The Maritime World of Ancient Rome 0472115812, 9780472115815

Citation preview

/';-=09

)(8*

=-0/']

The Maritime World of Ancient Rome

Memoirs of the American Academy in Rome

Supplementary Volume VI

Support for this publication was provided by the Lucy Shoe Meritt, FAAR '37, '50, Publication Fund of the American Academy in Rome.

The Maritime World of Ancient Rome Proceedings of "The Maritime World of Ancient Rome" Conference held at the American Academy in Rome 27-29 March 2003

edited by Robert L. Hohlfelder

PUBLISHED FOR THE AMERICAN ACADEMY IN ROME

by The University of Michigan Press Ann Arbor, Michigan

2008

Copyright © by the University of Michigan 2008 All rights reserved Published in the United States of America by The University of Michigan Press Manufactured in the United States of America @Printed on acid-free paper 2011

2010

2009

2008

4

3

2

No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, or otherwise, without the written permission of the publisher.

A CIPcatalogrecordfor this book is available/ram the BritishLibrary. Libraryof CongressCataloging-in-Publication Dataapplied/or. ISBN-13: 978-0-472-11581-5 ISBN-10: 0-472-11581-2 Every effort has been made to trace the ownership of all copyrighted material in this book and to obtain permission for its use.

(Photo by Chris Brandon)

In Memoriam Avner Raban, d. 11 February 2004 A pioneering underwater archaeologist, mentor to a generation of scholars, prolific scientist, respected colleague, and friend of the contributors to this volume . We miss you, Avner.

CONTENTS

PREFACE

RobertL. Hohl/elder

ix

INTRODUCTION

RobertL. Hohl/elder

1

PLENARYADDRESS

Ports, Ships, and Power in the Roman World

GeoffreyRickman

5

SESSIONI SHIPS AND SHIPPING IN THE ROMANWORLD

Roman Shipsheds

DavidBlackman

23

Cosa and Deep Sea Exploration

Anna MargueriteMcCann

37

Roman Shipsheds and Roman Ships

BorisRankov

51

SESSIONII MARITIMELIFE AND COMMERCE

Rome's Contribution to the Development of Piracy

Philip de Souza

71

Recent Archaeological Survey at Portus

Simon Keay,MartinMillett, and KristianStrut!

97

Imports at Ostia in the Imperial Period and Late Antiquity: The Amphora Evidence from the DAI-AAR Excavations

ArcherMartin

105

CONTENTS

viii

Testing the Waters: The Role of Sounding Weights in Ancient Mediterranean Navigation

John PeterOleson

119

Artifact Distributions and Wreck Locations: The Archaeology of Roman Commerce A. J.Parker

177

OchlosNautikos:Leisure Culture and Underclass Discontent in the Roman Maritime World NicholasK. Rauh, MatthewJ.Dillon, and T.DavinaMcClain

197

SESSION III HARBORS OF TIIE ROMAN WORLD

Roman Structures in the Sea: Sebastos, the Herodian Harbor of Caesarea

ChristopherJ.Brandon

245

Protection and Conservation of Ancient Harbor Structures in Tuscany: Two Examples on the Coast of Grosseto

PamelaGambogi

255

Cosa's Hydraulic Concrete: Towards a Revised Chronology ElaineK. Gazda

265

Response to "Cosa's Hydraulic Concrete: Towards a Revised Chronology" by E. K. Gazda

Anna MargueriteMcCann

291

The Roman Maritime Concrete Study: A Brief Summary of Fieldwork from 2002 to 2005

Robert L. Hohl/elder,ChrisBrandon,and J. P.Oleson

297

Archaeological Evidence for Ships and Harbor Facilities at Berenike (Red Sea Coast), Egypt Steven E. Sidebotham

305

The Expansion of Triumphal Imagery beyond Rome: Imperial Monuments at the Harbors of Ostia and Lepcis Magna

Steven L. Tuck

325

SESSIONI SHIPSAND SHIPPING IN THE ROMANWORLD

SESSION II MARITIME LIFE AND COMMERCE

SESSION III HARBORS OF THE ROMAN WORLD

It was not until the third century B.C. that geopolitical realities beyond Italy forced Rome to recognize the importance of the sea to its own fate. Two centuries later, following the fall of Egypt in 30 B.C., Rome emerged as the dominant maritime power. Once in place, Rome's dominance of the sea became an important component of its imperial history. No other power before or since has controlled the Mediterranean basin or exercised an imperial naval tenure to such an extent. Derived from the proceedings of the conference 'The Maritime World of Ancient Rome" held at the American Academy in Rome 27-29 March 2003, this volume was conceived to provide a forum for recent research on subjects relating to the maritime life of Rome and the vast empire it created. With contributions from eminent scholars from around the world, this volume builds upon and extends the scope of the American Academy in Rome's first volume on Rome's Commerce of Ancient maritime life, TheSeaborne Rome:Studiesin Archaeology andHistory.It will be of interest to scholars investigating maritime aspects of the Roman period and to upper level students studying the maritime affairs of the Roman period.

