Ships, Boats, Ports, Trade, and War in the Mediterranean and Beyond: Proceedings of the Maritime Archaeology Graduate Symposium 2018 9781407317021, 9781407355979

Table of contents :
Front Cover
Title Page
Copyright
Of Related Interest
TABLE OF CONTENTS
Foreword
1. Introduction
2. Before ‘Thalassocracies’: Reconstructing the ‘Longboat’ and Rethinking its Use and Social Implications in the 4th and 3rdMillennium South Aegean
3. Technology Behind the Mazarrón Boats: a Virtual 3D Approximation
4. Experimental Archaeology and the Contributory Reconstruction of a Roman Warship
5. Managing the Threat: A Maritime Archaeological Study of the Island of Menorca as a Key Ancillary in the Roman Mediterranean
6. The Roman Port of Berytus
7. The Study of the Port System of the Coast of Almería from the Analysis of the Maritime Cultural Landscape, Eighth – Twelfth Centuries AD
8. The Whole Story: Exploring the Transportation of Whole Olives in Antiquity Through Shipwreck Evidence
9. Piracy in the Hellenistic Period: A Misunderstood Phenomenon
10. Investigating the Role of the Sea in Roman Crete’s Prosperity:A Maritime Archaeological Perspective
11. Cultural Underwater Heritage of the South Pacific Armada
12. Maritime Archaeology in Biscay, Basque Country: Facts, Acts, Research and Opportunities
13. Pioneers of Maritime Activity: The Uses and Abuses of the Maritime Aspects of Phoenician Culture
14. Exploring Hominin Movement Patterns in the Lower Palaeolithic Aegean Dry Land: Methodological Challenges

Citation preview

210mm WIDTH

2020

This volume gathers papers presented at the Maritime Archaeology Graduate Symposium, held in Oxford in 2018. The event was an Honor Frost Foundation initiative dedicated to new and upcoming research focused on eastern Mediterranean Archaeology and realised by a committee from the Centre for Maritime Archaeology, University of Southampton, and the Oxford Centre for Maritime Archaeology, University of Oxford. These essays represent the proceedings of this conference and contain some of the latest research and fieldwork in the Mediterranean and beyond. Topics include ships, ports and port systems, maritime economics, chemical analysis of archaeological remains and legislation in maritime archaeology. This publication is the product of an endeavour to promote early career research for maritime archaeologists with unique foci, and to establish a platform for them to discuss their findings with the wider community.

‘[This] volume is important in the sense that it shows the scientific contributions of students and young researchers to maritime archaeology.’ Dr Éric Rieth, Centre national de la recherche scientifique, France

Naseem Raad is a PhD candidate at the University of Southampton, based at the Centre for Maritime Archaeology. He is interested in the characterisation of what has been dubbed ‘the Roman economy’. Naseem combines economic theory with archaeological methodology to explore production and distribution networks in the Roman Levant. Carlos Cabrera Tejedor has a trans-disciplinary background, having started in conservation, and completed two bachelor’s degrees, an MA in Nautical Archaeology from Texas A&M University and a DPhil in Archaeology at the University of Oxford. He is a maritime archaeologist with an interest in ancient shipbuilding and ports. Contributors: Cristina Agudo Rey, Margaret A. Amundson, Lucy Blue, Lisa Briggs, Carlos Cabrera Tejedor, Marta del Mastro Ochoa, Dimitris Karampas, José Manuel, Matés Luque, John McNabb, Joseph M. Pacheco Jr, Mateusz Polakowski, Naseem Raad, Lamia Sassine, Peny Tsakanikou, Panos Tzovaras

210mm WIDTH

B A R I N T E R NAT I O NA L S E R I E S 2 9 6 1

2020

297mm HIGH

‘The[se] proceedings are a valuable contribution to maritime archaeology. … [The volume] expands on previous research, providing some new primary data, and fresh analysis of secondary literature contributing to new approaches in how we consider the Mediterranean basin.’ Professor Kimberly Monk, Trent University, Peterborough, Ontario

BAR  S2961  2020  RAAD & CABRERA TEJEDOR (Eds)  Ships, Boats, Ports, Trade, and War in the Mediterranean and Beyond

B A R I N T E R NAT I O NA L S E R I E S 2 9 6 1

11.2mm

Ships, Boats, Ports, Trade, and War in the Mediterranean and Beyond Proceedings of the Maritime Archaeology Graduate Symposium 2018

EDITED BY

NASEEM RAAD AND CARLOS CABRERA TEJEDOR

Printed in England

210 x 297mm_BAR RAAD 11.2mm CPI ARTWORK.indd 3

21/1/20 9:44 PM

210mm WIDTH

2020

297mm HIGH

B A R I N T E R NAT I O NA L S E R I E S 2 9 6 1

Ships, Boats, Ports, Trade, and War in the Mediterranean and Beyond Proceedings of the Maritime Archaeology Graduate Symposium 2018

EDITED BY

NASEEM RAAD AND CARLOS CABRERA TEJEDOR

210 x 297mm_BAR RAAD 12mm TITLE ARTWORK.indd 1

15/1/20 10:56 AM

Published in 2020 by BAR Publishing, Oxford BAR International Series 2961 Ships, Boats, Ports, Trade, and War in the Mediterranean and Beyond ISBN 978 1 4073 1702 1 paperback ISBN 978 1 4073 5597 9 e-format doI https://doi.org/10.30861/9781407317021 A catalogue record for this book is available from the British Library © the editors and contributors severally 2020 Cover Image 3D reconstruction of an early Byzantine merchant vessel. Cover image by Grant Cox (ArtasMedia). The Authors’ moral rights under the 1988 UK Copyright, Designs and Patents Act are hereby expressly asserted. All rights reserved. No part of this work may be copied, reproduced, stored, sold, distributed, scanned, saved in any form of digital format or transmitted in any form digitally, without the written permission of the Publisher. Links to third party websites are provided by BAR Publishing in good faith and for information only. BAR Publishing disclaims any responsibility for the materials contained in any third party website referenced in this work.

BAR titles are available from: Email Phone Fax

BAR Publishing 122 Banbury Rd, Oxford, ox2 7bp, uk [email protected] +44 (0)1865 310431 +44 (0)1865 316916 www.barpublishing.com

Of Related Interest

SOMA 2008 Proceedings of the XII Symposium on Mediterranean Archaeology, Eastern Mediterranean University, Famagusta, North Cyprus, 5-8 March 2008 Edited by Hakan Oniz

Oxford, BAR Publishing, 2009

BAR International Series 1909

Navigated Spaces, Connected Places Proceedings of Red Sea Project V held at the University of Exeter, 16–19 September 2010 Edited by Dionisius A. Agius, John P. Cooper, Athena Trakadas and Chiara Zazzaro Sub-series: British Foundation for the Study of Arabia Monographs (formerly Society for Arabian Studies Monographs), 12 Oxford, BAR Publishing, 2012

BAR International Series 2346

SOMA 2010 Proceedings of 14th Symposium on Mediterranean Archaeology Taras Shevchenko National University of Kiev Ukraine 23-25 April 2010 Edited by Yana Morozova and Hakan Oniz

Oxford, BAR Publishing, 2013

BAR International Series 2555

Anciens peuplements littoraux et relations Homme/Milieu sur les côtes de l’Europe atlantique / Ancient Maritime Communities and the Relationship between People and Environment along the European Atlantic Coasts Edited by Marie-Yvane Daire, Catherine Dupont, Anna Baudry, Cyrille Billard, Jean-Marc Large, Laurent Lespez, Eric Normand and Chris Scarre, Avec la collaboration de / With the collaboration of Francis Bertin, Chloé Martin et Kate Sharpe

Oxford, BAR Publishing, 2013

For more information, or to purchase these titles, please visit www.barpublishing.com

BAR International Series 2570

Contents Foreword............................................................................................................................................................................ vi Carlos Cabrera Tejedor 1. Introduction.....................................................................................................................................................................1 Lucy Blue and Naseem Raad 2. Before ‘Thalassocracies’: Reconstructing the ‘Longboat’ and Rethinking its Use and Social Implications in the 4th and 3rd Millennium South Aegean..............................................................................................................3 Panos Tzovaras 3. Technology Behind the Mazarrón Boats: a Virtual 3D Approximation..................................................................23 Carlos Cabrera Tejedor 4. Experimental Archaeology and the Contributory Reconstruction of a Roman Warship......................................35 Mateusz Polakowski 5. Managing the Threat: A Maritime Archaeological Study of the Island of Menorca as a Key Ancillary in the Roman Mediterranean......................................................................................................................................51 Margaret A. Amundson 6. The Roman Port of Berytus.........................................................................................................................................61 Naseem Raad 7. The Study of the Port System of the Coast of Almería from the Analysis of the Maritime Cultural Landscape, Eighth – Twelfth Centuries AD...............................................................................................................73 Marta Del Mastro Ochoa 8. The Whole Story: Exploring the Transportation of Whole Olives in Antiquity Through Shipwreck Evidence........................................................................................................................................................................87 Lisa Briggs 9. Piracy in the Hellenistic Period: A Misunderstood Phenomenon.............................................................................99 Joseph M. Pacheco Jr 10. Investigating the Role of the Sea in Roman Crete’s Prosperity: A Maritime Archaeological Perspective.......107 Dimitris Karampas 11. Cultural Underwater Heritage of the South Pacific Armada................................................................................119 Cristina Agudo Rey 12. Maritime Archaeology in Biscay, Basque Country: Facts, Acts, Research and Opportunities.........................127 José Manuel Matés Luque 13. Pioneers of Maritime Activity: The Uses and Abuses of the Maritime Aspects of Phoenician Culture............143 Lamia Sassine 14. Exploring Hominin Movement Patterns in the Lower Palaeolithic Aegean Dry Land: Methodological Challenges...................................................................................................................................................................149 Peny Tsakanikou and John McNabb

v

Foreword Carlos Cabrera Tejedor Oxford Centre for Maritime Archaeology, University of Oxford [email protected] The papers presented in this monograph are the result of a two-day conference held at the University of Oxford and sponsored by the Honor Frost Foundation (HFF) and the Oxford Centre for Maritime Archaeology (OCMA). The Maritime Archaeology Graduate Symposium, or MAGS, was conceived as a place of encounter for graduate students and early career researchers in Maritime Archaeology. It was hosted, for the first time, at the University of Oxford on February 3 and 4, 2018.

bringing a refreshing approach. This enthusiastic style is perceived in some founding works such as those of Honor Frost and her contemporary pioneers, but has been largely lost in many other monographs and proceedings through which Maritime Archaeology has consolidated its position as an academic field in the last decades. Perhaps the best contribution of the MAGS conference is the facilitation of synergies among early career specialist eager to contribute to their fields and meet other colleagues with the same passion. I know that the same opportunity is offered in other senior symposia and international conferences but, in my opinion, in those meetings (as it happens with their presentations and proceedings) other factors such as seniority, politics, academic ‘clans’, diplomacy, etc., play an important role in the interactions among attendees. It is not that MAGS was completely free of those academic idiosyncrasies, but certainly their influence was less.

MAGS was originally envisaged to be a low-key event, short of a day of colleagues exchanging ideas without the peer pressure existing at more senior symposia or academic meetings. It was initially conceived to be a national gathering (UK based) but we decided to make an open call of papers so friends from other countries would not feel unwelcome. It was to our initial surprise, and later on pride, that the ‘low-key event’ that we originally planned ended up lasting two full days, gathering about 70 attendees, including 28 speakers, that came to Oxford from at least twelve different countries distributed among four different continents.

I believe that the MAGS 2018 conference, and the papers presented in these proceedings that derived directly from it, are brave, academically bright, vibrant, and a suggestive compilation of heterogeneous maritime archaeology studies in different fields and geographic spaces. Based on my personal experience, I believe (and hope) that the MAGS would be destined to become an essential academic meeting for ‘friends and colleagues’ who, nearly free from academic constraints, aspire to overcome the contrast between underwater and maritime, between shipwreck and landscape, between artefacts and history.

The papers presented at the conference showed a wide range of topics within the framework of the growing Maritime Archaeology field. Within the attendees, most of the sub-fields and the different disciplines that accompany the Maritime Archaeology field (e.g. shipbuilding, GIS, archaeometry, wood studies, anthropology, classical studies, naval history, conservation, etc.) were represented at some degree or level. The papers presented in this volume are a selection, yet provide a good example, of those discussed during the two days of MAGS 2018. It is not the intention of this foreword, and it would be impossible, to summarise in just a few lines the main contributions discussed in these papers, but I consider there are some points that should be emphasised.

It is therefore a great pleasure knowing that MAGS 2019, hosted at the University of Southampton, was another successful ‘low-key’ academic meeting with again an international attendance. At the time I am writing these lines, the MAGS 2020, to be hosted at the University of Oxford, is in preparation and looks to be a successful continuation of the two previous meetings.

These papers contain something that is, usually, rare to find in other more senior publications or academic journals: that is, enthusiasm. These papers show, in my humble opinion, the eagerness of their authors to contribute to a field that they are passionate about. That is, in my view, priceless since usually academic and disciplinary constraints often end up dissolving the passion and energy with which the papers presented here are written. If only because of that, I would argue that MAGS 2018 was a success that contributes to the field of Maritime Archaeology by

To conclude, we must acknowledge the Honor Frost Foundation as well as the Oxford Centre for Maritime Archaeology for generously supporting this first meeting in Oxford without which MAGS would have never become a reality. And from these entities, Lucy Blue and Damian Robison are the individuals to whom we must express our gratitude in particular. For their contribution to MAGS 2018, we must also thank the two keynote speakers Dr vi

Foreword Jon Henderson and Professor Robin Allaby who provided the conference with two enlightening talks. I finally would like to acknowledge the friends and colleagues that worked very hard planning, organising, coordinating, and delivering the event, making MAGS 2018 a reality: Crystal El Safadi, Veronica Walker Vadillo, Lisa Briggs, Ziad Morsy, Naseem Raad, Peter Fiske and myself. It is largely because of them, their endless enthusiasm, teamwork, humour, professionalism, diplomacy, resilience, and hard work that MAGS 2018 was, in the words of many of the attendees, a great success. I personally have had the immense privilege to briefly meet some of the international forefathers and pioneers of the field of Maritime Archaeology. It has been an honour learning directly from some of the world-leading experts (in different countries and research institutions) who, in recent decades, have gained our field (international recognition) within the wider academic world. Meeting early career colleagues full of enthusiasm and passion about what they want to study provides a different yet equally impressive and uplifting experience. We must, therefore, look forward to the contributions of those new generations that are arriving to the field of Maritime Archaeology, as they will not only consolidate and strengthen our field of research, but also take it to a new stage.

Figure 1. MAGS 2018 organising committee. From left to right, Veronica Walker Vadillo, Carlos Cabrera Tejedor, Lisa Briggs, Peter Fiske, Ziad Morsy, Crystal El Safadi and Naseem Raad.

vii

1 Introduction Lucy Blue1 and Naseem Raad2 1

Maritime Archaeological Director, Honor Frost Foundation [email protected] University of Southampton [email protected]

2

of their career could network with other similarly placed scholars. The first day focused on eastern Mediterranean maritime archaeology, and was divided into three sessions: ships and cargoes, maritime movement and trade, and the interpretation and reception of maritime archaeology in the eastern Mediterranean. The second day focused more broadly on methodological approaches, and maritime heritage and legislation. Throughout the weekend, it was clear that presenters were eager to publish their research in a self-contained and comprehensive volume.

This volume entitled Ships, Boats, Ports, Trade, and War in the Mediterranean and Beyond, is the product of a symposium held in 2018 that was initially conceived as a forum to encourage early career researchers to discuss a variety of topics related to maritime archaeology in the eastern Mediterranean. The premise revolved around a co-operation between the University of Southampton and the University of Oxford for students and early career researchers to exchange ideas in a supportive environment. This initial concept grew into a more substantial event that saw interest beyond UK institutions, sparking the development of The Maritime Archaeology Graduate Symposium (MAGS). The symposium was inspired and sponsored by the Honor Frost Foundation (HFF), and organised by a team of HFF scholars from the Centre for Maritime Archaeology at the University of Southampton, along with colleagues from the Oxford Centre for Maritime Archaeology.

Thus, the first MAGS publication was established. This volume strives to characterise the central philosophy of the conference, reflecting the complexity of eastern Mediterranean archaeology and maritime archaeology as a whole. Thirteen papers are presented by international authors, and have been divided into four sections: ‘ships and boats’, ‘ports and port-systems’, ‘maritime economics’ and ‘legislation, management and the perception of maritime archaeology’. All unified by the general theme of maritime archaeology as presented at MAGS, these works mirror the diversity of methodological approaches and foci of modern archaeology.

MAGS was founded to promote the collaboration between early-career researchers working on topics related to maritime archaeology, with a particular focus on the eastern Mediterranean. The organisers sent out the call for papers with this core concept in mind, anticipating a response from regional institutions within the UK. Several weeks later, they realised that they had underestimated the potential scale and scope of the event, with submissions from numerous international scholars expressing an interest in participating. What initially was conceived as an afternoon with a few colleagues chatting about maritime archaeology in a relaxed environment, gradually emerged into a two-day conference involving multiple international organisations.

The first section takes us through the potential of ship and cargo analysis, iconographic studies and digital reconstruction. Each paper gives a glimpse into Bronze Age, Iron Age and Classical shipbuilding techniques to shed light on the processes involved in maritime transportation through the ages. This is followed by a characterisation of the harbours that sheltered ships and boats and facilitated interconnected maritime networks. This section embodies an interdisciplinary approach that incorporates geomorphology, material culture and urban analysis to provide a comprehensive examination of ports and port systems. The third section centres on some of the economic systems within which ships and harbours operated. Themes such as indicators of interpreting ancient maritime commerce, including amphorae, shipwrecks and ancient texts, are discussed. The final three works in this volume tackle broad concepts related to heritage management and the perception of maritime culture in the modern world. Within today’s dynamic political geography, it is important to reflect on the role that maritime archaeologists should play in the preservation

The symposium proved to be a success, with 25 speakers sharing their research on a variety of topics and engaging one another in lively discourse. Beyond strictly academic collaboration, Masters students, Doctoral candidates and Postdoctoral researchers became familiar with an emerging generation of maritime archaeologists from all over the world, including Spain, Peru, Lebanon, America, France, South Korea and Greece, among others. Students at an earlier stage of their academic career could gain valuable experience by presenting their work in a professional setting, while those at a more advanced stage 1

Lucy Blue and Naseem Raad of underwater cultural heritage, as well as the ways it is perceived and presented in museums, in popular culture and across different countries. The range of topics and methods depicted in the symposium and this volume, demonstrates the high degree of specialisation in current day maritime archaeology. As will be seen, the authors utilise archaeometric techniques, geomorphological analyses, and network analysis, among other tools, to explore multifaceted questions. Each paper provides a unique perspective on complex maritime archaeological issues throughout the Mediterranean and beyond, incorporating multiple lines of evidence in a holistic approach. As a number of these authors are still at an early stage in their academic careers, it is the hope that the works presented here will continue to grow and expand into significant research with far-reaching implications. In this regard, the Honor Frost Foundation, in cooperation with the University of Oxford, strives to fuel this growth by helping establish MAGS as a yearly event. By promoting communication not simply across geographic space but across disciplines, it is possible to help early-career researchers build a network of scholars for academic collaboration as well as disseminate publications and results in meaningful ways. By centring the symposium around the general theme of eastern Mediterranean maritime archaeology, it becomes possible to have discussions between divers, architects, archaeobotanists, zoologists and 3D modellers amongst other scholars, from all over the world. Since the pioneering days of Honor Frost, the field of maritime archaeology has expanded and evolved extensively, technology can now get us deeper, we can document faster and more accurately, and are able to digitally visualise and model archaeological finds to an extend that permits greater interpretation and access. Although this humble publication represents a small step in the development of maritime archaeology, it will hopefully spark further discussion and continued engagement between young scholars who, inspired and supported by this symposium, will progress to become the maritime archaeologists of the future.

2

2 Before ‘Thalassocracies’: Reconstructing the ‘Longboat’ and Rethinking its Use and Social Implications in the 4th and 3rd Millennium South Aegean Panos Tzovaras University of Southampton – Southampton Marine and Maritime Institute [email protected] Abstract: In many respects, the Aegean boatbuilding tradition and the conceptualisation of boat types of the 4th and 3rd millennium BC is one of the most controversial subjects of the Aegean prehistory. Until recently, the image of the Aegean watercraft had survived through a quite limited iconographic database, leading many scholars to suggest that the only type existed was the longboat, an advanced form of a logboat. Inevitably, many associated this vessel with a heavyduty cargo boat without considering the limitations of the hull structure. Nevertheless, the last few decades many more iconographic boat evidence have been discovered around the Aegean Basin, such as at Vathy and Strofilas, painting a different picture concerning the Aegean boatbuilding tradition, types of boats and their uses. Having in mind the words of Thucydides regarding piracy, this endeavour exploits the aforementioned discoveries and reassess the corpus of the extant data through an interdisciplinary approach to discern boat types and examine their potential uses. To do so, it utilises aspects of naval architecture by respecting the vessels’ context, in order to reconstruct the so-called longboat with CAD software and test it in simulated conditions to create a performance envelope in which parameters for the boats can be discussed. Key words: Aegean; Neolithic; EBA; boatbuilding tradition; seafaring; naval architecture; reconstruction

΄…οἱ γὰρ Ἕλληνες τὸ πάλαι..καὶ ὅσοι νήσους εἶχον, ἐπειδὴ ἤρξαντο μᾶλλον περαιοῦσθαι ναυσὶν... ἐτράποντο πρὸς λῃστείαν, ἡγουμένων ἀνδρῶν οὐ τῶν ἀδυναωτάτων....οὐκ ἔχοντός πω αἰσχύνην τούτου τοῦ ἔργου, φέροντος δέ τι καὶ δόξης μᾶλλον·’ ―Thucydides Historiae, 1.5.2

evidence, ethnographic parallels and aspects of traditional boatbuilding which may illuminate issues of functionality. Moreover, the ‘longboat’s’ relationship in raiding acts will be examined by implementing ancient sources such as Thucydides (Historiae 1.5.2) who describes the islanders before the mythical king Minos as renowned pirates.

Introduction

Inevitably, a possible reconstruction and testing of the ‘longboat’ could also affect interpretations of the economy, trade and exchange and the nature of the organisation and social structure of maritime communities during the end of the Neolithic and the beginning of the Early Bronze Age (E.B.A.) (ca. 4500–2400/2300 B.C.). Additionally, it would provide a potential answer to the long-standing controversy of which end is the bow. To consider these crucial issues, a novel, interdisciplinary methodological framework of analysis will be proposed through the implementation of 3D CAD modelling software, such as MAXSURF, Autodesk 3ds Max, Rhinoceros 3D and its plugin Orca.

This study aims to address lacunae in research relating to prehistoric boatbuilding traditions and seafaring developed in the Aegean Basin and its adjacent coastland during the 4th and 3rd millennia B.C. Furthermore, it will attempt to bridge gaps in our current knowledge and challenge misconceptions by reassessing previous scholarship and implementing new evidence. The central premise will be to prove that the formation of an advanced boatbuilding tradition in the South Aegean islands was already welldeveloped by the end of the Neolithic, thus much earlier than currently believed. This notion is clearly indicated by the rock-art from the settlements of Strofilas in Andros (Televantou 2018) and Vathy in Astypalaia (Vlachopoulos in preparation). Additionally, by demarcating the chronological and geographical background as well as the context entailing the so-called ‘longboats’, a boatbuilding sequence will be proposed based on the published

Background Due to the island’s biogeography impact, there is a misconception regarding the myth of primitive isolate (Boomert and Bright 2007: 5) and insular island 3

Panos Tzovaras communities burdened by the vast sea (Eriksen 1993; Rainbird 1999; Terrell 2004). These outlooks, reinforced by a Childean conception of the prehistoric communities and the principle of Ex Oriente Lux (Childe 1936; 1942; 1958), afflicted the understanding of the Neolithic Aegean. However, the archaeological record suggests that the sea, whilst a burden, could also act as a connecting bridge (Broodbank 1993; Gosden and Pavlides 1994; Irwin 1980). Thereby, on islandscapes (Broodbank 2000: 21– 22) and especially on clusters of islands, it is common that not only isolation, but interaction may be increased to an extreme level (Broodbank 1993: 316).

chronological spectrum. We can, therefore, refer to complex inter and intra-Aegean spheres of interaction among the so-called ‘small-worlds’ comprised by islandscapes (Broodbank 2000; Tartaron 2013). Nevertheless, there is a wealth of tangible proof, suggesting a surge in seafaring and interconnectivity, consolidated by the use of a seaworthy type of watercraft. Its existence is exemplified by an extensive Aegean corpus of iconographic images of boats on rock engravings and pictorials, decorations on pottery and three-dimensional models and archaeological remains. It is within this context that the so-called ‘longboat’ emerged, leading Renfrew (1972: 170, 358) to postulate on the phenomenon of ‘international spirit’. Hence, some of the specialised Cycladic centres of the KerosSyros horizon emerged; by exploiting their shipbuilding prowess they managed to control and monopolise the local and perhaps the interregional maritime trade (Broodbank 1993: 316). Thus, it was assumed that the pinnacle of the FN/EBA I-II technological achievement, the ‘longboat’, instigated a surge in interconnectivity since it was deemed a vessel connected to commerce.

The Cyclades and the shipbuilding tradition emerging there provide a good example of the aforementioned notion. Seaborne mobility is evident even since the Middle to Upper Pleistocene (ca. 130.000BP) in Stélida in Naxos (Carter et al. 2017) and well-attested during the early Holocene (ca. 11,000 BP) era, considering the obsidian found at the Franchthi Cave, transported from the island of Melos (Perlés 1987) andthe end of Mesolithic (ca. 9000–8000 BP) when human presence is attested to Maroulas in Kythnos (Sampson et al. 2002), etc. This emerging seafaring ideology is evident from the intense colonisation of the Cyclades during the 5th millennium and the emergence of the Saliagos culture, whose traits can be identified beyond the Cycladic ambits (Cherry 1990; Kouka 2008: 211; Zachos 1996: 85). Aside from the pan-Aegean procurement of the Melian obsidian, complex distribution networks of material culture were well-established during the Late Neolithic (LN) period as the circulation of pottery (Tzovaras: forthcoming, a) and Spondylus gaederopus suggests (Halstead 1993: 607–608).

Previous scholarship on the study of the ‘longboat’ Many scholars have ascribed the role of a cargoboat to the ‘longboat’ (i.e. Casson 1971; Erkurt 2011; McGrail 2001; Papadatos and Tomkins 2013); this was a consequence of the absence of another type of watercraft, which, nonetheless, has been identified recently. The new depictions of boats pecked on rocks from the settlement of Strofilas in Andros (Televantou 2017; 2018) and these from the settlement of Vathy in Astypalaia (Vlachopoulos: forthcoming) may paint a very different picture about our knowledge of the FN/EBA I-II shipbuilding tradition and technology. However, before examining the extant data, a concise recount of relevant previous scholarship is necessary.

A proliferation of seafaring and shipbuilding activity is indicated during the transitional Final Neolithic (FN) (4500/4300–3200 B.C.) and the EBA I-II (3200–2400/2300 B.C.) periods (hereafter FN/EBA I-II). Various indirect evidence suggests these ideas, namely island colonisation (Broodbank 1999), the emergence of Aegean metallurgy (Zachos 2007), the presence of non-local products to a culturally-unified region, a pan-Aegean system of codes and symbols and the cargo remains of the Early Helladic (EH) Dokos (Papathanasopoulos 1976) and Giagana (Evaggelistis: forthcoming) shipwrecks. However, in order to grasp the wider implication of seaborne mobility, it is vital to be aware of the concept of the politics of motion (Cresswell 2010). For instance, there are specific agents driving mobility in the first place, entailing cultural, societal, technological, political and economic aspects (Renfrew 1993: 10–11).

The first encounter of the so-called ‘longboat’ occurred in 1899. In his work Kykladika II, Tsountas (1899: 90–92 Fig. 16–22) introduced a few terracotta vessels, the socalled frying pans (Figure 1). These were found at the Chalandriani cemetery in Syros and were decorated with various incised and impressed motifs including spirals, fish, female genitalia and boats. Tsountas described them as long and narrow watercraft propelled by oars, having a low-post with a projection and a high one decorated with a fish emblem and banner. Tsountas also read the tall-post as the stem and the low as the stern based on fish’s direction. This instigated the discussion regarding the early Aegean boats and their architecture. As a full analysis of this discussion is outside the scope of this paper, only a few of the works examining these boats will be mentioned.

As a result, all these internal processes led to major advancements in the economic, societal and political infrastructure paving the way to proto-urbanism, cultural ‘koinae’ and intricate maritime networks in the Aegean basin and its adjacent coastland (Tzovaras: in preparation, a). To some extent, Tartaron’s (2013: 186, Table 6.1) ‘multi-scalar framework’ applies to our geographical and

Sir Arthur Evans (1921; 1923; 1928) agreed with Tsountas’ remarks in relation to the issue of directionality and compared them with a clay model from Palaikastro 4

Before ‘Thalassocracies’

Figure 1. A drawing representation of the boats depicted on the Early Cycladic II “frying pans”. Courtesy professor John Coleman. Source: Coleman 1985: 199, Ill.5.

in Crete (Figure 2), as well as the logboats and log-rafts of the South Pacific. Svoronos (1923: 27) proposed that the bow is the lower end. Moreover, Svoronos suggested that the fish emblem had the role of a sacred wind-vane, also acting as an apotropaic image providing protection and tailwind. Elsewhere, Marinatos identified the highpost as the stern based on the principle of ‘corps en forme de poisons’ (1933: 170–235). Marinatos further consolidated his suggestion by accounting for the presence or absence of a steering oar. He also based his theory in the direction of the ship sign on the Phaistos

Figure 2. A drawing representation of the Early Minoan I-II clay boat model from the site Hellenika, found in ossuary Grave II at Palaikastro. Source: Evans 1923: 240, Fig. 137.

5

Panos Tzovaras Disk in accordance to the orientation of the writing in Minoan scripts. Finally, Marinatos sought a solution for the issue of directionality in the oar’s position that may indicate the start of the stroke. Additionally, Marinatos discerned different types, describing a large vessel built from one trunk where its projection at the bow offered protection during the beaching. Subsequently, it acquired a more angular form and eventually was developed with a circular hull, similar to the boat depicted on a sherd from Phylakopi I (Atkinson et al. 1904: Suppl. IV, Pls. V, c).

stemming from the South Aegean. Consequently, it would be prudent to present some of the available direct evidence (Figure 3), which, in conjunction with the various ethnographic parallels and the essentials of the traditional boatbuilding, yielded a trial digital reconstruction. However, it should be noted that the lead models from an Early Cycladic (EC) II cistgrave from Naxos (Renfrew 1967: Cat. nos. 12–14, Pls. I, 3; Thimme 1977: 546) have been excluded since they have been convincingly characterised as falsifications (Sherratt 2000: 102, 105–106). Additionally, the petroglyphs from the Asphendhou Cave at Sfakia in western Crete (Strasser et al. 2018: 105–106 Figs. 7–8) have been excluded due to their problematic dating. Finally, the painted boats on the walls of the K’noupi Hill rock-shelter in Imbros (Andreou and Andreou 2017: Figs. 53α-β) and the clay boat model of Thermi in Lesvos (Marangou 2001b: 751 Figs. 12–13) are also not included due to the different geographical context, as well as the clay models from Ftelia in Mykonos (Sampson 2002: Pls. 23.3–23.4) and Ayioi Anargyroi from Naxos (Doumas 1977: 18 Fig. 6b) owing to their fragmentary form.

Barnet (1958), described them as seagoing, mastless, paddled ‘long canoes’ with a high stem. Landström (1961: 26–27) argued for a logboat ancestry and for boats with a high, vertical stern-post and planks sewn on the sides. He suggested three stages of evolution; an early one, where the waterline is flat, a more advanced stage with a more curved hull, lifting the bow out of the water and a final one where a projection was added, acting as a cutwater. Casson (1971: 30–35, 41–42) also proposed a logboat origin. Casson argued for a hull with a vertical, narrow stempost as well as a stern, ending in a vertical transom. In dealing with the issue of directionality, Casson adhered to a high-end-as-bow theory mainly based on later examples. Johnston (1982; 1985) suggested that the ‘longboats’ are variations of an extended logboat preceding the fully plank-built boats. Alternatively, Kapitän (1999: 225–231) postulated that these were ‘side-sheltered log-rafts’ similar to the lashed log-rafts kattu-maram or the pegged ones called theppam.

Iconography The case of frying pans The first case study comprises fifteen ‘longboats’ (Figure 1) (for clarity, Coleman’s (1985) numerical order was used), incised on clay frying-pans; the majority of these originate from the Chalandriani cemetery in Syros (Figure 3), dating to the EC II period (Coleman 1985: 199 Ill. 5; Marthari 2017: 149–151 Figs. 4–9; Wedde 2000: Pls. 401–412, 417, 421). All are mastless, possessing an elongated shape, narrow and angled hull with an obtuse high-post, mounted with a fish emblem and a slightly raised low-post with a horizontal projection. In some instances, the ‘keel-line’ is entirely flat (Figure 1, no. 25) and in others asymmetrical (Figure 1, no. 14); occasionally, a steep angle is formed at the lower extremity. There are usually small oblique parallel incisions at the ‘keel-line’ and attached to the gunwale, indicating paddles or oars. However, in some of the examples, these are replaced by small triangles (Figure 1, no. 27). No. 27 also bears zigzag lines along the hull, interpreted as an indicator of additional strengthening, stabilisers or sewn planking (Basch 1987: 86; Broodbank 2013: 328; Landström 1961: 27–28; Wachsmann 1998: 74). In contrast to the previous examples, no. 54 (Figure 1) is crudely-made and has parallel lines along its hull, resembling the horizontal ladder motif of the Mycenaean vessels and possibly representing structural elements, such as stanchions (Wachsmann 1998: 131).

A different approach in determining directionality was implemented by Roberts (1987) and Vichos (1989; 1990), who concluded that the hydrodynamic-aerodynamic properties of the hull would have been enhanced with the tall-post as a stern. Various types of boats were differentiated by Basch (1987) based on the hull’s diagnostic features. Broodbank (2000: 96–106, 256– 258) suggested two canoe-based types, and a potential, partially plank-built boat with a high stern. Based on ethnographic parallels, experimental archaeology and the examination of hull remains he proposed that the smaller type would have been ca. 4–6 m long, manned by 1–4 paddlers, having a cargo capacity of ca. 50–150 kg and an average speed of 5 km/h. The bigger one would have been ca. 15–20 m minimum, manned by a crew of twenty-five, with a capacity of a tonne, having an average speed of 10 km/h. Finally, Wedde (2000) was the first to conceptualise the idea of an Aegean boatbuilding tradition. Wedde proposed that Minoan boats descended from the so-called ‘Syros cluster’ into which the ‘longboat’ belongs, thus implying a linear trajectory of evolution.

The case of Korphi t’Aroniou

Case studies: forming a database

The next example derives from Korphi t’Aroniou in Naxos (Figure 3), a sanctuary on the top of a steep hill. Two marble plaques – probably part of a frieze – were found depicting two scenes with ‘longboats’ (Figure 4),

This research is based on a comprehensive re-analysis of the extant archaeological data relating to boatbuilding and vessel types of the so-called ‘longboat’ type mostly 6

Before ‘Thalassocracies’

Figure 3. The location of the evidence mentioned on a Digital Elevation Model of the Aegean Basin and the adjacent coastland. [MAP]. The map was produced by merging DSM and Bathymetric data (26m. cell size) (ASTER GDEM v2 data). ASTER GDEM is a product of NASA and METI. Data Source: NASA/METI/AIST/Japan Space systems, and U.S./Japan ASTER Science Team. ASTER Global Digital Elevation Model. 2009, distributed by NASA EOSDIS Land Processes DAAC, https://doi.org/10.5067/ASTER/ASTGTM.002. Source: Map generated by the author. Using: ArcGIS for Desktop [GIS]. Version 10.6.

dating to the EC II period (Doumas 1965: 49, 53, Ill. 4, 7). These are morphologically connected to the previous examples but characterised by conscious schematism and substation. The first one has two protrusions amidships which may be identified as paddles or tholes and is laden with a quadruped. In the second plaque, two schematic figures are depicted aboard. The boat is simpler than the first one, with a vertical high-post and a lower one without a projection.

The case of the Ayio Gala Cave in Chios On a pithos sherd found in Ayio Gala (upper cave) dating to the LN period (Figure 3), a fragmentary representation of a vessel was identified (Vichos 1999: 488, Fig. 7) (Figure 5). Along the ‘keel-line’, schematic figures in profile are presented. It possibly represents the earliest example of a boat with human figures on-board, exemplifying the mode of propulsion. 7

Panos Tzovaras

Figure 5. A digital drawing of a boat with schematic human figures paddling (;), on a Late Neolithic pithos sherd found in the Ayio Gala cave in Chios. Drawn by the author by using CorelDRAW 2017, after Vichos 1999: 488, fig. 7.

Figure 4. Top: An Early Cycladic II grey marble fragment with an engraved representation of a boat with two figures onboard. Below: An Early Cycladic II white marble fragment with an engraved representation of a boat laden with a quadruped. Found the sanctuary of Korphi t’Aroniou (now at the Apeiranthos Museum). Drawn by the author by using CorelDRAW 2017, after Doumas 1965: 49, 53 Ills. 4, 7.

either part of the vessel’s decoration (Künze 1934: 25) or unrelated incisions (Renfrew 1972: 357). Three-dimensional evidence: Clay models from Palaikastro in Crete and Ayia Marina in Spetses

The case of Ayios Ioannis

The Palaikastro model (Figure 2), was found in the ossuary Grave II at the site Hellenika (Figure 3) and dates to the Early Minoan II period (Basch 1987: 83, figs. 170–171; Evans 1921: 75; 1923: 240; Wedde 2000: Pl. 103). It has a ‘tear-drop shape’, a pointed high-post and a projection at the wider, lower end; its two beams across its breadth were interpreted as thwarts or cross-bars stiffening the hull (Basch 1987: 83; Dawkins 1903; Johnston 1985: 7; Kapitän 1999: 225–231; McGrail 2001: 107–108; Wachsmann 1998: 71–73).

At Ayios Ioannis’ mines in Siphnos (Figure 3), Weisberger (1985: 109, Abb: 102) (Figure 6) identified a boat pecked on a shaft, analogous to the Korphi t’Aroniou boats. The case of the Dokos wreck An EC II strap of a jug with hastily-made, horizontal strokes was discovered within the cargo of the Dokos wreck (Figure 3). Vichos (1999: 487, Fig. 4a-b) identified it as a ‘longboat’, similar to these on the frying pans and recognised the vertical high-post as the stern. Above the hull, seven vertical lines could be interpreted as possible human figures. Vichos suggested that the Y-shaped object is a horned animal.

An EH II fragment of a boat model was found at the settlement of Ayia Marina in Spetses (Vichos 1999: 484– 486, Fig. 1a-d) (Figure 3) and preserved in a fragmentary condition. The surviving part is probably the point where the hull rises to a tall-post. The rising post above the gunwale, as well as the rest of the hull, is missing. A part of its bottom is flattened with a possible keel-like feature and a pointed extremity. Vichos suggested it is similar to the Palaikastro boat (Figure 2), considering the V-shaped cross-section of the post.

The case of the Orchomenos boat The last pictorial example is a vessel, depicted on a fragment of an askos jug from Orchomenos (Figure 3) dating to the EH II period (Künze 1934: 87; Van de Moortel 2017: 264 Fig. 1b; Wedde 2000: Pl. 415). Whilst analogous to the previous examples (cf. Van de Moortel (2017: 265) who argued that it belongs to a different type, her ‘Type A2’), it has no fish emblem or a tassel on the high-post and the bottom is more curved (Van de Moortel 2017: 264). There are seventeen parallel strokes representing paddles or oars attached to the gunwale. The two parallel, vertical lines above the hull have been interpreted as masts (Vermuelle 1964: 54–56). However, it has been proven that these are

Archaeological record: Boat outlines from Dispilio and Mitrou Although from different geographical and chronological contexts the imprints of four boats will be considered since they constitute the only tangible archaeological remains. Three of these are dated to the beginning of the FN period and they were found in the lake settlement of Dispilio, 8

Before ‘Thalassocracies’ both settlements provide evidence regarding the origins of the Aegean boatbuilding and tradition. During the Summer of 2017, the author conducted research on the iconography of the boats represented on the Vathy rock-art (Tzovaras: in preparation, b). This took place during the project assignment, ‘Justification research of the iconography of prehistoric ships’, within the bounds of the ‘Archaeological field studies and excavations on Vathy, Astypalaia’ programme of the Research Committee of the University of Ioannina and under the supervision of the excavation’s director, Professor Andreas Vlachopoulos. The site’s first habitational phase is attested to the FN period and continued up to a part of the EBA. Although it geographically belongs to the cluster of the Dodecanesian islands (Figure 3), during that period it fell within the ambits of the Cycladic cultural sphere and possibly to the Grotta-Pelos (Angelopoulou and Vlanti: forthcoming; Vlachopooulos 2017: 550–551). Rock-carvings of variously executed boats have been identified on rocks and complexes of boulders (Vlachopoulos: forthcoming) which have been wrongly classified as ‘longboats’ (Van de Moortel 2017: 264). Although some of the badly preserved boats pecked on the ‘Gate of Boats’ belong to a type with a flat/asymmetrical bottom, as well as a high and a lowpost with a projection, for the rest this is not the case. After a personal inspection, it has been identified that they possess a rounder profile and gentler lines; additionally, it appears that both ends have almost the same heightformation, thereby resembling double-ended boats. These traits are exemplified on the middle, better-preserved boat (Figure 7). It has eighteen paddles or oars attached to the gunwale and a steering oar, ending in a triangular blade. On the other side, the post is mounted either with a fish and hanging tassels or a banner. Similarly, the boat of the ‘South Coast’ (Figure 8) is a double-ender, having eight paddles or oars and possibly two steering oars. The last example of a boat with a slightly round bottom is pecked on a slanting rock upon which is supported the western base of the Hellenistic tower; it might belong to the category where the oblique parallels lines are depicted vertically and within the hull, probably indicating the vessel’s crew (Figure 9). Interestingly, both ends terminate in a transomlike manner. However, its identification as a boat is still under consideration.

Figure 6. A digital drawing of the Early Cycladic boat found pecked on a shaft of Ayios Ioannis’ mines. Drawn by the author by using CorelDRAW 2017, after Weisberger 1985: 109.

Kastoria (Hourmouziades 1996; Marangou 2001a: 199, Fig. 17.5; 2003: 16–17, Fig. 4.1a-c) (Figure 3). According t Marangou (2001a: 195; 2003: 14, 18) they could have been made of oak or pine and their dimensions could have been around 3–3.3–3.5 m in length and their maximum breadth 0.73–0.80–1.40 m respectively. Additionally, they possibly represent two types, one with two pointed extremities and another with just one. Interestingly, on the third imprint, close to its pointed extremity, a transversal element can be seen running across the vessel’s breadth. The last example comes from Mitrou, a coastal islet at East Lokris in Pthiotis (Figure 3), dating to the Middle Helladic II period (Van de Moortel 2012: 18, Figs. 3.2– 3.4). The excavator characterised it as an extended, leafshaped logboat, having a flattened-bottom and one gently rising blunt end (Van de Moortel 2012: 17–24; 2017). Interestingly, an exterior round timber was also identified, characterised as a possible bilge-keel. The vessel was made of oak and its original dimensions were estimated as ca. 6 m long, 1.3 m wide and 0.34 m high, reaching 0.50 m with the addition of a 0.16 m potential washtrake which was found nearby. Moreover, the remains of a possible paddle have been discovered. Finally, it was suggested that it could have carried four paddlers and had a cargo capacity of ca. 250 kg.

The FN site of Strofilas in Andros constitutes a unicum in the Aegean Basin (Figure 3); it is one of the most extensive fortified settlements of the period, resembling those of the following period (Televantou 2017: 389). The site’s importance is also reflected in the numerous rock-art representations and the hundreds of depictions of boats. From the first published evidence, it can be inferred that these boats do not belong to the same type as the ‘longboat’ as it has been wrongly suggested (i.e. Broodbank 2013: 327; Papadatos and Tomkins 2013: 355; Renfrew 2011: XL). The main difference is that their profile is clearly crescent-shaped; they also have parallel, vertical lines within the hull representing the paddlers or

New data, different tradition? Rock-art from the sites of Vathy and Strofilas In recent years, the settlements of Vathy in Astypalaia and Strofilas in Andros (Figure 3) yielded crucial information regarding the enigmatic transitional period of the FN/EBA I-II. The rock-art found in both sites – pinpointing the roots of the Aegean iconography – is of utmost importance. Additionally, the numerous depictions of boats found in 9

Panos Tzovaras

Figure 7. An orthomosaic of the “Gate of Boats”. [Detail]. To Scale and georeferenced. Generated by the author with Agisoft PhotoScan Professional. Source: Archaeological Fieldwork at Vathy, Astypalaia – The Archaeological Society of Athens, photogrammetry: Tzovaras, P.

Figure 8. An orthomosaic of the boat in the South Coast. [Detail]. To Scale and georeferenced. Generated by the author with Agisoft PhotoScan Professional. Source: Archaeological Fieldwork at Vathy, Astypalaia – The Archaeological Society of Athens, photogrammetry: Tzovaras, P.

Figure 9. The boat found close to the Hellenistic tower. Courtesy professor Andreas Vlachopoulos. Source: Archaeological Fieldwork at Vathy, Astypalaia – The Archaeological Society of Athens.

oarsmen. One of the best examples is the so-called ‘Large Ship’ (Figure 10). It has a round profile, a high-post and a low-post bearing evidence of a trapezoidal rudder or a steering oar.

Consequently, it can be inferred that two new types of boats are represented at both sites that could have different functions to those of the ‘longboat’. In general, they both clearly belonged to a broad tradition of boats with 10

Before ‘Thalassocracies’ and the sherd from the Dokos wreck (Vichos 1999: Figs. 4a-b), emphasising the depiction of the paddlers/crew and not just the paddles or oars. Finally, apart from the double-enders, a second category is attested to Strofilas where some of the boats have a crescent-shaped hull but dissimilar ends (Figure 10). They possibly resemble the ΕΒΑ boat pecked on a rock from Plaka on Andros (Televantou 2018: 58 Fig. 14b) or the one on a EH II clay disc from Dimini (Adrimi-Sismani, 2013: Table 10:3) (Figure 3). Before reconstruction: Data analysis and interpretation

Figure 10. The Final Neolithic “Large Ship”. Its hull is slightly crescent. On the left extremity, there is a triangular device, possibly a rudder. Copyright Dr Televantou Chr. A. Source: Image kindly provided by Dr Christina A. Televantou (personal archive).

Spotting the similarities Some principal characteristics of the so-called ‘longboats’ can be discerned based on the previously presented data. They are all mastless and their basic shape is that of a long, slender and angled hull. Their bottom is sometimes asymmetrical or flat with a slight drooping of the foot close to the lower end. In all cases, there is an extremity ending in a high-post, usually rising to an angle of ca. 70 degree and to a height of almost 1/3 of the hull’s total length, forming an almost obtuse angle with the keel (Johnston 1982: 363; Johnstone 1973: 9; McGrail 2001: 109; Morrison 1995: 136; Roberts 1987: 309; Wedde 2000: 50, 95). Occasionally, this high-post is mounted with a fish emblem with hanging tassels or a banner with chevrons ending in fringes. Their lower-post usually rises at an angle of 22 degrees to the horizontal and it ends in a horizontal projection at the keel-level (Vichos 1989: 21; Morrison 1995: 136; Wachsmann 1998: 71; Wedde 2000: 145; McGrail 2001: 109), to a size of almost 1/10 of the hull’s total length.

a crescent or nearly-crescent hull. The first category, the double-enders, includes most of the Vathy’s boats (Figures 7–9) and some of the Strofilas’ (Televantou 2018: Fig. 12). To this category we may add the boat models from Mochlos (Seager 1909: 279 Fig. 2; 1912: Fig. 48.31; Soles, 2012: 194 Figs. 21.8–21.11; 195 Fig. 21.12) (Figures 3, 11–12), from Tholos Tomb X at Khristos in Messara (Xanthoudides 1924: Pls. XL, b) and Kephala Petras in Sitia (Papadatos 2012: 157, Fig. 17.2) or the Phylakopi I sherd (Atkinson et al. 1904: Suppl. IV, Pls. V, c) (Figure 3). To a similar boatbuilding tradition, we could add a boat pecked on a rock at Kastri on Syros (Marthari, 2017: 158 Fig. 19) (Figure 3). The latter probably belongs to the same iconographic tradition as the boats from Strofilas (Figure 10), Ayio Gala (Figure 5), Orchomenos (Wedde 2000: Pl. 415), the boat found close to the tower of Vathy (Figure 9)

The stem-stern dispute: a hull’s shape examination The major problem regarding these boats is how to determine which end is the bow and which the stern. Whilst this controversy persists, it is widely accepted that the high-post represents the stern and the low represents the bow (Basch 1987: 85–88; Broodbank 2000: 99; 2013: 327; Doumas 1965; Gray 1974: 76–81; Johnston 1982;

Figure 11. An Early Minoan II double-ended (;) clay boat model found at the House’s A deposits in Mochlos, Crete. Source: Seager 1909: 279, Fig. 2.

Figure 12. An Early Minoan I-II double-ended (;) clay boat model (or horns of consecration) from the cemetery in Mochlos. Source: Seager 1912: Fig. 48 (31).

11

Panos Tzovaras 1985, Landström 1961: 26–27; Marinatos 1933; Roberts 1987; Svoronos 1923:27; Vichos 1989; 1990; Wedde 2000: 50). The adherents of the opposing theory mainly based their assumptions on the fact that the fish emblem points to the direction of travel (Casson 1971; Evans 1928: 240–241; Hutchinson 1962: 91–93; Tsountas 1899: 91). Some scholars also suggest that the high-prow theory is consolidated by two evidences: first, the square, transomlike lower end of the questionable lead boat models (Casson 1971; Renfrew 1967: 5; Wachsmann 1998: 73); second, the horizontal projections at their bows of the boats from the Akrotiri miniature frieze (Casson 1975: 4–9; Wachsmann 1998: 73). Additionally, it has been suggested that the horizontal projection is a fixed rudder, thus indicating the stern (Barnett 1958: 224; Evans 1921: 75; 1923: 240; Glotz 1923: 218–220; Hall 1928: 34–36; Hood 1967: 46).

Figure 13. Hodges W. 1777: The war boats of the island of Otaheite (Tahiti) and the Society Isles, with a view of part of the harbour of Ohaneneno, in the island of Ulieta, one of the Society Islands (BHC2374). Low-res. [Oil on canvas]. London: Royal Museums Greenwich. Courtesy © National Maritime Museum, Greenwich, London, Ministry of Defence Art Collection. Source: http://images.rmg.co.uk/en/ search/do_quick_search.html?q=The+war+boats+of+the+isl and+of+Otaheite+%28Tahiti%29+and+the+Society+Isles.

Nevertheless, the high-stern theory is more convincing, and it has been suggested that it explains the watercraft’s peculiar shape (Figure 1). From an aerodynamic perspective, the high-post would have been of light construction and wide, allowing the manipulation of the Aegean Northerly winds, therefore acting as a form of sail (Roberts 1987: 309–311; Vichos 1989: 21–23). Thereby, by exploiting the lift and drag forces of the wind, the steersman would have been able to steer at various angles from the wind, allowing the boat to come about. Otherwise, the boat would have turned downwind (Roberts 1987: 309; Vichos 1989: 22). Additionally, bows tend to be beamier than the stern since the speed is affected by turbulence at the stern, a fact consolidated by the Palaikastro model (Gardner 1984: 62; Johnston 1985: 8; Marinatos 1933: 182–184) (Figure 2). Furthermore, the higher position of the steersman not only provided them with better protection, but also meant that they could see beyond paddlers’ heads (Morrison 1972: 230–33; Svoronos 1923: 27). The rock-art from Korphi t’Aroniou depicting a man driving a goat into a boat moored astern consolidates the low-stem theory too (Doumas 1965: 49; McGrail 2001: 110) (Figure 4).

(Wylie 1985; Blue 2003). Hence, the Maori’s Waka-Taua paddlers (Figure 13), always face the bow when paddling (Tregear 1904: 122) and the equivalent Tahitian logboats have a high stern, consequently securing protection aft since they travel on a following sea (Doumas 1970: 285– 90). Estimating ‘longboat’s’ longitudinal-transverse dimensions There is a consensus that the parallel strokes represent paddles (Barnet 1958; Basch 1987: 84; Broodbank 2013: 327; Marangou 1990: 260–261; McGrail 2001: 109; Wachsmann 1980: 288; 1998: 71; Wedde 2000: 66–67), as these boats would have been too narrow for oars (Wachsmann 1998). This observation is reinforced by the findings from Mitrou, Dispilio and the sherd found at Ayio Gala (Figure 5), as well as the notion that most of the prehistoric logboats would have been paddled (McGrail 1987: 207). Although the boat depictions may not constitute a physically accurate watercraft representation, the number of the paddles is probably the only available way to determine the vessel’s longitudinal dimension and the mid-ship’s beam. However, we should always acknowledge the fact that the depiction of paddles is very problematic since their angle and number is always unstable (McGrail 2001: 110).

By being placed at the bow, the horizontal projection acquires hydrodynamic qualities as a cut-water (Landström 1961: 26–27; Marinatos 1933; Morrison 1995: 136). Moreover, as a continuation of the main log placed at the level of the waterline, the longitudinal stability would have been increased as well as the dynamic one and also the vessel’s performance. It also could have protected the stem during beaching (Basch 1987: 85–88; Doumas 1965; McGrail 2001: 108). Thus, it could also act as a bulbous bow which would have enhanced fore buoyancy and reduced the vessel’s resistance and pitching (Taylor 1980: 95). The V-shaped hull has a complementary role and would have maximised all the aforementioned qualities (Vichos 1989: 22). This projection offered another solution. The shape and weight of the tall-post may have weighted the vessel down. This force could have been balanced by adding a horizontal, wide extremity at the waterline (Vichos 1989: 21–22). Finally, the high-post as a stern can be further underpinned by implementing the principle of analogy

We can assume the vessel’s length by defining the paddlers’ interscalmium, the working space of a paddler, which constitutes a hypothetical cube (Wedde 2000: 102). Broodbank (1989: 329) and McGrail (1987: 206) suggested a 1 m distance, similar to the Pacific canoe paddlers (Best 1925). However, because this distance complies with the Vitruvian interscalmium (De Architectura I.2.4) intended for oars, the modified length of 0.75 m proposed by Marinatos (1974: 151; see also, Wedde 2000: 102) will 12

Before ‘Thalassocracies’ be chosen in an attempt to assume the length. Another parameter to be considered is the belief that the eastern Mediterranean wooden sea-going vessels are not believed to exceed 35 m (Wedde 2000: 107). By multiplying the number of paddlers of the aforementioned examples with the modified interscalmium (plus the percentage of the stem/stern-post and the horizontal projection), we attain a length varying between ca. 6–35 m. These estimations agree with those generally proposed by scholars (Broodbank 2000: 103, Table 3; 2013: 327; Johnston 1982: 363–364; 1985: 3–9; Marinatos 1933; Renfrew 1972: 357; Wedde 2000: 107).

259) coupled with references to ancient sources (Homer Il.; Od.; Plato De Leg. 1.705; Theophrastus His. Plan. 5.7, 1–3; Vegetius Ep. Rei Mil. 4.34), the optimal timber species for shipbuilding in general are fir, pine, cedar, oak, poplar, elm, chestnut, beech, alder, olive and cypress. However, the South Aegean belongs to the true Mediterranean vegetation zone with an annual rainfall below 400 mm. (Renfrew 1972: 267–268; Treuil et al. 1989: 98–99). Additionally, pollen analytical investigations suggested drier climatic conditions in the Aegean than today, thus thwarting woodland expansion (Bottema 1982). This, in conjunction with the low altitudes of most of the Aegean islands, resulted in mainly medium-sized trees (Treuil et al. 1989: 98–99). This is consolidated by the findings from Vathy in Astypalaia, (Ntinou: forthcoming). However, we do know that trees such as cypress, oak, fir, pine and possibly cedar were naturally growing in Crete (Bintliff 1977: 73–74; Hall 1928: 38; Meiggs 1982: 88–115). Some of these species, along with elm, thrived in Thera, Naxos, Samos, Kea, Lesbos and other Aegean islands (Asouti 2003: 473; Bintliff 1977: 72–73; Ntinou pers. comm.; Vlachopoulos pers. comm.).

The next step is to estimate the ‘longboat’s’ maximum breadth and height. We are working on the premise that the minimum space for paddlers paddling side by side is 0.85 m (McGrail 1987: 207). Nevertheless, it is almost impossible that they had a maximum breadth of ca. 1 m because of their elongated, lanceolate shape. Thus, the breadth would have decreased drastically close to the bow and stern, leaving no space for two paddlers abreast. If we consider that their beam to length ratio was around 1:10 (Wachsmann 1998: 74) or 1:12 (Marangou 1990: 260–261), then their beam would have been close to 2–3 m (Wedde 2000: 109). Intriguingly, the same ratio is attested to the similarly-shaped Chinese dragonboats usually manned by thirty-six or more paddlers (Johnstone 1980: 193; Wachsmann 1998: 74; Wedde 2000: 108). Likewise, McGrail (2001: 111) suggested a beam of ca. 1.8 m based on the Palaikastro model (Figure 2). Finally, the Mitrou and Dispilio boats are close to 1.5 m and the Maori’s similar Waka-Taua approaches 2 m (Haddon and Hornell 1936: 200–201).

In Antiquity, a potentially good source of timber could have been the island of Andros where flora possibly flourished in forests of oak, planes, eucalyptus, chestnut, alder and pine (Televantou personal communication). Another candidate is the island of Samos, renowned for its woodlands, especially pine trees, since Antiquity (Dimitriadis and Landros 2014). Interestingly, ethnographic studies in the Aegean suggest that all the shipbuilders preferred the Samian Pinus brutia and nigra (Damianidis 2013: 174). Hence, these larger trees, such as Pinus nigra possibly found in Vathy (Ntinou: forthcoming), Pinus Brutia, Quercus deciduous, found at Akrotiri in Thera (Mavromati 2016: 8, Table 1), or the Cretan cypress, were available and probably exported when needed. The trade of timber is consolidated by recent analyses of wood-charcoal from Dhaskalio, indicating the presence of exogenous species such as cypress, ash, poplar and black pine (Ntinou 2013: 419), as well in Vathy (Ntinou: in preparation). This evidence verifies the construction of 20–35 m ‘longboats’ from a single log. Of course, the local timber would have been sufficient for shorter boats.

Regarding their height, its calculation is imponderable; for instance, it is subject to the type of tree and trunk’s circumference and whether washtrakes are present. However, based on the premise of a low gunwale (Marangou 1990), the dimensions of the Mitrou boat, together with McGrail’s (2001: 111) and Broodbank’s (2000: 208) estimations, it can be suggested that it would have ranged between ca. 0.50–1.15 m. similar to the Neolithic Bracciano logboat (Fugazzola Delpino and Mineo 1995). Their shallow depth in conjunction with an average height of ca. 1.67 m (±0.20) of a male in the EBA Aegean (Dickinson 1994: 138) supports paddling from a kneeling position; additionally, this way of paddling increases stability (McGrail 1987: 206) and is exemplified by a Middle Minoan seal found in Knossos (Evans 1936: 521 Fig. 463).

Determining the construction method The aforementioned data allows us to assert that the ‘longboats’ were probably partially plank-built, extended and expanded logboats, perhaps with a composite, transom-like stern and stem, that would have increased the transverse stability at waterline and freeboard (Johnston 1982: 364; 1985: 6–7; Johnstone 1980: 49; Marangou 2001b: 743; McGrail 1987: 66). Besides, the requisite tools for the construction of a logboat were available since the Mesolithic period and its expansion-extension possible since the FN-EBA as the Italian, Neolithic Bracciano logboat indicates (Johnstone 1973: 4; Marangou 2001b: 741–742; McGrail 1987: 64, 86). Similarly, the Maoris built their 40 m long war-canoes by using stone-shell tools

Where did the timber come from? The most problematic aspect is timber availability because the islanders would have most probably been depended on the local flora (Casson 1971: 213; Renfrew 1972: 356). Based on ethnographic and archaeological evidence (Evans 1925; Hall 1928: 38; Hutchinson 1962: 91; Johnstone 1973: 4; Landström 1961:26–27; McGrail 1987: 59–60, Table 6.1; Meiggs 1982: 100, 111, 116–120; Steffy 1994: 256– 13

Panos Tzovaras and fire (Johnstone 1980: 206–207). Additionally, the expansion would have been a necessity to accommodate two paddlers working abreast. An indicator of expansion is the Palaikastro model (Figure 2), with two beams along the breadth of the boat, reinforcing the hull’s strength too (Kapitän 1999: 225–231; McGrail 2001: 108). However, when a log is expanded, its sides usually collapse; therefore, the freeboard is decreased (McGrail 1987: 58). This can be resolved by adding washtrakes (Johnstone 1980: 49; McGrail 1987: 58), as has been attested by the Mitrou’s boat.

like stern-post, added at the end and joined with lashes (Johnstone 1980: 208) (Figure 13). A new methodological framework and a pilot All of the aforementioned conclusions are based on hypothetical assumptions. The inability to test and validate the assumptions inferred is a problem attested to all of the previous endeavours. However, their digital reconstruction can be made possible through software like MAXSURF, Rhinoceros and Orca 3D used in naval architecture. Hence, MAXSURF Modeler and Rhinoceros 3D allowed the design of a 3D CAD surface model in real dimensions. Additionally, the Rhinoceros’ plug-in, Orca, allowed for the addition of material properties to the 3D models for the calculation of the waterline. Finally, MAXSURF Stability offered the opportunity to test the seafaring properties of the watercraft in simulated conditions with respect to the hydrostatics, stability and displacement, as well as to explore the issue of directionality.

Although it is possible that there was not any internal structure reinforcing the rigidity of the hull, the ladder motif represented in one of the vessels (Figure 1, no. 56) of the frying pans and the possible thwarts or crossbeams of the Palaikastro model (Figure 2) do not allow us to be absolute. Analogous, internal transverse partitions (possibly having the role of bulkheads) have been identified on Neolithic examples such as on the imprint of the Dispilio boat, the almost intact Bracciano logboat in Italy (Fugazzola Delpino and Mineo 1995)), a logboat found on the south bank of Ladoga lake near the village Korovkino in Russia (Chepelev 1998: 41), the LN Tsangli clay boat model (Basch 2004: 109 Fig 7.7), the AF346 boat model from the northern Slope of the Acropolis (Basch 2004: 104 Fig. 7.1) and perhaps on an enigmatic Middle Neolithic clay model from Knossos (Evans 1921: 39 Fig. 11) (Figure 3). Hence, the abovementioned transverse features that could act as an internal structure in conjunction with the keel-like features attested to the bottom of the Tsangli boat (Marangou, 2001b: 744), the Neolithic AF393 model from Acropolis (Basch 2004: 105–107 Fig.7.3), the EBA boat from Thermi in Lesvos (Marangou 2001b: 742) and the boat from Spetses (Vichos 1999: Figs. 1a-d ) would have reinforced the vessel’s longitudinal-transversal strength.

A first pilot By combining all the data analysed earlier in MAXSURF Modeler and Rhino 3D, an attempt at a pilot reconstruction of two ca. 34 m logboats was initiated. One reconstruction was conducted with a high-bow and another with a highstern so as to approximately estimate their deadweight, capacity and seafaring properties (Figure 14). Based on the aforementioned analysis, the mid-ship section is estimated at ca. 26 m long, its breadth at ca. 2 m and its height at ca. 1.15 m along with a washtrake. The crew consists of thirty-seven men and the vessel’s volume was estimated at ca. 20 m3. Additionally, by implementing Orca 3D, the vessel has been assigned with material properties of a pine tree native to the Aegean. This yielded a displacement of ca. 7–8 t. Its hydrostatics and stability properties have been tested with the software MAXSURF Stability. Although both reconstructions successfully passed the floatation test, the stability results indicate that the vessel with the high-end as the stern is better to that with the high-end as the bow (Figure 15). However, they both had stability issues (Figure 15). Finally, it is evident from the models produced (Figure 14) that this boat type is impossible to carry cargo – beyond the crew’s personal belongings and supplies – owing to a lack of space, poor stability, as well as the extremely low freeboard resulting in bilge water.

Further answers to boat construction can be sought in an analysis of Neolithic housebuilding carpentry and technology. An astonishing finding that could change our understanding regarding an advanced Neolithic boatbuilding tradition and technology of the Aegean are the planks assembled with pegs attested to a floor of an Early Neolithic building in Prodromos (Demoule and Perlès 1993: 375; Hourmouziades 1971: 164–175) and in Nea Makri (Pantelidou-Gofa 1996: 71). Additionally, the use of shaped planks alone is indicated at the LN Servia (Rhomiopoulou and Ridley 1973; Ridley and Wardle 1979), Saliagos (Coleman 1974: 334) and possibly in Dispilio (Marangou 2001a: 195–196). Hence, washtrakes could have been pegged on the boat with the use of the mortise-and-tenon technique or lashed if we consider the zigzag lines on some of the frying-pan ‘longboats’ (Basch 1987: 86) (Figure 1, no. 27). However, it should be noted that the zigzag lines constitute one of the prehistoric Aegean art conventions in representing the sea (Marinatos 1930: 124). Finally, the vessel’s ends could have been linked with transoms (Casson 1971; McGrail 1987: 74) and the stern’s high extremity could have had that role. The Maori’s Waka-Taua had a similar high, transom-

Discussion Although a pilot, this hypothetical reconstruction allows us to posit that the use of the ‘longboats’ with their highstern – at least for the larger ones was not related to commerce due to their inability to carry more than the crew and their belongings. The digital CAD model produced patently demonstrates the lack of room within the boat (Figure 14). Besides, a cargo-boat, the ‘maid-of-all-work’ is characterised by ample space in order to distribute cargo safely and speedily (Taylor 1980: 1–2; Tupper and Rawson 2001: 3). Additionally, if a boat is not designed 14

Before ‘Thalassocracies’

Figure 14. The 3D reconstruction of the ‘longboat’ with a high-end as stern and its lines plan. The model was generated by using the MAXSURF Modeler and Rhinoceros 6 software. The material properties were assigned by using the plugin Orca 3D. The paddlers, steersmen, lookout, crossbeams, fish emblem and cargo were modelled through the software Autodesk 3ds MAX (Credits: The base model for the lookout was produced by ‘avenirdesign’ and downloaded at www. sketchfab. com.). The stability assessment at the bottom examining the behaviour of a ship at 0.8 m. trochoidal waveform was conducted by using the software MAXSURF Stability. Source: The 3D models, the lines plan through and the stability analysis were generated by the author.Source: http://images.rmg.co.uk/en/search/do_quick_search.html?q=The+war+boats+of+the+island+ of+Otaheite+%28Tahiti%29+and+the+Society+Isles.

have been detrimental to the vessel’s transverse stability, which is inherently poor (Figure 15). These aspects would have rendered the vessel incapable of traversing a roughsea.

to transport cargo, it might be affected by the hogging and sagging effect, caused by the improper distribution of the weight. This effect would have been deteriorated by a potential absence of a keel Thereby, it would have been impossible for a ‘longboat’ to transfer vast amounts of Naxian marble to the settlement of Dhaskalio, where it was used as building material (Boyd 2013). Additionally, from a technical perspective, its presumably flat bottom, the meagre freeboard and ensuing bilge water (Figure 14) would have caused the flat-surface effect which would

Nevertheless, the new data emerging from the rock-art of Strofilas and Vathy and the older ones (i.e. the boat from Tsangli, etc.) possibly indicate a potentially more spacious, seagoing type of vessel that would have undertaken the role of Herodotus’ νῆες στρογγύλαι (Historiae, 1.163) or the 15

Panos Tzovaras

Figure 15. A comparative stability assessment between the ‘longboats’ with a high-end as the bow and the high-end as the stern. The graphs indicate that in the boat with the high stern the maximum heeling moment that the ship can sustain without capsizing is better on Test 3 than 2. Additionally, it has a larger range of stability and righting energy, hence it is able to heel to a larger angle before attaining negative stability. Source: Generated by the author by using the software MAXSURF Stability.

1989: 330–331; Wachsmann 1998: 76). This means that the most productive part of a community would have been absent and unable to contribute for a considerable amount of time if we consider the man-days required to construct the boat and that a substantial proportion of the settlement’s population would have been devoted in aiding them. The same would have happened with the manning of a 34 m ‘longboat’. This would have required at least forty of the most productive individuals, risking death at sea. Consequently, it would have a negative impact on the settlement’s population and a shift to its dynamics. Hence, an ensuing demographic, social and economic breakdown is risked (Broodbank 1989: 330). Therefore, the technical issues discussed, the labour-intensiveness and the obviously high ‘risk-to-benefit ratio’ suggest that

Homeric ἀμφιέλισσα (Il. 2.165). The word ‘ἀμφιέλισσα’ means a boat curved at both ends, or a double-ender, and ‘νῆες στρογγύλαι’ have been interpreted as ‘round-built merchant-ships’ (Liddell et al. 2011). Therefore, both categories are potentially exemplified by the rock-art of both sites. It is viable that the shipwrecks of Dokos and Giagana could have belonged to that type. The construction of a ‘longboat’ would have been a communal effort (Broodbank 2000: 256; McGrail 1987: 64) and therefore required a certain degree of organisation and abilities (Broodbank 1989: 330–332; Vermuele 1964), implying specialisation. Moreover, it has been calculated that such an activity would have bound a good number of the male population of an EBA II settlement (Broodbank 16

Before ‘Thalassocracies’ their interpretation as cargo-boats is invalid and opens the question as to what the ‘longboat’ was used for if it was not for trading.

crucial insights that allowed the realisation of this analysis. I am also indebted to Dr Christina Televantou and Dr Maria Ntinou for their valuable help in issues of Aegean shipbuilding and timber sources. I would also like to thank Dr Lucy Blue and Dr Julian Whitewright for their helpful comments and feedback. Finally, many thanks are due to the reviewers for their valuable comments and for making the reviewing process enjoyable.

The answer perhaps lies in Thucydides’ words (Historiae 1.5.2); these referred to piracy during an old age before Minos, where raiding was part of islanders’ every-day life as they searched for glory and acknowledgement of their valour from their peers. Thereby, by venturing into new lands, fetching exotic goods and recounting their triumphs, they satisfied their need for virtus et honor. Their virtus was represented by the same ‘longboat’, whose prowess to harness the Aegean Sea with an imposing appearance, resembling a huge sea-monster, would have brought terror to their ‘prey’. Additionally, the strong and intricate fortification systems demonstrated during the FN/EBA I-II (e.g. at Strofilas, Zagani, Kastri, Markiani, Panormos, etc.) point in that direction. Additionally, any artefact can acquire both technical and non-technical functions, based on the user’s needs (Crilly 2010). It is viable that, like Maori’s Waka-Taua (Tregear 1904: 119–122) (Figure 13), the monumentality of the ‘longboat’s’ construction would have taken a ritual form, where the special class of boatbuilders/seafarers would have been the focal point. In that way, one of the most important products of material reality, the ship, acquired a symbolic-ceremonial function too and was transformed into a powerful image. This is clearly attested to the abundance of the boat representations in various forms of art, very early in prehistory that possibly highlights aspects of social hierarchy and a hitherto unknown high prestige class of seafarers.

Works Cited Primary Sources HERODOTUS. The Persian Wars, Volume I: Books 1–2, trans. Godley A. D. 1920. Loeb Classical Library 117. Harvard University Press: Cambridge, MA. HOMER. Ilias, trans. Murray, A. T. Revised by Wyatt, W. F. 1924. Loeb Classical Library 170. Harvard University Press: Cambridge, MA. HOMER. Odyssey, trans. Murray, A. T. Revised by Dimock. G. E. 1919. Loeb Classical Library 104. Harvard University Press: Cambridge, MA. PLATO. Laws, Volume I: Books 1–6, trans. Bury, R. G. 1926. Loeb Classical Library 187. Harvard University Press: Cambridge, MA. THEOPHRASTUS. Historia Plantarum, trans. Hort, A. 1911. Loeb Classical Library 70. Harvard University Press: Cambridge, MA. THUCYDIDES. Historiae, trans. Smith, C. F. 1919. Loeb Classical Library 108. Harvard University Press: Cambridge, MA. VEGETIUS RENATUS FLAVIUS. Epitoma Rei Militaris, Reeve, M. D. (ed.) 2004. Oxford Medieval Texts: Oxford.

Conclusion Undeniably, the construction of a ‘longboat’ would have been technologically plausible. It was probably a watercraft that was not intended to conduct intra or interregional exchange or trade. Its use would have transcended the needs of every-day life and perhaps resembled a longship, ideal for raiding. Equally, by using their vigour and the speed of their ships as a springboard, this class of seafarers possibly managed to form the Homeric warrior/heroic ideal where the role of their boats transcended the quotidian and acquired an increased ideological significance. On the other hand, the assumed cargo-boats of Vathy and Strofilas represent the ‘engine’ that drove an Aegean extroversion, a cog in the wheel of their forthcoming prosperity and superiority that lasted for many millennia. The intricacy of this subject requires a methodological framework that exploits the tools offered by various fields of research. By utilising an interdisciplinary approach without disrespecting the aspects of traditional research, it was possible to test a range of possibilities of this watercraft and determine realistic aspirations of FN/EBA I-II boat construction.

VITRUVIUS. De Architectura, trans. Granger, F. 1931. Loeb Classical Library 251. C Harvard University Press: Cambridge, MA. Secondary Sources ADRIMI-SISMANI, V. 2013. Ο μυκηναϊκός οικισμός Διμηνίου: 1977–1997. 20 χρόνια ανασκαφών. Volos: Υπουργείο Παιδείας και Θρησκευμάτων, Πολιτισμού και Αθλητισμού, Γενική Γραμματεία Πολιτισμού, Αρχαιολογικό Ινστιτούτο Θεσσαλικών Σπουδών. ANDREOU, I. and ANDREOU, I. 2017. Ίμβρος: Ένα μικρό νησί με μεγάλη ιστορία. Patakis Publication: Athens. ANGELOPOULOU, A. and VLANTI, D., In preparation. ‘Η κεραμική των προϊστορικών φάσεων και των ιστορικών χρόνων’. Paper presented at: Βαθύ Αστυπάλαιας. Πέντε χρόνια συστηματικής έρευνας σε ένα διαχρονικό παλίμψηστο του Αιγίου. Επιστημονική ημερίδα, Σάββατο, 14 Ιουλίου 2018. Ροδιά Χώρας.

Acknowledgements

ASOUTI, E. 2003. ‘Wood charcoal from Santorini (Thera): New evidence for climate, vegetation and timber imports in the Aegean Bronze Age’. Antiquity, 77: 471–484.

I would like to express my gratitude to Professor Andreas Vlachopoulos for allowing me to record and study the important rock-art material from the Vathy’s excavation, as well as for his

17

Panos Tzovaras ATKINSON, T. D., BOSANQUET, R. C., and EDGAR, C. C. 1904. Excavations at Phylakopi in Melos: conducted by the British School at Athens. Supplementary paper No. 4. Published by the Council and sold on their behalf by Macmillan and Co., Limited: London.

BROODBANK, C. 2013. The making of the Middle Sea. A history of the Mediterranean from the beginning to the emergence of the Classical world. Thames and Hudson: London. CASSON, L. 1971. Ships and seamanship in the ancient world. Princeton University Press: Princeton.

BARNETT, R. D. 1958. ‘Early shipping in the Near East’. Antiquity 32: 220–230.

CASSON, L. 1975. ‘Bronze Age ships. The evidence of the Thera wall paintings’. International Journal of Nautical Archaeology, 4: 3–10.

BASCH, L. 1987. Le musée imaginaire de la marine antique. Athens: Institut Hellénique pour la Préservation de la Tradition Nautique.

CHEPELEV, V. R. 1998. ‘Finds of ancient dugouts in the USSR and Cossack dugouts « oaks »’. Archaeonautica, 14: 41–55.

BASCH, L. 2004. ‘Two Athenian ship models of the third millennium B.C’. In: F. M. Hocker and C. A. Ward (eds.), The Philosophy of Shipbuilding: Conceptual Approaches to the Study of Wooden Ship. Texas: Texas A&M University Press: College Station. Pp. 103–111.

CHERRY, J. F. 1990. ‘The first colonisation of the Mediterranean islands: a review of recent research’. Journal of Mediterranean Archaeology, 3: 145–221.

BEST, E. 1925. The Maori canoe. Dominion Museum Bulletin 7. W. A. G. Skinner: Wellington.

CHILDE, V. G 1936. Man makes himself. Watts & Co: London.

BINTLIFF, J. L. 1977. Natural environment and human settlement in prehistoric Greece, based on original fieldwork. International Series 28. BAR Publishing: Oxford.

CHILDE, V. G. 1942. What happened in history. Penguin: Harmondsworth. CHILDE, V. G. 1958. ‘The relations between Greece and prehistoric Europe’. Acta congressus Madvigiani, Hafniae 1954 / Proceedings of the Second International Congress of Classical Studies / Actes du Deuxième Congrès International des Etudes Classiques. General Part. Munksgaard: Copenhagen.

BLUE, L. 2003. ‘Maritime Ethnography: The reality of analogy’. In: C. Beltrame (ed.), Boats, ships and shipyards: proceedings of the Ninth International Symposium on Boat and Ship Archaeology, Venice 2000. Oxbow Books: Oxford. Pp. 334–338.

COLEMAN, J. E. 1974. ‘The Chronology and interconnections of the Cycladic islands in the Neolithic period and the Early Bronze Age’. American Journal of Archaeology, 78.4: 333–344.

BOOMERT, A. and BRIGHT, A. J. 2007. ‘Island Archaeology: in search of a new horizon’. Island Studies Journal, 2.1: 3–26. BOTTEMA, S. 1982. ‘Palynological investigations in Greece with special reference to pollen as an indicator of human activity’. Palaeohistoria, 24: 257–289.

