeletons were in a restricted space with heads to the northwest, reminiscent of the Moita mass burial where skeletons were roughly oriented in the same direction. However, the burial description suggests the varied dispositions of cadavers, unlike Moita. Da Costa noted dispersed bone, indicating purposeful burial over a lengthy period, with the disturbance of earlier by later interments (Pereira da Costa 1865: 16): he also noted that variations in damage were probably caused by the weight of the sediment (Pereira da Costa 1865: 14). We do not know whether da Costa actually saw most of the bones, but his sophisticated discussion of soft tissue decomposition resulting in voids into which bones could fall, with articulations flattening or undergoing torsion (Pereira da Costa 1865: 19) indicates that he saw some material en bloc, especially as he also discussed the development of calcareous breccia. With Ribeiro, he was one of the two directors of the Geological Survey of Portugal in the early 1860s, but Ribeiro had been in charge of the excavation: since this was unacknowledged it perhaps led to discord between the two men (Antunes 1986: 795). Ribeiro was a military man and civil servant in charge of mines with an interest in lithics. Although Pereira da Costa was Professor of Mineralogy, he had medical training and was certainly knowledgeable about human bones. Pereira da Costa (1865:14) simply states that a collector of the Geological Commission of Portugal provided information. At that time, there were four collectors (Anon. 1899) and one of them, Manuel Roque, would later work for Ribeiro at Arruda in 1880 (Jackes and Alvim 2006), and might have been on site in 1864. Perhaps only the material described in detail (Pereira da Costa 1865: 14) was taken to Lisbon. First, an almost complete skeleton was apparently still brecciated: the description of hyperflexion of the legs onto the trunk and other features of the disposition make it evident that this was similar to many burials found since (Jackes and Lubell, in press). Secondly, an almost complete skeleton with the skull broken after

Cabeço da Arruda in the 1860s

49

excavation was found near the first, probably a young female, buried with knees only slightly raised. Thirdly, there were a number of crushed skulls, some perhaps associated with non-conserved bones (suggesting that much material was not brought to Lisbon), some mixed pell-mell with other bones (Pereira da Costa 1865: 15). These skulls could be crushed bilaterally, or anterior-posteriorly; others were separated at suture lines. The conclusion was that the differences related to burial position and soil pressures. Pereira da Costa (1865: 15), while emphasizing the number of skeletons in such a small space, was at pains to disprove any idea of a massacre.

What We Know about the Human Finds Pereira da Costa (1865) published descriptions and illustrations of some of the Arruda finds. Dentitions were listed as follows (numbering here and in Figure 4.3 follows Pereira da Costa, 1865: 23-4): 1) a mandible of an elderly individual with heavy wear but no pathology (obviously incorrect); 2) a jaw of a younger individual with curved plane of wear; 3) a mandible broken at the symphysis during excavation, morphologically differing from the previous two, with unusual molar dentine, yellowed and deep below thick white enamel remnants; 4) maxilla with displacement of M1 and M2 into the sinus; 5) dentition with both maxilla and mandible, with a small lesion in the left upper M1, noted as very rare, but with major interest in unerupted left canines. We searched for casts in the Geological Museum in Lisbon in the 1980s but were told there were none. These have now been found and some have been identified as from the Arruda 1864 excavations, namely, maxillae 4 and 5 and pathological mandible 1 (Fig. 4.3). One maxilla has an original paper label “3”, and the other is labelled “4” (Fig. 4.3, no. 4). Of two additional mandible casts, one was illustrated by Delgado (1867, Plate I, 5), coming from a specific area in the Neolithic level at Casa da Moura, and associated with a skull (Delgado, 1867, Plate I, 1) which was also cast. A further mandible retaining seven teeth but with central incisors and left lateral incisor and canine missing, cannot be identified with certainty. Information on the casts was published in England when their receipt by the Ethnological Society was acknowledged (Busk 1869). The Royal Anthropological Institute took over the Ethnological Society collections

50

Chapter Four

but no longer holds the casts (Sarah Walpole, RAI Archivist, in litt. 17/4/2013) which were sent to the Royal College of Surgeons based on a letter dated 9 March 1897 asking that casts of skulls be returned to the RAI. They are no longer held by the RCS (Sarah Pearson, Curator, Hunterian Museum, in litt. 2/9/2013).

Fig. 4.3. Dentitions from Arruda 1864 excavations, after Pereira da Costa (1865, Plates I, II): numbering follows description in da Costa (1865:23-24).

Casts were also sent to the Museum of the Société d’Anthropologie de Paris (de Quatrefages and Hamy 1882: 33), as acknowledged in 1867 (Pruner-Bey 1868: 33). Three skulls were included in the gift of casts, one being described as too crushed to be informative on “the question of the prehistoric human races of western Europe”. The statement that most mandibles were from Casa da Moura, and that at least five had chins differing in morphology from others, raises the question of how many casts were sent to Paris. Finally, there is mention of a massive mandible from Arruda that Pruner-Bey believed to be from a dolichocephalic skull, unlike the mandibles with chins which he associated with brachycephalic skulls. De Quatrefages and Hamy (1882: 33-34) make it clear that only

Cabeço da Arruda in the 1860s

51

two maxillae were sent, both from Arruda, but confusingly say that two of the skulls were from Casa da Moura. Pereira da Costa (1865, Plates II-V) published several images of one skull (Fig. 4.4) that became famous, illustrated again by Paula e Oliveira (1884 Plate III 5 a, b and c; b printed reversed and c “rectified”) and by Cartailhac (1886: 319, Fig. 446) where it was published as a “rectified outline”, based on the cast. It was mentioned as Skull 2 (de Quatrefages and Hamy 1882: 134) and, while one cast of it in Lisbon is labelled “170”, another has an original paper label “2”. The label “170” is mentioned for a cast having an estimated cephalic index of 86.4 in a letter from Barbosa Sueiro to Mendes Correia 3/i/1931 (Abrunhosa 2012: 168).

Fig. 4.4. Skull 2 from Arruda 1864 excavation, after Pereira da Costa (1865, Plates II-V).

52

Chapter Four

The interest in Skull 2 was its shape, expressed by the cranial or cephalic index. Retzius introduced the concept in the 1840s (see Stewart, 1939, for an exhaustive history of this index). The maximum length and maximum width across the parietals are the relevant measurements, with the breadth expressed as a percentage of length. The maximum length would be measured from the glabella and any value above 80 was “brachycephalic”. Paula e Oliveira (1884: 294), discussing what he called “the brachycephalic male skull”, reported Skull 2 as having a cephalic index of 82.56 (Paula e Oliveira 1884: 296), stating that he “rectified” the norma verticalis image. A comparison of the illustrations demonstrates that brachycephaly was emphasized: Pereira da Costa’s image suggests a breadth ~81% of length and the 1884 image a breadth 89% of length, very slightly more than Cartailhac’s outline. All "rectifications" were based on casts. Vallois (1930) searched in vain for the actual skull in the 1920s. Arruda got caught up in convoluted arguments about the meaning of skull shapes and European racial history. Particular focus was on whether the Furfooz “race” was represented at Arruda (de Quatrefages and Hamy 1882). Of two skulls excavated at Furfooz in Belgium in 1867, one was said to be brachycephalic, the other mesocephalic. This led to a complex of theories, the simplest being that broad-headed people migrated into Europe, mixing with long headed Cro-Magnon survivors. Pereira da Costa noted that compression had no doubt affected Arruda cranial measurements, yet Hervé (1899) would say that Muge demonstrated “a first ethnic mixing” with newly-arrived brachycephalics and Coon (1939: 63-64, 558) was still discussing whether Muge was relevant to the European racial makeup. While discussions still continue on skull form (Gravlee et al. 2003a, b), no-one doubts that the cephalic index varies. Arruda adult skulls in the Geological Museum, Lisbon, have a mean index of 75.6 (female mean 76.6, n 6; range 70.2–83.2: male mean 74.6, n 6; range 69.4–81). Two females and one male have an index above 80. The Moita adult mean (75.9, n 23) is little different, but no values rise above 80 (male range 67.5–77.6, n 9; female range 66.3–79.5, n 14). A better understanding of normal variability and skeletal plasticity ensures that we no longer build ethnic history on minor variations of one trait. Pereira da Costa was, for many years, a professor at the Polytechnic School which became the Faculty of Sciences, University of Lisbon. It was perhaps for this reason that the Arruda bones from 1864 were stored in the Bocage Museum, which was attached to the Faculty of Sciences. We were told in 1985 that a 1978 fire had destroyed all the Arruda bones (Jackes and Meiklejohn 2004: 95), and in 2006 were informed that there

Cabeço da Arruda in the 1860s

53

were no remnants among the museum collections. Fortunately this was not true. Casts now located in the Geological Museum, Lisbon, can be compared with material in the Bocage. Neither casts nor bones are labelled, but we suggest that they are the remaining evidence from the 1864 Arruda excavations, citing: x a rough note with the Bocage collection saying "probably Muge"; x the distinctive breccia on the Bocage specimens with charcoal fragments; x accompanying samples of Cardium and Scrobicularia; x a large skull with a depressed nasal region identical to a cast in the Geological Museum. This skull (Fig. 4.5) included five adherent cervical vertebrae, confirming that it had slid forward and down under the pressure of the heavy deposits above it, typical for Muge burials (Jackes et al. 2013).

Fig. 4.5. Arruda skull 1 from the 1864 excavation conserved in the Bocage Museum, Lisbon.

It seems likely that this skull was mentioned (Pereira da Costa 1865: 14, 15) for its “fracture in the form of a cross” which is remarkable but not easily attributed to post-mortem compression. There are cut marks on the skull from cleaning, but certain features of breakage, cracking and marking, especially on the frontal, merit further examination which we were not able to undertake. Compression, excavation trauma and cleaning may not explain everything. The fracture was said to be on the right

54

Chapter Four

parietal, but no impressive cross fracture can be seen on the right parietal, whereas the Skull 2 image (Fig. 4.4) displays a definite cross-shaped fracture on the left parietal. The frontal of Skull 1 was said to be broken along the suture line, with the temporals slightly spread outwards. The skull was crushed onto the face with the nasal root pushed in and the cervical vertebrae were within the mandible. The separated temporals, detached frontal, compressed nasion and the placement of the cervicals, as described, all characterize the skull shown in Figure 4.5. This appears to be part of the fully described “Skeleton 1”, and there is at present a fragmentary paper label apparently marked “1” on the frontal of the skull, though not of the cast. In 1930, Vallois wrote that the Zoological Museum (the Bocage) had material characterized as "debris" which was labelled "Arruda", although this did not include the skull that da Costa had thought female and numbered “2”. In this "debris" Vallois included three damaged and deformed skulls. A fourth, a male, permitted some study, althoufh still partly in matrix and also deformed. His description matches the skull shown in Figure 4.5. We can therefore expect three additional Arruda skulls. Three groupings of further material retained in the Bocage are as follows. 1. A box with a scribbled note, "probably Muge", contains cranial fragments, including a robust left temporal. There is also a damaged proximal left femur and left mandibular ramus with M2 and M3. The breakages on both femur and mandible appear recent and could relate to being salvaged from the 1978 fire. The mandibular molars have typical wear for Muge Mesolithic dentitions, the M3 worn flat with the mesiolingual cusp worn into the dentine. M2 has all cusps removed with coalescent dentine exposure except for limited mesial and central retention of enamel. The disto-buccal cusp probably sustained trauma and there is alveolar resorption lingually below M3. While the mandible labelled 5 in Figure 4.3 could represent this specimen before breakage, this cannot be verified. 2. A further skull in another box is clearly different from the lost Arruda Skull 2 in Figure 4.4. Comparison with the Geological Museum cast makes the difference clear. While this Bocage skull has relatively gracile parietals, occipitals and temporals, the breakage pattern is quite different, and it lacks the frontal. 3. The final skull is that of a child. It appears to have lain on its right side and there is an indication that the skull was placed on a hand. The left side of the skull has been crushed against the right. The right frontal bears holes, one partially cleared of matrix, meriting further attention: they are

Cabeço da Arruda in the 1860s

55

reminiscent of pick marks, which can become filled with matrix if bone becomes wet during excavation (Roksandic and Jackes 2013), or during cleaning or soaking with fire hoses. However, the exposed margin of one does not suggest excavation trauma. The matrix is soft compared with that of the compressed male skull, and it is full of shell and charcoal. Charcoal lodged in the breccia within the skull was identified as Juniperus sp. by Monica Ruiz Alonso (Instituto de Historia, Madrid) and dated at 7060±40 bp (Beta 271927). This is close to the date for an Arruda canid (calculated without reservoir effect offset) excavated in 1880 from a similar depth. When calibrated, our date is 7960–7830 calBP (2ı), according well with other information on Arruda. While junipers may live for hundreds of years (Ward 1973: 918), there is no reason to think that a charcoal fragment, presumably associated with a hearth, would represent an extremely large old trunk. A date of around 7900 cal BP is consonant with a new date for Arruda 6, a burial excavated in the 1930s from close to the basal sands (Jackes et al., 2014).

Conclusion The Bocage Museum in Lisbon retains materials excavated from Cabeço da Arruda in 1864, based on the date of associated charcoal, nineteenth and twentieth century literature, and casts made in the 1860s in Lisbon. There is no evidence that the famous Skull 2 (variously identified as male or female) has been examined since the 1860s. It was certainly lost by the 1920s.

Acknowledgements Dr. Pablo Arias, Universidade de Cantabria, arranged to have the charcoal sample identified through the COASTTRAN project. We are grateful to Dr. M. Magalhães Ramalho, Coordenador, Museu Geológico, Laboratório Nacional de Energia e Geologia for permission to examine Muge materials. MJ acknowledges the bequest made to her by Esther Palmer, which funded the new charcoal date.

References Abrunhosa, A., 2012. As escavações arqueológicas nos concheiros mesolíticos de Muge: contributo para o estudo da historiografia das investigações da década de 1930. Master’s thesis, Faculdade de Letras Universidade do Porto.