ROBERT L. HOHLFELDER, Professor of History at the University of Colorado, Boulder, is a world-renowned expert in underwater archaeology. He has published widely on ancient maritime history, late Roman and early Byzantine history, and ancient numismatics. His many books Dream:Caesarea ontheSea. include KingHerod's He is currently at sea leading a National Geographic expedition to Rome, the AmalA Coast, and Sicily.

Cover illustration: Roman merchantman under sail entering or leaving Portus, ea. 3rd century A.D. Courtesy Fototeca Unione,AAR.

PREFACE

I

n the spring of 2001, I had the privilege of an extended stay at the American Academy in Rome (AAR) as a visiting scholar. An opportunity to join such a vibrant and diverse academic community is always an extraordinary experience in so many ways, and this one was made even more so by the opportunity to spend time with Professor Lester Little, its director. Although my research fields in ancient maritime history and marine archaeology are far from his own, his catholic interest in all aspects of classical antiquity is keen and informed. During one conversation in his office, I suggested that perhaps it was time for the AAR to host another conference on Rome and the sea. I reminded him that the last one, organized by the then-director John D' Arms and held over a two-year period in 1978-1979, had resulted in the publication in 1980 of The Seaborne Commerce of Ancient Rome: Studies in Archaeology and History, edited by J. H. D'Arms and E. C. Kopf£(= Memoirs of the American Academy in Rome 36). It was well received when published and has remained an important contribution to the literature in the field of maritime life in antiquity. I remember him asking me in his typically dry, humorous way if anything new had happened in the field since then. My response to him was in the same vein: "Yes, but not that much. We won't need two years for our conference. We can cover our material in two days." He said that I should give it a go and try to organize a symposium of leading scholars for the spring of 2003. From that exchange of banter "The Maritime World of Ancient Rome" (MWAR) was born. The road from this moment of conception to the actual conference and then to this volume has been a long one. In January of 2006, the final details regarding publication had been finalized, and this book moved closer to reality. Authors were then afforded the opportunity to incorporate recent research into their contributions, so the content of the various articles is more current than the date of the original conference would suggest. Without the help of many people, including all of the speakers who made time in their busy lives to come to Rome in March 2003 and then to revise their papers as articles for this volume, nothing would have come of the plans formulated in Professor Little's office in 2001. There are others to thank who were equally indispensable. At the AAR, Ingrid Rowland (who has since left this institution), Archer Martin, Anne Coulson, and Pina Pasquantonio all worked hard to nurture this effort and ensure that it became a reality worthy of the endorsement of their prestigious institution. Of course, the conference and this volume occurred under the enthusiastic patronage of Lester Little-sine qua non. I wish to thank him personally for his encouragement and support for our endeavors as well as his gracious public welcome at our opening session. Elaine Gazda, who serves as the chair of the AAR's publication committee, has been the inspiration for including this volume in the series Supplements to the Memoirs of the American Academy in Rome. Chris Hebert and others at the University of Michigan Press were extremely patient as I waited for the inevitable late submission of articles and figures that has delayed the final publication beyond their and my

X

PREFACE

own expectations. My thanks to them for the final copyediting and composition of this volume, as well as to the three anonymous reviewers whose suggestions for changes in the various articles have tightened and improved the final product. The publication of a book is always a momentous event for those involved in its composition. I hope all who contributed to this one feel that their efforts have helped capture in some measure the excitement and synergy of shared interests and professional comradeship that pervaded our conference. Rome's maritime life needs more attention, and as our authors demonstrate there is still much to learn about this often-neglected area of study. Robert L. Hohlfelder June2006

INTRODUCTION Robert L. Hohl/elder

R

ome came slowly and grudgingly to the sea. For most of her first five hundred years, her energies :were focused elsewhere and only intermittently on the Mediterranean. It was during this time that the Romans gained their independence as a city-state, created and sustained alliances with immediate and distant neighbors, expanded their influence in Italy and beyond, and forged a unique governmental system and way of life. Italy's geographical location in the approximate center of the Mediterranean provided the distance and safety from the great powers of the East to assure Rome's independent development in her formative centuries. Rome's site was near the sea, some 20 miles inland, and at about midpoint in a peninsula rich in the resources required by a preindustrial society. The city had been founded on the banks of the Tiber River, a navigable waterway that provided an access corridor both to the sea and to the inland regions to the northeast, at a convenient crossing point for north-south terrestrial travel. Her remove from the sea was great enough to provide a buffer of safety from opportunistic seaborne marauders, while her famous seven hills added another degree of protection for a young, developing city. Simply stated, her geographical position could not have been more ideal or advantageous. These natural blessings helped Rome pursue her early communal priorities free from the pressures of economic privations and with requisite access to transportation networks when necessity dictated their use. As their early history unfolded in Latium and then spiraled outward to other parts of Italy, the Romans never chose to embrace or to capitalize on the Mediterranean's potential as a trading highway or a means of projecting their growing military might. Mere proximity to this great sea, which they called the mareinternum or later, presumptuously but accurately, the marenostrum,did not in itself guarantee that Rome would become a maritime state or ultimately the greatest naval thalassocracy the ancient world ever produced. The legends and decipherable snippets of fact that form the canon of early Roman historical tradition do not dwell on the sea or hint that the city's destiny was linked to an intimate relationship with the Mediterranean. Rome's early vision was never maritime. It was not until the third century B.C. that geopolitical realities beyond Italy forced Rome to recognize the importance of the sea to her own fate. The city met the challenge of the Carthaginians in two wars that destroyed what arguably was the greatest navy at that time and propelled the victorious Romans to the forefront of Mediterranean history and naval supremacy. By the end of that century, Rome had achieved dominance in the western seas and had begun her involvement in the Hellenistic world to the east. Although the city that had evolved into an empire stood on the threshold of international supremacy, Rome remained uncomfortable and uncertain about her relationship with the sea. 1 Had 1 It

seems appropriate to remember Lionel Casson's (1991, 143) thoughts on Rome's rise to naval power: "The Romans

are an anomaly in maritime history, a race of lubbers who became lords of the sea in spite of themselves" or the pointed