COLEMAN, J. E. 1985. ‘Frying pans’ of the Early Bronze Age Aegean’. American Journal of Archaeology, 89: 191–219.

BOYD, M. J. 2013. ‘The Structure and Architecture of the Settlement’. In: C. Renfrew, O. Philaniotou, N. Brodie, G. Gavalas, and M. J. Boyd (eds.), The settlement at Dhaskalio. The sanctuary on Keros and the origins of Aegean ritual practice: the excavations of 2006–2008. Volume I. McDonald Institute for Archaeological Research: Cambridge. Pp. 341–385.

CRESSWELL, T. 2010. ‘Towards a politics of mobility’. Environment and Planning D: Society and Space, 28.1: 17–31. CRILLY, N. 2010. ‘The roles that artefacts play: technical, social and aesthetic functions’. Design Studies, 31.4: 311–344.

BROODBANK, C. 1989. ‘The longboat and society in the Cyclades in the Keros-Syros Culture’. American Journal of Archaeology, 93.3: 319–337.

DAMIANIDIS, K. A. 2013. Ελληνική παραδοσιακή ναυπηγική. 3rd ed. Piraeus Bank Group Cultural Foundation: Athens.

BROODBANK, C. 1993. ‘Ulysses without sails: trade, distance, knowledge and power in the Early Cyclades’. World Archaeology, 24.3: 315–331.

DAWKINS, R. M. 1903. ‘Excavations at Palaikastro. II: § 8. —The Pottery’. The Annual of the British School at Athens, 9: 297–328.

BROODBANK, C. 1999. ‘Colonization and configuration in the insular Neolithic of the Aegean’. In: P. Halstead (ed.), Neolithic society in Greece. Sheffield Academic Press: Sheffield. Pp. 15–41.

DEMOULE, J. P. and PERLÈS, C. 1993. ‘The Greek Neolithic: A New Review.’ Journal of World Prehistory, 7.4: 355–416. DICKINSON, O. 1994. The Aegean Bronze Age. Cambridge University Press: Cambridge.

BROODBANK, C. 2000. An Island Archaeology of the Early Cyclades. Cambridge University Press: Cambridge.

DIMITRIADIS, V. and LANDROS, C. (eds.) 2014. Tο Σαμιώτικο πεύκο και η παραδοσιακή υλοτομία στη Σάμο. Samos: Πνευματικό Ίδρυμα Σάμου «Ν. Δημητρίου». 18

Before ‘Thalassocracies’ DOUMAS, C. 1977. Early Bronze Age Burial Habits in the Cyclades. Paul Äström: Göteborg.

Arawe islands, Papua New Guinea’. Archaeology in Oceania, 29: 162–171.

DOUMAS, C. 1970. ‘Remarques sur la forme du bateau egeen a l’age du Bronze Ancien’. In E. Anati (ed.), Valcamonica symposium. Actes du Symposium International d’ Art Prehistorique. Capo di Ponte: Edizioni del Centro. Pp. 285–90.Doumas, C. 1965. ‘Κορφή τ’ Αρωνιού: μικρά ανασκαφική έρευνα εν Νάξω’. Αρχαιολογικόν Δελτίον, 20.Α: 41–64.

GRAY, D. 1974. Seewesen. Archaeologia Homerica, Band I Kapital G. Vandenhoeck & Ruprecht: Gottingen. HADDON, A. C. and HORNELL, J. 1936. The Canoes of Oceania. Bernice P. Bishop Museum: Honolulu. HALL, H. R. 1928. The civilisation of Greece in the Bronze Age (the Rhind lectures 1923). Methuen & Co. Ltd.: London.

ERIKSEN, T. H. 1993. ‘In which sense do cultural islands exist?’. Social Anthropology, 1.1B: 133–147.

HALSTEAD, P. 1993. ‘Spondylus shell ornaments from late Neolithic Dimini, Greece: specialized manufacture or unequal accumulation?’ Antiquity, 67: 603–609.

ERKURT, G. 2011. ‘Experimental Archaeology: Project for the re-animation of Cycladic boats’. In V. Şahoğlu and P. Sotirakopoulou (eds.), Across: The Cyclades and Western Anatolia during the 3rd Millennium BC. Sakip Sabanci Muzesi: Istanbul. Pp. 178–183.

HOOD, S. 1967. The home of the heroes: The Aegean before the Greeks. Thames and Hudson: London. HOURMOUZIADES, G. 1971. ‘Δυο νέαι εγκαταστάσεις της αρχαιότερης Νεολιθικής περιόδου εις την Θεσσαλίαν’. Αρχαιολογικά Ανάλεκτα Αθηνών, 4.2: 164–175.

EVAGGELISTIS, D. forthcoming. ‘Survey in Giagana bay Cephallonia’. In H. Tzalas (ed.), TROPIS IIX. International Symposium on Ship Construction in Antiquity, Ydra, 2002.

HOURMOUZIADES, G. 1996. Dispilio (Kastoria). The prehistoric lake-side settlement. Codex Publications: Thessaloniki.

EVANS, A. J. 1921. The palace of Minos: a comparative account of the successive stages of the early Cretan civilization as illustrated by the discoveries at Knossos. Vol. I. Macmillan and CO. limited: London.

HUTCHINSON, R. W. 1962. Prehistoric Crete. Penguin Books: Harmondsworth.

EVANS, A. J. 1923. The palace of Minos: a comparative account of the successive stages of the early Cretan civilization as illustrated by the discoveries at Knossos. Vol. II, part I. Macmillan and CO. limited: London.

IRWIN, G.J. 1980. ‘The prehistory of Oceania: colonization and cultural change’. In A. Sherratt (ed.), The Cambridge Encyclopaedia of Archaeology. Cambridge University Press: Cambridge. Pp. 324–332.

EVANS, A. J. 1925. ‘The early Nilotic, Libyan and Egyptian relations with Minoan Crete’. The Journal of the Royal Anthropological Institute of Great Britain and Ireland, 55: 199–228.

JOHNSTON, P. F. 1982 [2015]. ‘Bronze Age Cycladic ships: An overview’. In P. P. Betancourt (ed.), Temple University Aegean Symposium. A compendium. INSTAP Academic Press: Philadelphia, PA. Pp. 363–370.

EVANS, A. J. 1928. The palace of Minos: a comparative account of the successive stages of the early Cretan civilization as illustrated by the discoveries at Knossos. Vol. II, part II. Macmillan and CO. limited: London.

JOHNSTON, P. F. 1985. Ship and Boat Models in Ancient Greece. Naval Institute Press: Annapolis, MD. JOHNSTONE, P. 1973. ‘Stern first in the Stone Age?’ The International Journal of Nautical Archaeology and Underwater Exploration, 2.1: 3–11.

EVANS, A. J. 1935. The palace of Minos: a comparative account of the successive stages of the early Cretan civilization as illustrated by the discoveries at Knossos. Emergence of outer western enceinte, with new illustrations, artistic and religious, of the Middle Minoan Phase. Vol IV, part I. Macmillan and CO. limited: London.

JOHNSTONE, P. 1980. The sea-craft of Prehistory. Harvard University Press: Cambridge, MA. KAPITÄN, J. 1999. ‘What was the Early Bronze Age ship wrecked at Dokos’. In H. Tzalas (ed.), TROPIS V: 5th International Symposium on Ship Construction in Antiquity, Nauplia, 26, 27, 28 August 1993. Hellenic Institute for the Preservation of Nautical Tradition: Athens. Pp. 225–236.

FUGAZZOLA DELPINO, M. A. and MINEO, M. 1995. ‘La piroga neolitica del lago di Bracciano (“La Marmotta 1”)’. Bullettino di Paletnologia Italiano, 86: 197–266.

KOUKA, O. 2008. ‘Diaspora, Presence or Interaction? The Cyclades and the Greek Mainland from the Final Neolithic to Early Bronze II’. In N. J. Brodie, J. Doole, G. Gavalas and C. Renfrew (eds.), Horizon – Όρίζωυ: A Colloquium on the prehistory of the Cyclades, Cambridge, 25th-28th March 2004. McDonald Institute for Archaeological Research: Cambridge. Pp. 311–319.

GARDNER, J. 1977. Building classic small craft. Vol. 1. International Marine Publishing: Camden, Maine. GLOTZ, G. 1923. La civilisation Égéenne. La Renaissance du Livre: Paris. GOSDEN, C. and PAVLIDES, C. 1994. ‘Are islands insular? Landscape vs. Seascape in the case of the 19

Panos Tzovaras KÜNZE, E. 1934. Orchomenos, III. Die keramik der frühen Bronzezeit. Bayerischen Akademie der Wissenschaften: München.

MEIGGS, R. 1982. Trees and timber in the ancient Mediterranean world. Oxford University Press: Oxford.

LANDSTRÖM, B. 1961. Das schiff. Vom einbaum zum atomboot. Rekonstruktionen in bild und wort. C. Bertelsmann Verlag: Gütersloh.

MORRISON, J. S. 1972. ‘Ships and seamanship in the ancient world’. Casson, L. Reviewed in: International Journal of Nautical Archaeology, 1: 230–233.

MARANGOU, C. 1990. ‘Rowers paddling sailing ships in the Bronze Age Aegean’. In H. Tzallas (ed.), TROPIS II. 2nd International Symposium on Ship Construction in Antiquity. Delphi 28, 29, 30 August 1987. Hellenic Institute for the Preservation of Nautical Tradition: Athens. Pp. 259–270.

MORISSON, J. S. 1995. ‘Ships and boats in the Mediterranean from 3000 to 500 BC’. In B. Greenhill (ed.), The archaeology of boats & ships. An introduction. Conway Maritime Press: London. Pp. 131–150. NTINOU, M. 2013. ‘Wood Charcoal: Vegetation and the Use of Timber at Dhaskalio’. In C. Renfrew, O. Philaniotou, N. Brodie, G. Gavalasand, and M. J. Boyd (eds.), The settlement at Dhaskalio. The sanctuary on Keros and the origins of Aegean ritual practice: the excavations of 2006–2008. Volume I. McDonald Institute for Archaeological Research: Cambridge. Pp. 417–428.

MARANGOU, C. 2001a. ‘Neolithic watercraft: Evidence from Northern Greek wetlands’. In A. B. Purdy (ed.), Enduring records. The environmental and cultural heritage of wetlands. Oxbow Books: Oxford. Pp. 192– 205. MARANGOU, C. 2001b. ‘Neolithic craft: Evidence about boat, types and uses’. In Y. Bassiakos, E. Aloupi and Y. Farcorelis (eds.), Archaeometry issues in Greek prehistory and antiquity. Hellenic Society of Archaeometry and Society of Messenian Archaeological Studies: Athens. Pp. 737–751.

NTINOU, M., forthcoming. ‘Ξυλώδη φυτά στα αρχαιολογικά ιζήματα των ανασκαφών στο Βαθύ: προκαταρκτική αρχαιοβοτανική μελέτη μακροκαταλοίπων καμένου ξύλου στο πλαίσιο του ερευνητικού προγράμματος PlantCult-ERC’. Paper presented at: Βαθύ Αστυπάλαιας. Πέντε χρόνια συστηματικής έρευνας σε ένα διαχρονικό παλίμψηστο του Αιγίου. Επιστημονική ημερίδα, Σάββατο, 14 Ιουλίου 2018. Ροδιά Χώρας.

MARANGOU, C. 2003. ‘Neolithic watercraft in Greece: Circumstantial evidence and serious causes’. In C. Beltrame (ed.), Boats, Ships and Shipyards: Proceedings of the Ninth International Symposium on Boat and Ship Archaeology, Venice 2000. Oxbow Books: Oxford. Pp. 14–18.

PANTELIDOU-GOFA, M. 1996. ‘Central Greece’. In G. A. Papathanasopoulos(ed.), Neolithic Culture in Greece. N. P. Goulandris Foundation, Museum of Cycladic Art: Athens. Pp. 69–72.

MARINATOS, S. N. 1930. ‘Αί θαλασσογραφικαὶ παραστάσεις τῆς Κρητομυκηναϊκῆς ἐποχῆς’. Αρχαιολογική Εφημερίς. Pp. 108–126.

PAPADATOS, Y. 2012. ‘An Early Minoan boat model from Kephala Petras, Siteia’. In E. Mantzourani and P. P. Betancourt (eds.), PHILISTOR: Studies in Honor of Costis Davaras. INSTAP Academic Press: Philadephia, PA. Pp. 155–159.

MARINATOS, S. N. 1933. ‘La marine créto-mycénienne’. Bulletin de correspondance hellénique, 57: 170–235. MARINATOS, S. N. 1974. ‘Das Schiffsfresko von Akrotiri, Thera’. In D. Grey , Seewesen. Archaeologia Homerica, Band I Kapital G. Vandenhoeck & Ruprecht: Gottingen. Pp. 140–151.

PAPADATOS, Y. and TOMKINS, P. 2013. ‘Trading, the longboat and cultural interaction in the Aegean during the late fourth millennium BCE. The view from Kephala Petras, East Crete’. American Journal of Archaeology, 117.3: 353–381.

MARTHARI, M. 2017. ‘Aspects of pictorialism and symbolism in the Early Bronze Age Cyclades: a ‘frying pan’ with longboat depiction from the new excavations at Chalandriani in Syros’. In V. Vlachou and A. Gadolou (eds.), ΤΕΡΨΙΣ. Studies in Mediterranean Archaeology in honour of Nota Kourou, 10: 147–160.

PAPATHANASOPOULOS, P. 1976. ‘Το Πρωτοελλαδικό ναυάγιο της Δοκού’. Αρχαιολογικά Ανάλεκτα εξ Αθηνών ΙΧ.1: 17–22. PERLÉS, C. 1987. Les industries lithiques taillees de Franchthi (Argolide, Grece). Tome I: Presentation generate et industries paleolithiques. Excavations at Franchthi Cave, fasc. 3. Indiana University Press: Bloomington and Indianapolis.

MAVROMATI, A. 2016. ‘Landscape and wood-fuel in Akrotiri (Thera, Greece) during the Bronze Age’. Quaternary International xxx: 1–12. MCGRAIL, S. 1987. Ancient boats in North-West Europe: The archaeology of water transport to AD 1500. Longman Archaeology Series: London and New York.

RAINBIRD, P. 1999. ‘Islands out of time: towards a critique of Island Archaeology’. Journal of Mediterranean Archaeology, 12.2: 216–234. Et al.

MCGRAIL, S. 2001. Boats of the world: from the Stone Age to Medieval times. Oxford University Press: Oxford.

20

Before ‘Thalassocracies’ RENFREW, C. 1967. ‘Cycladic metallurgy and the Aegean Early Bronze Age’. American Journal of Archaeology, 71: 1–20.

cave art from Crete, Greece’. Journal of Archaeological Science: Reports, 18: 100–108. SVORONOS, I. 1923. ‘Τo λάβαρον των ‘Αθηναίων. Πέπλος, ίστός, κεραία’. Αρχαιολογικόν Δελτίον 6: 21– 25.

RENFREW, C. 1972. The emergence of civilization. The Cyclades and the Aegean in the third millennium B.C. Methuen and Co Ltd: London.

TARTARON, T. H. 2013. Maritime networks in the Mycenaean world. Cambridge University Press: Cambridge.

RENFREW, C. 1993. ‘Trade beyond the material’. In C. Scarre and F. Healy (eds.), Trade and Exchange in Prehistoric Europe: Proceedings of a Conference Held at the University of Bristol, April 1992 (Oxbow Monographs in Archaeology). Oxbow Books: Oxford. Pp. 5–16.

TAYLOR, D. A. 1980. Merchant ship construction. London: Butterworth & Co. Ltd. TELEVANTOU, C. A. 2017. ‘Strofilas, Andros: New perspectives on the Neolithic Aegean’. In S. Dietz, F. Mavridis, Z. Tankosić and T. Takaoğlu, (eds.), Communities in Transition: The Circum-Aegean Area in the 5th and 4th Millennia BC. Oxbow Books: Oxford. Pp. 389–396.

RENFREW, C. 2011. ‘Preface and Introduction’ (2010). In: The Emergence of Civilisation: The Cyclades and the Aegean in the Third Millennium BC. Oxbow Books: Oxford and Philadelphia. Pp. XXVII-LI. RHOMIOPOULOU, K., and RIDLEY, C. 1973. ‘Prehistoric settlement of Servia (W. Macedonia). Excavations 1972’. Athens Annals of Archaeology, 6: 419–424.

TELEVANTOU, C. A. 2018. ‘Οι ρίζες της εικονιστικής τέχνης στις Κυκλάδες: από τον Στρόφιλα στο Ακρωτήρι / The Roots of the Pictorial Art in Cyclades: from Strophilas to Akrotiri’. In A. Vlachopoulos, (ed.), ΧΡΩΣΤΗΡΕΣ. Η Τοιχογραφία και η Αγγειογραφία της 2ης χιλιετίας π.Χ. σε Διάλογο / PAINTBRUSHES. WallPainting and Vase-Painting of the 2nd millennium BC in dialogue. Ταμείο Αρχαιολογικών Πόρων: Athens. Pp. 42–65.

RIDLEY, C. and WARDLE, K. A. 1979. ‘Rescue excavations at Servia 1971–1973: A preliminary report’. Annual of the British School at Athens, 74: 185–226. ROBERTS, O. T. P. 1987. ‘Wind-power and the boats from the Cyclades’. The International Journal of Nautical Archaeology and Underwater Exploration 16.4: 309–311.

TERRELL, J. E. 2004. ‘Island models of reticulate evolution: the “Ancient Lagoons” hypothesis’. In S. M. Fitzpatrick (ed.), Voyages of Discovery: The Archaeology of Islands. Praeger: Westport, CT. Pp. 203–222.

SAMPSON, A. 2002. The Neolithic settlement at Ftelia, Mykonos. University of the Aegean: Rhodes.

THIMME, J. 1977. Art and culture of the Cyclades: Handbook of an ancient civilisation. The University of Chicago Press: Chicago.

SAMPSON, A., KOZŁOWSKI, J. K., KACZANOWSKA, M. and GIANNOULI, B. 2002. ‘The Mesolithic settlement at Maroulas, Kythnos’. Mediterranean Archaeology and Archaeometry, 2.1: 45–67.

TREGEAR, E. 1904. The Maori race. A. D. Willis, Printer and Publisher: Wanganui, NZ.

SEAGER, R. B. 1909. ‘Excavations on the island of Mochlos, Crete, in 1908’. American Journal of Archaeology, 13.3: 273–303.

TREUIL, R., DARCQUE, P., POURSAT, J-C. and TOUCHAIS, G. 1989. Les civilisations égéennes du néolithique et de l’Age du Bronze. Presses Universitaires de France: Paris.

SEAGER, R. B. 1912. Explorations in the island of Mochlos. American School of Classical Studies at Athens: Boston and New York.

TSOUNTAS, C. 1899. ‘Κυκλαδικά ΙΙ’. Αρχαιολογική Εφημερίς, 17: 73–134.

SHERRATT, S. 2000. Catalogue of Cycladic antiquities in the Ashmolean Museum. The captive spirit. Vol. 1. Oxford University Press: Oxford.

TUPPER, E. C. and RAWSON, K. J. 2001. Basic Ship Theory. Vol. 1. 5th ed. Butterworth-Heinemann: Oxford.

SOLES, S. J., 2012. Mochlos boats. In E. Mantzourani and P. P. Betancourt (eds.), PHILISTOR: Studies in Honor of Costis Davaras. INSTAP Academic Press: Philadelphia, PA. Pp. 187–199.

TZOVARAS, P. forthcoming (a). ‘Τεχνολογική και τυπολογική ανάλυση των πλοίων στις βραχογραφίες της Αστυπάλαιας’. Paper presented at: Βαθύ Αστυπάλαιας. Πέντε χρόνια συστηματικής έρευνας σε ένα διαχρονικό παλίμψηστο του Αιγίου. Επιστημονική ημερίδα, Σάββατο, 14 Ιουλίου 2018. Ροδιά Χώρας.

STEFFY, R. J. 1994. Wooden ship building and the interpretation of shipwrecks. Texas A&M University Press: College Station.

TZOVARAS, P. forthcoming (b). ‘Cultural ‘koinae’ and maritime networks in the 4th millennium Aegean and its adjacent coastland: Mapping the distribution of

STRASSER, T. F., MURRAY, S. C., VAN DER GEER, A., Kolbd, C. and Ruprecht, L. A. Jr. 2018. ‘Palaeolithic

21

Panos Tzovaras the material culture and sites’ intervisibility through ArcGIS’. Paper presented at: Islands in Dialogue (ISLANDIA) International Postgraduate Conference in the Prehistory and Protohistory of Mediterranean Islands. 14th, 15th and 16th of November 2018 Università degli Studi di Torino, 2018.

WACHSMANN, S. 1980. ‘The Thera waterborne procession reconsidered’. The International Journal of Nautical Archaeology and Underwater Exploration, 9.4: 281–295. WACHSMANN, S. 1998. Seagoing ships and seamanship in the Bronze Age Levant. Chatham Publishing: London.

VAN de MOORTEL, A. 2012. ‘Middle Bronze Age Boat of Mitrou, Central Greece’. In N. Günsenin (ed.), Between continents: Proceedings of the Twelfth Symposium on Boat and Ship Archaeology Istanbul 2009. Ege Yayınları: Istanbul. Pp. 17–25.

WEDDE, M. 2000. Towards a hermeneutics of Aegean Bronze Age ship imagery. Bibliopolis: Manheim and Möhnesee. WEISGERBER, G. 1985. ‘Bemerkungen zur prihistorischen und antiken Bergbautechnik’. In G. A. Wagner and G. Weisgerber (eds.), Silber, Gold Blei und Gold auf Sifnos: Prdihistorische und antike Metallproduction (Der Anschnitt, Beiheft 3). Deutsches Bergbau-Museum: Bochum. Pp. 86–112.

VAN de MOORTEL, A. 2017. ‘A new typology of Bronze Age Aegean ships: developments in Aegean shipbuilding in their historical context’. In J. Litwin (ed.), Baltic and beyond. Change and continuity in shipbuilding. Proceedings of the Fourteenth International Symposium on Boat and Ship Archaeology, Gdańsk 2015. National Maritime Museum: Gdańsk. Pp. 263–268.

WYLIE, A. 1985. ‘The reaction against analogy’. Advances in Archaeological Method and Theory, 8: 63–111.

VERMEULE, E. 1964. Greece in the Bronze Age. University of Chicago Press: Chicago.

XANTHOULIDES, A. 1924. The vaulted tombs of Mesará: an account of some early cemeteries of southern Crete. University Press of Liverpool Ltd.: Liverpool; Hodder & Stoughton Ltd.: London.

VICHOS, Y. 1989. ‘Aerodynamics and hydrodynamics of the Early Cycladic ships, illustrating pan-shaped utensils from Syros’. Αρχαιολογία, 32: 21–23.

ZACHOS, K. 1996. ‘The Cyclades and the Northeastern Aegean islands’. In G. A. Papathanasopoulos (ed.), Neolithic Culture in Greece. N. P. Goulandris Foundation, Museum of Cycladic Art: Athens. Pp. 85–87.

VICHOS, Y. 1990. ‘L’ extrémité haute des navires à l’âge du Bronze en mer Egee: La poupe ou la proue? Une approche nautique au problème de I’identification des extremités des navires cycladiques representés sur les “poelons” de Syros’. In H. Tzalas (ed.), TROPIS II. 2nd International Symposium on Ship Construction in Antiquity. Delphi 28, 29, 30 August 1987. Hellenic Institute for the Preservation of Nautical Tradition: Athens. Pp 363–370.

ZACHOS, K. 2007. ‘The Neolithic background: A reassessment’. In P. M. Day and R. C. P. Doonan (eds.), Metallurgy in the Early Bronze Age Aegean. Sheffield archaeological monographs Volume 7, Sheffield studies in Aegean archaeology. Oxbow Books: Oxford. Pp. 168–206.

VICHOS, Y. 1999. ‘Three unknown representations of ancient ships (from early Bronze Age to the late Geometric period)’. In H. Tzalas (ed.), TROPIS V: 5th International Symposium on Ship Construction in Antiquity, Nauplia, 26, 27, 28 August 1993. Hellenic Institute for the Preservation of Nautical Tradition: Athens. Pp. 477–490. VLACHOPOULOS, A. and ANGELOPOULOU, A. forthcoming. ‘Early Cycladic figurines from Vathy, Astypalaia’. In M. Marthari, C. Renfrew, and M. J. Boyd (eds.), Early Cycladic sculpture in context from beyond the Cyclades: Mainland Greece, the north and the east Aegean. Athens Archaeological Society 25–26 May 2015. Pp. 223–249. VLACHOPOULOS, A., forthcoming. Vathy Astypalaia. ‘An Early Cycladic site in the Dodecanese’. In S. Vitale and T. Marketou (eds.), The Southeast Aegean/ Southwest Coastal Anatolian Region. (SASCAR): Material Evidence and Cultural Identity I. The Early and Middle Bronze Age, Italian School of Archaeology, Scuola Archeologica Italiana, Parthenonos 14, 117 42 Athens May 12th to 14th, 2016.

22

3 Technology Behind the Mazarrón Boats: a Virtual 3D Approximation Carlos Cabrera Tejedor Oxford Centre for Maritime Archaeology, University of Oxford [email protected] Abstract: Unlike shipbuilding in the Greco-Roman era, which is moderately well-known through many examples, shipbuilding practises in the Iron Age remains largely conjectural. Shipwreck evidence for ancient Eastern Mediterranean ship construction is limited to only a few examples. Two shipwrecks discovered at Playa de la Isla in Mazarrón, Spain, provide exceptional information to improve our understanding regarding shipbuilding technology during the Iron Age. The purpose of this paper is to analyse, in detail, and present the design and construction of these two 7th-century B.C. vessels, the Mazarrón hulls. Key Words: Mazarrón shipwrecks; Phoenicians; Iron Age; Iberian Peninsula; Orientalizing Period; ancient shipbuilding; naval architecture Preface

world (Throckmorton, 1987: 50; Aubet, 1993). Except for written classical sources, scarce iconographical and model testimonies, and a few archaeological examples, virtually nothing was known about the vessels that allowed these people to explore and trade across the Mediterranean, and even beyond, for over a millennium. We know from ancient texts about these peoples’ expansion to the west of stemming from the Canaanites. These Greek texts describe several different Phoenician ships: Hippoi (Horses), named for their characteristically horse-head shaped stem and sternpost; and Gauloi (Tubs), bigger seagoing merchant vessels (Casson 1994: 41; Aubet 1993: 146).

This paper, originally presented in 2008 (vide infra), was the first and initial article on the subject of the Mazarrón boats written by the author. However, this initial work has been succeeded and expanded by other subsequent academic papers that explore more in detail the construction details of the Mazarrón boats, particularly those of the Mazarrón 1 shipwreck, see Cabrera Tejedor 2017 and 2018. The present study and reconstruction of the Mazarrón boats first consisted in a reconstruction of the hull with theoretical lines plans following the ‘Steffy method’ (Steffy 1994) (completed in 2006). A second phase (completed in 2007), involved a 3D reconstruction using 3D modelling software Rhinocerso 4.0. These initial reconstructions were created from thorough study of the publications about the Mazarrón boats published until then as final projects of different graduate courses at the Nautical Archaeology Program at Texas A&M University. The preliminary observations from those studies were presented in an oral presentation and submitted in an article in August 2008 at the tenth International Symposium on Ship Construction in Antiquity (TROPIS X). However, the TROPIS symposium no longer takes place and they have not yet published the proceedings of the last three symposia: TROPIS VIII (2002) TROPIS IX (2005), and TROPIS X (2008). Therefore, here is the same article that was submitted in 2008 to the TROPIS X proceedings yet was never published and thus is presented here for the first time.

We have few representations of these types of vessels as shown in examples such as: a relief at the Palace of Sennacherib (ca. 700 B.C.) showing King Luli and his people fleeing from Tyre to Cyprus, another relief from the Palace of Sargon (722–715 B.C.) at Khorsabad, and an Iron Age II clay model ship from Amathus, Cyprus (Basch 1984: 306–318).

Introduction

However, archaeological examples of Late Bronze Age and Iron Age shipwrecks are scarce. From the Late Bronze Age, two shipwrecks, at Uluburun, (ca. 1300 B.C. (Pulak 1998: 188–224 and Bass 2005: 34–47) and at Cape Gelidonya, (ca. 1200 B.C. (Bass 2005: 48–55, and Bass 1991: 69–82), provide our only archaeological remains and only scant remains of their hulls were preserved. From the Iron Age two more shipwrecks are known, the Ashkalon wrecks (Ballard 2002 and Stager 2004), Tanit and Elissa (ca. 750 B.C.), but these lay hundred of meters below sea level and thus have not been excavated.

It is well known that the Phoenicians were some of the most enterprising traders and mariners of the ancient

Consequently, the discovery in 1988 of a shipwreck with Phoenician pottery of the 7th century B.C. at Playa de la Isla 23

Carlos Cabrera Tejedor vessels have been assigned consecutive numbers starting from the stern and moving forward to the bow. As for the photographs that I examined, I will indicate the original article where they are published. The reason for this is to avoid confusion should the reader compare this paper with the original sources.

in Mazarrón (Roldán 1994 and Negueruela 1995), and a second shipwreck discovered in 1994, at the same location and in a state of near perfect preservation (Negueruela 2000), represented an important breakthrough for a better understanding of Iron Age vessels. These two shipwrecks are important for several reasons. Firstly, they are two new examples to add to our incomplete picture of Iron Age shipbuilding. Secondly, they are the only two shipwreck examples belonging to the 7th century B.C. (Parker 1992: 14). Finally, they are probably our only examples of western Phoenician related vessels, with the exception of the Marsala Punic shipwreck (Frost 1981).

Reconstruction of the boat For the reconstruction, I follow the suggestion that both vessels are fairly similar and were built using the same construction techniques (Negueruela 2004: 230; 2006: 24). I have studied each photograph and drawing in detail. I then compared and contrasted them and matched them with the site plan. Then, using all the data I compiled, I made a line-drawing of an approximation of the vessel (Figure 3). During this process, I found that some of the published measurements are not logical, and most likely are incorrect, and thus I revised them as needed. My reasons are discussed below.

Unfortunately, nowadays, decades after their discovery, their excavation, study, and publication remain incomplete. We only have scattered articles (many in obscure publications) that, regrettably, do not offer a complete explanation of the structural and shipbuilding features of these two vessels. Probably, this is due to the unfortunate fact that none of the excavation directors had a background in nautical archaeology.

Construction technique

The purpose of this paper is, therefore, to first summarize the scattered information, second, to discuss and provide a general overview of the construction details through an approximation of the Mazarrón wrecks. The hull of both Mazarrón 1 and 2 are reported as having similar structural elements, overall dimensions and general shape (Negueruela 2004: 230; 2006: 24). Therefore, the limited data available for Mazarrón 1 was supplemented by construction details drawn from Mazarrón 2 and used to create a preliminary computer-based three-dimensional approximation of the original hull of the Mazarrón boats.

The Mazarrón boats are reported to be built (Negueruela 2004: 244) in the shell-first technique (Basch 1972, 15– 18). Overall dimensions The overall preserved dimensions of the Mazarrón 2 remains are reported to be: length 8.15 m; width 2.20 m; and minimum height 0.90 m (Negueruela 2004: 235). But, for interpretation purposes, we must understand that the original dimensions of the vessel were larger, due to the loss of the upper planks on the port side and at bow and stern. Moreover, the study of the site plan and some photographs provide us with two more pieces of evidence that we must take into consideration: the hull lists to starboard at an inclination of no greater than 20 degrees (Negueruela 2004: figs. 9 and 19),furthermore, the hull has been flattened out slightly by the weight of the sand which covered it for the last 26 centuries.

Construction details of the Mazarrón boats The information regarding the construction of the hulls presented within the published articles, unfortunately, seems incomplete, poorly organized and, in certain cases, incoherent. Information regarding the construction of the ships is principally found in one article regarding Mazarrón 1 (Figure 1) (Negueruela et al. 2000) and regarding Mazarrón 2 in two published articles (Negueruela 2004; 2006). What follows is a brief summary of the main shipbuilding elements of the Mazarrón boats reported by the excavators in the consulted references along with the procedure that I followed to complete the approximation of the Mazarrón boats.

Reconstructing the half breath view Longitudinal axis To obtain the longitudinal axis of the boat, first I traced a line between the stern and the bow. Since the maximum length of Mazarrón 2 is known as (8.15 m and its beam is (2.20 m (Negueruela 2004: 235), I located the amidships line at the centre of the longitudinal axis.

Much of the basic information regarding scantling and construction details of the Mazarrón boats was obtained through a careful study of the published photographs, drawings and site plans. From these, I made an approximation of the boat with several scale drawings that completed the general construction details of the vessel (Figures 3 and 4).

Symmetry of the remains Tracing a line between the tips of the bow and the stern in one published site plan (Negueruela 2000b, 194, fig. 8), I studied the symmetry of the hull remains, realising that the two sides of the boat were not symmetrical. This seems to be the consequence of three factors aforementioned: the

For clarity, I follow the same numbering system used in the consulted references. Structural elements of the 24

Technology Behind the Mazarrón Boats

Figure 1. Mazarrón 1 in situ at the Playa de la Isla underwater archaeological site (after Negueruela et al. 2000: photograph 5).

Figure 2. Site plan of Mazarrón 1 (scale bar 1 m, after Negueruela et al. 2000: fig. 1).

disappearance of some upper planks, the list of the vessel, and its flattened state.

Cylindrical Frames The frames of both boats are roughly cylindrical in section and have diameters reported as 4 cm (Negueruela 2004: 149), 6 cm (Gómez-Gil 1996: 219), and 8 cm (Cabrera 1992: 3). In the Mazarrón 2 boat no additional information, including their total number, is provided for the frames, except that they were attached to the hull with lashings (Negueruela 2004: 245). Frames are reported to be made of fig wood, Ficus carica L. (Negueruela 2004: 236–237; 2006: 25, 29).

Curvature of the sheer line To define the curvature of the sheer line, in half breath view, I used two main factors: site plan drawings and the measurements of the beams. Since in the site plan this curvature is not symmetrical or smooth, an approximation was made to estimate this irregular curve. To redefine this curve, I took into consideration the amidships position and beam of the remains. For the extremities of the boat, mainly the stern, I used the measurements of through beams. I divided by two the maximum length of the beams and matched this reference with the longitudinal axis. In this way I obtained the position of the sheer line at every beam. Thus, the sheer line must be placed slightly above the extremities of the beams. Taking into consideration all these factors and using a flexible ruler, I traced the sheer lines. The result was a rather elongated vessel (Figure 3).

The cylindrical frames are barely mentioned in the published articles. Only the method used to attach them to the hull (lashing) is mentioned (Negueruela 2004: 245, 249). The frames were lashed to the hull in an ‘x’ shaped pattern at the point where two planks are joined. Since the total number of frames in the Mazarrón 2 shipwreck is not given, the first step is to determine how many frames the boat had. Examination of the published photographs 25

Carlos Cabrera Tejedor

Figure 3. Lines drawing of the approximation of the Mazarrón boats hull (scale bar 3m, ©Author).

of the hull (the extremities and amidships), the frames are spaced wider apart. These are the areas where rigging elements are secured to the hull and manipulated: the steering oars and brails in the stern, the mast, sail and halyard amidships, and the anchor(s) at the bow. Free space is needed in these areas to allow the crew to manoeuvre the vessel. Second, in the two areas within the hull where the bulk of the cargo is stowed, between amidships and both ends, frames are placed closer together (15–20 cm). In my view, this is not coincidental but responds to a deliberate design of the hull. The mast-step timber The mast-step timber of Mazarrón 2 is 98 cm long and slightly wider than the keel (Negueruela 2004: 241). The excavator of the Mazarrón 2 boat describes its mast-step timber as a keelson, sobrequilla in Spanish (Negueruela 2004: 241). According to its measurements and configuration, it seems more appropriate to refer to this construction element as a mast-step timber. The timber is 98 cm long and extends over frames 6 and 7, slightly forward of amidships (Negueruela 2004: 273, figs. 23–25).

Figure 4. Computer based three-dimensional approximation of the Mazarrón boats hull (©Author).

and site plan suggests that there are at least 12 preserved frames (see photos in Negueruela 2004; 2006). To define the original positions of the frames, I traced parallel lines between the tips of the frames that appear in the site plan (Negueruela 2000b: 194, fig. 8). Once traced, these lines were not perfectly parallel. Their study shows that frame 6 was parallel with the remains of beams 2, 3, and 4. Thus, I considered frames 6 and 7 as my master frames and transversal lines. I redefined the position of the rest of the frames according to these master frames. The resulting frame spacing varies from 30–40 cm in the forward part of the hull, 40–50 cm in the centre, which is consistent with the published data (Negueruela 2004: 249), and 15–20 cm in the aft end of the boat.

The fact that the mast-step of the Mazarrón 2 boat lies over the frames illustrates that it was put in place after the frames (contra the sequence of construction offered by the excavator, who proposed that the mast-step was installed immediately after the keel and before the planks, Negueruela 2004: 241). Negueruela (2004: 241) describes the mast-step as having a large rectangular mortise into which the foot of the mast was inserted. It is fastened to the keel with five mortise-and-tenon joints: four longitudinally arranged in a straight line, with the fifth, aft, orientated perpendicular to the others. This mortise is located precisely beneath beam number 5, an important detail that will be discussed further, below.

This configuration has no regular frame spacing. Rather, there are two distinct regions of the hull wherein the frames are spaced differently. First, in the working areas 26

Technology Behind the Mazarrón Boats contrary, by carefully examining two specific photographs (Negueruela 2004: 272, fig. 20 and Negueruela et al. 2004: 471 fig. 17), it appears that beam 4 is not located so low. Thus, for my reconstruction I considered that the position of beam 4 is between strakes 7 and 8.

Through-beams of Mazarrón 2 According to the excavator, there are seven through-beams in the Mazarrón 2 hull remains. They are secured to the hull with dovetails (Negueruela 2004: 241) that prevent the lateral displacement of the beams and fix them to the hull (Negueruela 2004:, 273, fig. 22). The excavator describes the aft preserved beams 1, 2, 3, and 4 (Negueruela 2004: 242–243, 272, figs. 19–22), but barely mentions beams 5, 6, and 7 indicating that beam 5 is only partially preserved, while 6 and 7 are not preserved (Negueruela 2004: 244). Thus the first step was to determine where each beam was originally located and to estimate the overall dimensions of those beams that have been lost.

This through-beam configuration (with one beam situated below all the others, in an inferior strake) is remarkable and unique. There is no parallel of this configuration in any known ancient vessel. The excavator offers the following hypothesis: ‘since beam 3 is deteriorated, it may be that beam 4 was installed as a repair or for reinforcement to prevent any further deterioration of beam 3.’ (Negueruela et al. 2004: 479). This interpretation seems flawed; if there was a weakened beam, it would be more logical to simply replace it. It seems illogical to add another through-beam to strengthen one in bad shape on a lower strake below the bad one. To do this, the strakes above the new beam would have to be taken off, and thus the deteriorated beam with them. It would only make sense, then, to discard the deteriorated beam and replace it with a new one.

The position of the through-beams The excavator reported that the through-beams of Mazarrón 2 are placed on the eighth strake (Negueruela 2004: 241). Beam 1: length 36 cm, width 11.5 cm, thickness 4 cm (Negueruela 2004: 242). Beam 1 is near the stern, but we do not know the exact distance. It is the smallest of the beams. It should be located at least at strake 9, since the side of the hull here is preserved up to strake 8, but there is no notch for the through-beam (Negueruela 2004: 242). As the maximum length of the beam is 36 cm, it could not have been located lower than the ninth strake. Thus, in the approximation, the beam has been located between strakes 9 and 10.

It is not clear what role beam 4 played, but it must have had some legitimate purpose that we do not yet understand. Nevertheless, penetrating the hull with a hole of this size so close to the waterline is not a good idea and the shipwrights must have had a good reason for doing it. The position of these two beams (3 and 4) is above the scarf between the keel and sternpost. It may be that a second beam was inserted at this specific position in order to reinforce the joint, as perhaps with the stanchion of beam 7, as discussed below.

Beams 2, 3: Beam 2 is located on strake 8, 98 cm forward the stern and is 96 cm long, 20 cm wide and 4 cm thick (Negueruela 2004: 243). Beam 3 is located on strake 8, 162 cm forward of the stern and is 132 cm long, 15 cm wide and 4 cm thick (Negueruela 2004: 243).

Beam 5: Reported to be partially broken and only the port-side half is preserved. No measurements reported in references. It is placed between frames 5 and 6. Its exact position could be estimated because it is placed directly next to frame 6 (Negueruela 2004: 268, fig. 11).

Matching their exact distance from the stern with the longitudinal axis, I determined the exact position of these beams in the half breadth view. Beam 2 has two square holes. These holes were probably part of the structure that supported two steering oars placed on either side of the boat, similar to that on some Egyptian vessels (Landström 1970: 107–108, figs. 332–334, 337). Negueruela (2004: 243) states that beam 2 is 15 cm wide. However, in the scaled site plan (Negueruela 2000b: 194, fig. 8), the beam is recorded to be approximately 20 cm wide. One photograph in Negueruela (2004: 273, fig. 22) shows that beam 3 is roughly the same width as beam 2 (i.e. 20 cm). Thus, the measurement for the beam 3 in the approximation is 20 cm wide.

As previously mentioned, the aft extremity of the maststep timber is positioned just below beam 5. A mortise cut into this extremity was probably meant to secure the foot of a stanchion. This configuration of beams, secured with stanchions from beneath, is similar to that of the Jules Verne 7 shipwreck (Pomey 2003: 64, fig. 11.8) and also the Ma’agan Mikhael shipwreck (Linder 2003: 105). Beam 6: Only its starboard side notch is preserved. However, a stanchion is set into the keel at the exact position of this beam and must surely have been placed there to support it (Negueruela 2004: 270–271, figs. 15– 18). Furthermore, the exact position of beam 6 is known from the one preserved notch where it was inserted onto strake 8 on the starboard side (Negueruela 2004: 271, fig. 17) next to frame 11.

Beam 4: Located 168 cm forward of the stern, it is 96 cm long, 20 cm wide and 4 cm thick (Negueruela 2004: 243). Beam 4 is reported to be located 35 cm underneath beam 3 (Negueruela 2004: 243) that is secured on the eighth strake. Since the average width of the plank is about 15 cm. That would put beam 4 at least two strakes below beam 3; in other words, between strakes 6 and 7. On the

Beam 7: Beam 7 was the foremost beam, and only its starboard side notch is preserved. It was placed between frame 12 and the stem. Its notch is preserved in the top edge of strake 8 (Negueruela 2004: 270, 271, figs. 15, 17). 27

Carlos Cabrera Tejedor I estimated its position within the hull to be around 50 cm from frame 12.

Gil 1996: 219). In two publications, however, it has been reported as 4.50 m long (Negueruela 2004: 230; 2006: 24), yet this is not compatible with the evidence. The keel is made of Cupressus sempervirens L. (Negueruela 2004: 236–237; 2006: 25), commonly known as cypress, yet it has been mistakenly described as cedar (Negueruela 2004: 237). The keel of Mazarrón 2 is not described in the consulted references. The site plan of Mazarrón I shows that the keel ends directly under one of the frames (Negueruela et al. 2000: 1675, fig. 1). The keel has 6 mortises just forward of its mid-point where a mast-step was probably secured. Thus, we know that this end is the side which matches with the stem.

The ‘missing’ through-beam: At this point of my reconstruction, I realized that the present configuration of through-beams did not make sense. Two main beams at the stern, one single beam aft amidships, and two more main beams at the bow seems an illogical and unbalanced configuration. A configuration with two at the stern, two at amidships, and two at the bow would seem to be more logical, thus I looked for evidence of another throughbeam. There is no mention in the consulted references of Mazarrón 2 of any mast support placed near the maststep. However, such a support would be expected and a mast-partner through-beam must had been used to lock the mast when raised, as documented in the Ma’agan Mikhael shipwreck (Steffy 1994: 40–41; and Linder 2003: 105).

The keel scarf One of the most interesting aspects of this boat and its construction is the keel’s scarf. Unfortunately, only one report briefly mentions its unusual shape (Negueruela et al. 2000: 1673). Photographs published in the NAVIS database offer some details of the scarf’s shape (NAVIS Project web site: photos 7, 8, 11). Unfortunately, it is impossible to say more about this scarf beyond the fact that seems to be unique with no known archaeological parallels.

A through-beam positioned next to frame 7 would fit this purpose well, and would work well with the position of the mast-step, whose forward end overlaps this frame slightly. Upon careful examination of one specific photograph (Negueruela 2004: 27 fig. 23), it appears that the notch where a through-beam was inserted is preserved, on the port side eighth strake. Thus, for my reconstruction the final number of through-beams is eight instead of the reported seven. The addition of another beam helps to explain other details. As previously described, there is a mortise in the fore extremity of the mast step. We have seen that the aft mortise probably held the foot of a stanchion that supported beam 5 directly above it. It only makes sense that a similar configuration (a stanchion supporting a through-beam) would exist at the forward end of the step, as suggested by the preserved mortise at that end of the mast-step timber.

Stem and sternpost Mazarron 1 has no post preserved and almost nothing is published regarding the endposts of Mazarrón 2. Neither stem (Negueruela 2004: 270, figs. 16, 17) nor sternpost (Negueruela 2004: 270, figs. 21, 22) of Mazarrón 2 protrudes into the interior of the hull (Negueruela et al 2004: 478). The sternpost is curved and has a cylindrical hole in its top end (Negueruela 2004: 273, fig. 21); careful observation of this photograph shows that the sternpost has a rabbet. Until Mazarrón 2 is extracted and disassembled, it is impossible to know if this rabbet extends into the keel as well. Both stem and sternpost are curved as we can see in one of the published drawings (Negueruela 2004: 464, fig. 10), some photographs of the interior of the boat (Negueruela 2004: 271, figs. 17 and 21). Additionally, iconographic representations of contemporaneous boats from the Korsabad relief (722– 705 B.C.E.) (Basch 1984 308, fig. 650) depict similar boats with curved ends.

Reconstructing the sheer plan view Sheer line To define the sheer lines I started with the fact that the caprail of Mazarrón 2 is reported to be partially preserved (Negueruela 2004: 277, fig. 30), in the central third of the starboard side (Negueruela 2004: 240). Since we know the minimum height of the hull amidships is 90 cm (Negueruela 2004: 235), we can reconstruct the sheer line at least in this zone. The sheer line must rise at the extremities due the curvature of the planks. Based on iconography and historical traditions, the resulting curve should be slightly higher at the stern, where the steering oars were placed.

Planking The planks are barely mentioned in the published references. The planks of Mazarrón 2 are 19–23mm thick (Negueruela 2004: 247). A total count of the strakes within the hull is not given although planking presents flat scarves (Negueruela 2004: 248). Planking is made of pine (Pinus sp.) (Negueruela 2004: 236–237; 2006: 25). This thickness is relatively small when compared to examples from other ancient wrecks (see Table 1).

The keel The keel of the Mazarrón 1 shipwreck is well preserved but broken at one extremity (Figure 2). It has the following measurements: length 3.98 m (Negueruela et al. 2000: 1672), side 17 cm (maximum), moulded 10 cm (Gómez-

A total count of the strakes in the Mazarrón 2 shipwreck is not given. The photographs and one drawing (Negueruela 28

Technology Behind the Mazarrón Boats Table 1. Mortise-and-tenon joint comparison of Bronze Age and Iron age known shipwrecks Shipwreck

Date (B.C.)

Length of the vessel Plank thickness Tenon spacing Tenon Dimensions (LxWxTh) cm

Peg Diameter

Uluburun

1300

15 m

6 cm

21 cm

30x6.2x1.6

2.2 cm

C. Gelidonya

1200

12 m

?

?

22x5.3x1.3

1.8 cm

Mazarrón

650

8.5 m

2 cm

15-23 cm

9-11x4x0.6

1-1.5 cm

Jules-Verne 7

525

15 m

2.5-3 cm

19-20 cm

14-15x3-3.5x0.5

?

Ma’agan Mikhael

400

13.4 m

4 cm

12 cm

14x4.5x1

0.9 cm

Kyrenia

290

15 m

3.6 cm

12 cm

15-20x4.3x0.6

1 cm

et al. 2004: 462, fig. 6) show that Mazarrón 2 seem to have had at least 10 strakes. There are 10 strakes preserved on the starboard side (up to the caprail) and 9 on the port side (see photos in Negueruela 2004). The position and average width of the planks can be determined from the study of one particular photograph (Negueruela 2004: 271, fig. 18). Additionally, the site plan of Mazarrón 1 (Negueruela et al. 2000: 1675, fig. 1) provides a good scaled drawing of the remaining planks of this wreck. For the approximation I estimated that the planks were between 12 to 20 cm wide.

visible in the two published photographs (Negueruela et al. 2004: 476, figs. 23, 24). Historical context and purpose of the boats During the surveys at the Playa de la Isla site more than 7300 pottery sherds were recovered; approximately 70% were identified as Phoenician, although in ceramics collected from the immediate surroundings of the Mazarrón 1 shipwreck the percentage increased to c. 80% (Arellano, et al. 1999: 221; Barba et al., 1999: 199; Negueruela et al. 2000: 1671). All the Phoenician pottery sherds found had a homogenous chronology dating to the second half of the 7th century B.C. (Arellano et al. 1999: 221; Negueruela et al., 2000: 1672; Negueruela 2004: 238). No Phoenician pottery sherds were found which definitely belonged to either the 6th or 8th century B.C. (Barba et al. 1999: 199; Negueruela et al. 1995: 191, 192; 2000: 1671). This led the excavators to assign the Mazarrón boats to the second half of the 7th century B.C. (Negueruela et al. 1995: 189; 2000: 1672; Negueruela 2000a: 182; 2004: 238; 2006: 22–23). Dating obtained from radiocarbon analyses conducted on wood samples from Mazarrón 2 timbers, roughly matches the chronological dates of the pottery (Negueruela 2004: 238; 2006: 24, 25; Negueruela et al. 2000: 1674).

Joinery The hulls of the Mazarrón boats were built with pegged mortise-and-tenon joints (Basch 1972: 23–29), assembling keel, garboards, and all strakes of planking. The mortises of Mazarrón 2 measure 40 x 6mm and the spacing of the tenons varies from 15 to 23mm (Negueruela 2004: 247). Tenons and pegs are made of olive (Olea europea L.) (Negueruela 2004: 236–237; 2006: 25). Waterproofing The interior of the Mazarrón 2 hull retains remnants of its original coating of burned pine resin or tar, used to waterproof the wood (Negueruela 2004: 251). Additionally, it is reported that the planks of Mazarrón 1 above strake 2 have beveled edges and are sewn. The excavators suggested that this stitching was used to reinforce the joints of the planks as well as to secured in place thin ropes used as waterproofing material (Negueruela 2000: 1673). Only two graphical documents depicting this unique feature are published (NAVIS Project web site: photos 4 and 15).

The Orientalizing Period in the south of the Iberian Peninsula is an important phase of the colonization and commerce of the Phoenician culture, not only in Spain but for the rest of the Mediterranean network. In the study of this period the archaeological record probably constitutes the best source of information. Much has been done in land archaeology since serious studies and excavations began to be conducted during the 1980s. The exponential increase of archaeological studies has permitted a better understanding of the relations between the Phoenician colonists and the indigenous peoples of the Iberian Peninsula during the Iron Age.

The anchor Excavators found one well-preserved anchor among the remains of Mazarrón 2, located approximately 1 m from the hull remains. It is a two-armed wooded anchor with a lead cored wooden stock (Negueruela et al. 2004: 478). The anchor has a hole at the end of one of its arms for attaching a recovery line. This is the oldest example of this type of anchor. The earliest other example that we know appears 250 years later, on the Ma’agan Mikhael shipwreck (Linder 2003: 140–46). However, the Ma’agan Mikhael anchor has only one arm. Unfortunately, the excavators give no additional information other than that

Assuming that the pottery remains found around Mazarrón 1 were its cargo, it would constitute an archetypical example of internal commerce and redistribution of pottery between settlements. The red-polished Phoenician pottery of the eighth to the sixth centuries was manufactured widely, but with different finishes. Technology is starting to be able to physically analyse pottery made in different workshops so that imports and local manufactures can be correctly attributed to their sources, and distributions 29

Carlos Cabrera Tejedor drawn up. Many centres of production quickly arose; all with a similar repertoire based on Phoenician decorations favouring a red surface, or grouped horizontal bands of red, black, and maroon paint applied with multiple brushes. It will be quite interesting to track the origin of Mazarrón 1 pottery in order to figure out its possible route.

second regulus (gold and silver) and finally separating the lead from it. All this reveals an excellent knowledge of metallurgy, particularly significant if we bear in mind that the Tartessians had only been initiated into silver metallurgy on a large scale from the eighth century B.C. All these processes demanded the presence of specialist personnel (technicians, miners, metallurgists, transport workers) in charge of the various stages of production. Since we have already seen that the ore was not always extracted in the place where it was smelted or put aboard ship. Probably the metal was transported in the form of ingots or bars, previously weighed and divided up, from the ports of Huelva or Gadir direct to Tyre or Greece, and perhaps even to Pirecusas or other Mediterranean centres where we have documentary evidence of he activity of silver- and goldsmiths during the eighth to seventh centuries B.C.’ (Aubet 1993: 240).

Mazarrón 2 constitutes by itself a milestone regarding ancient shipbuilding because of its construction and state of preservation. It is the best-preserved shipwreck build with pegged mortise-and-tenon fastenings from the Iron Age. Mazarrón 2 was discovered carrying a cargo of 2800 kg of litharge (lead oxide) cake ingots (Negueruela 2004: 235; 2006: 25). Based on its cargo and small size, Mazarrón 2 was probably designed for the purpose of shorthaul transportation of metals mined locally around Mazarrón to larger production centres (i.e. colonies). Mazarrón 2 cargo could give us important information regarding the Phoenician metallurgic industry. Almost three tons of unrefined lead (litharge) ingots could seem a not glamorous cargo or even a strange cargo since it is well known that the industrial or large-scale use of lead did not happen until Roman times. Hence, why a lead cargo in the seventh century B.C.?

Thus, if the addition of lead makes easier the process and extraction of silver and gold, through the process of cupellation, perhaps the transport of lead cargoes to metallurgic centres could be a plausible explanation for such effort.

Several explanations can be offered, and will now be explored. The cargo was moved from the lode to a metallurgic centre to refine it. But we still know that Phoenicians did not use lead in industrial quantities; so why refine it? In order to extract silver contained in its composition? Metallographic studies should be conducted in the Mazarrón 2 ingots to confirm this hypothesis.

The Mazarrón 2 boat seems to have sunk just after it was loaded, probably on the same beach, with a cargo of metal ingots from the mines of Mazarrón. The reason is unknown since the area is protected from the winds (Rondán 1994: 505). The boat was found intact and still secured to its anchor by a well-preserved hawser. The excavator has suggested the possibility that the boat was overloaded (Negueruela et al., 2004, 480). The ship was carrying a cargo of ingots weighing 2,800 kg (Negueruela 2004: 235). If we add the weight of the other shipboard items such as the anchor, ropes, personal items, and crew (at least two sailors), the total weight of the boat would be around three tons. The reconstruction of the hull and the hydrostatic calculations suggest that this vessel had a maximum load capacity of about 4 tons (see Table 2). It seems then that the vessel was not loaded to capacity and, therefore, overloading can be discarded as a cause of its sinking.

Aubet (1993: 40) gave as a possible hint: ‘The technique used at Huelva and San Bartolomé de Almonte for processing silver was based on fusion and cupellation, using basically gossan ore with a high gold, silver and lead content. The process consisted in placing the crushed ore together with the flux, and heating it so as to obtain the slag and the regulus (lead, silver and gold), the lead serving to capture the metals. After this first phase of fusion came the cupellation proper; the regulus was placed in a cupel and fired; the cupel absorbed the lead, releasing a

Table 2. Reconstructed measurements, calculations and hydrostatic coefficients of the Mazarrón boats hull (©Author) Reconstructed dimensions and calculations of Mazarrón boats approximation Depth amidships

Overall length

Overall beam

Standard draught

Standard displacement

Length-to-beam ratio

c. 1,00 m

c. 8,20 m

c. 2,20 m

c. 0,20 m

c. 0,500 tn.

c. 3.7:1

Different drafts and their correspondent dimensions, calculations and coefficients of the hull Draught depth

Length on waterline

Beam on waterline

Hull volume

Displacement (saltwater)

Block coefficient Prismatic coefficient

Waterplane coefficient

0,65 m

7,127 m

2,191 m

4,813 m3

4,933 tn.

0,474

0,580

0,659

0,60 m

7,018 m

2,166 m

4,305 m3

4,412 tn.

0,471

0,582

0,658

0,45 m

6,655 m

2,053 m

2,878 m

3

2,950 tn.

0,468

0,585

0,656

0,30 m

6,220 m

1,862 m

1,636 m

3

1,676 tn.

0,470

0,589

0,650

0,15 m

5,689 m

1,523 m

1,636 m3

0,667 tn.

0,500

0,591

0,630

30

Technology Behind the Mazarrón Boats If the cause of its sinking is undetermined, an explanation for why the cargo was not recovered is even more inexplicable. The boats were found lying in only two meters of water and within 40 m of the beach. Even though the shore had a different configuration in antiquity, it was nevertheless still in shallow water and protected from the prevailing winds (Rondán 1994: 505–6). Perhaps we will never know why these boats sank or why they remained untouched until the end of the 20th century.

the use of mortises-and-tenon technique were Uluburun (ca. 1300 B.C.) (Pulak 1998: 188–224 and Bass 2005: 34–47) and Cape Gelidonya (ca. 1200 B.C.) (Bass 1991: 69–82; 2005: 48–55) shipwrecks. Unfortunately, it is impossible to know if these hulls were entirely assembled with mortises-and-tenon, since the hull’s remains of these wrecks are scarce in the first and almost inexistent in the second. The first archaeological evidence of a ship entirely assembled with mortise-and-tenon joints was the Kyrena shipwreck (ca. 290–280 B.C.). Thus, the Mazarrón discoveries advance the date of mastered use of mortisesand-tenon by more than four centuries.

Shipbuilding discussion and conclusions The development of shipbuilding techniques in antiquity is a topic often discussed by classicists, historians, and archaeologists. Patrice Pomey, analyses the development of shipbuilding techniques in one particular article (Pomey 1997) where he proposes that assembly techniques in antiquity can be divided into two principal families, or currents, characterized by the particular method of edge joinery. The first of these used ligatures to assemble the hull of a vessel, whereas the second employed pegged mortises-and-tenons joints.

The Mazarrón boats are long, thin, stylized, and sharp vessels with high stem and sternpost. Guerrero Ayuso (2008: 59–60) has suggested that perhaps the Mazarrón boats are the only known archaeological examples of vessels similar to th hippoi class described by Strabo (II, 3, 4), or those carved in relief in the palace of Khorsabad (722–705 B.C.), with a ‘horse-shaped’ figurehead at the stem and a ‘papyri form’ shaped sternpost (Basch 1984: 306–310; see vessels 1 and 5 in fig. 650A and vessel 8 in fig. 650B). If that were the case, it would be a magnificent discovery, not only for the boats themselves, but also for the fact that the first archaeological evidence for these famous ancient crafts.

Pomey explains that during the 6th century B.C. these two currents start to influence one another, and by the end of that century, both coexist in particular ‘transitional’ examples, such as the Jules-Verne 7 shipwreck, in which the two assembly techniques are found (Pomey 1997: 199). At the end of the 5th century B.C., a shipwreck discovered at Ma´agan Mikhael (c. 400 B.C.), Israel, maintains the use of the two methods but demonstrates the prevalence of the mortise-and-tenon technique. Additionally, this shipwreck incorporates new and significant developments, such as a new vessel shape (wine glass) and the use of metal fastenings (Linder 2003). The final stage of this transition, according to Pomey, takes place at the end of the 4th century B.C., with the Kyrenia shipwreck (Steffy 1985). Until the discovery of the Mazarrón boats, the Kyrenia shipwreck represented the first known example of a complete hull constructed with mortise-and-tenon joints. Kahanov and Pomey (2004) have then proposed that the progressive assimilation of pegged mortises-and-tenons joints by other shipbuilding traditions for fastening hull planking took place in the Mediterranean. Their theory argues that this transition was accomplished in several stages from the 6th to the 4th centuries or beginning of the 3rd century B.C. (Kahanov and Pomey 2004: 20–26).

If these boats are indeed hippoi, it would help validate the history related by Strabo of the Phoenicians circumnavigating Africa (Casson 1994: 66, n. 115). However, the low height of the sheer makes it difficult to believe that these boats were seagoing vessels. It may be that we are in dealing with a smaller version of the vessels referred to by Strabo. To summarize, despite the limited information published to date by the excavators, these boats have many unique characteristics that make them invaluable for better understanding our past, particularly the Iron Age in the Iberian Peninsula. The information compiled in this study represents an initial attempt to organize and present the available data in such a way as to show that these vessels represent a magnificent example of ancient shipbuilding. I hope that in the near future these boats will be studied in detail and that the results of such work will be published in a final report befitting such important archaeological remains. Works Cited

Guerrero Ayuso (2008) has suggested, rightly in my opinion, that the Mazarrón boats are likely to be a regional type of vessel, made by descendants of the first Phoenician colonists, or perhaps by the Iberian peoples that intermingled with the newcomers over many generations, absorbing their traditions and technologies.

AUBET, M. E., 1993, The Phoenicians and the West: Politics, Colonies and Trade. Cambridge. Cambridge University Press. BALLARD, R. D., et al., 2002. ’Iron Age Shipwrecks in Deep Water of Ashkelon, Israel’. American Journal of Archaeology 106.2:151–68.

Regarding the construction technique of the Mazarrón boats, the most remarkable feature is their assembly method. The hulls were assembled completely with tenons pegged into their mortises. Before the discovery of the Mazarrón boats, the earliest archaeological evidence of

BASCH, L., 1984. La musé imaginaire de la marine antique. Athenes. Institut hellénique pour la preservation de la tradition nautique. 31

Carlos Cabrera Tejedor BASCH, L., 1972. ‘Ancient Wrecks and the Archaeology of Ships’. International Journal of Archaeology 1: 1–58.

LINDER, E., KAHANOV, Y., BLACK, E., 2003, The Maʿagan Mikhael ship : The recovery of a 2400-yearold merchantman. Israel Exploration Society, University of Haifa. Jerusalem.

BASS, G. F. (ed.), 2005. Beneath the Seven Seas:  adventures with the Institute of Nautical Archaeology. London. Thames & Hudson.

NAVIS PROJECT ‘Mazarrón (English)’ http://www2. rgzm.de/navis/Ships/Ship058/Ship058Engl.htm [Accessed: 30/08/2018].

BASS, G. F. 1991. ‘Evidence of Trade from Bronze Age Shipwrecks’. In N.H. Gale (ed.), Bronze Age Trade in the Mediterranean. SIMA 90: Jonsered. Pp. 69–82.

NEGUERUELA, I., 2000a, Protection of Shipwrecks: the experience of the Spanish National Maritime Archaeological Museum. In M.H. Mostafa, N. Grimal, and D. Nakashima (eds.), Underwater Archaeology and Coastal Management. Focus on Alexandría (Alexandría, 1997). UNESCO: Paris. Pp. 111–116.

CABRERA, P., PINEDO, J., ROLDAN, B., BARBA, J.S., PERERA, J., 1992. ‘Campaña de cubrición del yacimiento subacuático de la Playa de la Isla (MazarrónMurcia)’. In II Jornadas de Arqueología Subacuática en Asturias (Gijón 1991). Oviedo. Pp. 10–21.

NEGUERUELA, I., 2000b, ‘Managing the maritime heritage. The National Maritime Archaeological Museum and National Centre for Underwater Research, Cartagena, Spain’. International Journal of Nautical Archeology 29: 179–198.

CABRERATEJEDOR, C., 2017, ‘Apuntes sobre el barco de Mazarrón 1: estimación de dimensiones, reconstrucción preliminar del casco, cálculos hidrostáticos, función y origen de la nave’. In M. Martínez Alcalde, J.M. García Cano, J. Blánquez Pérez and Á. Iniesta Sanmartín (eds.), Mazarrón II. Contexto arqueológico, viabilidad científica y perspectiva patrimonial del barco B-2 de la bahía de Mazarrón (Murcia). En homenaje a Julio Mas García. Universidad Autónoma de Madrid: Madrid. Pp. 187–227.

NEGUERUELA, I., 2004, ‘Hacia la comprensión de la construcción naval fenicia según el barco Mazarrón 2 del siglo VII a.C’. In V. Pena, A. Mederos and Carlos and G. Wagner (eds.), La navegación Fenicia. Tecnología naval y derroteros, Encuentro entre marinos, arqueólogos e historiadores. Universidad Complutense: Madrid. Pp. 227–78.

CABRERA TEJEDOR, C., 2018, ‘The Mazarrón 1 Shipwreck: an iron‐age boat with unique features from the Iberian Peninsula’. International Journal of Nautical Archaeology 47.2: 300–24.

NEGUERUELA, I., 2006, ‘Coagmenta punicana e bagli. La costruzione navale a fasciame portante tra i Fenici del VIIe s. a.C’. In B. M. Giannattasio, C. Canepa, L. Grasso and E. Piccardi (eds.), Aequora, pontos, jam, mare… Mare, uomini e merci nel Mediterráneo antico. Atti del Convegno internazionale, Genova, 9–10 dicembre 2004, All’Insegna del Giglio: Florence. Pp. 22–41.

CASSON, L., 1986, Ships and Seamanship in the Ancient World. Princeton, N.J. Princeton University Press. CASSON, L, 1994. Ships and Seafaring in Ancient Times. Austin, Texas. University of Texas Press. FROST, H., 1981. ‘Lilybaeum : The Punic Ship: Final Excavation Report’, in Notizie degli scavi di antichità. Roma. Accademia nazionale dei Lincei. 30 (Supplement).

NEGUERUELA, I., GONZÁLEZ GALERO, R., SAN CLAUDIO, M., MÉNDEZ SANMARTÍN, Á., PRESA, M., MARÍN, C., 2004, ‘Mazarrón-2: el barco fenicio del s. VII a.C. Campaña de noviembre-1999/ marzo 2000’. In G. Matilla, A. Egea, and A. González (coords.), El mundo púnico: religión, antropología y cultura material: actas II Congreso Internacional del Mundo Púnico, Cartagena 6–9 de abril de 2000. Universidad de Murcia: Murcia. Pp. 453–83.

GÓMEZ-GIL, C. SIERRA, J.L., 1996. ‘Extracción y tratamientos del barco fenicio (barco 1) de la Playa de la Isla (Puerto de Mazarrón, Mazarrón)’. Cuadernos de Arqueología Marítima 4: 217–25. GUERRERO AYUSO, V.M., 2008, ‘Barcos aborígenes en el estrecho de Gibraltar’. In VIII Jornadas de Historia de Ceuta. Barcos, puertos y navegación en la historia de Ceuta. Instituto de Estudios Ceutíes: Ceuta. Pp. 33–65.

NEGUERUELA, I., PINEDO, J., GÓMEZ, M., MIÑANO, A., ARELLANO, I., BARBA, J. S., 1995, Seventhcentury B.C. Phoenician vessel discovered at Playa de la Isla, Mazarrón, Spain. International Journal of Nautical Archaeology 24.3: 189–97.

KAHANOV, Y. and POMEY, P., 2004, ‘The Greek Sewn Shipbuilding Tradition and the Ma’agan Mikhael Ship: A Comparison with Mediterranean Parallels from the Sixth to the Fourth centuries B.C.’. Mariner’s Mirror 90.1: 6–28.

NEGUERUELA, I., PINEDO, J., GÓMEZ, M., MIÑANO, A., ARELLANO, I. BARBA, J.S., 2000, ‘Descubrimiento de dos barcos fenicios en Mazarrón (Murcia)’. In M. Barthélemy, M.E. Aubet Semmler (coords.), Actas del IV Congreso Internacional de Estudios Fenicios y Púnicos: Cádiz, 2 al 6 de octubre de 1995. Universidad de Cádiz: Cádiz. Pp. 1671–1680.

LANDSTRÖM, B., 1970, Ships of the Pharaohs. Garden City. New York.

32

Technology Behind the Mazarrón Boats PARKER, A. J.,  1992, Ancient Shipwrecks of the Mediterranean and the Roman Province. International Series 580. BAR Publishing. Oxford. POMEY, P., 1997. ‘Un example d´évolution des techniques de construccion navale antique: de l´assembalge par ligatures à l´assembalge par tenons et mortaises’. In Techniques et économie antiques et médiévales: le temps de l´innovation (Aix-en-Provence 1996). Paris. Pp. 195–203. POMEY, P., 2003, ‘Reconstruction of Marseilles 6th century B.C.E. Greek ships’. in C. Beltrame (ed.), Boats, ships and shipyards: proceedings of the Ninth International Symposium on Boat and Ship Archaeology, Venice 2000. Oxford. Pp. 57–65. PULAK, C., 1998, ‘The Uluburun Shipwreck: An Overview’. International Journal of Nautical Archaeology 27: 188–224. PULAK, C., 2005, ‘Discovering a Royal Ship from the Age of King Tut: Uluburun, Turkey’. In G.F. Bass (ed.), Beneath the Seven Seas. Thames & Hudson. Pp. 34–47. ROLDAN BERNAL, B., PERERA RODRIGUEZ, J., SANTOS J. FRUTOS, B. PINEDO REYES, J., 1994, ‘El fondeadero de la Playa de la Isla, avance preliminar’. In J.L. Cunchillos and M. Molina (eds.), El Mundo Púnico. Historia, sociedad y cultura, Cartagena 17–19 de noviembre de 1990. Murcia. Pp. 503–16. STAGER, L.E., 2004. ‘Dos pecios fenicios en alta mar de la costa norte del Sinaí’. In Peña, V. (ed.), La Navegación Fenicia: tecnología naval y derroteros. Encuentro entre marinos, arqueólogos e historiadores. Madrid. Pp. 179–97. STEFFY, J.R., 1994, Wooden Ship Building and the Interpretation of Shipwrecks. College Station, TX. STEFFY, J. R., 1985. ‘The Kyrenia ship. An Interim report on its hull construction’. American Journal of Archaeology 89–1: 71–101. STIEGLITZ, R., 2006. Classical Greek Measures and the Builder’s Instruments from the Ma`agan Mikhael Shipwreck. American Journal of Archaeology 110: 195–203. THROKMORTON, P., 1987. The Sea Remembers:  Shipwrecks and Archaeology from Homer’s Greece to the Rediscovery of the Titanic. New York. Weidenfeld & Nicolson.

33

4 Experimental Archaeology and the Contributory Reconstruction of a Roman Warship Mateusz Polakowski Centre for Maritime Archaeology, University of Southampton [email protected] Abstract: Although ancient authors state that it was quinqueremes (fives) that took part in the Battle of the Egadi Islands (Diodorus 24.11; Polybius 1.60–61), archaeological evidence does not support this. It seems very likely that the ships represented by the thirteen Egadi rams found at the Battle of the Egadi Islands Site were smaller attack galleys, unlike the quinqueremes crewed by 300 men described by Polybius (1.61). This paper attempts a minimal reconstruction of the basic hull dimensions of a galley that sank at the battle of the Egadi Islands in 241 BC. It will then present an interpretation of the size and capacity of the warships that have been found at the site. Using the Egadi 10 ram, the presented reconstruction developed a possible hull shape, set a load waterline and introduced three different rowing arrangements to test hull capacity while comparing that capacity to historical accounts. Keywords: Experimental archaeology; ship reconstruction; warships; Rome; Carthage; First Punic War; Egadi Island Rowed warships played pivotal roles throughout the Mediterranean basin yet their construction, development and classification remain uncertain. The mid-third century BC Egadi rams have revealed a wealth of new information about three-finned rams, which in turn have the potential to yield new insights into warship construction (Royal and Tusa 2012). Using principals put forth through experimental studies (Coates et al. 1995; Crumlin-Pedersen and McGrail 2006), this research attempts a minimal reconstruction of an Egadi warship utilising evidence from the Battle of the Egadi Islands. The aim of the reconstruction is to propose a possible size of the crew of the Egadi 10 warship, and to present a possible classification of the rams found at the Battle of the Egadi Islands Site.

reconstruction of a vessel with scarce remains. The Egadi 10 ram, in conjunction with additional archaeological evidence, can provide an interpretation of the hull, which can then provide a hypothesis on the size of the crew and the vessel’s classification. The main characteristics of a ship are its shape, weight distribution, method of propulsion and its construction. Together, these factors affect the stability, strength and performance of a vessel (Marsden 1993: 137). Most approaches to ship reconstruction follow a pattern of developing ship lines and construction from archaeological remains, and then attempt to recreate characteristics of the ship in the water with the cargo found associated with the wreck. The Egadi Islands assemblage presents a problem when using this approach; since nearly all the organic remains have disappeared, it is not possible to follow this process. The only form of direct archaeological evidence pertaining to hull shape were the angles and interior contours recorded from the interior of the Egadi 10 ram. For this reason, this research had to approach the reconstruction of the vessel through reverse engineering. Potential load waterlines were set at the beginning of the reconstruction and the general hull shape was defined by the need to house rowers down the length of the hull, while maintaining a hull size and shape that fit into shipsheds and met the requirements for fast, sleek vessels.

The reconstruction hypothesis was based around the principal that only a minimum reconstruction would be possible (McGrail 1984: 23). Capital and contributory reconstructions, two general categories proposed by J. Richard Steffy (1994: 214–21), help to further define and distinguish this type of reconstruction. Capital reconstructions include detailed lines drawings, construction plans and models based on a substantial amount of archaeological remains. Contributory reconstructions supply new information and interpretation of hull remains that are too degraded or scarce to conclusively define the original construction of the hull (Steffy 1994: 215). A reconstruction of an Egadi warship is a contributory reconstruction. A capital reconstruction would require extensive portions of an intact hull, addressing the biases of modern ship building, and acknowledgement of the inevitable conjectures and assumptions that result from the

The Egadi Warships The Egadi rams are bronze three-fin waterline rams, protecting the integrity of the ship and increasing its 35

Mateusz Polakowski

Figure 1. The Egadi 10 Ram (a) Front (b Port side (c) Starboard side (d) Interior looking aft (e) Top (f) Bottom (Image courtesy of RPM Nautical Foundation).

Figure 2. Schematic of Egadi 10 Ram terminology.

damage potential. Evidence of timber impression on the interior of the rams indicates they were cast around the bow timber (Tusa and Royal 2012: 12–14). The impacting piece was flat-faced so that it did not become lodged in the enemy vessel. The rams share the five basic structural elements: a ramming head, driving centre, wales pockets, a cowl for the stem and a bottom plate for the keel (Figure 2). The twelve Egadi rams found to date are similar both in size and shape to one another (Tusa and Royal 2012: 18–19), meaning that the warships that sank at the battle belonged to a certain class of vessel.

old. Polybius writes that the Carthaginian fleet was routed with 50 ships sunk and 70 being captured by the Romans, totalling 10,000 prisoners (1.61.5). Historical accounts attesting to ship classes range from ‘ones’ all the way up to ‘forties’. The basic concept used to determine this classification system seems to have relied on the rowing arrangements of oarsmen. With each increasing number class, there seems to have been a corresponding increase in the number of rowers per rowing station (Casson 1995: 77–140). This means a trireme (three) had three rowers per side per oar bank, a quadreme (four) had four and a quinquereme (five) had five.

These warships sank during the Battle of the Egadi Islands in 241 BC when a Roman fleet intercepted a Carthaginian relief fleet destined to relieve Hamilcar Barca’s troops entrenched on Mt. Erice (Eryx). Polybius describes a decisive victory by a Roman fleet of quinqueremes, each of which was manned by 300 rowers (1.61). The Carthaginians faced a serious disadvantage in that the ships sent as a relief effort were heavily laden and quite

The problem of classification comes to the fore with the current interpretations of ancient Mediterranean warships, which equate the sizes of the Egadi rams with dimensions of triremes (threes), rather than quinqueremes (fives) (Murray 2012: 38–68). These claims are based on the extensive study of the Actium Monument, which 36

Experimental Archaeology and the Contributory Reconstruction of a Roman Warship once displayed the prows of Marc Antony’s defeated fleet (Murray 2012). This presented a problem because Polybius (1.61) states that only quinqueremes were present at the Battle of the Egadi Islands. Diodorus’s (17.115.1–5) discussion of Hephastion’s tomb describes the gilded prows of quinqueremes lining the sides of the tomb, each prow measuring about 3.26 metres (Morrison 1995: 69). This further complicates the issue of classification and its relation to ship size and rower arrangement. Could ships of two different sizes hold the same classification if they operated with the same number of rowers per file? Or were ships differentiated based on their size alone, or did it depend on a combination of both rower arrangement and size?

the dimensions of the Egadi 10 ram (Table 1). Modelling began by importing a three-dimensional scan of the Egadi 10 ram, collected during the 2014 Battle of the Egadi Islands field season using a handheld scanner, into Rhinoceros (Figure 3). Measurements taken by hand were compared with the point cloud generated by the threedimensional scan, in order to adjust the model within 1 mm of variance. Since only small fragments of wood remained inside the ram, the reconstruction of the interior relied upon the shapes of the sockets. Since the ram was cast onto the bow timbers directly in order to provide the best possible fit, the interior contours provided enough substantial evidence to accurately determine the basic timber structures, as per casting analysis from another ram that was found near Athlit, Israel (Eisenberg 1991: 40–50). The dimensions of the timbers inside the ram are given at their extremities at the exit of the ram and at their tapering end points inside the ram (Table 1).