56

Chapter Four

Anon., 1899. Les Services Géologiques du Portugal de 1857 à 1899. Comun. Serv. Geol. Port. 4(I), vii–xxi. Antunes, M.T., 1986. Sobre a história da paleontologia em Portugal. Mem. Acad. Ciênc. Lisb. História e desenvolvimento da ciência em Portugal II, 773–814. Busk, G., 1869. Notice Respecting Human Remains Discovered in the Carbeco da Arruda, by F. A. Pereira da Costa ; and in the Grutas da Cesareda, by J. F. N. Delgado. Trans. Ethnol Soc. Lond. 7, 39–44. Cardoso J.L., Rolão J.M., 1999/2000. Prospeçcões e escavações nos concheiros mesolíticos de Muge e de Magos (Salvaterra de Magos): contribuição para a história dos trabalhos arqueológicos efectuados, Estud. Arqueol. Oeiras 8. Cartailhac, E., 1886. Les Ages Préhistoriques de L’Espagne et Portugal. Reinwald Librarie, Paris. Coon, C.S., 1939. The Races of Europe. Macmillan, NewYork. Delgado, J. F. N., 1867. Da Existencia Do Homem No Nosso Solo Em Tempos Mui Remotos Provada Pelo Estudo Das Cavernas: Primeiro Opusculo. Noticia Ácerca Das Grutas Da Cesareda. Typographia Academia Real das Sciencas, Lisboa. Gravlee, C.C., Bernard, H.R., & Leonard, W.R., 2003a. Heredity, environment, and cranial form: a reanalysis of Boas's immigrant data. Am. Anthropol. 105, 125–138. Gravlee, C.C., Bernard, H.R., & Leonard, W.R., 2003b. Boas's "Changes in Bodily Form": The Immigrant Study, Cranial Plasticity, and Boas's Physical Anthropology. Am Anthropol. 105, 326–332. Hervé, G., 1899. Populations mésolithiques et néolithiques de L'Espagne et du Portugal. Rev. Ecole d'Anthrop. 9, 265–280. Jackes, M.K., & Lubell, D. in press. Muge Mesolithic burials, a synthesis on mortuary archaeology. In: J. Grünberg (Ed.) Mesolithic Burials – Rites, symbols and social organisation of early postglacial communities. Tag. Landesmus. Vorgesch. Halle/Congr. State Mus. Prehist. Halle. Jackes, M., & Alvim, P., 2006. Reconstructing Moita do Sebastião, the first step In: Bicho, N.F., Veríssimo, H. (Eds.), Do Epipaleolítico ao Calcolítico na Península Ibérica: Actas do IV Congresso de Arqueologia Peninsular, Faro, Universidade do Algarve, pp. 13–25. Jackes, M., Alvim, P., Anacleto, J.A., & Roksandic, M. 2013. New photographic evidence on the 1954 excavations at Moita do Sebastião. In: Roksandic, M., Mendonça de Souza, S., Eggers, S., Burchell, M., Klokler, D. (Eds.) The Cultural Dynamics of Shell Matrix Sites:

Cabeço da Arruda in the 1860s

57

Diverse Bioarchaeological Perspectives. University of New Mexico Press, Albuquerque. Jackes, M.K. & Cunha, M.J., 2015. A new resource on Muge archaeological sites: a 1932 film. Mesolith. Misc.. Jackes, M., Lubell, D., Alvim, P., & Cunha, M.J., 2015. Sources for the reconstruction of Cabeço da Arruda. This volume. Jackes, M.K., Lubell D., & Cunha M.J., 2014. Redating a Mesolithic skeleton from Cabeço da Arruda, Muge, Portugal. Mesolith. Misc. 22(2):40-44. Jackes, M., & Meiklejohn, C., 2004. Building a method for the study of the Mesolithic-Neolithic transition in Portugal. In: Budja, M. (Ed.), Neolithic Studies 11, Doc. Praehist. 31, 89–111. Paula e Oliveira M. F., 1884. Notes sur les ossement humains qui se trouvent dans le Musée de la section Géologique de Lisbonne. Congrès international d'anthropologie et d'archéologie préhistoriques. Compte rendu de la neuvième session a Lisbonne 1880. Typographie de l’academie royale des sciences, Lisbonne, pp. 291–305. Pereira da Costa, F.A., 1865. Da existencia do homem em epochas remotas no valle do Tejo. Primeiro opusculo. Noticia sobre os esqueletos humanos descobertos no Cabeço da Arruda. Imprensa Nacional, Lisbon. Pruner-Bey, F., 1868. Objets prehistoriques de Portugal. Congrès international d'anthropologie et d'archéologie préhistoriques. Compte rendu de la deuxième session, Paris 1867. Paris, C. Reinwald, pp. 33– 35. de Quatrefages, A., Hamy, E.-T., 1882. Crania Ethnica. J.B. Baillière et fils, Paris. Ribeiro, M. C., 1867. Note sur le terrain quaternaire du Portugal. Bull. Soc. Geol. Fr. Séance du 17 Juin 1867, Series 2, 48, 692–717. Ribeiro, C. 1884. Les Kioekkenmoeddings de la Vallée du Tage. Congrès international d'anthropologie et d'archéologie préhistoriques. Compte rendu de la neuvième session a Lisbonne 1880. Typographie de l’Academie Royale des Sciences, Lisbonne, pp. 279-290. Roksandic, M. & Jackes, M. 2013. The skeletal assemblage and burial ritual at the site of Cabeço da Amoreira: the 1960s excavations by Veiga Fereira and Roche In: Roksandic, M., Mendonça de Souza, S., Eggers, S., Burchell, M., Klokler, D. (Eds.) The Cultural Dynamics of Shell Matrix Sites: Diverse Bioarchaeological Perspectives, University of New Mexico Press, Albuquerque. Stewart, T.D., 1936. Anthropometry. Anthropometric nomenclature. I. The cephalic (length-breadth) index. Am. J. Phys. Anthropol. 22, 97–140.

58

Chapter Four

Vallois, H.V., 1930. Recherches sur les ossements Mésolithiques de Mugem. L’Anthropol. 40, 337–389. Ward, L. K., 1982. The conservation of Juniper: longevity and old age. J. Appl. Ecol. 19, 917–928.

CHAPTER FIVE ANTLER DEBITAGE IN MUGE SHELL MIDDENS: THE COLLECTIONS OF THE GEOLOGICAL MUSEUM MARINA ALMEIDA ÉVORA ICArEHB Universidade do Algarve [email protected]

Abstract As one of the most important Mesolithic sites in Europe the Muge shell middens have been known for 150 years. Much information has been published about the hundreds of human burials, but not much about the lithic and faunal assemblages. Although the osseous materials present in the Muge shell middens are a very important part of the material culture, their technological analysis is yet to be done. We present here the results of the technological analysis of a sample of antler artefacts that were recovered from Moita do Sebastião and Cabeço da Arruda shell middens in the late 19th and early 20th centuries, and deposited in the Geological Museum in Lisbon. We have registered two fracturing techniques and three wear techniques employed in the manufacture of this osseous industry and four categories of tools made by the Mesolithic hunter-gatherers.

Introduction As one of the most important Mesolithic sites in Europe the Muge shell middens have been known for 150 years. The survey in Tejo Valley began in 1863 with Carlos Ribeiro. In Muge, he found a large quantity of human remains, shells, animal remains, charcoal and quartzite in the Cabeço da

60

Chapter Five

Arruda shell midden. In the following year, he and Nery Delgado found Moita do Sebastião and Fonte do Padre Pedro shell middens (Cardoso and Rolão 1999/2000; Detry 2007). In 1880, Carlos Ribeiro carried out more excavations in the Moita do Sebastião and Cabeço da Arruda shell middens from which they recovered dozens of human skeletons and other artefacts and it was all then published in the Proceedings of the IX Session of the International Congress of Anthropology and Prehistoric Archaeology in 1884. After his death, the work in the Muge shell middens continued under the supervision of Paula e Oliveira who took great interest in studying the human remains. He then classified the archaeological site as prior to the Neolithic but still dating to the Quaternary (Cardoso and Rolão 1999/2000; Detry 2007). In 1888, Paula e Oliveira died and for many years the research into the Muge Shell middens stopped until 1930 when Mendes Correa decided to start the excavations again (Cardoso and Rolão 1999/2000). In the 1950s the site of Moita do Sebastião started to be destroyed and only a base layer of the shell midden remains to the present day. Jean Roche published several articles on the Muge shell middens until 1980 (Cardoso and Rolão 1999/2000; Detry 2007). Later, in 2001 J. Rolão began the work of cleaning, recovery and conservation in Cabeço da Arruda (Cardoso and Rolão 1999/2000; Detry 2007). The human remains are the most studied in the Muge shell middens, but the archaeological sites also provided several other artefacts like lithics, adornments, faunal remains and osseous industry. These analyses of artefacts were briefly detailed, although they were published in 1865 by Pereira da Costa, in 1884 by Carlos Ribeiro and in 1888 by Paula e Oliveira (Bicho et al. 2010a) and in 1947 Breuil and Zbyszewski published the first review of the Mesolithic industries from Muge also paying attention to the antler and mammal bone artefacts from Cabeço da Arruda and Moita do Sebastião (Breuil and Zbyszewski 1947). In 1986 Lentacker analysed the faunal remains and in 2007 Detry made an extensive study of the faunal remains from the Muge shell middens, focusing on biometrics, taphonomy, paleoecology and paleoeconomics. Along with beads, lithics and faunal remains, there is a large quantity of osseous materials that were intentionally modified. The recent excavations in the Cabeço da Amoreira shell midden began in 2008 under the research project “Last Hunter-Gatherers of the Tagus Valley–The Muge Shell Middens” directed by Nuno Bicho from the University of Algarve. Presently, the radiocarbon data indicates that the shell middens first appeared by 8200 cal BP in Cabeço da Arruda and their use by

Antler Debitage in Muge Shell Middens

61

Mesolithic hunter-gatherer-fishers ended by 7400 cal BP when a Neolithic population entered the Tagus Valley (Bicho et al. 2010b; 2013). Here, by paying special attention to bone surface modifications made during its manufacture and use, we present some preliminary results of the technological analysis of 100 Red deer (Cervus elaphus) antler artefacts that were recovered from the Moita do Sebastião and Cabeço da Arruda shell middens during the late 19th/beginning of the 20th centuries and deposited in the Geological Museum in Lisbon.

Materials and Methodology We have studied a collection of antler artefacts that belongs to old excavations. The sample is biased by the methods of excavation and by later preservation in the museum. The research objectives at that time were different from those of today–the main goal was to recover and study the human remains. Also the methods of excavation were different and the osseous industry has no indication of its stratigraphical provenience and the artefacts only have an indication of the archaeological site provenience (Bicho et al. 2013; Detry, 2007). Antler tools were analysed using macro- and micro-wear approaches with the use of a magnification lens (10x-40x) and not just macroscopically (Blumenschine et al. 1996; Dominguéz-Rodrigo et al. 2009; Évora 2008), otherwise some important features would be missed. We focused our observations on modifications made during production and use activities and also on post-depositional processes. The bone surface analysis methodology was done based on the previous works of several researchers, namely Averbouh (2000), Bertrand (1999), David (2005), d’Errico et al. (1984), d’Errico and Giacobini (1985), d’Errico and Giacobini (1986), d’Errico and Espinet-Moucatel (1986), d’Errico (1993), Goutas (2005), Lyman (1994), Maigrot (1997) and Semenov (1985). The sample is composed of 100 artefacts from Cabeço da Arruda (n=50) and Moita do Sebastião (n=50), all made from Cervus elaphus antler.

Results The antler artefacts are present in four categories of products: (1) finished objects; (2) blanks; (3) secondary blocks; and (4) debris. The number of finished objects (n=50) is far greater than the other categories, followed by the debris (n=36) (Fig. 5.1).

62

Chapter Five

Fig. 5.1. Cabeço da Arrudda and Moita do o Sebastião antller artefacts categories.

Finished Objects O This category is com mposed of sev veral types off tools (Table 5.1) and they were exxtracted from all parts of th he antler, but m mainly from th he tines.

Bevelled d tools =15) (Fig. One of tthe best repreesented tools is unifacially bevelled (n= 5.2). The unnifacially beveelled tools varry in total lenggth from 263 mm m to 37 mm and theeir cortical tisssue thicknesss varies from 87 mm to 3 mm. The bevel total llength varies from 60 mm to 9 mm andd its width vaaries from 375 mm to 5 mm. The bifaacially bevelled tool is alw ways represennted in the sample s in conjunction with a handlee or a wedge on the opposiite end of the tool. The tools vary inn total length from f 114 mm m to 101 mm, aand their cortiical tissue thickness rannges from 52 mm to 47 mm m. The secondd bevel is mucch shorter than the othher, ranging from 248 mm to t 10 mm in ttotal length. They T were made from tines and poortions of beaams. One peddicle (Fig. 5..3) raises doubts of itss being debriss, as a portion n of the beam is still attach hed and at its distal endd the spongy tissue t has an active a area, veery smoothed with fine striations.

palmation base beam B beam A and B beam A indeterminate beam middle tine bez tine eye tine indeterminate tine pedicle burr indeterminate total

Anatomical provenience

4

3

15

1

Uni/Bifacial

2 1

2

1

Uni/Bifacial

1 2 8

1 2 1

Unifacial

3

4

2 15

1

Pointed

4 4

3

1

1

Wedge

Table 5.1. Tools and Debris quantities and their anatomical provenance.

Antler Debitage in Muge Shell Middens

4

3

1

Hammer

2

1

1

Handle /

2 2 1 2 12 11 2 1 36

Debris 1 1 1

63

64

Chapter Five

Could it have been used as a digger tool or in processing hides?. There is a similar artefact from the Polderweg Mesolithic site in the Netherlands that according to Van Gijn was used as a hide scraper (Van Gijn 2007).

Fig. 5.2. Bevelled antler tools from Cabeço da Arruda and Moita do Sebastião shellmiddens.

Antler Debitage in Muge Shell Middens

65

Fig. 5.3. Pedicle with distal bevel.

Wedges Another type of finished object is the wedge. Wedges are mainly made from tines, and some could be determined as from the eye tine and from the middle tine. The total length varies from 189 mm to 49 mm, with a cortical tissue thickness ranging from 63 mm to 35 mm showing that some wedges are very robust. All tools used as wedges have small impact marks on their active part. One of these tools has impact marks and small withdrawals on the distal end and also a small lateral perforation on the mesial area (Fig. 5.4).

66

Chapter Five

Fig. 5.4. Wedge with negative withdrawals on the proximal end.

Pointed Tools Pointed tools are represented in small quantities (n=4) and are made from tines. Their total length varies from 201 mm to 77 mm. These tools have on their distal end an active area with fine striations that have a random or circular orientation and are quite localized around the distal end of the tool.

Hammers The hammers (n=4) are made from eye tine and beam. Their total lengths vary from 184 mm to 103 mm. They all have impact marks on the proximal end surface.

Antler Debitage in Muge Shell Middens

67

Handles/Axes Handles (n=2) are made from tine and beam. They have a total length of 126 mm and 85 mm. The one made from a tine had the spongy tissue extracted and a saw-type fracture. There are another two handles, but these are in conjunction with a uni/bifacially bevelled tool. The handle of one of these tools has the spongy tissue removed to insert another tool, and in that way, its bevel could have been used as an axe (Fig. 5.5). Their cortical tissue thicknesses are 63 mm and 56 mm, so they are robust tools.