2

ROBERTL. HOHLFELDER

her leaders decided to seize the maritime initiative that victories over Carthage had provided and to exploit fully the naval vacuum caused by her rival's demise, her future may well have unfolded in different ways. But the maritime indifference that had emerged early in her history prevailed, and Rome remained irresolute about the sea. A hinge-point of history slipped by the Romans, and the newest reluctant masters of the sea abdicated their naval dominance. Rome's fleet eroded and then vanished in the next century. The Romans chose to rely on their allies to address their maritime interests. They seemed unable to appreciate the potential or the magnitude of the prize they had won from the Carthaginians. But Rome had a second chance. Ultimately she became the only naval power in the Mediterranean late in the first century B.C. by virtue of her conquest of the states along its littoral, her naval response to a burgeoning pirate menace, and her increasing military use of ships in the final decade of the civil wars that brought about the collapse of the Republic. Hostilities between Octavian (later to be called Augustus) and Mark Antony and his ally, Cleopatra VII, the last Ptolemaic ruler, ended with the Battle of Actium in 31 B.C. After the fall of her Egyptian kingdom the following year, Rome faced no other major threats from the sea and emerged from the ashes of her internecine combat with the most powerful navy in her history. The entire Mediterranean belonged to the Romans, but this maritime hegemony had not been forged by a conscious foreign policy. Rome's thalassocracy came about more by inadvertency and contingency than by design. Once in place, Rome's dominance of the sea became an important component of her imperial history. No other power before or since has controlled the Mediterranean basin or exercised an imperial naval domination of this sea for such a long duration. Yet this aspect of the Roman experience, which might be called a pax maritima, has received less scholarly attention than it deserves. Our conference, "The Maritime World of Ancient Rome," was conceived with this deficiency in mind and to provide a forum for recent research on subjects relating to the maritime life of this city and the vast empire created by Roman expansion.2 All sessions were held at the American Academy in Rome during 27-29 March 2003. This colloquium began on Thursday evening, 27 March, with an opening plenary address by Geoffrey Rickman of St. Andrews University on "Ports, Ships, and Power in the Roman World." With his usual panoptic thoroughness, he presented the current state of research on the theme of this conference with a particular emphasis on the need to study and understand ports in a broad economic context. He wove the work of the scholars who would be offering papers into his survey in a flattering and relevant way. His inaugural lecture, and the reception and dinner that followed, set the perfect tone for the presentations to come. The first full day of papers on Friday, 28 March, was organized into two sessions: "Ships and Shipping in the Roman World" and "Maritime Life and Commerce. "3 David Blackman and Boris Rankov presented linked papers, "Roman Shipsheds" and "Roman Shipsheds and the Roman Navy," which both stressed the importance of finding clearly identifiable Roman shipsheds at archaeological sites, a task that has proven difficult even though these installations are frequently mentioned in epigraphical, literary, and pictorial evidence. Both papers served as an introduction to a new research project directed by these two scholars, "Shipsheds of the Ancient Mediterranean," which, as its name suggests, will be a comprehensive study covering all periods of antiquity. comment of Chester Starr (1960, 171) that throughout Rome's long history one constant attitude was "an inveterate Roman disinterest in the sea." 2

A report on this conference by David Blackman and Boris Rankov, two of its participants, appeared as Blackman and

Rankov 2004. 3

The titles of the papers presented may not correspond exactly with the articles submitted for publication in this volume. Additional coauthors, who were not present at the conference, also appear on some reports.