To determine a possible interpretation that factored both of these variables, this study applies the measurement parameters of the average known weight of materials and men with the Roman interscalmium measurement system, equivalent to two Roman cubits, or roughly 90 cm. The interscalmium marked the distance between tholepins, which were vertical pins used to hold oars in place and act as fulcrums for rowing (Shaw 1995: 164). This unit of measurement combined with contributory data from contemporary shipwrecks, shipsheds, iconography and literature will determine the basic proportions of the ship. Once a hull shape was developed, crew sizes, based on the number of oarsmen that could fit into the space between tholepins, could be tested to determine an acceptable load that would float the hull at a safe and operational depth.

Since the bow timbers provided the only direct archaeological evidence, it was then necessary to introduce secondary evidence to define a hull shape that conformed to the dimensions established through archaeological research. The hull form of the Egadi 10 warship was confined by constraints that would allow it to attain its intended performance characteristics. The maximum extent of the hull was guided by the sizes of 218 Carthaginian military harbour shipsheds, constructed during the third-second centuries BC, measuring between lengths of 27 to 35 metres with beams of 6.6 to 6.7 metres (Blackman and Rankov 2013; Hurst 1977, 1979; Hurst and Stager 1978). These shipsheds were selected due to their proximity in both space and time to the battle in 241 BC. The ships needed to be smaller than the maximum extents of the shipsheds and there needed to be enough room for crews to repair the hulls. As a result, a maximum length of 31 metres and beam of 5.5 metres was determined to guide the initial design of the hull. The two larger shipsheds at Carthage were omitted because they were likely used for larger capital ships. Although the Egadi 10 is most likely a Roman vessel that sank during the battle, Polybius’ (1.59) account of the Romans using a captured Carthaginian quadreme as a model for their ships permitted the use of the Carthaginian shipsheds for basic dimensions.

Reconstruction The Egadi warships were designed for the purpose of naval combat. Ramming warfare necessitated a specialised construction that produced a ship that was sleek, fast and adequately reinforced to withstand the physical strains of ramming and carrying crews of oarsmen. The ship needed to function properly while under sail and it needed to perform during pitched, rowed battles on the open sea. The ability of the bow to deliver and withstand ramming attacks was paramount to its ability as a warship. The hull design required it to be able to function during battle as a lightweight war galley, but it also needed to carry supplies, troops and make short overseas voyages.

The vessel’s draft was limited by its need for speed and manoeuvrability and by the placement of the ram, which roughly needed to sit at the waterline. The shape of the

Reconstruction of the ship shape begins with the extrapolation of bow timbers. This reconstruction utilises

Table 1. Bow Timber Dimensions Reconstructed from the Egadi 10 ram Interior Maximum Length Molded Dimension Molded Dimension Sided Dimension within Ram (cm) at Exit of Ram at Forward End at Exit of Ram (cm) (cm) (cm)

Sided Dimension at Forward End (cm)

Keel

70.0

14.5

4.0

13.7

10.0

Ramming Timber

54.5

20.5

16.0

52.0

4.0

Starboard Wale

61.0

19.9

16.0

8.0

2.0

Port Wale

60.5

20.5

16.0

11.0

2.0

Stem and Chock

57.2

35.5

4.0

16.3

4.7

37

Mateusz Polakowski

Figure 3. Modeling ram and bow timber in Rhinoceros (a) Initial lines and measurements taken by hand (b) Combining measurements with scan of Egadi 10 interior.

keel was limited by the 164-degree angle at its forward end as it became integrated into the bow structure. A slightly rockered keel would provide greater longitudinal strength by reducing the sloping angle of the keel at either end and was better equipped to withstand and disperse the force of shock resulting from ramming. Tests conducted in the 1980s and 1990s with the reconstructed Athenian trireme, Olympias, determined that a flat rise in the after portion of the keel was necessary to provide a surface upon which to haul the ship out of the water into shipsheds or along the shore (Morrison et al. 2000: 207). Therefore, the aft portion of the keel and sternpost were projected with a sharper and flatter rise. Coates (1995: 128) argues that a rowed galley needed to be long, slender and of shallow draft relative to its displacement in order to reach optimum performance. Hulls with low displacement and shallow drafts reduced the resistance created as the ship travelled through the water, while optimising oar system propulsion. On the other hand, lateral stability for sailing and accommodation of oarsmen required a sufficient breadth.

effective oared warships minimised the wetted area of the hull while maximising the number of effective rowing spaces, translating into long hulls with minimal waterline depths. These ratios are vital factors in keeping a ship’s effectiveness in the water. Since warships relied on speed and manoeuvrability, these factors would have been a priority to the ancient shipwright. They are found on the first century AD Oberstimm vessel (Hockmann 1990, 1991) and the Mainz vessels (Hockmann 1993), which were oared river galleys built by the Romans for their northern imperial conquests. These ratios are also present cross-culturally on oared galleys like the later medieval galleys of the Mediterranean (Alertz 1991: 144–48) and the Norse Skuldelev ships (Olsen and Crumlin-Pedersen 1978). These ratios needed to be maintained through the course of the hull design in order to produce a viable rowed warship. Lines drawing Combining the dimensions of the Carthaginian shipsheds and the necessary length to beam ratios, the Egadi 10 hull design was reconstructed with a 28.71 metre length and a 4.42 metre beam. This resulted in a 6.5:1 length to beam ratio and would allow the vessel to be safely berthed and repaired within the confines of the shipsheds. With the general hull constraints determined, first the stem was extrapolated from the dimensions taken from the interior of the cowl of the Egadi 10 ram. Since no archaeological evidence was available to determine the shape of the stem, the extension of the stem past the ram was conjectural and based on estimates taken from the Athlit ram, as well as the curving shape seen in numerous depictions of contemporary warships including the Carthaginian tomb relief, the Tiber Island ship and Roman coins depicting

Hull form A rowed warship is long, slender and shallow drafted relative to its displacement (Coates 1995: 128). The design of the Egadi 10 reconstruction needs to provide a practical balance of combat capability and performance under oar and sail while maintaining hull integrity and seaworthiness. The hull form of the Egadi 10 warship is confined by constraints that would allow it to attain its intended performance characteristics. The reconstruction relied on length to beam ratios between 6:1 and 7:1, based on the necessary hull coefficients determined by Coates (1995: 128–29) and Steffy (1991: 29–39). The most 38

Experimental Archaeology and the Contributory Reconstruction of a Roman Warship

Figure 4. Lines drawing of Egadi 10 hypothetical reconstruction.

the bows of warships (Casson 1991: 120–23, Figure 107; Morrison 1995: 67).

Table 2. Principal Hull Dimensions Lower Load Waterline

Section lines were projected every metre along the length of the keel rabbet, resulting in a total of 28 stations. The length and beam of the vessel was maintained by factoring in the necessity of allowing the maximum number of rowers within the dimensions provided by the Carthaginian shipsheds. For this reason, while the bow of the vessel was sleeker, the stern was designed with a rounder shape. This allowed for a greater number of rowers in the aft section while preserving the length to beam ratio of the hull. Once the keel, rabbet lines, and main wales were determined, section lines were constructed using shapes determined from the hull plans of contemporary Mediterranean construction. The wine glass shape, as it is commonly referred to, is the most common hull type found in the ancient Mediterranean. This shape is found on hulls as early as the Ma’agan Michael shipwreck (Steffy 1991: Figures 3–21), the Kyrenia Ship (Steffy 1991: Figures 3–32) and the larger ships of later periods like the Madrague de Giens (Pomey 1978: Plate 36).

Higher Load Waterline

Length Overall (m)

28.71

28.71

Maximum Beam (m)

4.42

4.42

Length on Waterline (m)

25.83

26.14

Beam at Waterline (m)

3.72

3.8

Draft (m)

1.25

1.35

Freeboard (m)

3.49

3.59

Table 3. Hydrostatic Calculations

All lines are shown to the interior of the hull planking. A total of 16 section line were spaced two metres apart, aside from two section lines at the bow and three at the stern which were spaced one metre apart. The turn of the bilge was constructed in order to align as closely to the main wales as possible to provide the most support to the area of the hull experiencing the greatest amount of longitudinal stress (Figure 4). Table 2 summarises the principal dimensions of the reconstructed Egadi 10 hull, while Table 3 presents the results of hydrostatic analyses of this hull shape.

Lower Load Waterline

Higher Load Waterline

Displacement in Salt Water (Metric Tons)

32.3

39.8

Waterplane Area (m2)

71.17

74.75

Prismatic Coefficient

0.62

0.62

Block Coefficient

0.32

0.34

Midship Coefficient

0.52

0.55

Length to Beam Ratio

6.95

6.5

Beam to Draft Ratio

3.64

3.4

Testing two different waterlines provided comparative data of two possible displacements, resulting in a better hypothesis of crew capacities. Additional waterlines were generated every 0.25 metres below the 1.25 metre load waterline, but only the two hypothetical load water lines were analysed for displacement purposes. The hull

The displacement of a vessel is the weight of the volume of water displaced by the underwater portion of the hull, and equals the weight of the ship at a given waterline. Two waterlines were set at the middle fin of the ram (1.25 metres) and at the top fin of the ram (1.35 metres), to produce more comparative hydrostatic results and a better understanding of the vessel’s displacement. Both waterlines allowed the ram to act as a cutwater while remaining in an optimal striking area against another hull.

As a ramming warship, the Egadi 10 needed to not only withstand the general stresses exerted upon the hull, such as hogging and sagging, it also required the structural integrity to deliver and withstand the shock generated during ramming battles. Shell-based construction, the main shipbuilding tradition of the Mediterranean during the third century BC, relied on tightly fitting mortise-andtenon joints to disperse shear forces along the length of the hull. Longitudinal forces generated during impact 39

Mateusz Polakowski Table 4. Mortise-and-Tenon Dimensions of Selected Vessels Shipwreck

Wood Type

Width x Thickness (cm)

Depth of Mortise (cm) Average Spacing (cm)

Ma’agan Michael

Oak

3.5 x 0.6

6.75

12.5

Kyrenia

Oak

4.3 x 0.6

9.0

12.0

Marsala

Oak/Maple

– x 0.9

-

-

Capestillo

Oak

5.0 x –

6.0

17.25

Athlit Ram

Oak

7.5 x 1.1

10.25

11.7

Madrague de Giens

Oak

5.6 x 0.7 / 8.3 x 1.4

9.4

15.0

Mahdia

Olive/Acacia

12.0 x 1.2

-

7.0

Caesarea

Oak

8.5 x 1.1

10.9

13.5

Anse des Laurons 2

-

7.5 x 1.0

12.7

11

would disperse along the length of the wales while the mortis-and-tenons would act effectively as a chain mail coat absorbing shock and dispersing the load across the entire hull (Morrison 1995: 131). Some of those standards are present in the mortise-and-tenon shell-based construction of Mediterranean hulls during the fourth and third centuries BC, and persist in the archaeological record through to the first century AD (Frey et al. 1978; Frost 1971, 1974; Gianfotta and Pomey 1981; Kahanov 1991; Linder 1989, 1992; Steffy 1985; Tchernia et al. 1978; Ucelli 1950). Archaeological, historical and iconographic examples were chosen according to building tradition and proximity to the third century BC (Crumlin-Pedersen and McGrail 2006: 55). The final projected dimensions are averages calculated for the purposes of this experiment and not intended as definitive representations.

Table 5. Tenon Dimensions of Egadi 10 Reconstruction Wood Type Width (cm)

Thickness (cm) Spacing (cm)

Oak

1.0

5.0–5.5

10.0

thousands of internal frames throughout the hull (Steffy 1985: 90). Therefore, their placement required careful calculation and marking. In order for the mortise-andtenon structure to work, tenons needed to be fashioned from hard oak, spaced regularly and tightly fitted in order to resist the crushing stresses exerted upon them. Treenails and pegs made of hardwoods expanded when impregnated with water, providing better grip and water-tightness. As shown in Table 4, mortise-and-tenon joinery was fairly uniform in ships from the ancient Mediterranean. Tenons were constructed of oak or other hard woods and were similar in width, thickness, length and spacing relative to the ship’s size. While contemporary tenons measure 4.2 cm wide, 0.6 cm thick, and are spaced approximately every 12.0 cm centre to centre, tenons from the Athlit ram are wider, thicker and more closely spaced. Based on these figures, a reasonable interpretation of the Egadi 10 tenon dimensions is about 5.0 to 5.5 cm wide, 1.0 cm thick and spaced 10.0 cm apart (Table 5). In addition to pegged mortise-and-tenon joints, the planking of ancient Mediterranean vessels was joined into strakes by diagonal scarfs. Scarfs were well-suited to resist the tension exerted on horizontal connections (Ulrich 2007:60).

Materials Theophrastus (HP 5.7.1–3) provides the most direct information for preferred shipbuilding timbers for triremes and long ships, which were fir and pine for planking, oak for the keel and ash, mulberry, or elm for the cutwater. Luckily, due to Rome’s recent conquest of the Italian peninsula, fir and pine were abundant in the regions of Etruria and Umbria. On the other hand, the North African coast supplied Carthage with an abundance of oak, Aleppo pine and cedar (Meiggs 1982: 14–142). Combining the wood fragments recovered from the Egadi 10 ram and historical accounts, it was possible to interpret the vessel’s component timbers with a high degree of confidence. The selected timber for components are discussed individually with each structural component.

Framing and the ram were held in place by metal fasteners and treenails. Bronze and copper nails were favoured because they were known to last longer than iron fasteners in the water (Vegetius 34). This is corroborated by archaeological evidence from the selected shipwrecks (Table 6). A copper fastener measuring 13.7 cm in length and 7.0 cm thick found inside the Egadi 10 ram and the remains of bolts 9 and 10 (Figure 5) support the use of bronze and copper fasteners. The pattern of fasteners used to hold the ram at the bow is based on the bolt holes in the bronze ram itself. The bolt holes in the Egadi 10 ram demonstrate that the ram was attached to the stem, wales and keel with five bolts per side. The fastening pattern of the Egadi 10 ram provides the best direct evidence of

Joinery and fasteners The integrity of Mediterranean long, shell-based ships depended on tightly fitting joinery and fastening, preventing planks from sliding against each other by carrying the shear forces exerted along the plane of the hull (Morrison 1995: 131; Ulrich 2007: 60). Mortise-and-tenon joints acted as a cohesive network providing strength and protection to the warship. Tenons held in place by oak pegs distributed lateral and longitudinal stresses, acting like 40

Experimental Archaeology and the Contributory Reconstruction of a Roman Warship assembly and spacing of fasteners of the bow timbers. However, without greater hull preservation of any ancient warship, it was not possible to determine the exact fastening pattern of the Egadi 10 hull.

Keel With maximum length determined using the 27 to 35  metre lengths of the Carthaginian shipsheds, the forward dimensions of the keel were determined from the shape of the keel trough and the 164-degree downward angle of the Egadi 10 ram. Extrapolating the basic dimensions of the keel’s size, shape and rabbet relied on the traditional construction of keels evident in ancient Mediterranean shipwrecks (Table 7). The moulded and sided dimensions, including the rabbet, were projected using measurements taken from the keel channel. As the keel extended aft, it tapered out slightly to a final sided dimension of 16.5 cm and 14.0 cm moulded. This provided additional strength and is reflected in the surviving keels of merchant vessels such as the Kyrenia Ship (Steffy 1994: 43).

Table 6. Fastener Types Found in Selected Vessels Shipwreck

Fastener Metal Treenails Type Wood Types

Clenched Nails Driven Through Treenails

Ma’agan Michael

Iron

-

Iron

Kyrenia

Copper

Pine

Copper

Marsala

Bronze

Oak

Iron/Bronze

Capestillo

Copper/Iron

-

-

Athlit Ram

Bronze

-

Bronze

Acqualadroni Lead/Copper Ram

-

-

Madrague de Giens

-

-

-

Mahdia

Copper

-

-

Caesarea

Copper Bronze Oak

Bronze

Nemi Barges

Copper

Copper

Anse des Laurons 2

Copper/Bronze -

Pine/Fir

During the trials of the Olympias, the builders noted that, in long ships, a rocker was only necessary towards the fore and aft sections of the keel (Morrison et al. 2000: 207). As mentioned earlier, a slightly rockered keel would provide greater longitudinal support for the ship. Tests conducted with the Olympias determined that a flat rise in the aft portion of the keel was necessary to provide a surface upon which to haul the ship out of the water into shipsheds or along the shore. Therefore, the aft portion of the keel and sternpost were projected with a sharper and flatter rise.

-

Based on the keel fragment found, the keel of the Egadi 10 was hewn from oak. Oak was fitting for a warship keel because it provided greater longitudinal stiffness and greater tensile strength during ramming attacks. The rabbet of the keel was projected out to 6.5 cm in order to fit the thicker garboards and its angle was based on the 78-degree angles of the Kyrenia Ship (Steffy 1994: Figures 3–25) and the Madrague de Giens wreck (Tchernia et al. 1978: Figure 11). It is very likely that the keel was originally constructed from two to three sections, but the

Figure 5. Copper fastener found in the interior of Egadi 10 (Image courtesy of RPM Nautical Foundation). Table 7. Keel Dimensions of Selected Vessels Shipwreck

Date (century) Wood Type

Length (m)

Sided (cm)

Molded (cm)

Cross Sectional Area (cm2)

Ma’agan Michael

5th/4th B.C.

Pine

8.25

11.0

16.0

176

Kyrenia

4th/3rd B.C.

Pine

9.33

13.0

20.3

263.9

Marsala

3rd B.C.

Oak

-

-

-

-

Capestillo

3rd/2nd B.C.

-

-

-

-

-

Athlit Ram

2nd B.C.

Pine

-

20.0

32.0

640.0

Acqualadroni

2nd B.C.

-

18.0

24.0

432.0

Madrague de Giens

1st B.C.

Elm

40.0

35.0

40.0

1400.0

Mahdia

1st B.C.

Elm

-

-

-

-

Caesarea

1st A.D.

-

-

-

-

-

Nemi Barges

1st A.D.

-

-

20.0

30.0

600.0

Laurons 2

2nd A.D.

-

-

16.0

20.0

320.0

41

Mateusz Polakowski reconstruction utilised a single timber, as the number of segments did not matter for the purposes of hydrostatic calculations and weight analyses (Table 8).

Planking and wales While Theophrastus’ (5.7.1–3) account of timber types sought in warship construction favoured the use of pine or fir, the best evidence of warship planking and pine wale construction remains the interior timbers of the Athlit ram (Steffy 1995: 10–28). Closely spaced mortise-andtenon joints would ensure that the stresses exerted along the lengths of the hull were distributed and supported accordingly. Thicker garboards and lower strakes provided necessary support and, in warships, would have served as defence against penetration by an enemy ram. Six planks were recovered from the Athlit ram, averaging 4 cm in thickness except for the bottom two planks which were 7.5 cm thick (Steffy 1983: 236). The main wales provided longitudinal support by taking the brunt of ramming impacts and dispersing them across a specifically designed network of planks and mortise-and-tenon joints. Secondary wales and upper wales were necessary for additional longitudinal support.

Stem and stern post The stem was extrapolated using the dimensions and angle of the cowling taken from the Egadi 10 ram (Table 9). At the point where the stem fit into the cowling, it had a trapezoidal shape with a rabbet to receive the upper planking, secondary wales and upper wales. The stem’s thickness was reconstructed to 17.0 cm based on the projections of the stem in the ram cowling with an extra centimetre of additional support for shocks sustained during ramming attacks. The sternpost is briefly discussed in connection with the design of the keel. Although the Marsala wreck is claimed to be a warship (Frost 1973), even the architects of the Olympias admittedly decided that the straight sternpost from the wreck was not an acceptable shape for the warship. Thus, a longer, gentler slope was needed to reduce the effects of sagging at launch and to improve the flow of water across the length of the hull (Morrison et al. 2000: 197). The sternpost followed the gentle upward curve of the keel in order to provide the needed surface area for hauling and launching the vessel. The basic shapes of both the stem and sternpost also rely on iconographic evidence that shows fairly standard shapes of the prows and sterns of ancient galleys (Casson 1971: Figures 119–33).

Using the collected information from hull remains, correlations between estimated original lengths, planking thickness and timber types produced potential structural dimensions (Table 10). The Ma’agan Michael (Kahanov 1991; Linder 1989, 1992), Kyrenia (Steffy 1985), and Capestillo (Frey et al. 1978) wrecks are fairly uniform in their length and softwood planking thickness. Both the double-planked Madrague de Giens (Gianfotta and Pomey 1981; Tchernia et al. 1978) and the Mahdia (Salies et al. 1994) hulls were constructed with hard deciduous woods with a plank thickness of about 4 cm. The single-planked Caesarea (Fitzgerald 1995) and the Nemi Barges 9 (Ucelli 1950: 155) employed soft woods with a plank thickness of about 9.5 cm. Shorter hulls, under 20 metres in length, tended to use softwood like pine for planking. The doubleplanked Madrague de Giens ship was constructed with hardwood inner and softwood outer planking layers, while both layers of the Mahdia ship were of hardwood. On the other hand, the Caesarea hull and the Nemi Barges were constructed with a single layer of softwood planking that was twice as thick as the individual planking layers of the Madrague de Giens and Mahdia vessels.

Table 8. Keel Dimensions of Egadi 10 Reconstruction Length Wood (m) Type

Sided Molded Angle at Exit Cross Sect. (cm) (cm) of Ram Area (cm2)

17.0

16.5

Oak

14.0

164

231.0

Table 9. Stem and Sternpost Dimensions of Egadi 10 Reconstruction Timber

Wood Type

Stem

Elm

Sternpost Elm

Length (m)

Max. Width Max. (m) Thickness (cm)

3.5

33.0

17.0

12.85

21.0

31.0

It is therefore possible to state that large hulls measuring 30 metres or more with a single layer were constructed with

Table 10. Planking Thicknesses and Vessel Lengths of Selected Shipwrecks Shipwreck

Date (century)

Planking Timber

Avg. Planking Thickness (cm)

Vessel Length (m)

Ma’agan Michael

5th/4th B.C.

Pine

4.0

14.4

Kyrenia

4th/3rd B.C.

Pine

3.7

14.0

Capestillo

3rd/2nd B.C.

-

4.5

20.0

Madrague de Giens

1st B.C.

Elm/Fir

4.0

38.0

Mahdia

1st B.C.

Elm

4.3

30.0

Caesarea

1st A.D.

Pine

9.4

40.0

Nemi Barges

1st A.D.

Pine

10.0

71.0

Anse de Laurons 2

2nd A.D.

Pine/Cedar

2.5

13.3

42

Experimental Archaeology and the Contributory Reconstruction of a Roman Warship that it would have been constructed with a single layer of planking, made of pine or fir averaging between 5.0 and 7.0 cm in thickness (Table 11). Garboards and bottom strakes were reconstructed with a 6.5 cm thickness in order to provide strength along the bottom of the hull. The wales were projected out to a maximum thickness of 12.0 cm and a maximum width of 21.0 cm and were constructed of pine following the dimensions of the Egadi 10 ram and the suggestions of using pine for warship planking by Theophrastus (HP 5.7.1–3) (Table 12). The larger size of the bottom stakes and the taper of the wales is also supported by the timbers of the Athlit ram (Steffy 1991: 6–15).

thick planking and mortise-and-tenons spaced no farther apart than their own width (Fitzgerald 1995: 117). Ships built with single layers of planking and lengths of about 30 metres or more were built with planking averaging about 9.7 cm thick. Double-planked hulls and hulls with lengths 25 metres or smaller were constructed with an average planking thickness of 4 cm or slightly less. This small sample of shipwrecks highlights previous correlative studies of plank thickness and timber usage relative to hull sizes (Fitzgerald 1995: 128–33). It seems unlikely for a warship intended to be fast and light to be constructed with a double-planked hull. Although it could increase the hull’s defence against a ramming attack, it would require greater resources and take longer to build a double-planked vessel. A double-planked ship would also require double sets of mortise-and-tenon joints, increasing the weight of construction materials which, in turn, would limit the ship’s speed and its manoeuvrability. Since there is no archaeological evidence of 9–10 cm thick planking in the mid-third century BC, potential planking thickness of the Egadi 10 warship needed to fall within a range of contemporary ship dimensions while factoring in comparative data regarding ship size to planking thickness ratios.

Frames Greek and Roman shell-based mortise-and-tenon built ships relied on a framing pattern of half frames alternating with floor timbers and futtocks, inserted along the length of the ship as the planking was built up. This framing system assisted the mortise-and-tenon joinery in dispersing hogging, sagging and transverse stresses exerted along the length of the hull. The Ma’agan Michael, Kyrenia and the Madrague de Giens hulls had framing sequences with paired floor timbers and futtocks alternating with half frames and top timbers.

If the Egadi 10 warship was approximately 28.7 metres in length with a beam of about 4.4 m, then it is probable

Frames were fastened to planking by clenched copper or iron nails driven through treenails from the outside (Table 13). Floor timbers spanned the keel and generally extended approximately to the turn of the bilge; from there, futtocks extended up to the sheer. Half-frames did not cross the keel and generally extended from the lower strakes up to the sheer. The Egadi 10 reconstruction follows this pattern. An estimate of 94 frame stations could fit along the length of the reconstructed hull.

Table 11. Planking Dimensions of Egadi 10 Reconstruction Timber

Wood Type

Width (cm)

Thickness (cm)

Strake Length (m)

Main Planking

Pine

21.0– 22.0

5.0

9.0

Garboards and Bottom Strakes

Pine

21.0

6.5

9.0

Calculating an average frame spacing between 25.0 and 35.0 cm for the selected merchant vessels, the Egadi 10 vessel was reconstructed with frames spaced at the lower end of 25.0 cm centre to centre (Table 14). The archaeological evidence of larger ships indicates they had proportionally larger framing that was more closely spaced, providing greater hull support (Fitzgerald 1995:

Table 12. Wale Dimensions of Egadi 10 Reconstruction Timber

Wood Type

Width (cm)

Thickness (cm)

Angle at Exit of Ram

Port Wale

Pine

21.0

12.0

166

Starboard Wale

Pine

21.0

12.0

166

Table 13. Framing Dimensions of Selected Vessels Shipwreck

Date (century)

Wood Type

Sided (cm)

Molded (cm)

Spacing (cm)

Ma’agan Michael

5th/4th B.C.

Pine

-

-

75.0

Kyrenia

4th/3rd B.C.

Pine

8.0

8.0

25.0

Marsala

3rd B.C.

Oak/Maple

-

-

-

Capestillo

3rd/2nd B.C.

-

16.0

10.0

15.0

Madrague de Giens

1st B.C.

Oak/Elm/Walnut

14.0

13.0

23.0

Mahdia

1st B.C.

Elm

20.0

20.0

60.0

Caesarea

1st A.D.

Pine

18.0

26.0

25.0

Nemi Barges

1st A.D.

Oak

30.0

40.0

50.0

Laurons 2

2nd A.D.

Pine/Oak

17.0

9.0

55.0

43

Mateusz Polakowski fastening patterns suggested a possible stringer about 8 cm thick (Fitzgerald 1995: 148–49).

Table 14. Framing Dimensions of Egadi 10 Reconstruction Frame Type Wood Type

Sided (cm)

Molded (cm)

Length (m)

Spacing (cm)

Floor Timbers

Pine

16.0

10.0

2.5

25.0

Half Frames Pine

16.0

10.0

1.13

25.0

Futtocks

12.0

8.0

1.6

25.0

Pine

Interior structures The last issue facing the completion of the vessel’s reconstruction was the interpretation of the interior structures. These included the rigging elements, crew accommodations and the hypozomata (a cable running the interior length of the ship to alleviate hogging and sagging stresses) (Morrison et al. 2000: 207–27). The crew accommodations included rowing benches, providing adequate space for efficient rowing and appropriate deck space for commanding officers and marines. Some of the best archaeological evidence of Roman deck structures comes from the Anse des Laurons 2 ship (Gassend et al. 1985: Figure 17c). A basic rowing station assembly was designed based on the Olympias rowing models (Morrison et al. 2000: 198–99). The rowing station had a width of 49.0 cm according to the interscalmium, including oars, benches for the rowers, 10 cm stanchions and 5 cm stringers to support the deck structures, and an outrigger to allow for the use of long oars. Since there is no archaeological evidence that could provide any better projections, the remaining interior structures of the Egadi 10 were modelled on the Olympias reconstruction (Figure 6) (Morrison et al. 2000: 194–98).

145). In order to provide additional support and to reinforce the hull against the violent nature of ramming warfare, a closely spaced pattern of thicker frames would have been necessary. On the other hand, if the framing was too thick or too closely spaced the vessel would be overweight and ineffective. Also, if the frames were placed too close together it would restrict space for rowers and oar ports. The reconstructed size and spacing seems a reasonable interpretation, given these limiting factors. In addition to floor timbers, futtocks and half frames, there is good evidence that stringers alternating with thinner ceiling planking provided additional longitudinal support. The Madrague de Giens ship has preserved stringers, alternating with thinner ceiling planking, measuring about 20.0–30.0 cm wide and 6.0–10.0 cm thick. A stringer preserved at the turn of the bilge measures 12.5 cm in thickness (Pomey 1978: 84). A timber found on the Capestillo wreck, measuring 30.0 cm wide and 6.0 cm thick, may have served as a stringer; however, due to the depth and conditions of the excavation, this is not certain (Frey et al. 1978: 293–94). The Nemi barge’s stringers, which also alternate with thinner ceiling planking, measure an average of 25.0 cm thick and 31.0 cm wide (Ucelli 1950: 155). The Caesarea wreck did not yield direct evidence of stringer preservation; however, study of

The reconstructed mid-ship sections are representations of the three-dimensional model used to determine the weight of the construction materials. The final hull had a total of six wales, three per side including the main wales projected from the Egadi 10 ram wale pockets. It had a total of thirty-two planks, sixteen per side with a total of twenty planks being of the thicker bottom planking dimensions

Figure 6. Reconstructed mid-ship section based on the Olympias design, showing hypothesized interior hull structure and projected superstructure.

44

Experimental Archaeology and the Contributory Reconstruction of a Roman Warship and twelve being of the thinner upper dimensions. Four stingers supported deck structures that included areas for the rowers. The decks structures, including stanchions, outrigger and projected superstructure were modelled using the designs developed for the Olympias trials (Morrison et al. 2000: Figures 54–57). Designing the craft any further at this point would result in greater conjecture and diminish the practicality of the results. Instead, individual questions and theories about the vessels found at the Egadi site should be addressed specifically, rather than attempting a general reconstruction, which would inevitably involve repeating the experiments conducted during the Olympias trials.

Table 16. Calculated Weights from Orca3D Analysis

Assigning weight properties Assigning proper weights to the structures of the reconstructed Egadi 10 vessel was the first step in determining the possible numbers of rowers that could effectively operate this hypothetical hull shape. A determined hull weight could be subtracted from the displacement tonnage and the remaining displacement was used to test hypothesised rower numbers. Using the remaining displacement as the unchanging factor, or control, the number of rowers became the variable. The reconstruction calculated the weight of the hull by estimating the weight of structural materials and subtracting that total from the displacement. By combining the standard weight of a Roman man and estimating the water needed to sustain him, it was then possible to compare various arrangements to present the best hypothetical size of the ship’s crew.

605.0

Oak (Quercus)

675.0

Pine (Pinus halepensis)

430.0

Silver Fir (Abies alba)

435.0

Cedar (Cedrus libani)

530.0

0.426

Stem

0.091

Ramming Timber

0.008

Chock

0.001

6 Wales and 4 Stringers

2.22

Planking

7.6

Frames

5.0

Egadi 10 Ram

0.163

Equipment and Fittings

4.0

Total Weight of Hull Components

19.5

For the purposes of weight calculations, rowers were represented by a square meter box following the interscalmium measurements provided by Vitruvius (1.2.4) and known cross-culturally for oared ships (Alertz 1991: 144–48; Coates 1995: 128–29; Olsen and CrumlinPedersen 1978). The average weight of a Roman male was set at 60 kg based on anthropological work of Roman burials and dietary habits (Roth 1999; Woodson 1981: 715, 737). With an interpreted vessel size and weight of timbers, the weights of rowers and the water necessary to

Table 15. Weights of Timbers Used in the Egadi 10 Reconstruction (based on Tsoumis 1991:111–127) Elm (Ulmus minor)

Keel and Sternpost

Maintaining oarsmen during operations required adequate hydration and caloric intake. During the Olympias rowing trials, researchers determined that 1 litre of water per rower per hour was needed to sustain basic rowing (Morrison 1995: 130). Therefore, the construction of a galley needed to factor in the weight of water and food required to sustain rowers. This also meant that an increase in rowers meant an additional weight of water and food needed per rower. Estimating that a warship that was cleared for battle was carrying enough water for four hours of action meant, at minimum, four litres of water were needed per rower in order to sustain a basic level of hydration. The circumstances of the battle provided the Romans with an advantage. While the Carthaginian ships were laden with supplies and weary from the sail from northern Africa, the Roman only needed to dispatch their ships once the Carthaginians were close enough to the islands. This meant that the Roman ships would have been weighted only with supplies needed for this single day of battle.

The calculation of equipment and fittings posed a problem since there is a lack of evidence regarding cargo size and

Weight (kg per cubic meter)

Individual Weights of Components (Metric Tons)

the necessities on-board warships (Table 16). Factored into these supplies was a general estimate of copper, bronze, iron and wood fasteners, as well as rigging. The weight of the frames was estimated by designing a pair of floor timbers and half frames with futtocks in Rhinoceros around the widest section of the hull. Calculations based on frame sizes and spacing estimated 94 frame stations along the length of the hull. Together all frame pairs within the hull resulted in a total of about 5.0 tons of timber. Each individual reconstructed rowing station weighed an average of 178.0 kg.

The Orca 3D software allows solids to be assigned properties including weight per cubic metre, volume, and cost estimates. For the purposes of this study, cost estimates were omitted because the focus of the study is the performance of the ship rather than the economic factors that constrained its construction. Weights were assigned to the keel, stem, sternpost, planking, ramming timber, framing and rower structure according to the wood interpreted for each type (Table 15). Using dried and seasoned weights of timbers was appropriate considering that the effectiveness, speed and efficiency of a warship depended on it not becoming heavily waterlogged (Morrison et al. 2000: 179–90).

Timber Type

Component Timbers

45

Mateusz Polakowski keep them alive, along with their corresponding rowing stations, became the variables used to determine the possible sizes of the crew. The rowing arrangements were based on designs developed by Coates (1995: 138). They were not meant to be definitive rowing arrangements or indicate a specific categorisation of the hull; instead they provided basic organisation structures for the interior of the hull.

the design and construction of a warship, the shipwright would need to account for the size of the crew and the position as well as the angle of rowing stations in order to produce an effectively-rowed warship. If the classification system of ancient warships was based on rowers, then these arrangements would represent triremes, quadremes and quinqueremes. Based on previous research (Coates 1995: 127–41), these arrangements were selected to provide a constant structure for the varying amounts of rowers. In order to fit the most potential rowers along the length of the vessel, partial rower stations were used towards the bow and stern where the narrowing of the hull no longer allowed for full rowing stations.

Rowing arrangements The aim of the reconstruction is to propose a possible size of the crew of the Egadi 10 warship and see how it compares to Polybius’s accounts. Polybius (1.61) claims that the warships that took part in the Battle of the Egadi Islands were quinqueremes, rowed by crews of 300 men. However, according to studies conducted on the Actium Monument (Murray 1989, 2012), the Egadi rams seem to belong to much smaller vessels. The monument, commissioned by Octavian in 29 BC, commemorated his naval victory over Marc Antony and Cleopatra. The monument held between 36 and 37 rams of varying sizes placed in sockets at the base of the temple running along the façade (Murray 2012: 38–39). The Athlit ram, which is about twice the size of the Egadi rams, fits onto the smaller sockets of the monument (Murray and Petsas 1989), strongly suggesting that the Egadi rams would have fallen into a much smaller class of vessels. In order to produce a comprehensive analysis, three, four and five rower arrangements were selected to represent potential rowing systems.

This maximised the use of ship’s hull space, allowing for more files of oarsmen to row effectively during battle when speed was of the utmost importance. It was necessary to test the possibility of these classification systems based on rower arrangement in order to determine what kind of ship the Egadi 10 potentially was. Since the shape and size of the reconstructed Egadi 10 hull remained constant, the number of rowers became the variable used to determine which arrangement was the best possible option. In order to characterise the oarsmen for weight calculations, each rower was represented by a 1 m2 box with the assigned 60 kg weight of an average Roman male (Figure 7) (Welsh 1988: 156). This provided an easy way to adjust the rower arrangements while maintaining the correct distance between rower stations based on the interscalmium. The first arrangement tested was a trireme, which had three rowers per station. This arrangement showed that a total of 120 rowers could fit into the confines of the hull, divided among 17 full rowing stations, 3 partial stations at the bow and 2 partial stations at the stern. These 120 rowers would add an additional weight of about 7.0 metric tons, while their rowing benches added another 4.0 metric tons. Testing a quadreme arrangement (four rowers per station) with 17 full rowing stations, 3 partial stations at the bow and 2 partial stations at the stern yielded a total of 160 rowers, equalling an average weight of 9.6 metric tons. Rowing benches added an additional 4.5 tons. Testing Polybius’s claim of quinqueremes (five rowers per station) gave a total of 192 rowers divided among 14 full rowing stations, 5 partial stations at the bow and 3 partial stations at the stern. This would equal a weight of about 11.6 metric tons, with rowing benches adding another 5.0 metric tons (Table 17).

‘Oarsmen geometry is dictated by the vital statistics of the oarsmen and not by the hydrodynamics’ (Welsh 1988: 156). Since the Egadi 10 vessel was an oar-powered galley, it required a design that factored the ship’s size proportional to the amount of propulsion produced by its oarsmen. Therefore, galleys like the Egadi 10 could only be as effective as the power provided by the crews operating them. The physical demands of the rowers constrained the development and construction of oar-powered warships. Rowers were the human engine that powered the Egadi 10 warship. A staggered arrangement of rowers was necessary to fit each individual rower while providing adequate space for efficient rowing. Experimentation of rowing spaces and human capacities was conducted during the Olympias trials. To date, these results still provide the best insight into potential rowing systems and their physical constraints (Morrison 1995: 63; Morrison et al. 2000: 211–30).

Displacement analysis

The placement of rowers resulted in higher centres of gravity that diminished optimal speed, but their location was necessary in order to provide rowing stations with appropriate rowing angles. Oar lengths of 9 (4.41 metres) and 9.5 (4.66 metres) cubits, based on the Naval Inventories from Piraeus, are most effective at a minimum 30-degree horizontal angle, limiting the angle and position of rowing benches above the waterline (Shaw 1995: 163). During

The next step was to test each hypothesised rowing arrangement to see which one provided the best interpretation for the Egadi 10 warship. This was done by analysing the displacement that each rowing arrangement resulted in when added to the weight of the reconstructed hull, equipment and supplies. To be accurate, the interpretation needed to sink the hull to a displacement 46

Experimental Archaeology and the Contributory Reconstruction of a Roman Warship

Figure 7. Rowing systems (a) three rowers per station (b) four rowers per station (c) five rowers per station. Table 17. Weight Estimates of Rower Arrangements Rowing Stations

Number of Rowers

Rower Weight (Metric Tons)

Rowing Bench Weight (Metric Tons)

Total Water Needed for 4 Hours of Rowing (Metric Tons)

Total Weight (Metric Tons)

3 Rowers per Station

120

7.0

4.0

0.48

11.5

4 Rowers per Station

160

9.6

4.5

0.64

14.2

5 Rowers per Station

192

11.5

5.0

0.77

17.3

Table 18. Weight and Displacement Calculations at the 1.25 m Waterline (LWL) Rowing Stations

Weight of Hull (metric tons)

Weight of Rowers and Corresponding Rowing Benches (metric tons)

Total Weight (metric tons)

1.25 m LWL Hull Displacement (metric tons)

Remaining Weight to Match Intended 1.25m Displacement (metric tons)

3

19.5

11.5

31.0

32.3

1.3

4

19.5

14.2

33.7

32.3

-1.4

5

19.5

17.3

36.8

32.3

-4.5

indicate the versatility of these warships. It was designated as a secondary waterline because the ram would have been fully submerged, reducing the performance of the ship in the water (Table 19). It also provides an estimate of the additional weight the ship could carry if it were transporting additional supplies or troops.

level that would place the ram at or very close to the intended load waterline (LWL). This concept can be thought of using the following equation: Weight of Hull, Equipment and Supplies + Weight of Rowers = Displacement at LWL

The resultant weights revealed that only two of the three proposed rowing arrangements were viable in a vessel of the proposed dimensions. More precisely, it provided an average number of men that could fit in a hull of the proposed size while retaining hull stability and performance. Arrangements of three and four rowers per station fell within an acceptable weight variation from the displacement of the vessel at the 1.25 metre LWL. Overweighting the vessel with 192 rowers in a quinquereme arrangement would sink the vessel too low and reduce its ability to ram effectively.

In this equation, total weight of the ship and displacement at the load waterline serve as constants against which to test variable weights of rowing arrangements. Results Once each rowing arrangement was designed, the weights of rowers and their corresponding oar stations were calculated using the Orca3D weight calculation tool. Both the 1.25 metre LWL and the 1.35 metre LWL were tested, but the 1.25 metre LWL was designated for the result. For this warship to be functional, it needed to be versatile; it needed to perform optimally during naval engagements, but it also needed to be able to sail with supplies and carry extra troops when not in combat. The 1.25 metre LWL represents the ship if it were stripped for battle carrying only its crew and supplies to sustain it for a day of battle, as was the situation of the Roman fleet at the Battle of the Egadi Islands (Table 18). The 1.35 metre LWL provides comparative evidence and can be used to

Discussion Based on the figures projected, the Egadi 10 and the other Egadi rams do not coincide with the ships described in Polybius’s descriptions of the quinqueremes with 300-member crews. Displacement analysis of the reconstructed hull shape reveals that the Egadi 10 warship would not have been large enough to house the number 47

Mateusz Polakowski Table 19. Weight and Displacement Calculations at the 1.35 m Waterline (LWL) Rowing Stations

Weight of Hull (tons)

Weight of Rowers and Corresponding Rowing Benches (metric tons)

Total Weight (metric tons)

1.35 m LWL Hull Displacement (metric tons)

Remaining Weight to Match Intended 1.35m Displacement (metric tons)

3

19.5

11.5

31.0

39.8

8.8

4

19.5

14.2

33.7

39.8

6.1

5

19.5

17.3

36.8

39.8

3.0

Works Cited

of rowers described by Polybius. Based on the combined weights of construction materials and rowers, the Egadi rams are more likely to have come from triremes or possibly quadremes, but not from quinqueremes. Although the weights of equipment, fittings, food and water could affect these results, both the three and four arrangements provided enough variable tonnage to make these results acceptable.

Primary Sources DIODORUS. 1970. Historical Library. Trans. Oldfather, C. H. Harvard University Press: Cambridge. HERODOTUS. 1988. The Histories. Trans. Grene, D. The University of Chicago Press: Chicago, IL. POLYBIUS. 1962. Histories. Trans. Shuckburgh, E. S. Indiana University Press: Bloomington.

Is there any way to reconcile this interpretation with Polybius’ statements? Although Polybius does not mention that any other vessel types took part in the battle, it is possible that smaller ships were omitted from the record in order to highlight the prominence of the bigger, more expensive quinqueremes. It is also very likely that the large fleets operating throughout the course of the war had contingents of support units which are omitted from the general overview provided by Polybius. The simplest explanation would be that Polybius was writing much later, and made either an anachronistic error or a deliberate attempt to enhance the histories he was writing about the Romans.

THEOPHRASTUS. 1953. The characters of Theophrastus. Trans. Edmonds, J. M. Harvard University Press: Cambridge, MA. VEGETIUS. 1996. Epitome of Military Science. Trans. Milner, N. P. Liverpool University Press: Liverpool. Secondary Sources ALERTZ, U. 1995. ‘The Naval Architecture and Oar Systems of Medieval and Later Galleys’. In R. Gardiner and J. Morrison (eds.), The Age of the Galley: Mediterranean Oared Vessels since pre-Classical Times. Conway Maritime Press: London. Pp. 142–62.

There is no claim that the current reconstruction attempt represents the Egadi 10 warship with complete accuracy. Instead, this contributory or minimal reconstruction should be viewed as a vehicle for exploring questions regarding ancient warship size and shape, and as a hypothesis to be tested against future archaeological evidence. Additional research is needed into the inter-related constraints that framed the construction of the Egadi 10 warship. More research is needed to understand warship building tradition, the technology available to the shipwrights (tools, metal working, timber carving techniques) and the economic requirements necessary to support such a large building program.

BLACKMAN, D. and RANKOV, B. 2013. Shipsheds of the Ancient Mediterranean. Cambridge University Press: Cambridge. CASSON, L. 1971. Ships and Seamanship in the Ancient World. Princeton University Press: Princeton. CASSON, L. 1991. The Ancient Mariners: Seafarers and Sea Fighters of the Mediterranean in Ancient Time 2nd Ed. Princeton University Press: Princeton, New Jersey. COATES, J. 1995. ‘The Naval Architecture and Oar Systems of Ancient Galleys’. In R. Gardiner and J. Morrison (eds.), The Age of the Galley: Mediterranean Oared Vessels since pre-Classical Times. Conway Maritime Press: London. Pp.127–41.

A reconstruction is incomplete unless it produces a link to analyse the society that built, operated and utilised the ship and must, therefore, be studied through a context of interrelated constraints that continuously influenced the construction and operation of this warship. For a reconstruction to be justified it must produce a finite series of interpretations judged by the standards of their prevailing contemporary shipbuilding techniques (Crumlin-Pederson and McGrail 2006: 57). The present hypothesis introduces a possible hull shape using contributory evidence to suggest potential size and classification of vessels that have been found at the Egadi Island’s Battle Site.

COATES, J., MCGRAIL, S., BROWN, D., GIFFORD, E., GRAINGE, G., GREENHILL, B., MARSDEN, P., RANKOV, B., TIPPING, C., and WRIGHT, E. 1995. ‘Experimental Boat and Ship Archaeology: Principals and Methods’. International Journal of Nautical Archaeology 24.4: 293–301. CRUMLIN-PEDERSEN, O. and MCGRAIL, S. 2006. ‘Some Principals for the Reconstruction of Ancient Boat Structures’. International Journal of Nautical Archaeology 35.1: 53–57.

48

Experimental Archaeology and the Contributory Reconstruction of a Roman Warship EISENBURG, S. 1991. ‘Metallurgical Analysis of the Ram’. In L. Casson and J. R. Steffy (eds.), The Athlit Ram. Texas A&M University Press: College Station, Texas. Pp. 6–39.

MARSDEN, P. 1993. ‘A hydrostatic study of a reconstruction of Mainz Roman ship 9’. International Journal of Nautical Archaeology 22.2: 137–41.

FITZGERALD, A. 1995. A Roman Wreck at Caesarea Maritima, Israel: A Comparative Study of Its Hull and Equipment. Ph. D. Dissertation, Texas A&M University, College Station, Texas.

MORRISON, J. 1995. ‘Hellenistic Oared Warships 399– 31 BC’. In R. Gardiner and J. Morrison (eds.), The Age of the Galley: Mediterranean Oared Vessels since preClassical Times. Conway Maritime Press: London. Pp. 66–77.

FREY, D., HENTCHEL, F. D., and KEITH, D. H. 1978. ‘Deepwater archaeology. The Capestillo Wreck Excavation, Lipari, Aeolian Islands’. International Journal of Nautical Archaeology 7: 279–300.

MORRISON, J. S., COATES, J. F., and RANKOV, N. B. 2000. The Athenian Trireme: The History and Reconstruction of an Ancient Greek Warship 2nd Ed. Cambridge University Press: Cambridge.

FROST, H. 1973. ‘First season of excavation on the Punic wreck in Sicily’. International Journal of Nautical Archaeology 2.1: 33–49.

MURRAY, W. M. 2012. The Age of Titans: The Rise and Fall of the Great Hellenistic Navies. Oxford University Press: New York.

FROST, H. 1974. ‘The Punic wreck in Sicily: 1. Second season of excavation’. International Journal of Nautical Archaeology 3.1: 35–54.

MURRAY, W. M. 2014. ‘Observations on the Ship Class of the Egadi Ram’. Paper presented at the 115th Annual Meeting of the American Institute of America, Chicago, IL.

GASSEND, J. M., LIOU, B., and XIMENES, S. 1985. ‘L’epave 2 de l’anse des Laurons (Martigues, Bouchesdu-Rhone)’. Archaeonautica 4: 75–105.

MURRAY, W. M. and PETSAS, P. M. 1989. ‘Octavian’s Campsite Memorial for the Actian War’. Transactions of the American Philosophical Society 79.4: 1–172.

GIANFROTTA, P. A. and POMEY, P. 1981. Archaeologia subacquea: Storia, techniche, scoperte e relitti. Mondadori: Milan.

OLSEN, O. and CRUMLIN-PEDERSEN, O. 1978. Five Viking Ships from Roskilde Fjord, translator Barbara Bluestone. Vikingeskipshallen: Roskilde.

GISELA HELLENKEMPER, S., von PRITTWITZ UND GAFFRON H.H., and BAUCHHENΒ G. 1994. Das Wreck: Der antike Schiffsfund von Mahdia. RheinlandVerlag GmbH: Kӧln.

SHAW, J. T. 1995. ‘Oar Mechanics and Oar Power in Ancient Galleys’. In R. Gardiner and J. Morrison (eds.), The Age of the Galley: Mediterranean Oared Vessels since pre-Classical Times. Conway Maritime Press: London. Pp. 163–71.

HÖCKMANN, O. 1990. ‘Roman Danube Vessel from Oberstimm, Germany’. In H. Tzalas (ed.), Tropis II: Second International Symposium on Ship Construction in Antiquity. Hellenic Institute for the Preservation of Nautical Tradition: Delphi. Pp. 215–24.

STEFFY, J. R. 1985. ‘The Kyrenia Ship: An Interim Report on Its Hull Construction’. American Journal of Archaeology. 89.1: 71–101.

HÖCKMANN, O. 1991. ‘Roman Danube Vessel from Oberstimm (Germany) as an Example of ‘ShellFirst’ Construction. Carvel Construction Technique: Skeleton-First, Shell-First’. In H. R. Reinders and K. Paul (eds.), Proceedings of ISBSA 5. Oxbow Monograph: Oxford. Pp. 14–18.

STEFFY, J. R. 1991. ‘The Ram and Bow Timbers: A Structural Interpretation. In L. Casson and J. R. Steffy (eds.), The Athlit Ram. Texas A&M University Press: College Station, Texas. Pp. 6–39. STEFFY, J. R. 1994. Wooden Shipbuilding and the Interpretation of Shipwrecks. Texas A&M University Press: College Station, Texas.

HÖCKMANN, O. 1993. ‘Late Roman River Craft from Mainz, Germany’. International Journal of Nautical Archaeology 22.2: 125–35.

TCHERNIA, A., POMEY, P., HESNARD, A., COUVERT, M., GIACOBBI, M. F., GIRARD, M., HAMON, E., LAUBENHERIMER, F., LECAILLE, F., CARRIERGUILLOMET, A., CHENE, A., GASSEND, J. M., REVEILLAC, G., and RIVAL, M. 1978. L’Epave Romaine De La Madrague De Giens. Centre National de la Recherche Scientifique: Paris.

HURST, H. 1977. ‘Excavations at Carthage 1976: Third Interim Report’. The Antiquaries Journal 57.2: 232–61. HURST, H. 1979. ‘Excavations at Carthage 1977–8. Fourth Interim Report’. The Antiquaries Journal 59.1: 19–49.

UCELLI, G. 1950. Le navi di Nemi, 2nd ed. Libreria dello Stato: Rome.

HURST, H. and STAGER, L. E.1978. ‘A Metropolitan Landscape: The Late Punic Port of Carthage’. World Archaeology 9.3: 334–46.

WELSH, F. 1988. Building the Trireme. Constable and Company Limited: London.

KAHANOV, Y.1991. ‘An Investigation of the Ma’agan Michael Ship’. C.M.S. News. 18: 5–6.

49

Mateusz Polakowski ZEEV, A. B., KAHANOV, Y., TRESMAN, J., and ARTZY, M. 2009. The Ma’agan Mikhael Ship Volume III: A Reconstruction of the Hull. Leon Recanati Institute for Maritime Studies, University of Haifa: Haifa, Israel.

50

5 Managing the Threat: A Maritime Archaeological Study of the Island of Menorca as a Key Ancillary in the Roman Mediterranean Margaret A. Amundson Universitat Rovira i Virgili [email protected] Abstract: This is a maritime archaeological study of the Roman period in Spain’s Balearic Island of Menorca in the Mediterranean Sea. The aim of this paper is to affirm that Rome’s annexation of Menorca’s ports and the pacification of its waters in 123 BC can be best understood as a key component in a region-wide initiative to ensure the security of critical western maritime routes and greater geopolitical stability in the Western Mediterranean. While imperialistic in scope, Rome’s objective concerning Menorca was not territorial, a fact which had bearing on its actual involvement in the island. Therefore, this paper also challenges the common conceptualisation of the Roman ‘conquest’ of the island, as well as the continued use of such terminology in describing that intervention. Finally, some of the assertions presented in this study may contribute toward discussions about defensive imperialism, globalising processes in the ancient world, and the roles of islands within the context of a globalising society. Key words: Menorca; piracy; Roman trade; globalisation; island archaeology; maritime archaeology

The strategic utility of the Balearic Islands in controlling western Mediterranean maritime routes was first recognised by Rome during the Punic Wars when a number of different campaigns to establish control over the islands were staged by both Rome and Carthage (Zucca, 1998). Concerning the island of Menorca, the object of both belligerents in the conflict was, at different times, to acquire and maintain military control of the island’s immense southern port. At the conclusion of the Punic Wars in 146 BC, a treaty was signed in which Carthage relinquished its dominion over the Balearic Islands to Rome (Liv. 30.30.24–25). However, it was not until piracy in the islands became a substantial threat to its maritime interests that Rome finally intervened in their autonomy. Given Menorca’s limited resources for export, and its poor connectivity and commercial prospects, it must be assumed that it held little inherent value for Rome. Nevertheless, the Romans deemed it necessary to establish and maintain a permanent military presence there. Therefore, the intention of this article is to ascertain the nature of Rome’s interest in Menorca, and ultimately illuminate its role as a key ancillary in the Roman epoch.

Empire was, in fact, far more complex, a process spanning as many as two centuries or more. Additionally, the specific motivations behind this intervention with respect to each of the three main islands in the archipelago may have differed significantly enough to warrant individual consideration of each island based on its strategic value, or threat, to the expanding Roman Republic. A more individualised approach will help reframe questions arising from archaeological evidence, which has revealed unexpected differences in the processes of Romanisation among the Balearic Islands, most especially in the case of Menorca. These differences owe much to the islands’ physical locations, nautical conditions and coastal morphologies. Under these premises, existing narratives about the Roman epoch in the Balearics have proven problematic in their scope, orientation and their continued use of the term, ‘conquest’. Therefore, another objective of this article is to discuss the suitability of such terminology in Roman archaeological studies of Menorca.

The incorporation of the Balearic Islands into the Roman Empire is typically regarded as a singular event identified as the ‘Balearic conquest’. While there was a military campaign launched in 123 BC under the command of Quintus Caecilius Metellus on behalf of Rome to secure the archipelago, Menorca’s integration into the Roman

The Balearic Islands form the eastern limits of the Baetic System, a mountain range which extends from the southwest to the northeast along the southern Mediterranean coast of the Spanish mainland. Easternmost in the archipelago, and situated centrally within the western Mediterranean, the island of Menorca was uniquely positioned to intercept

Geographical position and maritime conditions of Menorca in the Balearics

51

Margaret A. Amundson maritime traffic or to provide convenient anchorage for repairs or supplies along critical shipping routes between Rome and Hispania, particularly via the Strait of Bonifacio, and between the coasts of Massalia and the North African provinces (Guerrero Ayuso 1993; Keay 1988).

coastal morphologies. Although southern wind, known as ‘sirocco’, blows most strongly from the direction of the Sahara in summer, and is capable both of causing rough seas as well as depositing surprising amounts of sand and dust on the island (and even as far as the Spanish mainland), its frequency and severity hold no comparison with the tramuntana.

In addition to its central location, Menorca boasts the largest natural port in the Mediterranean. At nearly six kilometres in length, twelve hundred metres in width, and with a depth of up to thirty metres, the Port of Mahon would easily have accommodated an entire ancient merchant or naval fleet. Today, it hosts cruise ships and large freighters. Despite the island’s apparently convenient location and the attractiveness of this major port, the distance between Menorca and the nearest mainland is not insubstantial. This, combined with frequently unfavourable maritime conditions, seems to have kept Menorca from reaching its potential as a primary maritime network hub.

The effect of the weather pattern associated with the tramuntana is twofold: firstly, it renders nearly the entire northern coastline of Menorca largely inhospitable to ships seeking commerce or shelter there; and secondly, the resulting wave action has carved out numerous secluded inlets and coves from dramatic rocky cliffs. Given the need for specific knowledge about various marine hazards and safe anchorages in order to make use of that coast, it would have been unprofitably risky for merchants to try to negotiate this coastline on any regular basis, and for those seeking shelter from rough seas it could have been considered as a landing place only as a final option (Guerrero Ayuso 2010). The safest port along this coast is located at Sanitja, sheltered somewhat by the Cap de Cavallería and the Isla d’els Porros. These conditions undoubtedly contributed to Rome’s decision to use this location as a military port in order to secure the northern coastline, though for the reasons indicated above, the corresponding settlement never evolved into a commercial centre as did the Menorcan cities of Iamo and Magon (modern Ciutadella and Mahon, respectively), which could be reached more safely and easily from a southerly approach (see Figure 1).

Typically in winter, though occurring often and unpredictably throughout the year, hard North Atlantic winds blow southward over the north of Spain toward the Mediterranean. The Pyrenees normally provide a significant degree of shelter to Formentera, Ibiza and Mallorca from these winds and the tempestuous weather systems which often accompany them. This should be qualified by noting that, while Mallorca does enjoy generally calmer weather than Menorca, the north-easternmost coast of Mallorca is close enough to the edge of this sheltered Pyrenean zone that it does receive some of the high winds which often curve around the Cape of Creus, resulting in occasional shipwrecks such as the well-known wreck of Sant Jordi. Menorca, on the other hand, lies fully outside of the Pyrenean lee, and therefore receives the full impact of every North Atlantic system which forces its way past the eastern edge of the range. Such weather systems tend to grow in intensity as they funnel down through the southern coast of France and arc past the Cape of Creus on their collision course with Menorca. The effect of this northern wind, the ‘tramuntana’, is evident in a simple visual comparison of Menorca’s northern and southern

The pre-Roman maritime economy of Menorca Following the prehistoric colonisation of Menorca by peoples of unknown origin who later came to be identified with the Talayotic Culture, the island’s history was influenced largely by early trade contacts between those island settlers and Greek, Phoenician and later Punic merchants. As the winds of the tramuntana blow significantly less strongly toward the western islands of

Figure 1. Location of Menorca within the western Mediterranean, and its three Roman cities of Iamo, Sanisera and Magon (adapted from Google Earth 2017).

52

Managing the Threat the Balearics, the resulting calmer seas made the island of Ibiza, then ‘Ebusus’, a more ideal port of call along western shipping lanes. It was also well-positioned geographically to communicate with major river ports on the Iberian Peninsula, such as those of the Ebro and the Guadalquivir. Thus, it served as a nodal point between other occidental settlements on the coasts of Iberia and North Africa, and even the southern coast of France, thereby reducing a great deal of cumbersome coastal navigation.

Peninsula, much of which seems to have been regulated or negotiated by Ebusus (Belén and Fernández Miranda 1979; Orfila 2008; Puig Palerm 2005). Before the arrival of the Romans in the peninsula, Iberia had been known for a wealth of produce, but was not then known for olive oil or wine production. Following the Roman conquest of Hispania, Italic imports were therefore able to assume market dominance, with central Italic wine from Campania, Latium and Etruria becoming popular (Asensio 1996; Keay 1988). Transported in Dressel 1A and the Graeco-Italic style amphorae, examples of Italic commodities are also found in Menorca at Calascoves (Belén and Fernández Miranda 1979) and Lazareto (Nicolás Mascaró 1979), both in the south of the island near Mahon. Invariably encountered among these sites were also pieces of black-gloss Campanian fine ware, which formed a sort of secondary cargo, possibly intended for sale or for the personal use of the crews (Belén and Fernández Miranda 1979; Nicolás Mascaró 1979). The ultimate destination for these shipments has not been ascertained, and so it is unclear whether such ships were arriving at the south of the island to trade, seek shelter from bad weather, obtain water and supplies for a longdistance journey (the total load of the Lazareto shipwreck has been estimated at 300–400 amphorae, with a vessel capacity of 80–100 tons) (Nicolás Mascaró 1979: 7), or some combination thereof.

A Phoenician commercial settlement was established on Ebusus by the seventh century BC, and by the first half of the sixth century BC, the first evidence of Ebusitan self-production emerged, apparently not only for local consumption, but also for export (Fernández and Costa 2006). Transported in the first locally fabricated amphora, type PE-10, these commodities are believed to have included wine, olive oil, and salted fish products (ibid). Later, within the broader Punic commercial hegemony of the western Mediterranean, Ebusitan production evolved into a lucrative enterprise with the eventual establishment of satellite factories in Mallorca as well. While it is believed some sort of Punic trading outposts existed at Iamo and Magon in Menorca (Sanchez León 2003), no evidence of Punic production appears anywhere on the island. Two of the earliest known shipwrecks of Menorca, both containing Punic Ebusitan (PE) products, have been located off the northern coast of the island in the area of Punta de Sa Torre beside the Cap de Cavallería (Talavera Montes and Contreras Rodrigo 2015). The first of these sites, Sanitja II, contains a mixed cargo of indeterminate origin. Its shipment, which has been dated to the fourth century BC, consists primarily of containers from Carthage, likely produced in modern-day Tunisia, as well as items from Massalia and Ebusus (ibid). Less is known about the second shipwreck, located at Cala Torta. Both shipwreck sites were discovered as a result of underwater surveys conducted by students of the Sanisera Archaeology Institute, and, as the depths at which the fragments of the Cala Torta wreckage can be located typically exceed twenty metres, a more detailed survey would require the skills of more advanced divers. However, pieces from the Cala Torta wreck which have been identified have been consistently of PE manufacture, including amphorae and a base of a ceramic bowl which mimics common productions from the Catalan coast (ibid: 108). Additionally, an amphora of Iberian origin was located in conjunction with that shipment, so that the overall composition of the cargo has been likened to that of the fourth century wreck of Binissafúller, which contained similar items (Aguelo et al.. 2007; Aguelo Mas and Pons Machado 2011; Aguelo Mas et al.. 2013; Talavera Montes and Contreras Rodrigo 2015).

In spite of the apparent difficulties in reaching Menorca, the timeline of amphorae quantities and typologies found in Menorca do demonstrate clear and intense levels of connectivity between the island and its primary trading partners throughout its maritime history. That connectivity, however, seems to have been almost entirely dependencebased on the part of Menorca. This trade dependency is first evident in Menorca’s pre-Roman trade relationship with Ebusus. The hugely disproportionate amount of Ebusitan amphorae emanating especially from Mahon and Calascoves in the south of the island, and found in major Talayotic settlements, attest to the intensity of that relationship. The suggestion that access to much smaller quantities of Graeco-Massaliotic and Italic items was largely controlled by Punic Ebusitan merchants (Orfila 2008; Puig Palerm 2005) is reflected in the archaeological assemblages from that period. Motivations for the intervention of Q. Metellus in 123 BC A number of motivations likely influenced Rome’s decision to send Quintus Caecilius Metellus to the Balearic Islands in 123 BC. While the primary aim of the campaign was to establish control over the entirety of the archipelago and its surrounding sea lanes, Rome’s specific intentions and actions towards each island cannot be understood in a meaningful way by viewing the islands as a collective entity. Furthermore, there is no sufficient evidence to suggest the islands had indeed incorporated themselves into any kind of federation by that time, so there is little justification to regard them as such. The process of

From available evidence it is clear that, prior to Roman intervention, trade from and between the Balearic Islands was dominated almost entirely by the Punic Ebusitan industry. Pre-Roman finds from marine environments indicate a minimal trade relationship with the Italic 53

Margaret A. Amundson Romanisation in each Balearic island should instead be assessed individually in partial consequence of its relative value and utility to Rome.

Italy via the coasts of Massalia or the Strait of Bonifacio, and between Massalia and North Africa. As Rome’s empire began to grow rapidly, and demand for oil and wine increased, Hispania and North Africa quickly came to dominate those markets. In fact, by the height of the Empire, Rome’s estimated annual olive oil consumption may have been as much as twenty-five million litres or more, with a per capita consumption of twenty to twentyfive litres annually, perhaps comprising as much as a third of an individual’s yearly caloric intake (Hichner 2002).

Possible motives behind Rome’s action against the Balearics relating to economic concerns and the financial and political interests of the wealthy Roman elite should be considered. It is in this discussion that the historical destinations of the individual islands in the archipelago come into sharpest contrast. Referring to both Mallorca and Menorca, Mayer-Olivé (2017: 226) suggests the ‘need to obtain lands on which to settle colonists, and even to increase the possessions of the Roman aristocracy’ as such concerns. Mallorca was the largest of the Balearic Islands and therefore likely offered greater opportunity for settlement, cultivation, and resource acquisition. Centrally located within the archipelago, it was also the most likely site of Metellus’ settlement of the three thousand Hispanic colonists mentioned by Strabo (3.5.1) as it became the seat of Roman administration in the islands.

In 123 BC, Rome had reason to fear those waterways might be appropriated by pirates or insurgents, possibly even becoming impassable. The Tyrrhenian, Ligurian and Balearic seas had all come under threat in the immediate power vacuum following the Punic Wars. For example, Rome’s pacification of Sardinia between the years 126 and 122 BC resulted in some rebel holdouts entrenched in the island’s mountainous interior, and perhaps left others looking for a new island base from which to launch antiRoman attacks (Arribas Palau 1983: 7). It is possible the declining power of Massalia also contributed to the increase of piracy in its waters (Morgan 1969; Ormerod 1924). Prior to this, around 181 BC, a tribe known as the Ingauni had been engaging in piratical activities and disrupting maritime commerce from Liguria to as far as the Pillars of Hercules (Liv. 40.18.3–4, 40.25–29; Plut. Aem. 6.2–3). Therefore, a primary objective in combating maritime predation in the western seas would have been to assure that pirates successfully expelled from Sardinia and southern Gaul couldn’t simply retreat and regroup among the innumerable small and hidden coves of the Balearics.

If we consider Menorca individually, however, MayerOlivé (2017: 226) admits ‘the spoils that might be obtained were relatively small’, which seems correct. Despite ancient descriptions of the Balearic Islands’ fertility and abundance (Diod. 5.17.2; Strab. 3.5.1), the archaeological record in Menorca demonstrates only a subsistence-based cultivation of agriculture and livestock (Valenzuela et al.. 2017), and it offered very limited mineral resources (Guerrero Ayuso, Calvo Trias and Salvà Simonet 2007; Orfila Pons et al.. 2006). Numerous other locations in Hispania and the newly won coastal African provinces were much more attractive for large-scale agricultural development and resource acquisition. In contrast to Mallorca (Orfila 1993), the fact that no Roman villae have appeared anywhere in the archaeological record of Menorca, nor any new settlements dedicated to the cultivation of land, seems to support this conclusion (Nicolás Mascaró 2003).

Successfully securing Menorca would have posed obvious difficulties, as most of its ports and coves have small entry mouths which make them difficult to identify from the sea, and would have impeded the ingress of larger ships attempting to manoeuvre into them, making the process slower and more difficult. Two strategies stand out which Rome could have pursued in order to ensure the security of the western Mediterranean from piracy. An ‘active’ defence option would have involved the employment of a standing navy which would be relied upon to seek out the enemy before they approached the coast (Starr 1943: 56). As Rome would come to learn during subsequent anti-piracy initiatives throughout the Mediterranean (Troster 2009: 21–22), it makes little difference how many pirates are successfully defeated at sea if they are not also deprived of their bases and hideouts. Nor could it eliminate all pirate nests from countless islands and coves, and then expect they would not become infested yet again in just a short while if not monitored continuously.

Alternatively, cultural practices or belief systems peculiar to Menorca could have discouraged commercial development. While the Talayotic cultures on Mallorca and Menorca shared many similarities, the megalithic taula structures peculiar to Menorca suggest differences in ritual practices or religious beliefs. It has been noted that pre-Roman coins in Menorca are scarce, and that this may be attributed to a belief system which, according to Diodorus (5.17.3–4), rendered the importation of silver and gold money to the island taboo (Orfila 2008; Puig Palerm 2005). If this account is accurate, it could offer a possible explanation for the absence of export production or market emporia in Menorca.

Thus, the second strategy, a ‘passive’ defence, consisted in fortifying the shoreline itself (Starr 1943: 56). Considering Menorca’s coastal morphology, a passive defence would have been the most economical means of guarding the littoral. The advantage of the ‘active’ defence strategy is that it works best initially to address a sudden outbreak of piracy, whereas a ‘passive’ strategy is more economical

Historical references provided by Florus (1.43.2), Orosius (5.13.1) and Strabo (3.5.1) all indicate the threat of piracy as being the reason for military action taken by Metellus against the islands in 123 BC, and modern scholarship has thus far disclosed no reason to dispute this claim. Especially critical were the sealanes between Hispania and 54

Managing the Threat and effective as a long-term measure. Beginning with Metellus’ military campaign, Rome pursued a combined active and passive strategy towards combating its piracy problem in the occidental seas surrounding the Balearics during the second half of the second century. Rome’s acknowledgment of the necessity for a permanent military presence in Menorca, as opposed to just an active regional naval patrol, is evidenced by the creation of a military garrison at Sanisera on the north coast of the island (now Sanitja). Archaeological records at Ciutadella and Mahon show military personnel were garrisoned at those sites as well, but whereas those cities had more varied and commercial foundations, evidence from Sanisera suggests it was established solely on the basis of its utility as a military installation (Nicolás Mascaró 2003).

confirmed infrastructure constructed in Menorca by the Romans is a single road which bisects the island and links Magon and Iamo directly, with no visible subsidiary routes. This road would have been created for intercity communication and logistical purposes between the island’s two more accessible port cities, but the only records which confirm its construction date place it after the Flavian period (Orfila Pons et al.. 2006: 191). Zucca (1998: 160–61) mentions the possibility of another road between Iamo and Sanisera along the north, but provides no detail as to its chronology. Another road in the south connecting indigenous settlements to Magon is possible as well, but was likely dedicated during the time of Caracalla (198–217 A.D.), well after the implementation of Flavian municipal law (Orfila Pons et al.. 2006). In fact, the only indication of governance of Menorca was the assignment of a praefect responsible for directing military personnel and monitoring the security of all the islands and the shipping traffic within their orbit. Although it has been suggested that Metellus was the first governor of the Balearic Islands, based on epigraphic evidence, the first praefect with a certain date wasn’t until Lucio Titinio Glauco Lucreziano, administrator of the islands from 63– 65 A.D. (Zucca 1998: 111).

It is significant that just three Roman cities in Menorca are known from the historical records – Iamo, Magon and Sanisera – and that they were each located at natural harbours more or less equidistant around the island, clearly for the purpose of guarding the littoral and adjacent seas. The only known source which references all three cities, including Sanisera, comes from Pliny the Elder in the first century BC (Nat. 3.11). Whereas Iamo and Magon are identified as having Punic origins (Sanchez León 2003), it seems Sanisera was the only settlement founded by the Romans immediately following Metellus’ campaign (results from initial archaeological surveys indicated an indigenous settlement had existed in the area of the port of Sanitja/Sanisera, but that it was located some distance toward the interior) (Rita, Latorre and Orfila 1988). That no other Roman settlements were founded in the interior suggests their objective in Menorca was not territorialisation.

Other regional commercial interests may have offered more motivation than simple territorial expansion. While it has been established this was not likely the case with Menorca, it could very well have been a consideration with regards to Ebusus and, to some extent, Mallorca. Recall that Ebusus was the site of a Punic emporium with a wellestablished network of production and trade between the other islands in the archipelago, as well as with Punic and Greek colonies on the Iberian and Massaliotic coasts. Some PE exports even travelled as far as the Italic peninsula, as Pliny the Elder (Nat. 14.8.71) can be quoted praising the quality of Balearic wines.

Conversely, Roman ceramics and building remains are instead found within Talayotic settlements, and Roman inhumations and inscriptions have been found in Talayotic necropolises dating from the first and second centuries A.D. (Arribas Palau 1983: 48; Nicolás Mascaró 2003). Roman cultural artefacts in Talayotic settlements such as Torre d’en Galmés, Binicalaf, Torellonet Vell, Trepucó, Ses Talaies de N’Azina, Talatí de Dalt and Biniparratx Petit include lamps, jewelry, glass, tableware and Latin inscriptions (Nicolás Mascaró 2003). The incorporation of bricks, tiles, imbrices, opus signinum, etc., also demonstrate the introduction of new materials and construction techniques in some of these native settlements (ibid: 114). Taken together, and in the absence of villae or centuriae anywhere on the island, this evidence seems to suggest a level of Roman integration within indigenous communities not encountered elsewhere in the empire.

It would have been understandable then, given the Republic’s interests in expanding sales of Italic goods into new markets, that Rome may have wanted to appropriate the Ebusitan market, just as they had competed in the Greek and Carthaginian commercial territories. Indeed, immediately following Metellus’ campaign in 123 BC, the entire PE operation underwent a major reorganisation, resulting in a significant decrease in production on Mallorca and Ebusus. In fact, Mallorcan production centres indicate total abandonment for a time (Orfila Pons et al.. 2006). Na Guardis and neighbouring Colònia de Sant Jordi provide evidence that some of the PE production centres of Mallorca resumed economic activity in the first century A.D. (Fernández and Costa 2006; Orfila Pons et al.. 2006). A large number of PE amphora finds at other sites in the Balearics and on the coast of the Iberian Peninsula dating from the late first century BC to the first century A.D. attest to this revival of Ebusitan trade (Fernández and Costa 2006).

No evidence indicates any overt attempts toward subjugation of Menorca’s indigenous population in general. Until the first century A.D. when Vespasian introduced Flavian municipal law to all of Hispania (including the Balearics), nearly two hundred years after the ‘conquest’, no imperial infrastructure nor similar indicators of legal incorporation are apparent. The only

Given the timing, it appears the contraction of PE exports was a direct result of the Roman intervention in Ebusus 55

Margaret A. Amundson and Mallorca. Imports of PE products to Menorca all but disappeared around the same time as Roman exports began arriving to the island. As indicated above, the first Roman exports to Menorca at that time include Graeco-Italic amphorae and Campanian black varnished fineware from the Italic peninsula, as well as amphorae from Roman Hispania Citerior (Tarraconensis): Laietana 1 and, shortly after, Pascual 1 and Dressel 2–4. Marine sites in Menorca which contain these typologies are found at Addaya (Pasarius 1956), Lazareto (Nicolás Mascaró 1979) and Cala Vellana (dell’Amico and Pallarés 2007: 55; Corsi-Sciallano and Liou, 1985: 154–58). Whether, and to what extent, appropriation or supplanting of Ebusitan commercial interests may have been a motivator or simply a consequence of Metellus’ campaign in the Balearics is still unclear. Nevertheless, what is obvious is the inconsistency in the scope and intensity of Roman involvement in Ebusus and Mallorca in contrast with Menorca following the Balearic campaign.

The second part of the English definition of ‘conquest’ indicates the use of military force in order to assume control of a place or people. The installation of military garrisons at each of the three most practicable ports on the island, and nowhere else in the interior, is the clearest evidence that Rome’s interest in Menorca was entirely maritime oriented. It also demonstrates an ongoing awareness of potential threats to the security of Roman maritime interests such that a permanent military presence was warranted. Such establishments enabled military personnel to regularly patrol the island’s littoral, as well as make logistical and tactical use of the ports. However, whether such initiatives should be considered policing actions legitimised as defence of natural rights of property, or whether they represented the naval extension of imperialism (Anderson 1995: 185; Braund 1993; Harris 1979) is an important distinction. Control of Menorca’s maritime access points would still have translated into Rome’s total discretionary control over the island population’s extra-island movement and contact, potentially inhibiting their ability to access foreign goods and form alternative trade alliances.

‘Conquest’ and conclusions As indicated in the introduction, one objective of this paper is to address the accuracy and propriety of the term ‘conquest’ as applied to the Roman military intervention in the Balearic Islands in 123 BC Linguistically, words deriving from an etymology of the Latin root, ‘conquirere’ (meaning ‘gain’ or ‘win’), tend to be charged with historical and cultural meaning which is almost guaranteed to stimulate biases, whether conscious or unconscious. This alone renders the employment of such terminology highly problematic for the purposes of archaeological and historical investigation. It is therefore advisable that its use be treated with a much greater degree of caution and scepticism than has been broadly demonstrated within the field of Roman studies in the Balearic Islands to date.

If Rome felt it necessary to use military force to take control of Menorca, then it follows that it believed its enemies were physically established on the island. Indeed, Florus (1.43.8) describes how the islanders themselves were responsible for the piratical activities, preying on ships and hurling stones at those who came near portages. His description is consistent with the weaponry employed by the funditore, an elite warrior native to the Balearic Islands who employed slings to hurl lead ingots or stones with deadly force and accuracy. Such warriors had been enlisted as mercenaries in the Punic Wars, first by Carthage and then by Rome. The funditores of Mallorca and Menorca had ostensibly changed alliances during the Second Punic War (Liv. 22.20.9; 28.37.7–15). At the time of Metellus’ campaign, the available literary sources suggest the Balearics had last been allied to Rome.

In English, Oxford Dictionaries (2018) defines ‘conquest’ as ‘the subjugation and assumption of control of a place or people by military force.’ The term ‘subjugation’, while equally laden with cultural and historical meaning and biases, in its basic definition refers to the imposition of rules by one party or entity in order to assume control over a place or people. The literary record mentions no immediate institution of a legal framework for control in Menorca, but rather describes a simple military operation to rid the island of pirates. Epigraphical sources which do demonstrate Menorca’s eventual legal integration into the Roman Empire all seem to date to well after Metellus’ military intervention, mostly in the first and second centuries, A.D. (Nicolás Mascaró 2003; Sanchez León 2000, 2003). The process of organising a provincial framework and instituting a legal regime over the indigenous cities seems to have been significantly retarded in Menorca as compared to Mallorca and Ebusus. Although Romanisation in Menorca is a key topic in modern scholarship, the extent of its integration into the Roman world remains unclear at least until the declaration of Flavian municipal law in 73/74 A.D., at which time its cities and citizens were granted full legal status along with the rest of Hispania.