Blanks We have registered only two blanks in this sample. One came from the beam, measuring 93 mm in total length and 46 mm in its cortical tissue thickness. Another blank came from a tine, has a total length of 198 mm and a cortical tissue thickness of 54 mm.

Debris Debitage debris is the second most represented artefact category in this sample (n=36). The majority of the debris consists of pedicles and tines. The pedicles came from hunted Red deer antler (n=7) and from shed antler (n=5). Most of the pedicles from hunted animals are still attached to the cranial bone on one end and have the burr and outer burr on the other end (Fig. 6). The tines are mostly indeterminate and the debris, in a general way, originated from several parts of the antler as seen on Table 5.1.

Secondary Blocks This artefact category is represented by 6 objects, mainly extracted from the beam, bez tine and/or middle tine. Their total lengths range from 126 mm to 90 mm and their cortical tissue thicknesses vary from 6 mm to 5 mm.

68

Chapter Five

Fig. 5.5. Handle/Axe: with a circular perforation on the proximal end and saw type fracture on the distal end.

Fig. 5.6. Debris: Pedicle and Burr

Antler Debitage in Muge Shell Middens

69

Fracturing and Wear Techniques In the antler debitage we registered two breaking techniques (1) direct percussion; and (2) breaking by bending; and three techniques (1) sawing; (2) double longitudinal grooving; and (3) scraping.

Fracturing Techniques Those two techniques were used, mainly together, to separate the pedicles, the beams, the middle and eye tines and the burr. The stigmata can be seen in the artefacts and debris that present impact marks left by direct percussion with a cutting edge tool (Fig. 5.7) and the negative of the bending in their bone surfaces. The fracture planes are irregular.

Wear Techniques We have observed three techniques. The sawing (Fig. 5.8) technique was applied transversely to the longitudinal axis of the object, and in some artefacts sawing was used in conjunction with direct percussion and bending. It was used mostly to separate tines. The fracture plane varies according to the methods used. Double longitudinal grooving was registered in only 6 artefacts, all made from beams (Fig. 5.9), leaving a regular fracture plane. The scraping technique was applied on the manufacture of the bevelled tools and it is limited to its active part and does not extend to the rest of the bone surface.

Discussion Antler artefacts are present in small and medium knapped fragments and also as entire tools. All the antler parts are represented in this sample, indicating that hunter-gatherers took advantage of all antler morphology to manufacture their bone toolkit. The resources of Red deer antler came from shed antler and also from hunted animals, as attested by several pedicles and burrs.

70

Chapter Five

Fig. 5.7. Marks of direct percussion with a cutting edge tool on a tine.

Fig. 5.8. Sawing marks all around the proximal end of a tine and negative of bending.

Antler Debitage in Muge Shell Middens

71

Fig. 5.9.Antler beam: the arrows indicate the double longitudinal grooving.

These tools on osseous material represent several completed artefacts (n=47). Despite that, some tools are incomplete and represented as mesial or proximal parts. Distal fragments are less represented (n=8). Manufacture and use traces are common. The preferred fracturing techniques were removal by direct percussion with a cutting edge tool and breaking by bending. The wear techniques used were sawing, double longitudinal grooving and scraping. All these techniques left visible stigmata on the bone surface such as impact marks made from a cutting edge tool, negatives of bending, and the negative of longitudinal grooving. Some tools were used as wedges as their distal active area has small impact marks and on the proximal end there are small negative withdrawals of the bone surface. Together with wedges, the bevelled tool is the most representative in Muge shell middens, some carrying long and large bevels and the great majority with stigmata resulting from impacts on the distal end, characterized by saw type fractures and short withdrawals. Others have a smooth bevel area with fine striations that could have resulted from scraping or from their use. The functional aspects of the bevelled tools are controversial as noted by Griffitts and Bonsall (2001). The authors refer that most bevelled tools are thought to have been used for working wood, processing hides or grinding tools. After an experimental study they concluded that this type of tool could have been used to collect and process limpets that are the dominant components of

72

Chapter Five

the shell middens in Western Scotland (Griffitts and Bonsall 2001). In the Muge shell middens the dominant component shells are from estuarine areas: cockles (Cerastoderma edule) and clams (Scrobicularia plana) (Detry 2007). The length of the tools are also different: the Muge bevelled tools are much longer and more robust, made from the beam and from tines and have a thick cortical tissue on average, as opposed to the bevelled tools from Western Scotland shell middens, that are made from antler and mammal bone splinters. So, we can infer different tasks for the Muge bevelled tools. These artefacts may suggest intensive woodwork, due to the number and the large size of some wedges (axes?) and handles since it could reflect an exploitation of forest resources, common during the Tardiglacial (Childe 1944; David 2005). Muge bevelled tools share some characteristics with bevelled tools from the Mesolithic sites of Dudka 1, Pobiel 10 in Poland and Hohen Viecheln 1 and Friesack 4 in Germany, with the same fracturing techniques (percussion and double longitudinal grooving, although not the prevalent ones there), wear techniques (sawing), and also the antler portions used as hammers, handles, bevelled tools were made from tines (Pratsch 2011). Some other parallels are seen with the Mesolithic osseous industry of Mullerup 1 and Lundby II in Denmark (David 2005). These Danish sites have bevelled tools similar to some bevelled tools from Muge shell middens, mainly adzes with longitudinal hafting, although the Danish tool is made from an auroch’s metatarsals and the Muge tool is made from antler tine, but the hafting mode is the same and morphologically they are similar. Also similar are the hammers made from antler in both Portuguese and Danish sites. Muge bevelled tools may also suggest the treatment and processing of hides, since in the faunal analysis from the Muge shell middens, bones like the astragalus, tibia, radius and calcaneum are present in high frequencies and this could indicate the use of hides and tendons from hunted animals like Red deer, wild boar, rabbit and hare which are all present in the faunal assemblage from Muge (Detry 2007).

Conclusions We have presented here a preliminary analysis of a sample of osseous industry made from Red deer antler. The fact that these artefacts have no stratigraphic provenience prevents us from registering the evolution of the techniques in the manufacture of this equipment along with the human occupation of the Muge shell middens, as opposed to other studies of Mesolithic bone industries. On the other hand, its value is that it has permitted a reconstruction of the chaine operatoire used for the production

Antler Debitage in Muge Shell Middens

73

of antler tools. Artefacts show abundant features that ethnographically and experimentally were recurrently associated with direct percussion and breaking by bending. Also, it registers the importance of this industry in the daily life of Mesolithic hunter-gatherer-fishers, by showing the role of this resource in the manufacture of their toolkits, surely made on the site (attested by the high frequency of debitage debris) and the importance to their economy, since Red deer was the most hunted mammal in Muge (Detry 2007). More research is needed in this area, since there are also tools made on mammal bones that were superficially presented in articles published in the last century. Comparison with other Mesolithic sites with osseous industry in Portugal is, to our knowledge, not yet possible since that is not published.

Acknowledgements We would like to thank Prof. Miguel Ramalho for the permission to study the antler collection deposited in the Geological Museum in Lisbon and to José António Anacleto for his assistance in the museum. I also thank to C. Detry for her comments on the manuscript.

References Averbouh, A., 2000. Technologie de la matière osseuse travaillée et implications palethnologiques. L’exemple des chaines d’exploitation du bois de Cervidé chez les Magdaléniens des Pyrénées, Thèse de doctorat, Université de Paris I, 2 volumes Bertrand, A. 1999. Les Armatures de sagaies magdaléniennes en matière dure animale dans les Pyrénées, BAR International Series 773 Bicho, N. F., Pereira, T., Cascalheira, J., Marreiros, J., Pereira, V., Jesus,L. & Gonçalves, C. 2010a. Cabeço da Amoreira, Muge: resultados dos trabalhos de 2008 e 2009. Proceedings of the Workshop The last hunter-gatherers and the first farming communities in the South of the Iberian Península and North of Marocco. (2009) Universidade do Algarve, Faro, pp.11-17. Bicho, N., Umbelino, C., Detry, C. & Pereira, T. 2010b. The Emergence of Muge Mesolithic Shell Middens in Central Portugal and the 8200 cal yr BP Cold Event. Journal of Island & Coastal Archaeology, 5, pp.86–104. Bicho, N., Cascalheira, J., Marreiros, J., Gonçalves, C., Pereira, T.& Dias, R. 2013. Chronology of the Mesolithic occupation of the Muge valley,

74

Chapter Five

central Portugal: The case of Cabeço da Amoreira. Quaternary International 308-309, pp.130-139. Blumenschine, R. J., Marean, C. W. & Capaldo, S. D., 1996. Blind Tests of Inter-analyst Correspondence and Accuracy in the Identification of Cut Marks, Percussion Marks, and Carnivore Tooth Marks on Bone Surfaces, Journal of Archaeological Science 23, pp. 493–507. Breuil, H. & Zbyszewski, G. 1947. Révision des industries mésolithiques de Muge et de Magos (Collections du Service Géologique du Portugal). Comunicações dos Serviços Geológicos de Portugal, XXVIII:3-66. Cardoso, J. & Rolão, J., 1999/2000. Prospecções e escavações nos concheiros mesolíticos de Muge e Magos (Salvaterra de Magos): contribuição para a história dostrabalhos arqueológicos efectuados. Estudos Arqueológicos de Oeiras 8, pp. 83-240. Childe, V. G., 1944. Archaeological ages as technological stages. Journal of the Royal Anthropological Institute of Great Britain and Ireland 74, pp. 1-19. David, E. 2005. Technologie osseuse des derniers chasseurs préhistoriques en Europe du Nord (Xe-VIIIe millénaires avant J.-C.) Le Maglemosien et les technocomplexes du Mésolithique. Thèse de doctorat, Maison Archéologie Ethnologie, Nanterre, 667p. D’Errico, F. 1993. Identification des traces de manipulation, suspension, polissage sur l’art mobilier en os, bois de cervidés, ivoire, Traces et Fonctions: les gestes retrouvés, Colloque international de Liège, Editions ERAUL, vol 50, pp. 177-188. D’Errico, F. & Espinet-Moucadel, J., 1986. L’emploi du microscope électronique à balayage pour l’étude expérimentale de traces d’usure: raclage sur bois de cervidé, Bulletin de la Société Préhistorique Française, Tome 83, nº 3, pp. 91-96. D’Errico, F. & Giacobini, G. 1985. Approche méthodologique de l’analyse de l’outillage osseux, un exemple d’étude, L’Anthropologie, Tome 89, nº 4, pp. 457-472, Paris. D’Errico, F. & Giacobini, G. 1986. L’emploi des répliques en vernis pour l’étude de surface des pseudo-instruments en os, Artefacts 2, pp. 57-68. D’Errico, F., Giacobini, G. & Puech, P.-F., 1984. Varnish Replicas: A New Method for the Study of Worked Bone Surfaces, OSSA, vol 9-11, pp. 29-51. Detry, C. 2007. Paleoecologia e Paleoeconomia do Baixo Tejo no Mesolítico Final: O contributo do estudo dos mamíferos dos concheiros de Muge. Unpublished Ph.D. Thesis, University of Salamanca, 445p.

Antler Debitage in Muge Shell Middens

75

Domíngos-Rodrigo, M., Juana, S., Galán, A.B., & Rodrígues, M., 2009. A new protocol to differentiate trampling marks from butchery cut marks, Journal of Archaeological Science 36, pp. 2643–2654. Évora, M. A., 2008. Artefactos em haste e em osso do Paleolítico Superior Português, Promontória, vol 6, pp. 9-50. Griffitts, J., & Bonsall, C. 2001. Experimental Determination of the Function of Antler and Bone ‘Bevel-ended Tools’ from Prehistoric Shell Middens in Western Scotland, in: Choike, A. M and Bartosiewicz, L. (eds) Crafting Bone: Skeletal Technologies through Time and Space. BAR International Series 937, pp.207-220. Goutas, N. 2005. Caractérisation et Evolution du Gravettien en France par l’approche techno-économique des industries en matières dures animales (étude de six gisements du Sud-Ouest), Thèse de doctorat, Université de Paris I – Pantheon – Sorbonne, 2 volumes Lyman, R. L. 1994. Vertebrate Taphonomy, Cambridge manuals in Archaeology, Cambridge University Press Maigrot, Y., 1997. Tracéologie des outils tranchants en os des V et IV millénaires av. J.-C. en Bassin parisien : essai méthodologique et application, Bulletin de la Société préhistorique française, 94, 2, pp. 198-216. Pratsch, S., 2011. Mesolithic antler artefacts in the North European Plain, in: Baron, J. and Kufel-Diakowska, B. (eds) Written in Bones Studies on technological and social contexts of past faunal skeletal remains. Institute of Archaeology, University of Wroclaw, pp.79-92. Semenov, S. A., 1985. Prehistoric Technology. An experimental study of the oldest tools and artefacts from traces of manufactures and wear, Barnes & Noble Books, Totowa, New Jersey Van Gijn, A. 2007. The use of Bone and Antler Tools: Two Examples from the Late Mesolithic in Duch Coastal Zone, in: St-Gates, P. C. and Walker, R.B. (eds) Bones as Tools: Current Methods and Interpretations in Worked Bone Studies. BAR International Series 1622, pp. 81-92.

CHAPTER SIX READING THE LITHICS IN FLINT OF CABEÇO DOS MORROS SHELL MIDDEN ANABELA JOAQUINITO Portuguese Association for Archaeological Investigation (APIA) Portugal [email protected]

Abstract The Cabeço of Morros shell midden, located on the bank of the Magos River, was identified in 1863 by Carlos Ribeiro. Archaeological excavations were carried out between 1997 and 2001 under the direction of José Manuel Rolão, and raised a collection of over 1500 lithic materials, 85% of which are flint. The study of the lithic industry gives the impression that it was a seasonal occupation and possibly done systematically, with a similar set of goals and activities maintained throughout its existence. The main strategies for flint were to produce bladelets and microliths tools, with the cores abandoned at an exhaustion stage.

Introduction The Morros shell midden is located in the Ribatejo region in the municipality of Salvaterra de Magos and belongs to a set of 11 shell middens discovered in the district of Santarém in the 19th and 20th centuries (Fig. 6.1).The Morros shell midden is defined as a "support" location, receiving this designation as it had been temporarily occupied with specific objectives. Standing at the low Quaternary terraces, with an elevation of 2/3 m along the Magos, its area is about 200 m2 and its height is about 1.3/1.5 m. Its occupation was seasonal, as its location did not

78

Chapter Six

allow occupation during the autumn/winter, with the rise of the water level (Rolão 1999).

Fig. 6.1. Overview of Cabeço dos Morros shell midden.