INTRODUCTION

3

Andrea Camilli, the archaeologist from the Soprintendza per i Beni Archaeologici della Toscana currently in charge of the Pisa shipwreck excavations, gave a progress report on his recent work discussing the controversy that has arisen over the nature of the site itself. His paper, "Port, Harbor, or Archaeological Deposit: The Pisa Shipwreck Site Evidence," focused on the question of whether the find spot marks an actual harbor or simply a concentration of wrecks in a particularly dangerous river bend. 4 Anna M. McCann closed the morning session with a report on her deep-water research off Cosa during the summer of 2002. Her report, "Cosa and Deep Sea Exploration," referred to her earlier work at Cosa and the Skerki Bank but primarily was devoted to her explorations with an AUV (autonomous underwater vehicle) as a search tool off Cosa, fieldwork conducted in collaboration with Dr. Moccheggiani Carpano and the Soprintendza per i Beni Archaeologici della Toscana. The afternoon session, "Maritime Life and Commerce," featured papers by A.J.Parker, "Artifact Distribution and Wreck Locations: The Archaeology of Roman Maritime Commerce"; Archer Martin, "Imports at Ostia in the Imperial Period and Late Antiquity: The Amphora Evidence from the AAR/DAI Excavations"; John P. Oleson, "Testing the Waters: Sounding Weights and the History of Navigation and Maritime Trade"; and Simon Keay and Martin Millet, "Portus and Rome's Maritime Fac;ade." Parker stressed the positive role the sea played in facilitating trade and reminded us that trade routes crossed the open sea and were not simply in coastal waters. Martin provided a series of charts of amphora samples from four different periods of Ostia's floruit and discussed the implications of these distributions. Oleson expanded on his ongoing research on sounding weights in a lively and humorous paper. Keay and Millet reported on their recent surveys at Ostia, concluding that the Claudian harbor was much larger than expected while offering many new aerial views of Portus. This session concluded the following morning with two papers by Nicholas Rauh, "Ochlos Nautikos:Maritime Mobs and Seaborne Culture in the Roman Mediterranean World," and Philip de Souza, "The Roman Invention of Piracy." Rauh offered a vivid picture of maritime life in Roman ports, including accounts of the taverns, inns, and brothels, while de Souza described how Rome and individual Romans like Pompey and Octavian used the common fear and stigma of piracy for their own political ends. The next session, "Harbors of Antiquity," provided a different focus point for the examination of Rome's maritime world. Pamela Gambogi of the Soprintendza per i Beni Archaeologici della Toscana reported on "Protection and Conservation of Ancient Harbor Structures in Tuscany: Two Examples on the Coast of Grosseto," presenting recent conservation and protection efforts at Cosa and a Roman villa site at nearby Santa Liberata. Steven Sidebotham discussed an important gateway for maritime trade in the Red Sea and beyond in his report, "Archaeological Evidence for Ships and Harbor Facilities at Berenike." Elaine Gazda in her paper, "Cosa's Hydraulic Concrete: Towards a Revised Chronology," reviewed the published data from the Cosa excavations and offered a later date than McCann's second-century B.C. provenance for the site's hydraulic concrete installations. Her report produced a spirited rebuttal by McCann in the discussion session that followed. The result of this scholarly exchange was an editorial decision to include a response by McCann to Gazda's arguments, which is included in this volume immediately after Gazda's paper. Steven Tuck's "The Harbors at Ostia and Leptis Magna: Triumphal Monuments Outside Rome" stressed in particular the temples that adorned these harbors. Chris Brandon offered new insights into the construction 4 Dr. Andrea Camilli chose not to submit his paper for publication in this volume.

4

ROBERTL. HOHLFELDER

of the vast artificial harbor installations at Caesarea Maritima. He suggested in his paper, "Roman Structures in the Sea: The Harbor of Caesarea Maritima," that the building program included the construction of a spine of concrete pilae, or piers, on the southern breakwater that were joined by an in-fill of rubble. This method of breakwater construction is not mentioned by Vitruvius in his De architecturaand may have been implemented to meet the unique natural challenges confronting the builders of Caesarea's harbors. The last session of the formal conference ended with a report by R. L. Hohlfdder, J.P. Oleson, and C. Brandon on the first season of the Roman Maritime Concrete Study (ROMACONS) entitled "The Roman Maritime Concrete Study at Antium and Portus, 2002." It offered a discussion of the intended scope of this pan-Mediterranean research project, the methodology devdoped and used to collect concrete cores for physical and mechanical analysis, and a brief report on the scientific results obtained to that point in time. Since this oral report, several articles concerning this project have appeared, superseding this prdiminary account and rendering it redundant. The coauthors decided not to publish it in this volume but instead to include a summary of fi.ddwork from 2002 to 2005 along with references to recent ROMACONS's publications. In an effort to insure adequate discussion of all the papers, each set of reports was followed by 30 minutes devoted to questions and comments. In addition, at the end of both days of the conference, an open-ended session was hdd that began with short communications by postgraduate students(= "the next generation") and others to present aspects of their maritime research. Presenters included Catherine Bouras, Mamoru Ikeguchi, Maria Costanza Lentini, Christa Steinby, and Isto Vatanen. "The Maritime World of Ancient Rome" conference expanded our knowledge of Rome and the sea. Our collection of essays builds on and augments the scope of the AAR's first volume on Rome's maritime life, The SeaborneCommerceof Ancient Rome:Studiesin Archaeologyand History(edited by J. H. D'Arms and E. C. Kopff as MAAR 36). That book became an immediate standard in the literature on Roman interaction with the Mediterranean and has remained so since its publication in 1980. All who are associated with this volume hope that it will mirror the success, impact, and longevity of its predecessor.

WorksCited Blackman, D., and B. Rankov, "The Maritime World of Ancient Rome," InternationalJournalof Nautical Archaeology33.1 (2004) 173-175. Casson, L., The Ancient Mariners(Princeton 1991). Starr, C., The Roman ImperialNavy (Cambridge 1960).