Strabo (3.5.1) and Diodorus (5.17–18), on the other hand, provide a very different portrayal of the Balearic islanders, describing them as generally peaceful people, owing to the favourable climate and fruitfulness of their islands. Diodorus indicates that in fact they were more in the habit of paying ransoms to pirates for kidnapping their women than they were to fighting for them. Strabo’s narrative seems to provide the most likely reconciliation of the disparate accounts of peaceful yet belligerent natives: ‘But certain malefactors, though few in number, having associated with the pirates in those seas, they all got a bad name, and Metellus, surnamed Balearicus, marched against them’ (3.5.1). This version makes the most sense in light of the other accounts, what we know about the political and economic uncertainty of the time, and the types of circumstances which tend to beget piracy. Piracy is a high-risk occupation, so it could be reasoned that those with naval training who were left suddenly without employment might be compelled to commit acts of piracy 56

Managing the Threat in order to provide for themselves and their families. The end of the Punic Wars would likely have left many mercenary funditores in such a situation. Additionally, piracy tends to follow economic cycles wherein it will typically reach its apotheosis during times of peace, when trade levels are also high (Anderson 1995; Rauh 2003), especially following a change in political regime or other similar political or social instability (Mejia et al.. 2012; de Souza 1999). Finally, given the nature of maritime crews of all occupations to be of mixed nationalities, and the fact that Menorcan funditores would have by that time gained experience fighting in heterogeneous yet well-organised armies, it makes sense that at least some of them might have joined up with pirate crews of disparate backgrounds.

of Metellus’ campaign, the Balearic Islands were clearly still important for providing shipping ports-of-call. Each island featured numerous natural coves and ports which could potentially provide safe harbour in rough seas, or anchorages for making repairs or replenishing food, water and supplies. What is important to remember, however, is that differing interests in the islands directly influenced their individual administration and consequently, their processes of Romanisation. With regards to Ebusus, Roman interests and concerns may have been influenced by an awareness of the former Punic colony’s previous loyalty to Carthage, and by the competitiveness of its products and the strength of its trade alliances in the western markets. Major withdrawal of Ebusitan products from the market immediately after 123 BC could reflect Rome’s interest in regional commercial markets, or it could simply have occurred as a consequence of a major reorganisation undertaken for administrative or political reasons. In Mallorca, the indigenous population had already established an alliance with Rome before the end of the Punic Wars (Liv. 28.37.8–12). As both the largest and most central island, it probably offered the better opportunity for colonisation, exploitation and administrative oversight. In any case, sources both ancient and modern agree that the problem of piracy in and amongst the archipelago had, at that time, risen to such a degree that military action was deemed necessary in order to eliminate that threat. That, above all else, can be considered the primary motivation behind Rome’s armed intervention in the Balearic Islands.

To suggest the general population of Menorca was conquered by military force, however, is a generalisation which is not supported by the physical evidence. On the contrary, archaeological sources indicate the indigenous people of Menorca had a more or less cooperative relationship with the Romans who came to their island. Evidence suggesting subsequent Roman and indigenous domestic integration, and the apparent continuity of Talayotic culture with the selective incorporation of elements of Roman religion and customs (Moreno Pérez 2012; Zucca 1998), renders a unified island resistance unlikely. More probably, military force was used to clear the island, mainly the littoral, of pirate gangs of mixed origins which included foreign sailors as well as funditores native to Menorca. From the end of the third through the second centuries BC some of the indigenous Menorcan coastal fortifications were abandoned (Belén and Fernández Miranda 1979; Prados et al.. 2017). It remains unclear to what extent this abandonment was due to demographic factors related to the Punic Wars, or was the result of later Roman reorganisation. Some Talayot villages, such as Talatí de Dalt, did go into notable decline after 123 BC. However, given its proximity of just four kilometres to the Roman settlement at the port of Magon, it is possible the village inhabitants migrated there in order to take advantage of new opportunities made available by the arrival of the Romans and the security they effected. Although habitation may have declined or basic functions altered in some areas of Menorca (in many cases severely), surveys of indigenous sites including villages and necropolises have shown that a majority continued to experience some degree of continuity throughout the entirety of the Roman epoch, including some reconstructions, remodelling and additions to old structures in the villages (Mas et al.. 2007; Nicolás Mascaró 2003). Such additions and modifications cannot be attributed necessarily to Romanisation, or at the very least they do not represent significant changes (Nicolás Mascaró 2003).

Concerning Menorca in particular, however, it appears to have been Rome’s only motive. For as many as two centuries afterwards, there is little indication of territorialisation of the island interior or the subjugation of its population. The integration of Roman religion and social organisation within Talayotic society appears to have been a more or less gradual process. As a type of global institution, the army would have served as a proxy for Roman practices and customs, especially with regards to religion, urbanisation and social hierarchies. Available evidence from excavations of Talayotic villages supports the suggestion that this integration developed over time in consequence of globalising processes, rather than as a direct result of conquest. Therefore, to use the term, ‘conquest’, to define Rome’s intervention in Menorca is at least questionable. As the eventual integration of the native Menorcans into the Roman world resulted from consequent but separate processes, to speak of the event which catalysed those processes and the processes themselves as being concurrent is a misleading oversimplification. That is, as it is an ‘historical error to equate conquest with cultural assimilation’ (Belén and Fernández Miranda 1979: 201), so it should also be considered erroneous to equate cultural assimilation with conquest.

Nevertheless, Rome sent forces in 123 BC to secure the entirety of the Balearic archipelago, not only Menorca, a fact which should not be neglected entirely. If the acquisition of territory or resources was not the purpose

It is also important to consider the broader geopolitical situation at the time of Rome’s military intervention in the Balearics. The coincidence of other campaigns 57

Margaret A. Amundson in the region from 124–120 BC, waged with the intent to secure coastal terrestrial and marine routes between Italy and the new western provinces, is impossible to ignore. For example, the establishment of the province Gallia Narbonensis in 120 BC was significant in that the destruction of Entremont and the foundation of Aquae Sextiae (now Aix-en-Provence) by the consul C. Sextio Calvino, both in 123 BC, secured a terrestrial connection to Hispania (Orfila 2008). At the same time, a war in Transalpine Gaul must have placed greater importance on the sea routes to Hispania (Liv. Epit. 60; Morgan 1969; Puig Palerm 2005). Regardless of whether trouble would develop in Hispania, the ability to transport troops and supplies to Roman garrisons stationed there was still necessary to regional security and basic governance. Considering as well, the incidents of unrest in Sardinia between 126 and 122 BC, Rome’s action in the Balearic Islands may best be understood as an integral component in a region-wide initiative toward the suppression of piracy and the maintenance of critical maritime networks and security in the western Mediterranean. Menorca’s strategic importance in the ancient world, therefore, lay not so much in its facilitation of connectivity as in its capacity to disrupt maritime transmissions and abet actors in regional destabilisation. The very occurrence of Rome’s intervention and its continued occupation of the island’s ports thereafter for defence purposes demonstrate Menorca’s significance as a key ancillary in the assurance of both maritime security and Roman imperial hegemony.

PLINY THE ELDER. The Natural History. Trans. Bostock, J. and Riley, H. T. 1855. Taylor and Francis: London. PLUTARCH. Lives, Volume 6: Dion and Brutus. Timoleon and Aemilius Paulus. Trans. Perrin, B. 1918. Loeb Classical Library 98. Harvard University Press: Cambridge, MA. STRABO. Geography, Volume I: Books 1–7. Trans. Hamilton, H. C. and Falconer, W. 1903. George Bell & Sons: London. Secondary Sources AGUELO, X., PALOMO, A., PONS, O., and DE JUAN FUERTES, C. 2007. ‘El pecio de Binissafúller’. In J. Pérez Ballester and G. Pascual (eds.), Actas V Jornadas de Arqueología Subacuática: Comercio, redistribución y fondeadores. La navegación a vela en el Mediterráneo. Gandia, Spain, 8–10 November 2006. Servei de Publicacions: Valencia. Pp. 199–207. AGUELO MAS, X. and PONS MACHADO, O. 2011. ‘El pecio de Binissafúller. Resultados de la campaña 2011’. Actas de las Jornadas de ARQUA. Pp. 96–101. AGUELO MAS, X., PONS MACHADO, O., DE JUAN FUERTES, C., RAMON TORRES, J., MATA PARREÑO, C., SORIA COMBADIERA, L., PIQUÉ, R. and ANTOLÍN, F. 2013. ‘El pecio de Binissafúller. Estado de las investigaciones’. I Congreso de Arqueología Náutica y Subacuática Española: Navegación y comercio en la Antigüedad. Cartagena, Spain, 14–16 March 2013. Ministerio de Educación, Cultura y Deporte: Madrid. Pp. 67–85.

Given the importance of islands both in past and present security, mobility and economic interests, maritime archaeological study offers the potential for informing current policies based on a better understanding of past experiences. The physical accessibility of Menorca’s maritime archaeological heritage and the wealth of material discovered there as a result of ongoing survey efforts (Pons Machado and Aguelo 2012) make it an ideal location to further this understanding.

ANDERSON, J. L. 1995. ‘Piracy and world history: An economic perspective on maritime predation’. Journal of World History. Pp. 175–99. ARRIBAS PALAU, A. 1983. La Romanització de les Illes Balears. Universitat de Mallorca: Palma de Mallorca. ASENSIO, D. 1996. ‘Les amfores d’importació de la ciutadella iberica d’Alorda Park o les Toixoneres (Calafell, Baix Penedes, Tarragona)’. Revista d’arqueologia de Ponent, 6: 35–79.

Works Cited Primary Sources DIODORUS SICULUS. Library of History, Volume III: Books 4.59–8. Trans. Oldfather, C. H. 1939. Loeb Classical Library 340. Harvard University Press: Cambridge, MA.

BELÉN, M. AND FERNANDEZ-MIRANDA, M. 1979. El fondeadero de Cales Coves (Alayor, Menorca). Ministerio de Cultura: Madrid. BRAUND. D. C. 1993. ‘Piracy under the Principate and the ideology of imperial eradication’. In J. W. Rich and G. Shipley (eds.), War and Society in the Roman World. Routledge: London. Pp. 195–212.

FLORUS. Epitome of Roman History. Trans. Forster, E. S. 1929. Loeb Classical Library 231. Harvard University Press: Cambridge, MA. LIVY. History of Rome. Trans. Rev. Canon Roberts. 1912. E. P. Dutton and Co.: New York.

CONTRERAS, F., MÜLLER, R., and VALLE, F. J. 2006. ‘El asentamiento militar romano de Sanitja (123–45 a.C.): una aproximación a su contexto histórico’. Mayurqa, 31: 231–49.

LIVY. History of Rome by Titus Livius: the epitomes of the lost books. Trans. McDevitte, W. A. 1850. Henry G. Bohn: London.

CORSI-SCIALLANO, M. and LIOU, B. 1985. ‘Les épaves de Tarraconaise à chargement d’amphores Dressel 2–4’. Archaeonautica, 5.1: 5–178.

OROSIUS. Historiarum Adversum Paganos Libri VII, ex recogn. Zangemeister, C. 1889. Teubner: Leipzig. 58

Managing the Threat DE SOUZA, P. 1999. Piracy in the Graeco-Roman World. Cambridge: Cambridge University Press.

MORGAN, M. G. 1969. ‘The Roman conquest of the Balearic Isles’. California Studies in Classical Antiquity, 2: 217–31.

DELL’AMICO, P. and PALLARÉS, F. 2007. ‘Le anfore della Laietania. Appunti e riflessioni’. Archeologia Marittima Mediterranea. An International Journal on Underwater Archaeology, 4: 53–133.

NICOLAS, J. C. 1985. ‘Sanitja: a Roman Settlement North of Es Mercadal’. Festes Majors d’Es Mercadal. Pp. 21–30.

FERNÁNDEZ, J. H. and COSTA, B. 2006. ‘Ibiza FenicioPúnica’. Historia de las Islas Baleares, 3. El Mundo-El Día de Baleares: Palma de Mallorca.

NICOLÁS MASCARÓ, J. C. de 1979. ‘La nave romana de edad republicana del puerto de Mahon’. Arqueología clásica en Baleares, 1. Centro de Investigaciones Submarinas de Menorca: Mahon.

GUERRERO AYUSO, V. M. 1993. ‘Navegación y Comercio en las Baleares Romanas’. In E. Lipinski, Ç. G. Wagner, A. Domínguez, P. Fernandez, A. Ma Vázquez, and V. M. Guerrero (eds.), Estudis d’Història Econònomica: Economia i Societat a la Prehistòria i Món Antic. Prensa Universitaria: Palma de Mallorca. Pp. 113–38.

NICOLÁS MASCARÓ, J. C. de 2003. ‘Els primers segles de l’ocupació romana de l’illa de Menorca: continuïtats i ruptures en l’àmbit rural’. Mayurqa, 29: 111–28. ORFILA, M. 1993. ‘Construcciones rurales romanas en Mallorca’. Homenatge a Miquel Tarradell. Estudis Universitaris Catalans: Barcelona. Pp. 793–805.

GUERRERO AYUSO, V. M. 2010. ‘¿Foceos en el comercio tardoarcaico al norte de Baleares?’. Mayurqa, 33: 131–60.

ORFILA, M. 2008. ‘La intervención de Q. Cecilio Metelo sobre las Baleares (123 a 121 a. C.). Condiciones previas y sus consecuencias’. Pyrenae, 39.2: 7–45.

GUERRERO AYUSO, V. M., CALVO TRIAS, M., and SALVÀ SIMONET, B. 2007. ‘Insularity and the indigenous world on the periphery of the system: the Balearic Islands between the 6th and the 1st c. B.C.’. In P. van Dommelen and N. Terrenato (eds.), Articulating local cultures. Power and identity under the expanding Roman Republic. JRA Supplementary Series 63: 71–84.

ORFILA PONS, M., CHÁVEZ ÁLVAREZ, E., MERINO, J. and ARANEGUI GASCÓ, C. 2006. ‘Las Baleares en Época Romana y Tardoantigua’. Historia de las Islas Baleares 4. El Mundo-El Día de Baleares: Palma de Mallorca. ORMEROD, H. A. 1924. Piracy in the Ancient World: An Essay in Mediterranean History. Reprint 1997. Johns Hopkins University Press: Baltimore.

HARRIS, W. V. 1979. War and imperialism in Republican Rome, 327–70 BC. Oxford University Press: Oxford.

OXFORD UNIVERSITY 2018. Oxford Dictionaries. [online] Oxford: Oxford University Press. Available from . [Accessed 03 January 2018].

HITCHNER, R. B. 2002. ‘Olive Production and the Roman Economy: The Case for Intensive Growth in the Roman Empire’. In W. Scheidel and S. von Reden (eds.), The Ancient Economy. Routledge: New York. Pp. 71–86.

PASARIUS, M. 1956. ‘El yacimiento de ánforas en las Islas de Addaya’. Noticiario arqueológico hispánico, 5: 209–11.

KEAY, S. 1988. Roman Spain. University of California Press: Berkeley.

PONS MACHADO, O. and AGUELO, A. 2012. ‘15 Años de Cartas Arqueológicas Subacuáticas de Menorca, 1996–2011’. In A. Castillo (ed.), Proceedings of the First International Conference on Best Practices in World Heritage: Archaeology. Mahon, Spain, 9–13 April 2012. Universidad Complutense de Madrid: Madrid. Pp. 975–82.

MAS, C., CAU, M. Á., GURT, J. M. and SALES, M. 2007. ‘El poblamiento rural de Menorca durante la Antigüedad tardía: primera aproximación’. Revista de Menorca, 90(1): 177–215. MAYER-OLIVÉ, M. 2017. ‘Shipping and the Movement of Materials and Products in the Roman Mediterranean, with particular reference to their reflection in the Balearic Islands’. In J. Velaza (ed.), Insularity, Identity and Epigraphy in the Roman World. Cambridge Scholars Publishing: Newcastle upon Tyne.

PRADOS, F., JIMÉNEZ, H., LEÓN, M. J., ADROHER, A. M., NICOLÁS, J. C. De and MARTÍNEZ, J. J. 2017. ‘Menorca Entre Cartago y Roma: Avance de la Excavación Arqueológica del Proyecto Modular en el Poblado de Son Catlar (Ciutadella)’. VII Jornades d’Arqueologia de les Illes Balears. Mahon, Spain, 30 September-2 October 2016. Consell Insular de Menorca: Maó. Pp. 153–60.

MEJIA JR., M. Q., CARIOU, P. and WOLFF, F. C. 2012. ‘Piracy in shipping’. In W. K. Talley (ed.), The Blackwell Companion to Maritime Economics. John Wiley and Sons Ltd.: Chicester, UK. Pp. 346–70. MORENO PÉREZ, A. S. 2012. ‘La Religión Romana en Menorca: Aproximación desde las Fuentes Iconográficas’. Lucentum, 31: 177–94.

PUIG PALERM, A. 2005. ‘L’entrada de Mallorca en l’esfera imperialista de la República romana. Una contextualització històrica de l’evidència arqueològica anterior a la conquista romana del 123 aC’. XXIII

59

Margaret A. Amundson Jornades d’Estudis Històrics Locals. Institut d’Estudis Baleàrics: Palma de Mallorca. Pp. 247–59. RAUH, N. K. 2003. Merchants, Sailors and Pirates in the Roman World. Tempus: Stroud, UK. RITA, C., LATORRE, J. L. and ORFILA, J. A. 1988. ‘Las excavaciones arqueologicas en el yacimiento de Sanitja (Menorca) hasta 1984’. Les Illes Balears en temps cristians fins als arabs. Institut Menorqui d’Estudis: Mahon. Pp. 41–49. SANCHEZ LEÓN, M. L. 2000. ‘Movilidad geográfica y élites municipales en la Menorca romana (Siglo II dC.)’. Bolletí de la Societat Arqueològica Lul·liana: Revista d’estudis històrics, 56: 35–44. SANCHEZ LEÓN, M. L. 2003. ‘La Ciudad de Mago (Maó, Menorca): Continuidades y Rupturas’. Mayurqa, 29: 97–110. STARR, C. G. JR. 1943. ‘Coastal Defense in the Roman World’. The American Journal of Philology, 64.1: 56– 70. TALAVERA MONTES, A. J. and CONTRERAS RODRIGO, F. 2015. ‘Dos Naves Púnicas en el Norte de Menorca en el siglo IV a.C. (Puerto de Sanitja, Es Mercadal)’. Rev. digital, @rqueología y Territorio 12: 105–19. [online]. Accessed 3 October 2016. Available from . TRÖSTER, M. 2009. ‘Roman Hegemony and Non-State Violence: A Fresh Look at Pompey’s Campaign against the Pirates’. Greece & Rome, 56.1: 14–33. VALENZUELA, A., ALCOVER, J. A. and CAU, M. Á. 2017. ‘The impact of Roman conquest on the pattern of livestock exploitation on the Balearic Islands’. Archaeofauna, 26: 127–42. ZUCCA, R. 1998. Insulae Baliares. Le Isole Baleari sotto il Dominio Romano. Carocci: Rome.

60

6 The Roman Port of Berytus Naseem Raad Centre for Maritime Archaeology, University of Southampton [email protected] Abstract: Though Berytus served as a key port throughout the Roman Period, the Classical harbour is still not well understood. This paper proposes a new interpretation of published material regarding the Roman city of Berytus through an examination of archaeological and geomorphological data uncovered over the past several decades. It will be shown that despite the fragmentary state of evidence and the lack of focus on Roman remains, a holistic approach can allow us to situate the harbour basin in the Roman Period, more definitively characterise and date maritime installations uncovered in the Beirut Central District excavations, and ultimately shed light on the evolution of an important maritime hub in the Roman Empire. Key Words: Beirut archaeology; Roman economics; Roman trade; Roman ports; Eastern Mediterranean Introduction

now able to better understand the layout of the port city and how it facilitated the loading and unloading of ships, the transportation of shipped goods to the market and the subsequent exchanges that took place within the ancient souks.

Located in central Lebanon, the city of Beirut is characterised by a long history of continuous occupation thanks to the site’s access to water, strategic geographical location and effective harbour. Even in ancient times, the city was seen as infinitely old, as suggested by Nonnus sometime between the fourth and fifth centuries AD, stating Beirut to have been ‘the nursemaid of cities… first to appear, born with time, old as the universe’ (Dionysiaca 41.361–7). Recent excavations have explored the nature of the city and its development over time (Perring et al. 2003), but there remains the need for in-depth analysis of the harbour and its relationship with the city, as well as the maritime networks within which Beirut was involved.

In this paper, the author examines harbour installations found in Beirut to situate and date the harbour basin in the Roman period. This is done by first providing historical context and outlining the time period of focus, then presenting and discussing archaeological and geomorphological data uncovered in the city (see Figure 1), and briefly comparing results to other maritime sites in the eastern Mediterranean. History

In the Roman period, the port city of Berytus served as one of the main centres of exchange in the eastern Mediterranean. After the settlement of the veterans of the Battle of Actium, the city was elevated to the status of colonia, resulting in an expansion of the existing city limits, a lavishing of public spaces and a refurbishing of the harbour (Elayi 2010: 160–61; Hall 2004: 95; Marriner 2009: 210; Millar 1993: 36; Perring et al. 2003: 204, 220; Seeden and Thorpe 1997: 236; Stuart 2002: 98–104, Fig. 5). In addition, the city began producing several new amphora types starting in the first century BC and packaging Berytus wine and oil to be subsequently distributed (Reynolds 1999; 2000b; Reynolds et al. 2010). These developments suggest the port city to have benefited economically from Roman colonisation, which translated into urban growth and expansion. With evidence of harbour installations utilised in the Roman period having been uncovered along the northern coast of Beirut (Butcher and Thorpe 1997: 299, Fig. 8; Perring 1997: 25–26), we are

The Phoenician coast was under Persian rule before being conquered by Alexander the Great, and passed to Seleucid control after his death (Butcher 2003: 22). After years of internal strife and local disputes, Antiochus III ‘the Great’, the ruler of the Seleucid kingdom from 222–188 BC, began the first of many clashes with Rome (Butcher 2003: 27). These feuds, among other reasons, led to the decline of the Seleucid kingdom. As a result, later Seleucid rulers were forced to grant independence to some of the more powerful cities, such as Tyre in 126/125 BC, Tripoli in 112/111 BC and Sidon in 111 (Butcher 2003: 29). Beirut received its independence from Seleucid control several decades later by Tigranes, the king of Armenia, in 81 BC (Butcher 2003: 23–26; Hall 2004: 45; Lauffray 1977). Thus, the city existed for several decades in the ‘power vacuum’ that had been left after the withdrawal of Tigranes and before the arrival of Pompey in 66 BC. Shortly 61

Naseem Raad

Figure 1. Map of Beirut depicting excavated sites from the BCD excavations (data provided by Hans Curvers).

thereafter, Beirut was incorporated as a Roman territory in 64–63 BC with Pompey’s deposing of Antiochus IV and reorganisation of the political structure of the region (Hall 2004: 45; Sartre 2005: 43). In 42 BC, Mark Antony took control of the eastern provinces after the battle at Philippi, and eventually gifted a selection of land and cities (the Bekaa Valley and coastal towns) to Cleopatra (Josephus Ant. 15.95; Plutarch 51.2.1–2, 3.1; Hall 2002: 142). This appears to have included most of the Levantine coast ‘with the exception of Tyre and Sidon, which he knew to have been free from the time of their ancestors, although she earnestly pleaded that they be given to her’ (Josephus Ant. 15.95). After the defeat of Mark Antony at the hands of Augustus at the Battle of Actium, it is believed that Augustus settled two legions in Beirut around 31 BC based on numismatic evidence and Strabo’s reference to the settlement of soldiers after Actium (Strabo 16.2.19; Hall 2004: 46; Millar 1993).

a second period of intense private and public construction in the city, and marks an important urban transition. Therefore, the main period of focus in this paper resides between the ‘independent’ period and the later Roman Period to compare the physical evidence to transitional phases in the city’s history. Harbour installations Archaeological data can help in shedding light on the abovementioned developments. Excavations in Beirut have revealed several phases of a quay (BEY 039), tanks/ vats (BEY 007) and a series of ashlar blocks that may be indicative of a quay in the north-western quarter of the Roman city (BEY 007), the analysis of which may provide crucial insight regarding the location of the ancient coastline and the continuous use of the harbour before and after Roman colonisation. This section examines these features to better situate the Roman harbour, and analyse construction methods and associated material.

After colonisation, Berytus quickly rose in status. Local rulers favoured the city, and at different times were said to have bestowed monumental structures in the form of statues, sculptures and a theatre. Marcus Julius Agrippa II, the client king ruling in the coastal provinces, was said to have given the people grain and olive oil and presented the population of Berytus with annual spectacles (Josephus Ant. 20.211–13). Over time, the city grew and expanded its territory, acquiring large tracts of land in the Bekaa Valley and further south, reaching some point slightly south of Jiyeh. In the reign of Septimius Severus (193–211 AD), Berytus was included in the newly organised province of Syria Phoenice, which was a further subdivision of the Syrian province (Hall 2002: 149–51). This came during

BEY 007 BEY 007, located in the northwest region of the Beirut Central District (BCD) excavations, was excavated in 1996 under the direction of Helga Seeden and Reuben Thorpe (Thorpe et al. 1998: 31). Exploring the area offered a rare opportunity, as the site is situated on a high promontory that runs N-S, and appears to make up the western limit of the ancient harbour basin (Thorpe et al. 1998: 32). Excavators were hopeful to better understand the waterfront of the Hellenistic/Roman city and associated maritime activity. 62

The Roman Port of Berytus 1998: 36). Much of this pottery has been roughly dated to the Classical period by the excavators, and similar vats dated to the late Roman or Byzantine period were found at BEY 006 (Thorpe et al. 1998: 38). Thus, it seems that the tanks can be tentatively dated to the late Roman/early Byzantine period.

As mentioned earlier, excavations at BEY 007 uncovered maritime installations that can be tentatively dated to the Classical period; however, preservation of archaeological material from this phase is quite poor (Butcher and Thorpe 1997: 299). Most of these deposits were fragmentary and truncated by later periods of activity in the Ottoman period as well as modern construction (Thorpe et al. 1998: 36, 43, 46). Regardless, the data still sheds light on the Roman port when taken in context with other BCD sites.

One possible function for these tanks could be as a fish tank, as described by Columella in his treatise on agriculture: …the best pond is one which is so situated that the incoming tide of the sea expels the water of the previous tide… for a pond most resembles the open sea if it is stirred by the winds and its water is constantly renewed… The pond is either hewn in the rock, which only rarely occurs, or built of plaster on the shore… If the nature of the ground permits, channels should be provided for the water on every side of the fish-pond… It will be well to remember that gratings made of brass with small holes should be fixed in front of the channels through which the fish-pond pours out its waters, to prevent the fish from escaping (Columella 8.16.7).

Tanks/vats The eastern portion of the main area of excavation of BEY 007 turned up two large rock-cut tanks or vats of an unknown function, one of which lies in close proximity to the other features (Figure 2). The only surviving portions of the tanks are those that were cut directly into the bedrock, with later phases having been truncated by Ottoman occupation and modern construction (Thorpe et al. 1998: 36). These remaining parts of the vats are around 3 by 1 metres in size and reach around 1.5 metres in depth (Thorpe et al. 1998: 36). The insides of the vats were plastered with a coarse, pink mortar with pottery and tile inclusions and subsequently covered by a fine-grained pink mortar to give a more refined finish (Thorpe et al.

It is difficult to say definitively if these were indeed fish tanks, as the upper portions have been heavily truncated

Figure 2. Harbour installations uncovered in the eastern area of BEY 007; Phase 1 has been roughly dated to the Roman period, but may have been constructed in the Hellenistic Period; sea wall phases 2–4 represent the Ottoman quay (data provided by Hans Curvers).

63

Naseem Raad in the Ottoman period and from modern construction. However, they are reminiscent of basins found in Sarepta in the south of Lebanon, which are dated to the late Roman period (Pritchard 1971). They are identical in size, are carved into the bed rock, and lined with plaster. Similar examples have also been found at Chersonisos and Mochlos on the northern coast of Crete (Pritchard 1971). In Sarepta, the fish tanks were fed by carved channels, and located adjacent to the quay. This differs from those observed in BEY 007, as the tank closest to Sea Wall Phase 1 (SWP1) is actually about 15 metres southwest of the feature. If the two features (quay and tank) are to be interpreted as contemporaneous, there may have been a problem in water flow reaching the tank consistently. Additionally, the tanks may have been used in the growing of murex, from which a purple dye could be produced that the coastal Levant was famous for throughout history. Another possibility is the utilisation of the tank as a basin for fresh water, as observed in Sarepta (Pritchard 1971: 47). This will be discussed in more detail later.

the tank observed in BEY 007 from SWP1 supports this proposition, as this would have prevented sea water from flowing into the basin. Sea wall In the eastern extension of the main area of excavation, a series of ‘sea walls’ were observed. SWP1 has been tentatively dated to the Classical period based primarily on construction technique (Seeden and Thorpe 1997: 228; Thorpe et al. 1998: 38), and phases 2–4 were built in the Ottoman period. Phase 1 of the sea walls was uncovered in the south-eastern corner of excavation and consists of large, ashlar limestone blocks set on roughly a northwestsoutheast axis. The width of the wall varies from 0.3 to 0.7 metres (Thorpe et al. 1998: 38), though this variability could be due to the heavy truncation and poor preservation of the upper courses of the wall. Most of the ashlar blocks seem to be closer to 0.7 metres in width, with several outliers in the extreme south-eastern area of excavation. The upper courses of the wall lay between 1.83 and 2.50 metres above sea-level (Thorpe et al. 1998: 38).

Rock-cut steps In between these tanks, a sequence of ten rock-cut steps were uncovered on an east-west axis. They cut through the natural break of slope of the bedrock, with the lowest step on or just below the waterline (Thorpe et al. 1998: 36). More specifically, the lowest step cut through the level bedrock, and the sides of this cut were plastered with a pink mortar similar to that used in the vats described earlier (Thorpe et al. 1998: 36). The steps descend from west to east, eventually opening up and leading into a natural cove in the north-western corner of the harbour basin in the late Roman period (Thorpe et al. 1998: 36–38).

This construction technique of ashlar blocks being laid adjacent to one another and perpendicular to the coastline is quite typical of harbour construction in the eastern Mediterranean throughout the Hellenistic and Roman periods. Examples include Amathous in Cyprus, Elaia in modern-day Turkey, Sarepta and Tyre in south Lebanon. However, without ceramic material or numismatic evidence, it is difficult to narrow in on a specific phase based solely on this site.

At a later phase, a wall of squared and roughly faced sandstone blocks reinforced with a thick, weak, orange, sandy mortar overlay the plaster at the base of the rock-cut steps (Thorpe et al. 1998: 37). This wall has not been dated, though it was most definitely erected after the formation of the steps. At the north-eastern end of the cove, two courses of squared limestone blocks were observed. These were joined with a thick, pink mortar with sandstone packing and faced on the southern edge, and the northern side of these walls was bonded to the projecting bedrock spur (Thorpe et al. 1998: 37). Since only the southern side of the wall is faced, and the northern side is bonded to the bedrock, this indicates that the cove south of the wall would have been intended to be a closed-off space. As the rock-cut steps lead to this designated space which lies adjacent to the ancient coastline, it is quite likely that these features can be dated to the same period. Furthermore, it is possible that the rock-cut tanks discussed in the previous section are also of a similar time period, as attested by the presence of a similar pink mortar.

In BEY 039, a sounding of 68 metres2 was excavated on the western side of Allenby Street (Elayi and Sayegh 2000: 225). The sounding revealed what appears to be a quay with a possible mooring post. The quay is characterised by several rows of rectangular, limestone, ashlar blocks associated with three phases of construction. The blocks are situated longitudinally and oriented in an east-west direction (Elayi and Sayegh 2000: 229). Though the stratigraphic sequences are complex and the area was quite disturbed, three strata can be roughly distinguished.

BEY 039

Stratum I Stratum I is situated about 0.95 metres above actual sea-level and was uncovered about 2.45 metres beneath Allenby Street. It is composed of ashlar blocks consisting of ‘ramleh’, a type of limestone found locally. The stones are 0.60 metres long by 0.30 metres wide, and are bonded together by a greyish mortar, composed in part by lime and ash (Elayi and Sayegh 2000: 230). About 1.40 metres from the edge of the quay, a cylindrical mooring post of the same ramleh material was uncovered, with two deep grooves on either side, likely from the usage of moored ships. It was also found slightly inclined towards the port, which may be the result of repeated use (Elayi and Sayegh 2000: 230).

A parallel can again be drawn with Sarepta, as similar rockcut steps were observed leading from the edge of the quay to a nearby basin that was possibly used to purify water or hold fresh water (Pritchard 1971: 47–48). The distance of 64

The Roman Port of Berytus This mooring post was the only one found; however, gaps in the array of limestone blocks were located at regular intervals in relation to the mooring post. Finally, several Roman bronze coins were found in situ in context with the blocks in stratum I along with a needle for repairing fishing nets (Elayi and Sayegh 2000: 230). Unfortunately, while ceramic sherds were noted in the excavators’ overall stratigraphic sequences (Elayi and Sayegh 2000: 226–31), they have not been specifically linked to the three strata of the quay, nor have they been published or analysed in detail.

However, the distinction between each stratum (mortar, dovetail joints, diagnostic material) supports a clear chronological sequence. Stratum III Stratum III was found at 0.15 metres above actual sealevel and 3.25 metres beneath Allenby Street. This layer is quite similar to stratum II in terms of the dimensions of the ashlar blocks in the same layout and orientation (Elayi and Sayegh 2000: 231). These blocks are joined in a similar fashion to that of stratum II with the overlapping stones, with certain blocks joined with dovetail joints. Several large hollow cavities were observed with traces of reddishbrown material at the interior, indicating the utilisation of iron joints, presumably coated in lead to prevent damage to the stones (Elayi and Sayegh 2000: 231). At the borders of the quay, it appears that the blocks are faced towards the sea, though they have very much eroded (Elayi and Sayegh 2000: 231). This layer was quite moist, likely due to the fact that it is close to actual sea-level. This stratum likely represents Iron Age/Early Hellenistic installations, especially since the adjacent site of BEY 010 revealed extensive remains of an Iron Age/Hellenistic settlement.

The use of mortar to bind the stones together is typically associated with Roman engineering, though it seems to have come into use earlier (Blackman 1982b: 197). As this stratum overlays two other rows of ashlar blocks which were not bonded with mortar, and did not provide similar dating material, it is likely that this layer represents the refurbishment of older harbour installations in the Hellenistic or Roman periods with new construction techniques. As observed at Atlit, Tyre, Sidon and Akko, it seems that older harbour works were often refurbished and continued to be used in the Roman period (Galili and Rosen 2008; Haggi 2010; Marriner et al. 2014). This proposition is corroborated by the large number of sigilatta and other ceramic sherds from the Roman period that were observed throughout BEY 010 (Elayi and Sayegh 2000: 196–99), the settlement immediately adjacent to the quay of BEY 039. This phase is also near identical in the placement of the ashlar blocks, the size of the stones and the use of mortar to SWP1 in BEY 007.

Sedimentary cores In order to clarify the complicated situation detailed in previous sections, geomorphological analyses provide crucial insights. The ancient harbour basin of Beirut currently lies landlocked beneath the modern city along the northern coastline. This coastal progradation is largely due to silting up from a lack of upkeep in the harbour, as well as construction works since the nineteenth century (Marriner et al. 2008: 2504). Urban developments have made excavation difficult and limited to specific times and locations in the city. Thus, geomorphological analysis traces the development of the ancient harbour over time using a non-destructive method.

Stratum II Stratum II, located about 0.65 metres above actual sea-level and about 2.75 metres beneath the lower part of Alleby Street, is characterised by a similar row of ashlar blocks in the same orientation (Elayi and Sayegh 2000: 230). These blocks are considerably larger, measuring on average 0.60 wide, 0.50 high, and more than 1 metre in length (Elayi and Sayegh 2000: 230). They are assembled ‘joint vifs’, joined together without mortar through overlapping stones to reinforce the structure. More specifically, not all the ashlar blocks are perfectly rectangular and situated regularly; rather, certain stones are carved to fit into each other. Several blocks were joined together with lead-enforced dovetail joints, possibly those that would have been most exposed (Elayi and Sayegh 2000: 230). Another bronze needle for the repair of fishing nets was also discovered between these blocks (Elayi and Sayegh 1998; 2000: 230).

The western façade of the city is exposed to the dominant winds and waves as well as dangerous eddies, such as the one near Ramlet el Bayda (Davie 1987: 147). The western shores did not benefit from the natural protection of the rocky promontory of Ras Beirut, and any port would have been quickly filled in, especially as the sea is quite shallow in this area. Thus, the primary focus for archaeologists and geomorphologists studying the ancient harbour of Beirut has always been the northern coastline, specifically the cove of Ain el Mreisseh, the cove of Hotel Saint Georges and the Bay of Saint André (Figure 3). As the archaeological material and ancient city largely centred around the Bay of Saint André, it was proposed to begin geomorphological analysis in this area (Marriner et al. 2008). 25 cores were drilled in and around the hypothesised location of the ancient harbour basin in Beirut in collaboration with the BCD excavations (Carayon et al. 2011: 51). 20 of these cores, depicted in Figure 4, were undertaken by Marriner et al. (2008). This campaign was based largely on Davie’s proposed location for Beirut’s ancient harbour (1987). The

The use of the dovetail joint is usually associated with Hellenistic construction (Elayi and Sayegh 2000: 231; Martin 1965: 254–55; Raban 1991; Seeliger et al. 2013), which appears to be consistent with the possible phases of the maritime installations. This does not suggest the technique to have changed or gone out of use in later periods, as the technique remained in use for a long period of time (Blackman 1982b: 197). Thus, more precise dating is not possible solely through architectural analysis. 65

Naseem Raad

Figure 3. The bays of Beirut.

Figure 4. Coastal change in Beirut based on core analysis and archaeological work (after Carayon et al. 2011: 52, Fig. 9).

cores were drilled to the east and west of the ancient tell to test this hypothesis, and the results have been analysed to better understand geomorphological processes at play

along Beirut’s shores. This section summarises the analysis of these cores and discusses some of the implications regarding harbour upkeep in the Roman period as well as 66

The Roman Port of Berytus the transition from a high-energy to a low-energy marine environment.

and specify any upkeep that may have taken place in the past.

Eastern basin

Based on the quay uncovered in BEY 039 (see below), as well as a Middle to Late Bronze Age shoreline in BEY 069 (Marquis 2004), Marriner et al. (2008) have estimated a 70-metre progradation of the coastline between the Early Bronze Age and Roman period. As this figure is based on excavations, it is largely dependent on the dating of associated archaeological material. In this case, the feature uncovered in the BEY 039 sounding is often described as an Iron Age III/Persian quay, which has allowed it to be a marker for the rate of coastal progradation and the reformation of the ancient coastline from the Bronze Age through to the Roman period. It must be recalled that this identification is tentative, and the possibility of reuse of maritime installations in later periods could extend this timeline to the Hellenistic or Roman period, proposing a smaller degree of progradation. Thus, the Roman coastline proposed by Marriner et al. was likely much closer to BEY 039 than has been estimated. In the northern portion of the port basin, it appears that the coastline remained fairly stagnant over the past 6000 years (Marriner et al. 2008: 2508).

Cores Be III, Be V, and Be XX, located in the anchorage east of Burj al-Mina, have all revealed ‘medium grain marine sands’ which reflect an area not significantly sheltered (Marriner et al. 2008). It is possible that this sandy area could have been used as a fair-weather shelter from the Bronze Age onwards for shallow draught vessels, but based on recent analyses, it is certain that there did not exist a well-protected harbour comparable to that observed in the western basin. These results seem to corroborate the general situation of the Roman city with the main street leading to the harbour basin (Figure 5). The western edge of the eastern harbour basin lies near the outskirts of the city, and no harbour installations have been found in this area. Western basin The western basin is located west of Burj al-Mina and was hypothesised to be the ancient city’s main harbour (Davie 1987). As mentioned earlier, this basin underwent significant transformations over different periods of occupation and it is necessary to better understand the rate of coastal progradation to situate the harbour more accurately, assess the nature of the marine environment

Cores Be VIII, Be IX, and Be X were analysed in detail and samples were taken from individual phases of each core for radiocarbon dating. Each core depicts a change from a high-energy marine environment to a low-energy marine environment based on the sequence of sediments (Marriner

Figure 5. Roman street grid of Beirut based on recent excavations; Decumanus Maximus West (1) and East (2), Cardo Maximus South (3) and North (4); Imperial Thermae BEY 045 (5); Central Forum BEY 009 (6); Temple and large, domestic dwellings BEY 004 (7); Roman amphora kilns BEY 015 (8) (data provided by Hans Curvers).

67

Naseem Raad actually match those observed in BEY 007 quite closely. Thus, it is possible that the feature observed in stratum I in BEY 039 could be associated with the ‘sea wall’ observed in BEY 007.

et al. 2008: 2508). The transition from coarse-grained sand to silts and clays reflects the implementation of artificial harbour works in the Iron Age and Hellenistic/early Roman periods, which is corroborated by the faunal data observed in each core (Marriner et al. 2008). Interestingly, in Be VIII and Be X, there is a hiatus in sedimentary sequences observed between the Iron Age and Roman Period, which has been interpreted as the result of dredging practices in the Roman period which removed earlier strata from the geological record (Marriner et al. 2008: 2508). This pattern seems to be consistent with Roman harbours in general, as dredging and regular upkeep of harbours became much more widespread (Oleson 1988; Rickman 1988).

In stratum II, the construction technique implemented large headers set parallel to each other. This technique is quite typical of Phoenician construction (Iron Age to Hellenistic), though the actual manifestation of this technique is variable across the Levantine shore. In the harbour of Tyre, for example, large ashlar blocks of comparable size to those in strata II and III in BEY 039 are laid in the same fashion. Two courses have been observed in the underwater surveys conducted by Noureddine and Mior (2013), with a third protruding through the sediments at certain places. However, Tyre is notorious for the lack of consensus in the dating of the mole. For Carayon, the feature could be associated with a later period, possibly no earlier than Roman (2008: 651). Based on preliminary sedimentary soundings, it is supposed that there existed a previous mole, possibly dated to the Iron Age/Persian period (Carayon 2008: 651). However, the study of 70 pottery sherds collected during the excavation of the mole by Descamps and Sicre indicates an earlier date of construction, possibly between the 6th and 4th centuries BC (Castellvi et al. 2007: 68). The nature of the ashlar blocks used to construct the mole in Tyre and the technique utilised are quite similar to strata II and III of BEY 039. However, unfortunately, it is difficult to definitively date the features based on this comparison.

Ultimately, the results of the cores suggest Davie’s proposed location to have been quite accurate, with a continuous utilisation of the harbour since at least the Iron Age. The coastline in 1840 is based primarily on Davie’s map. The recreations by Marriner et al. of the Hellenistic-Roman and Bronze Age-Iron Age coastlines were constructed based on a combination of the rate of progradation, sediment analysis of the discussed cores and archaeological material. However, the interpretation of the quay uncovered in BEY 039 has proved to be key in this hypothesis, and the ceramic data from the cores and archaeological work in BEY 143 and 147 provide corroborative evidence. If the same installations were refurbished and reutilised in the Roman period, it would suggest the Roman and most definitely the Hellenistic coastline to be farther inland. The author has proposed this to have been the case, as depicted in the Roman coastline in Figure 1, which follows the quay uncovered in BEY 039, as well as the installations uncovered in BEY 007. Given that the ashlar blocks in BEY 007 and BEY 039 are almost identical, lie on a similar axis and were joined with mortar, the author proposes here that they be taken as contemporaneous.

The construction technique of joining two blocks together with a joint, sometimes reinforced with lead, is also observed at a number of sites throughout the eastern Mediterranean such as Elaia, Sarepta, Akko, Araq el-Amir and Dor (Elayi 2010: 160; Pritchard 1978; Sharon 1987: 38). As discussed earlier, the dove-tailed joint is often associated with Hellenistic construction, though it may have been utilised over a long period of time. At Elaia in modern-day Turkey (see Figure 6), the dove-tailed joint is used at a harbour with an estimated date of construction at the end of the third century BC (Seeliger 2016). Dovetailed joints are also observed in the massive harbour at Amathous, where the estimated date of construction is around 315 BC (Empereur 1987, 2016). The port of Amathous is also characterised by large, ashlar masonry and heavy, lead-enforced joints.

Construction and continuity This paper has combined multiple lines of evidence for a brief overview of the harbour in the Roman period. It is clear from sedimentary analysis of the cores undertaken in the western harbour basin, the archaeological evidence uncovered in BEY 007 and BEY 039, and the orientation of the city grid that the Roman harbour basin was located roughly in the area between Burj al-Mina and the Ottoman/ Late Roman quays identified in BEY 007. Furthermore, based on this evidence, it seems that harbour installations that have been unequivocally attributed to the Iron Age and early Hellenistic period were likely refurbished and reutilised in the Roman period. More specifically, stratum I of the quay identified in BEY 039 appears to represent a phase of construction associated at least with the Hellenistic Period and likely the Roman Period. The use of mortar differentiates this layer from strata II and III, and the Roman coins and ceramics suggest a later date than the previous two. The blocks used in stratum I are smaller in size than the other two, and

Figure 6. Example of dove-tailed joint observed in breakwater at Elaia (photo courtesy of Nicholas Carayon).

68

The Roman Port of Berytus However, in the Roman port at Sarepta, a number of blocks of the Roman jetty were joined together using a similar joint. The date of construction of the first phase at Sarepta has been dated to the first century AD (Pritchard 1971). These blocks were of a comparable size to those observed in BEY 007 as well as stratum I in BEY 039, and excavations at Sarepta also uncovered a number of large tanks reminiscent of those observed in BEY 007, though the ones in Beirut were severely truncated. Thus, it is difficult to utilise a single factor to date construction phases at the quays in Beirut; rather, the combination of all maritime installations as well as the geomorphological analysis point to an initial construction date in the Iron Age for the first phase, with refurbishment in the Hellenistic period and later in the Roman period.

Figure 7. Northern section of sounding in BEY 143 depicting amphora sherds at bottom layer (photo provided by Hans Curvers, ceramic dating conducted by Paul Reynolds).

The evidence indicates a significant degree of continuity in the utilisation of the harbour at Beirut from the Iron Age to the Roman period. Previous courses served as the foundation for later Roman ones. Despite the lack of archaeological preservation of the upper remains, also resulting in the truncation observed in BEY 007, the available evidence points to an Iron Age instalment of a quay near the Iron Age settlement in BEY 010, a refurbishing of this quay in the Hellenistic period and again in the early Roman period. Strata II and III of BEY 039 both revealed large, ashlar blocks (larger than stratum I and the maritime installations in BEY 007) joined together with the dove-tailed joint, and found in association with Iron Age and early Hellenistic archaeological material. Given that ashlar blocks of stratum I of BEY 039 and the quay in BEY 007 were not joined together with dove-tailed joints, and were bonded with mortar, the author proposes that these features can be differentiated from the other strata discussed and represent late Hellenistic/early Roman developments. This would also match the increase in private and public construction seen in the city itself in the Augustan period, probably associated with the arrival of Roman veterans to the newly-formed colony.

were taken roughly in the vicinity of BEY 143 and 147, which are located at a key intersection within the Roman city between land and sea. Specifically, BEY 143 lies just outside the abovementioned archaeological remains in BEY 039, and BEY 147 lies in the centre of the presumed ancient harbour basin. Therefore, although no maritime installations were uncovered on either site, they are in an important position to shed light on the stratigraphic sequences within the ancient harbour basin and provide comparative data for strata I-III from BEY 039. The deepest soundings at these sites have revealed Hellenistic diagnostics, first-century AD bowls and other Roman ceramics (Figure 7) (Curvers and Stuart 2007: 189). Crucially, this material underlays a series of strata, one of which may be directly correlated with a low-energy marine environment sediment. The overlaying stratum has been dated to about 500–800 AD, indicating a gap in the sediments from the first century AD to the 6th century AD, and further proof of cleaning operations in the harbour basin (Curvers, personal communication). A similar pattern has also been observed at BEY 147 (Curvers and Stuart 2007: 191). The fact that ceramics from the first century AD have been uncovered even after dredging activities in the Roman period indicates that this portion of the harbour would have been use in the Roman Period.

Thus, the Roman coastline proposed by the author in Figure 1 is largely based on the work of Marriner et al. (2008), as well as the research of Davie (1987), but adapted according to the archaeological analysis in this paper. The author suggests the coastline in the Roman period to have followed the installation uncovered in BEY 007 (as it appears that the feature can be tentatively dated to the Roman period), as well as the quay in BEY 039 (given that stratum I appears to be a refurbishment in the early Roman period). If stratum I from BEY 039 and SWP1 from BEY 007 are to be understood as contemporaneous, the actual coastline in the Roman period would have been about 70 metres farther inland than initially anticipated.

Concluding thoughts Though some of the data is fragmentary, a holistic examination of BEY 007, 010 and 039 suggests an active harbour throughout the Hellenistic and Roman Periods. Having established that the harbour was in use in the Roman period and well-maintained, it would be expected that commercial connections should reflect this activity. However, it must be stated that future work that undertakes this endeavour cannot definitively indicate causation, but only correlation between commercial success, urban growth and harbour upkeep. In other words, it is a fairly straightforward process in tracing the distribution of a Beirut product around the Mediterranean, quantifying the finds and comparing commercial patterns to fluctuations in

BEY 143 and 147 This hypothesis can be further corroborated when considering BEY 143 and 147. The sedimentary cores 69

Naseem Raad CASTELLVI, G., DESCAMPS, C., PORRA-KUTENI, V., SALVAT, M., and SICRE, J. 2007. ‘Recherches sousmarines à Tyr’. Bulletin d’archéologie et d’architecture Libanaises 11: 57–102.

urban development. It is much more difficult making the jump to stating Beirut grew and expanded during a certain time period because of an expansion in commercial distribution. Regardless, these are considerations that will be undertaken in the next step of the author’s research, and will hopefully provide another line of evidence with which to better understand the Roman port of Berytus, the primary node in a complex and multi-faceted commercial maritime network.

CURVERS, H. H. and STUART, B. 2007. ‘The BCD Archaeology Project 2000–2006’. Bulletin d’archéologie et d’architecture Libanaises 9: 189–221. DAVIE, M., 1987. ‘Maps and the Historical Topography of Beirut’. Berytus Archaeological Studies 35: 141–64.

Acknowledgments

ELAYI, J. 2010. ‘An unexpected archaeological treasure: The Phoenician quarters in Beirut city center’. Near Eastern Archaeology 73.2–3: 156–68.

The author would like to thank the Honor Frost Foundation for generously sponsoring this research at the University of Southampton. The foundation’s support has been invaluable in furthering our understanding of maritime commercial networks in the eastern Mediterranean.

ELAYI, J. and SAYEGH, H. 1998. ‘Un quartier du port phénicien de Beyrouth au Fer III/Perse: Les objets’. Transeuphratène. Supplement no. 6. Gabalda: Paris.

Thank you also to Hans Curvers for kindly providing metadata and sharing information about his work in collaboration with Paul Reynolds in key areas of Beirut.

ELAYI, J. and SAYEGH, H. 2000. ‘Un quartier du port phénicien de Beyrouth au Fer III / Perse’. Transeuphratène. Supplement no. 6. Gabalda: Paris.

Works Cited

EMPEREUR, J.-Y. 1987. ‘The Underwater Excavation at the Ancient Port of Amathus in Cyprus’. The International Journal of Nautical Archaeology and Underwater Exploration 16.1: 7–18.

Primary sources JOSEPHUS. 1926–65. Antiquitates Judaicae. Trans. Thackeray, H. St. J. Heinemann: London. LUCIUS JUNIUS MODERATUS COLUMELLA. 1968. On Agriculture. Trans. Ash, H. B. Harvard University Press: Cambridge, MA.

EMPEREUR, J.-Y. 2016. Ancient Port of Amathus. Accessed January 2018. Available from .

NONNUS. 1940. Dionysiaca. Trans. Rouse, W. H. D. Harvard University Press: Cambridge, MA.

GALILI, E., and ROSEN, B. 2008. ‘Akko Harbor: New Finds Revealed While Deepening the Port’. The New Encyclopedia of Archaeological Excavations in The Holy Land 5: 1558–61.

PLUTARCH. 1920. Plutarch’s Lives. Trans. Perrin, B. Harvard University Press: London. STRABO. 2001. Geographia. Trans. Jones, H. L. Heinemann: London.

HAGGI, A. 2010. ‘Report on Underwater Excavation at the Phoenician Harbour, Atlit, Israel’. The International Journal of Nautical Archaeology 39.2: 278–85.

Secondary sources

HALL, L. J. 2002. ‘Berytus Through the Classical Texts: from Colonia to Civitas’. ARAM 13–14: 141–69.

BLACKMAN, D. J. 1982a. ‘Ancient Harbours in the Mediterranean. Part 1’. International Journal of Nautical Archaeology and Underwater Exploration 11.2: 79–104.

HALL, L. J. 2004.  Roman Berytus: Beirut in Late Antiquity. Routledge: London and New York.

BLACKMAN, D. J. 1982b. ‘Ancient Harbours in the Mediterranean. Part 2’. International Journal of Nautical Archaeology and Underwater Exploration 11.3: 185–211.

LAUFFRAY, J. 1977. ‘Beyrouth archéologie et histoire, époques gréco-romaines. I. Période hellénistique et Haut-Empire romain’. In H. Temporini and W. Haase (eds.), Geschichte und Kultur Roms im Spiegel der neueren Forschung. Berlin and New York. Pp. 135–63.

BUTCHER, K. 2003. Roman Syria and the Near East. The British Museum Press: London.

MARQUIS, P. 2004. ‘Les fouilles du centre-ville’. In C. Doumet-Serhal (ed.), Decade: A Decade of Archaeology and History in the Lebanon. Archaeology and History in Lebanon. The Lebanese British Friends of the National Museum: Beirut. Pp. 266–279.

BUTCHER, K. and THORPE, R. 1997. ‘A note on excavations in central Beirut 1994–96’. Journal of Roman Archaeology 10: 291–306. CARAYON, N. 2008. Les ports phéniciens et puniques. Géomorphologie et infrastructures, these de doctorat non publiée. Université March-Block: Strasbourg.

MARRINER, N. 2009. Geoarchaeology of Lebanon’s Ancient Harbours. International Series 1953. BAR Publishing: Oxford.

CARAYON, N., MARRINER, N., and MORHANGE, C. 2011. Geoarchaeology of Byblos, Tyre, Sidon and Beirut. Rivista di Studi Fenici 39.1: 45–55.

MARRINER, N., MORHANGE, C., and BEYDOUN, M. 2008. ‘Geoarchaeology of Beirut’s Ancient Harbour, 70

The Roman Port of Berytus Phoenicia’. Journal of Archaeological Science 35: 2495–2516.

REYNOLDS, P. 2003. ‘Amphorae in Roman Lebanon 50 BC to AD 250’. Archaeology and History in Lebanon 17: 120–30.

MARRINER, N., MORHANGE, C., KANIEWSKI, D., and CARAYON, N. 2014. ‘Ancient harbour infrastructure in the Levant: tracking the birth and rise of new forms of anthropogenic pressure’. Sci. Rep. 4: 5554.

REYNOLDS, P., CURVERS, H., LEMAÎTRE, S., NSOULI, B., ROUMIÉ, M., and WAKSMAN, S. Y. 2010. ‘An Early Imperial Roman Pottery Production Site in Beirut (BEY 015): Chemical Analyses and a Ceramic Typology’. Berytus 51–52: 71–115.

MARTIN, R. 1965. Manuel d’architecture grecque. Paris.

RICKMAN, G. E. 1988. ‘The Archaeology and History of Roman Ports’. The International Journal of Nautical Archaeology and Underwater Exploration 17.3: 257– 67.

MILLAR, F. 1993. The Roman Near East, 31 B.C.-A.D. 337. Harvard University Press: Cambridge. NOUREDDINE, I. and MIOR, A. 2013. Archaeological Survey of the Phoenician Harbour at Tyre, Lebanon. Honor Frost Foundation, Publication forthcoming.

SARTRE, M. 2005. The Middle East Under Rome. Belknap Press: Cambridge and London.

OLESON, J. P. 1988. ‘The Technology of Roman Harbours’. International Journal of Nautical Archaeology 17: 147–57.

SEEDEN, H. and THORPE, R. 1997. ‘Beirut from Ottoman sea walls and landfills to a twelfth century BC burial: Report on the archaeological excavations in the Souks northern area (BEY 007)’. Berytus archaeological studies 43: 221–54.

PERRING, D. 1997. ‘Excavations in the Souks of Beirut: an Introduction to the Work of the Anglo-Lebanese Team and Summary Report’. Berytus Archaeological Studies 43: 9–34.

SEELIGER, M. 2016. Elaia, the Maritime Harbour City of Ancient Pergamon (Turkey) Coastal Evolution and Human Impact over the past Eight Millenia. PhD Dissertation. Universität zu Köln.

PERRING, D., REYNOLDS, P., and THORPE, R. 2003. ‘The Archaeology of Beirut: A Report on Work in the Insula of the House of the Fountains’. The Antiquaries Journal 83: 195–229.

SEELIGER, M., BARTZ, M., ERKUL, E., FEUSER, S., KELTERBAUM, D., KLEIN, C., PIRSON, F., VÖTT, A., and BRÜCKNER, H. 2013. ‘Taken from the Sea, Reclaimed by the Sea: The Fate of the Closed Harbour of Elaia, the Maritime Satellite City of Pergamum (Turkey)’. Quaternary International 312: 70–83.

PRITCHARD, J. B. 1971. ‘The Roman Port at Sarafand (Sarepta). Preliminary Report on the Seasons of 1969 and 1970’. Bulletin du Musée de Beyrouth 24: 39–56. PRITCHARD, J. B. 1978. Recovering Sarepta, A Phoenician City. Excavations at Sarafand, Lebanon, 1969–1974, by the University Museum of the University of Pennsylvania. University of Princeton: Princeton.

SEMAAN, L. 2015. Desk based assessment of maritime archaeological sites on the Lebanese coast. Honor Frost Foundation, UK.

RABAN, A. 1991. ‘Minoan and Canaanite Harbours’. In R. Laffineur and L. Basch (eds.), Thalassa: l’Egée préhistorique et la mer. Liège. Pp. 129–46.

SHARON, I. 1987. ‘Phoenician and Greek Ashlar Construction Techniques at Tel Dor, Israel’. Bulletin of the American Schools of Oriental Research 267: 21–42.

REYNOLDS, P. 1999. ‘Pottery production and economic exchange in 2nd century Berytus: some preliminary observations of ceramic trends from quantified ceramic deposits from the Anglo-Lebanese Excavations in Beirut’. Berytus 43: 35–110.

STUART, B. 2002. Cemeteries in Beirut. ARAM, 13–14: 87–112. THORPE, R., BEYHUM, A., KOULY, S., and BEAYNO, F. 1998. ‘BEY 007 The Souks Area. Preliminary Report of the AUB/ACRE Project’. Bulletin d’archéologie et d’architecture Libanaises 3: 31–55.

REYNOLDS, P. 2000a. ‘Baetican, Lusitanian and Tarraconensian amphorae in classical Beirut: some preliminary observations of trends in amphora imports from the western Mediterranean in the AngloLebanese Excavations in Beirut (BEY 006, 007, and 045)’. Congreso Internacional ‘Ex Baetica Amphorae’ (Sevilla-Écija, December 1998). Editorial Gráficas: Écija. Pp. 1035–65. REYNOLDS, P. 2000b. ‘The Beirut Amphora Type, 1st century BC-7th century AD: an outline of its formal development and some preliminary observations of regional economic trends’. Rei Cretariae Romanae Fautorum Acta 36 (Selçuk, 27 September-6th October 1998) 36: 387–95. 71

7 The Study of the Port System of the Coast of Almería from the Analysis of the Maritime Cultural Landscape, Eighth – Twelfth Centuries AD Marta Del Mastro Ochoa Universidad de Cádiz [email protected] Abstract: The study of the maritime cultural landscape of a medieval port system requires a multidisciplinary approach. It implies not only an extensive theoretical knowledge about the area and the chronological framework of study, but also the understanding of the object of study as a maritime culture, with the nautical perspective that this requires. This paper explains some concepts for the study of the maritime cultural landscape of the port of Almería, and sheds light on some of the elements that were part of the port system in medieval times. These elements have been identified through an analysis of a mosaic of sources: written, iconographic, cartographic and archaeological. These sources have been studied with a diachronic perspective, in order to understand why this landscape was so useful to be the main port of south Al-Andalus in caliphal and taifas times. The project includes in-depth space analysis, which was developed through geographic information systems (GIS). Key Words: Almería; maritime cultural landscape; Middle Ages; space analysis; geographic information systems (GIS) Introduction

chronological period. A complete historiography on the concept is found in the thesis of the archaeologist Felipe Cerezo (Cerezo 2016: 17–19). From all the available information, we can conclude that the maritime culture of a past society is noticed in activities such as traditional fishing, the construction and maintenance of wooden boats by riparian carpenters or shipbuilders (both more difficult to find and preserve), the anchorage and berthing of boats, coastal defences, the construction of buildings destined to manage the administrative documentation (if it exists) of maritime trade, or even the management of the port activity itself.

This work, presented during the second day of the Maritime Archaeology Graduate Symposium 2018, consists of an extract of the preliminary conclusions of the MA project entitled ‘The study of the port system of Almería between the eighth and twelfth centuries AD’, which was defended on February 2018 at the University of Cádiz. In this paper, I explain how we used the methodology of a maritime cultural landscape approach in order to shed light on the maritime culture of Almería in a specific chronological framework. In this way, I will justify why the port of Almería was located in the area that it was, despite its evident natural weakness that left it open to human attacks by the sea, as well as how it tried to solve this problem while also developing port activities.

Subsequently, the imprint left behind by these activities in the landscape is what we call the ‘maritime cultural landscape’. This term was used and developed by Westerdahl (1992), though recent work has seen the advent of new perspectives (Ford 2011). Nowadays, one of the most interesting discussions is about the cognitive landscape approach, which has been shown by Westerdahl to be a very useful concept to understand, for instance, the borders of a maritime cultural landscape (Westerdahl 2006: 7–54). Thus, we could define it as the remains left by the alteration, adaptation, anthropisation or exploitation of a maritime zone by a human group, whether it alters the landscape in a physical or visual way, and whose main objective is to appropriate that space with a maritime function. This can range from obtaining maritime resources, to facilitating navigation, grounding

First of all, we need to explain the meaning of some concepts in order to set a theoretical framework in which to present our arguments. One of the most important is, of course, what we understand as maritime cultural landscape. A maritime culture can be defined as the material, ideological, graphical or written manifestation of the relationship of a human group with the sea upon which its survival depends. This is a relatively recent term. During the 1930s, it was mentioned by Scandinavian researchers Ake Campbell and Olaf Hasslöf, who defined the concept as the sum of all material remains, structures of economic exploitation, hunting and fishing that form a society independent of its 73

Marta Del Mastro Ochoa and anchoring, to defending an area which is susceptible to being attacked by human groups that don’t exercise the dominant authority in the geographical space.

were articulated around three classifications: coastal anthropic elements, natural elements and underwater anthropic elements. The first mainly corresponded to elements which were built expressly for a nautical function; the second one refers to those which were used for nautical purposes, but non-altered. The anthropic underwater ones are those archaeological remains that were documented in underwater contexts, and characterised as part of an anchorage, port or shipwreck.

The study of maritime cultural landscapes allows, and usually demands, the development of several lines of research, particularly: the study of nautical conditions, the naval architecture and typology of the ships according to the historical period, studies on types of navigation according to the chronology and the zone, and the diachronic study of port systems. The landscape must be analysed from several perspectives with a focus on the interrelation of the sources, and subsequently, the data. In this way, the data has been divided into three large contexts. The first one is called a micro-spatial context: analysis of the material remains by zones (pottery, metal or isolated elements). This provides information regarding the dispersion of materials in specific areas.

Thirdly, when possible, a spatial analysis was carried out using geographic information systems (GIS) in order to verify the ‘maritime quality’ of some of the elements. GIS has turned out to be a tool increasingly used in recent years. Their importance lies in the possibility of using qualitative data as quantitative data, and integrating them into a cartographic system, which allows for establishing patterns with a considerably more objective criterion. In this sense and only in those cases in which it was possible, a basic spatial analysis of visibility and accessibility was undertaken in order to ascertain the ‘maritimacy’ of some of the elements and spaces. With this study, I sought to gather enough evidence to explain the role of these elements in the port system of Almería.

The second, specified in this paper as the territorial context, is important as well. By linking the spaces, an analysis of the impact of human action on the territory can be carried out (construction of docks, shipyards, towers, shipping and fishing structures). This provides information on settlement patterns, catchment areas, landscape transformation and the palaeo-topography or palaeo-landscape.

‘Good’ visibility conditions have been defined as situations where a terrestrial signal can be observed from any point of the sea based on a specific maximum distance directly proportional to its height (Izquierdo i Tugas 1996: 300). This visibility can be analysed through GIS, as shown by Cerezo in his thesis on the Cartagena port system (Cerezo 2016: 691–701). In the case at hand, it is intended to propose some hypotheses about the functionality of some elements of the maritime cultural landscape, both natural and anthropic. The primary aim is to shed light on which spaces were visible from the sea in coastal navigation and at what maximum distance they would be visible. This should provide us with quantifiable data about the suitability of building towers (which could also have been used as lighthouses) that were recognisable from the coast, including some mentioned in ancient sources that have not survived to this day. On the other hand, it is also important to analyse some of the anthropic elements found on the sea front of Almería (the ribats of La Chanca, the alcazaba fortification and the tower of al-Idrῑsῑ), in order to check their suitability for the function that they traditionally fulfilled, especially those that comprised, according to ancient sources, part of the early medieval coastal defensive system.

The last one is called the cognitive context. The analysis of the cognitive landscape consists of studies that explore the degree of social impact of the maritime landscape on the population, that is, how it is perceived and how it is adapted or transformed. It provides information on reference points or delimiting elements, for example. One of the best ways to get this type of information is through the analysis of toponymy, both functional and/or sacralised (Westerdahl 2006). Methodology Firstly, one the most extensive and important steps was the identification and reinterpretation of the sources of study for the maritime cultural landscape, with a nautical perspective in order to get nautical information. The sources included were: written, cartographic, iconographic, epigraphic and archaeological, from both land and underwater archaeology. These were selected based on a diachronic perspective; this implies that information which came from older and later sources than our chronological framework was compared with the analysed material. In the case of ancient sources, the objective was to try to reconstruct the palaeo-landscape of the eighth century AD in an interpretative way to allow us to understand why that space, and not another, was chosen to have a port function. In the case of the later sources, the objective was also to carry out a regressive study that would allow the identification of elements of the maritime cultural landscape that, nowadays, have disappeared.

We must bear in mind that, although it is not represented in the analysis exposed here, there are other factors that have been recently studied, besides distance, that can alter visibility, such as water temperature, atmospheric suspension, or cloudiness, which affect the perception of the observer with respect to the visual line (Cerezo 2016). The analyses are subject to optimal visibility conditions, so they only provide binary data (visible and not visible). The visible spaces are those coloured on the map, while the non-visible spaces are called ‘shadow’ spaces. The

Secondly, the identification and classification of the elements in databases was carried out. These databases 74

The Study of the Port System of the Coast of Almería Height of the Viewer, that is, the difference of the observer with respect to the level of the ground, has been introduced manually in each anthropic element with the same average value (1.7 metres).

and wine trade have also been documented (Blánquez et al. 1998). It seems to indicate that the area was a point where nautical activities would take place between the first and fifth centuries AD.

A privileged geographical situation

Moreover, if we pay attention to the current orography of the sea front of Almería, we can see that the location of the modern city, as well as its port, is the most suitable area to perform anchoring activities. As seen in Figure 1, the port area of Almería is currently located east of Sierra de Gador, which is between 100 and 190 metres above sea level up to two kilometres away from the coastline. This situation offers privileged protection from winds from the southwest (they are most frequent during the summer, with speeds exceeding 8 m/s at a frequency of 20–24 per cent). The Sierra de Gador is a natural element which already existed during the Medieval Period, as observed in geomorphological maps of the area. Furthermore, the winds from the southwest are a recurrent nautical condition in the sources that wrote about Almería (Tofiño de San Miguel 1787), so it is assumed that this point was the most interesting area to structure the port. Urban archaeology has confirmed that the medieval settlement was located in the same emplacement as the current city (Alcalá et al. 2005: 94–102, Cara et al. 2005: 167–92).

Today, Almería is the south-eastern Spanish province of the Iberian Peninsula. It is a territory with around 219 km of coastline, whose coastal landscape has two very different areas: the coastal landscape of the Alborán Sea, which would mainly include the Gulf of Almería, located between the geographical features of Cabo de Gata and Punta Entinas, and the coastal landscape of the east coast that extends between the promontory of Cabo de Gata and the municipality of Águilas. This zone is the easternmost point of the peninsula adjacent to the Mediterranean Sea, probably making it conducive for maritime activities since antiquity. Referring back to the comparative diachronism in the study exposed here, we can see that the area corresponding to the Gulf of Almería is mentioned in sources long before our chronological frame mark. In the case of the Gulf of Almería, toponymical analysis has proved useful in explaining the transcendence of this natural element within the maritime cultural landscape was. Both Pomponius Mela, first century AD, and Claudius Ptolemy, second century AD, refer to this maritime space with place names which were describing good conditions to carry out nautical activities, specifically for anchoring and sheltering boats.

The location of Almería was strategic not only for its useful anchorage points, whose deepness allowed the anchorage of many types of ships for a long period (mainly between the ninth and nineteenth centuries AD), but also for its watering points as well. Watering points are crucial during sailing, so any important port needed some way to provide this for traveling sailors. As far as this issue is concerned, Casson’s analysis (1989) provides one of the most accurate maritime itineraries of antiquity. During the voyage, watering points are mentioned as a very important element to anchor in a port. This type of indicator is also recurrent in the Medieval Period (al-Idrῑsῑ ‘Nuzhat), and in the sailing courses of modern times (Tofiño de San Miguel 1787). This fact highlights the importance of identifying watering points to understanding the most useful places to anchorage or to build a port, especially in geographically isolated areas (Almería is a region surrounded by two mountains ranges). The written sources of the Middle Ages highlights Almería as a city with some baths (alIdrῑsῑ ‘Nuzhat 15.4), so it indicates that access to water was probably relatively easy. This access was likely provided by ‘ramblas’. Rambla is a Spanish word which describes a natural channel formed by water when it rains over a long period. They are very common in the south and east coast of the peninsula. Furthermore, they are generally considered seasonal, so their flow is not constant over time, and they were occasionally torrential and dangerous. This implies that there may be times in history when they could have had a stronger flow, though this does not mean they were navigable (we should remember they are torrential currents, ergo erratic). Rather, a rambla is an abrupt opening in a sharp coast, which allows a ship to disembark materials or people in a more stable area to reach inland.

Pomponio Mela, in De Chorographia Liber Secundus, called this area ‘sinus urcitanus’ (Mela 2.84), which probably referred to the area of Chuche, Pechina (Lirola 2005: 45), while Ptolemy, in his Geographia, did the same with the Latin concept ‘Portus Magnus’ and ‘magnos liman’, the original Greek term (Nobbe 1843: 76). Both terminologies can be translated as ‘the great port’. This indicates that, on the one hand, we should be inclined to think that the Gulf of Almería was already, in ancient times, an interesting point for the development of nautical activities, and that, by making this term a reference to the geomorphological characteristics of the space and not the monumentality of its built structures, this area wouldn’t need to have a particular anthropised space, because its orography gave it the natural characteristics needed to be a maritime space of great interest. To reaffirm this hypothesis, we need to highlight the existence of the archaeological complex of Turaniana or Ribera de la Algaida, located between the present municipality of Roquetas de Mar and Aguadulce (at the western part of the gulf), which were possibly occupied from the first to the fourth centuries AD. At the same time, we must point out the existence of some salting rafts found a few metres away from the coast near medieval shipyards, also of Roman chronology, possibly dating to sometime between the first and third centuries AD. The underwater archaeological remains associated with the salt 75

Marta Del Mastro Ochoa There are three significant ramblas close to the city which need to be assessed as part of the maritime cultural landscape. The first of them is called the Rambla of Almería. It is located approximately 3.6 km east of the modern city of Almería, and is the river mouth of the Andarax River in the gulf. Its seasonal and torrential character, as well as the urbanisation of the immediate areas next to it, has caused it to have become practically dry. The sources that have led us to locate this element have been considerably scarce. One of these sources, the sailing course of Vicente de Tofiño points out the place where the rambla is located as the ‘Punta of the river of Almería’, referring to the delta cone of this rambla (Tofiño de San Miguel 1787: 48). However, when Tofino is referring to the watering points of Almería and its harbour, he did not write about this specific element, but of a number of wells and fountains that existed in the city. On the other hand, this element was not mentioned in the sailing courses of Alonso de Contreras, nor in the Kitab of Piri Reis from the sixteenth century, nor in the portolan Compasso de Navegare from the twelfth century AD (Debanne 2011). It is probable, then, that their utility as a point to get drinking water would have been overshadowed by the supply that, according to the sources, provided the city of Almería.

This fact indicates that these points were easily accessible, so the city likely needed to protect them with a structured defensive system. A self-defenced port: a way to protect an important but naturally unsafe maritime area Defensive structures analysed through the written sources The importance of defensive structures within the maritime cultural landscape of Medieval Almería was evident from the beginning of the analysis. When data collection was completed, we realised that the number of defensive structures (among which are walls, towers, shipyards and ribats) was 47 per cent of the total (see Figure 2). Practically all were in the settlement of Almería, as well as along the gulf (see Figure 3). The fact that almost half of the anthropic elements that emerged from the documented landscape had a mainly defensive function reveals several elements. One possibility is that the nautical, anthropic elements would had been built with perishable materials, so they have not been preserved. One of the materials could have been wood, as evidenced by the port of Comacchio in Italy (Gelichi 2012), despite the fact that the defensive structures had been built with stone. But it is also possible that the unsafe area of Almería (the coast without defensive natural barriers, such as reefs) required the construction of some defensive structures in

The other two ramblas are known as the Chancla’s and Portillo’s ramblas. It is important to point out that the Medieval city was articulated between both ramblas, which limited it to the west and east, respectively. In addition, the walls surrounded them, as we can see in the modern plan designed by Felipe Crame in 1740 (Figure 1).

Figure 1. Medieval city of Almería (after Crame 1740).

76

The Study of the Port System of the Coast of Almería to a defensive and organised structure off the coast of the Gulf of Almería. This gulf, as discussed in this paper, has been mentioned since the time of Pomponius Mela and Ptolemy as an optimal space for the anchorage of boats and developing of nautical activities. Moreover, in its origin, this place name did not specifically refer to the city but to a defensive system of the port, which was associated with the city of al-Bayyᾱna (Pechina), located about eight kilometres from the sea.

order to keep the port safe. This is crucial for this study, because it shows that the port was important enough to grow and be actively developed, despite the fact that the area was not naturally defensible. The importance of the defensive structure as an older element of the maritime cultural landscape of Almería is shown, for example, in the place name of the city itself. This has an Andalusian origin; the first mention we have about it is in the tenth century, from the historian Ibn Ḥawqal, who called the city al-Mariya (Ibn Ḥawqal Kitᾱb 2.3). Some interpretations about the etymology of the word Al-Mariya were collected by the archaeologist Cara Barrionuevo (1990: 3). In this paper, it is considered that the only sufficiently argued case is that of the watchtower (Atalaya in Spanish) defended by Yaqut (‘Abd al Karim 1974: 284). Nowadays, this is the largely supported hypothesis about the origin of Almería’s place name. In this way, we can see how, in its origins, it seems to refer

The documentation of Ibn Ḥawqal’s mention is the oldest we have regarding this settlement. Al-Ḥimyarῑ, in the fourteenth century AD, describes an interesting historic context about the settlement of Almería in the ninth century AD. Al-Ḥimyarῑ points out that during the emirate (eighth to tenth centuries AD), a group of ‘al-Ándalus rabble people’ prepared some ships to act in a ‘piratical’ way, attacking the unprotected coasts, which indicates that, indeed, there was no particularly strong defence of the coasts at that time. These people were called albaḥriyyῡn (the sailors) by al-Ḥimyarῑ (al-Ḥimyarῑ Kitᾱb) which is also telling us that they were accustomed to live and subsist mainly from the sea. These al-baḥriyyῡn, according to al-Ḥimyarῑ, were the Yemenis who settled down in the Urṧ / Arṧ al-Yaman and founded the settlement of Pechina. Thus, it seems likely that the port was erected first and, to keep it protected, some defensive structures were built, which gave rise to the origin of the name of the current city of Almería.

Figure 2. Graphic of the anthropic elements of the Maritime Cultural Landscape of Almería.

We have also documented two decisive attacks on this port city: the attack of a Fatimid fleet in the ninth century

Figure 3. Map of the anthropic elements documented on the coast of Almería.

77

Marta Del Mastro Ochoa Defensive structures examined through visibility analysis

and the attack of the Genoese and Catalan fleet of the twelfth century. These moments were pivotal for the maritime history of Almería because they explain why the city was equipped with such concrete defensive structures such as towers, walls, ribats and shipyards, and imply that the port was a significant influence for the development of al-Ándalus.

As we can see in the following simple and strictly theoretical visibility analysis, the two ribats were located strategically to cover a great distance (Figure 4). The ribats located in the Chanca and the Alcazaba offer an almost complete visual basin of the Gulf of Almería (41 km from Punta Entinas to Cabo de Cat horizontally, and 34.9 km in a straight line perpendicular to the ribāts). The zone of ‘shadow’ to the west (where the current city of Aguadulce is located which is referred to by the sources as a rural settlement) is especially significant, where it would be expected that there were other coastal constructions that complemented the visual defence of the coast.