The objectives of this temporary change to this site were: a) the gathering of shellfish, which included scrobicularia plana, cardium edule, ostrea and solen marginatus; b) the use of natural conditions for the practice of fishing activities, such as the fishing species myliobatis aquila, sparidae and sparus auratus; and c) a possible place for the start of the journey to fetch flint which headed to the right bank of the Tagus in the municipality of Alenquer, where nodules of flint were present; the raw material chosen for tool production.

The Rediscovery of the Cabeço Morros Shell midden In 1935 Hipólito Cabaço sought a shell midden in Quinta da Sardinha: it had been discovered by Carlos Ribeiro and had since disappeared. Hipólito Cabaço did not find the shell midden Quinta da Sardinha, now believed to be the Cova da Onça shell midden, but rediscovered five shell middens, including the Cabeço dos Morros shell midden, on the left bank of the Magos stream. The shell midden is currently the only one on the left

Reading the Lithics in Flint of Cabeço dos Morros Shell Midden

79

bank of the Magos stream; the other two identified, Magos de Baixo and Cabeço da Barragem, were destroyed (Rolão and Cardoso 2002/2003). In an article published in 1936, Afonso do Paço refers to a survey conducted by Hipólito Cabaço after the discovery of the site and the removal of the archaeological material: "But in Cabeço dos Morros were found many shells of shellfish, animal bones, a human skull and some bones, diverse lithic material, like knives, percussors, awls, scrapers, many flakes of flint and quartzite, and some trapezoidal microliths, and the industries and the skull of Muge type" (Afonso do Paço 1970). In the 1980s, the surface layer of the shell midden was razed about 60 cm, so the original stratigraphy was probably about 1.6 m. Between 1997 and 2000, the team of the Centre for Archaeological Studies of Autónoma de Lisboa University directed by José Manuel Rolão initiated a project of archaeological work in Cabeço dos Morros (Rolão et al. 2006). The shell midden features 4 layers of occupation in which three structures stood out: a paleosol of habitat and combustion structures previously identified in the shell middens of Moita do Sebastião, Cabeço da Arruda and Cabeço da Amoreira. The structures are associated with the higher occupancy archaeological level (layer 2) which provides about 60% of the total of the lithic industry.

Stratigraphy About 50/60 cm of the top layer of this shell midden were destroyed by earthworks. The following description of the four layers of occupation only refers to 120 cm of the stratigraphy where archaeological excavation was carried out. Layer 1 corresponds to the superficial layer, the rummaged disorder that was suffered when the earthworks provoked the intrusion of materials between layers. The occupation layer is 10 to 30 cm deep, and presents variations in the sections, as a result of the situation referred to above, and is composed of a layer of brown-grey sediment, with a level of an average density of pebbles. It began to emerge that these layers had a high concentration of the remains of malacological fauna and lithic artefacts. Layer 2 features darker sediment and a high concentration of malacological fauna and pebbles, revealing clear traces of occupation. The “bags” of malacological fauna are in greater numbers, many of them from the previous layer, showing some variations as an increase in its diameter and a greater concentration of coal. The paleosol structure of habitat

80

Chapter Six

appears in its entirety in this layer in squares S and T, consisting of a layer of breccia and including a combustion structure. Layer 3 varies between 20 and 30 cm in height. It is composed of light yellow-brownish sands, showing scant traces of occupation constituted only by a reduced number of lithic artefacts and pebbles. Layer 4 was excavated to a depth between 80 and 120 cm and has finegrained clear sand and some pebbles, it is a sterile layer where there is no archaeological material. The nature of the shell midden sediment allowed good preservation and conservation of the original positioning of the structures, giving rise to an accurate picture of the spatial organization that developed in this temporary site.

The Lithic Industry The Cabeço dos Morros shell midden presents an important lithic industry divided between a majority of microlithic tools (bladelets, geometrics, points and microscrapers) and a minority of macrolithic tools, supported by local rocks (quartz and quartzite) and giving rise to retouched tools (massive scrapers and thick flakes) and without retouching (flatter and percussor). The lithic industry is based on exploiting preferential flint (88.7%), sidelined by local rocks, quartzite (10.4%) and quartz (0.9%). Quartz products compose only 17 pieces, including only 3 instruments. The quartz is presented in two types: 82.3% of the pieces were produced in the milky quartz and the remaining 17.7% were in crystal quartz. The core groups are formed by a subprismatic in crystal quartz and a core fragment in milky quartz, with lamellar support production, similar to the flint cores. Only a fragment of proximal bladelet and two thin flakes were produced in crystal. The artefacts include two retouched flakes and a percussor, all in milky quartz.

Debitage Technology Two exploitation strategies for the flint cores were identified in this shell midden. In the first strategy which includes 90% of the cores, the cores are subprismatic, subpyramidal, crossed debitage, informal and flanc de nucléus, with negatives of parallel, opposed and perpendicular bladelets. The second strategy regards the production of flakes, coming from flake cores and formless cores.

Reading the Lithics in Flint of Cabeço dos Morros Shell Midden

81

The exploitation of the cores reaches its limits, with nearly all of the cores at their final stage weighing on average 15 grams, 40% less than 9 grams, and only one core of more than 100 grams. Only 3% correspond to the 1st debitage phase, most coming from small pebbles. The average negative recorded in the cores is situated at 21 mm, being possibly established in parallel with the dimensions of the retouched and unretouched bladelets whose higher percentage values recorded are at 1519 mm and 20-24 mm. The blade/bladelet group is composed of 277 pieces and corresponds to 19.6% of the debitage material; the blades are just 5, which represent 0.47%. The proximal fragments are those that present a higher percentage value, with 40.1% of the total followed by the distal fragment gills with 28.5%. The flint bladelets have the following metric medium parameters; 19.4 mm long, 7.7 mm wide and 2.4 mm thick. Comparing them with the negatives from the core, there is a difference of 7 mm, being the entire slides from an initial phase of core recovery. The fragments of distal and proximal slides reveal identical average measurements both in the length of 15 mm, as in the width of 7.9 mm and a thickness between 2.1 mm and 2.3 mm. The bladelets of larger dimensions, quality and thickness, with trapezoidal sections and parallel edges were removed from the cores in the initial phase, when the cores had a length/width exceeding 35 mm. The bladelets with smaller metric parameters were rarely retouched however the choice of continuing production (3.8% of the debitage products) indicates a single use or a part of a composite tool. The analysis of the attributes of bladelets with and without retouching lets us establish that most of the bladelets have a triangular section (57.9%), parallel edges (63.2%), a curved profile (57.9%) and are without the presence of a cortex (72%). The butts are mostly smooth (36.8%) and faceted (29.8%). The plains of percussion are preferably smooth and faceted with reduced dimensions of the lugs (Table 6.1). There is a clear strategy of standardized production in debitage lamellar supports, both morphological and metric. The flakes were divided into two subgroups, with and without cortex; the first group totals 345 parts and the second 103, representing 31.7% of the total debitage material. The high number of flakes presupposed the existence of a greater number of cores for flakes, but such is not the case, a situation that was probably due to the intensive exploitation of the cores which in the final stages has opted to remove bladelet products (Fig. 6.2).

82

Chapter Six

Table 6.1. Distribution of bladelets by layer according to their attributes. BLADELETS Section Triangular Trapezoidal Edges Parallel Winding Convergent Talão Flat Facetted Jammed Cortical Punctiform Indeterminade Profile Arched Right Córtex Without cortex Parcial cortical

Layer1

Layer2

Layer3

12,30% 7%

36,80% 29,80%

8,80% 5,30%

10,50% 5,30% 3,50%

43,90% 10,50% 12,30%

8,80% 5,30% 0%

7% 3,50% 1,75% 5,30% 0% 1,75%

24,60% 21,10% 7% 5,30% 7% 1,75%

5,30% 3,50% 3,50% 1,75% 0% 0%

12,30% 7%

36,80% 29,80%

8,80% 5,30%

12,30% 7%

49,10% 17,50%

10,50% 3,50%

Microlithic Technology-Tools At the Cabeço dos Morros shell midden 40% of the artefacts are trapezes, while the triangle represents a mere 3%, and the segment is nonexistent. The microlithic geometric group comprises 87.2% of trapezes; 7% of triangles and 5.8% fragments of undefined geometric. The choice of one type of geometrical shape in the shell middens is often at the almost complete expense of any other type of geometric. In the Cabeço dos Morros shell midden there is an absence of geometrics, except for crescents, because all typological groups in the shell middens of the Tagus Valley are present in this site. This suggests that the segments are the last stage of the technological evolution of geometrics, not present in

Reaading the Lithicss in Flint of Cab beço dos Morroos Shell Midden n

83

this site or just emergedd at the last phase p of use which was destroyed d before the arrchaeological excavation.

Fig. 6.2. Perccentage chart off flint debitage material. m

Their tyypology includes short and d long symm metric and asy ymmetric trapezes, annd scalene andd isosceles triangles. With regard to theeir length, the main feaatures of the trapezes t are the measurem ments of 10-14 4 mm and 20-24 mm. The analysis of the dimenssions of the trrapezes shows that the trapezium tyype is the asyymmetrical sh hort of 15-19 mm in length h and 5-8 mm in widthh, with 47.8% % followed by the t asymmetrric long with 38.8%. 3 The trappezes have diiverse truncation associatioons, marks off use and different typpes of breaks and a the main features f are: a) Sm mall concave truncation t and d big oblique truncation: th his is the subgroup w with the highesst percentage. This associaation is the staandard in asymmetric short and long; b) The ggeometrics witth higher dim mensions arise when both tru uncations are oblique or in the asssociation of small concavve truncation n and big oblique trunncation in the asymmetric a lo ong trapeze; c) Whhen at least onne truncation is convex the triangle is sm maller (14 mm): this is an association that is rare on o this site; d) In 995% of the traapezes, the big truncation iis on the rightt, and this is the activve and functional part off the compossite tool and d hunting weapons;

84

Chapter Six

e) The identified breaks in the truncations are of two types: the right truncation is either fractured or broken between 1 to 3 mm from the right end; f) The marks of use are common in the big and small bases, at 73.6% and 48.8% respectively; g) The big base has a partial retouching in 34.5 % of trapezes, generally on the obverse and rarely on the inverse, with short and parallel removals, usually located at the right end and occasionally with a notch in the centre. The possibility that both truncations of the trapeze could be active and one could replace the other in case of need is of course a consideration, but it looks unlikely or at least not common in Cabeço dos Morros. The small left truncation is rarely broken, in only 16% of cases and only 2 trapezes have both truncations broken. This seems to indicate a one-way use of trapeze microliths and the adaptation of these to composite instruments (Fig. 6.3).

Microburin Technique There is an exclusive use of the microburin technique for obtaining geometrics; the quality of flint use in the shell midden facilitates the use of that technique. The microburins represent 11% of the debitage material, corresponding to the second highest percentage in the shell middens of the Tagus valley and a ratio of 1.9 microburin to 1.0 geometric. The distal microburin has the highest number of pieces, i.e. 78, while there are 56 proximal microburins and 14 opposed to fracture. The existence of only 2 double microburins is justified by the reduced size of the bladelets (Fig. 6. 4).

Reaading the Lithicss in Flint of Cab beço dos Morroos Shell Midden n

Fig. 6.3. Perccentage chart off wear marks in trapezes.

Fig. 6.4. Corees, retouched blladelet, microbu urins and trapezze.

85

86

Chapter Six

Remaining Tools The remaining lithic tools support the well-defined metric parameters, according to the economy of the microlithic industry and the lithic resources mentioned above. The flint microscrapers in Portuguese shell middens, with an emphasis on those of the Tagus Valley, are reduced in size, at an average of 20 mm, at an average not exceeding 36 mm long and 20 mm wide, and as to their thickness, present a greater diversity between the metric 2 and 15 mm. Metrical parameters of the 4 microscrapers are 17 mm long, 11.4 mm wide and 3.4 mm thick, including one that is a reuse of a microburin, transformed in an end microscraper. The group is completed with 15 scrapers that correspond to 7% of debitage material, and the typology includes sidescraper, scraper on flake and unguiform scraper. The group of retouched flakes consists of 22 pieces of small size, which translates to 10.9% of the tools. The lengths of the flakes are a minimum of 10-14 mm and a maximum of 35-39 mm, with the length more representative of 15-19 mm with a percentage value of 29%, followed by 20% at 25-29 mm and 16% of flakes at 10-14 mm. Seventeen burins have been identified that represent 7.9% of the tools and as regards their typology there are only two types: a dihedral burin and a truncation burin. The first type is the most significant, with 14 pieces and the second with just three. About 50% of the dihedral burins have a medium length between 15 and 24 mm, while the remaining burins have diversified, with metric parameter values between 10-14 mm and 35-39 mm. There are only 4 points, three of these are not retouched and include one short ogival which is only one of two in existence, and the fourth presents alternating retouching. Lastly there are 6 awls with metric parameters of an average length of 23 mm and an average width of 14 mm. Their thicknesses are more variable, with values between 2 mm and 10 mm.

Interpretations and Parallelism x The cores are chosen for the production of bladelets and flakes; the bladelets’ morphologies are regular, thanks to the positional stability of the cores, the preparation of the point of impact and the percussor’s quality, when used with the pressure technique; x The reduced size of the cores at their final stage shows a policy based on the saving and rentabilising of the raw material and the presence

Reading the Lithics in Flint of Cabeço dos Morros Shell Midden

87

of cortex on small bladelets and micro-tools supports the idea of an intense exploitation of that raw material; x The microliths’ geometrics outnumber other implements due to their multi-functionality and easy replacement; x The high number of flakes presupposed the existence of a greater number of cores; of the 470 flakes, only 22 are retouched. Most of the flakes would have been obtained in carving operations in a first phase of debitage; x There are significant typological differences between Cabeço dos Morros and the other two small shell middens Vale da Fonte da Moça I and II (Santos et al. 1990); the choice of different geometric types: the trapezoids’ geometric election in Cabeço dos Morros, with an almost total absence of the other, with the exception of 5 triangles. The Vale da Fonte da Moça I features a reverse situation; the segments are the geometric type and trapezes with only 3 and 9 triangles. In the Vale da Fonte da Moça II there is another picture: an equal number of triangles and circle segments and only 4 trapezes. x The percentage relationship between debitage and instruments is different between the Vale da Fonte da Moça I and II, with values of 73% and 27%, respectively, while the retouched tools of Cabeço dos Morros are only 13%. x The tools such as scrapers, arrows, retouched flakes and retouched bladelets are featured in higher densities in the Vale da Fonte da Moça shell middens. In debitage material, about 53% is populated by debris in the VFM shell middens, a value that corresponds to just 36% in Cabeço dos Morros. x The proportions seem to indicate a higher need for instruments and production in Vale da Fonte da Moça shell middens, which also gave rise to a greater diversification of the flint typology.