PLENARY ADDRESS

PORTS, SHIPS, AND POWER IN THE ROMAN WORLD Geoffrey Rickman

I

n 1979 I was privileged to take part in the conference held here in the American Academy in Rome entitled "The Seaborne Commerce of Ancient Rome." That was twenty-four years ago, and it is an honor to be invited to participate again in this conference on "The Maritime World of Ancient Rome," which seems to me to be an even bigger subject. Of course, at the age of seventy I know less, and am less certain of what I do know, than when I was forty-six. Then I talked about the grain supply of Rome under the Empire. 1 This time I want to talk more generally about Roman ports and their maritime world, on which I started to work from the early 1980s-work much interrupted since, I might add-though, as you will see, the role of grain in this world still preoccupies me. I offer a sort of general meditation on our theme, perhaps provocative, perhaps rather dull. My aim is to give an overview of our subject as I see it, without a mass of detail (and without illustration), to pave the way for the more specific contributions by others that are to come over the next two days. If one looks at the history of the study of Roman ports, one quickly gets a measure of the great difficulties that are involved and of the exciting possibilities. The earliest studies tended to concentrate on one port, such as Dubois (1907) on Puteoli or Jondet (1916) on Alexandria, and they can still have great value. Only rarely was there an attempt to cast the net wider, as with Columba's I porti antichidellaSicilia(1906) or the even earlier work by De Fazio in the 1820s and 1830s on harbors in the Bay of Naples. Sometimes such work grew out of a desire to improve contemporary port design and construction, or even the feasibility of reviving an ancient port for modern use, as in 1835 with Carlo Fea and the Neronian port at Anzio (Antium). It was Lehmann-Hartleben ( 1923) who made the first proper academic study of ancient harbors in general in the Mediterranean. This remarkable work, discussing and cataloguing some 300 or so harbor sites in the Greek and Roman world, is now naturally out of date. It relied for its evidence essentially on ancient literary sources, combined where possible with travelers' reports of surface remains that were, or had been, visible. Moreover, as its subtitle revealed, its purpose was to look at harbors as part of the history of town planning. Valuable though this may be, it is only one of the many standpoints from which ports and harbors can be viewed. The nearest thing to a replacement for Lehmann-Hartleben, at least in terms of comprehensiveness-though he would, I think, disclaim such an honor-is David Blackman's survey of 1982, published in two parts in the InternationalJournalof NauticalArchaeology.But we see at once when we read that survey that we are in a different world from that of Lehmann-Hartleben. Blackman, after all, had to take on board new categories of evidence: the work of underwater archaeology, opened up by Cousteau and his colleagues after the Second World War, in the 1940s and '50s;2 of

1

Rickman 1980a.

2

Muckdroy 1980.

6

GEOFFREY RICKMAN

aerial photography, pioneered by Poidebard at Tyre in the mid-1930s, and at Sidon in 1945, and vastly developed since;3 and new work on geomorphology, which air photographs in particular revealed to be of the utmost significance to the understanding of coastline changes and of the location of ports. 4 The range and type of evidence, therefore, have changed profoundly. But the truth is also that we want to ask quite different questions from those Lehmann-Hartleben asked. We want to know much more, in a more sophisticated way, not just about structural remains but about social and economic history, about administration, about religion, about ideology and the projection of power. Above all, we want to understand properly the connectedness of the Mediterranean world, centered on the sea, which has recently been the subject of an intensely interesting and exciting book by Peregrine Harden and Nicholas Purcell, The CorruptingSea (2000).5 The scope is simply immense. The thoughtful introduction to Anna McCann's great book on The Roman Portand Fisheryof Cosa (1987) gave us perhaps the first real idea of it. When I started to think about Roman ports in the early 1980s, already the scale of the difficulties led me to call for "imaginative co-operation between historians and archaeologists." "Roman ports," I argued at the time, "will then take their proper place in the changing history of Roman government and society."6 I later discovered that the French scholar Chevallier was already calling in 1967 for interdisciplinary and international cooperation in the study of navigation and ancient ports. 7 In 1987, twenty years after that, Jean Rouge-whose enormous book, Recherchessur l'organisation du commercemaritimeen Medite"anee sous!'Empireromain(1966), is the great quarry from which we all work-said bluntly in an article that the redoing of Lehmann-Hartleben was "beyond the work of one man"-"un travail enorme qui ne pourrait etre l'oeuvre d'un seul homme." 8 Hence the importance of international gatherings such as this one here in Rome. So what have we learned so far? I think it is that everything is much more complicated even than we thought it was. Just as Puteoli's port in Campania was not simply the famous arcaded mole, and the few walls of the so-called docks studied by Dubois or Lehmann-Hartleben, hut the vast complex of the ripa hortensianasweeping north to Portus Iulius and round toward Misenum, shown on Giuseppe Camodeca's plan; 9 so Portus, on which we have exciting new work by Simon Keay and Martin Millet, was not just the Claudian and Trajanic basins on the coast of Latium but the spread south to the river Tiber at Ostia and east up the river to Rome itself, with satellite harbors both north and south at Centumcellae, Antium, and Terracina; 10 so Alexandria's port was not just the great harhors on either side of the causeway, the heptastadionleading to the island of Pharos with its astonishing lighthouse, but stretched back southwards, as Jean-Yves Empereur has emphasized, to the even greater wharves of Lake Mareotis behind the city and eastwards to the harbor of Canopus and the other mouths of the Nile-the river that disgorged goods not just from Egypt but from the Red Sea, Arabia, India, and the East; 11 so Carthage's Roman port, as Henry Hurst has reminded us, was not just the famous internal double set of harbors, circular and rectangular, whose location and 3

Poidebard 1939;Poidebard and Lauffray 1951.