Regarding these defensive structures, we identified four specific types, according to the function they had in our chronological framework. We acquired information about walls (surrounding the urban settlement of Almería), ribats and towers. They are analysed in this paper in order to understand their function and character in the maritime cultural landscape of Almería during the Middle Ages.

On the other hand, and given that our study is based on the perception of the landscape from the sea, it seemed significant to us to do the same analysis, but in reverse (Figure 5). That is, from several random points located in the sea towards land, in order to analyse which areas of the coast are visible from the sea under favourable conditions, and, therefore, could constitute points of orientation in coastal navigation. As can be seen in the analysis, the zones of the ribats are only visible from the sea at approximately 9 km from the coastline (Figure 6). This detail is interesting since, if we combine both analyses, we see that the point from the sea that we have called Point 4 (P4) is in the shadow zone of both ribats, but oddly, these and the city are visible from the point in question.

The ribats were defined as a specific type of defensive structure, not necessarily dependent on a city’s construction, but rather, with their own architectural characteristics (Azuar 2004: 23). In our database, the concept of ribat has been understood as a defensive space in which, following the definition of Lirola (2005), surveillance functions were carried out in service to the community in the context of a spiritual retreat (in our case, the monitoring of adjacent coasts). The ribat and their complex structure in our chronological framework could be studied with archaeological methodology, mainly through comparison with the archaeological complexes of Ribat al-Munastīr in Guardamar del Segura (province of Alicante) and Ribat al-Rayhāna in Algarve (Portugal). However, in our case, practically all the mentions that have been documented are only evidenced in the written record, so we do not know the specific architectural characteristics of the ribats in our area. Through the analysis of the sources, we could identify two ribats, both in the area of the present city. One of the first written testimonies we found about them was by al-Ḥimyarῑ. He points out that, after the arrival of the Normans on the northern coast of the Alboran Sea (that is, in the middle of the ninth century AD) some watchtowers were built, whose functions were carried out through the ribat. The historians Tapia Garrido (Tapia 1986: 118) and Andrés García Lorca (García 1990: 39) pointed out that one of them could have been in the corner of the district La Chanca, which is close to the old part of the city of Almería in the west. They considered this area the most suitable for a main anchoring zone without the need for port infrastructure, a condition that Tofiño highlighted in his sailing course and that we have explained before. Both authors, Tapia Garrido and García Lorca call it ‘The ribat of San Roque’, probably referring to the hermitage of San Roque (today San Roque’s church), in front of which Alonso de Contreras, in his seventeenth century AD sailing course, indicated a deep enough area to shelter the galleys. It is possible that the location of this ribat corresponds to one of the rábita de la despedida (the farewell church).

Figure 4. Visibility analysis of the Alcazaba (red circle) and Chanca (green circle) ribats, in which it is possible to see the shadow zone in the north-west of the gulf; the viewer height used is 6.7 m (5 metres for the building and 1.7 for the viewer).

78

The Study of the Port System of the Coast of Almería

Figure 6. Visibility analysis with the al-Idrisi and shipyard towers, and the Alcazaba and la Chanca ribats; the visible zone passes through the al-Idrisi tower (in pink), which means the shadow zone in Figure 4 is covered.

Figure 5. Visibility analysis of the points from the sea; the viewer height used is 1.7 m.

This is another piece of information that suggests that there was some coastal construction in relation to these ribats, around Aguadulce and Roquetas de Mar (the western part of the Gulf of Almería): the towers. In this sense, the testimony provided by the Andalusian author alIdrῑsῑ of the twelfth century proves useful, which tells us of the existence of a tower 9 km from Aguadulce. Currently, the distance mentioned by al-Idrisi between the tower and the rural settlement of Aguadulce is six miles. This was combined with Lirola’s hypothesis about the conversion, which is based on Ptolemy’s miles, and not Arabic ones (Lirola 2005: 63). With this conversion, we realised that the distance between Aguadulce and Roquetas de Mar were quite close (8.87 km), especially if we utilised the distance over the Santa Ana’s castle in Roquetas de Mar, which was built in recent times in order to control the port. If we look at the visibility analysis (Figure 6), we see that, effectively, the existence of this tower eliminates the ‘shadow’ zone (but the seafront of Almería is not visible), thus covering the entire coastline.

Figure 7. Visibility analysis with the al-Idrisi and shipyard towers, and the Alcazaba and la Chanca ribats. The visible zone through the al-Idrisi tower (in purple) shows that it is possible to see the shipyard tower and the shipyard from there.

Another tower was identified in the urban settlement of Almería, which has been dubbed ‘the Shipyards Tower’ due its location. The only evidence we know of its existence is provided by the geographer al-ᶜUḏrῑ from the eleventh century, who, in his story about the shipyards of Almería, states that, at the door of the tower and towards the sea, a tower existed that was destroyed by Aflah (alᶜUḏrῑ Tarṣi 7). Although the document lacks details, we can approximate a date of destruction of the tower based on the history of Almería. Aflah was killed by Jayrán in 1014 AD, as related by al-ᶜUḏrῑ himself. Considering that Aflah governed the city of Almería between 1010 and 1014 AD, this tower had to have been destroyed at the beginning of the eleventh century AD, being one of the

defensive structures of the seafront of Almería before the reform of the walled enclosure of Jayrán. In Figure 7 we see how the tower of al-Idrῑsῑ was visible from the Alcazaba and from the Shipyards Tower, but not from the Chanca. From the sea, it is visible in optimal conditions up to 13.9 km perpendicular to the tower and 26.4 km parallel to the coast. The minimum visual distances covered (in a straight line perpendicular to the 79

Marta Del Mastro Ochoa Table 1: Minimum visual distances covered (in a straight line perpendicular to the element Element

Viewer height (m) Distance (km)

al-Idrῑsῑs Tower

5.7

14

Ribat of Alcazaba

6.7

34.9

Shipyards Tower

5.7

34.9

Defensive structures examined through archaeological analysis The walls and the shipyards are other significant elements of the maritime cultural landscape of Almería. This is largely based on the oldest iconographic representation known of the city: a miniature of its view from the sea (Figure 8), known as the graphite of the Genoese Admiral of the twelfth century AD, Caffaro di Rustico da Caschifellone, collected in the Annali Genovesi. Currently. One of the most interesting interpretations about this graffiti can be found in the paper of Garzón Osuna (2009: 146–65). The presence of the walls in the iconographic representation is quite telling, as the city had to retain a fortified frontal view from the sea, which would have had to be overcome by the attacking coalition in order to take the city.

element) from the ‘Alcazaba’, the Shipyards Tower and al-Idrῑsῑs Tower are visible in Table 1. The existence of so many defensive structures on the coast of Almería since the establishment in Pechina indicates a period of instability and insecurity with respect to the coasts, but an interest in this area as a maritime point despite its natural weaknesses. We have seen that the ribats and towers practically cover all the visible area of the Gulf of Almería, because they are located so that they are not only visible at any point in the gulf, but also in order to cover all the blind points of the coast. However, it must be recalled that visibility worsens as distance increases from the point of the observer. This, in addition to natural conditions such as temperature or bad weather (storms, rains or winds), can negatively influence the visibility of these points.

Some of the sections of these walls have been identified by urban archaeology, but they are also identifiable in modern cartography (Figures 1 and 8). In this sense, we believe that it is especially revealing to point out the existence of a double door in the southwest section of the wall. The rampart wall and the small brick door were located during the commercial archaeological excavation of Inés Relaño Street, located in the south-western area of Almería, parallel to the sea and 600 metres from the current western dock. Both date to the eleventh century

Figure 8. Plan of the city of Almeria (after Oviedo 1621).

80

The Study of the Port System of the Coast of Almería AD and, therefore, to the first phase of fortification of the urban settlement. This wall closed the city off from the sea, thus evidencing the existence of a period of certain maritime insecurity. On the other hand, the existence of a door (and therefore of an opening in the wall that weakens it) in this section indicates a wish to connect the city with the maritime space. Thus, an access to the best sheltered area from the southwest winds is opened, where anchoring activities could have been carried out. These activities have persisted in this area until today.

the archaeological remains, both terrestrial and underwater, documented along the Gulf of Almería during the last decades of the twentieth century AD. The large number of defensive structures has been equally significant, essentially since the beginning of the settlement in Pechina in the middle of the ninth century AD. To shed light on these structures, the defensive walls of the city (the ribats, the fundῡqs and the watchtowers) have been documented through archaeological, cartographic and written sources. Their presence is curious because it indicates that, despite the favourable maritime conditions that the territory of Almería benefitted from, its accessibility also presented a weak defensive point, as evidenced by the Fatimid and Normand attacks of the ninth century. This defence system surrounding the city seems to be trying to solve a situation of vulnerability from the sea which supplied it, but is weakened by two open accesses to the beach from the commercial district, where maritime activities would have taken place. However, this did not prevent the city from growing to be one of the most significant ports of al-Ándalus for three centuries.

Moreover, these factors are closely related to the commercial activity associated with this port until the middle of the twelfth century. Regarding this subject, there are different points to be highlighted. Firstly, we should explain the concept of alhóndiga. These buildings are especially interesting to analyse the importance of a port city. The historian Blanca Garí describes them as the place where outsiders settled with their merchandise (Abulafia and Garí 1996: 98). It was also where wholesale goods were stored, which were then sold in auction that were specialised for each commodity. These buildings were quite important because it was here that exports and imports were traded and distributed next to the souks or in the suburbs. According to Garí, who translates them from the word fundῡq, it could have been the etymological origin of the old Spanish word fonda.

The importance of the port of Almería for the growth of the city is evidenced by the fact that, originally, a port area was established, around which a city grew, and not vice versa. For this reason, it is logical to think that the place where the city of Almería is located was chosen for its maritime characteristics, prioritising the port. This is despite the fact that the same natural conditions that favoured navigation and anchorage also made it accessible to be attacked, which is why it was necessary to build a structured defensive system. Of course, with the data we have, we cannot make a theory about the factors that made these events possible, since they correspond to different periods and problems. However, we can point out that these structures were intended to solve a situation of insecurity regarding the sea.

Regarding the city of Almería specifically, al-Idrῑsῑ reports the existence of a large suburb called Rabad alHawd (Arrabal de la Hondonada), which contained souks, houses, baths and fundῡq, totalling 970 structures and features (al-Idrῑsῑ ‘Nuzhat) based on the taᶜtῑb tax registration (tax intended to repair the walls according to Molina) (Viguera 1997: 252). If we consider that alIdrῑsῑ makes a regressive description of Almería before the attack of 1147 AD, it is evident that, before that date, the city experienced intense commercial maritime traffic, based on the existence of an urban area dedicated to the lodging of retailers and movement of merchandise. The fact that it was uninhabited after 1147 AD (Lirola 2005: 192) suggests that the city lost much of the maritime trade potential that had characterised it over the previous two centuries. This was the main motivation for the attack on the city: the neutralisation of its port and, therefore, its maritime pre-eminence.

In conclusion, through the analysis of the maritime cultural landscape of Almería, we can contextualise the city and the maritime culture that created it. It is the port that explains the development of the city, and it is its location and nautical conditions that explain the interest of creating a port there. This is especially important because it indicates a maritime knowledge of the area (probably oriented more towards North Africa, as suggested by al-Bakri, eleventh century AD, al-Idrῑsῑ, twelfth century AD, and al-Rusati, fourteenth century AD), but also of an experience of the populations that settled there. They comprise a maritime culture that, during at least three centuries, focused the foundation and development of its city on its maritime functions.

Conclusions As has been shown throughout the text, the maritime characteristics of Almería seem to explain the situation of a port and a city at this point. The interrelation of the data provided by the analysis of the different types of sources has allowed us to present some information about its maritime cultural landscape. Furthermore, the zone of Almería is identified by the sources of antiquity with references to its good natural conditions to carry out anchoring activities. We have been able to analyse them through the place names of the space indicated by ancient authors such as Ptolemy and Pomponius Mela, but also through direct analysis of

Acknowledgements This study has been the result of so many efforts that it is difficult to expose everyone in their due place. I will try anyway. First of all, I must thank all the professors and coordinators of the first promotion of the Master in Nautical and Underwater Archaeology of the University of Cádiz. It goes without saying

81

Marta Del Mastro Ochoa Yᾱqῡt al-Ḥamawῑ. Muᶜyam al-buldᾱn. Trans. Lirola, J. 2005. Instituto de Estudios Almerienses: Almería.

that without your effort and your guidance I would not have acquired the necessary training to carry out the writing of this study. Thanks also to the staff of the Underwater Archaeology Center of Cádiz, especially Lourdes and Carlos, for their dedication and interest in this work, which resulted in the cessation of data that constitute one of the fundamental axes of the investigation.

Secondary sources ABULAFIA, D., GARÍ, B. 1996. En las costas del Mediterráneo occidental. Las ciudades de la Península Ibérica y del reino de Mallorca y el comercio mediterráneo en la Edad Media. Ediciones Omega: Barcelona.

Jorge Lirola, for guiding me and highlighting my mistakes so I could fix them properly, and also for his incredible work on Andalusian sources, without which it would have been impossible to do this work. Thanks, in retrospect for all these years dedicated to the Andalusian maritime culture.

ALARCÓN, M., GARCÍA DE LINARES, R. 1940. Los Documentos árabes diplomáticos del Archivo de la Corona de Aragón. Editados y traducidos por Maximiliano A. Alarcón y Santón y Ramón García de Linares. Editorial Maestre: Madrid.

To the R&D project ‘Pimalborán’ of the University of Granada, especially Professor Bilal Sarr, for the interest he has always shown both in the results of this research, and in its continuity. To Carles Aguilar, for having introduced me to the interesting world of nautical and underwater archaeology, and Xavier Nieto, who despite his lack of time, has always had a moment for coffee and words of encouragement.

ALCALÁ, F., MONTILLA, I., NAVARRO, M., BARBA, V. 2005. ‘Intervención arqueológica preventiva en los números 5 y 7 de la calle Arco de Almería’. Anuario Arqueológico de Andalucía 2005. Junta de Andalucía: Sevilla. Pp. 94–102.

Works Cited Primary sources

ARTEAGA, O., HOFFMANN, G. 1987. ‘Investigaciones geológicas y arqueológicas sobre los cambios de la línea costera en el litoral de la Andalucía Mediterránea’. In Anuario arqueológico de Andalucía 1986. Junta de Andalucía: Sevilla. Pp. 194–95.

Abῡ l-ᶜAbbᾱs al-ᶜUḏrῑ. Tarṣi al-ajbᾱr wa-tanwi al-ᾱṯᾱr wa-l-bustᾱn fῑ garᾱ’ib al-buldᾱn wa-l-masᾱlik ilà yami al-mamᾱlik. Trans. Lirola, J. 2005. Instituto de Estudios Almerienses: Almería. Abῡ ᶜUbayd al-Bakrῑ. Kitᾱb al-Masᾱlik wa-l-mamᾱlik. Trans. Lirola, J. 2005. Instituto de Estudios Almerienses: Almería.

AZUAR, R. 2016. ‘La taifa de Almería en el comercio mediterráneo del siglo XI’. In L. Cara Barrionuevo (ed.), Cuando Almería era Almariyya. Mil años en la historia de un reino. Instituto de Estudios Almerienses: Almería. Pp. 293–312.

Abῡ Muḥammad al-Ruṧᾱtῑ. Kitab Iqtibᾱs al-anwᾱr. Trans. Lirola, J. 2005. Instituto de Estudios Almerienses: Almería.

AZUAR, R. 2004. ‘Rutas y comercio de al-Andalus con las Repúlicas italianas de Génova, Pia y Amalfi, durante los siglos XI al XIII’. In L. De Maria and R. Turchetti (eds.), Rotte e porti del Mediterraneo dopo la caduta dell’impero romano d’Occidente. Continuità e innovazione tecnologiche e funzionali. Rubbettino Editore: Genova. Pp. 77–106.

Abῡ ᶜAbd Allᾱh al-Idrῑsῑ. Nuzhat al-muṧtᾱq fῑ jtirᾱq alᾱfᾱq. Trans. Lirola, J. 2005. Instituto de Estudios Almerienses: Almería. Abῡ ᶜAbd Allᾱh al-Zuhrῑ. Kitᾱb al-Ῡuᶜrᾱfiya. Trans. Lirola, J. 2005. Instituto de Estudios Almerienses: Almería. Abῡ l-Walῑd al-Ṧaqundῑ. Risᾱla fῑ Tafḍῑl al-Andalus,. Trans. Lirola, J. 2005. Instituto de Estudios Almerienses: Almería.

AZUAR, R. (ed.) 1989. La rábita califal de las dunas de Guardamar (Alicante). Cerámica. Epigrafía, Fauna. Malacofauna. Diputación Provincial de Alicante: Alicante.

Abῡ ᶜAbd Allᾱh al-Ḥimyarῑ. Kitᾱb al-Rawḍ al-miᶜṭᾱr fῑ jabar al-aqṭᾱr. Trans. Lirola, J. 2005. Instituto de Estudios Almerienses: Almería.

BARRERA, J. I. 2008. ‘Grafitos históricos en la casa morisca de calle San Martín, 16 (Granada)’. Arqueología y Territorio Medieval. Universidad de Jaén 15: 91–126.

Ibn Faḍl Allᾱh al-ᶜUmarῑ. Masᾱlik al-abṣᾱr fῑ mamᾱlik al-amṣᾱr. Trans. Lirola, J. 2005. Instituto de Estudios Almerienses: Almería.

BLACKMAN, D. J. 1982a. ‘Ancient harbours in the Mediterranean. Part 1’. International Journal of Nautical Archaeology and Underwater Exploration 11.2: 79–104.

Ptolemy. Claudii Ptolemaei Geographia. Edited by Nobbe, K. F. A. 1843. Lipsiae: Sumptibus et typis Caroli Tauchnitii.

BLACKMAN, D. J. 1982b. ‘Ancient harbours in the Mediterranean. Part 2’. International Journal of Nautical Archaeology and Underwater Exploration 11.3: 185–211.

Ibn Ḥawqal. Kitᾱb Ṣῡrat al-arḍ. Trans. Lirola, J. 2005. Instituto de Estudios Almerienses: Almería. Pomponio Mela. De Chorographia. Liber Secundus. Trans. Ranstand, G. 1971. The Latin Library.

BLACKMAN, D. J. 1973. ‘Evidence of sea level change in ancient harbours and coastal installations’. In D. J.

82

The Study of the Port System of the Coast of Almería CARA BARRIONUEVO, L. 1994. ‘Estructura urbana’. In L. Cara Barrionuevo (ed.), Roquetas de Mar: Arqueología e Historia. Desde la Prehistoria hasta inicio de la Edad Moderna. Instituto de estudios almerienses: Almería. Pp. 122–30.

Blackman (ed.), Marine Archaeology. Colston Papers: 115–22. BLÁNQUEZ, J., ROLDÁN, L., MARTÍNEZ, S., SÁEZ, F., and BERNAL, D. 1998. La carta arqueológica subacuática de la costa de Almería (1983–1992). Consejería de Educación y Cultura de la Junta de Andalucía y Universidad Autónoma de Madrid: Madrid.

CARA BARRIONUEVO, L., and CARA RODRÍGUEZ, J. 1989. ‘Puertos romanos en la costa meridional’. Revista Arqueológica 93: 8–19. CARA BARRIONUEVO, L., CARA RODRÍGUEZ, J. and RODRÍGUEZ LÓPEZ, J. M. 1988. ‘Las cuevas de la reserva (Roquetas) y otras factorías pesqueras de época romana en la provincia de Almería’. Actas del Congreso internacional del estrecho de Gibraltar. Universidad Nacional de Educación a Distancia: Ceuta. Pp. 919–35.

BOETTO, G. 2010. ‘Le port vu de la mer: l’apport de l’archéologie navale à l’étude des ports antiques’. In Bollettino di Archeologia Online. Ministero per i beni e le attività cultural Volume speciale B/B7/9. CABRILLANA, N. 1978. Documentos notariales referentes a los moriscos (1569–1571): Archivo Historico Provincial de Almería. Universidad de Granada Secretariado de Publicaciones: Granada.

CARRILLERO, M., GARRIDO, O., EGA, J. J., DÍAZCANTÓN, A., PADIAL, B., LÓPEZ, J. J., and GRACIA, M., 1987. ‘Memoria de la prospección arqueológica superficial del bajo Andarax (fase 2) y piedemonte de Sierra Alhamilla (Almería)’. Anuario Arqueológico de Andalucía 1986. Actividades Sistemáticas II. Junta de Andalucía: Sevilla. Pp. 66–68.

CALVO, M., JAVALOYAS, D., ALBERO, D., GARCIAROSSELLÓ, J., and GUERRERO, V. 2011. ‘The ways people move: mobility and seascapes in the Balearic Islands during the late Bronze Age (c. 1400–850/800 BC)’. World Archaeology. Taylor & Francis, Ltd. 43.3: 345–63.

CASADO BAENA, M. 2007. ‘Localización de la Antigua Ciudad de Urci y delimitación de la Frontera Interprovincial entre las Provincias Bética y Tarraconense en Tiempos de Tolomeo’. Gerión. Revista de Historia Antigua. Publicaciones Universidad Complutense de Madrid 25.1: 391–400.

CAPEL MOLINA, J. J. 1999. ‘La presión atmosférica y los vientos en la Península Ibérica: Reflexiones sobre el monzón ibérico’. NIMBUS: Revista de climatología, meteorología y paisaje. Universidad de Almería Servicio de Publicaciones 4: 5–60. CARA BARRIONUEVO, L. 1990. La Almería Islámica y su alcazaba. Editorial Cajal: Almería.

CASSON, L. 1989. The Periplus Maris Erythraei: Text with Introduction, Translation, and Commentary by Lionel Casson. Princeton University Press: Princeton.

CARA BARRIONUEVO, L., GARCÍA LÓPEZ, J. L., and MORALES SÁNCHES, R. C. 2000. ‘Arqueología urbana e historia de la ciudad. El caso de Almería medieval’. In L. Cara Barrionuevo (ed.), Ciudad y territorio en Al-Andalus. Athos-Pérgamos: Granada. Pp. 167–92.

CATALÁN, D., and DE ANDRÉS, Mª S. (eds.), 1974. Crónica del moro Rasis. Versión del Ajbar muluk al-Andalus de Ahmad ibn Muhammad ibn Musà alRazi, 889–955: romanzada para el rey Don Dionís de Portugal hacia 1300 por Mohamad, alarife y Gil Pérez, clérigo de Don Perianes Porçel. Editorial Gredos: Madrid.

CARA BARRIONUEVO, L. 1989a. ‘Informe sobre las actuaciones llevadas a cabo en la Alcazaba de Almería’. Anuario de Arqueología de Andalucía. Actividades de Urgencia III. Junta de Andalucía: Sevilla. Pp. 7–12.

CISNEROS CUNCHILLOS, M. 1988. Mármoles hispanos: su empleo en la España romana. Portico Librerias: Zaragoza.

CARA BARRIONUEVO, L. 1989b. ‘La alcazaba de Almería. Primeras intervenciones (MayoDiciembre 87)’. Anuario de Arqueología de Andalucía. Actividades de Urgencia III. Junta de Andalucía: Sevilla. Pp. 13–21.

CONTI, S. 1985. ‘Portolano e carta nautica: confronto toponomastico’. In C. Clivio Marzoli, G. Corna Pellegrini, and G. Ferro (eds.), Imago et mensura mundi. Atti del IX Congresso Internazionale di Storia della Cartografia. Istituto della Enciclopedia Italiana II: Roma. Pp. 55–60.

CARA BARRIONUEVO, L. 1994. ‘El puerto de los Bajos’. In L. Cara Barrionuevo (ed.), Roquetas de Mar: Arqueología e Historia. Desde la Prehistoria hasta inicio de la Edad Moderna. Instituto de estudios almerienses: Almería. Pp. 138–57.

CONTRERAS, A. 1996. Derrotero universal del Mediterráneo: manuscrito del siglo XVII. Estudio preliminar de Ignacio Fernández Vial. Editorial Algazara: Málaga.

CARA BARRIONUEVO, L. 1994. ‘La Ribera de la Algaida’. In L. Cara Barrionuevo (ed.), Roquetas de Mar: Arqueología e Historia. Desde la Prehistoria hasta inicio de la Edad Moderna. Instituto de estudios almerienses: Almería. Pp. 44–49.

CRAME, P. 1740. Plano de la Plaza y Castillo de Almería en que se demuestran las tres Baterías y camino cubierto que se propone para la Defenza y cubrir el

83

Marta Del Mastro Ochoa GUICHARD, P. 1995. ‘Actividad marítima y poblamiento’. In J. Vernet (ed.), Al-Andalus y el Mediterráneo. Lunwerg Editores: Madrid. Pp. 37–46.

Frente de su Marina Málaga, 30 de Mayo de 1740. [MPD, 36, 045]. Archivo General de Simancas. DEBANNE, A. 2011. Lo Compasso de navegare. Edizione del codice Hamilton 396 con commento linguistico e glossario. European Interuniversity Press: Belgium.

GUICHARD, P. 1976. Al-Andalus. Estructura antropológica de una sociedad islámica en Occidente. Universidad De Granada: Granada.

ELSNER, H. 1994. ‘Wikinger Museum Haithabu: Schaufenster einer frühen Stadt’. Archäologisches Landesmuseum der Christian – Albrechts – Universität: Neumünster.

GURRIARÁN, P. and MÁRQUEZ, S. 2009. ‘Aparejos constructivos de la Alcazaba de Almería. Lectura, análisis, interpretación’. Monografías del Conjunto Monumental de la Alcazaba nº 2. Junta de Andalucía: Almería. Pp. 233–66.

FANSA, G., and PIQUERAS, G. 2008. ‘La Península Ibérica en el gran Atlas de al-Idrisi’. Treballs de la Societat Catalana de Geografia 65: 465–77.

HOCKER, F., and WARD, C. A. 2004. The Philosophy of Shipbuilding. Conceptual approaches to the study of wooden ships. Texas A&M University Press: Texas.

FORD, B. 2011 (ed.). The Archaeology of Maritime Landscapes. Springer: New York.

HOFMANN, C., RICHARD, H., and VAGNON, E. (eds.), 2012. L’Âge d’or des Cartes Marines. Quand l’Europe découvrait le monde. Éditions Seuil: París.

FOSTER, R. 1970. ‘Agoranomos and Muḥtasib’. In Journal of the Economic and Social History of the Orient. Brill 13: 128–144.

IZQUIERDO I TUGAS, P. 1996. ‘Los condicionantes de la navegación en la antigüedad: Una aproximación al caso de la provincia Hispania Citerior Mediterránea’. I Simposio de Historia de las técnicas. La construcción naval y la navegación, Santander, (1995). Distribuciones-editorial Breogan S: Santander. Pp. 299–306.

GARCIA LORCA, G. 1990. El puerto de Almería. Junta del puerto de Almería: Almería. GARZÓN OSUNA, D. 2005. ‘Almería 1146. Evolución de la estructura urbana y defensas medievales’. Revista ph88. Proyectos, actuaciones y experiencias. Instituto andaluz de Patrimonio Histórico 88: 146–65.

JÉZÉGOU, M-P., JONCHERAY, A., and JONCHERAY, J-P. 1997. ‘Les épaves sarrasines d’Agay et de Cannes’. Archaeologia 337: 32–39.

GELICHI, S. 2012. ‘Las estructuras portuarias de Comacchio y los puertos altomedievales en Italia’. In A. Fábregas (ed.), Navegación y puertos en época medieval y moderna. Grupo de investigación Toponimia, Historia y Arqueología del Reino de Granada: Granada. Pp. 15–33.

LAFUENTE, J. 1949. ‘Traducción del poema de Avieno “Ora marítima” y localización de sus citas geográficas’. Estudios geográficos. Consejo Superior de Investigaciones Científicas: Madrid. Pp. 5–32, 209– 50.

GILMAN, A., MARTÍN, C., and FERNÁNDEZ-POSSE, M-D. 1993. ‘El sustrato neolítico en la cuenca de Vera (Almería)’. Trabajos de prehistoria 50.1: 57–85.

LÉVI-PROVENÇAL, E. 1950. España musulmana (711– 1031): La conquista, el Emirato, el Califato. Editorial Espasa-Calpe: Madrid.

GOITIEN, S. D. 1976. Letters of Medieval Jewish Traders. Princeton University Press: Princeton.

LÉVI-PROVENÇAL, E. 1953. ‘La “Description de l’Espagne” d’Ahmad Al-Rᾱzi’. In Al-Andalus, Vol. 18: 51–108.

GOLDBERG, J. L. 2012. Trade and Institutions in the Medieval Mediterranean: the Geniza Merchants and their Business World. Cambridge University Press: Cambridge.

LIROLA, J. 2005. Almería andalusí y su territorio: Textos Geográficos. Instituto de Estudios Almerienses: Almería.

GONZALBES CRAVIOTO, C. 2000. ‘La costa de Almería en los itinerarios de peregrinos y portulanos medievales’. Revista del Instituto Egipcio de Estudios Islámicos 32: 119–41.

LIROLA, J. 1991. El poder naval de al-Andalus en época del califato omeya (siglo IV de la hégira/X era cristiana. Departamento de Estudios Semíticos Facultad de Filosofía y Letras Universidad de Granada: Granada.

GONZALBES CRAVIOTO, C. 2000. ‘La frontera oriental nazarí en los portulanos medievales’. In P. Segura Artero (ed.), Actas del Congreso la Frontera Oriental Nazarí como Sujeto Histórico (S.XIII-XVI): Lorca-Vera, 22 a 24 de noviembre de 1994. Instituto de estudios almerienses: Almería. Pp. 451–66.

MAIER, J., and SALAS, J. 2000. Comisión de Antigüedades de la Real Academia de la Historia: catálogo e índices, Andalucía. Real Academia de la Historia (España): Madrid.

GONZÁLEZ PONCE, F. J. 1995. Avieno y el Periplo. Editorial Gráficas Sol: Sevilla.

MARTÍN, M. G. 2014. Cartagena Medieval. Cuadernos monográficos Museo del Teatro Romano: Cartagena.

84

The Study of the Port System of the Coast of Almería PICARD, C. 2012. ‘Las atarazanas en los puertos de alÁndalus, símbolos de poder musulmán’. In A. Fábregas (eds.), Navegación y puertos en época medieval y moderna. Grupo de investigación Toponimia, Historia y Arqueología del Reino de Granada: Granada. Pp. 34–50.

MEDAS, S. 2004. De Rebus Nauticis: L’arte della navigazione nel mondo antico (Studia Archaeologica). L’Erma di Bretschneider: Roma. MEDAS, S. 2008. ‘Lo Stadiasmo o Periplo del Mare Grande e la navigazione antica: commento nautico al piú antico testo portolanico attualmente noto’. In Gerión. Revista de Historia Antigua. Publicaciones Universidad Complutense de Madrid 26.12: 5–225.

PICARD, C. 2009. ‘Le port “construit” sur les littoraux du monde musulman méditerranéen et atlantique (VIIIe au XVe siècles), d’après les sources arabes’. In G. Fabre, D. Le Blévecn, and D. Menjot (eds.), Les ports et la navigation en Méditerranée au Moyen Âge: actes du colloque de Lattes, 12, 13, 14 novembre 2004, musée archéologique Henri Prades. Éditeur du Manuscrit: París. Pp. 213–24.

MEDAS, S. 2009. ‘Il piu antico testo portolanico attualmente noto: Stadiasmo o Periplo del Mare Grande’. Mayurqa: revista del Departament de Ciències Històriques i Teoria de les Arts. Universitat de les Illes Balears Servicio de Publicaciones 33: 333–364. MEDAS, S. 2009. ‘Andature e manovre con la vela quadra’. Monografíes del CASC: Arqueologia Nàutica Mediterrània. Generalitat de Catalunya: Barcelona. Pp. 419–26.

PICARD, C. 2007. ‘Architecture, institutions et sociétés portuaires des cités maritimes musulmanes de la Méditerranée médiévale, bilan et enjeux’. Histoire urbaine 19.2: 77–92.

MOLINA, E. 1997. ‘El retroceso territorial de al-Andalus’. In Historia de España T. VIII. Editorial Espasa-Calpe: Madrid.

PICARD, C. 2004. ‘Les arsenaux musulmans de la Méditerranée et de l’océan Atlantique (VIIe-XVe siècle)’. In D. Coulon (ed.), Chemins d’outre-mer: Études d’histoire sur la Méditerranée médiévale offertes à Michel Balard. Éditions de la Sorbonne: París. Pp. 691–710.

NIETO, X. 1998. ‘El transport naval com a condicionant del comerç marítim: l’exemple empordanès’. XI Col·loqui d’Arqueología de Puigcerdà. Comerç i vies de comunicació 1000 a.C – 700 d.C (1997). Institut d’Estudis Ceterans: Puigcerdà. Pp. 151–63.

POMEY, P., KAHANOV, Y., and RIETH, E. 2012. ‘Transition from Shell to Skeleton in Ancient Mediterranean Ship Construction: Analyses, problems and future research’. International Journal of Nautical Archaeology and Underwater Exploration 41.2: 235– 314.

NIETO, X. 1997. ‘Le commerce de cabotage et de redistribution’. In P. Pomey, La navigation dans l’Antiquité. Editeur Édisud: Aix-en-Provence. Pp. 146– 59.

PRYOR, J. H. 1995. ‘From Dromōn to Galea: Mediterranean Bireme Galleys A.D. 500–1300’. In R. Gardiner (ed.), The Age of the Galley: Mediterranean Oared Vessels since pre-Classical Times. Conway Maritime: Michigan. Pp. 101–16.

NIETO, X. 1988. ‘Cargamento principal y cargamento secundario’. Cahiers d’Histoire 3–4: 379–95. ORTIZ SOLER, D., MORALES SÁNCHEZ, R., and LÓPEZ BUSTOS, F. 1993. ‘Excavaciones de urgencia de apoyo a la restauración en la alcazaba de Almería. Primeros resultados’. Anuario de Arqueología de Andalucía. Actividades de Urgencia III. Junta de Andalucía: Sevilla. Pp. 9–19.

PUJOL I HAMELINK, M. 2012. La construcción naval a Catalunya a l’Edat Mitjana. Editorial Base: Barcelona. PUYOL, J. 1924. ‘Viaje por España y Portugal entre los años 1494 y 1495’. Boletín de la Real Academia de la Historia 84.1: 31–119.

OVIEDO, J. DE. 1621. Map of the City of Almería. [MPD, 29, 022]. Archivo General de Simancas.

LECHUGA DE SERANTES, J.E. (ed.) 2007. Kitab-I Bahriye. Libro para navegantes. Centro Nacional de Información Geográfica: Madrid.

PASCUAL Y ORBANEJA, G. 1699. Historia de Almería en su antigüedad, origen, y grandeza. Noticias recogidas de diversos fragmentos, que han estado retirados en el silencio. Facsimile Edition 1975. Ateneo de Almería: Almería.

REMIE CONSTABLE, O. 1997. Comercio y comerciantes en la España musulmana: la reordenación comercial de la Península Ibérica del 900 al 1500. Ediciones Omega: Barcelona.

PIRENNE, H. 1927. Les villes du Moyen Age, essai d’histoire éc on omique et sociale. Lamertin: Bruxelles.

RIVERA MEDINA, A. M. 2015. ‘Espacios urbano y portuario: Las dinámicas de gestión’. In E. Aznar and R. J. González (eds.), De mar a mar. Los puertos castellanos en la Baja Edad Media. Servicio de Publicaciones de la Universidad de La Laguna: La Laguna. Pp. 93–122.

PICARD, C. 2014. La mer des Califes. Éditions du seuil: París. PICARD, C. 2014. ‘Pechina-Almería aux IXe-Xe siècles La naissance d’un port omeyyade en Méditerranée’. In E. Malamut, and M. Ouerfelli, Villes méditerranéennes au Moyen Âge. Presses universitaires de Provence: Aixen-Provence. Pp. 135‑46. 85

Marta Del Mastro Ochoa ROCA, J. R., and JULIÁ, R. 1997. ‘Late-glacial and Holocene lacustrine evolution based on ostracode assemblages in Southeastern Spain’. Geobios 30.6: 823–30.

VERNET, J. 1958. ‘La carta magrebina’. BRAH 142.2: 495–533.

ROUGÉ, J. 1966. Recherches sur l’organisation du commerce maritime en Méditerranée sous l’empire Romain. S.E.V.P.E.N: París.

VERNET, J., and ORÚS, J.J. 1950. ‘La transformación de las coordenadas árabes’. La Gaceta Matemática: revista publicada por el Instituto “Jorge Juan” de Matemáticas y la Real Sociedad Matemática Española 3: 78–82.

SAMSÓ MOYA, J. 1995. ‘La ciencia andalusí y el Mediterráneo’. In J. Vernet (ed.), Al-Andalus y el Mediterráneo. Lunwerg Editores: Madrid. Pp. 21–22.

VIEDMA, M. 2002. ‘El ritmo estacional del viento en el Arco Mediterráneo Español e Islas Baleares’. Papeles de Geografía. 35: 171–92.

SCHULTEN, A. 1959. Geografía y etnología antiguas de la Península Ibérica. Tartessos: Madrid.

WESTERDAHL, C. 1992. ‘The maritime cultural landscape’. In International Journal of Nautical Archaeology and Underwater Exploration. Nautical Archaeology Society 21.1: 5–14.

SCHULTEN, A. 1955. Ora marítima (Periplo Massaliota del siglo VI a. de J.C.) junto con los demás testimonios anteriores al año 500 a. de J. C., Avieno. Librería Bosch: Barcelona.

WESTERDAHL, C. 2006. ‘Maritime Cosmology and Archaeology’. Deutsches Schiffahrtsarchiv 28: 7–54.

S. KHALILIEH, H. 2006. Admiralty and Maritime Laws in the Mediterranean Sea (ca. 800–1050). The Kitᾱb Akriyat al-Sufun vis-à-vis the Nomos Rhodion Nautikos. Brill: Leiden.

WILLENSEN, A. 2010. Dorestad in an International Framework. New research on Centres of Trade and Coinage in Carolingian Times. Brepols: Turnhout.

SOLORZÁNO TELECHEA, J.A. 2015. ‘La influencia del mar en la conformación de los puertos de la marisma y costera de España en la Baja Edad Media’. In E. Aznar and R. J. González (eds.), De mar a mar. Los puertos castellanos en la Baja Edad Media. Servicio de Publicaciones de la Universidad de La Laguna: La Laguna. Pp. 59–92. SUÁREZ MARQUEZ, A. 1987. ‘Memoria de la excavación de urgencia realizada en el solar de la calle Reina. Parque Nicolás Salmerón’. Anuario de Arqueología de Andalucía. Actividades de Urgencia III. Junta de Andalucía: Sevilla. Pp. 22–28. TAPIA GARRIDO, J.A. 1982. Historia General de Almería y su provincia: Colonizaciones, vol. II. Editorial Cajal: Almería. THROWER, N. J.W. 2002. Mapas y civilización. Historia de la cartografía en su contexto cultural y social. Ediciones del Serbal: Barcelona. TOFIÑO DE SAN MIGUEL, V. 1787. Derrotero de las costas de España en el Mediterraneo: y su correspondiente de Africa para inteligencia y uso de las cartas esféricas presentadas al rey nuestro señor. Editorial la Vda. de Ibarra e Hijos: Madrid. TORRES BALBÁS, L. 1946. ‘Atarazanas hispanomusulmanas’. Al-Andalus 11: 175–209. VALENCIA, R. 2010. ‘La cartografía y Andalucía en la Edad Media’. In F. Olmedo, and J. Cortés (eds.), Andalucía: la imagen cartográfica hasta fines del siglo XIX. Instituto de Cartografía de Andalucía (ICA): Sevilla. Pp. 92–105. VERNET, J. 1999. Lo que Europa debe al Islam de España. Editorial el Acantilado: Barcelona.

86

8 The Whole Story: Exploring the Transportation of Whole Olives in Antiquity Through Shipwreck Evidence Lisa Briggs Oxford Centre for Maritime Archaeology, Research Laboratory for Archaeology and the History of Art, University of Oxford [email protected] Abstract: The production, trade, and consumption of olive oil in the ancient world has been widely explored. Often overlooked, however, was the role of the whole olive in the Mediterranean diet and long-distance maritime trade. Remains of whole olives are frequently recovered from ancient shipwrecks that range in date from the Late Bronze Age to the Roman period. This paper examines the whole olive in the ancient Mediterranean as a trade item, provision, and dietary element, and further explores how the production of whole olives is not reliant on the complex infrastructure necessary to produce olive oil. Keywords: Maritime trade, olive, shipwreck, archaeobotanical remains, waterlogged

Introduction

What is interesting about the discovery of whole olives in shipwreck amphorae is that they are rarely expected: a tendency to assume vessel contents based on their shape and geographic origin can expedite the interpretation of a shipwreck cargo. This approach will be called the ‘normative approach’ to amphorae interpretation. While excavators on terrestrial sites must look to ancient texts, charred remains, phytoliths, amphorae stamps, and residue analysis in an effort to deduce what the nowempty amphorae found on land previously contained, underwater archaeologists occasionally are given the luxury of studying the actual contents of amphorae due to the superior preservation of organic remains on some maritime sites.

Olea europaea L., the olive tree, is a species emblematic of the Mediterranean littoral, and so inexorably linked with this region that some have argued for the renaming of this species as Olea mediterranea (Mattingly 1988). Underwater archaeology can provide unique insights into olive cultivation and exploitation in the ancient Mediterranean. Excavations off the Levantine coast at the submerged Late Pre-Pottery Neolithic village of AthliYam have yielded olive pollen and carbonised olive wood dating from 6100–5500 B.C.E., while abundant remains of olive endocarps (commonly called ‘stones’ or ‘pits’) and evidence for olive oil processing come from the nearby underwater Late-Neolithic to EarlyChalcolithic site of Kfar Samir (Galili et. al. 1993; 1997). Shipwreck excavations have provided the opportunity for significant archaeological discoveries of olives; the largest single deposit of Late Bronze Age olive remains comes from a Canaanite amphora recovered from the Uluburun shipwreck, in which over 2,500 olive stones were found (Haldane 1993), while the cargo of an Etruscan shipwreck from the 6th century B.C.E. found near Giglio, Italy included an amphora full of olives (Bound and Vallintine 1983, Lanza 2011). Similarly, the Mazotos shipwreck discovered off the south coast of Cyprus carried a cargo of amphorae from the Greek island of Chios, which was a famed producer of wine in the ancient world: a reasonable assumption followed that these amphorae contained the wine for which Chios was renowned, yet several amphorae recovered from this wreck contained not wine, but thousands of whole olives (Demesticha 2011).

Unexpected foodstuffs and objects have been found in amphorae recovered from many shipwreck sites dating from a vast range of time periods. Most of the Canaanite jars found on the Late Bronze Age Uluburun shipwreck contained resin (Pulak 1998, Stern 2008), but some contained thousands of olives (Haldane 1993). On the Classical-period Tektaš Burnu wreck a pseudo-Samian amphora that was assumed to have contained wine was found to contain butchered beef bones, offering interesting insight into not only amphorae use in the Classical and Hellenistic world, but also into the consumption and preservation of meat products (Carlson 2003). Several Samian amphorae discovered on the Kyrenia wreck were found to contain whole almonds, and when counted at the end of the excavation the total number of almonds amounted to over 10,000 (Katzev 2005, 76). The preservation of such ‘macro-remains’ allows for simple, optical identification of vessel contents. In this case, without visually inspecting

87

Lisa Briggs the vessel contents, it is unlikely that any archaeologist would have guessed from their outward appearance that the amphorae from Samos recovered from the Kyrenia shipwreck contained whole almonds.

transportation and trade of whole olives left behind as a distant second, if discussed at all. Here, two shipwrecks sites whose excavators have thoroughly recorded the olives discovered on board will be explored: the Late Bronze Age Uluburun shipwreck discovered off the coast of Turkey (circa 1300 BCE), and the 4th century BCE Mazotos shipwreck discovered off the southern coast of Cyprus. Both ships share one interesting feature: the olive stones found on these ships were contained in ceramic transport containers assumed to have carried a different cargo.

Even associating amphorae exclusively with the transportation of foodstuffs and agricultural products would not reflect their diverse uses: a shipwreck discovered off the Levantine coast carried an amphora filled with scrap metal from a broken figurative statue (Galili et al, forthcoming), while one Canaanite jar on the Uluburun wreck contained thousands of glass beads (Pulak 1998). Olives, almonds, butchered beef, scraps of metal statues and glass beads; the evidence grows with each new shipwreck excavation that amphorae were multi-use containers, not limited to only wine and olive oil, the contents with which they are so often associated.

Olives on the Uluburun shipwreck The Uluburun shipwreck, its origin, trajectory, crew, and cargo have been much discussed (i.e. Bass 1987, Pulak 1988; 1998; 2000; 2005; Monroe 2010; Payton 2013) and need not be reviewed in detail here. The shipwreck, which is thought to date from around 1300 BCE (Pulak 1998), contained a wide variety of organic remains (Haldane 1993). Questions were raised as to whether mixing and contamination may occur by the introduction of artefacts from later periods, but an experiment conducted on the Uluburun shipwreck site itself saw 500 olives place in an open-mouthed ceramic vessel upslope of the excavation area; 90 days later only three olive stones had disappeared, suggesting that loss of olive stones or seaborne contamination is most likely not a primary factor in the composition of archaeobotanical samples found on shipwrecks (Haldane 1993, 349).

espite the preponderance of whole olive remains found on ancient shipwrecks, the trade in whole olives in antiquity has not been sufficiently explored. In this paper, I will argue that evidence from ancient shipwreck sites can offer new and relevant insight into the transportation and use of whole olives in antiquity. In addition, it will be shown that the scholarship surrounding the cultivation and consumption of olives in the ancient world is dominated by discussions surrounding the trade in olive oil, admittedly an important industry, but not the exclusive purpose of Olea europaea cultivation. By ignoring the consumption and trade of whole olives, we may be misinterpreting important facets of past agricultural developments. Finally, what purpose the thousands of whole olives found on ancient shipwreck serves will be explored; were they intended as gifts, as trade items, or provisions for the crews aboard the ships on which they are found?

While the cargo of the Uluburun was highly diverse, consisting of copper ingots, glass ingots, pithoi, and sundry items, there was a substantial cargo of over 149 Canaanite jars on board (Pulak 1998, 201). The majority of these jars carried an amber coloured resin interpreted as Pistacia resin (Mills and White 1989, Stern et al 2008), but at least two of the Canaanite jars contained whole olives (Haldane 1993). Numbering over 2,500, this is the largest single deposit of olive stones from the Late Bronze Age (Haldane 1993), and as such constitutes clear evidence as to how shipwrecks can provide important data on the use of whole olives in the Late Bronze Age, which terrestrial sites lack.

Shipwreck evidence for the transportation of whole olives The discovery of whole olives in shipwreck amphorae is often mentioned in passing in publications on ancient shipwreck discoveries, but not properly explored. Ballard and his team make a brief mention of a Classical era wreck containing Sinopian amphorae in the Black Sea, one of which was found to contain thousands of olives (Markey 2003). No images of the olives or the amphorae in which they were found appear to be have been made available to date. Similarly, the discovery of a fifth-century BCE Chian amphorae off the coast of Knidos which was found to contain olive stones has not been explored further, with only a brief mention made of this discovery in the Institute for Nautical Archaeology newsletter (Carlson 2003, 590; citing Frey 1982, 5). Reports from the Etruscan shipwreck found near Giglio, Italy that mention an amphora full of whole olives (Bound and Vallintine 1983, Lanza 2011) also fail to go into sufficient detail as to the number, size, and condition of the olive remains recovered. As will be discussed below, this seems to follow a pattern of scholarship focussing almost exclusively on the trade in olive oil, with the

On the outset, there is little about the appearance of the Canaanite jars recovered from the Uluburun shipwreck that contained olives that would immediately distinguish them from the majority of jars which contained Pistacia resin. However, it is interesting to note that one of the jars, KW 1553 has a small hole towards the bottom of the jar. This appears to be a firing error: such holes often occur when a small limestone inclusion is saturated with water then expands in the kiln causing the limestone inclusion to explode, leaving a hole in the vessel wall. While the exact contents of this jar are unclear, several olive stones were recovered from inside this vessel. Vessel KW 1565 on the other hand, contained several over 2,000 olives. The ambiguity as to the exact contents of KW 1553 88

The Whole Story been detected without a systematic investigation into the contents of each and every jar. This highlights issues with the normative approach to amphora interpretation, which would have assumed every jar on the Uluburun contained resin. While the normative approach is admittedly less time-consuming and more cost-effective (as it does not require the retrieval of every ceramic transport container prior to making declarative statements regarding their contents), the drawback of the normative approach is that it does not reflect the diverse and nuanced reality of amphorae use on ancient ships. In this case, if the normative approach to amphora interpretation had been employed on the Uluburun shipwreck, the largest single deposit of Late Bronze Age olive stones would have remained undiscovered. Olives on the Mazotos shipwreck In 2006, a 4th century B.C.E. shipwreck was discovered off the southern coast of Cyprus near the village of Mazotos. During the subsequent investigations, it was noted that the majority of extant cargo consists of amphorae in the shape and style associated with Chios, therefore this location was named as the most likely origin for the cargo of the ship (Demesticha 2011). Ancient shipwrecks are often found in pieces on the seafloor with a spread of artefacts surrounding where the hull of the ship once lay. The Uluburun shipwreck discussed above settled onto a steeply sloping seafloor, causing the cargo items to slowly roll down the underwater cliff during the many millennia spent underwater (Pulak 1988). In the case of the Mazotos wreck, however, the ship settled onto a flat, sandy, and relatively featureless seafloor, leaving the compact nature of the finds to outline the shape of the ancient ship (Demesticha 2011).

Figure 1. Canaanite jar 1553, Uluburun shipwreck.

Olive remains have been recovered from both the proposed bow and stern portions of the wreck. As the mid-ship area of the wreck has not yet been explored, it remains unclear to what extent the cargo in this area did, or did not contain whole olives. A full report on the olive remains recovered from the Mazotos shipwreck is being prepared (Pagnoux, Margaritis and Demesticha, in preparation), while here a select sub-sample is discussed. Olive stones were recovered from at least three amphorae: two amphorae in the distinctive style of Chios, and one in a Southern Aegean style that features a flared, mushroomshaped rim. Thus far, more than 10,000 olive stones have been recovered from excavated amphorae, as well as numerous stones collected through sieving the excavated seafloor sediment within the shipwreck area. The sheer number of olive stones present on this wreck is intriguing for two reasons: first, it yet again causes us to challenge the normative approach to amphora interpretation which would have assumed all the Chian amphorae contained wine, and second, the volume of olive stones recovered suggests that this was not a meaningless bit of cargo, but rather an element of the ship’s conveyance that was deliberately collected and included for a designated

Figure 2. Canaanite jar KW 1553, with hole.

remains, as a considerable mixing of the organic material was noted throughout the excavated area, allowing for the possibility that the olive stones may have floating into this jar, which may have originally contained something else. On the other hand, is it possible this jar was deliberately selected to contain whole olives, as this cargo item would be less likely to be lost through this unintended hole in the ceramic vessel? As the majority of the Canaanite jars recovered from the Uluburun shipwreck contained a resinous material, the presence of olives in two of the jars would not have 89

Lisa Briggs

Figure 3. Approximate location of olive stone recovery areas, Mazotos shipwreck.

neck and extant handles on amphora P0465 have been entirely lost on amphora P0488. If one compares the respective aperture size, the damaged amphora P0488 has an artificially enlarged aperture due to the break or removal of its long narrow neck. This brings up the intriguing possibility that damaged or  defective amphorae could have been deliberately selected to hold olives. A liquid product, like wine or olive  oil, would not be suitable for the Canaanite jar KW  1553 as the hole near the toe would allow a liquid product to escape, while an amphora like P0488 pictured above would allow a liquid product to slosh out of the broken top. In addition, if the olives in amphora P0488 were conveyed on the ship as provisions for the crew, it is possible that a larger aperture could have allowed easier access to the olives throughout the journey.

Figure 4. Chian amphora P0488 containing olive stones, Mazotos shipwreck. Photo by the author.

purpose. Yet more intriguing is the state of amphora P0488, recovered from the proposed stern portion of the shipwreck.

We can also deduce pertinent information about the olives themselves. The Mazotos olive stones show signs of consumption by marine organisms. The soft kernel which acts as the embryonic olive tree and is housed in the protective casing known as the endocarp (what is commonly known as the olive stone), has been accessed by marine organisms and consumed.

This amphora, like the Canaanite jar KW 1553 which also contained olives, appears to be a damaged amphora. The above photograph shows the area of the amphora just below where the cylindrical neck meets the shoulder of the amphora. Here, we see a greatly enlarged aperture, with no amphora neck extant. Interestingly, the break appears to have happened in antiquity, as the broken edges appear quite smooth and worn. This does not have the sharp and distinctive appearance of recently broken ceramic fabric.

It appears that marine organisms have accessed the soft, lipid-rich kernel through the stem aperture. Unfortunately, this could potentially compromise attempts to study shipwreck olive stones through certain scientific methods, like isotope analysis and DNA analysis, if the kernel, which represents a significant portion of the olive stone, is no longer present.

For comparison, a complete Chian amphora from the Mazotos shipwreck is shown below; the narrow cylindrical 90

The Whole Story

Figure 5. Chian amphora P0465, Mazotos shipwreck.

Metric analysis of Uluburun and Mazotos olives

Uluburun olive remains appear to be larger in size than olive remains recovered from terrestrial sites which date from the same period (Haldane 1993).

The gross anatomical morphology of whole olives which have been waterlogged could offer better insight into their original shape and size than olive remains which have been charred. Charred olive remains found in terrestrial sites are found in either fragmented pieces or whole, resulting in ambiguity surrounding the nature of their use as either left-overs from oil processing or charred remains of whole olives which were eaten (Margaritis and Jones 2008, 382). Whole olive remains recovered from shipwreck amphorae, however, appear to be less ambiguous in nature as it seems clear these olives were intended to be eaten whole.

The charts on the following pages have been recreated based on data Haldane collected through metric analysis of the Uluburun olive stones, which was compared to the limited metric data available for olive remains found on contemporary terrestrial sites. As can be seen below, only the desiccated remains from Late Bronze Age Timna are larger than the Uluburun olive stones. The comparatively large size of the Uluburun olive stones caused Haldane to surmise that these olives were a luxury trade product, and had most likely been deliberately selected for their size (Haldane 1993: 353).

In addition, charring experiments have shown that in oxidising conditions between 250°C and 400°C olive stones shrink slightly in size, while, interestingly, when temperatures exceed 450°C the olive stones retain their size and shape (Margaritis and Jones 2008, 387). As charred olive stones constitute the vast majority of olive remains found in the archaeological record, it is possible that taphonomic processes such as charring have altered their size and shape sufficiently to obscure their original dimensions. If this is the case, and waterlogging is shown to have less effect on the size and shape of archaeological olive remains, it may be that waterlogged olive stones recovered from shipwrecks can offer more accurate information on the size and shape of ancient olive stones.

Preliminary results from the Mazotos olive stones offer yet another interesting parallel to the Uluburun olive stones: the comparatively large size of the Uluburun olive stones of 15.3mm x 8mm was interpreted by Haldane as an indication that these olives were a luxury product intended for trade, while the average size of 13 Mazotos olive stones is nearly identical, at 15.2mm x 8mm. In a similar fashion to Haldane’s comparison with contemporary terrestrial sites, results from the metric analysis of the olive stones from the Mazotos shipwreck are compared with olive remains from two Classical / Hellenisitic sites: both located on mainland Greece, charred olive remains from the excavation of an ancient tavern in Corinth and an ancient farmhouse at Tria Platania are measured here. To add additional context to this analysis, a group of modern olive stones from Chios are also measured and compared to the ancient samples.

Here, data obtained by Haldane during her study of the Uluburun shipwreck olive stones is compared with new data from the Mazotos shipwreck olive stones. Despite the fact that the sinking of these two shipwrecks is separated in time by almost 1,000 years, interesting parallels emerge in the metric analysis of their olive stone remains. During analysis of 100 randomly selected olive stones from the over 2,500 recovered specimens, Haldane found that the

Like the Uluburun olive stones, the olive stones recovered from the Mazotos shipwreck are larger than their terrestrial parallels. Even the modern olive stones from Chios are 91

Lisa Briggs

Figure 6. Mazotos olive stone 9.1 with a hole bored into the top of the endocarp.

Figure 7. Mazotos olive stone 8.3 with only an empty seed coat remaining of the kernel.

only marginally larger; an average of 20 yielded a mean length and width of 17mm x 8mm, with the Mazotos olive stones not far behind with the average of 13 stones yielding a mean length and width of 15.2mm x 7mm. Yet, given the propensity of olive stones to shrink during charring in an oxidising environment (Margaritis and Jones 2008) the possibility that the charred olive stones from Corinth and Tria Platania have shrunk cannot be excluded. It does appear noteworthy, however, that the modern olive stones from Chios, grown under modern agricultural conditions are still only marginally larger than the Mazotos olive remains. Were the Mazotos olives deliberately selected for their large size, as Haldane has postulated was the case for the Uluburun olive remains?

may also be able to add insight into the crop processing of olives for olive oil. If a process of deliberate exclusion of large olives is undertaken at olive oil processing sites (or prior to transporting olives to such a site) this could add an interesting new dimension to our understanding of the production of olive oil in antiquity. However, it should be noted here that the 13 Mazotos shipwreck olive stones analysed in this study comprise a fraction of the 10,000 recovered. A comprehensive study of the Mazotos shipwreck olive stones is underway which will analyse a far greater percentage of the recovered olive remains (Pagnoux, Margaritis, Demesticha in preparation).

Another interesting aspect of the 13 Mazotos shipwreck olive stones included in this study is their consistent size: is this further evidence that these olives were deliberately selected for their large size, while their smaller counterparts were consigned to the oil mill? If this is the case, shipwreck evidence for the transportation of whole olives

In the ancient world, olives were used to produce oil for cooking and cleaning (Foxhall 2007), as a snack or an ingredient of a more substantial dish (Margaritis 2014), or as a fuel source as the many charred examples of postoil processing olive stones attest (Margaritis and Jones 2008). Despite the many uses of olive products, it is the

Challenging the exclusive focus on olive oil

Chart 1. Size of Late Bronze Age olive stones, chart recreated from data in Haldane (1993), citing Helbaek 1962; Hjelmqvist 1979; Kroll 1984; Kislev 1988.

92

The Whole Story

Chart 2. Metric analysis of Mazotos olive stones compared to ancient charred remains from Corinth and Tria Platania, and modern samples from Chios.

Chart 3. Length and width of individual olive stones from the Mazotos shipwreck.

production and transportation of olive oil that dominates the scholarship surrounding olive cultivation, trade, and consumption.

as a fuel source (Margaritis and Jones 2007, Margaritis 2013, Margaritis 2014), and these charred remains form the greater part of olive remains found in the archaeological record. As such, either the architectural remains of olive oil processing sites, or charred olive material have been the focus of most scholarship surrounding olive cultivation and consumption. It is possible that this focus on the artefacts from oil processing sites has led to a perceived diminished importance of whole olives as a food source and trade good in antiquity. Shipwreck evidence for the trade in whole olives may help to balance the data for olive exploitation in the ancient world.

Oil processing sites equals olive cultivation? Perhaps not… While the ancient Mediterranean trade in olive oil is well documented, commentary on the movement of whole olives in the ancient world is frustratingly elusive. Detailed and comprehensive studies have been undertaken of olive oil pressing sites (e.g. Waliszewski 2014), as the remains of olive oil processing infrastructure is far more likely to survive in the archaeological record than the remains of whole olives. Charred remains of olive stones found at oil production sites indicate that, once pressed for oil, the remains of whole olives could serve a secondary function

It has been argued that cultivation of the olive and the grape (Vitis vinifera) was vital to the development of marginal landscapes and the emergence of complex societies in the Early Bronze Age Mediterranean (Renfrew 93

Lisa Briggs 1972, Blitzer 1993) while others countered this view with claims that insufficient evidence for the processing of olive products has been found to justify the importance of olives at this early period (Runnels and Hansen 1986, Hamilakis 1996). Yet, if whole olives for consumption as a food source were the intended products of Early Bronze Age cultivation of Olea europaea, one would not require oil processing equipment to meet this end. According to Foxhall (2007), prior to the classical period, table olives for food consumption may have been more accessible than olive oil to all socio-economic spectra due to the fact that they were cheaper and easier to produce. Significantly, the nutritional value of wholes olive exceeds that of olive oil in levels of carbohydrates, calcium, magnesium and iron (Zipori et al. 2015). If widespread consumption of tables olives occurred, evidence for oil processing equipment at Early Bronze Age sites becomes irrelevant, as the successful cultivation and consumption of table olives by a large societal demographic would not require this heavy equipment. In addition, experimental archaeological research conducted on ancient olive processing methods showed that a simple mortar and pestle were sufficient to produce a small amount of oil (Warnock 2007), which further undermines the argument that heavy processing equipment was essential to olive utilisation. Even if producing small amounts of oil was the desired outcome, it appears little more than a pestle and mortar may remain in the terrestrial archaeological record. Significantly, the largest deposit of olive remains from the Late Bronze Age comes not from charred remains found at an oil-processing sites, but from whole olives aboard the Uluburun shipwreck.

that tribute may have come in the form of both whole olives and olive oil. On a Late Bronze Age stele from Egypt, olives were included in a list of rations for the messenger of Pharaoh Seti I (circa 1323 – 1279 BCE) (Kebler 2009: 344; citing Breasted 1906: 208). The Harris papyrus which details the pious deeds of Ramesses III (circa 1186 – 1165 BCE) mentions gifts of whole olives in jars given to the temple of Amun-Ra, while jar dockets from Deir el-Medina dating from the reign of Ramesses II (circa 1279 – 1213 BCE) mention centres in Egypt where whole olives were produced (Kebler 2009: 344). Even in Egypt where olives were not well-suited to the climate, and during this early period, it appears whole olives were functioning as an important source of nutrition for those conducting the pharaohs business, as well as gifts for the temple. In the Odyssey, the poet Homer’s Iron Age masterpiece, the hero Ulysses describes being moved by witnessing Tantalus’ unquenched desire for ‘pears, pomegranates, apples, sweet figs and juicy olives’ (Butler 1994). Thus, in the Bronze Age, as well as the Iron Age, references to whole olives for eating, trading, gift-giving, and temple tribute are known. As the demand for olive products increased throughout the millennia that separate the sinking of the Uluburun shipwreck from the Mazotos, textual references relating to the trade and movement of these products become more frequent. Several ancient authors mention laws concerning the trade in olive oil: Plutarch records in his Life of Solon that the Athenian statesman (638–558 BCE) banned the export of all Attic goods with the notable exception of olive oil. This may indicate that during the Greek Archaic period in which Solon lived, oil was still a semi-luxury commodity (Foxhall 2007). This means that regardless of the amount produced (low yield or high) the people of Attica did not rely on their local production of olive oil as an essential food source. Interestingly, this legislation does not mention restrictions on the export of whole olives.

By the Hellenistic period, olive cultivation was sufficiently established throughout the Mediterranean to constitute a major industry. However, levels of cultivation and exploitation differed temporally and spatially in Greece, a major producer of olives (Margaritis 2013), and adjacent regions. In addition, the biennial cycle of large harvests compounded the issue of supply. Horden and Purcell discuss in their influential work, The Corrupting Sea, that varied micro-climates of the Mediterranean contributed to complex rhythms of supply and demand (Wilson and Bowman 2018: 2–4; citing Horden and Purcell 2000). The biennial nature of olive trees contributes to this fluctuation in the supply of whole olives. It could be argued that this specific feature of the Olea europaea fruiting cycle could encourage the movement of whole olives around the Mediterranean region. Thus, the need for seaborne trade of olive products throughout the Mediterranean to fulfil the ebb and flow of surplus and deficit is clear (Davies 2006).

Conflict and warfare could have a direct effect on olive production: during Peloponnesian invasions and Boeotian incursions considerable damage was done to the olive groves of Attica, which Aristophanes seems to suggest led to a shortage in olive oil in 422 B.C.E. (Hopper 1979). It appears that political entities throughout time took a keen interest in the supply of olive products available in their domain: in the 2nd century C.E. a law was passed by the Roman Emperor Hadrian (76 – 138 C.E.) which states that olive oil producers in Attica could keep two-thirds of the oil produced in a given year only in the event that the Roman state did not require it elsewhere (Sallares 1991). This suggests that Attic producers could expect a fluctuation in yields, and that in an ‘off’ year of the tree’s biennial cycle olive oil producers themselves may have been expected to purchase oil. In addition, these contradictory laws suggest that in the intervening 600 years between the Archaic Greek period and the height of the Roman Empire olive oil consumption had increased considerably throughout the Mediterranean.

Textual evidence for the trade in whole olives Literary, epigraphic, and papyrological data can supplement the evidence from shipwrecks for the trade in whole olives in the ancient world. Bronze Age Linear B tablets from Crete have different ideograms for olives, olive trees, and olive oil (Palaima 2012: 12), suggesting 94

The Whole Story Whole olives as a trade item

But what of the movement of whole olives? While legislative measures and international trade wars appear to have been more concerned with olive oil, there is ample evidence for the movement of whole olives in this period. Demosthenes refers to an ‘olive-merchant’ in Against Aristogiton I (Murray 1939), suggesting that the trade in whole olives was common enough for a merchant to make his living from this pursuit. A letter from the Zenon archives (circa 259 BCE) requests that provisions for a journey be delivered including whole olives, pickled fish, pounded lentils, and wine (P. Mich Nos 1, 2 in Edgar 1931), while whole olives also frequently appear in customs house receipts from Roman Egypt, where ‘green olives’ are listed among textiles, wine, camels and a variety of other trade goods as being monitored and taxed through the harbour of Memphis and other customs checkpoints in Egypt (e.g. P.Customs 34; P.Customs 466 in Sijpesteijn 1987). Another papyrus which dates from around 100 CE asks explicitly about purchasing whole olives – ‘Tell me at what price for the artaba [a unit of measurement] you wish me to buy succulent olives (SB 6 9025 in Youtie 1944).’

It has been argued that the 2,500 olive stones found in one Canaanite amphora on the Late Bronze Age Uluburun shipwreck are, based on their large size (15.3mm x 8mm on average), an additional luxury item in a cargo that included amber beads, gold jewellery, elephant tusk, copper and glass ingots and a metric tonne of Pistacia resin (Haldane 1993). The average size of the 13 Mazotos olive stones selected as a sub-sample for this study is nearly identical to that of the Uluburun remains (15.2mm x 7mm). Does this suggest that the Mazotos olives were intended for trade? As has been discussed above, the size of the Mazotos olive stones is remarkably consistent, strengthening the argument that these were deliberately selected for their size, and were intended to be a trade item. One wonders if provisions for the crew would be so carefully selected. On the other hand, the damaged nature of both the Canaanite jar KW 1553 which contained whole olives on the Uluburun and Chian amphora P0488 from the Mazotos shipwreck may undermine the notion that these vessels were to be sold. Would provisions for the crew be more likely to be stored in damaged amphorae than cargo intended for sale?

Yet another personal letter preserved on papyrus may offer insight into the recovery of whole olives on the Mazotos shipwreck from both inside amphorae and in the surrounding seafloor sediment: a man writes to his mother that he has received the basket of olives she sent to him, then later in the letter asks for more olives – ‘Send me an extra jar of olives for a friend; do not fail to do so (P. Mich 3 203, SB IV 7356 in Edgar 1931).’ What is interesting here is that, in the same letter, whole olives are referred to as being transported in both perishable baskets (which would not survive on an ancient shipwreck) and in ceramic jars. Is it possible that both were carried aboard the Mazotos ship, which would explain the volume of olives recovered from the seafloor sediment, as well as from within amphorae?

The ample evidence from textual sources discussed above certainly shows that whole olives were transported as a trade item in the ancient Mediterranean, so the purpose of whole olives on board ancient ships as an object of trade is an obvious possibility. Whole olives as gifts Kebler has conducted a thorough study on the import of olives and olive products to Egypt during the New Kingdom (1550 – 1150 BCE) and has postulated that prior to their widespread cultivation in Egypt, olive products were an elite gift exchanged among royalty to cement international alliances (Kebler 2009). The cargo of the Uluburun shipwreck, which is contemporary with this period, has often been discussed as a royal consignment being sent from one great king to another, due to the similarity of this cargo to shipments of goods described in the Amarna tablets (Bass 1987). If this is the case, the presence of whole olives among such an elite cargo may reinforce the idea that olives in this period, not as widely cultivated at this time, were a prestige item and designed to be a gift to a distant ruler.

On ancient shipwrecks, we may see physical evidence for the trade and transportation of whole olives that ancient sources describe. The growing importance of underwater archaeology and the increasing interest in organic remains offers a new line of evidence for the movement and trade of whole olives which is currently understudied and little understood. By utilising evidence from both ancient shipwrecks and textual sources, a more complete picture of the trade and transportation of whole olives can be seen. Exploring the purpose of whole olives on ancient ships: trade items, gifts, or provisions?

For later periods, explicit references to baskets of olives as gifts, as discussed above, shows that the gift-giving of olives was not solely an elite practice, and that the practice of giving olives as gifts had great longevity in the ancient Mediterranean from the Late Bronze Age, to Roman Egypt. Yet, even in giving olives as gifts they must first be procured, and in the letters mentioned above it is not clear if the olives sent as a gift were grown by the giver or purchased at market prices. Thus, whole olives found

Several purposes have been suggested for olive remains recovered from ancient shipwrecks. The first is as a proper element of the cargo in the form of a trade product meant to be sold in markets abroad. The second is as a gift designed to cement alliances. The third possible function of olives on board were as provisions for the crew. Here, these three possible functions are explored, and quantification exercises are employed to test their plausibility. 95

Lisa Briggs aboard shipwrecks could be simultaneously a gift, and an object of trade.

from the Mazotos shipwreck? If we are postulating a crew of five for this ship as well, and which would again be consuming 10 olives per person per day, a supply of 10,000 olives would last for 200 days. While results from the quantification exercises for the Uluburun are less clear, the results from the Mazotos shipwreck seem to suggest that either the crew were transporting provisions which they hoped would last for 200 days, or that these olives were a proper element of the cargo, and were intended for sale.

Whole olives as provisions for the crew In recent scholarship, long-distance trade conducted across open water, and not dependant on the coast, has replaced coastal tramping as the predominant model of trade in the ancient world (Arnaud 2011: 61). As such, our models regarding the length of journeys and the provisions required to sustain sailors through open-water voyages must be amended. It is a distinct possibility that the whole olives found aboard the Uluburun and Mazotos shipwrecks were intended to be provisions for the sailors undertaking the journey. The damaged nature of the two ceramic transport containers discussed above seems to strengthen this argument. Specifically, in the case of the damaged Mazotos amphora, the well-worn collar of the amphora where the neck has been broken off or removed appears to be an old break, and may have been broken in antiquity. The ease of removal is greatly enhanced for these olives, as one imagines the great difficulty involved in removing whole olives from an amphora with an intact, and narrow, neck. Perhaps quantification exercises can help to establish the purpose of these olives.

Conclusion It is not the contention of this paper that olive oil production and transportation was secondary to that of whole olives. Rather, what this work has attempted to show is that the growth, trade, packaging, transportation, and consumption of whole olives was wide-spread in antiquity and has been largely overlooked in modern scholarship. Shipwreck evidence for the trade and consumption of whole olives supports this contention, as whole olives included in transport containers are frequently found on ancient Mediterranean shipwrecks, the earliest of which dates to the Late Bronze Age. In addition, literary, epigraphic, and papyrological evidence attests to the widespread consumption, movement, trade, and gift-giving of whole olives. While the mention of whole olives is not as frequent as mentions of olive oil, this should not negate their role as an important element of the Mediterranean diet. Further, exercises in quantification may suggest that the sheer volume of whole olives transported (over 10,000 on the Mazotos shipwreck) indicate that these olives were objects of trade, rather than provisions for the crew. Further work is underway on this large collection of olive remains, with publication pending (Pagnoux, Margaritis, Demesticha in preparation), which may help clarify the purpose of whole olives on the Mazotos shipwreck.

Quantification of the olives recovered from ancient shipwrecks In Aristophanes’ Ecclesiazusae, written around 392 BCE, the Chorus bemoans the greediness of citizens motivated by the three-obol compensation for public service, recalling when ‘during the archonship of generous Myronides, none would have dared to let himself be paid for the trouble he spent over public business; each one brought his own meal of bread, a couple of onions, three olives and some wine in a little wine-skin. But nowadays we run here to earn the three obols, for the citizen has become as mercenary as the stonemason’ (O’Neill 1938). We could take the modest rations described here as lunch to be a typical, if slightly frugal, lunch from the Classical period. If we take a lunch that contains three olives in addition to an assortment of other foods (in this case bread, onions, and wine) as a typical lunch, and further suppose that three additional olives may have been eaten at both breakfast and dinner, that would give us a suggested number of nine olives per day for an adult male in Classical Athens. For ease of multiplication (and for the possibility that four olives may have accompanied an evening meal) let us say 10 olives per day. That would be a hypothetical consumption of 3,650 olives per year for an adult male.

As has been shown in this work, discussions surrounding the production and transportation of whole olives has been overshadowed by that of olive oil. Yet, fewer elements of infrastructure and heavy apparatuses are needed to produce whole olives. The nutritional value of whole olives exceeds that of olive oil in levels of carbohydrates, calcium and magnesium (Zipori 2015), and it has been suggested that whole olives were more widely available to non-elite members of society prior to the Classical period (Foxhall 2007). We find ample evidence for the movement and trade in whole olives in both ancient literature and in ancient documents relating to trade such as customs receipts, tax records, and personal letters requesting provisions.

Over 2,500 olive stones found in a Canaanite jar on the Uluburun shipwreck could have either been an item of trade, a gift, or provisions for the crew. If a crew of five sailors were each eating 10 olives a day they would consume 50 olives a day. A jar containing 2,500 olives would, in this scenario provide a supplemental food source for 50 days. And what of the over 10,000 olives recovered

In addition, questions about crop processing and the social organisation of agricultural pursuits in antiquity have often focused on charred remains and phytoliths (Hillman 1981, Jones 1984, Jones 1996, Harvey and Fuller 2005, van der Veen 2007, Lightfoot and Stevens 2012) but perhaps the waterlogged agricultural products that are the result of such processes may provide a different source of information 96

The Whole Story as they are subject to a different preservation bias (e.g. Colledge and Conolly 2014). A focus on charred remains from oil processing, and the perceived prerequisite of oil processing equipment to be present for olive cultivation to have occurred, must be questioned and supplemented with evidence from ancient shipwrecks.

changes in plant diet breadth.’ Quaternary Science Reviews 101:193 – 206. DAVIES, J.K. 2006 ‘Hellenistic Economies’. In Bugh, G (ed), The Cambridge Companion to the Hellenistic World. Cambridge University Press: Cambridge. DEMESTICHA, S. 2011. ‘4th Century BC Mazotos Shipwreck, Cyprus: a preliminary report.’ International Journal of Nautical Archaeology 40: 39–59.

When we find whole olives on ancient shipwrecks, this serves to strengthen the contention that whole olives formed an important part of the Mediterranean diet and culture. Whether these olives were a luxury trade item as suggested by Haldane (1993), a royal gift designed to cement international relations as suggested by Kebler (2009), rations for workers, provisions for the crew of a ship, a modest gift sent from one relative to another, or as a supplement to a larger dish, their presence on ancient shipwreck allows us to infer that whole olives, and not just olive oil, were an important enough commodity to be collected, selected, and stowed in a cargo assemblage intended for transport across the sea.

EDGAR, C.C. 1931. Zenon Papyri, Nos. 1 – 120. University of Michigan Studies, Humanistic Series 24. Ann Arbor: Michigan. FOXHALL, L. 2007. Olive Cultivation in Ancient Greece: Seeking the Ancient Economy. Oxford University Press: Oxford. GALILI, E., WEINSTEIN-EVRON, M., HERSHKOVITZ, I., GOPHER, A., KISLEV, M., LERNAU, O., KOLSKA-HORWITZ, L., LERNAU, H. 1993. ‘AtlitYam: A Prehistoric Site on the Sea floor off the Israeli Coast.’ Journal of Field Archaeology 20.2: 133–157.

Works cited

GALILI, E., JEAN STANLEY, D., SHARVIT, J., WEINSTEIN-EVRON, M. 1997. ‘Evidence for Earliest Olive-Oil Production in Submerged Settlements off the Carmel Coast, Israel.’ Journal of Archaeological Science 24: 1141–1150.

ARNAUD, P. 2011. ‘Ancient sailing-routes and trade patterns: the impact of human factors.’ In Maritime Archaeology and Ancient Trade in the Mediterranean, Robinson, D., Wilson, A. (eds.). Oxford Centre for Maritime Archaeology, Institute of Archaeology: Oxford.

HALDANE, C. 1993. ‘Direct Evidence for Organic Cargoes in the Late Bronze Age.’ World Archaeology 24.3: 348–360.

BASS, G. 1987. ‘Oldest known shipwreck reveals splendors of the Bronze Age.’ National Geographic Magazine, National Geographic Society: USA. 172: 693–734.

HAMILAKIS, Y. 1996. ‘Wine, oil and the dialectics of power in Bronze Age Crete: a review of the evidence.’ Oxford Journal of Archaeology 15: 1–32.

BASS. G. 1991. ‘Evidence of Trade from Bronze Age Shipwrecks.’ In Bronze Age Trade in the Mediterranean. Papers presented at the conference held at Rewley House, in December 1989. In Gale, N.H. (ed.). Paul Astroms Forlag: Sweden. Pp. 69–82.

HAMILAKIS, Y. 1999. ‘Food technologies/technologies of the body: The social context of wine and oil production and consumption in Bronze Age Crete.’ World Archaeology 31.1: 38–54.