References Cardoso, J. L. e Rolão J. M. F., 2002/2003. Muge Estudos Arqueológicos, Centro de Estudos de Arqueologia da Universidade Autónoma de Lisboa/Câmara de Salvaterra de Magos. Rolão, J. M. F., 1999. Del würm Final al Holocénico en el Bajo Valle del Tajo (complejo arqueológico mesolítico de Muge). Ph.D: dissertation, Universidade de Salamanca. Rolão, J. M., Joaquinito, A. & Gonzaga, M., 2006. O complexo Mesolítico de Muge: novos resultados sobre a ocupação do Cabeço da Amoreira, In Do Epipapelolítico ao Calcolítico na Península Ibérica. Actas do IV

88

Chapter Six

Congresso de Arqueologia Peninsular, pp. 27-41, Faro: Universidade do Algarve Santos, M. F., Marques, G. D. e Rolão, J. M. F. 1990, I e II campanhas no concheiro do vale da Fonte da Moça I. in Actas do I Congresso do vale do Tejo, Lisboa.

CHAPTER SEVEN MARINE INVERTEBRATES AND MODELS OF THE ECONOMIC ORGANIZATION OF THE COASTAL ZONE DURING THE MESOLITHIC: FRENCH AND PORTUGUESE EXAMPLES CATHERINE DUPONT1 AND NUNO BICHO2 1

CNRS UMR 6566 CReAAH Université de Rennes 1, Campus Beaulieu, bât. 24-25, 35042 Rennes, France, [email protected] 2 FCHS, Universidade do Algarve Campus de Gambelas 8005-139 Faro, Portugal, [email protected]

Abstract In spite of the extremely high frequency of marine invertebrates in Mesolithic shell middens, their study was not systematic before the end of the 20th century. The development of archaeological disciplines such as archaeozoology helped to change the method of excavation involved in Mesolithic shell middens. Finally, the diet of Prehistoric populations from old excavations was only known from relatively large artefacts collected by hand, and thus, only large fish and more resistant shells were collected. So our knowledge of the components of food was totally distorted and their description was reduced to a list of species and to the largest animals. During the last twenty years this outcome has caused the “reexcavation” of some of the most famous Mesolithic shell middens, including Beg-an-Dorchenn and Beg-er-Vil in France and those from the Muge and Sado complexes in Portugal. These new excavations offer a renewed vision of the diet and residence of these populations. They also allow us to discuss their place in the spatial organization of the coast during the Mesolithic.

90

Chapter Seven

Introduction On the scale of the European Atlantic facade, more than 300 Mesolithic sites near the coast or subject to the marine influence testify to the consumption of shells (Fig. 7.1). Many of them have been known for a long time and only from survey (Gutiérrez-Zugasti et al. 2011) while only about 45% of them have been radiocarbon dated. The old excavations focused on structures such as graves, hearths, or paved areas (GutiérrezZugasti et al. 2011). On the archaeozoology side, the big mammal remains are the best described. The presence of other animal remains is not always referred to. When this is the case, the recorded species are the ones which leave the biggest fragments and are the best known to archaeologists. Paradoxically, while shells represent the most visible volume of these Mesolithic accumulations, the mollusc is not the food component that has been the most studied. Beg-er-Vil can be cited as an example. In the 1980s, the publications of the excavations refer to the flat oyster Ostrea edulis as the main species (Kayser 1990). Indeed, this bivalve is massive and is still the most visible in the section eroded by the sea. But, when we sort out the components of this accumulation in terms of the minimum number of individuals, it is the mussel Mytilus edulis which is mainly present (MNI, Dupont 2006). This difference is linked to the fact that people tend to find what they know best when they excavate. In addition, mussel shells badly handle the pressure of sediment. In Beg-er-Vil, in spite of counting a thousand mussels, not a single whole valve was discovered. Likewise at Cabeço da Amoreira, a single Scrobicularia plana is represented by 58 fragments and a Cerastoderma edule by 36. This admission of fragmentation can authorize two conclusions: excavation without sieving does not allow an estimation of animal diversity and only the application of a fine mesh allows a reliable quantification of what is nowadays preserved. In this paper, we will focus on two archaeozoological groups: the marine molluscs and the crabs.

The French Model The General Context Mesolithic shell middens are scarce along the French coast (Fig. 7.1). This is not connected to the lack of use of the sea by these populations but to all phenomena which accompanied the last transgression (Dupont 2006). Only six Mesolithic sites are known for their consumption of shellfish. Beg-an-Tour, dated from the Middle Mesolithic was unfortunately

Marine Invertebrates and Models of Economic Organization

91

destroyed before it could be excavated (Du Châtellier 1881, Fig. 1). It was a shell midden but nothing is known about its shell composition. Five others shell middens are dated to the Late Mesolithic. This late presence of Mesolithic people on the French coast is not linked to a massive arrival of Mesolithic populations on the coast in this period. In fact, a taphonomic bias can be at the origin of this distorted archaeological vision. This period corresponds to the slowdown of the rise of sea level. Among these sites, the most famous are Téviec and Hoëdic excavated during the first half of the twentieth century (Péquart et al. 1937; Péquart and Péquart 1954; Fig. 1). Unfortunately, the exact composition of shell species in these shell middens is not known. A return to the field is, at present, impossible for administrative reasons. Beg-er-Vil, Beg-an-Dorchenn and Saint-Gildas 1b are better known (Fig. 1). The two former sites were excavated in the 1980s (Kayser 1985, 1990). Sedimentary samples have been collected at Beg-er-Vil. They had been sieved but only studied more than ten years later. This site, which suffers from sea erosion, has been under excavation since 2012 by G. Marchand and C. Dupont. Only a small number of shells were collected in Beg-an-Dorchenn during the excavation. A complementary sampling was necessary to better estimate the real malacofaunal composition of this accumulation. So, in 2001 a total sampling of 1 m² was realized (Dupont et al. 2010). The “pointe de Saint-Gildas” is known for its density of microliths (Marchand 1999; Tessier 1984). Two shell accumulations have been observed: Saint-Gildas 1b and 1c. They were described during survey and surface survey (Tessier 1984). In 2003, the alarming degradation of one of them (Saint-Gildas 1b), led us to excavate this accumulation (Dupont and Marchand 2008).

A Review of the Applied Methodology In order to take advantage of the most recent archaeological surveys, a protocol was set up between various specialties of archaeozoology. This consists of the sieving of the sediment of archaeological sites with 2 and 4 mm meshes. The sorting of sediment allows for the visibility of archaeozoological remains (shells, fishes, micromammals, birds, sea urchins, goose barnacles, crabs and barnacles) in addition to the bones of larger mammals collected by hand during the excavation. All the fragments greater than 4 mm can be counted. The number of fragments (NISP) found in the 2 mm mesh can be estimated by the integral study of a fraction of the sample and by applying a rule of proportion.

92

Chapter Seven

Fig. 7.1. Distribution of Mesolithic shell-middens along the Atlantic European façade.

The total number of remains from the 2 and 4 mm mesh allows the NISP to be obtained. The Minimum Number of Individuals (MNI) is obtained by taking into account the highest number of apex or apertures for the gastropods, the number of apex associated with a muscular print for the limpet and the highest number of left or right valves for bivalves. For crabs, the MNI corresponds to the number of claws (Dupont and Gruet, in press). An important point is that the sorting is not only a thankless task; it

Marine Invertebrates and Models of Economic Organization

93

is a stage that will prepare the quality of future analyses. So it is crucial to do the sorting in association with the different fauna specialists.

Saint-Gildas: a Fast but Repetitive Passage of Mesolithic on the Coast? The first description of the shell deposit of Saint-Gildas 1b was that it was 2.50 m long, with a width of no more than 0.80 m and a thickness of 0.25 m above ground (Bellancourt 1980: 17). In 2003, a salvage excavation showed that the site was totally destroyed by the sea and we can study within the Mesolithic level of the cliff a lens of Scrobicularia plana. In spite of this huge destruction, the composition of the sample studied in 2003 is similar to precedent descriptions (Dupont 2006). The archaeological site of Saint-Gildas only consists of shells and lithic artefacts (Dupont and Marchand 2008). No other anthropogenic structures or animals have been observed. Thirteen species have been identified. Only two species (Scrobicularia plana and Patella sp.) are frequently present as whole shells–more than ninety percent are whole shells (Fig. 7.2). According to the lithic industry in the northwest of France, satellite sites could be complementary to habitats used for prolonged residence (Marchand 2005). If we apply this model to the sites with shells, SaintGildas b is more the reflection of specific expeditions (Dupont and Marchand 2008; Table 1). This result is based on the volume of the shell midden, the shells being the only food component, the high proportion of one species, and the absence of burials and of habitat structures (Table 7.1). The paleoenvironmental reconstruction shows that the Saint-Gildas site was located near the estuary of the Loire. This represents a stage in the migration of various animal species (birds and fishes). Thus, the human groups could exploit these resources, which would be more abundant at certain times of the year, and consume shellfish while at their stopping place before returning to a residential base (Dupont and Marchand 2008).

Beg-er-Vil: a Specific Image of Stability The original surface of Beg-er-Vil is difficult to estimate. It is clear that the site has been damaged by marine erosion since its first observation. At least 100 m² would have been observed with a thickness of 50 cm (Kayser 1987, 1990). The same thickness is currently still present (excavations by Marchand and Dupont). Contrary to Saint-Gildas, the layer of shells is diversified: lithic artefacts, mammals, birds, charcoals, crabs, and fish are present. Thirty-three species of mollusks have been

94

Chapter Seven

identified: three times as many species as at Saint-Gildas. Four species of crabs are also present. We have to add the MNI percentages of nine species to obtain ninety percent of shells found at Beg-er-Vil (Fig. 7.2). This does not correspond to a unique environment. Both rocky and sandy shores have mainly been exploited. Muddy ones have also been used. Thanks to marine invertebrates we can see that various environments have been exploited by this Mesolithic group. The diversity of species and of exploited environments is not restricted only to shells. If we look at other animals we find the same scenario for marine and terrestrial birds, fish, and crabs (Fig. 7.2) both from marine and terrestrial environments. This result and the diversity observed are verified for all kinds of animals. This behaviour at once reflects a great knowledge of the coastal environment by these Mesolithic people and a high dependence of this population on the sea resources. The exploitation of all the diversity of resources within a radius of five kilometres can be opportunistic behaviour. If we look at the potential seasonal accessibility of resources exploited at Beg-er-Vil, these Mesolithic people could have stayed more than one season at the site (Dupont et al. 2009). The accessibility to diverse resources, both terrestrial and marine, could be a reason for the beginning of the residential stability of these populations. If we confronted the criteria of Beg-er-Vil with the model of the economic organization of the territory during the Mesolithic, this shell midden can correspond to a residential base (Dupont and Marchand 2008). This integration of Beg-er-Vil in this category is linked to the wider diversity of species and animal classes and the presence of an anthropogenic structure (paved area, Table 7.1). So, the logistic model based on the lithic industry seems to be exercisable to French shell middens. This model would be allowed if only invertebrates could be taken into account in order to easily separate the residential base from the satellite sites. For the former, fishes, mammals and birds accompany shells and crabs in the food registers. More than 4 species of mollusks are necessary to reach more than 90% of the MNI and several species of crabs are also represented to more than 10% of the MNI. Several marine environments were exploited. For the more temporary sites, a single species represents more than 90% of the NMI at the same time for crabs and mollusks. These data are compared with those from the Portuguese Mesolithic shell midden of Cabeço da Amoreira.

Marine Invertebrates and Models of Economic Organization

95

Tab. 7.1. Cabeço da Amoreira in comparison with the French Late Mesolithic shell-middens (in light grey common points with Cabeço da Amoreira). Littoral residential base (after Dupont and Marchand 2008) Beg-an-Dorchenn Téviec Hoëdic Beg-er-Vil The addition of 4 to 5 species is necessary to obtain 90 % of the total weight of seashells A wider exploitation of the environment: others faunal remains are represented (mammals, fishes, crabs...) Residence at least during the half of a year at Began-Dorchenn and at the beginning of the spring and of the autumn at Beger-Vil Habitat structures (hearths, paved surfaces) Burials at Téviec and Hoëdic Full lithic reduction sequences Raw materials available locally

Littoral locations (after Dupont and Marchand 2008)

Littoral multifunction

Saint-Gildas 1b et 1c

Cabeço da Amoreira

The addition of 1 to 2 species is necessary to obtain 90 % of the total weight of seashells Seashells are the only faunal remains represented

The addition of 1 to 2 species is necessary to obtain 90 % of the total weight of seashells A wider exploitation of the environment: others faunal remains are represented (marine and terrestrial mammals, birds, fishes, crabs...) Unknown time of residence, but likely long.

Unknown time of residence

No burials

Habitat structures (hearths, post-holes and pits) Burials

Full lithic reduction sequences Raw materials available locally (imports from the inland are rare)

Full lithic reduction sequences Raw materials available locally (imports from the inland are rare)

No habitat structures

Chaapter Seven

moreira Fig. 7.2. Marrine invertebrates spectra in Beg-er-Vil, Saint-Gilddas 1b and Cabeço da Am

96

Marine Invertebrates and Models of Economic Organization

97

Beginning the Comparison with Cabeço da Amoreira Muge Mesolithic Shell Middens Cabeço da Amoreira belongs to the complex of the Muge Mesolithic shell middens (Fig. 7.1). Its geographical position is interesting because it is localized along the river Muge, which was under marine influence during the Mesolithic. In a similar way to Saint-Gildas (Dupont and Marchand 2008), the sites of Muge were probably near vast mudflats (van der Schriek et al. 2008). Among the sites of Muge, at least four were totally destroyed. Cabeço da Amoreira was found in 1864 and has recently been the subject of work coordinated by Bicho. It is one of the largest middens in the Muge area with an exposed mound of c. 60 metres in diameter and close to 3.5 metres in height at the centre (Bicho et al. 2011). This site was discovered by Carlos Ribeiro and Pereira da Costa (Ribeiro 1884). It was then excavated several times between 1884 and 1885 by F. Paula de Oliveira, from 1930 to 1933 by Mende Corrêa, Serpa Pinto and Santos Junior, from 1952 to 1965 by Jean Roche and Veiga Ferreira (Zbyszewski et al. 1980-1981), José Rolão from 1998 to 2001 and Nuno Bicho since 2008 (Bicho et al. 2011). It was during this last excavation that we were able to apply the same protocol as the one that had been applied to French shell middens.