For example, Davidson and Shackley1976;cf.Barker and Mattingly 1999.

4

5

Cf. Malkin 2003; Harris 2005.

6

Rickman 1985, 112.

7

Chevallier1967.

8

Rouge 1987, 163.

9

Camodeca 1994;Frederiksen 1984,324, 353.

10

Rickman 1991.

11

Empereur 1998,214.

PORTS,SHIPS,AND POWER IN THE ROMANWORLD

7

shape went back to Punic times, but embraced the vastly wider waters of the Lake of Tunis much closer to the south and east of the city then than now 12-something that makes Carthage's maritime complexity under the Roman Empire much more understandable. The same pattern can be seen repeated empirewide. Antioch in Syria, 24 km inland up the river Orontes, not only had its river docks but a broad lake nearby and a harbor down at the river mouth at Seleucia-in-Pieria, as well as a support harbor to the south at Laodicea. Ades, inland on the Rhone in Gaul, had access to the sea by river but also by canal, the Fossae Marianae, dug by Marius's soldiers during the Republic, to another outport. Narbonne and Aquileia both had complicated networks of river, the Aude and the Natiso respectively, inland lagoons, and canals to link up all their port facilities. Something similar can be posited for the setup on the Guadalquivir in southern Spain from Hispalis (Seville) down to Gades (Cadiz) on the coast.13The great ports of the Roman world could without doubt be extensive, intricate, and complex beyond our previous dreams. They must be thought of as great clusters of facilities, set in wide webs of communication by road and by water. To some extent this perception is all part of the knock-on effect of the new categories of evidence. Underwater archaeology has revealed in a way not previously understood the importance of the sea in the Roman world and the sheer density of sea traffic. We have known, of course, for a long time that the relative costs and difficulties of transporting goods by land and by water were heavily in favor of movement by water, and in particular by sea. 14 Even so, the density of sea traffic newly revealed is a surprise. Caution has to be exercised in drawing conclusions and making generalizations from shipwreck evidence, as Toby Parker has properly stressed: 15 because some areas, for example off the south of France, have been more intensively investigated than others; because some of the excavations in the early days, such as those at Grand Congloue, were less expert than they might have been; and because deep sea areas of the Mediterranean are beyond the reach of the Aqua-Lung, only operational, I think, to a depth of some 200 feet. Nevertheless it is clear from the more than 1,500 wrecks so far discovered that in the period from 200 B.C. to A.D. 200 there was an intensity of traffic by sea that was not to be matched again for a thousand years. That is astonishing. It means that the Mediterranean Sea was of the utmost importance to the world of Rome and to its empire. This is not necessarily the impression we would get from reading our traditional sources, which exaggerated Roman ignorance of and indifference to the sea. Yet on reflection there are some indicators from the beginning of a different sort of story. From perhaps the sixth century B.C. Rome had had a series of treaties with Carthage, the great maritime power, which defined among other things where Roman ships might and might not go. There is reason to believe that Roman trade by sea in the early Republic was more vital than has been thought. In 313 B.C. the Pontine islands, one of the keys to navigation in the area, had been occupied, and by 311 B.C. Rome had created the office of duumvirinavales"to repair and equip a fleet." In 267 B.C. were added the quaestoresclassici "quaestors of the fleet," stationed in various coastal towns such as Ostia at the Tiber mouth. 16 More important still, Rome possessed navalza,a set of docks and shipsheds for her warships, located at a significant point on the edge of the Campus Martius. Both David Blackman and Boris Rankov are to talk about shipsheds and their role in the Roman world. ill-informed though we are about Rome's own shipsheds, we know that they already existed by the fourth century B.C., when 12 Hurst

2002; cf. Lancel 1995, 172-192, esp. 185-188.

13 Rickman

1996.

14 Hopkins

1983a.

15 Parker 16

1984; 1992a.

Cornell 1995, 388.