BLITZER, H. 1993. ‘Olive cultivation and oil production in Minoan Crete’. In La production du vin et de l’ huile en Mediterranee´, Symp. Intern, Aix en Provence 1991, Amouretti M.C., Brun J.P., Eitam D (eds.). Bulletin de Correspondance Hellenique 26 (Suppl): France. Pp. 163–175.

HOPPER, R.J. 1979. Trade and Industry in Classical Greece. Thames & Hudson Ltd: London. HORDEN, P., PURCELL, N. 2000. The Corrupting Sea: a study of Mediterranean History. Blackwell: Oxford. KATZEV, S. 2005. ‘Resurrecting an Ancient Greek Ship: Kyrenia, Cyprus’. In Bass, G (ed.), Beneath the Seven Seas. Pp. 72–9. Thames and Hudson: London.

BOUND, M., VALLINTINE, R. 1983. ‘A wreck of possible Etruscan origin off Giglio Island.’ International Journal of Nautical Archaeology and Underwater Exploration 12.2: 113–122.

KEBLER, J.M. 2009. ‘Royal Gift Exchange between Mycenae and Egypt: Olives as ‘Greeting Gifts’ in the Late Bronze Age Eastern Mediterranean.’ American Journal of Archaeology 113.3: 339–352.

BUTLER, S. 1994. The Odyssey by Homer. Retrieved from classics.mit.edu, 2018.

LANZA, F. 2011. Olive: A Global History. Reaktion Books Ltd: London.

CARLSON, D. 2003. ‘The Classical Greek Shipwreck at Tektaš Burnu, Turkey.’ American Journal of Archaeology 107.4: 581–600.

MATTINGLY, D.J. 1988. ‘Olea Mediterranea?’ Journal of Roman Archaeology 1: 152–161.

COLLEDGE, S., CONOLLY, J. 2014. ‘Wild plant use in European Neolithic subsistence economies: a formal assessment of preservation bias in archaeobotanical assemblages and the implications for understanding

MARGARITIS, E., JONES, M. 2007. ‘Crop processing of Olea europaea L.: an experimental approach for

97

Lisa Briggs RUNNELS, C.N., HANSEN, J. 1986 ‘The olive in the prehistoric Aegean: the evidence for domestication in the Early Bronze Age.’ Oxford Journal of Archaeology 5: 299–308.

the interpretation of archaeobotanical olive remains.’ Vegetation History and Archaeobotany 17: 381–392. MARGARITIS, E. 2013. ‘Distinguishing exploitation, domestication, cultivation and production: the olive in the third millennium Aegean.’ Antiquity 87.337: 764– 757.

SALLARES, R. 1991. The Ecology of the Ancient Greek World. Gerald Duckworth & Co. Ltd: London. SIJPESTEIJN, P. 1987. Customs Duties in Graeco-Roman Egypt. Stud.Amst. XVII. Zutphen: Holland.

MARGARITIS, E. 2014. ‘The Kapeleio at Hellenistic Krania: Food Consumption, Disposal, and the Use of Space.’ The Journal of the American School of Classical Studies at Athens 83.1: 103–121. MARKEY, S. 2003. ‘Ancient Greek Wreck found in Black Sea.’ Retrieved from: https://news.nationalgeographic. com/news/2003/01/0110_030113_blacksea.html

STERN, B., LAMPERT MOORE, C.D., HERON, C., POLLARD, A.M. 2008. ‘Bulk stable light isotopic ratios in recent and archaeological resins: towards detecting the transport of resins in antiquity?’ Archaeometry 50.2: 351–370.

MILLS, J., WHITE, R. 1989. ‘The identity of resins from the Late Bronze Age shipwreck at Ulu Burun (Kas).’ Archaeometry 31.1: 37–44.

WARNOCK, P. 2007. Identification of Ancient Olive Oil Processing Methods Based on Olive Remains. International Series 1635, BAR Publishing: Oxford.

MONROE, C. 2010. ‘Sunk Costs at Late Bronze Age Uluburun.’ Bulletin of the American Schools of Oriental Research 357: 19–35.

WALISZEWSKI, T. 2014. Elaion: olive oil production in Roman and Byzantine Syria-Palestine. University of Warsaw Press: Warsaw.

MURRAY, A.T. 1939. Demosthenes with an English translation. Harvard University Press: Cambridge.

WILSON, A., BOWMAN, A. 2018. Trade, Commerce, and the State in the Roman World. Oxford University Press: Oxford.

O’NEIL, E. 1939. Aristophanes’ Ecclesiazusae. The Complete Greek Drama, Volume 2. Random House: New York.

WINTER, J.G. 1936. Miscellaneous Papyri, Nos. 131 – 221. University of Michigan Studies, Humanistic Series 40. Ann Arbor: Michigan.

PALAIMA, T. 2012. ‘Linear B.’ In E. Cline (ed.), The Oxford Handbook of the Bronze Age Aegean. Oxford University Press: Oxford.

YOUTIE, H.C. 1944. ‘Papyrus SB VI 9025.’ American Journal of Philology 65: 251–255. ZIPORI, I., YERMIYAHU, U., EREL, R., PRESNOV, E., FAINGOLD, I., BEN-GAL, A., DAG, A. 2015. ‘The influence of irrigation level on olive tree nutritional status.’ Irrigation Science 33: 277–287.

PAYTON, R. 2013. ‘The Ulu burun writing-board set.’ Anatolian Studies 41: 99–106. PULAK, C. 1988. ‘The Bronze Age Shipwreck at Ulu Burun, Turkey: 1985 Campaign.’ American Journal of Archaeology 92.1:1–29. Archaeological Institute of America, USA. PULAK, C. 1998. ‘The Uluburun shipwreck: an overview.’ International Journal of Nautical Archaeology 27.3: 188–224. PULAK, C. 2000. ‘The Balance Weights from the Late Bronze Age Shipwreck at Uluburun.’ In Metals Make the World Go Round: The Supply and Circulation of Metals in Bronze Age Europe: Proceedings of a Conference Held at the University of Birmingham, University of Birmingham Press: Birmingham. PULAK, C. 2005. ‘Who Were the Mycenaeans Aboard the Uluburun Ship?’ In R. Lafhneur and E. Greco (eds.), EMPORIA: Aegeans in the Central and Eastern Mediterranean: Proceedings of the 10th International Aegean Conference.  Italian School of Archaeology: Athens. RENFREW, C. 1972. The emergence of civilization. The Cyclades and the Aegean in the third millennium BC. Methuen: London.

98

9 Piracy in the Hellenistic Period: A Misunderstood Phenomenon Joseph M. Pacheco Jr University of Haifa [email protected] Keywords: Piracy; social banditry; Mithridates VI; pirate archaeology; mercenary warfare

Introduction

except those of their allies. This practice of violent seizure or reprisal (rhysia) was not only a way for individuals to gain recompense or revenge, but it also gave private citizens an opportunity to conduct legitimised piratical raids to fill their coffers (Thucydides 5.1; Gabrielsen 2013: 141).

The term ‘pirate’ invariably evokes a sense of fear and suspicion, as well as other overwhelmingly negative emotions. Despite recent attempts by Hollywood to romanticise pirates and piracy in such film franchises as Pirates of the Caribbean, contemporary news reports describing hijackings and kidnappings carried out by pirates operating in the waters around the Horn of Africa and other piracy hotspots have reinforced our long held belief that, in the real world, pirates mean trouble. Throughout history, most people, at least those kind enough to leave historical records, have thought the same. Even as far back as Antiquity, historians, orators, playwrights and diplomats were mentioning the grievous misdeeds of pirates.

This practice of ‘part-time’ pirating likely originated centuries (if not millennia) ago. A custom paralleling those of the Greeks of Ozolian Locris and the opportunists mentioned by Aristotle can be found in the Amarna letters, a series of correspondences between the Egyptian pharaohs and the other Mediterranean and Near Eastern potentates dating to around the middle of the fourteenth century BC. One particular Amarna letter (EA 38) was written by the king of Alashiya (now known as Cyprus) to the Egyptian pharaoh, and describes incursions by ‘…the men of the land of Lycia…’ (Rainey 2015: 351). These attacks took place on an annual basis and involved the marauders raiding a Cypriot town and then returning to their homeland in what is now coastal southern Turkey. Because these attacks were carried out by sea, they can certainly be labelled as acts of piracy. However, the raiders carrying out the attacks were most likely merely supplementing their main source of income, which may well have been farming, fishing, trading or a combination of these and other activities. Of course, it is entirely possible that they chose other places to raid at different times of the year and thus piracy would have accounted for the lion’s share of their profits.

However, while antique pirates did engage in numerous raids on merchant ships as well as coastal communities at least as far back as the fourteenth century BC (Gilan 2013) and have continued to do so up until modern times, is this really the whole story or are we just scratching the surface and showing one side of the coin? It is my intention in this article to show pirates in a different light and explain the varied ways that piracy and pirates benefitted society as a whole and especially those most powerful communities during the specific temporal range of the Hellenistic Period and early Roman Period (approximately 323 BC31 BC). Before getting into the details, it might be useful to understand a bit more about the fellows doing all of this helpful pirating and where they possibly operated from.

It is very instructive to look at these men of Lycia and the later ‘part-time’ pirates as social bandits. This specific category of raiders stole for the common good and were seen as robbers and outlaws by most groups (especially those targeted by their assaults), but were praised within their own communities as heroes and even saviours and were fully supported in their activities by their family and neighbours rather than being cast out for their malicious acts against other settlements (Hobsbawm 1969: 13).

Pirate identity Aristotle mentions that pirates, at least in his time (fourth century BC) and presumably earlier, were just farmers who would seasonally embark on raids to supplement their income and stave off possible starvation during lean years (Aristotle Pol. 1.1256b). Piracy seems, to him, to be a natural mode of livelihood carried out by small, disorganised bands of ‘weekend’ pirates who would congregate to pillage, and then head back to their farmsteads with their plunder. As late as the fifth century BC, the Ozolian Locrian cities of Chaleion and Oianthia allowed their citizens to conduct raids against all shipping

Things had changed drastically by the mid-second century BC, completely altering the face of Mediterranean piracy and ushering in what can definitely be called the golden age of piracy in Antiquity. Professional pirate groups had begun to appear, most notably in Rough Cilicia, a 99

Joseph M. Pacheco Jr wild, mountainous area in what is modern-day southwest Turkey. Sources say that piracy began in Rough Cilicia after a pretender to the Seleucid throne named Didodotus Tryphon installed mercenaries at Coracesium, an easily defensible fortress with a protected anchorage (Strabo 14.5.2). Once Tryphon was defeated, these mercenaries turned to piracy and were joined by large numbers of other members of the maritime underclass or ‘ochlos nautikos’, a melting pot of disaffected sailors, fugitive slaves and renegade nobles among others (Rauh et al.. 2008: 197).

In addition, the distribution of sparse Hellenistic ceramics found on the island points to it being used during that period as an ‘…occasional, opportunistic port and refuge for sailors.’ (Varinlioğlu et al.. 2017: 58). This is just the type of location that would have been favoured by pirates who wanted to be isolated and protected, but still close enough to maritime routes to carry out sorties quickly and easily. The site that can be pointed to as an archetypal Hellenistic pirate bastion is located across the eastern Mediterranean on the western coast of Crete. Phalasarna, a once thriving harbour town, was probably resettled by pirates during the same period pirates were active in Rough Cilicia (late-second/early-first centuries BC). The excavators have argued that slipshod renovations to the impressive fortifications of the cothon-style harbour are a sure sign of a pirate presence (Frost and Hadjidaki 1990: 524). Even more convincing is the fact that evidence of a punitive Roman campaign against the harbour town can still be seen in the form of a deliberately blocked harbour entrance (Hadjidaki 1988: 475). Arrowheads and catapult stones excavated from this site are further signs of a violent incursion. Plutarch relates that a Roman commander named Metellus carried out just such a punitive expedition against Cretan pirates in support of Pompey the Great’s larger anti-pirate campaign in 67 BC (Plutarch Pomp. 29).

It is hard to say exactly how these piratical groups were organised but it is quite possible that they were much more egalitarian than society at large. Possibly, arch-pirates or pirate chiefs were elected to lead the bands. Just as Hellenistic dynasts can be seen as ‘stationary bandits’ who basically exploited the populations under their control in a legitimate manner with the help of a group of friends (φίλοι) (Manning 2010: 59), arch-pirates can be seen as ‘roaming bandits’ who no doubt had bases from which they operated but also conducted wide-ranging raids across the Mediterranean. In any case, these hodgepodge and polyglot groups of buccaneers repurposed and refurbished old Ptolemaic and Seleucid forts and outposts, in some cases leaving archaeological traces of their haphazard renovations. Possible pirate bases

Piracy and the slave trade

One such supposed pirate haunt was Iotape. While the town was not officially founded until 52 AD by Antiochus IV of Commagene, archaeological investigation has shown that the site was occupied earlier. Hellenistic ceramics attest to a presence in this period (Brock et al.. 2000: 174) and a crudely built wall surrounding structures built with local limestone set without mortar and showing signs of haphazard reconstruction (rough rubble masonry using pebbles set in mortar), possibly point to a pirate occupation. Presumably, the pirates were more adept at sailing and fighting than building. The Hellenistic Period also saw a dramatic increase in rural land use in Rough Cilicia including an increase in forestry (Brock et al.. 2000: 158), which also hints at a pirate presence. Of course, this archaeological evidence taken by itself does not conclusively point to a pirate occupation in Rough Cilicia during the Hellenistic Period, but taken together with Strabo’s attestation that this area was a breeding ground and haven for pirates (Strabo 14.3.2), it is a safe bet.

Now that the pirates were more or less organised and had bases from which to operate, they could carry out raids to pillage and plunder. One of the most sought after prizes was slaves which could be sold at a premium to that ravenous consumer of human chattel: the Roman Republic. Rome required enormous amounts of slaves to labour on ever growing latifundia, or plantations, owned by rich landowners who had supplanted independent farmers (Boese 1973: 52). The Roman thirst for slaves was seemingly unquenchable and it is possible that as many as tens of thousands of slaves were sold at the free emporium at Delos on any given day in the second half of the second century BC (Garlan 1987: 9). Piracy was a major source of slaves, perhaps second only to the human spoils gained in battle. This led to a sort of tacit agreement between the Romans and the pirates: each side would not unduly harass the other and business would continue to be brisk with the pirates providing slaves in exchange for coin or other goods such as wine.

Furthermore, Dana Island, just off the Rough Cilician coast near the modern-day town of Silifke in Turkey, might have served as a base or waystation for pirates in the mid-late second/early-first centuries BC. A 2016 survey found the possible remains of numerous slipways on the northwest coast of the island that could have accommodated light pirate craft (Varinlioğlu et al.. 2017: 54). Hellenistic Period ceramics have also been found associated with a fortress on the southwest of the island that could very well have been repurposed by pirates (Varinlioğlu et al.. 2017: 57).

This theory prompted one researcher, Andre Tchernia, to propose that Will Type 10 transport amphorae, which contained wine, and have been found off the coast of one of the main pirate strongholds in Rough Cilicia, Antiochia ad Cragum (Hoff and Townsend 2013: 66), were exchanged for slaves at Delos. Will Type 10 amphorae were produced in large quantities along the Adriatic coast of Italy and, thus, provide a direct archaeological connection between the Romans and the pirates.

100

Piracy in the Hellenistic Period As long as slavery was so essential to the Roman Republic, so were pirates, since they were able to provide a steady stream of slaves to work in Roman fields. Besides Delos, the pirates of Cilicia also utilised nearby cities in Pamphylia, such as Side and Attaleia, as slave trading centres which would have benefitted this region economically at least until the Romans under M. Antonius targeted Side specifically during an anti-pirate operation in 102 BC (Strabo 14.3.2; Brock et al.. 2000: 153).

The Rhodians were lauded before the Romans, perhaps somewhat falsely, as the anti-pirate police. After a bit of analysis, it becomes clear that the Rhodians also maintained something of an indirect relationship with the pirates operating in the eastern Mediterranean. Rhodes really only acted against pirates when they threatened its interests. Rhodes even saw the benefit of copying some of the pirates’ ship designs and using piratical tactics such as forming small squadrons of light triemiolia type ships to launch surprise hit and run attacks (Gabrielsen 2001: 224).

Horden and Purcell made the persuasive point that pirates rely on busy trade routes to maintain themselves (Horden and Purcell 2000: 157), and a disruption in trade would lead inevitably to a downturn in the profits available to pirates. A major portion of the goods seized in pirate raids (besides the possible hoards the pirates kept in their strongholds) were recirculated back into the mainstream economy. A sort of symbiosis existed between those practicing piracy and merchants dealing in (mainly) above-board, legitimate commerce. A good indicator of this fact is that maritime loans were still being given out in Asia Minor during the Mithridatic Wars, when Cilician pirates were privateering for the Pontic king, and grain prices remained stable from 130 BC to 50 BC, right through the infamous pirate periods in Cilicia and Crete (Avidov 1999: 18–19). Individual entrepreneurs were, of course, ruined, as well as being possibly enslaved or killed if they happened to be personally travelling with their cargo, but the Hellenistic economy and those later and previous economies which also had to function in the midst of a pirate ‘menace’ were not necessarily negatively affected by piracy in a significant way. As Gabrielsen wisely points out, ‘From an economic point of view it was immaterial whether a given cargo reached the market onboard the ship of the “armed” or the “unarmed” kind of merchant.’ (Gabrielsen 2013: 148–49).

Moreover, when it better suited Rhodes’ political interests, she turned a blind eye to piracy. For example, Strabo relates that the pirates were harassing Syrians (Seleucids) in the eastern Mediterranean and, as the Rhodians were unfriendly towards this dynasty, they let the pirates operate without any interference whatsoever (Strabo 14.5.2). Rhodes also put a great deal of effort into protecting her own merchant shipping, even outfitting specialised protection boats (phylakides nees) to help escort lumbering trading vessels (Gabrielsen 2001: 230). However, Rhodes very rarely aggressively rooted out piracy. The island state did profit from the protection market, however, and made her allies pay large sums in exchange for security. Delos paid Rhodes five thousand drachmas for protection (Gabrielsen 2001: 234) and the Hellenistic kingdoms at varying points supplied money and war material to support Rhodian antipirate operations. Rhodes also signed treaties with multiple Cretan cities, including Hieraptyna and Olous, stating that if pirates were captured along with their ships or booty, then Rhodes would receive the majority of the spoils. Clearly Rhodes benefitted from piracy or at least the threat of piracy and was able to maintain a reputation as a naval protector or προστάτες ‎(Latin prostates, ’one who stands before, protector, guardian’) (Gabrielsen 2001: 219) while actually doing quite little to interdict piracy for the sake of the common good. On a few occasions, the Romans were able to persuade or coerce the Rhodians to act against pirates. For instance, Livy recounts that a Roman praetor sent two Rhodian ships to deal with Hybristas, a Lacedaemonian pirate leader who was operating in the waters around the island of Ithaca (Livy 37.13).

‘Clearers of the Sea’ In addition to the economic benefits of a booming slave trade, the presence of pirates in the Eastern Mediterranean indirectly benefitted the Romans in another important way: they wreaked havoc. At least initially, before around 100 BC, the pirates were mainly marauding in areas controlled by rivals or enemies of Rome, so there was no political will or reason to try to clear the seas of piracy. Additionally, as Shaw points out, the presence of pirates in the waters of the eastern Mediterranean provided the Romans with a ‘… perpetual justification for Roman military intervention in the region.’ (Shaw 1984: 39). The Romans could always use the pirates as an excuse to campaign in the east and eventually Pompey the Great did use the pirate menace to gather a force of men and ships of great strength to root them out. While the Cilician pirates were indeed a serious problem for Rome and her allies during the early/ mid-first century BC, Pompey of course did not stop after their strongholds were taken, but rampaged throughout the Near East.

Pirates as mercenaries Pirates and Hellenistic powers also interacted in an even more direct way, which benefitted both parties greatly. Pirates were regularly recruited as mercenaries into Hellenistic armies and navies which had need of fighting men with particular seaborne and amphibious fighting skills. The pirates could count on a more stable income and take advantage of the numerous conflicts which plagued the Hellenistic Mediterranean if they were willing to join as mercenaries. Mercenaries abounded in the Hellenistic Period. According to G.T. Griffith, mercenaries were being hired in large numbers by armies already in the Saite period (approx. mid-seventh century BC to mid-sixth century 101

Joseph M. Pacheco Jr BC) (Fantalkin and Lytle 2016) and continued to be used right through the Hellenistic Period (Griffith 1968: 3). The numerous conflicts of the late Hellenistic Period led to large numbers of trained fighting men facing destitution, who led transient lifestyles trying to find service as mercenaries (Griffith 1968: 238). The large number of conflicts, of course, also created great demand for mercenary soldiers and sailors, and it seems that Hellenistic recruiters could easily assemble an army of wandering warriors willing to fight for a price. The same warriors who turned to mercenary service could just as easily turn to piracy or terrestrial banditry, perhaps even more lucrative ways of supporting themselves and their families. They could then, as will be shown, hire themselves out as mercenaries when the opportunity presented itself. Thus, the piratemercenary became a feature of Hellenistic warfare.

willingness of Athens and other later powers to accept pirates into their ranks wholeheartedly was perhaps due in part to the historical Greek ambivalence about and acceptance of piracy in the Archaic period. As mentioned, some cities still allowed legitimised piracy to be practiced as late as the fifth century BC. As will be shown, sometimes this trust in the loyalty of pirates was woefully misplaced and Hellenistic commanders were betrayed by their recent hires, but that was not uncommon mercenary practice in any period. Indeed, some modern scholars have commented on the fact that the line between pirate and mercenary was quite blurry in the Hellenistic Period. Ormerod wrote that pirates and mercenaries on Crete were nearly indistinguishable (Ormerod 1978: 143). Similarly, Janice Gabbert remarked on the interchangeability of pirates and mercenaries during the third century BC (Gabbert 1986: 156).

Surely, more traditional seaborne mercenaries had been hired much earlier than the Hellenistic Period. Perhaps the first recorded use of this type of nautical mercenary comes from the Amarna letters (Moran 1992: 178), specifically a letter in which a King of Byblos named Rib-Addi was targeted by the Amurru. While the Amurru besieged the coastal city on the landward side, the Arwadites, who had been hired by the Amurru for their naval prowess (Vidal 2008: 7), blockaded the harbour and intercepted ships sent out by the king. Besides merely being paid for their services, the Arwadites might have had an ulterior motive. Arwad (Ruad island off the coast of modern-day Syria) was a Late Bronze Age community built on trade, and was in competition with other Levantine city-states which also traded widely. A blow to Byblos (or Tyre or Sidon) meant that the Arwadites could step in and fill the commercial vacuum.

A famous example of the large-scale use of piratemercenaries involves Demetrios Poliorcetes, who attempted an ultimately unsuccessful siege of the island of Rhodes in 304 BC. Demetrios had recruited an archpirate (pirate leader) named Timokles who was defeated in battle by the Rhodian commander Amyntas, showcasing pirates’ inability to best regular naval squadrons in pitched sea battles. Timokles commanded three undecked pirate vessels and was merely the strongest of the pirates who had joined up with Demetrios (Diodorus 20.97.5). In addition to Timokles and his men, who were perhaps quasiprofessional pirates, a large number of merchant captains seconded themselves into the service of Demetrios. No doubt these merchants were chomping at the bit to capture some booty from the wealthy Rhodians, as well as kidnap the Rhodians themselves, who could be ransomed for 1,000 drachmas according to a prior agreement between Demetrios and the Rhodians (Diodorus 20.84.6). In effect, these merchants, while not previously pirates, became essentially privateers for the duration of Demetrios’ illfated campaign, and highlight the importance of both mercenaries and privateers to Hellenistic armies and navies. This episode also shows that a grey area existed, wherein pirates could become merchants selling slaves in exchange for wine, while merchants could morph into pirates or privateers if a particularly auspicious opportunity afforded itself.

Before this new type of seaborne mercenary which appeared on the scene in the Hellenistic Period is discussed further, it should be made clear what each side (military commander and pirate) had to gain from such a relationship. Hellenistic armies and navies were provided with seasoned fighters with a repertoire of specialised skills and tactical knowledge as well as equipment such as fast and light ships. These vessels could be used to great effect to swarm and catch off guard the larger quinquiremes, which composed most of the Hellenistic battle fleets, and whose emphasis was on harbour siege warfare (Murray 2012: 23). The pirates (mercenaries) gained an opportunity to seize war booty (including captives which could be sold into slavery) as well as regular payment for their services and perhaps an agreement that their piratical activities would be overlooked by the power that they served well.

And it was not only Demetrios who realised the immense strategic benefits of taking on pirates as mercenaries or at least allying with them in pursuit of a common goal. Mithridates VI Eupator, the Poison King, was one of Rome’s gravest enemies. Besides massacring considerable numbers of Italians during the Asian Vespers in 88 BC, he successfully invaded and occupied large parts of Rome’s eastern territories. He shrewdly allied himself with the Cilician pirates who worked as intelligence gatherers for the Pontic king rather than actual mercenaries fighting in regular sea or land battles (del Hoyo 2014: 410). Using his piratical spies, Mithridates was able to gain valuable information about Sertorius’ revolt in Spain (80 BC-72

Before the official start of the Hellenistic Period (323 BC), the mid-fourth century career of Charidemus proves that pirates could seamlessly transition into the mercenary role. Charidemus was a pirate who successfully captured three cities in the Troad region and later appears as a general in the service of Athens according to Demosthenes (23.152–60). Charidemus’ career path was not unique and was followed by others, as will be shown below. The 102

Piracy in the Hellenistic Period BC) at the other end of the Roman world, and possibly coordinated his offensives accordingly. The pirates also provided more tactical intelligence as they were used to sailing throughout the Mediterranean and had a keen eye for detecting which ports or coastal communities would be easy pickings for a Pontic war fleet.

was not wholly satisfied with this arrangement and was open to bribery by a Lysimachean general named Lykon. The conspiratorial pair designed to use a ruse to capture Ephesus, and Andron pretended to have captured some Lysimachean troops and led them into the city. In reality, this group carried concealed weapons and were able to open the gates to the rest of the army, and thus Ephesus was captured by subterfuge with the help of a traitorous pirate. Lykon wisely decided to disband Andron’s group of pirate-mercenaries because their loyalty was no doubt questionable. This episode goes to show that pirates, and mercenaries in general, during the Hellenistic Period, and throughout history, were not the most loyal of troops because they fought for money and not ideology.

As Strabo (14.1.32) makes clear, pirates already had their own wide-ranging intelligence networks in place to try to gain knowledge about rich cargoes and their destination. Pirates could move in circles that soldiers could not. As briefly mentioned, they came from and moved about stealthily in the maritime understratum of society. Of course, the pirates were probably not ideologically supportive of Mithridates in his wars against Rome, even though the existence of a naval hegemon would definitely not be in the pirates’ best interest. In return for their services, Mithridates most likely gave them the freedom to raid as they pleased against enemy or neutral shipping, provided them with coin and, as Appian relates (Mith. 9.63), fitted them out with warships.

However, pirates turned mercenaries did serve their paymasters well in some cases. Demetrios’ successor and son, Antigonos II Gonatas, was just as warlike as his father, and was determined to succeed where his predecessor had failed. Undoubtedly with the thought of Demetrios’ painful loss at Rhodes in the back of his mind, Gonatas would stop at nothing to prevail during his siege of Cassandreia in north-eastern Greece in 276 BC. Cassandreia was a hard nut to crack and after ten months, the tyrant Apollodorus was no closer to surrendering to Gonatas’ besieging army. The king of Macedon then decided to turn to a famed pirate named Ameinas to help him win the city by the use of stratagem. Polyaenus once again relates how pirates used cunning rather than force to outmanoeuvre their opponents (Strat. 4.18). Ameinas was sent to discuss terms with Apollodorus and brought with him a large quantity of food and wine as peace offerings. Unsurprisingly, the exhausted and undernourished defenders gorged themselves on this bounty and consequently lowered their guard. On one particular morning after feasting, a large of body of Aitolian pirates under the command of another arch-pirate named Melotas were able to use secretly constructed ladders to scale the walls at a lightly defended point and take the city. Apparently Ameinas proved his worth time and again during his mercenary service and eventually became a general in the Antigonid army according to Plutarch (Pyrrh. 29.6).

Pirates, and those more rightly termed privateers, who operated under a thin veneer of legitimacy, were also used as proxies to engage in undeclared conflict. Two instances involve Philip V of Macedon who was perhaps reticent to enter into direct, open war with Rome, at least for a time. Appian (Ill. 2.8) describes how Demetrios of Pharos, a minor ruler turned self-legitimised pirate, sought shelter with Philip V after the Romans were hunting him for committing acts of piracy in both the Cyclades and the Adriatic Sea. The Aitolian Dikaiarchos was another military adventurer who was supported by the King of Macedon during his ultimately fruitless struggle against Rome and her allies. During the Cretan War (205–200 BC), Philip overtly supported the Aitolian by giving him twenty ships with which to raid commercial shipping belonging to Rhodes (Polybius 28.1). Indeed, the Macedonian navy at this time was relatively green and untested. Going up against such a seaworthy opponent as Rhodes, even in such an indirect way, was risky business, and Philip did well to outsource some of his naval needs to an Aitolian pirate who had the necessary tactical experience and skills to complete the mission.

Pirates were clearly very useful additions to Hellenistic forces and were able to join the ranks of Hellenistic armies in quite large numbers, blurring the line between pirate and mercenary in this period. Very likely, pirates became mercenaries when dynasts or generals could afford them during campaigns and, in the opposite way, mercenaries turned to banditry and piracy once the war chests were emptied. It has already been mentioned how Tryphon’s mercenaries eventually turned to sea raiding once they were abandoned by their leader. Also, pirates were not merely used as cannon fodder or galley slaves. In contrast, it seems that Hellenistic commanders appreciated and admired their unique skill sets and their apparent gift for the use of stratagems and ploys.

Pirates, however, were not always the most reliable of allies and their loyalties were notoriously drawn to whichever king or commander had the fullest purse or the best opportunity to gain victory and plunder. Even quite early in the Hellenistic Period, generals recruited pirates to serve in specialist roles. Polyaenus relates (Strat. 5.19) that in around 300 BC, an arch-pirate turned mercenary named Andron double-crossed his patron, a general named Ainetos, in the service the aforementioned Demetrios Poliorcetes-Ainetos had generously let Andron operate out of Ephesus’ harbour. This was done perhaps on the condition that his band of pirates interdict enemy shipping and communication routes. However, Andron

Nicander of Samos was yet another pirate-mercenary who served with Hellenistic forces, this time with the Seleucid 103

Joseph M. Pacheco Jr navy. Nicander was recruited by a Rhodian admiral named Polyxenidas, who betrayed his city and went over to the Seleucids. During the Roman-Seleucid War (192–188 BC), Nicander and Polyxenidas decided to attack the Rhodians stationed on Samos under the command of Pausimachus in a combined land and sea assault. The pirates and perhaps others under the command of Nicander were again used in an irregular fashion as part of a ruse to trick the enemy into thinking they were surrounded on both land and sea. Appian (Syr. 5.24) describes that when Polyxenidas rowed towards the Rhodians with the main body of ships, Pausimachus decided to make a stand on land. However, meanwhile, Nicander had infiltrated behind enemy lines on the landward side and his presence in the rear so confused the Rhodians that they were mostly captured or killed in the ensuing chaos and carnage.

(r. 231–227 BC), the Illyrians became much bolder and turned seaborne banditry into a sort of national enterprise. They already possessed the ships and the naval expertise to transition to large-scale piracy, and during the midlate third century BC. the Illyrians were blessed with the aggressive leadership to turn these skills and technological developments into major gains for their community. For example, the Illyrians under Teuta were able to capture and plunder the Aitolian stronghold of Medion. Polybius (2.3) describes a raid on a massive scale, where an amphibious attack combined with the element of surprise allowed the Illyrians to take large amounts of material and human booty. This successful attack emboldened Teuta’s subjects to raid ships indiscriminately throughout the Adriatic (Polybius 2.8), and eventually led to Roman intervention in the region and the First Illyrian War in 229 BC. Ambitious and opportunistic Illyrian dynasts were able to realise the political as well as the economic benefits of piracy, and created a strong and expansive state (at least for some time) by practicing state-supported seaborne banditry at an unprecedented level of organisation and sophistication.

State-sponsored piracy Piratical tactics were indeed effective and were adopted by the Rhodians to an extent as has been discussed. The Romans also copied some of the pirates’ ship designs and tactics and learned a lot from their example and put this knowledge to good use during the course of their many conflicts, both internal and external, in the first century BC. For example, Sextus Pompey, the son of the famous Pompey Magnus, used piratical tactics to his advantage during the Sicilian War (42–36 BC). This conflict, while in reality a civil war fought between Octavian (later to be known as Augustus) and Sextus Pompey who controlled Sicily, was later characterised by Augustus in his Res Gestae (Augustus 25) as a police action pitting true Romans against pirates. Pirates recruited as mercenaries by Sextus Pompey probably made up only a small contingent of his force, but he used pirate style, hit and run tactics and raids to great effect against the Italian mainland, until finally being defeated by Marcus Agrippa at the battle of Naulochus in 36 BC.

Conclusion There is no doubt that pirates were a major feature of life in the eastern Mediterranean during Antiquity in general and the Hellenistic Period in particular. Piracy reached its highest peak, as mentioned, during the second and first centuries BC though it had been and remained a constant concern for a wide variety of chroniclers, dramatists and philosophers. This work has attempted to show that such concern was not warranted in all cases. While fully aware of the negative impact pirates made on seafarers and coastal communities throughout history, it would be misleading to not acknowledge the benefits pirates also provided. Pirates themselves were not the bloodthirsty buccaneers we perhaps picture them as and they were, more likely than not, up until at least the Classical period, ordinary farmers and craftsmen who took to the sea occasionally to raid in support of their communities and with their full blessing. Even during the closing centuries of the pre-Christian era, after centralised states began the process of the total monopolisation of the legitimate use of violence, people turned to piracy for a myriad of reasons other than merely to kill and plunder without prejudice (to escape slavery for instance), and sought refuge on the fringes of the civilised world in remote yet well-positioned bases and outposts.

The Romans sometimes even perpetrated acts which were virtually indistinguishable from piracy except for the fact that they were indeed carried out by members of the Roman navy, and were thus legitimised and could be said to be examples of state-sponsored piracy. For example, in 217 BC, a Roman fleet commander extorted ten talents from the islanders of Kerkina (Polybius 3.96.12). If the defenceless inhabitants had not paid, then they would have been at the mercy of the Roman ships. This was undoubtedly unabashed piracy and clearly illustrates that the Romans, as well as the Rhodians, adopted the pirates’ ‘raid mentality’ (Gabrielsen 2001: 224) and used it to their own advantage.

For all their faults, pirates were major engines of the slave trade, especially during the Hellenistic Period when Rome was in sore need of forced labour, and as such were constructive contributors to the mainstream economy. Pirates also, as can be witnessed in countless examples provided by ancient authors, acted as mercenaries in the service of commanders throughout Antiquity, sometimes after having been given license to privateer. Piratical tactics and ship designs were also hastily adopted by supposed anti-pirate states such as Rhodes and Rome, which, at

Another group, the Illyrians, wholeheartedly adopted the ‘raid mentality’ and practiced piracy on a wide scale. The Illyrians possessed a type of ship known as a Λέμβος (Lembus) (skiff) which was light and fast, perfectly suited for piracy, and it was even eventually copied by some naval powers such as the Romans. During the reign of King Agron, and later during that of his wife Queen Teuta 104

Piracy in the Hellenistic Period the same time, used real or perceived pirate threats as an excuse to exert their dominance ever further afield. Piracy in Antiquity, as it turns out, perhaps was not always the dire threat we so often see it as.

Hesperia: The Journal of the American School of Classical Studies at Athens. American School of Classical Studies at Athens: Athens. Pp. 513–27. GABBERT, J. 1986. ‘Piracy in the Early Hellenistic Period: A Career Open to Talents’ Greece and Rome. Cambridge University Press: Cambridge. Pp. 156–63.

Works cited

GABRIELSEN, V. 2001. ‘Economic Activity, Maritime Trade and Piracy in the Hellenistic Aegean’ Revue des Études Anciennes. Bordeaux University: Bordeaux. Pp. 219–40.

Primary sources APPIAN. Appian’s Roman History. Trans. White, H. A. 1962. Cambridge University Press: Cambridge. ARISTOTLE. Aristotle’s Politics Trans. Apostle, H. G. and Gerson, L. P. 1986. Peripatetic Press: Grinnell.

GABRIELSEN, V. 2013. ‘Warfare, Statehood and Piracy in the Greek World’ Seeraub im Mittelmeerraum. Wilhelm Fink: Paderborn. Pp. 133–53.

AUGUSTUS. Res Gestae divi Augusti. Trans. Cooley, A. 2009. Cambridge University Press: Cambridge.

GARLAN, Y. 1988. Slavery in Ancient Greece. Cornell University Press: Ithaca.

DEMOSTHENES. Demosthenes. Trans. Vince, J.H. 1958. Heinemann: London. LIVY. Ab Urbe Condita. Trans. Conway, R. S. and Walters, W. C. F. 1960. Clarendon Press: Oxford.

GILAN, A. 2013. ‘Pirates in the Mediterranean- A View from the Bronze Age.’ Mittelmeerstudien. Wilhelm Fink: Paderborn. Pp. 49–66.

PLUTARCH. The Lives of the Noble Grecians and Romans. Trans. Clough, A. H. 1938. J.M. Dent: London.

GRIFFITH, G. T. 1968. The Mercenaries of the Hellenistic World. Bouma’s Boekhuis N.V. Publishers: Groningen.

POLYAENUS. Polyaenus’ Stratagems of War. Trans. Shepherd, R. 1974. Ares: Chicago.

HADJIDAKI, E. 1988. ‘Preliminary Report of Excavations at the Harbor of Phalasarna in West Crete.’. American Journal of Archaeology. Archaeological Institute of America: Boston. Pp. 463–79.

POLYBIUS. Polybius. Trans. Walbank, F. 1972. University of California Press: Berkeley.

HOBSBAWM, E. J. 1969. Bandits. The Trinity Press: Worcester.

SICULUS, D. Diodorus of Sicily. Trans. Oldfather, C. H. 1963. Heinemann: London.

HOFF, M. C., and TOWNSEND, R. F. 2013. Rough Cilicia: New Historical and Archaeological Approaches. Oxbow Books: Oxford.

STRABO. The Geography of Strabo. Trans. Jones, H. L. 1961. Heinemann: London. THUCYDIDES. Thucydides. Trans. Rusten, J. S. 2009. Oxford University Press: Oxford.

HORDEN, P., and PURCELL, N. 2000. The Corrupting Sea: A Study of Mediterranean History. Blackwell: Oxford.

Secondary sources

MANNING, J. G. 2010. The Last Pharaohs: Egypt Under the Ptolemies, 305–30 BC. Princeton University Press: Princeton.

AVIDOV, A. 1997. ‘Were the Cilicians a Nation of Pirates?’. Mediterranean Historical Review. Routledge: Tel Aviv. Pp. 5–55.

MORAN W. L. 1992. The Amarna Letters. The John Hopkins University Press: Baltimore.

BOESE, W. E. 1973. A Study of the Slave Trade and the Sources of Slaves in the Roman Republic and the Early Roman Empire. University of Washington: Seattle.

MURRAY, W. M. 2012. The Age of Titans: The Rise and Fall of the Great Hellenistic Navies. Oxford University Press: Oxford.

BROCK, R., CORNELL, J., HODKINSON, S., and OLIVER, G. J. 2000. The Sea in Antiquity. International Series 899. BAR Publishing: Oxford.

ORMEROD, H. 1978. Piracy in the Ancient World. Liverpool University Press: Liverpool.

DEL HOYO, T. N. 2014. ‘Roman and Pontic Intelligence Strategies: Politics and War in the Time of Mithradates VI’. War in History. Sage: Newbury Park. Pp. 401–21.

RAINEY, A. F. 2015. The El-Amarna Correspondence: A New Edition of the Cuneiform Letters from the Site of El-Amarna Based on Collations of All Extant Tablets. Brill: Boston.

FANTALKIN, A. and LYTLE, E. 2016. ‘Alcaeus and Antimenidas: Reassessing the Evidence for Greek Mercenaries in the Neo-Babylonian Army’. Klio: Beiträge zur Alten Geschichte. De Gruyter: Berlin. Pp. 90–117.

RAUH, N., DILLON M. J., and MCCLAIN T. D. 2008. ‘Ochlos Nautikos: Leisure Culture and the Underclass Discontent in the Roman Maritime World’. The Maritime World of Ancient Rome. The University of Michigan Press: Ann Arbor.

FROST, F. J. and HADJIDAKI, E. 1990. ‘Excavations at the Harbor of Phalasarna in Crete: The 1988 Season.’. 105

Joseph M. Pacheco Jr SHAW, B. D. 1984. ‘Bandits in the Roman Empire’. Past & Present. Oxford University Press: Oxford. Pp. 3–52. VARINLIOĞLU, G., KAYE, N., JONES, M. R., INGRAM, R., and RAUH, N. K. 2017. ‘The 2016 Dana Island Survey: Investigation of an Island Harbor in Ancient Rough Cilicia by the Boğsak Archaeological Survey’. Near Eastern Archaeology. American Schools of Oriental Research: Chicago. Pp. 50–59. VIDAL, J. 2008. ‘The Men of Arwad, Mercenaries of the Sea’. Bibliotheca Orientalis. The Netherlands Institute for the Near East: Leiden. Pp. 5–15.

106

10 Investigating the Role of the Sea in Roman Crete’s Prosperity: A Maritime Archaeological Perspective Dimitris Karampas University of Oxford [email protected] Abstract: Crete came under Roman rule in the year 67 BC, and remained under Roman supervision for at least four centuries. The island’s strategic location made Crete an important province for Rome, while offering Crete a unique role in eastern Mediterranean trade and in maritime connections. As a result, the island’s participation in the Roman trade network of the eastern Mediterranean offered Cretans a range of economic and social opportunities. Certainly the island’s history goes hand in hand with the seafaring and seaborne trade, which is shown via the numerous coastal sites and shipwrecks found along the Cretan shores. These, in combination with terrestrial evidence, show some great changes in the local economy, architecture, technology and perhaps even in the social structure. Furthermore, the overall growth of the coastal centres during Roman occupation, which can be seen in the archaeological record, and the abundance of Cretan amphorae discovered in various regions around the Mediterranean, constitute further evidences of Crete’s trade influence. This paper examines the development that occurred in Crete under Roman rule from a maritime perspective, while scrutinising the representation of Roman Crete’s maritime character. Key words: Roman Crete, Roman Empire, shipwrecks, ports, eastern Mediterranean

Introduction

activities in the Roman coastal centres and the power of trade in the economic and social progress of Crete under Roman rule.

The study of Roman archaeology in Crete has shed light on a variety of questions, but there are still some aspects, such as the rapid development that occurred on the island during the Roman era, that need to be addressed. As an island, Crete was connected with the rest of the Roman Empire through seafaring. From the Neolithic period to Roman occupation, the island’s history has been intertwined with the sea. Ships have been a critical asset to Cretan society, especially during the Roman occupation, when Crete had its unique role in Mare Nostrum. Archaeological evidence provides researchers with a range of information about the role of the sea in the region’s commercial expansion. The plethora of Roman sites across the coasts, along with a number of shipwrecks around the island (most of which have been discovered in the past ten years), constitute significant material for further research in the fields of Roman history and archaeology in Crete. Through the combination of new data alongside with previously known archaeological evidence, a new perspective for Roman Crete is born. This paper aims to present the island’s maritime character, and the key role it played in Crete’s overall development during the Imperial Period. Archaeological evidence such as the plethora of coastal Roman sites and shipwrecks scattered around the island, have contributed to gleaning a new light on Roman Crete. Below, I will demonstrate the influence of seaborne

Previous studies The field of Roman archaeology in Crete is understudied, with several questions from this period remaining unanswered (Francis 2016: 3). Besides the site of Gortyn (Francis 2016: 4), which has been regularly excavated by the Scuola Archeologica Italiana di Atene, the surveys conducted in Knossos by the British School at Athens (Francis 2016: 4), and the excavation of Phalasarna’s port (Hadjidaki 1988), there is still a number of Cretan Roman antiquities that have not been examined properly. Only after Sander’s thesis publication in 1982 did more scholars approach Roman archaeology in Crete (Baldwin Bowsky 1983; Marangou-Lerat 1991). Several other studies appeared in the following years, concentrating primarily on the Romanisation of the island and the connection between Roman Crete with the Hellenistic Period (Harrison 1993). Within the last two decades, the edited volumes ‘From Minoan Farmers to Roman Traders’ (Chaniotis 1999) and ‘Roman Crete: New Perspectives’ have shed further light on the history and archaeology of Roman Crete (Francis and Kouremenos 2016). However, no systematic research regarding Crete’s maritime character during the Roman period has ever been conducted. 107

Dimitris Karampas Historical background

Interestingly, the Roman era in Crete saw major development along the coast. The majority of coastal sites replaced the traditional inner centres, due to the new circumstances. With the end of piracy, the coasts were safer. The sea was no longer seen as something to be feared, but as a mechanism of economic development and expansion. The open sea became a medium of connection rather than a border. It is highly probably that this was the main reason people moved closer to the coast. The end of piracy enabled Crete to participate in the Imperial trade, whose sea routes passed along the southern coast of the island. This development led to the creation of new occupations. Moreover, foreign merchants began arriving in Crete. The peace that was established in the Mediterranean during the Imperial period was instrumental for the Imperial trade’s prosperity.

During the first century BC, the Romans occupied a large number of territories in the Mediterranean basin. Crete became one of these territories, coming under Roman rule in the year 67 BC. The island united with the province of Cyrenaica around the year 27 BC (Chevrolier 2016: 15), and remained a single province for at least three centuries (Figure 1). The Romans acknowledged the benefits of controlling Crete. Firstly, it strengthened their war against piracy. It is well known that during the Hellenistic period, Cretan harbours and coastal sites hosted pirates from Eastern Mediterranean regions, particularly from Cilicia (Chevrolier 2016: 15; De Souza 1999: 59). The piracy caused an intense disruption in Rome’s trade with the eastern Mediterranean. By taking over Crete, Rome eradicated piracy in a great measure. Moreover, the Cretan agriculture and general economic potential made it a desirable conquest (De Souza 1998: 113). The Roman economy suffered a lot during the Mithridatic Wars and new sources of income were more than welcome in Rome. Another benefit of occupying Crete was the strategic location that the island offered. Crete acted as a natural border between the Aegean and the Libyan seas, while simultaneously acting as a gateway between the eastern and western Mediterranean; allowing a range of economic, social and military advantages.

The Imperial trade and the role of Crete The strategic location of Crete offered Romans a great deal of options for trade. The Cretan passage comprised the quickest voyage for merchant ships travelling to and from Rome to the East. The sea route along the southern coasts of Crete was the most popular choice among the ship-owners and merchants, since it served the needs of trade between the two largest centres of the period: Rome and Alexandria (Rickman 1980: 129, 131). However,

Figure 1. The province of Crete and Cyrenaica.

108

Investigating the Role of the Sea in Roman Crete’s Prosperity Crete’s southern route was probably well known to ancient sailors long before the Roman invasion (Broodbank 2013: 35). Further information about the southern trade route in Crete is provided in a passage from Paul in his Acts of the Apostles 27–8, where he describes the voyage he took from Alexandria to Rome. According to Paul, the Alexandrian ship on which he was aboard passed via southern Crete since it was a quick route to reach Rome (Davis 2009: 22; Rickman 1980: 131). Paul also describes some of the Cretan places he came across, like the Salmonium, and the refuelling stops that the ship made, including the coastal site of Kali Limenes, south Heraklion, Crete. Furthermore, he mentions a number of winds which seem to be wellknown among ancient sailors, and how they affected their journey.

merchant ships. Moreover, an interesting type of religious structure was built along Cretan coasts, which Sanders refers to as ‘coastal temples’. These include the Diktynna Sanctuary on the north side of the Rodopou peninsula, Athena’s sanctuary at Samnonium Point, and the temple of Vritomartis on the Kastri at Chersonissos (Sanders 1982: 40). Due to their unique locations, these coastal temples might have been operated as navigation points, similar to lighthouses, for passing ships (Sanders 1982: 40). Cretan Products: The Cretan Amphora Meanwhile, Cretan agriculture was booming. Among the most desired Cretan exports was the local wine. The ‘Cretan passum’ was well known all over the Empire and references to its wine quality and fame are made by various historical sources, including Polybius (MarangouLerat 1999: 270), and the Expositio Totius Mundi et Gentium, the most well-known geographical book from the Imperial period (Chaniotis 2008: 89). There is even an inscription discovered in Pompey, which characterises the ‘Cretan passum’ as ‘excellent Cretan wine’ (CIL IV 5526). Cretan amphorae, which were mainly used for the wine’s packaging and transportation (Marangou-Lerat 1995: 129), have been discovered in many regions across Europe and North Africa. Six different types of Cretan amphorae are known to have been manufactured during the periods discussed. Most of the examples date to between the first and third century AD (Marangou-Lerat 1999: 275). The Cretan amphora AC1 is considered the most common, with an overall capacity of around 20 litres (MarangouLerat 1999: 276). The AC1 was in use for at least three centuries. They have also been discovered in most of the Cretan amphora workshops.

Until the first century AD, only the southern half of the island seems to have flourished from the Imperial trade. The north coast of Crete became more economically active during the first and second centuries AD with the growth of the northern coastal centres (Sweetman 2013: 148). As a result, foreign merchant ships started sailing more frequently along the north coast because ships were now able to make restocking stops and trade with the north coastal sites. A piece of papyrus (P.Mich. inv. 5760a) discovered in Egypt in 1929 by the University of Michigan provided scholars with information about a number of possible sea routes in the eastern Mediterranean, including the exchanges conducted along the northern coasts of Crete. The papyrus describes the sea routes followed by a Roman merchant ship in order to reach Alexandria. Among the areas where the ship stopped was the Cretan coastal site of Attalion. Further details about the types and the quantities of products traded are also provided in this papyrus (Hailporn 2000: 340). The terminus post quem of the papyrus is 117 AD (Hailporn 2000: 341).

In Rome, large quantities of Cretan amphorae were imported, consisting of more than 10 per cent of the overall wine vessels that arrived in Rome during the Flavian Dynasty (69–96 AD) (Rizzo 2003: 205–07). In Pompey, among the hundreds of Cretan amphorae discovered, 41 of them carried the inscription ‘Lyttios’ (Marangou-Lerat 1995: 130–54). This discovery constitutes the biggest group of inscribed amphorae ever discovered in the region (Chaniotis 2008: 90). Cretan amphorae have also been discovered in shipwrecks hundreds of miles away from Crete (Gallimore 2016: 178). The Cretan amphora is tied with the development that occurred along the coasts, where they were exported from the Cretan ports. Due to the high demand for the wine, large numbers of amphorae workshops were located only a short distance from the shores (Marangou-Lerat 1999: 274). Presumably this made it easier to pack them into the ships.

Overall development Many changes occurred in the island’s economic and social structure during Roman occupation, which upgraded Crete’s role in Imperial trade. Meanwhile, even the Cretan landscape changed, forcing some scholars to adopt the term ‘Romanisation’ in order to explain the Roman influence in the region. The Roman occupation modified Crete, which gave the island and its people a new cultural identity. Furthermore, roads were built which improved connections to the inner parts of the island, facilitating the movement of goods (Chaniotis 2013: 63). Once known for its furious pirates, its mercenary archers, and, of course, its glorious Minoan past, Crete transformed into a commercial Roman base, and a stepping stone for the Imperial trade.

Other Exports

Roman constructions like aqueducts, baths, cisterns, temples and theatres were erected all over the island (Sanders 1982: 57), particularly in the coastal centres. Port facilities were created in order to serve the growing trade. Marine constructions such as breakwaters, piers and moles were made to host both local and foreign

Crete was not just a land of vineyards. Olive trees were visible all over the island, just as they can be seen today in most Mediterranean regions. Wine and olive oil were the key exports of Crete, but not the only ones. Goods like the Cretan palms, quinces and the large Cypriot figs were 109

Dimitris Karampas also produced and traded during the Roman Era (Sanders 1982: 32). In addition, according to Athenaeus, Cretan herbs appear to have been well-known all over the Empire (Raab, 2001: 13; Sanders 1982: 32). In some Roman texts, there are even references to Cretan honey and wax (Francis 2016: 84–90). Another profitable activity on the island was the Porfira processing, namely, the extraction of purple dye from murex shells (Papadakis 1997: 11–12). Evidence for this activity comes mainly from the eastern half of the island. Other commercial activities in Crete include a number of whetstone quarries discovered along the northeast coast (Harrison 1993: 190). Pliny mentions the exportation of the Cretan whetstone to the Roman Republic during the Hellenistic years (Sanders 1982: 18). Equally important was the production and exportation of timber. According to Strabo, Crete must have been a wellwooded island in Antiquity (10.4.4) and until Medieval times, when the Venetians deforested most of the island for the construction of ships (Meiggs 1982: 33). In 1415 AD, the traveller Bueldomondi arrived in Crete and saw an island full of cypress, from the mountains to the coasts. Pliny and Vitruvius also referred to cypress and Cretan cedar as a popular Cretan export, which was probably used for ship construction purposes. Pliny considered Crete as the greatest source of cypress in the Mediterranean (Meiggs 1982: 33).

number of veterans (Harrison 1993: 182). Romans were likely the first to get involved in the trade by exporting Cretan products. This not only proven by the plethora of epigraphic evidence referring to Roman names, but can also be seen in the fact that Roman names are inscribed in Cretan amphorae that have been discovered outside Crete. This indicates that the producer or the merchant was probably Roman (Baldin Bowsky 1999: 318). Cretans became active in the Imperial trade and in other economic activities later on by adopting Roman names themselves (Gallimore 2009: 150). Archaeological evidence The series of modifications and improvements in Cretan society was greatly influenced by maritime activities occurring on the island during the Roman period. Ports and ships had a critical role in the development and improvement of the economic network as described above, a role that has not been properly examined in previous studies. Harbour remains and shipwrecks contain critical archaeological evidence, since they constitute a tangible image of past maritime events. Marine facilities on the Cretan coasts Many natural anchorages were active prior to Roman occupation. There is evidence that during the Imperial times, marine facilities were constructed all over Crete in order to serve foreign and local ships. These installations were vital in covering the trade’s needs. It is not a coincidence that the most advanced facilities were found in areas with large coastal centres and on critical maritime highways. Examples can be seen on both the northern and southern coasts of Crete. The harbour remains in the sites of Chersonisos, Kissamos and Matala constitute great examples of the Roman influence on the Cretan coasts. Especially the remains in Chersonisos, which provide researchers with plenty of information about the marine establishments of the period.

Social changes The Romans changed the way of life on Crete, and, subsequently, the social structure of the Cretan society was also affected. A rising wealthy oligarchy appears to have replaced the aristocratic families from the Hellenistic period (Chaniotis 2008: 91). Roman rule, and the series of changes that altered the foundations of Cretan society, led to the reorganisation of social hierarchy. That was the result of the Imperial trade’s impact and the increased movement of foreign populations to Crete. Moreover, occupations like merchants and craftsmen became crucial to the new flourishing economy. These changes boosted the local economy, since more people were now involved in jobs related to the Imperial trade and sea activities (Chaniotis 2008: 86). In addition, coastal development was enhanced by the participation of foreigners who interacted with the Cretan trade.

The Roman port in Chersonisos constitutes the most characteristic marine construction of the north coast. The port was 270 metres long and 150 meters wide, and its entrance was 90 metres wide (Harrison 1993: 193). Furthermore, the harbour’s facilities were such that it could have moored at least 30 ships (Leatham and Hood 1958– 9: 266–73). Chersonisos had a critical role in the Cretan exports, and those packaged in the Cretan amphora. It is not a coincidence that close to the port, several amphora workshops have been excavated (Marangou-Lerat 1999: 272). The port of Chersonisos had strong ties with the city of Lyttos as described by Strabo ‘Λύττου δέ, ἧς ἐμνήσθημεν καὶ πρότερον, ἐπίνειόν ἐστιν ἡ λεγομένη Χερρόνησος’ (As for Lyttos, which we have already mentioned before, its port is Chersonisos) (10.4.14). Lyttos, which was one of the biggest centres in Crete since the Hellenistic period, was famous for its wine production (Figure 2) and used to export its products via the Chersonisos port. Furthermore, one of the few known Aqueducts from the Roman period,

Most of the foreigners who participated in the economic activities in Crete were Romans (Hadjisavvas 2012: 169). Sources of their presence in Crete is provided by the study of the available epigraphic record from around the island. Furthermore, the appearance of the latifundium (Chaniotis 2008: 87) and the familiaries (Baldin Bowsky 1999: 307) provide solid evidence of Roman immigrants in Crete. Latifundium was an extensive land owned by a wealthy person dedicated to the mass production of agriculture products, while the term familiaries was used to describe an area owned by one person (the paterfamilias), who was in charge of a group of people who worked and served in his fields. Romans and their families arrived in the colony erected in Knossos, while Cretan land was given to a 110

Investigating the Role of the Sea in Roman Crete’s Prosperity

Figure 2. The area around ancient Lyttos.

Figure 3. Port of Chersonisos main quay.

111

Dimitris Karampas Phalasarna’s destruction by the Romans in the years 68– 66 BC (Hadjidaki 1988: 513), and its port’s abandonment during the first century AD (Baika 2011: 42), Kissamos turned into the leading port of western Crete. Evidence of the large-scale trading at Kissamos is strengthened by the number of shipwrecks that have recently been discovered in the Kissamos Gulf (Theodoulou 2014: 46).

whose remains still stand, connected Lyttos with the port in Chersonisos. Further evidence about Chersonisos’ commercial activity comes from the study of 92 inscribed ostraca, which revealed the plethora of products shipped out of Chersonisos. The ostraca were dated to the second half of the second century AD, and the first half of the third century AD (Litinas 2008: 34). The Chersonissos port consisted of the main quay, a breakwater and three moles made out of concrete and rubble (Leatham and Hood 1955: 267). The quay is still visible from the modern port (Figure 3), while the breakwater and the southeast mole lie underwater, and the northeast moles are probably located below the modern port. A number of fish tanks are also located close to the roman port, which indicate yet another profitable activity for the region (Davaras 1974: 87).

Gortyn’s official ports were located in the area of Matala and the area of Lebena (Baldin Bowsky 1999: 336). The two regions were probably chosen by the Romans because of their proximity to the capital, and because both of the sites had been used as harbours in earlier periods. In Matala, Roman influence can easily be seen by the curved tombs in the northern part of the bay. Matala’s ancient port facilities were also described by Francesco Barrozi, an explorer who first visited the area in the year 1577 AD (Kaklamanis 2004: 80). Today, the unique evidence of port remains in Matala can been seen along the southern part of the bay (Figure 6). A small path leads into a site which indicates a possible part of a marine construction (Figure 7). A slipway located in the southeast of Matala was surveyed by Blackman in 1970 and 1971 (Blackman et al. 2013: 389). Further evidence of the maritime character of Matala during the Roman period comes from three caches of Roman amphorae discovered close to the bay representing three distinct shipwrecks (Crile 1962: 48). The first impression of Matala’s facilities is that they are

The strategic location of Kissamos made the region quite successful in Roman times (Figure 4). Straddling two peninsulas, the Gramvousa and Rodopou, Kissamos enjoyed protection from a fortified landscape and the privilege of controlling the trade with mainland Greece. Kissamos constitutes a great example of the coastal development, since a variety of Roman constructions existed in the area, among which were a theatre, a bath, various mosaics and a number of wealthy residences (Paton 1999: 286–87). The city’s harbour must have been located in the area of ‘Mavros Molos’, where the Roman breakwater still stands today (Figure 5). After

Figure 4. The gulf of Kissamos.

112

Investigating the Role of the Sea in Roman Crete’s Prosperity

Figure 5. Kissamos’ Roman breakwater ‘Mavros Molos’ from the NW.

Figure 6. The coastal site of Matala.

113

Dimitris Karampas

Figure 7. Possible port facilities in Matala located on the southwest side of the bay.

quite small for serving such a big city like Gortyn. Ian Sanders first suggested this (1982: 23), stating that there might be a chance that the official port might be located somewhere else, proposing somewhere in the Gulf of Messara.

Most of the wrecks are located in sites not far away from the large commercial centres, where ships restocked goods. Additionally, the fact that there are shipwrecks scattered all around the island indicates that Crete served and had relationships with different geographical regions across the Mediterranean or at least active trading along the island’s coasts. Future research will shed light on information regarding the purposes and the identity of the ships, and about seafaring and trade in the territory in general. Below, case studies of three different shipwrecks are presented, with the aim of highlighting the importance of Cretan waters in the Imperial trade.

Shipwrecks in the Cretan Seas A number of shipwrecks lying in the Cretan seas further supports the important commercial role of the island during the Roman Period. The number of Roman shipwrecks lying on the seabed of the Mediterranean comprises the largest amount, compared with the number of known wrecks from other periods (Parker 1992; Wilson 2011: 35). Furthermore, most of the shipwrecks date to the first century AD (Wilson 2011: 35), indicating how busy the sea routes were at that time, the critical role of trade and seafaring for societies and, of course, the unique role that ships had in the prosperity and expansion of the Roman Empire.

The shipwreck of Agia Galini The wreck of Agia-Galini was first discovered in 1937, 2 km east of the village of Agia-Galini in Rethimno, Crete (Figure 9). Part of the wreck’s cargo was accidentally discovered by two local fishermen. The site is located 200 metres offshore, at a depth of less than 5 metres. The artefacts were discovered in a radius of about 50–60 metres. Many bronze objects have been recovered, among them a number of statuettes that resemble local gods (Platon 1960: 508). No amphora remains were recovered. During the late 1950s, the search conducted in the area brought to the surface a total of 259 Roman coins dating from the time of Augustus (first century BC), with later examples portraying emperor Probus (276–282 AD). No remains of the hull have been recovered. The wreck is believed to have served as a recycling vessel, which

More than ten underwater sites identified as possible Roman shipwrecks have been discovered in Cretan waters (Figure 8). The large quantity of amphorae discovered in their cargoes have merchant stamps which can determine their origins. Furthermore, the presence of Cretan amphorae in most of the sites shows an active trade along the Cretan coasts (Theodoulou 2014). As previously discussed, Cretan amphorae present the strongest evidence for Roman Crete’s trading activity within the Empire. 114

Investigating the Role of the Sea in Roman Crete’s Prosperity

Figure 8. Roman shipwrecks around Crete.

have sunk while trying to moor in the bay, probably after a crash into the cape’s rocks (Theodoulou et al. 2015: 619). Part of the cargo likely remains buried in the seabed. The site dates probably to the first to second centuries AD.

gathered old bronze items from the coasts. Scholars came to this conclusion as most of the artefacts recovered were in poor condition (Brokalakis 2016: 133; Theodoulou, 2014: 32). Objects similar in style to those discovered in the Agia Galini wreck have been found in different regions around the Mediterranean (Brokalakis 2016: 129). The wreck further indicates the trade connections that existed within the Roman Empire. Moreover, some of the artefacts recovered appear to have Gothic origins, which can be connected with the Gothic presence in the Aegean and the Cretan waters in 269 AD (Wolfram 1990: 53).

The shipwreck of Dionisades Islets On the east side of the island, close to the islets of Dionisades, a shipwreck, probably from the first or second centuries AD, has been surveyed by the Ephorate of Underwater Antiquities (Preka-Alexandri et al. 2012: 104). The cargo contained two types of amphorae: the Cretan amphora AC2, comprising the majority of the assemblage, and a type is similar to the Tripolitanian I type from Tripolitania, Libya. This type of amphora dates to roughly the first and second centuries AD (PrekaAlexandri et al. 2012: 104). The ship’s cargo was scattered in a radius of 200 metres. A likely hypothesis is that the vessel crashed in a reef, which smashed the hull of the vessel, thus the cargo flowed into the sea, moments before the ship sank (Preka-Alexandri et al. 2012: 103). Part of the cargo probably still remains buried in the seabed (Preka-Alexandri et al. 2012: 104). The presence of the African amphorae indicates a merchant identity for the ship, while simultaneously confirming the use of the north coasts of Crete during the late Roman Period.

The shipwreck in the Gulf of Herakleion The second wreck was discovered in the Gulf of Herakleion in 2011. The site was located through research in the region by the Greek Centre of Underwater Research and the Woods Hole Oceanographic Institution (Theodoulou 2014: 44). The wreckage was discovered close to the western cape of the Agios-Georgios Bay. The fragments recovered from the seabed are similar to the Cretan amphora AC4. Furthermore, some of the intact amphora bodies, which remain on the seabed, resemble Cretan ones. The type of the amphorae indicate that they might had been produced in amphora workshops from Herakleion (Marangou-Lerat 1999: 272). According to the researchers, the ship must 115

Dimitris Karampas

Figure 9. Rock wall on the southern coast of Crete, close to Agia-Galini.

Conclusions

next centuries to come, with the most obvious alteration being the fact that the administrative centres of Crete remain on the coast (even to this day), while the interior’s centres diminished. It is not a coincidence that most of the Roman wrecks were discovered in areas very close to commercial centres, where ship transactions with the island occurred frequently. Furthermore, the island’s location allowed for seafaring to a number of destinations including the Aegean and Greece, the North African coasts and the provinces of the eastern Mediterranean.

This paper has attempted to present the key elements in the relationship between Roman Crete and the sea, and the impact of this on the island’s cultural, economic and social identity. The purpose of the examination of Roman Crete’s maritime character, and the island’s general relationship with the sea aims to justify the presence of sea and seaborne activities as crucial mechanisms in Crete’s overall development and prosperity during the Imperial times.

Keeping in mind how important the Cretan waters were at this point, in combination with the coastal development that Crete underwent, the great number of Cretan amphorae exportations and the numerous shipwrecks, an important question arises concerning seafaring in the eastern Mediterranean. Could the Cretans have produced their own vessels during the Roman Period? The marine facilities that were constructed, the increased number in exportations, the huge public works erected all over the island, in addition to the large wood processing and the cypress exports (Meiggs 1982: 33), all lead to the inquiry of whether Crete could not have produced its own ships and boats, in order to cover the needs of the trade. Crete had the products, the means of packaging, the facilities to export them and the material to construct the mean of transportation. The only missing factor is a means of distribution. Future research might provide researchers with answers to this challenging query that might greatly affect the overall thinking about Roman Crete.

Moreover, the progress which occurred in the coastal sites on the north and south shores of the island was much greater compared to the interior of Crete. The number of Roman shipwrecks located in the Cretan seabed, and the range of port facilities discovered in the Cretan shores, constitute the main evidences in this research. To a great measure, the transactions occurring, and especially the great range of Cretan exports, were the main instruments in the development of the Cretan economy. People’s wellbeing and prosperity came from the sea, and more specifically via the safe Imperial trade routes. During the Roman Period, ports and coastal sites became the most important areas for Crete and surpassed the traditional powers from the Hellenistic Period. More specifically, the alterations that occurred in the Cretan economy and society during Roman occupation created a new identity for Crete and its people, while shaping the Cretan landscape for the 116

Investigating the Role of the Sea in Roman Crete’s Prosperity Future research

BUELDOMONDI, C. 1983. Ένας Γύρος της Κρήτης στα 1415: Περιγραφή της Νήσου Κρήτης. Translated by Aposkiti. Σύλλογος Πολιτιστικής Αναπτύξεως Ηρακλείου: Herakleion.

A more in-depth study in the trade relationships and contacts between Roman Crete and other Roman provinces from the eastern Mediterranean, and especially from the Greek mainland, the Aegean and Asia Minor, would be useful. Further archaeological research should be applied to coastal sites, their port facilities remains and the available wrecks. Moreover, geospatial research in areas with high archaeological potential is also suggested. Additional studies in the epigraphic record from the Roman period can shed further light on the social impact of Roman occupation, while simultaneously providing a clearer structure of the Cretan society in the Roman Period, and how it was formed compared to the Hellenistic Period. Additionally, a stronger theoretical background concerning the economic maritime activities of the island should be applied in order to understand the nature of the transactions that occurred on Cretan shores, including taxation and administration.

CHANIOTIS, A. 2008. ‘What difference did Rome Make? The Cretans and the Roman Empire’. In: B. Forsen and G. Salmeri (eds.), The Province strikes back: Imperial Dynamics in the Eastern Mediterranean. Papers and Monographs of the Finnish Institute at Athens 13. Foundation of the Finnish Institute at Athens: Helsinki. Pp. 83–106. CHANIOTIS, A. 2013. ‘Hadrian, Diktynna, the Cretan Koinon, and the Roads of Crete: A New Milestone from Faneromeni (Crete)’. In W. E. B. Fehér and P. Kovác (eds.), Studia Epigraphica in memoriam Géza Alföldy. Dr. Rudolf Habelt Gmbh: Bonn. Pp. 59–68. CHEVROLLIER, F. 2016. ‘From Cyrene to Gortyn. Notes on the relationship between Crete and Cyrenaica under Roman Domination (1st century BC – 4th century BC)’. In J. N. Francis and A. Kouremenos (eds.), Roman Crete: New Perspectives. Oxbow Books: Oxford. Pp. 11–26.

Furthermore, the connection and relationship between Crete and Cyrene comprises a very interesting and challenging question. Unfortunately, the archaeological evidences are not enough in order to provide researchers with reliable answers about the relationship between the two regions as a single province.

DAVARAS, C. 1974. ‘Rock cut fish tanks in eastern Crete’. Annual of the British School at Athens 69: 87–93. DAVIS, D. L. 2009. Commercial Navigation in the Greek and Roman World. The University of Texas at Austin: Texas.

The maritime character of Crete played a key role in its transformation during the Roman period. Of course, that was not something new for the island. Dating back to the Bronze Age, sea activities offered much in the prosperity of island. The difference is that during the Imperial Period, Crete showed its importance in a much greater and more competitive network, taking advantage of the island’s unique location and local production. As a result, the new circumstances irreversibly changed the island’s identity, which led future generations into an inextricable bond with the sea.

DE SOUZA, P. 1998. ‘Late Hellenistic Crete and the Roman Conquest’. Ιn W. G. Cavanagh, M. Courtis, J. N. Coldstream, and A. W. Johnston (eds.), Post Minoan Crete. British School at Athens: London. Pp. 112–16. DE SOUZA, P. 1999. Piracy in the Greco – Roman World. Cambridge University Press: Cambridge. FRANCIS, J. 2016. ‘Apiculture in Roman Crete’. In J. N. Francis and A. Kouremenos (eds.), Roman Crete: New Perspectives. Oxbow Books: Oxford. Pp. 83–100.

Works cited

GALLIMORE, S. 2016. ‘Crete’s Economic Transformation in the Late Roman Empire’. In J. N. Francis and A. Kouremenos (eds.), Roman Crete: New Perspectives. Oxbow Books: Oxford. Pp. 175–88.

BLACKMAN, D. J., RANKOV, B. 2013. Shipsheds of the Ancient Mediterranean. Cambridge University Press: New York. BOWSKY, B. M. W. 1999. ‘The Business of Being Roman: The Prosopographical Evidence’. In A. Chaniotis (ed.), From Minoan Farmers to Roman Traders: Sidelights of the Economy of Ancient Crete. Franz Steiner Verlag Stuttgart, Harrison: Stuttgart. Pp. 305–47.

GRILE, G. and DAVARAS, C. 1964. ‘The possible site of Menelaus’ shipwrecks’. Κρητικά Χρονικά 17. Pp. 47–49. HADHIDAKI, E. 1988. ‘Preliminary Report of Excavations at the Harbor of Phalasarna in West Crete’. American Journal of Archaeology 92.4: 463–79.

BROKALAKIS, G. 2016. ‘The use of bronze on Roman ships and the shipwreck of Agia Galini’. In A. GiumliaMair and C. C. Mattusch (eds.), Proceedings of the XVIIth International Congress on Ancient Bronzes. Editions Mergoil: Izmir. Pp. 127–39.

HADJISAVVAS, S. and CHANIOTIS, A. 2012. ‘Wine and olive oil in Crete and Cyprus: Socio-economic Aspects’. In G. Cadogan, M. Iacovou, K. Kopaka, and J. Whitley (eds.), Parallel Lives: Ancient Island Societies in Crete and Cyprus. British School at Athens Studies: London. Pp. 157–73.

BROODBANK, C. 2013. The Making of the Middle Sea: A History of the Mediterranean from the Beginning to the Emergence of the Classical World. Thames & Hudson Ltd: London. 117

Dimitris Karampas University of Crete and 28η ΕΒΑ-ΑΙΚΣ: Rethimnon. Pp. 147–57.

HAILPORN, P. 2000. ‘Registre de Navires Marchands’. In R. Smet, H. Melaerts, and C. Saerens (eds), Papyri in Honorem Hohannis Bingen Octogenarii, P.bingen 77. Peeters Publishers: Leuven. Pp. 340–50.

THEODOULOU, T., FOLEY, B., EVAGGELISTIS, D., KOUTSOUFLAKIS, G., SAKELLARIOU, D., and TOURTAS A. 2013. ‘Crete Project 2011: Underwater archaeological survey at the area between Herakleion and Dia. A preliminary report’. In P. Karanastasi, A. Tsigounaki, and X. Tsigonaki (eds), Αρχαιο- λογικό Έργο στην Κρήτη III. Πανεπιστήμιο Κρήτης – ΚΕʹ ΕΠΚΑ – 28η ΕΒΑ: Rethimnon. Pp. 615–25.

KAKLAMANIS, A. 2004. Barozzi, Francesco 1537– 1604: Descrittione dell’isola di Creta, Heraklion: Βικελαία Δημοτική Βιβλιοθήκη. LEATHAM, J., and HOOD, S. 1958. ‘Sub-Marine Exploration in Crete, 1955’. Annual of the British School at Athens, 53/54. Pp. 263–80.

WILSON, A. 2011. ‘Developments in Mediterranean shipping and maritime trade from the Hellenistic period to AD 1000’. In A. Wilson and D. Robinson (eds.), Maritime Archaeology and Ancient Trade in Mediterranean. Oxford Centre for Maritime Archaeology, Institute of Archaeology: Oxford. Pp. 33–60.

LITINAS, N. 2008. Greek Ostraca from Chersonesos, Holzhausen Verlag Gmbh: Wien. MARANGOU-LERAT, A. 1995. Le vin et les amphores de Crète de l’époque classique à l’époque impériale. École Française d’ Athènes: Athens. MARANGOU-LERAT, A. 1999. Wine in the Cretan Economy. In A. Chaniotis (ed.), From Minoan Farmers to Roman Traders: Sidelights of the Economy of Ancient Crete. Franz Steiner Verlag Stuttgart, Harrison: Stuttgart. Pp. 269–79.

WOLFRAM, H. 1990. History of the Goths. University of California Press: California.

MEIGGS, R. 1982. Trees and Timber in the ancient Mediterranean world. Oxford University Press: New York. PARKER, A. J. 1992. Ancient Shipwrecks of the Mediterranean & Roman Provinces. International Series 580. BAR Publishing: Oxford. PLATON, N. and DAVARAS, K. 1960. ‘Αρχαιολογικήν Κίνησις εν Κρήτη το Έτος 1960’. Κρητικά Χρονικά 14. Pp. 504–27. PREKA-ALEXANDRI, K., THEODOULOU, T., ARGIRI, K., and MENTOYIANNIS, V. 2012. ‘Προκαταρτική έρευνα σε τρία άγνωστα ναυάγια στη βορειοανατολική ακτή της Κρήτης’. In M. Ανδριανάκης, Π. Βαρθαλίτου, and Ι. Τσαχίλη (eds.), Αρχαιολογικό Έργο Κρήτης 2. University of Crete and 28η ΕΒΑ-ΑΙΚΣ: Rethimnon. Pp. 103–11. RAAB, H. A. 2001. Rural Settlement in Hellenistic and Roman Crete: The Akrotiri Peninsula. International Series 984.BAR Publishing: Oxford. RICKMAN, G. 1980. The Corn Supply of Ancient Rome. Clarendon Express: Oxford. SANDERS, I. F. 1982. Roman Crete. Aris & Phillips Ltd: Warminster. SWEETMAN, R. J. 2013. The Mosaics of Roman Crete: Art, Archaeology and Social Change. Cambridge University Press: New York. THEODOULOU, T. 2014. ‘Underwater and coastal archaeological research in Crete during 2011–2013. Acts of the Department of Underwater Antiquities of Crete’. In P. Karanastasi, A. Tsigounaki, and X. Tsigonaki (eds), Αρχαιολογικό Έργο στην Κρήτη.

118

11 Cultural Underwater Heritage of the South Pacific Armada Cristina Agudo Rey University of Cadiz, Spain [email protected] Abstract: The history of the Viceroyalty of Peru and its naval defence is a key part of the underwater heritage of the South Pacific Ocean. The Viceroyalty navy is rarely recognised in history, but it is unique because of the ship construction which was influenced by the deep waters of the Pacific Ocean. These galleys and galleons were built not only for defending the Viceroyalty coasts, but as commercial ships as well. These armed ships were used to transport gold and silver and ultimately formed the South Pacific Armada. However, climate conditions caused several maintenance problems, along with multiple incursions and pirate raids, which resulted in a number of shipwrecks. Many of these Spanish ships are under the oceanic waters, which has kept them preserved because of climate conditions of the water and its low temperatures. This paper outlines a number of the vessels and provides some historical context in an attempt to further the conservation of maritime heritage in Peru. Keywords: South Pacific Ocean; Viceroyalty of Peru; South Pacific Armada; Spanish colonialism; underwater archaeology; shipwreck; shipwreck map

An introduction to the protection of underwater heritage in Peru

National Port Authority; companies involved in the construction sector in marine and coastal areas; offshore oil companies and underwater mining companies; and companies which operate in the dredging of ports, reservoir construction management, the use of lagoons, pools for aquaculture and shellfish farms, among others • There are conflicts of interest with the Peruvian Navy regarding underwater findings in the case of ship discoveries

In Peru, the protection of underwater cultural heritage is guaranteed by the General Law of Cultural Heritage. Underwater cultural heritage is defined as: All objects that have the importance, value and meaning referred to in Titles II and III of the first section of the Act, which are submerged under water, whether they be at sea within Peruvian territory, or in lakes and coastal areas, and other aquatic national territory, partly or wholly, periodically or continuously, for at least 50 years… (UNESCO 2004).