What about Marine Invertebrates at Cabeço da Amoreira? The majority of marine invertebrate species are often listed by archaeologists, but the quantifications are not. In 1986 A. Lentacker gave the number of 25272 Cerastoderma glaucum for 2581 Scrobicularia plana, which corresponds to 90% of cockles in Cabeço da Amoreira. Also, A. Mendes-Corrêa in 1933 quoted the cockle Cardium edule as the most plentiful shell, while Scrobicularia plana was cited as very frequent. The cockle is quoted at Cabeço da Amoreira with the terms Cerastoderma glaucum and C. edule (formerly called Cardium edule). In archaeology, the differentiation between Cerastoderma glaucum and C. edule is very difficult because it is based on the shape of the shell and the variation in their morphologies–C. glaucum supports a lower degree of salinity than C. edule (Poppe and Goto 1993). In the present state of the study, the term C. edule has been chosen because shells already observed are thick and mainly symmetrical. The result obtained is in total opposition to previous publications (Lentacker 1986; Mendes-Corrêa 1933; Fig. 2). Indeed, after the application of the

98

Chapter Seven

new protocol, it is the peppery furrow shell which dominates the corpus with about 95% and more than 86381 individuals studied at present, against only a MNI of 4263 for cockles which is close to 5% of the represented shells. How can we explain this major difference? The peppery furrow shell is one of the most fragile shells and it is rarely discovered whole in archaeology. It is thanks to a fraction of its shell, the ligamentar print or chondrophore, that we were able to estimate the number of right and left valves of this species. It is true that if we had deducted only the complete or sub-complete valves the result would have been close to that of our predecessors. So the sea-shells from Cabeço da Amoreira correspond to the quasi-exclusive exploitation of a mudflat by Mesolithic people as a supplement to the collection of cockles in a sandy area. Two species of crabs were identified at Cabeço da Amoreira: the green crab Carcinus maenas and the fiddler crab Gelasimus tangeri (Dupont and Gruet, in press). They had been identified before then but never quantified. The quality of the recent excavations of Cabeço da Amoreira was recently underlined by the discovery of part of the mandible of these animals (unpublished, C. Dupont). The profile of the faunal spectrum is identical to the shells (Fig. 7.2). A single species of crab widely dominates the corpus: the green crab. This species is also present in the brackish areas of estuaries and bays, as well as in the highly salty lagoons (Gruet 2002). The fiddler crab can live in the same environments, in muddy marshes, often near the mouths of rivers (Lloris and Rucabado 1998). So, at once, shells or crustaceans found at Cabeço da Amoreira indicate the exploitation of a unique environment: a muddy area with a variation in salinity. This result could be interpreted as a selective and indeed an expeditious behaviour by the occupants of Cabeço da Amoreira if the accessible resources were not taken into account. Contrary to Beg-er-Vil, localized at the interface between different marine environments, Cabeço da Amoreira is, in a sense, trapped in an estuary. If this environment is very productive, the number of accessible species was likely restricted. So, it could be the access to a reduced panel of marine invertebrates which led to a selective spectrum rather than just anthropogenic choice.

Marine Invertebrates and Models of Economic Organization

99

Discussion What about Cabeço da Amoreira in the Theoretical French Model? If only invertebrates are compared between the littoral residential base and the satellite littoral locations of the French sites, Cabeço da Amoreira shows some similarities with Saint-Gildas: a single species represents more than 90% of the MNI (Fig. 7.2). Should we, then, assign Cabeço da Amoreira as a temporary camp? An initial difference is nevertheless visible in the marine invertebrates. Contrary to Saint-Gildas, crabs were consumed at Cabeço da Amoreira. The synthesis of the different criteria also shows more common points with residential bases: many animal classes were consumed, and hearths and graves were discovered in the midden (Table 7.1). As at Beg-er-Vil, the crossing of the accessibility of different resources shows that Mesolithic people were able to live a great part of the year, even all year long, at Cabeço da Amoreira (Lentacker 1986: 22). But this possibility is not proof of the permanence of a settlement. The large number of graves discovered in the shell midden (more than 20) was also used to give the hypothesis of a residential base occupied all year long (Cunha and Cardoso 2001). The last excavations however come to question the contemporaneousness between the graves of Cabeço da Amoreira and the shell midden (Bicho et al. 2011, 2013). “There were five main phases of anthropogenic deposition separated by at least two long hiatus” (Bicho et al. 2013). There are clear residential phases and others with a funerary function. Some of them can be contemporaneous but often spatially separated. So the re-excavation of Cabeço da Amoreira is crucial. Shell middens had been taken into account as a single occupation, but they are perhaps spatially and stratigraphically more complex than previously thought. It is exactly this argument and this report which caused the reopening of several shell middens along the European Atlantic coast (Gutiérrez-Zugasti et al. 2011).

Are the Shell Middens Single Occupations? It is not an easy task to understand all the complexities of the French Mesolithic sites. They are all coastal sites which had already been eroded by the sea when first described. The archaeologists often focused on the hearts of the shell middens without testing their peripheries. Unfortunately for Téviec and Hoëdic, new excavations are not possible at present. The old excavations tend to describe graves globally as contemporaries to the

100

Chapter Seven

shell midden with the reopening of some of those graves (Péquart et al. 1937; Péquart and Péquart 1954). It is, however, not impossible that different areas from the site were used sequentially for several activities (waste disposal, rest area, area of ornament or tool confection, tools, areas of cooking, etc.…), as recently found in Cabeço da Amoreira. The zones of excavation at Beg-an-Dorchenn and Beg-er-Vil are small and have thus likely reduced the possible diversity of activities: 53 m² for the first and (Kayser 1990) some twenty square metres for the second (Marchand and Dupont, 2014). Only shell levels were excavated but no graves were found there. An area of burned granitic blocks of about 4 m in diameter with a hearth in the middle was discovered at Beg-an-Dorchenn (Kayser 1986). Pits were observed at the base of the stratification for Beger-Vil and in the lower part of the shell midden a level with oysters and burnt stones was also found (Kayser 1990; Marchand and Dupont, 2014). The recent excavations will perhaps allow us to know the function of these structures but also to know if the Mesolithic presence can be identified outside the shell midden in the part opposed to the coast. The dates obtained from the base to the top of the accumulation are homogeneous (between 6200 and 6000 cal BC). These underline the fact that there are no huge breaks in the construction of this shell midden. Micromorphological studies should allow us to know if there are short interruptions in this accumulation. Lots of questions can be asked at Saint-Gildas because the shell level was excavated very late, at a time when the site was mostly eroded by the sea. Can this degradation be at the origin of this composition based only on shells as faunal remains? Indeed, we were able to see this at Beg-anDorchenn and it seems to be true at Beg-er-Vil that all the peripheries of these shell middens have been destroyed (Dupont et al. 2010). A more important fragmentation of shells is also present, possibly due to a more acid pH and doubtless also causing a more important dissolution of the osseous remains. The 1980 survey does not show a presence of bones when the shell level was still great (Bellancourt 1980; Tessier 1984). These observations tend to confirm the unique presence of shells at this site. The production of flint tools, including the manufacture of hunting weapons, is probably one of the motivations for the settlement of prehistoric communities on the Pointe Saint-Gildas (Dupont and Marchand 2008). The access to the diverse animals which frequent estuaries seasonally could also be another cause for change.

Marine Invertebrates and Models of Economic Organization

101

The Theoretical Model: Above all, a Tool for the Archaeologist So, even if the archaeologists try to get as close as possible to reality, the time since the abandonment of the site inexorably erases its various components: the disappearance of some areas of the sites by erosion or anthropogenic action, the faster dissolution of the faunal fragments in the less dense zones of remains, the natural erosion of the midden in the peripheral zones, etc… The classification of archaeological sites is nevertheless necessary for the archaeologists to discuss their place in the daily life of Mesolithic people. The more that shell middens are studied the greater appears the diversity of human behaviour. So, two populations having access to similar environmental resources can react differently. Is this what we observe at Saint-Gildas and Cabeço da Amoreira? The topographic position of these two settlements differs and could be the key to their various functions. But it is not impossible that the same population acted differently in a given time and location according to its real life and its contacts. The new shell midden excavations which have been increasing recently will help to unmask this complexity.

Acknowledgements To all the Portuguese team which organized the meetings of Muge and to the persons in charge of excavation who allowed us to begin the study of mollusks from the field.

References Bellancourt, G., 1980. Le kjokkenmodding de la pointe Saint-Gildas et les sociétés à microlithes de l’intérieur et des rivages de Loire-Atlantique. Société Nantaise de Préhistoire. 1980-2:5-28. Bicho, N., Cascalheira, J., Marreiros, J. & Pereira, T., 2011. The 20082010 excavations of Cabeço da Amoreira, Muge, Portugal. Mesolithic Miscellany. 21(2):3-13. Bicho, N., Cascalheira, J., Marreiros, J., Pereira, T. & Dias R., 2013. Chronology of the Mesolithic occupation of the Muge valley, central Portugal: The case of Cabeço da Amoreira. Quaternary International. 308-309:130-139. Cunha, E. & Cardoso, F., 2001. The osteological series from Cabeço da Amoreira (Muge, Portugal). Bulletins et mémoires de la Société d’Anthropologie de Paris. 13(3-4 ):323-333.

102

Chapter Seven

Du Châtellier, P., 1881. Exploration des tumulus de Run-Aour et de la Torche en Plomeur (Finistère) et du kjökkenmödding de la Torche. Mémoires de la Société d’Emulation de Côtes-du-Nord. 19 :175-182. Dupont, C., 2006. La malacofaune de sites mésolithiques et néolithiques de la façade atlantique de la France : Contribution à l’économie et à l’identité culturelle des groupes concernés, British Archaeological Reports, International Series 1571. Dupont, C. & Gruet, Y., in press. Crab remains as an indicator to reconstruct human Mesolithic activities and the palaeoenvironment: the example of the European Atlantic façade, in: MESO 2010. Dupont, C. & Marchand, G., 2008. Coastal exploitation in the Mesolithic of western France: la Pointe Saint-Gildas (Préfailles). Environmental Archaeology. 13(2):143-152. Dupont, C., Marchand, G., Carrion, Y., Desse-Berset, N., Gaudin, L., Gruet, Y., Marguerie, D. & Oberlin, C., 2010. Beg-an-Dorchenn : une fenêtre ouverte sur l’exploitation du littoral par les peuples mésolithiques du sixième millénaire dans l’ouest de la France. Bulletin de la Société Préhistorique Française. 107(2):227-290. Dupont, C., Tresset, A., Desse-Berset, N., Gruet, Y., Marchand, G. & Schulting, R., 2009. Harvesting the seashores in the Late Mesolithic of north-western Europe. A view from Brittany? Journal of World Prehistory. 22:93-111. Gruet, Y., 2002. Reconnaissance de quelques espèces communes de crustacés (crabes et balanes) : application au site mésolithique de Beger-Vil (Morbihan, France). Revue d’Archéométrie. 26 :125-139. Gutiérrez-Zugasti, I., Andersen, S.H., Araujo, A. C., Dupont, C., Milner, N. & Monge-Soares, A.M., 2011. Shell midden research in Atlantic Europe: state of art, research problems and perspectives for the future. Quaternary International. 239 :70-85. Kayser, O., 1985. A propos de la fin du Mésolithique en Bretagne : l’amas coquillier de Beg-an-Dorchenn (Finistère). Note préliminaire. Travaux de l’Institut d’art préhistorique, Université de Toulouse-Le-Mirail, 8092. —. 1986. Les amas coquilliers d’Armorique. Archéologia. 218 :68-74. —. 1987. Beg-er-Vil : Quiberon (Morbihan), Campagne 1987, Sauvetage programme n°1400 site, n°56 186 007. —. 1990. Beg-er-vil. Gallia information, 64. Lentacker, A., 1986. Preliminary results of the fauna of cabeço de Amoreira and Cabeço de Arruda (Muge, Portugal). Trabalhos de Antropologia e Etnologia. 26(1-4) :1-26.

Marine Invertebrates and Models of Economic Organization

103

Lloris, D. & Rucabado, J., 1998. Guide d’identification des ressources marines vivantes du Maroc. Guide FAO d’identification des espèces pour les besoins de la pêche, Organisation des nations unies pour l’alimentation et l’agriculture, Rome. Marchand, G., 1999. La néolithisation de l’ouest de la France : caractérisation des industries lithiques. British Archaeological Reports, International Series 748. —. 2005. Le Mésolithique final en Bretagne: une combinaison es faits archéologiques, in: Marchand, G., Tresset, A. (Eds.), Unité et Diversité des Processus de Néolithisation sur la Façade Atlantique de l’Europe (6e–4e millénaires avant J.- C.). Société Préhistorique Français, Paris. pp. 67–86. Marchand G. & Dupont, C., 2014 - Maritime hunter-gatherers of the Atlantic Mesolithic: current archaeological excavations in the shell levels of Beg-er-Vil (Quiberon, Morbihan, France). Mesolithic Miscellany, 22/2, 3-9. Mendes-Corrêa, A., 1933. Les nouvelles fouilles à Muge (Portugal). Paris, Librairie E. Nourry. Péquart, M. & Péquart, S.-J., 1954. Hoëdic, deuxième station nécropole du Mésolithique côtier Armoricain, De Sikkel, Anvers. Péquart, M., Péquart, S.-J., Boule, M. & Vallois, H., 1937. Téviec : station nécropole mésolithique du Morbihan. Masson, Paris, 18. Poppe, T. & Goto, Y., 1993. European Seashells: Scaphopoda, Bivalvia, Cephalopoda. Verlag Christa Hemmen, Germany. Ribeiro, C., 1884. Les kjoekkenmoedings de la vallée du Tage, in : Comptes rendus de la IXème session du congrès International d’Anthropologie et d’Archéologie Préhistorique, Typographie de l’académie royale des sciences, Lisbonne, pp. 279-290. Tessier M., 1984. Les sites a microlithes entre Vilaine et marais poitevin. Études préhistoriques et historiques des Pays de la Loire. Association d’étude préhistorique et protohistorique des pays de la Loire, Nantes. van der Schriek, T., Passamore, D., Mugica, F., Stenvenson, A., Boomer, I. & Rolão, J., 2008.Holocene Paleoecology and floodplane evolution of the Muge tributary, Lower Tagus basin, Portugal. Quaternary International, 189:135-151. Zbyszewski, G., Penalva, C. & da Veiga Ferreira, O., 1980-1981. Découverte d’un instrument préparé sur bois de Cervus elaphus au Cabeço da Amoreira (Muge). Setubal Arqueologica. 6-7:39-45.