8

GEOFFREY RICKMAN

some of the ships of defeated Antium were lodged there, and that they were greatly embellished in the second century B.C. Given the association of the building of such dockyards with the assertion of a claim to power, seen most clearly in the case of Athens and her naval empire, Rome's possession of navaliais, to say the least, interesting. 17 Intermittent though Rome's actual efforts by sea may be during the late Republic, a claim to power subsisted, I suspect, making her wary, and intolerant, of attempts by others at maritime dominance. Underwater archaeology has provided not just such general stimulus to renewed reflection but all kinds of specific new knowledge, which itself provokes further thought. Cargoes have in some cases protected hulls, which can, therefore, be minutely studied for their carpentry. It is Richard Steffy preeminently-who calls wooden ships and boats "the most marvellous structures ever built by humankind" -who has shown the immense sophistication and strength of this carpentry. 18 Ancient shell-first constructed ships, with their double planking, edge-joined, held tightly by mortise and tenon joints, were built almost like furniture or fine examples of cabinetmaking. Their construction drew on a long tradition of shipbuilding, which Steffy himself has helped to illuminate, not least with his extraordinary work on the Kyrenia ship, a small, long-serving Greek cargo vessel of the fourth/third century B.C. The maritime world of Rome took such technical sophistication for granted and developed it in significant ways. In particular, though never deserting shell-first construction, the Romans, by wedding frames, and half-frames, to keels, could create ships with rounder hulls and flatter bottoms, and therefore holds for cargo of greater volume with less draft. But Steffy is not so obsessed by such technical details that he misses the wider picture of human endeavor into which they all fit. He puts matters succinctly when he says, "Ships and boats were always the means to an end and usually that end was profit, convenience, security and victory." Immediately we are reminded that in the world of Rome-which developed a great variety of ship types-the needs of war, or, I suppose, of piracy, about which Philip de Souza willtalk, may differ significantly from those of trade and the usages of harbor boats from those of sea-going vessels. We have new evidence on this from discoveries at Ostia, and spectacularly now at Pisa, about which Andrea Camilli is to speak. Not just the hulls of ships but the cargoes themselves allow us to speculate about the question of ship sizes. The French scholars Tchernia and Pomey, who helped to carry out the scrupulous excavation of the wreck at Madrague de Giens off Marseilles, with its hold full of wine amphoras, have written what seems to me the best study of maximum tonnages carried by Roman ships. 19 From this emerges a believable picture of a variety of sizes from quite small, 70 tons or less, which would have been the vast majority, to quite large, 300-400 tons being not uncommon. What is as yet unproven is Lionel Casson's belief that some of the grain freighters on the regular run from Alexandria to Puteoli could carry as much as 1,200 tons. That was deduced from literary sources and is as yet unconfirmed by underwater archaeology. Maybe such confirmation willnever be forthcoming since wheat as a cargo would hardly protect the hull; or perhaps the answer lies in deep water, which such ships must have crossed, and about the survey of which by means of robotics Anna McCann is to speak. Tchernia and Pomey believe, sensibly, that there was nothing in the technology of antiquity that would prevent the construction of such exceptional ships-and we know that the transport of obelisks and other heavy material for the building schemes of the emperors could demand, and secure, ships of such a size. 17 Purcell 18

1996, 269; cf. Coarelli 1997, 345-361.

Steffy 1994, 23-78, esp. 23, 77.

19

Pomey and Tchemia 1978; cf. Houston 1988.

PORTS,SHIPS,AND POWER IN THE ROMANWORLD

9

The deciding factor in their regular use would be economic convenience or necessity. But on that too there can be differing views, with Tchernia and Pomey not averse, I think, to such a possibility of regular use for grain but Keith Hopkins skeptical because of what he saw as the ruinous cost that would be involved in the loss of such a ship and the difficulty of finding ports it might use if it got into trouble on the voyage. I have argued elsewhere that these are not insuperable problems and that the Roman annona,at the very heart of the exercise of Roman power under the Empire, was exceptional. 20 I would not yet rule out the possibility of such grain carriers. The essential movement of such foodstuffs, after all, to compensate for the chronic variability of staple crops in the Mediterranean, was the oldest of truths, which underlay the very creation of cargo shipping itself in the Greek and Roman world. 21 The Roman annona embodied that ancient truth but in a form writ large, and with all the resources and compulsion of imperial power. But cargoes allow speculation about more than ship sizes. What was carried, in what quantity, with what regularity, in what containers, how packed and stowed, provoke questions about patterns of trade. Toby Parker, who is the great expert on all this, has already shown how even in quite small loads the cargoes were often very mixed-with, for example, amphoras, pottery, millstones, and ingots from different places in the same hold. 22 That has implications for the complexity of commercial activity, transshipment, and harborside wheeling and dealing; no simple linear model, of straight-line tramping from port to port, will do. More important, it also allows, with suitable caution, the development of views about circulation patterns of certain goods, whether relatively localized or spread further afield. This is a building up from the bottom, as it were, of a certain amount of hard evidence, on a microeconomic scale, to match or temper the macroeconomic speculations of historians about how the circulation of goods worked in the Roman Empire. A famous example of what can be achieved in this way is the story of the Sestii from the area near Cosa. Careful study of the amphoras-more of which have now, amazingly, turned up in Anna McCann's deep-water survey at Skerki Bank off northwest Sicily-has allowed the reconstruction of a trade in wine, which flourished for a century or more in the late Republic, particularly with southern Gaul. 23 But other names, such as Publius Veveius Papus from Fundi near Formia in south Latium, Trebius Loisios from either Ischia or Pompeii, or the Pirani family from Campania, can now be cited to help fill out the picture. The rise and fall in such trading patterns can be plotted with some confidence, even changes from shipping special Italian vintages to that of vin ordinaire, carried not in amphoras at all but in great dolia bedded in the holds of ships. Something similar can be repeated to an extent in the study of trade from Spain in wine and oil in the early Empire and later in the second century A.O. from Africa.24 They form particular case studies-and both Toby Parker and Archer Martin are to talk on these matters-which help to give welcome precision to some of our generalizations. But, intriguingly and importantly, I think, they also shadow the workings of empire and imperial power. What underwater archaeology and aerial photographs-particularly those in the great collection by General Giulio Schmiedt of the Italian coastline-have really allowed us to see is that the Roman possession and enjoyment of the sea came ultimately to be expressed in the fullest possible way.25 Maritime villas, for example, in favored areas, like the Bay of Naples, flaunted the wealth and taste of their owners to the passing shipping and with their towers and colonnades helped in 20

Rickman 1991.