In this paper, I outline some of the main principles currently in place in the maritime cultural sector of Peru, present an early example of the preservation of maritime cultural heritage and give some suggestions for how to properly manage this site and future related sites. The examined data provides a good example of the richness of Peruvian maritime heritage, presenting a preliminary map of shipwrecks in the region which can help us to take some first steps towards proper conservation and management.

Although these guidelines generally protect underwater cultural heritage, in terms of the regulations, there is no effective form of control. Similarly, there is no mention or consideration of other factors that affect their protection: • There is no mention in the regulations of how to deal with underwater projects • The delimitation of these sites is not regulated • The Ministry of Culture does not have staff who are qualified in this area • The definition of heritage could be confusing in the sense that it includes the Pre-Hispanic, Colonial and Peruvian Republican eras • The institutions that are charged with managing maritime and underwater issues do not know the regulations and their scope. These institutions are the

Maritime archaeology is still a very new area of research. Up until 2006, the only existing investigations had been carried out by amateurs, treasure hunters and diving enthusiasts. Famous findings include the Goleta Covadonga in Chancay to the north of Lima, and the Navio San Martin shipwreck (a ship dating back to the seventeenth century) (Figure 1), discovered off the shores of Lima. All the discoveries have suffered from the improper recovery of objects and this has affected the integrity of the sites as well as the archaeological context. 119

Cristina Agudo Rey

Figure 1. Plan of a seventeenth-century galleon: A. Bow, B. Stern, C. Keel, D. Draft, E. Beam; 1. Bowsprit, 2. Foremast, 3. Main mast, 4. Mizzen mast, 5. Decks, 6. Bulky stern.

In the 2013 UNESCO conference, the Regulations of Archaeological Research set out the guidelines for the implementation of evaluation, research, monitoring, conservation, enhancement and museology projects, including all activities concerning movable, immovable, underwater and other related cultural heritage. These regulations specify that, for monitoring plans and chance findings, the preferred option for submerged items is in situ conservation.

national inventory of underwater heritage properties took place in September 2013. These activities were carried out in different departments in Lima, Arequipa and Puno. The sites were chosen due to the fact that they were non-inventoried coastal and underwater properties. The activities identified four submerged heritage sites, three in the sea and one in Lake Titicaca. Although Peru has not yet ratified the UNESCO Convention on the protection of its underwater cultural heritage, it has enacted a legal framework and active national management policy which takes into account that which was stipulated in the convention. This includes: the obligation to preserve underwater heritage as world heritage; the priority of preserving underwater heritage in situ; preventing the use of underwater heritage for commercial purposes; helping to develop archaeological activities using specialised archaeologists; and promoting international cooperation in the development of new academic and scientific projects (Villar Astigueta et al. 2015: 44).

The Peruvian State, through the Ministry of Culture and the General Heritage Division, has implemented a management policy that includes different priorities and among these is the taking of a national inventory of the nation’s underwater heritage. All inventories are an important asset because of the detailed knowledge they provide regarding the composition, location and status of heritage items, and because they can be used to design and implement management and conservation plans for the protection thereof. The first works seeking to make a 120

Cultural Underwater Heritage of the South Pacific Armada But why is it so difficult to develop academic/ maritime archaeological activities and projects in Peru? Unfortunately, there is no right answer to this question and the debate is still open. There are different theories in this regard, such as the fact that the country is rich in Pre-Columbus sites which attract the largest investments and the big projects, the Caribbean is attractive in terms of its treasures related to maritime archaeology, and this type of archaeological activity in the Pacific Ocean is quite costly, a problem that is further exacerbated by the lack of fundraising in the country.

European countries did not acknowledge the physical, human and military situation in the Viceroyalty. Even after the first British incursions, the Viceroyalty still believed that the widespread ignorance of the reality of the country was its best card, as can be seen in with a letter sent to the Spanish king in 1597 by the Viceroy Don Luis de Velasco: Toda la defensa de las Indias en general, consiste más en la ignorancia que en los enemigos tienen de las cosas particulares de ellas y en la contrariedad de la tierra y de los tiempos, que en las fuerzas que acá hay para resistirles (The entire Indies defensive system is based more on our enemies’ ignorance of particular aspects of the Indies and on its setbacks and the weather, rather than on the forces we have to defend it) (Bradley 1976: 79).

The South Pacific Armada in the Viceroyalty of Peru The Viceroyalty of Peru was a Spanish colonial administrative district which, during the seventeenth century, extended from Central America to Tierra del Fuego in Chile. The Viceroyalty of Peru district comprised most of the western coast and covered the modernday countries of Panama, Colombia, Ecuador, Peru and Chile. Its market and mineral trading made the city of Arica, formerly in the Peruvian Republic, one of the most important ports of its time, along with the Chilean city of Valparaiso, which contained the exporting port of silver from the Bolivian mine, Potosí. The Viceroyalty of Peru was the main colonial district from where Spain received most of its silver. In addition to safeguarding the Caribbean area and its gold market, keeping the route and the galleons safe from pirates and other countries at war with Spain was a crucial issue for the Spanish crown. The silver route from the Viceroyalty of Peru was a long journey. It started from the aforementioned Potosí, and the cargo was transported by sea from the port of Arica in Peru to Panama. In Panama, the silver was then transported by the main galleons sailing back to Spain from the Caribbean over the Atlantic Ocean.

However, this situation had altered drastically in 1578 due to the change in the political defence of the Viceroyalty itself. The British privateer, Sir Francis Drake, had managed to cross the Strait of Magellan. The continuous British incursions led to the idea of building various fortifications on both sides of the Strait, thus keeping the pass to the ‘Spanish lake’ closed. This task was given to a sailor, Sarmiento de Gamboa, who was commanded to explore the area by boat and find the best locations to build the aforementioned fortifications. However, the expedition was a complete disaster and the entire crew died. The idea of the Pacific Ocean as a Spanish Lake, as a ‘mare clausum’ and as a natural defence against the European countries, came to an end when a new route via Cape Horn was finally discovered in 1616 (Fernández Duro 1972). The armed defence When addressing the main question – when was the Pacific Armada created – we cannot really give a proper answer yet. In his most recent studies, Pedro Emilio Pérez Mallaína has tried to give to the South Pacific Armada a date of birth. The only document found related to its origins is a royal letter from 1581 signed by King Philip II, who said:

Initially, the Viceroyalty of Peru did not encounter many attacks. It was believed that the Pacific Ocean was a closed sea, which meant that the main access to its waters was through Panama, which was also under Spanish control. Another route was through the Strait of Magellan, but all the expeditions that took place early on ended in disaster, and this made it easy to believe that it was a ‘doomed passage’ (Pérez Mallaína and Torres 1987). The cost of these expeditions was extremely high and most of the first explorers died as a result of hurricanes, strong currents, a lack of knowledge about the territory, etc. This created the idea of a closed sea, one that the Spaniards even took to calling mare nostrum or ‘our sea’, using the natural defence of the strait itself to create the idea of a Spanish lake. However, the discovery of Cape Horn in 1616 changed this belief.

Conviene que los navíos en que se trae a la provincia de Tierra Firme la plata y el oro, vengan del Perú juntos y en forma de Armada, bien artillados y apercibidos para cualquier ocasión que pueda ofrecer (It is convenient for the ships bringing the silver and gold from Tierra Firme to sail from Peru together, in the form of an Armada, well-armed and prepared for any situation that may arise). It seems as though the origin of this Armada was, in fact, a mere backup plan resulting from the failure of the previous defence systems which were based on natural defences and the building of fortresses in the Magellan Strait, as well as continued attacks from privateers. Thomas Cavendish was one of the first British privateers to cross the Strait and attack the Spanish colonies, followed by Richard Hawkins just before the end of the

Initially, the Viceroyalty of Peru surprisingly had no policy on defensive tactics. It felt secure in the fact that the Viceroyalty was a very long distance away from other enemy countries and it relied on the latter’s lack of knowledge about cartographic maps of that part of the world. In addition to this, the reality was that the other 121

Cristina Agudo Rey Guayaquil was where the ships for the armada would take shape but this always served the premises and interests of the merchants, due to a lack of funding from the crown. Ultimately, the South Pacific Armada was built with merchants’ money.

sixteenth century. These attacks reinforced their only remaining option: the mobilisation of a fully armed fleet not only for defending the silver and gold route and the coast from attacks, but also to carry the precious metals. This was the will of King Philip II, as he wrote in a royal document on 28 November 1590:

Guayaquil was located in Ecuador and was founded in the origins of the Viceroyalty in 1538, but it gained more prevalence as the Viceroyalty grew. This port-based city stood out because of its commercial links with various colonies and Spain, and its development was based on four principal colonial activities: the wood and timber market, the shipyard, the harbour and, to a lesser extent, the cocoa market. All these activities were linked with the maritime environment and it is no surprise that the city became one of the most important ports within the Pacific Spanish Empire.

Que en las costas del Sur se esté con cuidado, por si parasen corsarios a aquel mar. […] en tal forma y prevención, que si pasasen algunos enemigos o corsarios, hallen resistencia bastante y sean castigados (We must take care on the Southern Coasts, for privateers may be present on that sea. (...) this way and with precautions, so that if any enemy or privateer is seen, they will encounter sufficient resistance and they will be punished) (King Philip 2.2). However, it is an assumption that the origin of the building of a South Pacific Armada was based on these privateer attacks. Most of the royal documents regarding this armada were more focused on the protection of the American cargo, metals and trade, rather than the building of a strong and fully-armed fleet. Virtually all their concerns were about the defence of the silver on its way to Panama (King Philip 2.8). The idea was to build large merchant ships which would be able to carry guns so they would be able to defend themselves in the event of an attack. This was a new concept of an armada and these armed ships were used for two different tasks: not only for transporting silver for trade, but also as battleships. However, this brought more problems than solutions for the defence of the Viceroyalty of Peru. The commercial aspect of the ships would soon overtake their defensive role and they did not follow the royal legislation entirely. The aforementioned royal ordinances prioritised the defence of the coasts, but the reality would turn out to be very different, and the ships carried more silver in their holds than guns on their decks. (Pérez Mallaina and Torres Ramírez 1987).

In terms of dating the start of the fleet’s construction, there is not sufficient data to specify a date. The closest date related to the construction of the first ship in Guayaquil is around 1557, which is 20 years after the shipyard opened for business. The first ship built was a small ship, possibly a galley, commissioned by Viceroy Don Andrés Hurtado de Mendoza, Marquis of Cañete (Document has not been found after it was lost by the researchers who were consulting it, but it is supposed that it is in the General Archive of the Indies (AGI) in Lima). The reality is that by 1560, the shipyard was fully dedicated to the building of battleships; however, it is difficult to determine what the battleships’ class were at the beginning. Regardless, shipbuilding itself became more important and was an entire industry by the end of the sixteenth century. The situation would be different and more irregular between the end of the sixteenth century and the seventeenth century. Every ship from the armada had a different need and so the construction of new ships in Guayaquil was constantly interrupted, but this is also difficult to prove due to the lack of resources and official documents regarding the shipyard’s operations and ordinances.

Royal or specific documents about the purpose of the armada are difficult to find, if indeed there are any, due to the various modifications the armada underwent throughout its lifespan. The viceroyalty changed the ships’ purpose and even their shape, depending on the current viceroy’s opinion. Sometimes the fleet was more focused on the defence of the coasts and viceroyalty waters, as it was meant to be, but on other occasions, the fleet became more engaged with the silver trade. Moreover, there are no documents or records regarding the crew regulations of the South Pacific Armada, i.e. how the ships were manned or the existence of royal marines on board. In addition to this, the fleet was not initially formed of new battleships, but rather, private ones that were either rented or purchased. This situation would not change until the use of the shipyard in Guayaquil, which became the main location in the Viceroyalty charged with building and repairing the armada. However, even after the battleship construction activities started in Guayaquil, the renting of private vessels would always be used to bolster the fleet.

The South Pacific Armada faced a lot of difficulties due to the dockyard’s problems concerning funding the shipbuilding. In addition to the timber, most of the other materials were imported from Europe and were very expensive. The fund also changed depending on the whims of the viceroy in power at the time. The main source of income came from the royal coffers in Lima and Guayaquil, but private money was also provided. One example of this is the arrangements made between Viceroy Don Baltazar, Count of Castellar and the local shipbuilder named Goyonete who had to finish his initial work of a regular cargo ship as a battleship by viceroyalty order (Clayton 1978: 235–36). Shipbuilders had to accept private financing from traders who were looking out more for their own benefit and business than for the defence of the territory or the trading route. Another option for adding more ships to the fleet in the event of 122

Cultural Underwater Heritage of the South Pacific Armada special requirements or attacks was to convert personal or private ships into battleships, but these measures were merely temporary and relatively ineffective. Finally, the main issue in comparison with the rest of the Spanish Armada was that the South Pacific Armada was very expensive to build due to most of the materials not being from the region and the extremely high taxes. The abovementioned problems were so significant that the viceroyalty had to look for other, cheaper alternatives in terms of materials and timber, and they therefore created a new commercial exchange circuit within the region, including nearby colonies and the rest of the empire. Iron, copper nails, ammunition and supplies were brought from Europe, with these being the most expensive items due to the import taxes. Copper, cotton and ropes for the riggings were brought from Chile, while timber was gathered from the Viceroyalty of Peru itself.

to Panama (Pérez Mallaína and Torres Ramírez 1987: 140). The Spanish ships would begin to outlast the more advanced ones by the middle of the seventeenth century. Meanwhile, the British Royal Navy already had ships weighing up to 1500 tonnes. The ships of the South Pacific Armada encountered different enemies and suffered numerous attacks as well. Depending on developments in international politics in Europe, the South Pacific was attacked and raided by different nationalities at different periods of time, and these attacks increased in number after the DutchSpanish war. Royal orders were established to always keep a minimum of five battleships in the Viceroyalty’s Navy. However, the Armada was normally formed of three ships which were used mostly for the transport of precious metals; two ships carried the royal coffers from Arica to Panama and the third ship carried mercury from the Huancavelica mines. Hence, the coasts of the Viceroyalty often remained undefended. In 1609, the local administration decided that the quickest solution to this situation was to reinforce the Armada with private ships after modifying them as battleships. However, public criticism was turned against the Armada, especially after the loss of the Battle of Cañete. In this battle against the Dutch navy in 1615, one of the ships, the Santa Ana, was defeated and sunk. Critics began to define the ships as ‘cajones huntados[sic] de brea’ (big casks spread with tar) (Clayton 1978: 52). Holland’s continued attacks convinced the local administration to build more ships for the Armada, but nothing changed about their characteristics, nor were any improvements made. By the end of the seventeenth century, we start to see some changes in terms of their weight and cargo, in an attempt to modernise the Armada by Viceroy Conde de Alba de Liste. He ordered two new galleons: the San José and the Nuestra Señora de Guadalupe, each weighing 825 tonnes and carrying 32 guns (Pérez Mallaína and Torres Ramírez 1987: 95). These ships ran into a lot of problems when trying to keep the coastline secure, since their size made them more difficult to manoeuvre. Most of the South Pacific coast is made up of open bays with rough waters, and this issue was later used as an argument about the difficulty of using ships weighing more than 500 tonnes (Clayton 1978: 52).

Regarding the physical characteristics of the armada, sources and records are more concerned with periods of high tension or attacks and the viceroy had to write instructions to build and acquire more ships. After Francis Drake’s attack in 1577, Viceroy Francisco de Toledo started the construction of two galleys, thus encouraging the first works in Guayaquil. We know that only one of these galleys was built: the Santísima Trinidad, which was launched in 1579. As the British attacks continued, the next viceroy, Martín Enríquez, ordered a new bigger ship named San Pedro y San Pablo y el Apóstol Santiago, which was launched in 1586 and which joined the Armada as the flagship (Clayton 1978: 78). This ship was much bigger and more robust than those of the Mediterranean fleet, and the majority of the South Pacific Armada would soon have the same characteristics: a weight of between 350 and 600 tonnes and carrying between 20 and 36 guns. These standard ideal galleon characteristics were fixed by Viceroy Conde de Chinchon in 1634 in a document that can be found in the ancient merchants’ exchange in Seville. In his studies, Pérez Mallaína describes the ships as weighing 500 tonnes, and having a 30-yard keel, a 13yard beam and a 10-yard draft. The main wood used for the ship structure was a local wood called Guachapelí which is durable and dark in colour (Clayton 1978: 80). The guachapelí was the main tree used by the timber industry due to its hardness and longevity. Most of the ships built in the main dockyard in Guayaquil used this wood and they lasted for over 50 years. The ships would also be given another layer of oak for structural protection. Most of the ships from the South Pacific Armada had two gun decks, a poop deck, as well as a forward castle. They had four masts and the sails were mostly square. However, the main characteristic of these ships was their curvy and bulky sterns. Two clear examples of this are the Visitación and the Jesús María, both of which served for 15 years (Clayton 1978: 80–82). If we compare them to the Atlantic or West Indian ships, the latter never lasted for more than 4 round trips to America. Both sea journeys – the South Pacific silver trading journey and the one from Spain to the Caribbean and South America – were as difficult and long as the South Pacific journey from the Port of Arica

Studies on why the ships were not very good at fighting or defending are still very limited, but it seems as though one of their main difficulties in defending against attacks from other ships came about because the main users were merchants who financed bigger ships, thus favouring more space for the transport of goods over security and manoeuvrability. As stated by Pérez Mallaína, they were suited to the general needs of a very particular commercial sea route. The reason why some of them were sunk was, in addition to economic problems, the fact that the ships were simply too big and difficult to handle. To quote Mallaína once more, ‘It was too suspicious that they were better at transporting cargo than firing their guns’ (Pérez Mallaína and Torres Ramírez 1987: 162). 123

Cristina Agudo Rey Underwater archaeology shipwreck map Having discussed some of the issues of maritime cultural heritage in Peru and outlined an important data set, this section proposes some steps forward. As previously mentioned, there are a limited number of studies and fieldwork is lacking. Most of the studies are based on written sources, although these do not make it clear why the South Pacific Armada Ships could not stand up to any external attack and were all sunk. According to the documents from the period, the shipbuilding techniques used should have given them good stability, defence and overall performance. Due to the lack of underwater archaeological activities and studies, the South Pacific Armada has not garnered enough historical recognition and we are thusly also missing all the social and cultural facts related to it. Maritime archaeological activities will help us to study the remains and, by analysing the shipwrecks, we may be able to get some idea of the techniques used in their construction. This, in turn, could then be compared with the existing documentation. There are some organisations, such as the Institute of National Maritime Archaeology (INAM) which, since 1992, has been developing various projects to rescue Peru’s underwater heritage. Their work has been focused on compiling all the documentation from the time and all the information gained from dives carried out in the past 30 years. During the 1990s, the INAM rescued various artefacts from shipwrecks: shots, coins, medals, anchors and even one bronze cannon. Today, they are focusing on locating the main shipwrecks and identifying the ships and their cargo, in order to create a database archive.

Figure 2. Peruvian coast map showing the main areas of sunken ships and underwater sites; numbers refer to the various coastal sections in the country: 1. Tumbes, 2. Piura, 3. Lambayeque, 4. La Libertad, 5. Ancash, 6. Lima, 7. Ica, 8. Arequipa, 10. Tacna; the circled areas show where there are a big number of archaeological discoveries or active regions during the Viceroyalty.

Without a prior study about the South Pacific Armada, it would not have been possible to recognise the underwater heritage from the old viceroyalty, especially in the face of the lack of underwater archaeological activities. We know about the existence of this underwater heritage, and it will therefore be very worthwhile and interesting to develop future projects that will fill gaps in the maritime history of the Viceroyalty of Peru.

My research has been focused on using this data to create a project which will involve the drawing up of a map of shipwrecks. This map will pinpoint the main sunken ships of the South Pacific Armada and it will also be usable for archaeological surveys. This is an ongoing endeavour, and preliminary results are presented in this conclusion.

Works cited

Clearly noticeable are two main areas where there are more shipwrecks and these correspond to the two main dockyards at Palta and Callao Bay, probably due to their strategic location. They are located in the department of Piura, where important harbours operated during the seventeenth century, and the department of Lima (see Figure 2).

BRADLEY, P. 1975. ‘Some considerations on the defence at sea in the viceroyalty of Peru during the seventeenth century’. Historia de América 79: 77–97. CLAYTON, L. 1978. Los astilleros de Guayaquil Colonial. Guayaquil Historic Archive: Ecuador. FERNÁNDEZ DURO, C. 1972. Armada Española desde la unión de los reinos de Castilla y Aragón, VOL III. Sucesores de Rivadeneyra: Madrid.

This project will help to create a database containing all the archaeological sites and shipwrecks. In consequence, this will help to develop potential future projects and also implement appropriate protection and conservation measures given that their locations will now be marked. The project was based on an existing shipwreck map of Andalusia, in Southern Spain, where 81 shipwrecks from different historical periods were found, ranging from Phoenicians times to the Battle of Trafalgar.

GARCÍA RIVERA, C., ALZAGA GARCÍA, M., MARTÍ SOLANO, J., RODRÍGUEZ MARISCAL, N. 2009. ‘El centro de arqueología subacuática de Andalucía. La protección del patrimonio arqueológico subacuático’. Arqueología Náutica Mediterránea. Gráficas Varona: Salamanca.

124

Cultural Underwater Heritage of the South Pacific Armada MORALES, E. 1932. Sarmiento de Gamboa: un navegante español del siglo xvi. Araluce: Barcelona. OYARZUN, J. 1999. Expediciones españolas al Estrecho de Magallanes y Tierra de Fuego. Agencia Española de Cooperación internacional: Madrid. PÉREZ MALLAÍNA, P. E., and TORRES RAMIREZ, B. 1987. La Armada del Mar del Sur. Consejo Superior de Investigaciones Cien: Sevilla. PHILIP II, KING OF SPAIN. 1590. West Indies Laws, Book IX, Law No. 2 and 3. General Archives of Indies, Seville. UNESCO. 2004. The General Law for Cultural Heritage, Law No. 28296. Chapter 7. VILLAR ASTIGUETA, R., POZZI-ESCOT, D., and HOYLE, A.M. 2015. ‘Peru Underwater Heritage’. Underwater Cultural Heritage in Latin America and the Caribbean. UNESCO: Havana. Pp. 42–45. IAPH. (Andalusia Institute of Historic Heritage). National Department. Outstanding projects from the Center of Subaquatic Archaeology. Accessed 10 June 2018. Available from .

125

12 Maritime Archaeology in Biscay, Basque Country: Facts, Acts, Research and Opportunities José Manuel Matés Luque UPV/EHU University of the Basque Country, Arqueocean [email protected] Abstract: The purpose of this paper is to present how maritime archaeology has been carried out in the province of Biscay (Basque Country, Spain) in the last few years, from the first casual discoveries to the more prepared projects, either from an academic point of view or as a consultancy to the civil engineering works developed. Shipwrecks, port structures and other elements have been found and recorded. However, this is not enough if maritime archaeology wants to stand out and present itself as a discipline which can not only provide us with information about our past, but also used as a resource by many stakeholders. Opportunities are here if those in charge know how to use them efficiently. Keywords: Maritime archaeology; Biscay; Basque Country; research opportunities; management Introduction

found and researched so that the current situation of maritime archaeology can be understood in a broad sense (i.e. underwater archaeology, intertidal and foreshore archaeology) in the province. It will assess the current stage of research and present the opportunities that can be found in managing maritime archaeology wisely so that new sites can be found, researched and managed. Therefore, gaining new knowledge on the maritime past of the province will benefit all of us and guide the development of maritime archaeology in Biscay. What follows are the most important different projects and archaeological activities which have allowed maritime archaeology in Biscay to find its place, step by step, for the last 20 years.

Traditionally, maritime archaeology has not been an issue dealt with until recently in the province of Biscay, Basque Country. The finding of Roman coins at the beginning of the twentieth century while dredging the Bilbao port (Churruca 1904) cannot be considered maritime archaeology, but a lucky strike. However, this lucky strike does not take away the information gained about the Roman presence in the province. Besides such casual findings of coins, some other finds were discovered, exposed or found during the twentieth century, but little or no information about them is known other than the type of finding and where it took place. In fact, no pictures, drawings or images are known (MartínBueno et al. 1985). This clearly shows how little maritime archaeology has been a topic to be researched.

Urbieta wreck After reading the above paragraphs about how scarce maritime archaeology has been dealt with in Biscay, it is no surprise that it was not until the summer of 1998 when the first maritime archaeological find was found. The discovery took place while performing watching briefs in an old riverbed of the Oka river. Here, besides some old pottery and raw timber (which was suggested to have been cut for shipbuilding), a late-fifteenth century clinker built boat was found (Izaguirre et al. 1999, 2001; Izaguirre and Valdés 1998) (Figure 1). This was, and still is, an important discovery, as it was the first example of nautical remains properly excavated and researched (Rieth 2006; Rieth and Izaguirre 2004) and it seems it belongs to the tradition of boat and shipbuilding carried out in the Basque Country from the eleventh century (possibly as early as the tenth, but certainly by the twelfth century), for which the Basque Country is famous for. In terms of the boat’s maritime activities, based on Basque Country

It should be remembered that the province of Biscay is related to a long tradition of maritime history, but maritime archaeology has been an overlooked theme here until recently. We need to wait until 1998, when a late-fifteenth century clinker built boat, known as the Urbieta wreck (Izaguirre et al. 1999), was found in an old river course while conducting watching briefs during the repair of the old river course to be converted into a park. From that moment on, other maritime archaeological projects have been carried out. Some of them were commissioned by the Archaeological Service of the Biscay Regional Council, while others were done under private initiative. The purpose of this paper is to present the maritime archaeological activities which have been carried out in the province of Biscay, in an attempt to present the sites 127

José Manuel Matés Luque

Figure 1. Side clinker built view of the Urbieta wreck.

maritime traditions, some possible functions are whale hunting (Berraondo 1932; Ciriquiain Gaiztarro 1961; Laburu 1983) or navigation and trade between Spain and the north of Europe during the Middle Ages and later, when the expansion to America and the rest of the world took place (Alberdi Lonbide 1996; Alberdi Lonbide and Aragón Ruano 1998; Casado Soto 1975, 1995, 2004, 2006; García Fernández 2005; Laburu 1987; Lema Pueyo 2004; Odriozola Oyarbide 2002, 2003; Ruiz De La Peña 1998; Solórzano Telechea 2005). This boat is the first one in a continuous list of Medieval findings, outside the Basque Country, whose shipbuilding origin is in the Basque Country, or in what is called the Basque-Cantabrian region (L´Hour and Veyrat, 1989; Nayling and Susperregui 2014; Pujol i Hamelink and Soberón Rodríguez, 2011; Soberón et al. 2012; Soberón Rodríguez 2010; Zallo Uskola and Gómez Bravo 2016).

were some cannons just a few metres underwater below his boat in the estuary of the Oka river, not far away from its mouth. Thus, a research team was contacted and they found not one but two ships belonging to the nineteenth century that were involved in the civil wars in Spain at this time (Gómez Bravo and Zallo Uskola 2012, 2013, 2014a, 2014b, 2015, 2016, 2017, 2018; Zallo Uskola and Gómez Bravo 2012). From the moment of their discovery, regular monitoring dives are being performed on such shipwrecks to assess whether they are exposed or covered by sand to protect them. Once more, these wrecks show that there are many more to be found in the region. In fact, historical sources points to several vessels throughout history that were wrecked or foundered in the area (Casabán 2016).

What this boat also demonstrates is that vessels can appear anywhere and that any type of waterfront (either lake, river, estuary or coast) and nearby areas (in case land reclamation took place in the past and we are not aware of that until a find is exposed) should be protected in such a way that whenever a civil work takes place, there is an archaeologist on duty to overview such work. One never knows when a tiny piece of information might provide us with another important clue to understand the past better (see below Laida Beach redevelopment).

Due to the construction of a new sewage system to meet the requirements of an important area, an archaeological project was carried out. This was due to the fact that a nearby historical hospital might have been affected by the works performed on the beach and that several Palaeolithic sites are nearby. Unfortunately, no archaeological artefacts or features were found (Aragón Núñez 2011). However, the lack of archaeological finds does not mean that no archaeological data can be found nearby. The recovery during the twentieth century of some iron anchors, possibly dated to the late-nineteenth century or earlytwentieth century, and now on display in the Museum of Placentia de Butrón, is a clue that some shipwrecks might be in the area.

Gorliz sewage archaeological watching brief

Laida I and Laida II wrecks Speaking to the aforementioned point, by chance, a boat owner was cleaning his boat when he realised that there 128

Maritime Archaeology in Biscay, Basque Country Bou Nabarra underwater search

hopefully, will be amended soon when properly-managed archaeological projects take place.

The bou Nabarra was a fishing boat armed to be part of the Basque Auxiliary fleet during the Spanish Civil War. On one occasion, she faced a battlecruiser while protecting a convoy of supplying ships and after several hours of combat, she was sunk; however, her captain decided to remain on board and go down with her. This ship has been looked for several times as it has a romantic and heroic imprint that has attracted several research projects. However, despite the different campaigns, her whereabouts is still unknown, notoriously marking her as a ‘Grail Quest’ (Matés Luque and Armendariz 2018: 536–37).

Besides these historical wrecks, there have been some other modern wrecks, some still underwater but others which have been broken up (Benito and Mazpule 2017; Torres Goiri 1992). Thus, the importance of them as archaeological assets is, at the moment, very weak, but they should be kept in mind for future activities. Bakio shipwreck This wreck is known from 1985 when some cannons were seen on the seabed; unfortunately, no archaeological recording was carried out. Later, in 1999, local resident Miguel Uriarte, with the support of the town council, retrieved several cannons and other artefacts to the beach (Figure 2). It was not until 2004 when the archaeological recording of artefacts and a non-disturbance underwater survey was carried out. The results of such research point to a late-eighteenth century small British warship en route to or from Gibraltar, which was under siege from Spain in 1779–1783. This has been specified based on the fact that one of the two flintlock muskets is a 1779 Shortland Pattern, the regular musket used by British troops at this time. The other musket is a Hanoverian one, whose troops were also fighting alongside the British at this time (Matés Luque 2005, 2006, 2012, 2013a). This hypothesis has to be confirmed when more underwater findings turn up. A new underwater project is being prepared to continue researching the site and, therefore, it is hoped that the results might shed light on their origin and date.

Other twentieth-century wrecks While looking for the bou Nabarra, for which a side scan sonar was used, some other anomalies were found. Among those are a German cargo ship, named the Nordfels, which lies at 90 metres deep. Additionally, the Nazi patrol boat V-408 Hantelbank was found and so were the later-identified Norwegian cargo ships SS Tiger and SS Leikanger, sunk by a German submarine in 1917. Today, technical sport divers dive these sites. Besides these, some other Spanish cargo ships are also regularly dived by different sport divers as they are in depths ranging from 10 to 70 meters deep (Matés Luque and Armendariz 2018: 538). The deepest one are dived regularly by technical divers. Archaeologically, little information is available about any of them and the data is erratic and inconsistent. This lack of information,

Figure 2. Guns from the Bakio wreck.

129

José Manuel Matés Luque

Figure 3. Lekeitio pier which zig-zags along the river course so that the sediments do not end up in the town fishing harbour (to the left of the photo) but flow out to sea (to the right of the island, where the pier ends).

Lekeitio pier

raised, it was clear that it was not a cannon but a ship’s vent. It might have been related to the shipyard just metres away up the river on the same shore and, after falling out, the current left it there to be buried until it was found (Matés Luque 2013b). Despite the little archaeological interest in this item due to its recent date, it shows that there are items buried under the sand on beaches and riverbanks awaiting to be found and researched.

This archaeological activity was forced due to the need to repair a pier which was destroyed by several sea storms. The pier worked to channel the river Lea so that the sediments did not silt up the fishing harbour of Lekeitio (Figure 3). Since the pier seems to have been designed and/or built by one of the enlightened men of the Basque Country, Pedro Bernardo Villarreal de Berriz (Villarreal de Berriz 1973; VV.AA. 1990), an archaeological recording of the fabric and a watching brief of its reconstruction and repair was carried out. Although it could not be ascertained that the oldest available part of the pier belongs to Villarreal´s design, the archaeological research allowed a better understanding of the different stages of its building throughout time (Matés Luque 2016a) (Figure 3).

Laida beach redevelopment The watching brief on the regeneration of the estuary of the Oka river, just some metres down the findings of Laida I and Laida II wrecks, turned up quite unexpected findings. Due to severe storms and bad weather, the Laida beach, a popular destination for the summer season, needed to be regenerated as a large volume of sediment disappeared. In order to achieve this, caterpillars were deployed to get sand from different areas of the intertidal zone on the mouth of the estuary. Since the Laida wrecks (see above) were an important heritage resource, a boundary zone around them was established. Similarly, as the estuary is known to have suffered from several historical ships being foundered and run aground, the regeneration needed supervision in case any artefacts like cannons, anchors or part of a new wreck were exposed by the caterpillars.

Ship fitting found in Kurlutxu This fitting was buried under the sand of the river Lea, just metres away from the Lekeitio pier, and was revealed when the sand was removed by the current. Due to its position and shape, it was thought to be a cannon as there was also some big thick pieces of timber which might have been part of the gun carriage. Additionally, some sherds of pottery were recovered nearby. A team was set up to excavate the area nearby quickly and to ascertain whether it was a cannon. More pottery, timber pieces and a bowl of a clay pipe were recovered. However, when the item was

The archaeologist responsible for the watching brief not only overviewed the movement of sand but also used a metal 130

Maritime Archaeology in Biscay, Basque Country

Figure 4. Recovering sand from the estuary at low tide; wrecks are supposed to be buried in the area, like the Laida I and Laida II.

detector to scan the big trenches created by the caterpillars after most of the caterpillars were removed (Figure 4). Since cannons were found in the Laida wrecks, the use of the metal detector was a reasonable tool to deploy. While surveying the trenches, a clinker built plank was found, preliminary dated to the fifteenth century (Figure 5). The Urbieta wreck, roughly 10 kilometres up the Oka river, was the reference vessel to compare such a plank against. In order to ascertain the date, a sample was cut along the plank and sent to the dendrochronological laboratory, the same laboratory in which the identification of the Newport ship as being a Basque vessel was undertaken (Nayling and Susperregui 2014). Unfortunately, the rings were distorted and no date could be obtained. Therefore, the same sample was sent to a C14 laboratory, the same which dated the Urbieta wreck (Izaguirre et al. 2001). The results were unexpected and surprising as they dated the plank to sometime between 1210 and 1275 AD (Matés Luque 2016b). This date makes the Laida beach plank to be an important finding, despite being small, as it belongs to the moment when Basque coastal towns and maritime activities (fishing, whaling, trading, shipbuilding, war…) were being firmly established. Although maritime developments start to take place in 1000 AD, it is some time before iconographic images of Basque vessels appear with fishing villages in their seals, such as in San Sebastian or Bermeo, depicting, respectively, both a merchantman and a whaling boat (Alberdi Lonbide 1996; Alberdi Lonbide and Aragón Ruano 1998; Berraondo 1932; Casado Soto

Figure 5. Clinker built plank radiocarbon dated to 1210– 1275 AD; it is the oldest medieval piece of watercraft found in the Basque Country, probably one of many yet to be uncovered.

1975, 1995; Ciriquiain Gaiztarro 1961; García Fernández 2005; Laburu 1983, 1987, 2006; Lema Pueyo 2004). Due to the 200-year gap between the date of this plank and the date of the Urbieta wreck, it seems clear that this plank points to a thirteenth-century wreck that may be found eventually somewhere within the estuary. Therefore, this is a new Medieval shipwreck discovered in the Basque Country (Matés Luque 2016e). For this reason, whatever civil engineering works are carried out anywhere in the estuary should be archaeologically 131

José Manuel Matés Luque brief of the area both underwater (up to 30 metres depth) and on the beach. The whole route required the report of the underwater seabed and, thus, the magnetometry geophysical data was assessed. No human-made artefact was identified in the area of the cable or near the cable. However, the Spanish Ministry of Culture suggested checking four anomalies on the seabed near the beach and near the cable. Nothing unusual was found on or near those anomalies. Similar results were obtained on the seabed before and after the cable was buried and on the beach (Figure 7).

overviewed. Additionally, protocols should be deployed in case any other piece of timber is found. The archaeologists in charge of the monitoring of the Laida wrecks check the estuary regularly for timbers which they keep in a store room. Mendexa stone anchor Just on the other side of the river Lea is the territory of the Mendexa town council. Here, on the Mendexa beach, a classic three-stoned anchor was found by a tourist. Although no date can be ascertained due to lack of other data, this is the first properly-identified stone anchor (Matés Luque 2015a, 2016d) (Figure 6). It might also be the first of many stone anchors in the region. This proposition is made based on the nearby province of Gipuzkoa, where a number of stone anchors have been found (Benito Domínguez 2000). Therefore, it is unusual that so many have been found in Gipuzkoa but only one in Biscay. In order to identify future finds, as has been said above, it is necessary that any civil works which take place on beaches or sandy areas are archaeologically overviewed.

The reason for the archaeological activities is due to the fact that the Spanish government takes the impact on underwater heritage due to big projects such as pipelines, cables and similar projects very seriously. Additionally, it is the responsibility of the Regional County, in this case Bizkaia, to license the project. This is due to the fact that the nearby area is very rich in Palaeolithic and Neolithic sites, and the underwater area is poorly researched, similar to the Gorliz sewage archaeological project. Unfortunately, again, no archaeological artefacts or features were found (Matés Luque 2018a). However, this lack of archaeological data can be compensated with the fact that the available resources have not been overlooked. If more archaeological watching briefs or surveys related to assist any civil engineering works become regular practice, our knowledge of the underwater historical environment will expand. Eventually, some artefacts or useful archaeological data will be found.

Marea Project: underwater communication cable lying In June 2017, a new submarine communication cable was laid between Virginia Beach (USA) and Sopelana (Spain). This project required an archaeological watching

Bilbao fish weirs This research project strove to study fish weirs which have been in use at least from the early-twentieth century (Figure 8). They are used to capture baby eels, which are a delicacy, valued particularly for Christmas meals. Despite being quite modern, little is known about them in archaeological terms and, therefore, the purpose of the project was to get some information about their fabric, orientation and any other information which could be obtained during recording (Matés Luque 2015b). Although this archaeological information might be scarce, it will shed light on this type of fishing activity which is a bit better understood in other areas of the Basque Country (Merino 2003). This is particularly true since this type of fishing might have been more widely used through time. Indeed, fishing resources have been used by the famous inhabitants of the Basque Country in the Palaeolithic period, where there are many caves containing evidence of fishing and shellfishing remains (Altuna and Apellaniz 1976; Berganza et al. 2012; Gutiérrez Zugasti et al. 2011; Merino 2003; Roselló Izquierdo and Morales Muñiz 2011). It is clear that fishing activities must have been an important resource in the Roman Period and Medieval Period. However, no archaeological evidence is available for Biscay other than historical data on some portable fishing traps used in rivers (Lema Pueyo 2004). However, we should not forget that bigger fishing traps existed in the Middle Ages in other areas of the Basque Country (Alberdi

Figure 6. Classical three-holed stone anchor; first one that has been clearly identified.

132

Maritime Archaeology in Biscay, Basque Country

Figure 7. Cable ship during the landing of the telecom cable by the beach; despite the lack of archaeological data, it is a great achievement that the underwater watching brief was carried out as this has not been traditionally done in the past, thus, we do not know what has been lost due to the lack of archaeological activities.

Figure 8. Fish weir made of different types of fabric; this represents the reuse of the item over time.

133

José Manuel Matés Luque point out how little we know about such intertidal areas unless there are big structures (mills, piers, shipyards…) in the area (Ibáñez Gómez et al. 1988). Ultimately, we know little about any other remains in the intertidal area. Thus, we have tried to record all features found on the waterfront of estuaries (coastal area will be dealt with at a further stage), not only mills, piers and shipyards, but also shipwrecks, causeways, ramps and staircases to access the rivers, landing and loading places, dykes, quays, dams, slipways, irongrids, piers, berths, jetties, timber posts for reclaiming land and artefacts, if any are found (Figure 9).

Lonbide and Pérez Centeno 2004; Pérez Centeno and Alberdi Lonbide 2006). Therefore, it is not unreasonable to suspect that in rivers of Biscay, similar, albeit smaller, fishing traps might have existed. Could the current ones still available up the river from the Bilbao city centre have taken the place of Medieval or post-medieval weirs which were more widely common? Certainly, it is an interesting issue; if they were more widely spread out along the river course, the industrialisation which took place from the mid-nineteenth century destroyed them all. In their place, wharfs, docks, shipyards, embankment areas, cargo landing places and so on were built. This is why it is so important to understand current fishing activities by using such weirs.

As we have pointed out, this project is currently taking place and it is hoped that once finished, a better understanding of many of the remains will be achieved, thus allowing us to shed light on the relationship between human beings and their maritime (either sea or fluvial) environment (Matés Luque 2018b; 2018c).

Intertidal archaeology This project is the writer´s PhD topic which started in the academic year of 2016–17. Therefore, it is still ongoing and, thus, no final results can be presented. Nevertheless, we can offer some preliminary inputs on the fieldwork carried out during 2016–17. Intertidal archaeology has not been treated from a holistic point of view. The sites investigated so far have been done in an isolated way. For instance, when dealing with a tidal mill, research is done on such mill without paying too much attention to other features or sites nearby. This issue persists when dealing with piers, shipyards or landing areas. This is not to be critical against other colleagues´ work, but rather, to

Bilbao waterfront redevelopment Despite being a very industrialised city, some maritime archaeological works have been carried out in Bilbao, mainly the watching briefs of several parts of the waterfront. This is linked to the fact that Bilbao Town Council has been ordered to repair many quays at the waterfront in the last few years. In such repairs, watching briefs have been performed. Although not much of the waterfront was altered, some interesting data were retrieved. For instance,

Figure 9. Irongrid use for repairing ships; the river of Bilbao contains several of this type of item which provides information about repair and boatbuilding.

134

Maritime Archaeology in Biscay, Basque Country different types of rails for tramways, cranes and trains were found. Certain rails, specifically the Phoenix and the Vignole, allow us to understand the commercial busy life of cargo loading and unloading. They also show how the waterfront was extended towards the Bilbao estuary in an attempt to gain more loading area, even if this meant reclaiming the natural river course to add the seventeenthcentury Uribitarte island (which in Basque means the place between two water courses). This island and all the above-mentioned features were identified (Figure 10). In short, the watching briefs allowed us to understand how a

city grew vertically and expand horizontally (Matés Luque 2013b, 2014a, 2014b, 2016b) (Figure 11). Deusto Canal opening This canal is an old project, as the opening started in the 1950s and ended in the late 1960s without performing the full opening of the canal. Since then, the opening idea has been an ongoing project involving several authorities and government bodies to fulfil several purposes. Among them is to have a canal which allows any possible rise of the river due to strong rain and high tides to have an extra area to flow, hence avoiding any floods, like the tragic one in 1983. Additionally, the area will be redeveloped, with new modern buildings and bridges to link the island with the opposite waterfront from which it will be separated. The opening started in 2015 but it was not until 2018 that the proper archaeological watching brief was undertaken. At the time of writing this paper, little can be said about significant results, other than the fact that some pottery sherds have been found, the most important of which is a tiny piece of a Bellarmine rim. Additionally, several stems of clay pipe have been found, but no date can be determined for them as they lack any marking or decoration which could be used to date them. Some other sherds of pottery, from the late-nineteenth century and early-twentieth century have been recovered. Some timbers were also exposed. Many might be reused pieces

Figure 10. View of the Uribitarte Island which was later covered with cobble stones in 1870.

Figure 11. View of the Arriaga waterfront where several features were identified: the 1870 waterfront (middle of the picture where the ranging pole is flat) and the tram railways on both sides of the 1870 waterfront; the left one is from the 1870s and the right one is from the 1920s when the river was reclaimed for gaining docking space.

135

José Manuel Matés Luque

Figure 12. The first pieces of timber found during the watching brief; some spikes and boat pieces can be recognised.

and they consist of planks, knees and, perhaps, frames or other types of structural pieces which have been cut to have a pointed edge to be buried in the ground. Some raw timbers which have the shape of a ship piece to be converted have been also identified (Figure 12). Some timber samples were sent to the dendrochronological laboratory but, unfortunately, rings were insufficient in some of the samples or not useful in acquiring a date. However, it is thought that they might be from the end of the nineteenth century or beginning of the twentieth century. Having said this, we think that the line of timber posts buried in the ground which we were able to record in situ is part of the slipway of a late-nineteenth century shipyard (Figure 13). This is supported by the fact that two frames have been recovered. They have a Roman numeral ‘V’ inscribed on one of the surfaces (Figures 14 and 15). They reflect the past boatbuilding activities in the area, in use until approximately the 1930s, when the area was heavily industrialised with new mechanical and chemical workshops.

If we take a holistic approach, we can see that despite the lack of deep data and knowledge about all of them, it is clear that the assessed maritime interconnected sites are part of an overarching issue which, if faced properly, will enhance our knowledge of the relationship between human beings and the maritime seascape. The sites which we have presented here are those for which we have more data, as they have been dealt with an archaeological methodology. Similarly, it is quite clear that the maritime seascape is a rich area with plenty of opportunities to enhance our knowledge. Such knowledge could be enlarged further if, when a construction takes place, archaeological plans are in place to obtain knowledge, even if results are negative. In such a case, negative results would allow us to label an area as poor in archaeological terms. Authorities and developers should be more aware that the absence of evidence is not evidence of an absence and that, subsequently, archaeological information (artefacts, ecofacts, structures…) can be found in unexpected places. Archaeology is full of lucky casual strikes around the world. Many times we are told that nothing is expected to be found in an area where redevelopment is taking place. However, in many of these cases, if an archaeologist had been on site, the recording of such redevelopment might have provided us with interesting data, even if such data is not going to alter what we know about the history of a place.

Conclusion It is clear that there is still much more to be done in maritime archaeology in Biscay, either underwater or on the foreshore, coastal or intertidal zones. Furthermore, there are some archaeological activities which, for some, are not recognised as maritime and they are presented, many times, as a land activity. However, they deal with coastal landscapes as they are on a cliff’s edge or some metres away from the current shoreline, currently in parks or beneath roads.

If we analyse the different projects discussed above, we can see that they can be divided into several minor topics. For instance, shipwrecks are dealt with in the following projects: Urbieta, Laida I and II, Bou Nabarra, 136

Maritime Archaeology in Biscay, Basque Country

Figure 13. View of the timber posts which makes part of the slipway; more posts were retrieved once the area was affected by caterpillars removing the soil; old riverbed is just a few metres away from the end of the postline, at the top of the picture.

Figure 15. Possible first futtock found during the watching brief of the Canal opening, from the area where most the timbers have been recovered; piece for a ship and related to a shipyard.

It is quite clear that shipwrecks are the most common topic/project. When people think about maritime archaeology, it is always easier to say underwater archaeology, and this means shipwrecks. Very few people understand that underwater archaeology is a part of maritime archaeology, a wider field where other subtopics are included, such as seascapes, harbour infrastructure, coastal landscapes, fish traps, slipways, piers, shipyards, submerged aircraft, etc. Shipwrecks are the most common subtopics discussed, and they are also well-represented here in Biscay, despite the lack of historical and archaeological information for many, particularly the most recent ones. This means that due to the lack of positive data (Gorliz sewage and Marea Project did not provide any), little is known about other maritime seascapes which can be identified. Despite the effort, again, not much is known about any other subtopic. For example, the fish weirs found are quite modern and they should be researched further to gain any other information. Furthermore more comparative material is needed to better understand fluvial fishing.

Figure 14. Floor frame found during the watching brief of the canal opening, retrieved from area where most the timbers have been recovered; clearly a piece from a ship and related to a shipyard.

other twentieth-century wrecks, Bakio and Laida Beach redevelopment. Projects such as Gorliz sewage, Lekeitio pier, Marea Project, Bilbao Fish weirs and Bilbao waterfront can be characterised as related to the theme ‘maritime seascape’. The Lekeitio pier and Bilbao waterfront can be categorised under the subject of harbour infrastructure. The Mendexa stone anchor might be part of a shipwreck, although it might be more related to maritime seascape as it is largely related to fishing. Fishing can be taken as the main theme for the Bilbao fish weirs. Finally, the intertidal archaeology project, run as a PhD project by the writer, has a bit of everything (wrecks, harbour infrastructure, maritime seascape). 137

José Manuel Matés Luque It is hoped that the intertidal project, which seems to have found some new fish weirs, could provide such information.

shipwrecks, or part of them, which should be recorded properly. The suggested steps are just the foundation for enhancing our knowledge on maritime archaeology in Biscay. If we all (recreational divers, authorities, bodies, professional archaeologists, amateurs) want to better understand our common maritime past, we need to work together since whatever is done by some will be useful for the others. If a diver finds an artefact or shipwreck, the authorities and bodies involved in management will have new sites which should be assessed when a new civil project might affect them. This way, the developer will know beforehand what to expect. Similarly, if a developer´s plan is going to affect a maritime place, contingency plans should be put in place to assess such a site. If nothing of archaeological interest turns up, the location can be declared sterile; on the other hand, if something is found, the developer will know how the finding will affect his/her project before it is too late, if proper contingency archaeological policies are applied to mitigate the impact. It is clear that working together is beneficial for all parties involved. Otherwise, maritime archaeology will suffer and nothing new will be achieved. Let us all be wise before it is too late.

Another important subtopic to have in mind is the redevelopment of harbours and ports, particularly their waterfront. As we have pointed out, the redevelopment in Bilbao has allowed us to understand how a riverine city has been developing throughout time. Obviously, the deeper and further away from the shore of the river, the further back in time one can delve. However, even if the redevelopment only takes place in the first metres from the waterfront, useful information can be obtained. Therefore, it is an urgent call to developers and authorities, either local, regional or national, to pay attention to such areas when designing new sewage system, pavement, walking paths, embankments or any other infrastructure which might provide an excellent opportunity, a unique window, to get even just a glimpse of what archaeological data has been buried there for ages. As a society, we cannot afford losing such an opportunity. Thus, Biscay has still a long way to go to achieve its full potential in maritime archaeology. There are several steps which should be put in place within a reasonable period of time, to reach stages where we can report that good management systems are being deployed and that they are working efficiently. Some of the steps suggested are as follows:

Acknowledgments The author would like to thank many people and institutions who have collaborated throughout the years to make these projects possible. Thanks are given to the Archaeological Body of the Biscay Regional Council for their support in different underwater projects. Similarly, thanks are due to the Department of Cultural and Language Policy of the Basque Government for supporting our project on intertidal archaeology. Finally, special thanks are due to Xabier Armendariz and Kiko Bañuelos. Thank you to the countless other collaborators that are, unfortunately, too numerous to list here. All errors are our own.

• Set up clear policies agreed by and with different authorities and bodies about the presence of an archaeologist at any civil engineering work which might take place on or near the shoreline (sea, river or lake). Unfortunately, sometimes it is thought beforehand that no archaeological artefacts or data will be retrieved. It should be stated that finding no archaeological data is useful data in itself. Thus, such statements about the ‘unimportance’ of a place to be redeveloped should be discarded, as many times artefacts are found or, in the event that no artefact is found, a negative response can be recorded. • Educate recreational divers on how important it is to respect underwater heritage resources and involve them in work so that their skills can be used for the benefit of everybody. This way, the next time they go diving in a site, they can have access to more information about a site provided by previous divers reporting on the monitoring of such sites. Working with divers has been successfully done in many different places around the world (Aguilar 2014; Pydyn and Flatman 2008). This idea of working with volunteers has also been performed with people on the foreshore with excellent results (Beattie-Edwards and Satchell 2011; Satchell et al. 2009). • Set up protocols for reporting finds which might be found unexpectedly either by recreational divers (see above) or by people walking along the coastline. It is not unusual that artefacts might turn up after storms on beaches. Many times such findings are unknown

Works cited AGUILAR, C. 2014. ‘La colaboración de las federaciones deportivas de buceo en la actualización de las Cartas Arqueológicas Subacuáticas y la protección del PCS. Arqueología subacuática españolaʼ. Actas del I Congreso de Arqueología Náutica y Subacuática Española. Cartagena, 14, 15 y 16 de marzo de 2013. II. Pp. 299–304. ALBERDI LONBIDE, X. and PÉREZ CENTENO, J.M. 2004. ‘Estructura lígnea de la playa de Ondarreta (Donostia-San Sebastián)ʼ. Arkeoikuska 2003. Pp. 149. ALBERDI LONBIDE, X. 1996. ‘Orígenes de la construcción naval en el Cantábrico a través de las representaciones iconográficas románicasʼ. Actas de las I Jornadas sobre Minería y Tecnología en la Edad Media peninsular. León 26–29 de Septiembre de 1995. Pp. 453–62. ALBERDI LONBIDE, X. and ARAGÓN RUANO, A. 1998. ‘La construcción naval en el País Vasco durante la Edad Mediaʼ. Itsas Memoria. Revista de Estudios Marítimos del País Vasco 2: 13–33.

138

Maritime Archaeology in Biscay, Basque Country ALTUNA, J. and APELLANIZ, J. Mª. 1976. ‘Las figuras rupestres paleolíticas de la cueva de Altxerri (Guipuzcoa)ʼ. Munibe XXVIII.

Edad Media. Encuentros Internacionales del Medievo. Nájera 27–30 de Julio 2004. Pp. 257–94. GÓMEZ BRAVO, M. and ZALLO USKOLA, J. C. 2012. ‘Evaluación de los pecios Laida I y II (Ría de Mundaka)ʼ. Arkeoikuska 2011. Pp. 258–62.

ARAGÓN NÚÑEZ, E. 2011. ‘Emisario submarino en Gorlizʼ. Arkeoikuska 2010. Pp. 243–44. BEATTIE-EDWARDS, M. and SATCHELL, J. 2011. The Hulks of Forton Lake, Gosport. The Forton Lake Archaeology Project 2006–2009, BAR British Series. BAR Publishing: Oxford.

GÓMEZ BRAVO, M. and ZALLO USKOLA, J. C. 2013. ‘Pecios Laida I y II (Ría Mundaka)ʼ. Arkeoikuska 2012. Pp. 190–91. GÓMEZ BRAVO, M. and ZALLO USKOLA, J. C. 2014a. ‘Pecios Laida I y II (Ría Mundaka)ʼ. Arkeoikuska 2013. Pp. 216–17.

BENITO, A.Mª. and MAZPULE, J. 2017. Naufragios en la costa vasca 1976–2016, Begira. Txertoa: Donostia. BENITO DOMÍNGUEZ, A. Mª. 2000. ‘Aproximación a la historia de las anclas líticas a partir de las recuperadas en la costa gipuzkoanaʼ. Itsas Memoria. Revista de Estudios Marítimos del País Vasco 3: 665–84.

GÓMEZ BRAVO, M. and ZALLO USKOLA, J. C. 2014b. ‘Nuevos hallazgos en la ría de Gernika-Mundaka (Bizkaia): los pecios Laida I y Laida IIʼ. I Congreso de Arqueología Náutica y Subacuática Española. Cartagena 14, 15 y 16 de Marzo de 2013. Pp. 809–20.

BERGANZA, E., ARRIBAS, J. L., and RUIZ IDARRAGA, R. 2012. ‘Estudio tecnológico de los moluscos marinos perforados de los yacimientos de Lumentxa y Santa Catalina (Lekeitio, Bizkaia)ʼ. MUNIBE (Antropologia – Arkeologia) 63: 93–104.

GÓMEZ BRAVO, M. and ZALLO USKOLA, J. C. 2015. ‘Pecios Laida I y II (Ría Mundaka)ʼ. Arkeoikuska 2014. Pp. 231–32. GÓMEZ BRAVO, M. and ZALLO USKOLA, J. C. 2016. ‘Pecios de Laida I y Laida II (Ría de Mundaka)ʼ. Arkeoikuska 2015. Pp. 211–12.

BERRAONDO, R. DE. 1932. ‘Sellos medievales de tipo navalʼ. Revista Internacional de Estudios Vascos XXIII. Pp. 130–38. CASABÁN, J. L. 2016. ‘The Wreck of the ‘Apostle’ San Bartolomé (1597)ʼ. The Mariner´s Mirrow 102.2: 206– 10.

GÓMEZ BRAVO, M. and ZALLO USKOLA, J. C. 2017. ‘Pecios Laida I y II (Ría de Mundaka) e inspección de los arenales de Laida y Txatxarramendi (Ibarrangelu, Sukarrieta)ʼ. Arkeoikuska 2016. Pp. 312–15.

CASADO SOTO, J. L. 1975. ‘Arquitectura naval en el Cantábrico durante el siglo XIIIʼ. Altamira. Revista del Centro de Estudios Montañeses I. Pp. 345–73.

GÓMEZ BRAVO, M. and ZALLO USKOLA, J. C. 2018. ‘Pecios Laida I, Laida II y Txatxarramendi (Ibarrangelu / Sukarrieta)ʼ. Arkeoikuska 2017. Pp. 328–31.

CASADO SOTO, J. L. 1995. ‘Euskal Herria eta itsasoa. Hasieratatik XVIII menderarte. El País Vasco y el mar. Desde los orígenes hasta el siglo XVIIIʼ. Itsas aurrean. El País Vasco y el mar a través de la historia. Pp. 16– 75.

GUTIÉRREZ ZUGASTI, F. I., CUENCA SOLANA, D., CLEMENTE CONTE, I., GONZÁLEZ SAINZ, C., and LÓPEZ QUINTANA, J. C. 2011. ‘Instrumentos de trabajo y elementos de adorno en conchas de molusco de la cueva de Santimamiñe (Kortezubi, Bizkaia)ʼ. La Cueva de Santimamiñe: Revisión y Actualización (2004–2006), Kobie. Bizkaiko Arkeologi Indusketak. Excavaciones Arqueológicas en Bizkaia. Bizkaiko Foru Aldundia. Diputación Foral de Bizkaia: Bilbao. Pp. 155–70.

CASADO SOTO, J. L. 2004. ‘Euskaldunen Bizkaiko Itsasoko urpeko arkeologiari buruzko ohar batzuk, espainar testuinguruan. Notas sobre la arqueología subacuática cantábrica vasca en el contexto españolʼ. Urpeko Oroimena. Euskal Herriko Urpeko Arkeologia Eta Ondarea. La Memoria Sumergida. Arqueología y Patrimonio Subacuático Vasco. Untzi Museoa-Museo Naval: Donostia-San Sebastián. Pp. 14–19.

IBÁÑEZ GÓMEZ, M., SANTANA EZKERRA, A., and ZABALA LLANOS, M. 1988. Arqueología industrial en Bizkaia. Eusko Jaurlaritza. Kultura eta Turismo Saila. Gobierno Vasco. Consejería de Cultura y Turismo. Universidad de Deusto. Deustuko Unibertsitatea. Deiker: Bilbao.

CASADO SOTO, J. L. 2006. ‘Barcos para la guerra. Soporte de la Monarquía Hispánicaʼ. Cuadernos de Historia Moderna. Anejos V. Pp. 15–53.

IZAGUIRRE, M. and VALDÉS, L. 1998. ‘Avance de excavación del pecio del siglo XV de Urbieta (Gernika)ʼ. Itsas Memoria. Revista de Estudios Marítimos del País Vasco 2: 35–41.

CHURRUCA, E. de, 1904. ‘Monedas romanas en la ría de Bilbaoʼ. Euskal-erria XXV. Pp. 210–12. CIRIQUIAIN GAIZTARRO, M. 1961. Los vascos en la pesca de la ballena. Biblioteca de Autores Vascos. Ediciones Vascas: San Sebastián.

IZAGUIRRE, M., VALDÉS, L., and MATÉS, J. M. 1999. ‘Avance de excavación del Pecio del s. XV de Urbieta (Gernika)ʼ. Arkeoikuska 1998. Pp. 392–98.

GARCÍA FERNÁNDEZ, E. 2005. ‘Las cofradías de mercaderes, mareantes y pescadores vascas en la Edad Mediaʼ. Ciudades y villas portuarias del Atlántico en la 139

José Manuel Matés Luque IZAGUIRRE, M., VALDÉS, L., MATÉS LUQUE, J.M., and PUJANA, I. 2001. ‘State of the excavation works of the 15th century shipwreck in Urbieta (Gernika, Spain)ʼ. Trabalhos de Arqueologia 18: 49–54.

MATÉS LUQUE, J. M. 2013c. ‘Muellesʼ. Arkeoikuska 2012. Pp. 146–52. MATÉS LUQUE, J. M. 2014a. ‘Documentando el crecimiento de la ciudad: arqueología marítima portuaria en los muelles de Bilbao (Bizkaia)ʼ. Arqueología Subacuática Española. Actas Del I Congreso de Arqueología Náutica y Subacuática Española. Cartagena, 14, 15 y 16 de Marzo de 2013. Servicio de Publicaciones de la Universidad de Cádiz: Cádiz. Pp. 375–85.

L´HOUR, M. and VEYRAT, E. 1989. ‘A mid‐15th century clinker boat off the north coast of France, the Aber Wrac’h I wreck: A preliminary reportʼ. International Journal of Nautical Archaeology 18.4: 285–98. LABURU, M. 1983. ‘El bote ballenero de Bermeo en el siglo XIII (Notas previas a la reconstrucción de un prototipo del bote ballenero similar al que aparece en el sello de Bermeo del año 1297)ʼ. Bermeo 3. Pp. 319–31.

MATÉS LUQUE, J. M. 2014b. ‘Muelle de Arriagaʼ. Arkeoikuska 2013. Pp. 194–96.

LABURU, M. 1987. ‘La construcción naval vasca en la Edad Media y el Renacimientoʼ. VI Cursos de verano en San Sebastián. VI Udako Ikastaroak Donostian. El medio físico y humano en la historia de Euskal-Herria. Pp. 173–202.

MATÉS LUQUE, J. M. 2015a. ‘El ancla lítica de tres orificios de la playa de Karraspio (Mendexa, Bizkaia)ʼ. Kobie, Paleoantropología 34: 283–99. MATÉS LUQUE, J. M. 2015b. ‘Estructuras de pesca de la ría de Bilbao, entre San Antón y La Peñaʼ. Arkeoikuska 2014. Pp. 196–98.

LABURU, M. 2006. De mare vasconum. La memoria perdida, Ensayo y Testimonio. Pamiela, PamplonaIruñea.

MATÉS LUQUE, J. M. 2016a. ‘Malecón de Lazunarriʼ. Arkeoikuska 2015. Pp. 236–43.

LEMA PUEYO, J. A. 2004. ‘La actividad industrial, la pesca y el comercio del País Vasco en la Edad Mediaʼ. Historia Del Pais Vasco. Edad Media (Siglos V-XV). Hiria: San Sebastián. Pp. 311–76.

MATÉS LUQUE, J. M. 2016b. ‘Playa de Laida. Seguimiento de su restauraciónʼ. Arkeoikuska 2015. Pp. 207–10. MATÉS LUQUE, J. M. 2016c. ‘Maritime archaeology as a tool to understand a river town: the example of Bilbao (Basque Country, Spain)ʼ. Actas del V Congreso Internacional de Arqueología Subacuática. IKUWA V. Un patrimonio para la humanidad. Cartagena, 15–18 de octubre de 2014. Pp. 781–90.

MARTÍN-BUENO, M., IZAGUIRRE, M., CASADO, J. L., MEJUTO, R., and SENÉN LÓPEZ, F. 1985. ‘La arqueología subacuática en las costas del Norte y Noroeste peninsular: Estado de la cuestiónʼ. VI Congreso Internacional de Arqueología Submarina. Cartagena 1982. Pp. 33–58.

MATÉS LUQUE, J. M. 2016d. ‘Ancla lítica de tres orificios, en la playa de Karraspioʼ. Arkeoikuska 2015. Pp. 246–49.

MATÉS LUQUE, J.M. and ARMENDARIZ, X. 2018. ‘Arqueología subacuática en Bizkaia: 35 años despuésʼ. Actas de las IX Jornadas de Jóvenes en Investigación Arqueológica, Santander 8–11 junio. JIA 2016. Pp. 533–42.

MATÉS LUQUE, J. M. 2016e. ‘Una traca a tingladillo hallada en la playa de Laida (Ibarrangelua, Bizkaia). Un nuevo pecio medieval en el País Vascoʼ. Kobie (Serie Paleoantropología) 35. Pp. 205–24.

MATÉS LUQUE, J. M. 2005. ‘Pecio de Bakioʼ. Arkeoikuska 2004. Pp. 370–73.

MATÉS LUQUE, J. M. 2018a. ‘Cable de telecomunicaciones entre USA y la playa de Arrietara (proyecto Marea)ʼ. Arkeoikuska 2017. Pp. 316–17.

MATÉS LUQUE, J. M. 2006. ‘Pecio de Bakioʼ. Arkeoikuska 2005. Pp. 95–99. MATÉS LUQUE, J. M. 2012. ‘Arqueología subacuática en Bizkaia. El pecio de Bakio: investigación y puesta en valorʼ. Actas de las Jornadas de Arqua, Cartagena 3 y 4 de Diciembre 2011. Pp. 131–38.

MATÉS LUQUE, J. M. 2018b. ‘Arqueología intermareal en Bizkaiaʼ. Arkeoikuska 2017. Pp. 319–27. MATÉS LUQUE, J. M. 2018c. ‘Arqueología marítima en la ría de Bilbao: cuando los árboles dejan ver el bosqueʼ. La Ribera de Deusto y Zorrotzaurre en la Ría de Bilbao. Análisis urbano, sociocultural y estético del espacio marítimo. Maritimidades 5. Museo Marítimo Ría de Bilbao. Bilboko Itsasadarra Itsas Museoa: Bilbao. Pp. 299–328.

MATÉS LUQUE, J. M. 2013a. ‘¿Patrimonio sumergido=patrimonio escondido? El panel interpretativo del pecio de Bakio como ejemplo de concienciación social de un patrimonio que no ha de esconderseʼ. Arqueología para el s. XXI: Actas de las V Jornadas de Jóvenes en Investigación Arqueológica. Santiago de Compostela, Mayo 2012. JAS Arqueología: Madrid. Pp. 250–55.

MERINO, J. M. 2003. ‘Pesca artesanal de angulas desde “zotalak” ʼ. Itsas Memoria. Revista de Estudios Marítimos del País Vasco 4. Pp. 651–56.

MATÉS LUQUE, J. M. 2013b. ‘Kurlutxuʼ. Arkeoikuska 2012. Pp. 206–07.

NAYLING, N. and SUSPERREGUI, J. 2014. ‘Iberian dendrochronology and the Newport medieval shipʼ. 140

Maritime Archaeology in Biscay, Basque Country International Journal of Nautical Archaeology 43. 2: 279–91.

SOBERÓN, M., PUJOL, M., LLERGO, Y., RIERA, S., JULIÁ, R., and DOMÍNGUEZ, M. 2012. ‘El Barceloneta I. Una embarcación medieval a tingladillo en Barcelonaʼ. Itsas Memoria. Revista de Estudios Marítimos del País Vasco 7. Pp. 363–71.

ODRIOZOLA OYARBIDE, L. 2002. Construcción naval en el País Vasco, siglos XVI-XIX. Evolución y análisis comparativo. Diputación Foral de Gipuzkoa. Departamento de Economía y Turismo: San Sebastián.

SOLÓRZANO TELECHEA, J. A. 2005. ‘Élites urbanas y construcción del poder concejil en las cuatro villas de la costa de la mar (siglos XIII-XV)ʼ. Ciudades y villas portuarias del Atlántico en la Edad Media. Encuentros Internacionales del Medievo. Nájera 27–30 de Julio 2004. Pp. 187–230.

ODRIOZOLA OYARBIDE, L. 2003. La construcción naval en Hondarribia. 1203–2002. Ayuntamiento de Hondarribia: Hondarribia. PÉREZ CENTENO, J. M. and ALBERDI LONBIDE, X. 2006. ‘Nasa de Ondarreta (Donostia-San Sebastián)ʼ. Arkeoikuska 2005. Pp. 469–70.

TORRES GOIRI, M. 1992. Aproximación a la historia del salvamento de náufragos en el Puerto de Bilbao. EDIGETXO. Aula de Cultura de Getxo: Getxo.

PUJOL i HAMELINK, M. and SOBERÓN RODRÍGUEZ, M. 2011. ‘El pecio de Barceloneta I, una embarcación atlántica del siglo XV en Barcelonaʼ. Actas de las Jornadas de ARQUA 2011. Cartagena, 3 y 4 de diciembre de 2011. Pp. 117–23.

VILLARREAL DE BERRIZ, P. B. 1973. Maquina hidraulicas de molinos y herrerias y govierno de los arboles y montes de Vizcaya. Sociedad Guipuzcoana de Ediciones y Publicaciones de la Real Sociedad Vascongada de los Amigos del País. Caja de Ahorros Municipal de San Sebastián: Madrid.

PYDYN, A. and FLATMAN, J. (eds.), 2008. Collaboration, communication and involvement: maritime archaeology and education in the 21st century. University Press. Nicolaus Copernicus University: Torun.

VV. AA. 1990. Pedro Bernardo Villarreal de Berriz. Euskaldunon ekarpena XVIII,eko injeniaritzari. La aportación vasca a la ingeniería del XVIII. Bizkaiko Foru Aldundia. Kultura Saila. Diputación Foral de Bizkaia. Departamento de Cultura: Bilbao.

RIETH, E. 2006. ‘L’épave d’Urbieta (Gernika): une embarcation à clin du milieu du XVe siècle. Étude préliminaireʼ. Itsas Memoria. Revista de Estudios Marítimos del País Vasco 5: 603–16.

ZALLO USKOLA, J. C. and GÓMEZ BRAVO, M. 2012. ‘Evaluación arqueológica de los pecio Laida I y Laida II (Ría de Mundaka, Bizkaia)ʼ. Kobie, Paleoantropología 31: 207–26.

RIETH, E. and IZAGUIRRE, M. 2004. ‘Urbietako Erdi Aroko pezioa (Gernika). El pecio medieval de Urbieta (Gernika)ʼ. Urpeko Oroimena. Euskal Herriko Urpeko Arkeologia Eta Ondarea. La Memoria Sumergida. Arqueología y Patrimonio Subacuático Vasco. Untzi Museoa-Museo Naval: Donostia-San Sebastián. Pp. 142–51.

ZALLO USKOLA, J. C. and GÓMEZ BRAVO, M. 2016. ‘Nuevas embarcaciones a tingladillo en la ría de Gernika (Bizkaia)ʼ. Actas del V Congreso Internacional de Arqueología Subacuática. IKUWA V. Un patrimonio para la humanidad. Cartagena, 15–18 de octubre de 2014. Pp. 757–68.

ROSELLÓ IZQUIERDO, E. and MORALES MUÑIZ, A. 2011. ‘Evidencias de pesca en las ocupaciones de Santimamiñeʼ. La Cueva de Santimamiñe: Revisión y Actualización (2004–2006). Kobie. Bizkaiko Arkeologi Indusketak. Excavaciones Arqueológicas en Bizkaia. Bizkaiko Foru Aldundia. Diputación Foral de Bizkaia: Bilbao. Pp. 239–46. RUIZ DE LA PEÑA, J. I. 1998. ‘La apertura de la fachada costera cantábrica a las rutas de la navegación atlántica (siglos XII-XIII)ʼ. Actas del II Congreso Internacional de Estudios Jacobeos. Rutas atlánticas de peregrinación a Santiago de Compostela. Ferrol, Septiembre 1996. Pp. 19–31. SATCHELL, J., BEATTIE EDWARDS, M., MADDOCKS, J., and HARVEY, M. 2009. Forton Lake Archaeologiy Project. Forton´s Forgotten Fleet. NAS. HWTMA: Gosport. SOBERÓN RODRÍGUEZ, M. 2010. ‘El port baixmedieval de la ciutat de Barcelona: una visió des de l´arqueologia. L´escullera de 1477 i la troballa d´un vaixel tinglatʼ. Quarhis II. Pp. 134–63.

141

13 Pioneers of Maritime Activity: The Uses and Abuses of the Maritime Aspects of Phoenician Culture Lamia Sassine The University of Sheffield [email protected] Abstract: The Phoenicians were famously one of the most celebrated maritime populations of the ancient world. In fact, their seafaring reputation was such that this aspect of their identity has been appropriated or manipulated by many. This paper seeks to investigate the perceptions of Phoenician maritime activity, from those of their Mediterranean and Near Eastern Empires contemporaries to different nationalities in nineteenth-century Europe, and the modern countries that were part of its realm. It raises the question of the importance of the maritime aspect of Phoenician identity from a Mediterranean and long-historical perspective, in terms of commerce and exchange as well as in terms of imperialism and the idea of ‘civilising populations’. It also addresses the causes and consequences of identifying with specific cultural traits and the implications of this in terms of stereotyping and bias. Key words: Phoenicians; Phoenicia; identity; appropriation; maritime; perception Introduction

One of the most potent stereotypes linked with Phoenician culture is that they are associated with almost legendary seafaring. This is a crucial idea because archaeological evidence has suggested maritime exchange had been established in the Mediterranean since the Bronze Age, that is at least two millennia before the time of the Phoenicians, as we believe we know them. No earlier culture has been as blatantly celebrated in terms of its maritime activity. This paper, thus, will seek to investigate the origins of this association and gain a better understanding of the perceptions of Phoenician maritime activity, and their consequent implications. The paper will first give an overview of the sources available regarding Phoenician seafaring, then move on to discuss the perspectives of empires regarding this particular aspect. This will ideally set the stage for a discussion on the importance of commerce, exchange, and imperialism within a maritime context, and, finally, the specificities of the Mediterranean context. To conclude, this paper will address the question of the causes and consequences of identifying with specific cultural traits and the implications of this in terms of stereotyping and bias.

Ever since Antiquity, the Phoenicians have been a population celebrated for their navigational skills. The term Phoenician has many different implications, as it is still a relatively misunderstood and understudied culture. However, it is most commonly agreed upon that the Phoenicians were the Iron Age (1200–333 BC) population of the central Levantine coast, stretching north from Arwad and south to the region around Dor. The Phoenicians also established colonies across the Mediterranean; a part of which, under the influence of Carthage, develops a distinct material culture starting from the seventh century BC and is usually referred to as Punic (Sommer 2007: 103). For the sake of convenience and ease of communication, I will use the word ‘Phoenician’ in this paper to designate both Phoenician and Punic cultures. Phoenician culture is one that has been defined and conceptualised in a number of ways; transmitted mainly from foreign sources and exacerbated to the point of becoming stereotyped. For instance, the Phoenicians are believed to have invented the alphabet. While it is true that they developed it significantly and furthered its diffusion, the Phoenician alphabet emerged from a series of earlier scripts (Markoe 2005: 111). Similarly, early studies on the Phoenicians in the nineteenth century by scholars such as Rawlinson (1889), Renan (1864), and Perrot et al. (1884) circulated the idea that the Phoenicians were greatly skilled craftsmen, but that their art was mostly an art of ‘borrowing’. These thoughts persist today and they are reflected in many museum displays and more recent studies on the Phoenicians.

The sources The internal sources for the study of Phoenician seafaring are scarce and scattered. Up until the fifth century BC and the beginning of the use of coins, most representations of Phoenician ships come from foreign sources, the majority of which are found on Assyrian reliefs (Sauvage 2007: 97). These reliefs allow us to differentiate between two types of ships: commercial ships and warships. Commercial ships, attested on representations since the ninth century BC, 143

Lamia Sassine are usually symmetrical boats featuring hippoi on either one or both extremities (Sauvage 2007: 94). Depictions of warships are also attested from Assyrian reliefs since the eighth century BC. Warships differ from commercial ships in that they are much longer and consisted of two decks, with the majority of the crew assigned to rowing on the lower deck, a curved bow and a stern featuring a ram that was often covered in bronze (Bartolini 1995: 287). However, external representations such as Assyrian reliefs have to be considered carefully and cannot be taken for granted as often, the sculptors would be relying on descriptions rather than first-hand encounters with the ships (Sauvage 2007: 97). In fact, it has been suggested that the wall-reliefs of Korshabads, which have been used to reconstruct Phoenician ships, depicted fluvial rather than maritime scenes (Linder 1986).

studies, most authors have mentioned the famous maritime aspect of Phoenician identity, but there is surprisingly very little literature on the implications of this association. The following paragraphs will move away from background information on Phoenician navigation and focus instead on attitudes towards it. Empires and the sea We have seen that much of the information on Phoenician seafaring was transmitted through classical authors. We have also seen that these authors credited the Phoenicians with the invention of navigation altogether. Knowing that Greeks and, later on, Romans fostered complicated relationships with Phoenicia, to say the least, one realises that this is no ill feat. In fact, the Greeks evolved in the same Mediterranean network with similar commercial and colonial agendas, and it is, therefore, quite remarkable that they would praise the Phoenicians for an aspect of their culture they were in direct competition with. The same can be said of the Romans, who were repeatedly at war with Carthage. Although much older seafaring is archaeologically attested, the fact that the Phoenicians were given credit for it by Greek and Latin authors must indicate that there must be some intrinsic basis for their doing so. It could be that the term ‘Phoenician’ as used by classical authors included a larger time frame than the usual scholarly consensus, and encompassed the Bronze Age populations of the central Levant, as well. It could also be that the Iron Age Phoenicians did indeed take seafaring to new levels. In any case, this credit given to them by classical authors seems to be one of the main sources of the association of Phoenicians with seafaring.