CHAPTER EIGHT WHAT’S NEW? THE REMAINS OF VERTEBRATES FROM CABEÇO DA AMOREIRA— 2008-2012 CAMPAIGNS: PRELIMINARY DATA RITA DIAS, CLEIA DETRY AND ALEXANDRA PEREIRA ICArEHB Universidade do Algarve

Abstract The new excavations at Cabeço da Amoreira since 2008 have allowed a more thorough recovery of animal remains. Thus, zooarchaeological analyses have benefited from a more detailed description of their spatial distribution and from the recovery of smaller bone fragments both in situ, and whilst sieving. In this article we will describe the animal remains from vertebrates recovered in layers 1 and 2 from these excavations in the central area of Cabeço da Amoreira. Species found in previous excavations continue to be present, such as rabbit (the most abundant), hare, red deer and wild boar. Dog, wild cat and red fox have also been found in these new analyses. The results are then discussed to understand the paleoeconomy, social complexity and landscape use. In spite of the long-lasting research on Muge’s faunal remains (Pereira da Costa 1865; Lentacker 1986; Detry 2007), the new excavations have brought out interesting new data that we preliminarily analyse here. Together these all contribute to the knowledge of the subsistence, economy and social organization of these communities.

106

Chapter Eight

Introduction The Muge shell middens have been known since the 19th century and have been excavated since early times. These excavations brought to light the site’s great potential and revealed numerous human burials. The first excavations took place in the Cabeço da Arruda shell midden in 1864 (Cardoso & Rolão 1999/2000), followed in subsequent years by the other shell middens both in the Muge and Magos valleys, a little to the south. The most recent archaeological excavations of the “Last Huntergatherers of the Tagus Valley–The Muge Shell Middens” project started in 2008 with excavations of the Cabeço da Amoreira shell midden, followed by the “The Last Hunter-gatherers of Muge (Portugal): the Origins of Social Complexity”, starting in 2011 until today (Bicho et al. 2011). This first fieldwork aimed to delimitate the shell midden (about 2400 m2), as well as locate other loci, the latter by doing several auger and test pits in the contiguous areas of the shell midden (Bicho et al. 2011). A homogeneous presence of archaeological materials appeared which led to the beginning of an excavation of 5 cm artificial spits. From the excavation of the first spits in the main area, several lithic artefacts were recovered (fire-cracked clasts, cores, flakes, blades and bladelets as well as retouched tools like endscrapers, notches and geometric microlithic tools), as well as mammal remains, bird remains, fish remains, reptile remains, several small fragments and complete shells, perforated shells and crustacean legs, as well as a few clay fragments. Parallel to these excavations, some radiocarbon dates were obtained (Bicho et al., 2010) and some paleobotanical (Wollstonecroft et al. 2006) and shell samples were taken, which are currently under study. In this article we will describe the animal remains from vertebrates recovered in layers 1 and 2 from these excavations in the central area of Cabeço da Amoreira. The results will then be discussed to understand the paleoeconomy, social complexity and landscape use. In spite of the longlasting research on Muge’s faunal remains (Pereira da Costa 1865; Lentacker 1986; Detry 2007), the new excavations have brought out interesting new data that we preliminarily analyse here, focusing especially on layer 1.

What’s New? The Reemains of Verteebrates from Caabeço da Amoreeira

107

Faunal Remains R Land Ressources A prelim minary analyssis of materiaals from layerr 1 (new exccavations) allowed a m more detailed insight i into all the small veertebrate remaains, since they were ppiece-plotted (those ( larger than 2 cm orr complete an natomical elements) or found in a piece-plotted d sieved buckket (Bicho et al. 2011, 2012). Fish,, amphibians and reptiles were w also inccluded in this analysis. 785 (NISP) remains of a total of 1589 were idenntified to a sp pecies or Family/Genuus. The vast majority of the remains w were from Oryyctolagus cuniculus. O On the other hand, h Lepus sp p. is in comparrison considerrably less represented.. Other small mammals are also prresent like Erinaceus E europaeus, SSciurus vulgaris, unidentifiied Rodentia aand Microtus sp., and a juvenile Caastor fiber fem mur, as well as Aves: Annas platyrhynchos and other unidenntified water fowl f (Table 8.1, Fig. 8.1).

Figure 8.1. A All small vertebrrates NISP percentage.

108

Chapter Eight

It should be noted that man-made traces are almost non-existent in the assemblage of small vertebrates, apart from some fire marks. However, some interesting treatments were applied to rabbit metapodials as shown in Figure 8.2. Reptile and amphibian data produced expected results, consistent with other coeval sites and territory with the presence of Bufo bufo (the most numerous), Bufo calamita, Pelobates cultripes, Lacerta lepida, Testudo hermanni and unidentified Bufonidae, Testudinae and Colubridae (Haws et al. 2011; Haws et al. 2013).

Figure 8.2. Examples of anthropogenic marks on small vertebrate bones recovered at Cabeço da Amoreira. Rabbit’s metapodials sawed in tubes (1) or fractured by bipartition (2).

Aquatic Resources In regard to fish, Myliobatiformes represents a considerable percentage of total NISP, but they are however over-represented in comparison with MNI since most rays have 1-9 teeth plates (depending on age and species). Mugilidae are by far, the most important species in the assemblage and this demonstrates the importance of its several identified species in the local hunter-gatherer economy and subsistence (Table 8.2, Fig. 8.1). Other species and/or families like Argyrosomus regius, Trachurus trachurus, Gadidae, Chondrichthyes and particularly several Hippoglossus hippoglossus teeth come to show the diversity of pursued species, and habitats (Table 8.2, Figure 8.1). The latter is especially noticeable since it is a marine oceanodromous species which is mostly benthic (occasionally pelagic) (Nielson 1986), which means there would have been incursions to coastal environments and the use of species from those habitats in human

What’s New? The Remains of Vertebrates from Cabeço da Amoreira

109

subsistence. Trachurus trachurus further supports this supposition, since this is also a marine oceanodromous species present in pelagic-neritic environments (Smith-Vaniz 1986) also suggesting the use of a variety of environments other than the estuarine. Argyrosomus regius, although mainly marine found mostly inshore, is known for the juveniles and subadults entering estuaries and coastal lagoons (Chao and Trewavas 1990). They also feed, among others on mugilidae, which are largely present in the Tagus estuary. Table 8.1. Land resources NISP. Species Rodentia Bufonidae Colubridae Testudinae Soricidae Leporidae Oryctolagus cuniculus Lepus sp. Lepus europaeus Lepus granatensis Erinaceus europaeus Sciurus vulgaris Castor fiber Microtus sp. Columba sp. Anser sp. Anas platyrhynchos Testudo hermanni Chelonia sp. Lacerta lepida Bufo bufo Bufo calamita Pelobates cultripes Total

Frequency 2 4 16 2 1 1 435 7 7 6 5 4 1 1 1 1 1 2 2 13 17 1 1 785

% .3 .5 2.0 .3 .1 .1 55.4 .9 .9 .8 .6 .5 .1 .1 .1 .1 .1 .3 .3 1.7 2.2 .1 .1 100.0

The fact that the preliminary analysis of layer 2 has shown an impressive amount of Mugilidae remains could indicate that a complete analysis of layer one would return similar results. However, the percentage of unidentified fish, due to poor preservation, could eventually alter the

110

Chapter Eight

species balance and could give an interesting outlook on site formation and taphonomy, especially as there is still very little information on fish taphonomy. Table 8.2. Marine resources NISP. Species Ciprinidae Mugilidae Chondrichthyes Gadidae Liza sp. Liza aurata Liza ramada Chelon labrosus Mugil cephalus Argyrosomus regius Raja sp. Trachurus trachurus Hippoglossus hippoglossus Total

Frequency 1 48 1 1 11 44 40 15 11 5 72 1 2 252

% .4 19.0 .4 .4 4.4 17.5 15.9 6.0 4.4 2.0 28.6 .4 .8 100.0

Bird Remains There are only three specimens from the Class Aves recovered from layer 1, identified as: Columba sp., Anser sp., and Anas platyrhynchos, making it impossible to draw significant conclusions about their presence (Figure 1). What is clear nevertheless, is the dissimilarity in the number and variety of species in an earlier analysis by Detry (2007). This is probably a consequence of the lack of stratigraphic information available at that point, leading to a grouping of realities that do not characterize layer 1, but a set of different realities that is erroneously congregated, since the earlier analysis comprises what we now know to be different layers. Also, that could mean that there are differences between layers. Columba sp. is a resident and well-distributed species of southern Europe which does not provide significant seasonality information. The other two anatidae species are aquatic species, which make sense in the Cabeço da Amoreira environment. The mallard will stay in the same place all year round in warmer, milder climates, which seems to be the case in Muge in this period; therefore, this also does not provide much information on site seasonality (Carboneras 1992).

What’s New? The Remains of Vertebrates from Cabeço da Amoreira

111

Big Game A preliminary analysis of layers 1 and 2 (2133 bones analysed so far) (Figs. 8.2, 8.3 and 8.4, Table 8.3) evidently demonstrated the severe fragmentation of the assemblage of large vertebrates, since it was not possible to identify over 90% of the analysed remains. Typically, the animal remains from bigger vertebrates in the Muge shell middens are intensively fractured for the extraction of bone marrow (Detry 2007). The debris from Cabeço da Amoreira shows that fact more clearly since all the sediment is thoroughly sieved, and smaller diaphysis fragments are recovered. The intensive use of bone marrow by Mesolithic huntergatherers is observed by more authors (Pérez-Ripoll 1992). The majority of identified remains in this preliminary analysis are from artiodactyls, followed by carnivores (Figs. 8.3-8.5, Table 8.3). Cervus elaphus and Sus scrofa are the most numerous species, representing more than 80% of identified remains (see Table 8.3). Capreolus capreolus (a total of 11) was also existent in smaller numbers and from aurochs (Bos primigenius) there were only two remains: one cuneiform found in Layer 1 and one metapodial fragment in Layer 2. Carnivores were represented by a significantly smaller number of remains than artiodactyls but are important not only for providing meat but also fur and eventually as hunting trophies. Canids were represented by one distal phalanx (Layer 1) and one mandible (Layer 2) from Vulpes vulpes and one phalanx II in Layer 1 from Canis familiaris. Felids were represented by Lynx pardinus with two metapodials in Layer 1 and two phalanx II in Layer 2 and by Felis silvestris with a mandible, distal humerus and one phalanx II in Layer 2. The dog (C. familiaris) is the first domesticated species and the only one domesticated before the Neolithic. The purpose of this domestication is still discussed, was probably as a pet, but certainly used to help in hunting and for protection. It is clear that it was an important species for these populations since several remains were found in several of the sites, including a complete skeleton buried in the same layer as human burials (Detry & Cardoso 2010). These animals have been buried by huntergatherers since c. 12.000 BP (Davis & Valla 1978) and are frequent in Mesolithic sites (Ledge & Rowley-Conwy 1988, Larsson 1990; Morey 2006).

Chapter Eight

112

Big game NISP % layer 1 70 60 50 40 30 20 10 0

59.57

19.15

14.89 2.13

2.13

2.13

Figure 8.3. Big game NISP, layer 1.

Big game NISP % Layer 2 100 76.892

80 60 40 20

16.552 3.712

1.892

1.032

Feliz silvestris

Lynx pardinus

0 Capreolus capreolus

Cervus elaphus

Figure 8.4. Big game NISP, layer 2.

Sus scofa

What’s New? The Remains of Vertebrates from Cabeço da Amoreira

90 80 70 60 50 40 30 20 10 0

113

76.892

16.552 0

0

3.712

Layer 1

1.892 1.032

Layer 2

Figure 8.5. Percentage of remains by species of big game in layer 1 and 2 of Cabeço da Amoreira.

Table 8.3. Number of identified specimens (NISP) by species of big game recovered in Layers 1 and 2 of Cabeço da Amoreira – campaign 2008-2011. Layer 1 Bos primigenius Cervus elaphus Capreolus capreolus Sus scrofa Castor fiber Canis familiaris Felis silvestris Lynx pardinus Total

Layer 2

Frequency

%

Frequency

%

1 28 7 9 0 1 0 1 47

2 60 15 19 0 2 0 2

0 89 4 19 1 0 2 1 116

0 77 3 16 1 0 2 1

These populations relied in part on large-sized species and red deer and wild boar were the main contributors to the diet. Both species are also frequent in other Mesolithic assemblages in Portugal (Valente 2008). These species are typical of the Mediterranean moderate climate, with mixed forests and open areas, and they would therefore be abundant on the Iberian Peninsula in the first half of the Holocene.

114

Chapter Eight

Cut marks, pressure marks, fire marks, oblique fractures and longitudinal fractures (in phalanxes) were consistently present (Fig. 8.6) in both red deer and wild boar remains. Other than anthropic, animal-made marks were also verifiable, like punctures and digested bones, indicating the presence of carnivores which is confirmed by the presence of remains.

Discussion and Conclusion Cabeço da Amoreira’s faunal analysis has thus far shown a diverse hunting, fishing and recollection subsistence and economy, facilitated by an elevated aquatic (the consequence of an ample and deep estuary) and terrestrial biodiversity (both in terms of fauna and flora) that allowed the establishment of a systematic occupation over 200 years. The results, although preliminary, are consistent with data from previous excavations in terms of identified species (Lentacker 1986; Detry 2007). They actually add to these with the discovery of new species (Castor fiber) and details about their stratigraphic provenance, allowing a much finer time-distribution. The importance of fish remains is significantly increased by the introduction of new recording and sieving methodologies, more accurately demonstrating their importance in Mesolithic hunter-gatherer subsistence and economy and standing in line with the isotopic analysis of human remains (Bicho et al. 2011; Umbelino 2006).

Figure 8.6. Examples of anthropogenic marks on bones recovered at Cabeço da Amoreira. One Red deer antler soft hammer and one astragalus with cut marks.

What’s New? The Remains of Vertebrates from Cabeço da Amoreira

115

The fact that there are very few anthropic traces in small vertebrate remains in comparison with big vertebrate remains (Figs. 8.5 and 8.6) does not mean that the latter were a preferred food resource, but that smaller bones like those of the rabbit were either more likely to be all cooked and then broken by hand, without the use of stone or other tools, or their more fragile epiphysis, which would eventually bear those traces, were destroyed during cutting. On the other hand, the natural introduction of e.g. rabbits to archaeological sites, in consequence of dens, would result in the appearance of complete skeletons in articulation or in close proximity. That does not happen in the Muge assemblage.