There are some native representations of Phoenician ships, which, when compared to foreign sources such as the aforementioned reliefs and Egyptian paintings, have been able to generate respectable interpretations (Ruiz Cabrero 2007: 98). Examples of these native sources include cave paintings found in Spain and Sicily (Lopez-Bertran et al. 2008), and representations of ships on coins. Recently, Iron Age shipwrecks were discovered at Mazarron and Bajo de la Campana on the Spanish coast, as well as off the coasts of Ashkelon and Malta, which also helped contribute to the growing knowledge of Phoenician ships. In addition to the aforementioned evidence, scholars have strongly relied on literary sources to learn about Phoenician navigation. Despite the fact that some ten thousand Phoenician inscriptions are attested, most of them are dedicatory or funerary in nature. Therefore, there is surprisingly little terminology related to navigation in the surviving Phoenician inscriptions, and while there are some terms that seem to designate captains or seamen, there is a lack of attested terms referring to ships themselves (Ruiz Cabrero 2007: 93). As with the iconography, foreign sources are more explicit. In the Bible, the city of Tyre is compared to a Phoenician ship by Ezekiel in his prophecy (Bartoloni 1995: 285). This source has proven particularly useful in terms of retracing cargoes and building techniques.

Let us now take a leap forward in time to nineteenthcentury Europe. In the early days of archaeology, there were diverging attitudes towards the Phoenicians. On the one hand, British opinion was quite favourable, and perhaps even enthusiastic (Bernal 1987: 337). In fact, Britain, in its imperial ideology, was drawing parallels between its naval and industrial expansion and that of the Phoenicians (Champion 2001: 456). The English saw themselves almost as spiritual descendants from the Phoenicians, who Rawlinson called ‘the great pioneers of civilisation’ (1889: 552). These views offer some insight into the mindset of the English at the time: the Phoenicians were perceived as peaceful traders who ‘civilised’ the rest of the Mediterranean (and, by extension, the world). This was a convenient parallel to draw with British Empire which was expanding at the time, even going as far as claiming that the Phoenicians had reached the Cornish coast: ‘In the ensuing debate, it was suggested that the Phoenicians had been responsible not only for clotted cream, but also for the use of saffron in cakes in Cornwall’ (Champion 2001: 461).

Homer also mentions Phoenician and Sidonian seafarers in his epics. Additionally, many classical authors such as Strabo, Herodotus, Pliny and Thucydides have written about Phoenician seafaring and even credited the Phoenicians with the invention of navigation altogether (Bikai 1990: 97). I believe these sources are at the origin of the attribution of a maritime identity to the Phoenicians. Many of the studies previously done on Phoenician maritime activity have been concerned with the religious and ritualistic aspect of it (Brody 1998, 2005; Christian 2013; Ruiz Cabrero 2007). Others have been more general studies about the typologies of ships and technicalities of navigation in the Iron Age (Bartoloni 1995; Markoe 2005; Sauvage 2007). Ever since the early days of Phoenician

However, there are some contradictions to address with the identification of the British with the Phoenicians. First of all, there is the fact that the British, with their imperialistic ideology, also sometimes compared themselves to the 144

Pioneers of Maritime Activity Romans, whom they viewed as having civilised the primitive British populations. With the Romans being well-known enemies of the Carthaginians, this was a rather uncomfortable oxymoron. Another issue was that of the interlinked rise of Romanticism, Orientalism and Anti-Semitism. Phoenician religion, especially, was perceived as barbaric, and this did not exactly suit the political interests of Britain (Champion 2001: 460). These contradictions contributed to the decline of Britain’s association with the Phoenicians, alongside new archaeological studies in British prehistory where the native inhabitants of England started to be perceived under a more flattering light (Champion 2001: 462). Champion concludes his study with the claim that one does not need to claim descendance to identify with a people. Rather, the Phoenicians in nineteenth-century Britain were perceived as ‘role-models’ (Champion 2001: 462), and the aspect of their identity that was highlighted was the maritime one because it was suitable to Britain’s agenda. This statement applies clearly to this case study, but as we will see later, descendance is a great part of the modern Lebanese identification with the Phoenicians.

Phoenicians. Romanticism, for instance, was particularly well-suited to anti-Semitic views because it placed central importance on northern white cultures, praising highlands landscapes and snowy mountains. What Bernal calls ‘the Aryan model’ is very much linked to this idea of northern Indo-European origins, and henceforth a complete denial of any attachment to Semitic populations. In the face of the British identification with the Phoenicians, France identified with the Roman Empire. This illustrates the historic rivalry between the French and the British, echoed in their respective appropriations of Roman and Phoenician identities. In perceiving themselves as a renewed form of the Roman Empire, the French had in mind to conquer the whole of the Mediterranean. Archaeology comes into play here because it was used to relate modern populations to ancient ones, and hence associate the populations France was conquering to those that had been conquered by the Romans. Naturally, the people of Lebanon became linked to the Phoenicians, and, by extension, to all the pejorative qualities associated with them, stemming as we have just seen from classical and biblical perceptions, romanticism, orientalism and the anti-Semitic climate. This is especially clear in Renan’s volume, where the locals are depicted as having no interest whatsoever for science, the higher arts or any form of culture. According to him, the Semites were a hopeless case, a good-for-nothing people, the same as their barbaric ancestors.

On the other side of the Channel, perceptions of the Phoenicians in the nineteenth century were quite different. However much anti-Semitism may have contributed to the depreciation of the Phoenicians in Britain, it had a significantly higher impact in continental Europe (Bernal 1987: 337). Anti-Semitism was stronger in France because of their historic rivalry with Britain (Bernal 1987: 363). Phoenician was already a deciphered language, and it was evident that it was very closely related to Hebrew. Moreover, the Jews were also a population with a strong history of trading and commerce, which also worked in favour of their being associated with the Phoenicians. Let us stress once again that the main sources for the study of the Phoenicians at the time were the classical authors and Homer. Taking the perceptions of the Phoenicians in these sources and applying them to their own experience of the Jews, Europeans therefore amplified the Phoenician/ Semitic stereotypes. Particularly emphasised were notions of an ‘other’ different, uncivilised, barbaric, opportunistic and corrupted population. In this attitude, the Phoenicians became merged with the Jews and the anti-Semitic atmosphere meant that they were widely disregarded or rather mis-regarded. In fact, according to Bernal, the fall of anti-Semitism and the aftermath of World War II played a major role later on in the re-establishment of Phoenician studies (Bernal 1987: 400). Before we move on though, it is important to remember that the Phoenicians were heavily condemned in the Bible and that the Jews themselves never identified with them. Moreover, Bernal’s historiography does not take into consideration the very important and condescending role played by later Biblical archaeology and its effect on perceptions of the Phoenicians (Bikai 1990). The anti-Semitic attitudes are therefore a reflection of the Euro-centrist model that prevailed on most of the perceptions of the Phoenicians for a long time in the nineteenth and early-twentieth centuries. Some artistic and literary currents also affected the perceptions of the

No matter how dismissed the Phoenicians were though, most scholars could not completely deny their influence in the ancient world. As a consequence, some scholars plainly refused to attribute a Semitic character to the Phoenicians. Renan, for instance, opposed the characteristics of the ‘civilised’ ancient Phoenicians to his contemporary Semites; he viewed the Phoenicians as savage, in a sense (Liverani 1998: 6). In the early-twentieth century, some scholars such as Autran joined him in advocating a nonSemitic identity for the Phoenicians (Liverani 1998: 14). These perceptions have undoubtedly contributed to the rise of Phoenicianism, a political movement founded in Lebanon in the 1920s. The founders of Phoenicianism were young, French-speaking, Maronite Christian entrepreneurs, majoritarily educated at the Jesuit Universite Saint Joseph. The Jesuits played a major role in ties with France and the subsequent establishment of Phoenicianism (Kaufman 2001: 173). One could trace back the idea of Phoenician heritage in the modern population of Lebanon in the writings of clergymen in the nineteenth century, with the earliest association attested in the writings of Tannus Al-Shidiaq in 1859 (Quinn 2018: 4). Some of these authors were heavily influenced by French researchers and adopted the views that the Phoenicians were early pioneers of civilisation. In fact, Renan reports that some Maronite populations he encountered in the region of Byblos were ‘shocked’ to learn there were nonChristian populations established in the region before them (Renan 1964: 217). These ideas were initially more 145

Lamia Sassine popular within the Lebanese diaspora, helped by foreign perceptions. For example, the Lebanese US immigrants were called ‘Christian descendants of the Phoenicians’ in 1911 (Quinn 2018: 4). However, it is not until the creation of Phoenicianism that the Phoenician identity was used in adversity to Arab identity. The 1920s witnessed the early days of pan-Arabism, and Phoenicianism was initially a political reaction to this climate. Phoenicianists claimed that they were descended directly from the Phoenicians (who had converted to Christianity as soon as the opportunity presented itself to them), and therefore met the French ideology denying a Semitic origin, both for themselves and for their ancestors. Phoenicianists such as Al-Saouda referred to Lebanon as ‘the cradle of civilisation’ and contributed to the separation of Lebanon from Syria by claiming this distinct heritage (Kaufman 2001: 181; Quinn 2018: 7). Unlike the British, though, the Lebanese who started Phoenicianism claimed to be genetically descended from the Phoenicians. From then on, Phoenician identity became a staple of Lebanese identity, and Phoenician ships were featured on early Lebanese coins. This heritage is still deeply rooted in the minds of many Lebanese to this day.

little is known about it, but it has played a crucial role in the manipulation of Phoenician identity over time and space. The first aspect in which this maritime aspect was instrumental is in terms of the reification of commerce and exchange. The Phoenicians have been viewed by many as the founders of large-scale commerce. Even the French, who did not necessarily always have favourable views towards them, created visuals depicting them as pioneers of trade. In the early 1900s, a brand of packaged meat was giving out collectable cards in series, one of which was on the evolution of trade and commerce and featured a card with Carthaginian merchants. This aspect has also served the British empire in its association with the Phoenician culture, as they also saw themselves as traders linking a variety of networks together. The Phoenician commercial network was established in the broad Mediterranean world, whereas the British one was larger but still included a Mediterranean core. Similarly, the founders of Phoenicianism were not scholars but businessmen, and they especially recognised themselves in the great traders that were the Phoenicians. They saw the Phoenicians as ‘above all men of the sea, sailing as far as Great Britain, liberal, peaceful; bringing the world civilisation, commerce, and industry’ (Quinn 2017: 7). Coincidentally, this view is actually very close to the British perspective, with the subtle difference that the British saw the Phoenicians more as role-models while Phoenicianists considered them to be their genetic ancestors. In the mercantile sense though, the rationale was the same. Phoenicianists considered trade to be a skill engraved in their genes, transmitted to them by their celebrated ancestors. Lebanon remains a tax haven to this day, and its people are still proud of the fact that their supposed ancestors were the largest traders in Antiquity. Interestingly enough, this view seems to disregard at least some of the more negative perceptions of the Phoenicians as traders in classical sources. In Homer, for instance, the Phoenicians are often seen as crooks and slavers, but as it is often the case with nationalism, local memory is selective.

Interestingly enough, Tunisia had a similar form of identification with the Phoenicians, which came as a reaction to the French colonisation. As we have seen, the French identified with the Romans, historical rivals of the Carthaginians. The French colonial policy over there was much more aggressive than it was in Lebanon, with much more emphasis put on the legitimacy of the colonisation justified by the French perception of their empire as a revival of the Roman Empire (Quinn 2018: 12). Therefore, the natural reaction of the Tunisians was to associate with the Carthaginians and fight, in a sense, their Fourth Punic War. Tunisians used this image to promote their independent post-colonial identity and, as in Lebanon, Phoenician ships have become a symbol represented on Tunisia’s national emblem (Quinn 2017: 12). Although nowadays both Lebanon and Tunisia acknowledge a mosaic of cultures that have shaped their national identities (and have done so since the early days of independence), the Phoenician aspect of it has been a strong marker in both countries. In Lebanon, even after the demise of Phoenicianism, Phoenician heritage is still used to assert difference with the rest of the Arab world (Quinn 2018: 11). It comes as no surprise that the modern countries with the strongest identifications to the Phoenicians were the ones where the most important centres were located, Lebanon covering roughly 90 per cent of the ancient Phoenician homeland, and Tunisia home to Carthage, which gained considerable importance in the ancient history of the Mediterranean, establishing its own colonies.

Another aspect that has been used mostly by the British is what I call the ‘civilising movement’. In the minds of the nineteenth-century British, the Phoenicians sailed across the Mediterranean not only to gain access to resources and enrich themselves but, more importantly, to bring the greatest gift to the ‘primitive’ natives: civilisation. This was symbolised by the myth of Kadmos, a Phoenician prince offering the alphabet to the Greeks. This aspect of the maritime identity of the Phoenicians was especially important to imperial Britain because, through this parallel, Britain could justify its colonising movement. Just like the Phoenicians, the British were seen as civilising every nation they conquered. This model was later rejected with the rise of Helleno-centrism and Anti-Semitism (Bernal 1987), but it still served colonial Britain for a while, even making it to its national iconography, as illustrated by Leighton’s painting ‘Phoenicians trading with Ancient Britons on the coast of Cornwall’ in the Royal Exchange.

Why water matters From British imperial ideology to post-colonial Phoenicianism, all of the above perceptions and identifications seem to stem from a common root: the maritime nature of the Phoenician people. Ironically, very 146

Pioneers of Maritime Activity There was also a proposal for a series of paintings, one of which was a comparison of Phoenicians on the coast of Cornwall with Captain Cook in Tahiti (Champion 2001: 457). Interestingly enough, this aspect of having been pioneers of civilisation is also reflected in post-colonial countries. It forms a strong part of national identity in Lebanon, where stamps showing the Phoenicians as having disseminated knowledge and civilisation around the world are often released. This is just one example of the ways in which this image has been used in the construction of post-colonial national identities. It is even more blatant in popular culture, where every so often social media posts claiming that ‘you are reading and commenting on this post cause the Phoenicians, our ancestors, taught you how by creating the alphabet’ surface in the feeds of Lebanese users. This idealised view of the Phoenicians is directly linked to their seafaring: without it, they would never have been able to civilise the rest of the world. Whether these perceptions are a reaction to colonialism, in the sense that third-world countries felt the need to claim that they did not need to be ‘civilised’ since their ancestors had been through the reverse process before, or whether they are a consequence of stereotypes and selective memory remains to be examined in more depth.

and appropriated as navigation by both foreign and local cultures. In addition to the reasons cited above, this also has to do with the agency of ships. In Phoenician times, ships and navigation were linked to a series of specialised rituals (Brody 1998, 2005; Christian 2013). The vessels were often connected to a deity to the extent that they almost had their own personality. This special relationship between sailors and their ships is something that has been attested throughout history. It seems, therefore, that any population with a link to seafaring could relate to the Phoenicians on an almost emotional level. Moreover, the fact that boats were and still are represented on national artefacts such as coins is yet another indicator of their agency and symbolic significance. This aspect is actually reflected in all the cultures that have identified with the Phoenicians, or contributed in some way to the fabrication of Phoenician identity, starting with the Greeks, who encountered other seafarers, first in the Aegean and later all over the Mediterranean, and constructed their own identity against that of those seafarers. The British also have a strong naval culture and were, therefore, able to easily connect with the Phoenicians. Even before the British appropriation, the Dutch have been linked to the Carthaginians because of their maritime commercial activities (Quinn, 2017). In short, the Phoenicians, probably because the classical authors credited them with the invention of navigation, have always been associated with the sea. They became the staple of a maritime population, and any other group with a connection to maritime activity could naturally identify with them.

Let us now consider the setting of the maritime activity of the Phoenicians: the Mediterranean, which is important for its many significances. In terms of scale, it can be argued that the Mediterranean is a mosaic of cultural entities linked together on different levels. It has always been perceived as a connecting entity between the three old continents, and, thus, has always been a coveted centre of power. Whoever had control over the Mediterranean could benefit from an advantage of resources, trade routes and, consequently, economic control of what has always been one of the world’s most important networks. The prevalence of the Phoenicians in this network is that they have been perceived as its earliest unifier. They managed to create a cultural entity connected from East to West for the first time in history, and this is why the maritime aspect of their culture held such a primordial place in many later civilisations. Although exchange in the Mediterranean predates them, the Phoenicians are seen as its first conquerors in terms of establishing trading posts all over the sea. In the classical sources, we can pinpoint this aspect very clearly. Some Greek authors felt indebted to the Phoenicians because of the transmission of the alphabet. The Roman perspective might be linked to this as well, but it seems that the admiration from the part of the Roman authors also comes from the fact that the Phoenicians were seen as the first unifiers of the Mediterranean, having established colonies on all its shores and beyond, and that this feat was greatly admired by the later inhabitants of mare nostrum.

Conclusion: Beyond the sea In this paper, I have tried to go over some of the uses of the maritime aspect of Phoenician culture. I have started with an overview of the sources for the study of Phoenician seafaring, which are majoritarily external due to the scarcity of Phoenician evidence, both in the literature and in the material culture. I then looked at the perceptions of Phoenician maritime identity, first forged by the classical authors, and then appropriated by the British in the nineteenth century. I have also examined those perceptions in nineteenth-century Europe and in post-colonial Lebanon and Tunisia. Finally, in the last part of the paper, I have tried to explain where these perceptions might stem from with a focus on commerce, the civilising movement, the significance of the Mediterranean and, finally, the agency of ships themselves. Let us keep in mind that many of the people we call Phoenicians are likely to have never boarded a ship, and the seamen that are so celebrated were most probably a minority. When the Greeks met them in the Aegean, they began to construct the identity of the only proportion of the Phoenician population that was visible to them (and given evidence from earlier shipwrecks, it is quite likely that these sailors did not all originate from the same place). It may very well be that this reduced segment of the population fits the mould shaped by later perceptions, but it is by no means a representative sample of Phoenician

The final point I would like to raise is that of the symbolic significance of ships. As mentioned in the introduction, the Phoenicians were famous for deeds other than their seafaring; including craftsmanship and the diffusion of the alphabet. However, these have not been as highly celebrated 147

Lamia Sassine culture overall. The issue with this is that most of our current ideas about the Phoenicians have been shaped around perceptions of the seafaring segment of the population. While I am not denying that seafaring must, of course, have been a primary aspect of Phoenician culture, I am raising the question of representation and generalisation, especially given the flimsiness of the definition of ‘Phoenician’ itself. Therefore, I believe it is important to consider the implications of appropriating a single aspect of a culture. Selective identification can serve many purposes, but in the end, it might be doing more harm than good, as it can very much increase stereotypes as well as cultural cleavage. The sea is often seen as a connecting entity, bringing people together; but figuratively, the sea can also serve as a divider between one population. The history of the uses (and abuses) of Phoenician maritime identity may well, in fact, reflect one of the most overlooked sampling biases of archaeology.

LÓPEZ‐BERTRAN, M., GARCIA‐VENTURA, A., and KRUEGER, M. 2008. ‘Could You Take a Picture of my Boat, Please? The Use and Significance of Mediterranean Ship Representations’. Oxford Journal of Archaeology 27.4: 341–57. MARKOE, G. 2000. Phoenicians. British Museum Press: London. MOITY, M., RUDEL, M., and WURST, A.-X. 2003. Master seafarers: the Phoenicians and the Greeks. Periplus: London. PEÑA, V. (ed.), 2004. La navegación fenicia. Tecnología naval y derroteros. Encuentro entre marinos, arqueólogos e historiadores. Centro de Estudios Fenicios y Púnicos: Madrid. QUINN, J. C. 2017. In Search of the Phoenicians. Princeton University Press: Lawrenceville.

Works cited BARTOLONI, P. 1995. ‘Navires et navigation’. In V. Krings (ed.), La civilisation phénicienne et punique. E. J. Brill: Leiden. Pp. 282–89.

RAWLINSON, G. 1889. Phoenicia. Hardpress Ltd. RUIZ CABRERO, L. A. 2007. ‘La marina de los fenicios, de la creencia en la vida a las naves de la muerte’. Gerión 25.1: 91–119.

BERNAL, M. 1987. Black Athena: the Afroasiatic roots of classical civilization. Free Association Books: London.

SOMMER, M. 2007. ‘Networks of commerce and knowledge in the iron age: The case of the Phoenicians’. Mediterranean Historical Review 22.1: 97–111.

BIKAI, P. M. 1990. ‘Black Athena and the Phoenicians’. Journal of Mediterranean Archaeology 3.1: 67–76. BRODY, A. J. 1998. ‘Each man cried out to his God’: the specialized religion of Canaanite and Phoenician seafarers. Scholar’s Press: Atlanta. BRODY, A. J. 2005. ‘Further Evidence of the Specialized Religion of Phoenician Seafarers’. In J. Pollini (ed.), Terra Marique: Studies in Honour of Anna Marguerite McCann on the Receipt of the Gold Medal of the Archaeological Institute of America. Oxbow Books: Oxford. Pp. 177–82. CHAMPION, T. 2001. ‘The appropriation of the Phoenicians in British imperial ideology’. Nations and Nationalism 7.4: 451–465. CHRISTIAN, M. A. 2013. ‘Phoenician Maritime Religion: Sailors, Goddess Worship, and the Grotta Regina’. Die Welt des Orients 43.2: 179–205. DÍAZ-ANDREU GARCÍA, M. and CHAMPION, T. C. 1996. Nationalism and archaeology in Europe. UCL Press: London. LINDER, E. 1986. ‘The Khorsabad Wall Relief: A Mediterranean Seascape or River Transport of Timbers?’. Journal of the American Oriental Society 106.2: 273–81. LIVERANI, M. 1998. ‘L’immagine dei Fenici nella storiografia occidentale’. Studi Storici 39: 5–22.

148

14 Exploring Hominin Movement Patterns in the Lower Palaeolithic Aegean Dry Land: Methodological Challenges Peny Tsakanikou and John McNabb Centre for the Archaeology of Human Origins (CAHO), University of Southampton [email protected] Abstract: Recent work on the submerged Aegean landscapes challenges the current perception of this region as a barrier during the Lower Palaeolithic. Instead, exposed terrains during the Middle and perhaps during the Early Pleistocene offered attractive lands for occupation with viable pathways towards Western Europe, across a previously unrecognised area for potential hominin migration and settlement. The lost (now submerged) lands, joining Asia to Europe, emerge as a new research arena for the Lower Palaeolithic, holding cues for answering wider questions regarding the early settlement of Europe – from an ‘eastern gateway’ perspective. However, the Aegean is a highly dynamic region that provides only few windows of opportunity for studying systematically the Lower Palaeolithic evidence. The assessment of the archaeological implications is a quite complex task that requires novel methodologies. Topographic roughness and spatial analyses (GIS) are combined here to explore aspects of hominin presence, mobility and activity across the Middle Pleistocene Aegean. Keywords: Hominin dispersal, Aegean submerged landscapes, Middle Pleistocene, GIS Introduction

reason behind this paucity correlates with landscape dynamics and the active geomorphic processes that have been negatively affecting availability, accessibility and visibility of the Early and Middle Pleistocene material (Tourloukis 2010; 2016; Tourloukis and Karkanas 2012). However, other parameters should not be excluded. The paradox observed here i.e. great expectations for Lower Palaeolithic evidence vs weak Lower Palaeolithic signal, could also be the product of research and/or interpretation bias – the eastern and north-eastern Mediterranean until recently was perceived as a periphery rather than a core area during the Lower Palaeolithic. It could even reflect a sporadic and episodic hominin presence in the area during the Early and Middle Pleistocene.

The Aegean lies on a key-geographical position, located between Africa, Europe and Asia, in an area that has proved to be crucial for the survival and dispersal of populations during the Early and Middle Pleistocene. The wider eastern and north-eastern Mediterranean served as a refugium during the glacial periods (Tzedakis 2004), but also as a core area hosting source populations for the repopulation of the northern and western parts of Europe during the interglacials (Dennell et al. 2011; Hewwit 1999; Steward et al. 2010). The palaeofaunal record shows that the Balkans in particular operated as a reception and diffusion area for the dispersals towards Europe, providing multiple and multidirectional routes during the Pleistocene for several mammalian species (Kahlke et al. 2011; Kostopoulos and Koulidou 2015; Kostopoulos et al. 2002; 2007; MartinezNavarro 2010; Palombo et al. 2006). Recent evidence suggests that Europe may have been reached by hominins at – or slightly prior to – 1 Mya via the Peri-Pontic routes and the Bosporus passage (through Asia Minor), during a faunal westward expansion around 1.3–1.2 Mya (Koufos and Kostopoulos 2016; Spassov 2016).

The absence of evidence in the eastern part of Europe came in total contradiction with the rich Lower and Middle Pleistocene records from the Iberian and Italian Peninsulas in the west. Led by the Iberian evidence, the prevailing scenario for the early colonisation of Europe highlights the western part of the continent (Garcia et al. 2014; Oms et al. 2000; Toro-Moyano et al. 2013). Until very recently, the Aegean region remained out of the discussion, representing a hiatus on the Lower Palaeolithic map (e.g. Jöris 2014). Yet, recent finds from Rodafnidia site on Lesbos Island (Galanidou et al. 2013; 2016) and Marathousa 1 site in Megalopolis basin (Panagopoulou et al. 2015; Tourkoukis et al. 2018) provide unequivocal evidence for the presence of hominins in the Aegean region around – and possibly prior to – 400 Kya. Furthermore, recent ongoing research

Despite the high research potential, the Lower Palaeolithic signal from this part of Eurasia remains surprisingly low, with sparse archaeological and palaeoanthropological evidence, problematic stratigraphic contexts, extensive spatiotemporal gaps and dating inaccuracies (for a recent overview see Harvati and Roksandic 2016). The main 149

Peny Tsakanikou and John McNabb on submerged landscapes (Lykousis 2009; Sakellariou and Galanidou, 2016; 2017) revealed the existence of extended sub-aerially exposed landmasses with favourable conditions in the Aegean during the Middle – and possibly during the earlier parts of the – Pleistocene. The archaeological implications are enormous, raising questions in relation to the role of the Aegean during the early colonisation of Europe. Was the Aegean a barrier/ ‘cul de sac’ during the Middle (and Early?) Pleistocene or an extended terrestrial landscape with attractive habitats?

uplifted and subsiding blocks. The sea-level fluctuations, following the glacial-interglacial alterations during the Pleistocene and the Holocene, repeatedly changed the paleogeography of the Aegean, at times flooding the area, and at times exposing the seabed as dry land. The current configuration was established around 9 Kya (Lambeck 1996). Human populations that occupied the area, or were moving across this region diachronically, must have been drastically affected by the dynamic character of this landscape – either as a terrestrial landscape or as a seascape – in terms of adaptability and behavioural flexibility (Harff et al. 2016). This later aspect is examined here against the Aegean dry land hypothesis, proposed by Lykousis (2009) and further refined by Sakellariou and Galanidou (2016; 2017).

The Aegean dry land hypothesis is starting to have an important impact on archaeological thought beyond the Greek Palaeolithic. Several authors (e.g. Galanidou et al. 2016; Roksandic 2016; Strait et al. 2016), recognise and consider the existence of the exposed Aegean landscape as a dynamic agent in the history of movements during the Early and Middle Pleistocene. Strait et al. (2016: 76) specifically refer to it as the “Trans-Aegean pathway”, one of the possible dispersal routes during the Middle Pleistocene. Still, the idea of the Aegean as a traversable terrain is very generic and needs further exploration. Is it possible to trace hominin movement patterns across the Aegean dry land during the Middle Pleistocene?

In 2009 Vasilis Lykousis published a reconstruction of the Aegean for the last 400 ka, based on the identification of successive prograding sedimentary sequences (deltaic fans) laid down in glacial periods, in different parts of the Aegean (i.e. palaeo – shelf break glacial delta sediments), using seismic reflection profiles and sediment gravity cores. From the vertical displacement of the topsetto forset transitions, subsidence rates were calculated suggesting continuous and gradual subsidence for the Aegean margins during the last 400Ka. According to Lykousis’ palaeogeographical reconstruction, (a) the largest part of the central and northern Aegean was an exposed landscape, from at least MIS 10–12 until at least MIS 8 i.e. during both glacial and interglacial periods from at least 450 Kya to at least 250 Kya, and (b) fluvio-lacustrine conditions prevailed there until MIS 8. The extent of the Aegean palaeolandscape during its maximum exposure (MIS 10–12) has been calculated to be comparable to the current extent of mainland Greece (Tourloukis and Karkanas 2012: 9), and three to five times larger than the Last Glacial Maximum land exposure (~41,399 km2) (Tourloukis 2010:166). Furthermore, the rich and variable water resources (lacustrine, fresh water and marine resources) – suggested by geological data – attribute a high ecological value to the exposed landmass of the Aegean, allowing us to envision favourable conditions for the survival and dispersal of populations throughout the glacial-interglacial cycles of the Middle Pleistocene.

The aim of this paper is to discuss methodological challenges and limitations emerging when addressing these questions, and propose methodological schemes in order to explore aspects of hominin presence, mobility and survival across the Middle Pleistocene Aegean. An interdisciplinary approach is followed, having archaeology and spatial analysis (GIS) as the two main pillars. The research questions refer to wider areas and larger scales over which movement can be observed and conceptualised (Kuhn et al. 2016). In that sense, the dry Aegean is perceived not solely as a natural landscape, but rather as a record and testimony of hominin presence and activity, following Ingold’s “dwelling perspective” (Ingold 1993: 152). The Aegean palaeogeography and implications for hominin movement The Aegean region is dominated by the Greek archipelago and by the Greek mainland, which separates the Aegean Sea (to the east) from the Ionian Sea (to the west). The archipelago comprises of about 6000 islands and islets, with only 227 being inhabited. Its coastline covers 7500 km out of the total 16000 km of the Greek territory’s coastline. The topography of the wider area (including the mainland) is diverse and complex, combining coastal plains, lowlands and mountainous areas, and hosting variable environments/habitats. The Aegean has been developed as an extensional basin (Angelier 1978), due to the geotectonic evolution of the wider eastern Mediterranean (e.g. Le Pichon and Angelier 1979; Taymaz et al. 2007). Active geodynamics, plate movements and on-going geological processes create a complicated tectonic structure, resembling a geotectonic puzzle (Mascle and Martin 1990), with alternating

Founded upon Lykousis’ (2009) work, Sakellariou and Galanidou (2016; 2017) moved further, integrating evidence on the geology, tectonics, morphology and hydrogeology of the shallow coastal and shelf areas in order to reconstruct the palaeogeography of the Aegean. Taking into account the variable nature of the active morpho-tectonic processes, evolution and configuration of the coastal and submerged landscape, the Aegean has been divided into nine geographical areas with discrete geotectonic and morphological characteristics. This study set a frame of reference for assessing land-routes and natural resources available to hominins at different times of the Pleistocene, and allowed the Aegean region to enter the discussion of the early occupation of Europe, via a south-eastern route. 150

Exploring Hominin Movement Patterns in the Lower Palaeolithic Aegean Dry Land Under the light of these reconstructions, the Aegean region is now emerging as a new arena for the Lower Palaeolithic research and the study of the early colonisation of Europe, possibly holding valuable information for the hominin presence and activity during the Middle and perhaps the Early Pleistocene.

Topographic roughness is used as a proxy to identify potential areas favourable to hominins (occupation areas and natural pathways for movement) by measuring irregularities found in tectonically active environments. By recording surface roughness on modern dynamic landscapes, we can identify areas with high values of topographic complexity. Because of the continuous character of geological processes, modern terrain complexity will perpetuate areas of ancient roughness/ complexity. Modern features may still offer rough approximations and/or indications of the mosaic character of the past environments and their affordances, i.e. areas with high research potential, according to the concept described above.

Methodological challenges One obvious problem is that the topography and the environments of the now submerged Aegean landscapes are largely unknown. Moreover, available information from the Aegean, is characterised by (a) scarce and discontinuous evidence (archaeological, paleoenvironmental and paleogeographical), (b) temporal limitations, with the vast majority of available evidence going only as far back as the last glacial cycle (ca 130 Kya), and (c) the dynamic character of the tectonically active Aegean landscape. These limitations pose several methodological challenges in the study of the hominin movement across the Aegean dry land. The same limitations should be expected to affect any modelling attempt, with low resolution and accuracy, due to the poor quality of available data. Specific methodological approaches and alternative tools – consistent with the nature of available evidence – need to be developed in order to (a) study, model and conceptualise hominin movement and mobility over the Aegean and (d) target specific areas for the Lower Palaeolithic research.

The idea of the complex topography has been developed during the 1990’s, through work in Epirus (NW Greece), examining the location of Middle Palaeolithic sites – on a complex topographic setting – in relation to the seasonal movement of prey animals (Bailey, King and Sturdy 1993). It has further been tested in Lower Palaeolithic contexts in eastern Africa (along the African Rift) and the Red Sea, in areas with abundant Lower Palaeolithic evidence, and with applications primarily in terrestrial contexts (Bailey and King 2011; Bailey et al. 2011; King and Bailey 2006). The application of this concept on submerged landscapes, using current bathymetric data, has, so far, only been attempted in the southern Red Sea at the Farasan Islands (Bailey and King 2011; Meredith-Williams et al. 2014), enabling the identification of some elements of a complex topography “with localised barriers and basins that would be very familiar to human populations adapted to conditions in the African Rift” (Bailey and King 2011: 1550). These results, however, promising as they might be, should be treated with caution due to the poor resolution of data.

The ‘complex topography’ concept, developed by Bailey and King (2011; King and Bailey 2006) offers a rigorous approach to overcome the limitations that are inherent in the Aegean record. A dynamic landscape such as the Aegean region, shaped by extensive tectonics and volcanic activity, has a complex topography with pronounced variability and heterogeneities, e.g. fault -bounded basins, uplifted areas and volcanic landscapes. This fragmentation and transformation of the landscape is a continuous phenomenon, due to active geomorphic processes. It has been considered a negative factor in the study of hominin mobility and stasis. However, according to the complex topography concept (King and Bailey 2006: 266) the topographic complexity actually helped in the maintenance of mosaic environments, rich in natural resources and raw material availability and variability, providing the local conditions for hominin survival and the development of successful subsistence strategies. Those mosaic environments seem to be attractive for early hominins, as shown by available evidence from eastern Africa (e.g. Kübler et al. 2015; 2016). Key early-hominin localities of the Early Pleistocene are found in areas with high topographic complexity, while dispersal routes seem to be dictated / determined by pronounced landscape features generated by tectonic and volcanic activity (Reynolds et al. 2011). Lava flows, ridges and water barriers created naturals routes followed by dispersing populations during the major events of the Early Pleistocene. The opportunities provided by topographic complexity have even been associated with the emergence of anatomical features associated with bipedalism (Winder et al. 2013).

This project represents the first time that this concept has been applied to the Aegean region, using current elevation and bathymetric data to measure surface roughness onshore and offshore. The same practices developed for the analysis of terrestrial topography have been used for mapping the Aegean seabed topographic complexity (after Bailey and King 2011). The topographic roughness will set the background against which available evidence (palaeoenvironmental, palaeogeographical, palaeoclimatic and archaeological/ palaeoanthropological) will be projected and synthesised, via spatial analysis (GIS applications), in order to answer the two main research questions: 1. Can we suggest possible areas of hominin activity – as defined by Bailey and King (2011: 1533) under the ‘site region’ term – taking into account: (a) the topographic complexity of the landscape, and (b) the suggested richness of natural resources during the Middle Pleistocene over the Aegean exposed landscapes? 2. Can we produce scenarios of hominin movement and discuss possible dispersal routes traversing the Aegean dry land? 151

Peny Tsakanikou and John McNabb The Aegean, as mentioned before, is a tectonically active region divided by numerous major and minor fault systems into subsiding and uplifted areas, both onshore and offshore (see fig. 6 in Sakellariou and Galanidou 2016). The subsiding areas correlate with tectonically-controlled basinal systems (grabens), with gentle morphology that acted as sediment traps – receiving and accumulating fluvial, lacustrine and/or marine deposits. The uplifted areas on the other hand, have a harsh morphology e.g. mountain summits, steep slopes and deeply incised river valleys, and coincide with areas subjected to erosion – thus, supplying the sediment traps with eroded material. This geotectonic mechanism creates constant landscape transformation and instability. Still, there are areas that escape the main regimes e.g. the Chalkidiki Peninsula and the Thasos and Samothraki Islands remain uplifted, despite the general trend of subsidence over the Northern Aegean shelf (Sakellariou and Galanidou 2016). Other areas appear to be relatively stable, in terms of tectonics, such as the Cycladic Plateau, which has been characterised as aseismic (e.g. Maley and Johnson 1971: 114–16). There are also areas where the main landscape features persist through time, despite the action of the geomorphic processes; a characteristic example is the basinal system along the North Aegean Trough, in Northern Aegean, initiated during the Early Pliocene (Sakellariou et al. 2013). The areas that appear to persistently retain some basic elements of their past landscape nature – including sub-aerially exposed parts – are the main targets in the application of the complex topography concept in this study.

DEV measures the topographic position of a central point using TPI and the standard deviation of the elevation (Gallant and Wilson 2000). It measures the topographic position as a fraction of local relief normalised to local surface roughness (De Reu et al. 2013: 42). SL calculates the maximum rate of change of the surface in the horizontal and vertical directions, from every cell of the raster within an eight – cell neighbourhood (around the central cell each time). The maximum change in elevation over the distance between the cell and its eight neighbours identifies the steepest downhill descent from the cell (Burrough and McDonell 1998: 190). Maps of surface roughness have been produced by testing the three topographic metrics, mentioned above, on modern elevation and bathymetry from the Aegean. The study areaincludes, from east to west, the Aegean coasts of Turkey, the area now covered by the northern and central Aegean Sea and mainland Greece. It is worth highlighting at this point that throughout this study, the Middle Pleistocene Aegean landscape is perceived and treated as a terrestrial landscape, from at least MIS 10–12 until at least MIS 8, following the recent palaeogeographical reconstructions noted above (Lykousis 2009; Sakellariou and Galanidou 2016; 2017). According to Lykousis (2009: 2041, 2043), geological evidence suggests that the first marine intrusion and the gradual breaking up of the dry land did not happen before MIS 9. Fresh and brackish water sediments show the prevalence of lake environments in the central-northern Aegean during MIS 8 and prior to MIS 10. For that reason, sea-level fluctuations will not be considered in this work.

Measuring and recording topographic roughness in the Aegean

Topographic roughness for the study area has been calculated and recorded using the ArcMap software, a facet of the ArcGIS desktop suite. Terrain data was modelled using the ASTER Global Digital Elevation Map, version 2 (ASTERGDEM V2) (30m resolution), available at NASA Land Processes Distributed Active Archive Centre (LP DAAC), and bathymetric data was modelled using the Eastern Mediterranean Bathymetric Map (2016) (250m resolution) by courtesy of the Hellenic Centre for Marine Research. The spatial reference used throughout the analysis is the WGS84 geographic coordinate system, with linear units in metres, and all data is projected in the Web Mercator Auxiliary Sphere projection system (consistent with available online data). The DEM and the bathymetry map have been combined into a new raster map (Mosaic raster thereafter; Mosaic to New Raster tool; Figure 1). This new Mosaic raster has been used as the basic raster for calculating topographic roughness and applying spatial analyses.

GIS offer many different ways for measuring and visualising topographic roughness. Three different methods have been tested on the Aegean region: the Topographic Position Index (thereafter TPI), the Deviation of Mean Elevation (thereafter DEV) and the Slope Analysis (thereafter SL). All are topographic metrics, measuring the topographic position of a central point in relation to its surroundings in a predetermined radius, or to put it another way, the differences in elevation between a central point and its neighbouring points in a given neighbourhood. This concept is based on the topographic prominence idea as expressed by Lliobera (2001: 1007). TPI measures the difference between elevation at the central point (i.e. the cell elevation value) and the average elevation around it within a predetermined radius (Galland and Wilson 2000; Jenness 2006; Weiss 2001). The topographic position of each cell in the raster, is identified with respect to its local neighbourhood, thus its relative position. The index is useful to identify landscape patterns and boundaries that may relate with geomorphic processes, soil characteristics, rock types, vegetation, and air drainage. The index is applicable to bathymetric data as well (Bathymetric Position Index or BPI) (e.g. Verfaillie et al. 2007).

The TPI, for the study area, is calculated from modern elevation and bathymetry grids, using the formula developed by Arthur Crawford of ESRI. Three new elevation rasters are produced from the mosaic raster (modern elevation and bathymetry combined) (Spatial Analyst>Neighborhood> Focal statistics): 152

Exploring Hominin Movement Patterns in the Lower Palaeolithic Aegean Dry Land

Figure 1. The Mosaic raster combining elevation and bathymetry over the study area. Terrain data: ASTER Global Digital Elevation Map, version 2 (ASTERGDEM V2) (30m resolution), available at NASA Land Processes Distributed Active Archive Center (LP DAAC). Bathymetric data: Eastern Mediterranean Bathymetric Map (2016) (250m resolution) by courtesy of the Hellenic Centre for Marine Research. The map is produced in ArcMap 10.6.

The DEV, is calculated using, again, modern elevation and bathymetry grids, as before. The mosaic raster is the input in the focal statistics tool, to produce two new elevation rasters (Spatial Analyst>Neighborhood> Focal statistics):

• Mean elevation: calculates the mean (average value) of the cells in the neighbourhood, • Minimum elevation: calculates the minimum (smallest value) of the cells in the neighbourhood, • Maximum elevation: calculates the maximum (largest value) of the cells in the neighbourhood The following expression is then entered in the raster calculator (Spatial analyst> Map algebra> Raster calculator):

• Range elevation: calculates the range (difference between largest and smallest value) of the cells in the neighbourhood, • Mean elevation: calculates the mean (average value) of the cells in the neighbourhood (the same as before).

(“radius in metres” – “minDEM”) / (“maxDEM” – “minDEM”)

The following expression is entered in the raster calculator (Spatial analyst> Map algebra> Raster calculator):

The “radius in metres” represents the mean (smoothed) elevation raster, the “minDEM” represents the minimum elevation raster, and the “maxDEM” represents the maximum elevation raster.

(“meanDEM” – “DEM”) / “rangeDEM” The “meanDEM” represents the mean elevation raster, the “rangeDEM” represents the raster containing range of elevation values and “DEM” represents the mosaic raster.

The output raster provides an index of values expressing topographic roughness. Positive TPI values indicate that the central point is located higher than its average surroundings, while negative values indicate a position lower than the average. Values near zero represent either flat terrain (slope is near zero) or areas of constant slope (slope significantly greater than zero).

In the output raster, as in the TPI, positive values indicate that the central point is situated higher than its surroundings (neighbourhood) and negative values indicate that the central point is situated lower. The output values are usually between +1 and -1 and values outside this range usually indicate anomalies within the DEM. 153

Peny Tsakanikou and John McNabb slope of the cell (at the central point) is used to classify the cell into a slope position. If it is significantly higher than its surroundings, then it possibly lies on or near the top of a hill or a ridge. If it is significantly lower than its surroundings, then it possibly lies at or near the bottom of a valley (Jenness 2006: 6). However, the characterisation of landscape features may vary according to the scale over which they have been studied. The Weiss classification takes into consideration the variability of elevation values within the predetermined radius, enabling a better visualisation and a better understanding of the topographic complexity in different scales. Larger scales reveal larger landscape features while with smaller scales, a more detailed local topography emerges.

For the Slope analysis, the slope tool is used in the spatial analyst toolkit (Spatial analyst>Slope). Slope values are calculated form the mosaic raster. The output raster, shows the rate of change in elevation in classes of values; the lower the slope value the flatter the terrain, the higher the slope value the steeper the terrain. For the TPI and DEV three different radii have been tested: 1km, 3km and 10km, representing different scales within the exploitation range (10km) suggested for the Lower Palaeolithic (Bailey and King 2011: 1533; Feblot – Augustins 1999) (Figure 2). Slope analysis presented some important problems when applied on the Mosaic raster. More specifically, the result over the area covered by the bathymetric data is quite crude – possibly due to the resolution of available data (250m resolution). Slope analysis has been applied separately on each raster (bathymetry and DEM) in order to achieve better results in terms of visualisation.

Through testing the three different methods for measuring landscape roughness, some initial results are included here, in relation to the process itself and the function of the different metrics on the specific dataset from the Aegean: 1. TPI is a more adequate method for recording the Aegean topographic roughness in comparison to DEV (Figure 3). Examples from literature (e.g. De Reu et al. 2013) show that (a) DEV works better in a local scale, identifying local topographic features/patterns), rather than at a regional scale, and (b) TPI produces better results in relatively homogenous landscapes, as its range is highly

For the classification of the landscape into morphological classes, the Weiss (2001) classification is followed (Figure 3-right). The landscape is classified into discrete slope position classes using standard deviation of TPI. The degree to which the central point (cell) is higher or lower than its surroundings (in a predetermined radius), plus the

Figure 2. TPI applied on the Aegean bathymetry in three different radii (from left to right: 1km, 3km and 10km). Terrain data: ASTER Global Digital Elevation Map, version 2 (ASTERGDEM V2) (30m resolution), available at NASA Land Processes Distributed Active Archive Center (LP DAAC). Bathymetric data: Eastern Mediterranean Bathymetric Map (2016) (250m resolution) by courtesy of the Hellenic Centre for Marine Research. The map is produced in ArcMap 10.4.

154

Exploring Hominin Movement Patterns in the Lower Palaeolithic Aegean Dry Land

Figure 3. DEV (Jenks classification – natural breaks) and TPI (Weiss classification) applied on a 3 km radius. Terrain data: ASTER Global Digital Elevation Map, version 2 (ASTERGDEM V2) (30m resolution), available at NASA Land Processes Distributed Active Archive Center (LP DAAC). Bathymetric data: Eastern Mediterranean Bathymetric Map (2016) (250m resolution) by courtesy of the Hellenic Centre for Marine Research. The map is produced in ArcMap 10.4.

Exploring the ‘hominin factor’ through spatial analysis

influenced by surface roughness. In the Aegean, the scale of analysis is regional rather than local, and the area could be characterised homogenous in terms of hosting extreme landscape features, such as high mountains, ridges and deep depressions.

How do hominins fit into the picture? Can we generate hypotheses on hominin movement patterns through spatial analysis? Least-cost path analysis is undertaken to explore the potential of the Aegean exposed landscapes as a traversable terrain, and test recently made suggestions about possible trans-Aegean passages (land-bridge) between W. Anatolia and Europe via the Greek mainland, during the Middle Pleistocene (Strait 2016; Tourloukis and Karkanas 2012b).

2. The importance of the radius (R) size. Large R-values mainly reveal major landscape units, while smaller values highlight smaller features, such as minor valleys and ridges (De Reu et al. 2013: 42; Jenness 2006; Llobera et al. 2011). As such, the R-size depends on the research question and the scale of the study (regional scales = larger scales). In the Aegean context, the 3km radius has been selected as adequate for (a) identifying satisfactorily topographic features in a regional scale (Figure 2) and (b) complying with the spatial range of exploitation territories for the Lower Palaeolithic groups, as mentioned before.

The least-cost path travels from an origin to a destination, guaranteeing that this is the cheapest route, in terms of cost, to traverse the landscape, in relation to a cost surface/ friction surface. The cost surface is a raster where the value of each cell represents a cost of traversing the specific surface. The cost is determined according to the research questions(s) and the examined variables. What is of interest here, is to get an idea about costeffective possibilities to cross the Aegean, based on the modern landscape structure, and assuming a continuous terrain over the northern and central parts. In this case,

3. When combining the results from TPI and SL, it appears that TPI captures roughness more efficiently, even in areas that SL identifies as flat terrain (Figure 4). This is extremely important, since topographic roughness refers to surface irregularities found in both high and low elevation areas.

155

Peny Tsakanikou and John McNabb

Figure 4. SL (left) and TPI overlaying SL (right) over the same area of the northern Aegean (top- Mosaic raster). Notice that TPI captures surface roughness more efficiently in relation to SL. Terrain data: ASTER Global Digital Elevation Map, version 2 (ASTERGDEM V2) (30m resolution), available at NASA Land Processes Distributed Active Archive Center (LP DAAC). Bathymetric data: Eastern Mediterranean Bathymetric Map (2016) (250m resolution) by courtesy of the Hellenic Centre for Marine Research. The map is produced in ArcMap 10.4.

In order to use the hiking function in a given terrain, first the slope needs to be calculated and then converted to walking velocity. Slope (in percentage) has been already calculated for the Aegean (previous section).

topographic parameters are examined and the cost will refer to the time of travel. Step 1 – Cost surface

The percentage slope formula is:

A cost surface has been created representing traveling time, i.e. the time it takes to cross each grid cell to walk from point A to point B. Velocity is the indicator we are seeking – a measure of distance per time. Tobler’s hiking function (Tobler 1993), using empirical data from Imhof (1950), predicts human walking speed based on slope. The speed is affected by topography; smoother terrain permits fastest walking, while complex terrain reduces walking speed. Palaeogeographical features, representing natural barriers to movement, have been included in this modelling.

slopeperc = 100 . dh / dx Where dh stands for terrain elevation and dx for horizontal length. Notice that in Tobler’s function: S is not the same as slopeperc but S corresponds to dh / dx (S = dh / dx). To get the dh / dx value, the raster of the slopeperc is divided by 100 (raster calculator). The new slope dh / dx raster is then used to calculate velocity. In the raster calculator, the expression of the hiking function is entered, where S = slope dh / dx raster.

The hiking function is expressed as: W = 6 * exp {-3.5 * abs (S + 0.05)}

The output raster represents velocity in kilometres per hour. That needs to be changed into hours per metre; what we need to know is the amount of time it took to traverse the landscape, thus the ratio km/hr needs to be inverted. Furthermore, the linear unit in the coordinate system of the mosaic raster, from which slope has been calculated, is in

Where: W = walking velocity (km / hr) S = slope of the terrain (dh/dx)

156

Exploring Hominin Movement Patterns in the Lower Palaeolithic Aegean Dry Land metres (not km). These alterations are easily made through the raster calculator.

the archaeological evidence from the selected sites, due to the fragmented records and the extensive spatiotemporal gaps characterising the Lower Palaeolithic from this region. However, wherever such similarities are present, they have been considered in our models.

From the final time-of-travel raster (hours per metre), specific areas need to be excluded, corresponding to palaeolandscape features, which would have acted as natural barriers to hominin movement. To do so, first these areas need to be determined. Information from available palaeogeographical reconstructions (Lykousis 2009) is used to produce two new rasters with the areas covered by major water bodies in the northern and central Aegean during MIS 10–12 and during MIS 8 respectively. In order to determine no data for the cells corresponding to water coverage, the Is Null tool is used. To exclude water coverage areas from the time of travel raster, the following expression is entered in the raster calculator:

Least-cost paths are calculated here for two different time periods MIS 10–12 and before (> 480–350 Kya) and MIS 8 (300–250 Kya). For the MIS 8 period, the first origin is the Yarimburgaz cave in Bosporus (Turkey) (Arsebük and Özbaşran 1999), the second origin is the Karain cave (SW Turkey) (Otte et al. 1995) and the destination is the Rodia – FS 30 in mainland Greece (Runnels and Van Andel 1993b; 1999) (Figure 5 – top). The specific sites have been chosen due to their similar chronology, dated within MIS 8. For the Yarimburgaz and Rodia similarities in the stone tool assemblages have been reported as well, in terms of the raw material (quartz) and the Mode 1 stone tool tradition (Tourloukis 2010: 41).

Con (is_null raster, hr_m raster) The output rasters correspond to two cost surfaces (travel time hours per metre), excluding areas covered by water bodies during MIS 10–12 and during MIS 8.

For the MIS 10–12 period, the least-cost path has been produced between the sites of Rodafnidia (Lesbos island, NE Aegean) – origin (Galanidou et al. 2013; 2016) and Marathousa 1 site (Megalopolis basin, Peloponnese) – destination (Panagopoulou et al. 2015) (Figure 5 – bottom). Both sites have been securely attributed to the Middle Pleistocene, around or prior to 400 Kya. The stone tool assemblages demonstrate differences in their technology and typology, with Marathousa 1 assemblage following the Mode 1 tradition, while in Rodafnidia, an Acheulean/Mode 2 industry with possible African affinities is present. Nevertheless, hominin groups, even with different traditions, were present at roughly the same time in different areas of the Aegean.

Step 2 – Cost distance (spatial analyst> distance>cost distance). In the cost distance raster, a cost-accumulated surface has been created from the cost surface raster, where each cell represents the minimum accumulation of cost from an origin (Llobera et al. 2011: 844). As such, the cost distance raster identifies the accumulative cost for each cell to return to the closest source location (origin). In our case, two cost surfaces have been produced representing time of travelling for two different timeperiods: MIS 10–12 and MIS 8. Over the accumulated cost surface, the vertical changes in topography do not actually represent changes in height (elevation and bathymetry) but rather, changes in the cost accumulated from the origin. Ridges represent locations where the accumulation of cost is locally maximum, while valleys and/or channel-like features represent areas where the accumulation of cost is locally minimum. In other words, it would be more costly (i.e. it would take more time) to walk on steep terrain than on flatter/smoother ground. Movement over the accumulated cost surface, from an origin to a destination, follows the least-cost direction i.e. cells with the minimum accumulation of cost.

Results and Discussion At this stage of the study, the work being done on topographic roughness and least-cost path analyses permit some initial comments on the effectiveness of the practices followed, and a preliminary discussion on suggested areas with increased research potential across the Aegean region. These target-areas encompass the possibility rather than the certainty of preserving archaeological material, or information on hominin preferential environments, indicated by landscape patterns and the analysis and synthesis of available evidence, through the complex topography conceptual ‘filter’. Three different methodologies for recording and visualising surface roughness have been tested specifically in the Aegean context. The Topographic Position Index (TPI) performed better at the regional scale of this study and provided optimum results (in comparison to Deviation of Mean Elevation and Slope Analysis), capturing effectively roughness in low and high elevation areas, both onshore and offshore. However, recording and visualising landscape roughness in the Aegean region is biased due to several factors:

Step 3 – Cost path (spatial analyst> distance>cost path) The origins and destination in our analysis correspond to the locations of known Lower Palaeolithic archaeological sites, found on both sides of the Aegean (mainland Greece and W. Turkey). The selection of the sites is made on the basis of general chronological associations i.e. evidence (archaeological and/or palaeoanthropological) for the presence of hominins in the Aegean region during the Middle Pleistocene. In that sense, it is not always possible to draw strict or detailed associations, e.g. similarities in 157

Peny Tsakanikou and John McNabb

Figure 5. Top: Least-cost paths for MIS 8: (a) Yarimburgaz to Rodia, and (b) Karain to Rodia. Bottom: Least-cost path for MIS 10–12: Rodafnidia to Marathousa 1. The location of the water bodies and the position of the southernmost border of the Aegean exposed landscape during MIS 8 and MIS 10–12 follow the palaeogeographical reconstruction by Lykousis (2009). Terrain data: ASTER Global Digital Elevation Map, version 2 (ASTERGDEM V2) (30m resolution), available at NASA Land Processes Distributed Active Archive Center (LP DAAC). Bathymetric data: Eastern Mediterranean Bathymetric Map (2016) (250m resolution) by courtesy of the Hellenic Centre for Marine Research. The map is produced in ArcMap 10.6.

• A palaeolandscape reconstruction, taking into account all the main landscape transformation factors, for the Early and Middle Pleistocene Aegean is not yet available. • Therefore, active geomorphic processes (uplift, subsidence, sedimentation, erosion) have not been calculated and/or included in our models. • Topographic metrics have been applied on current bathymetric and elevation data.

features persist in time despite the action of geomorphic processes, such as the deep depressions at the northern Aegean (on the south of the Chalkidiki Peninsula), or escape the active geomorphic processes, being relatively stable, such as the Cycladic Plateau (at the south-central Aegean). Topographic complexity appears to be high in both areas (Figure 2), which are further highlighted when roughness is examined against, or in combination with, other landscape variables: (a) the abundance of water resources across the northern Aegean and (b) the distribution of volcanic rocks along the Southern Aegean Volcanic Arc suitable for knapping into tools. The selection of the variables is not random, but follows specific criteria suggesting a preferential association

Nevertheless, even from current topographic evidence, information can be deduced, at a first order, about the topographic complexity of the past landscapes and the nature of the past environments. Specific landscape 158

Exploring Hominin Movement Patterns in the Lower Palaeolithic Aegean Dry Land between hominins and water resources, as well as volcanic landscapes. The working hypothesis emerging now, which will be further explored, is that these areas could provide viable pathways, in terms of resources securing hominin survival, connecting W. Anatolia and Europe through mainland Greece: (a) the Northern and the Southern Aegean Affordance Corridors.

our understanding of wider processes occurring during the Lower Palaeolithic, such as the early peopling of Europe; with specific implications for (a) the origin of the first hominins that populated Europe, (b) the routes they followed and (c) the timing of these early dispersal events. We suggest that available evidence should be viewed under the prism of two main concepts:

The results from the least-cost path analysis are promising. The produced paths are traversing the Aegean exposed landscapes, despite the fact that (a) current elevation and bathymetry are used, and (b) the continuous subsidence pattern of the Aegean region throughout last 400 Ka (Lykousis 2009). In that sense, these paths should not be viewed as specific routes followed by hominins during the Middle Pleistocene, nor as indicative of specific movement patterns, but only as an indication that the concept of the Aegean as a traversable terrain during the Middle and possibly during the Early Pleistocene could be a viable hypothesis. The least-cost path analysis is only a first step towards the development of arguments supporting this suggestion. Other archaeological examples, as well as more variables, both palaeoenvironmental (e.g. climatic evidence) and biological (e.g. energy/calorific costs) could be incorporated in the model, to further inform aspects relating to hominin mobility per se (e.g. energetic costs to move across a landscape along a proposed route), and hominin mobility patterns (e.g. flexibility and/or adaptability in relation to environmental and climatic variability).

(a) Our perception of the palaeogeography of eastern and north-eastern Mediterranean and specifically of the Aegean region, needs to be re-adjusted (e.g. Roksandic 2016: 30). The existence of the exposed Middle Pleistocene landscape over the area now covered by the Aegean Sea which could be traversed by hominins needs to be embedded in the way we investigate, study, and interpret the archaeological and palaeoanthropological evidence from this region. (b) Instead of adopting the prevailing ‘terrestrial Eurocentric’ point of view in the discussion of the early dispersals and colonisation of Europe and the dispersal routes followed by early hominins, the Balkans and Anatolia are viewed here as a core area for the survival and dispersal of populations during the Middle Pleistocene (e.g. Dobos and Iovita 2016; Ivanova 2016; Strait et al. 2016), offering an alternative or perhaps a complementary scenario. The exploration of hominin movement and occupation patterns in the Aegean region during the Middle Pleistocene presents several serious methodological problems. Landscape dynamics have been highlighted as a major factor, affecting the preservation and discovery of Lower Palaeolithic material. The same bias is also responsible for limitations in modelling hominin mobility and activity. Landscape fragmentation, active geomorphic processes, spatiotemporal discontinuities, and a general paucity of evidence for the Early and Middle Pleistocene impose further difficulties, affecting the quality of available evidence. However, this biased context may still retain valuable archaeological and palaeoanthropological information. The complex topography concept has been used here as a filter in helping us to trace and identify areas with high research potential, while spatial analyses offer a valuable tool not only to map the Aegean palaeolandscape and its affordances, but also to conceptualise possibilities of hominin activity and mobility.

The Aegean holds a central geographical position between Africa and Eurasia, starting now to have an imprint on the archaeological and paleoanthropological records as well. New finds and the reassessment of older ones enhance significantly our knowledge about the archaeology and palaeoanthropology of this particular geographical region during the Early and Middle Pleistocene. Until recently (e.g. Runnels 1995; Galanidou 2004), the presence of hominins in the Greek territory was not securely established before the later parts of the Middle Pleistocene and undisputedly only from the Upper Pleistocene onwards. Current evidence from Rodafnidia, Marathousa 1 and the re-examination of the Megalopolis hominin remains (Harvati 2016) suggest that the time frame for the Lower Palaeolithic research could be stretched back to the earlier parts of the Middle Pleistocene, and potentially to the Early Pleistocene. This suggestion would be in good chronological agreement with secure evidence found elsewhere in the southern European Peninsulas with sites over 900 Kya (e.g. Mincel et al. 2017; Toro Moyano et al. 2013; Arzarello et al. 2007). According to current data, it could not be argued with certainty that hominins were present in the Aegean region before the Middle Pleistocene, but such a suggestion is at least now open to future research. The exposed landscape of the Aegean adds a new strand in this discussion. An important element of the paleogeography of the eastern part of Europe is now starting to affect the wider Lower Palaeolithic narrative, challenging

Acknowledgements This paper is the result of ongoing PhD research at the University of Southampton (Centre for the Archaeology of Human Origins), under the supervision of Dr John McNabb and Dr Helen Farr. Financial support has been provided by the University of Southampton (The Vice Chancellor’s Award) and the Greek Archaeological Committee UK (GACUK). The authors are grateful to Prof Nena Galanidou for the opportunity to participate in the excavation at Rodafnidia site on Lesbos Island and for facilitating the writing of the paper during the current fieldwork season (July-August 2018).

159

Peny Tsakanikou and John McNabb Works cited

Greece: 2010–2012’. In K. Harvati and M. Roksandic (eds.), Palaeoanthropology of the Balkans and Anatolia: Human evolution and its context. Springer: Dordrecht. Pp. 119–38.

ANGELIER, J. 1978. ‘Tectonic evolution of the Hellenic Arc since the Late Miocene’. Tectonophysics 49.1: 23–36.

GALANIDOU, N., COLE, J., ILIOPOULOS, G., and MCNABB, J. 2013. ‘East meets west: the Middle Pleistocene site of Rodafnidia on Lesvos, Greece’. Antiquity 87.336.

ARSEBÜK, G. and ÖZBAŞARAN, M. 1999. ‘Pleistocene archaeology at the Cave of Yarimburgaz in Eastern Thrace/Turkey: preliminary results’. British School at Athens Studies 3: 59–72.

GALLANT, J.C., WILSON, J. P. 2000. Primary topographic attributes. In: Wilson, J.P., Gallant, J.C. (Eds.), Terrain Analysis: Principles and Applications. Wiley: New York. Pp. 51–85.

ARZARELLO, M., MARCOLINI, F., PAVIA, G., PAVIA, M., PETRONIO, C., PETRUCCI, M., ROOK, L., SARDELLA, R. 2007. ‘Evidence of earliest human occurrence in Europe: the site of Pirro Nord (Southern Italy)’. Naturwissenschaften 94: 107–112.

GARCIA, J., MARTÍNEZ, K., CUENCA-BESCÓS, G., CARBONELL, E. 2014. ‘Human occupation of Iberia prior to the Jaramillo magnetochron (> 1.07 Myr)’. Quaternary Science Reviews 98: 84–99.

BAILEY, G.N. and KING, G. C. P. 2011. Dynamic landscapes and human dispersal patterns: tectonics, coastlines, and the reconstruction of human habitats’. Quaternary Science Reviews 30.11–12: 1533–53.

HARFF, J., BAILLEY, G., and LÜTH, F. 2016. ‘Geology and archaeology: submerged landscapes of the continental shelf: an introduction’. In J. Harff, G. Bailley, and F. Lüth (eds.), Geology and archaeology: submerged landscapes of the continental shelf. Geological Society London, Special Publication 411: 1–8.

BAILEY, G., KING, G., and STURDY, D. 1993. ‘Active tectonics and land-use strategies: a Palaeolithic example from northwest Greece’. Antiquity 67.255: 292–312. BURROUGH, P. A., and MCDONELL, R. A. 1998. Principles of Geographical Information Systems. Oxford University Press: New York.

HARVATI, K. 2016. ‘Paleoanthropology in Greece: Recent findings and interpretations’. In K. Harvati and M. Roksandic (eds.), Paleoanthropology of the Balkans and Anatolia: Human evolution and its context. Springer: Dordrecht. Pp. 3–14.

DENNELL, R.W., MARTINON-TORRES, M., BERMUDEZ DE CASTRO, L.M. 2011. ‘Hominin variability, climatic instability and population demography in Middle Pleistocene Europe’. Quaternary Science Reviews 30: 1511–24.

HARVATI, K. and ROKSANDIC, M. 2016 (eds.). Palaeoanthropology of the Balkans and Anatolia: Human evolution and its context. Vertebrate Palaeobiology and Palaeoanthropology Series, Springer: Dordrecht.

DE REU, J., BOURGEOIS, L., BATS, M., ZWERTVAEGHER, A.,GELORINI, V., DE SMEDT, P., CHU, W., ANTROP, M., DE MAEYER, P., FINKE, P., VAN MEIRVENNE, M., VERNIERS, J., and CROMBÉ, P. 2013. ‘Application of the topographic position index to heterogeneous landscapes’. Geomorphology 186: 39–49.

HEWITT, G. M. 1999. ‘Post‐glacial re‐colonization of European biota’. Biological Journal of the Linnean Society, 68.1‐ 2: 87–112.

DOBOŞ, A., and IOVITA, R. 2016. ‘The Lower Paleolithic of Romania revisited: New evidence from the site of Dealul Guran’. In K. Harvati and M. Roksandic (eds.), Paleoanthropology of the Balkans and Anatolia: Human evolution and its context. Springer: Dordrecht. Pp. 171–86.

IMHOF, E. 1950. Gelande und karte. Eugen Rentsch Verlag: Zurich. INGOLD, T. 1993. ‘The Temporality of Landscape’. World Archaeology 25.2: 152–74. IVANOVA, S. 2016. ‘A route through the Balkans and implications for the earliest settlement of Europe’. In K. Harvati and M. Roksandic (eds.), Paleoanthropology of the Balkans and Anatolia: Human evolution and its context. Springer: Dordrecht. Pp. 187–211.

FÉBLOT-AUGUSTINS, J. 1999. ‘Raw material transport patterns and settlement systems in the European Lower and Middle Palaeolithic: continuity, change and variability’. In W. Roebroeks and C. Gamble (eds.), The Middle Palaeolithic Occupation of Europe. Leyden University. Pp. 193–214.

JENNESS, J. 2006. Topographic Position Index (tpi_ jen.avx) extension for ArcView 3.x, v. 1.2. Jenness Enterprises. Available at: http://www.jennessent.com/ arcview/tpi.htm. [Accessed: April 2017]

GALANIDOU, N. 2004. ‘Early hominids in the Balkans’. In Balkan Biodiversity. Springer: Dordrecht. Pp. 147– 65.

JÖRIS, O. 2014. ‘Early Palaeolithic Europe’. The Cambridge world prehistory 3: 1703–46.

GALANIDOU, N., ATHANASSAS, C., COLE, J., ILIOPOULOS, G., KATERINOPOULOS, A., MAGGANAS, A., and MCNABB, J. 2016. ‘The Acheulian Site at Rodafnidia, Lisvori, on Lesbos,

KAHLKE, R.-D., GARCÍA, N., KOSTOPOULOS, D. S., LACOMBAT, F., LISTER, A. M., MAZZA, P. P. 160

Exploring Hominin Movement Patterns in the Lower Palaeolithic Aegean Dry Land A., SPASSOV, N., and TITOV, V. V. 2011. ‘Western Palaearctic palaeoenvironmental conditions during the Early and early Middle Pleistocene inferred from large mammal communities, and implications for hominin dispersal in Europe’. Quaternary Science Reviews 30.11–12: 1368–95.

LLOBERA, M., FÁBREGA-ÁLVAREZ, P., and PARCERO-OUBIÑA, C. 2011. ‘Order in movement: a GIS approach to accessibility’. Journal of Archaeological Science 38.4: 843–51. LYKOUSIS, V. 2009. ‘Sea-level changes and shelf break prograding sequences during the last 400ka in the Aegean margins: Subsidence rates and palaeogeographic implications’. Continental Shelf Research 29.16: 2037–44.

KING, G. and BAILEY, G. 2006. ‘Tectonics and human evolution’. Antiquity 80.308: 265–86. KOSTOPOULOS, D., VASILIADOU, K., and KOUFOS, G. 2002. ‘The beginning of Pleistocene in the Balkan area according to the mammal record; palaeozoogeographical approach’. Annales Géologiques des Pays Helléniques. Pp. 253–78.

MALEY, T. S. and JOHNSON, G. L. 1971. ‘Morphology and structure of the Aegean Sea’. Deep Sea Research and Oceanographic Abstracts 1: 109–22. MARTÍNEZ, K., GARCIA, J., CARBONELL, E., AGUSTÍ, J., BAHAIN, J. J., BLAIN, H. A., BURJACHS, F., CÁCERES, I., DUVAL, M., FALGUÈRES, C., and GÓMEZ, M. 2010. ‘A new lower Pleistocene archeological site in Europe (Vallparadís, Barcelona, Spain)’. Proceedings of the National Academy of Sciences 107.13: 5762–67.

KOSTOPOULOS, D.S., PALOMBO, M. R., ALBERDI, M. T., and VALLI, A. M. 2007. ‘Pliocene to Pleistocene large mammal diversity and turnover in North Mediterranean region: the Greek Peninsula with respect to the Iberian and Italian ones’. Geodiversitas 29.3: 401–19. KOSTOPOULOS, D.S. and KOULIDOU, I. 2015. ‘An early mammoth maxilla from north-western Greece’. Quaternary International 379: 155–163.

MASCLE, J. and MARTIN, L. 1990. ‘Shallow structure and recent evolution of the Aegean Sea: a synthesis based on continuous reflection profiles’. Marine Geology 94.4: 271–99.

KOUFOS, D. G and KOSTOPOULOS, D. S. 2016. ‘The Plio-Pleistocene Large Mammal Record of Greece: Implications for Early Human Dispersals into Europe’. In K. Harvati and M. Roksandic (eds.), Palaeoanthropology of the Balkans and Anatolia: Human evolution and its context. Springer: Dordrecht. Pp. 269–280.

MICHEL, V., SHEN, C. C., WOODHEAD, J., HU, H. M., WU, C. C., MOULLÉ, P. É., KHATIB, S., CAUCHE, D., MONCEL, M. H., VALENSI, P., and CHOU, Y. M. 2017. ‘New dating evidence of the early presence of hominins in Southern Europe’. Scientific reports 7.1: 10074.

KÜBLER, S., OWENGA, P., REYNOLDS, S. C., RUCINA, S. M., and KING, G. C. 2015. ‘Animal movements in the Kenya Rift and evidence for the earliest ambush hunting by hominins’. Scientific Reports 5: 14011.

OMS, O., PARÉS, J. M., MARTÍNEZ-NAVARRO, B., AGUSTÍ, J., TORO, I., MARTÍNEZ-FERNÁNDEZ, G., and TURQ, A. 2000. ‘Early human occupation of Western Europe: paleomagnetic dates for two paleolithic sites in Spain’. Proceedings of the National Academy of Sciences, 97.19: 10666–70.

KÜBLER, S., RUCINA, S. M., REYNOLDS, S. C., OWENGA, P., BAILEY, G., and KING, G. C. 2016. ‘Edaphic and topographic constraints on exploitation of the Central Kenya Rift by large mammals and early hominins’. Open Quaternary 2.5: 1–18.

OTTE, M., YALCINKAYA, I., KOZLOWSI, J., BARYOSEF, O., TAŞKIRAN, H., and NOIRET, P. 1995. ‘Evolution technique au Paléolithique ancien de Karain (Turquie)’. Anthropologie 99: 529–61.

KUHN, S. L., RAICHLEN, D. A., and CLARK, A. E. 2016. ‘What moves us? How mobility and movement are at the center of human evolution’. Evolutionary Anthropology 25.3: 86–97.

PALOMBO, M. R., VALLI, A. M. F., KOSTOPOULOS, D. S., ALBERDI, M. T., SPASSOV, N., and VISLOBOKOVA, I. 2006. ‘Similarity relationships between the Pliocene to middle Pleistocene large mammal faunas of Southern Europe from Spain to the Balkans and the North Pontic region’. CourierForschungsinstitut Senckenberg 256: 329–47.

LAMBECK, K. 1996. ‘Sea-level change and shore-line evolution in Aegean Greece since Upper Palaeolithic time’. Sea 70.269: 588–611.

PANAGOPOULOU, E., TOURLOUKIS, V., THOMPSON, N., ATHANASSIOU, A., TSARTSIDOU, G., KONIDARIS, G. E., GIUSTI, D., KARKANAS, P. and HARVATI, K. 2015. ‘Marathousa 1: a new Middle Pleistocene archaeological site from Greece’. Antiquity Project Gallery 343: 1–7.

LE PICHON, X. and ANGELIER, J. 1979. ‘The Hellenic Arc and Trench system a key to Neotectonic evolution of the Eastern Mediterranean area’. Tectonophysics 60: 1–42. LLOBERA, M. 2001. ‘Building past landscape perception with GIS: Understanding topographic prominence’. Journal of Archaeological Science 28.9: 1005–14.

REYNOLDS, S. C., BAILEY, G., and KING, G. C. P. 2011. ‘Landscapes and their relation to hominin habitats: case studies from Australopithecus sites in eastern 161

Peny Tsakanikou and John McNabb and southern Africa’. Journal of Human Evolution 60: 257–80.

of the Balkans and Anatolia: Human evolution and its context. Springer: Dordrecht. Pp. 69–78.

ROKSANDIC, M. 2016. ‘The role of the Central Balkans in the peopling of Europe: Paleoanthropological evidence’. In K. Harvati and M. Roksandic (eds.), Palaeoanthropology of the Balkans and Anatolia: Human evolution and its context. Springer: Dordrecht. Pp. 15–33.

TAYMAZ, T., YILMAZ, Y., and DILEK, Y. 2007. ‘The geodynamics of the Aegean and Anatolia: introduction’. Geological Society, London, Special Publications 291.1: 1–16. TOBLER, W. 1993. Three Presentations on Geographical Analysis and Modeling: Non-Isotropic Geographic Modeling; Speculations on the Geometry of Geography; and Global Spatial Analysis. National Centre for Geographic Information and Analysis, Santa Barbara, Technical Report (93–1).

RUNNELS, C. 1995. ‘Review of Aegean Prehistory IV’. American Journal of Archaeology 99 (Oct): 699–728. RUNNELS, C. and VAN ANDEL, H. T. 1993. ‘The lower and middle Paleolithic of Thessaly, Greece’. Journal of Field Archaeology 20.3: 299–317.

TORO-MOYANO, I., MARTÍNEZ-NAVARRO, B., AGUSTÍ, J., SOUDAY, C., De CASTRO, J. M. B., MARTINÓN-TORRES, M., FAJARDO, B., DUVAL, M., FALGUÈRES, C., OMS, O., and PARÉS, J. M. 2013. ‘The oldest human fossil in Europe, from Orce (Spain)’. Journal of Human Evolution 65.1: 1–9.

RUNNELS, C. and VAN ANDEL, H.T. 1999. ‘The Palaeolithic in Larisa, Thessaly’. British School at Athens Studies, 3 (The Palaeolithic Archaeology of Greece and Adjacent Areas: Proceedings of the ICOPAG Conference, Ioannina, September 1994). Pp. 215–20.

TOURLOUKIS, V. 2016. ‘On the spatio-temporal distribution of Mediterranean Lower Paleolithic sites: A geoarchaeological perspective’. In K. Harvati and M. Roksandic (eds.), Paleoanthropology of the Balkans and Anatolia: Human evolution and its context. Springer: Dordrecht. Pp. 303–23.

SAKELLARIOU, D. and GALANIDOU, N. 2016. ‘Pleistocene submerged landscapes and Palaeolithic archaeology in the tectonically active Aegean region’. Geological Society, London, Special Publications 411.1: 145–78.

TOURLOUKIS, V. 2010. The Early and Middle Pleistocene Archaeological Record of Greece: Current status and future prospects. Doctoral thesis. Archaeological Studies Leiden University (ASLU) 23.

SAKELLARIOU, D. and GALANIDOU, N. 2017. ‘Aegean Pleistocene Landscapes Above and Below Sea-Level: Palaeogeographic Reconstruction and Hominin Dispersals’. In G. Bailey, J. Harff, and D. Sakellariou (eds.), Under the Sea: Archaeology and Palaeolandscapes of the Continental Shelf. Springer: Cham. Pp. 335–59.

TOURLOUKIS, V. and KARKANAS, P. 2012. ‘The Middle Pleistocene archaeological record of Greece and the role of the Aegean in hominin dispersals: new data and interpretations’. Quaternary Science Reviews 43: 1–15.

SAKELLARIOU, D., MASCLE, J., and LYKOUSIS, V. 2013. ‘Strike slip tectonics and transtensional deformation in the Aegean region and the Hellenic Arc: Preliminary results’. Bulletin of the Geological Society of Greece, XLVIII (Proceedings of the 13th International Congress, Chania, Sept. 2013) 47.2: 647.

TOURLOUKIS, V., THOMSON, N., PANAGOPOULOU, E., GUISTI, D., KONIDARI, G. E., KARKANAS, P. and HARVATI, K. 2018. ‘Lithic artifacts and bone tools from the Lower Palaeolithic site Marathousa 1, Megalopolis, Greece: Preliminary results’. Quaternary International 497: 47–64.

SPASSOV, N. 2016. ‘Southeastern Europe as a Route for the Earliest Dispersal of Homo Toward Europe: Ecological Conditions and the Timing of the First Human Occupation of Europe’. In K. Harvati and M. Roksandic (eds.), Palaeoanthropology of the Balkans and Anatolia: Human evolution and its context. Springer: Dordrecht. Pp. 281–90.

TZEDAKIS, P. C. 2004. ‘The Balkans as prime glacial refugial territory of European temperate trees’. In H. I. Griffiths, B. Kryštufek, and J. M. Reed (eds.), Balkan Biodiversity: Pattern and Process in the European Hotspot. Springer: Netherlands. Pp. 49–68.

STEWART, J. R., LISTER, A. M., BARNES, I., and DALÉN, L. 2010. ‘REFUGIA revisited: individualistic responses of species in space and time’. Proceedings of the Royal Society B: Biological Sciences 277.1682: 661–71.

VERFAILLIE, E., DOORNENBAL, P., MITCHELL, A.J., WHITE, J., and VAN LANCKER, V. 2007. ‘The bathymetric position index (BPI) as a support tool for habitat mapping’. Worked Example for the MESH Final Guidance, 14.

STRAIT, D. S., ORR, C. M., HODGKINS, J., SPASSOV, N., GUROVA, M., MILLER, C., and TZANKOV, T. 2016. ‘The human fossil record of Bulgaria, and the formulation of biogeographic hypotheses’. In K. Harvati and M. Roksandic (eds.), Paleoanthropology

WEISS, A. 2001. ‘Topographic position and landforms analysis’.  Poster presentation, ESRI user conference, San Diego, CA (Vol. 200).

162