References Bicho, N., Umbelino, C., Detry, C. & Pereira, T., 2010. The Emergence of Muge Shell Middens in Central Portugal and the 8200 cal yr BP Cold Event. Journal of Island & Coastal Archaeology, 5: 86-104. Bicho, N., Cascalheira, J., Marreiros, J. & Pereira, T. 2011. The 20082010 excavations of Cabeço da Amoreira, Muge, Portugal. Mesolithic Myscellany, vol. 1, issue 2: 3-13. Bicho, N., Cascalheira, J., Marreiros, J., Gonçalves, C., Pereira, T., Dias, R., 2012. Chronology of the Mesolithic occupation of the Muge valley, central Portugal: the case of Cabeço da Amoreira. Quaternary International, 308-309: 130-139. Carboneras, C., 1992. Family Anatidae (Ducks, Geese and Swans). Del Hoyo, Josep; Elliott, Andrew & Sargatal, Jordi (eds.). Handbook of Birds of the World. Volume 1: Ostrich to Ducks: 536–629. Cardoso, J.L. & Rolão, J., 1999/2000. Prospecções e escavações nos concheiros mesolíticos de Muge e de Magos (Salvaterra de Magos): Contribuição para a história dos trabalhos arqueológicos efectuados. Estudos Arqueológicos de Oeiras 8: 306-357. Chao, L.N. & E. Trewavas, 1990. Sciaenidae. p. 813-826. In J.C. Quero, J.C. Hureau, C. Karrer, A. Post and L. Saldanha (eds.) Check-list of the fishes the eastern tropical Atlantic (CLOFETA). JNICT, Lisbon; SEI, Paris; and UNESCO, Paris. Vol. 2. Davis, S. & Valla, F., 1978. Evidence for domestication of the dog 12,000 years ago in the Natufian of Israel. Nature. 276: 608-610. Detry, C., 2007. Paleoecologia e Paleoeconomia do Baixo Tejo no Mesolítico Final: O contributo do estudo dos mamíferos dos concheiros de Muge. PhD thesis. Universidad de Salamanca/ Universidade Autónoma de Lisboa.

116

Chapter Eight

Detry, C. & Cardoso, J. L., 2010. On some remains of dog (Canis familiaris) from the Mesolithic shell-middens of Muge, Portugal. Journal of Archaeological Science. 37(11): 2762-2774. Haws, J., Dias, R., Pereira, V. & Hockett, B., 2011. Archaeofaunal records of MIS 3 and 2 environmental change in Lapa do Picareiro. Paper presented at the INQUA 2011 – International Union for Quaternary research, Berna (Switezerland). Haws, J., A., Benedetti, M., Hockett, B., Bicho, N., Friedl, L., Pereira, V., Kubátová, L. & Dias, R., 2013. Lapa do Picareiro: a 50,000-year record of human occupation and environmental change in central Portugal. GSA Annual meeting. Denver, Colorado, U.S.A. Larsson, L., 1990. Dogs in Fraction - Symbols in Action. In P. M. Vermeersch & P. Van Peer (eds.), The Mesolithic in Europe, pp. 153160. Leuven: Leuven University Press. Legge, A. J. & Rowley-Conwy, P. A., 1988. Star Carr revisited. London: Archaeological Laboratory. Centre for Extra-Mural Studies. Birbeck College. University of London. Lentacker, A., 1986. Preliminary results of the fauna of Cabço de Amoreira and Cabeço de Arruda (Muge, Portugal). Trabalhos de Antropologia e Etnologia 26, 9-26. Morey, D., 2006. Burying key evidence: the social bond between dogs and people. Journal of Archaeological Science, 33: 158-175. Nielsen, J.G., 1986. Pleuronectidae. p. 1299-1307. In P.J.P. Whitehead, M.-L. Bauchot, J.-C. Hureau, J. Nielsen and E. Tortonese (eds.) Fishes of the North-eastern Atlantic and the Mediterranean. UNESCO, Paris. Vol. 3. Pereira Da Costa, F.A., 1865. Da Existência do Homem em épocas remotas no vale do Tejo. Notícia sobre os esqueletos humanos descobertos no Cabeço da Arruda. Imprensa Nacional, Lisboa. Pérez-Ripoll, M., 1992. Marcas de carníceria, fracturas intencionadas y mordeduras de carnívoros en huesos prehistóricos del Mediterrânico Espanol. Alicante: Instituto de Cultura Juan Gil-Albert. Smith-Vaniz, W.F., 1986. Carangidae. p. 815-844. In P.J.P. Whitehead, M.-L. Bauchot, J.-C. Hureau, J. Nielsen and E. Tortonese (eds.) Fishes of the north-eastern Atlantic and the Mediterranean. UNESCO, Paris. vol. 2. Umbelino, C., 2006. Outros Sabores do Passado. As análises de oligoelementos e de isótopos estáveis na reconstituição da dieta das comunidades humanas do Mesolítico Final e do Neolítico Final / Calcolítico do território português. PhD Thesis. Universidade de Coimbra.

What’s New? The Remains of Vertebrates from Cabeço da Amoreira

117

Valente, M.J., 2008. As últimas sociedades de caçadores-recolectores no Centro e Sul de Portugal (10.000 - 6.000 anos BP): aproveitamento dos recursos animais. PhD thesis. Universidade do Algarve. Wollstonecroft, M., Snowdon, V., Lee, G.-A. & Austin, P., 2006. Archaeobotanical Sampling at Cabeço da Amoreira: preliminary results of the 2003 Field Season. In N. Bicho and H. Verissimo (eds.), Do Epipaleolítico ao Calcolítico na Península Ibérica. Actas do IV Congresso de Arqueologia Penínsular. Promontoria 04: 55-62.

CHAPTER NINE PRELIMINARY TECHNO-TYPOLOGICAL ANALYSIS OF THE LITHIC MATERIALS FROM THE TRENCH AREA OF CABEÇO DA AMOREIRA (MUGE, CENTRAL PORTUGAL) JOÃO CASCALHEIRA1, EDUARDO PAIXÃO, JOÃO MARREIROS, TELMO PEREIRA AND NUNO BICHO ICArEHB FCHS, Universidade do Algarve Campus de Gambelas 8005-139 Faro 1 [email protected]

Abstract One of the most remarkable aspects of the recent fieldwork at the Cabeço da Amoreira shell mound (Muge, Central Portugal) was the identification of a series of Neolithic and Mesolithic archaeological horizons located just outside the mound limits. These previously unreferenced occupations were exposed in various test pits around the midden and, more strikingly, in a 12x1 metre trench where it was possible to confirm that a total of five archaeological layers were formed during and after the construction of the shell deposit. Thousands of lithic materials, some ceramics, and a couple of fire structures, were found in this area but, with the exception of the shell midden layers, very few faunal remains were recovered. Here, we present the first results of the techno-typological analysis of the lithic materials from the Trench area. A functional and chronological dichotomy seems to exist between the shell midden layers and the

120

Chapter Nine

remaining horizons, mostly represented by differences in the frequencies of raw materials and some of the technological classes.

Introduction The Cabeço da Amoreira shell mound is one of the largest sites of the Muge Mesolithic complex, with an elliptical shape of c. 60 metres in diameter at its longest axis, and a stratigraphic sequence of roughly 3.5 metres (Bicho et al. 2011). The eastern part of the shell midden was extensively excavated in the 1930s and in the 1960s by the teams of A. Mendes Corrêa and of J. Roche and O. Veiga Ferreira, respectively. Although very detailed profile drawings are available from the 1960s excavations (Roche 1967), both teams divided the stratigraphy during excavation into three broad layers of occupation: Lower, Middle and Upper (Cardoso and Rolão 1999-2000). Indeed, even though some details are known about the spatial (horizontal) distribution of findings from older campaigns, very little information is available on its stratigraphic provenience (Bicho et al., in press). The exception would be the José Rolão excavation in the early 2000s, which had fairly good stratigraphic control, but unfortunately within only a very small excavated area, both in area (12 sq. metres) and in depth (1 metre) (Rocksandic 2006). The common focus of the previous work was the excavation of the dense shell layers, with a lot of the effort mainly directed towards the identification and analysis of burials. In order to fill some of the gaps left by the previous work, two research projects were carried out in the course of the last six years (entitled “The Last Hunter-gatherers of the Tagus Valley: the Muge Shell Middens” and “The Last Hunter-gatherers of Muge (Portugal): the Origins of Social Complexity”), both coordinated by one of us (NB) and funded by the Portuguese National Science Foundation (FCT). The focus of these projects was mainly the development of new excavations at Cabeço da Amoreira, where results have so far permitted us (1) to establish an absolute chronology for the time-span of site formation, now proved to be between c. 8000 and 7350 cal BP (Bicho et al. 2013); (2) to identify a spatially patterned basal occupation, below the shell layers, that likely had a residential and burial function (Bicho and Gonçalves, in press; Bicho et al. 2013); (3) the recognition of a top layer in the midden mostly composed of small pebbles and fire-cracked rocks, interpreted as a symbolic protective cairn for the burials below (Bicho et al., in press), dated to 7515-7350 cal BP (Bicho et al. 2013); and (4) to identify previously unreferenced “shell-free” archaeological horizons (including

Preliminary Techno-Typological Analysis

121

Neolithic) in a peripheral area contiguous to the mound itself (Bicho et al. 2011). The latter is perhaps the most interesting since, besides being virtually unique in the context of Mesolithic shell middens in Portugal and in the rest of Europe, it provides evidence for the existence of spaces with different functionalities in that specific locale that are other than midden deposits. It also confirms the pattern of the continuity of site use from the Mesolithic to the Neolithic (already attested by the radiocarbon dates of some of the top-layer burials), thus, re-launching the debate on the cultural interactions between the last hunter-gatherers and the first farming communities of Central Portugal. These layers were first exposed in a series of 1 m² test pits which were excavated around the site and later on larger excavation areas where good preservation conditions were found, called Area 1 and Area 2 (Fig. 9.1). Area 1 presented a sequence of layers that can be roughly divided into two different horizons that, despite the ongoing status of the material analysis, can be attributed to the Neolithic, with abundant (some decorated) ceramics and a low frequency of geometrics (specifically segments and triangles), and to the Mesolithic, without ceramics but with high frequencies of geometrics. Unfortunately, very few, poorly preserved, charcoal fragments were collected, and these were not enough to provide reliable absolute dating. Faunal remains were almost absent throughout the sequence from both areas, the exception being the shell midden layers in the Trench. The identification of these horizons at a small distance from the limits of the midden deposit raised the question of the relationship between these horizons and the construction of the shell mound as well as its use as a burial ground. To answer this, a 1x12 metre trench was opened between the northeastern corner of Area 1 and the southern limit of the midden (Fig. 9.2). The result was the identification of several Neolithic and nonmidden Mesolithic horizons from which thousands of lithic material finds and some ceramics were recovered, most of them overlaying the shell layers of the midden and with at least one possibly interstratified occupation between the top two layers of the midden. In this paper we present a preliminary analysis of the lithic assemblages recovered from the Trench. Particular attention is paid to the differences between the midden layers and the remaining horizons, concerning raw material presence and techno-typological variability.

122

Chapter Nine

Fig. 9.1. Cabeço da Amoreira topographic plan with indication of the excavation areas.

Trench Stratigraphy A total of eleven litho-stratigraphic layers were identified in the Trench. Three of these layers (2, 3 and 5) located towards the Northeastern side (Fig. 9.2), are shell midden levels, corresponding to Layers 1, 2 and 3 of the Midden area (Bicho et al. 2011). The remaining levels are essentially composed of a sandy matrix, whose colour varies depending on the presence of organic contents, although faunal remains were not recovered.

Preliminary Techno-Typological Analysis

123

With the exception of the V12 unit, all other units were excavated to the base of the sequence, a sterile late Pleistocene eolian sand-dune that covers the vicinity of the site. In methodological terms, excavation was carried out by 5 cm artificial spits simultaneously in all twelve units. Sediments were totally screened with a 1 mm mesh and all materials, excluding fire-cracked rocks and pebbles, were collected. From top to bottom the complete stratigraphic sequence of the Trench is: Layer 11–Fine sands resulting from the screening of the sediments from Area 1 (present in units V1 to V3); Layer 10–An organic layer of dark sand (7.5.YR2.5/1) with small shell fragments, fragments of small pebbles and no other artefacts (present in units V1 to V12); Layer 9–Well sorted and compact sands (7.5YR4/6) without roots or other natural inclusions. The presence of ceramic fragments and lithic industry that indicate a Middle Neolithic chronology; Layer 8–Well sorted sands (7.5YR4/3), similar to Layer 3 but lighter, and compact without roots or other natural inclusions. The presence of rare archaeological materials, probably resulting from episodes of colluviums from the surface of the midden (present in units V4 to V8); Layer 7–Well sorted sands (7.5YR4/4) without roots or other natural inclusions. The presence of ceramic fragments and some lithic implements, probably of Early Neolithic age (present in units V1 to V7); Layer 6–Well sorted fine dark brown sands (5YR2.5/2) with many lithic artefacts and without ceramic. It partially covers the top of the midden (present in units V1 to V7) and it is certainly of Mesolithic age; Layer 5–Dark sands (5YR2.5/2) similar to Layer 10, but with more shell fragments, with lithic artefacts and perforated shells. This layer is the southern limit of the upper layer of the midden, or "cairn”, which appears in the main area of intervention (present in units V5-V11); Layer 4–Well sorted fine sand, of a very dark colour (7.5YR2/1), burnt, very compact and hard, with many lithic artefacts, which intercalates between the top two layers of the midden and probably corresponds to the time of the production the of the cairn (present in units V1 to V7); Layer 3–A layer with a fine grey matrix (10YR3/1) and very fragmented shells with some pebbles and fire-cracked rocks (present in units V7-V11);

Chapter Nine

Fig. 9.2. Trench. West profile with indication of layers and excavation units.

124

Preliminary Techno-Typological Analysis

125

Layer 2–Fine light brown sands (7YR4/2) with shell fragments and frequent whole shells, rare lithic clasts or artefacts. In its lower part is presented a series of depressions that continue to Layer 1 and that were recognized as pits (present in units V8 to V11); Layer 1–Fine well sorted sands of a very light colour (10YR5/4) and sterile, showing the development of clay lamellae and stains in several areas, especially under the zone of contact with Layer 4. In addition to the negative structures detected in the bottom of Layer 2, two combustion structures were also identified in the north profile of units V9 and V11. Both were formed by circular or semi-circular burnt quartzite pebbles and revealed some poorly preserved charcoal fragments in their interiors. These hearths were located in the intersection between Layers 4 and 5 (units V7/V8) (in this case of Mesolithic age and covered by the concentration of shells from the shell midden layers) and in Layer 7 (unit V3) of Early Neolithic age according to the typology of the stratigraphically associated potsherds. According to the stratigraphic description presented above it is clear that, with the exception of Layer 4, all the layers identified in the border of the midden postdate the deposition of the packed shell levels. Concerning Layer 4, given the reduced size of the excavated area, its relation with the midden levels is more difficult to establish. It is certainly more recent than Layer 2 but, although it seems to be interstratified between Layers 3 and 5, it may also be contemporaneous with the latter.

Lithic Assemblage A total of 2821 lithic artefacts were analysed, of which the more frequent classes are flakes (c. 39%, including fragments), bladelets (c. 26%, including fragments) unclassifiable fragments (c. 16%) and