23

See now Will 2001.

21

Parker 1996, esp. 99.

24

Throckmorton and Parker 1987.

22

See, in addition, Parker 1992b.

25 Schmiedt

1970; Horden and Purcell 2000, 125-126.

GEOFFREYRICKMAN

10

tum to provide recognizable day-marks for navigation by such shipping. Such villas had their own miniature harbors and mooring facilities, just as ordinary workshops and craft industries in places like the Istrian peninsula in the northern Adriatic had jetties and moles for loading their goods. 26 Along the edges of the sea were great areas of salt pans, as at Ostia, or fish farms, as at Cosa; these are in fact to he found all over the place, from the Black Sea to the Straits of Gibraltar, with the greatest concentration perhaps along the south coast of Spain, in Baetica. They could be large commercial establishments for the production of both salt fish and garum, fish sauce, which came to be exported all over the Roman world. 27 But puzzles persist. When we say, "came to he exported all over the Roman world," what exactly do we mean? and what mechanisms are we implying? Clearly there was a dense scatter of material objects and exported goods throughout thisworld, found not just by the sea but deep inland. But was the circulation really empirewide and the result, directly or indirectly, of the state's need to levy its taxes, pay its soldiers, feed its armies and the privileged city of Rome at the center of it all, as Keith Hopkins argued? Or was the circulation broken down into smaller, more regional areas, as Duncan-Jones has insisted-though he also, I think, would see the state as crucial, wittingly or unwittingly, in creating even such regional flows.28 The role of the state, and its power, is hard to define properly, pervasive yet elusive, permeating the background but not, overtly at least, in control of much that went on. The state was most palpably present in this maritime world in the form of the Roman navy. Augustus after the defeat of Antony at Actium collected together the remnants of various battle fleets for a period at Forum Iulii (Frejus) in southern France. He then established new permanent naval bases at Misenum on the Bay of Naples and Ravenna on the Adriatic. Smaller naval detachments were placed elsewhere, for example on the coast of North Africa and on the Black Sea, so as to maintain a permanent naval control. The aim not least was of course to secure the safety, and the power, of the emperor himself by controlling all the approaches to Italy; as Tacitus tartly pointed out: "Armies, provinces, fleets, the whole system was interrelated. "29 But a permanent navy also helped inhibit piracy, sea raiding, and general lawlessness at sea, even if it did not end them. Yet there was no state merchant marine, unless the so-called Alexandrian fleet, the stolosAlexandrinos,for transporting grain supplies to Rome from Egypt, might he deemed one-which is highly doubtful. The state was certainly concerned to see that its imperatives in this sphere were met, whether feeding Rome or the armies, but this was achieved not least by leasing contracts to shippers and merchants. 30 It was part of a pattern of "farming out" almost everything, from building contracts to tax collection, which went back to the earliest days of the Republic. The actual machinery of state, magistrates and officials, was quite small, which made such a pattern of behavior inevitable, and that hehavior ensured in its tum that state machinery remained on a modest scale. In the late Republic and under the Empire there was an attempt to create greater security in grain supplies by offering privileges to those who served the annonafor certain periods of years. In the end, I think, that did help to create the system typical of the late Empire, where groups of shippers organized in collegia and corporawere, or were meant to be, instruments of the state's will. But that system was a long time coming and typified a period when Roman maritime control was in fact beginning to slip. Curiously, one area where direct intervention and control by the state has been suspected in the 26

Degrassi 1955; Bezeczky1995.

27 Curtis

1988; 1991.

29

Starr 1989, 67-81; Tac.Ann. 1.9.

30 Rickman 1980b, 129-132 (Alexandriangrain fleet),87-93

(contracts and privileges). 28 Hopkins

1980; 1983a;1983h;Duncan-Jones 1990,48-58. See now Temin 2001.

PORTS,SHIPS,AND POWER IN THE ROMANWORLD

11

early Empire is in the trade with the East, which Lionel Casson and Steven Sidebotham, who is to talk on the port of Berenike, have done so much to illuminate. 31 A mixture of remarks in Pliny the Elder about the drain of wealth from the empire to Arabia, India, and China and our knowledge of the careful provision of guard posts and drinking wells on the caravan roads from the Red Sea ports across to Coptos on the Nile (which we know to be the main route for oriental goods to Alexandria) has provoked the vision of the state with a central and beady eye on this traffic. But though the state certainly took a hefty 25 percent tax, portoria,on all the goods passing to and from the East-much higher than the 2.5--4percent levied elsewhere at portoriaboundaries in the empire-and though the emperors and upper classes enjoyed the fruits of this trade, spices and peppers, precious stones and textiles, to the full, the trade itself seems to be in private hands. Papyri make clear that wealthy private families in Egypt, as well as others, maintained their own agents in the Red Sea ports for handling their family interests in this overseas trade with India, just as Palmyrene families managed, and grew rich on, the goods coming overland from the east through Palmyra, a "port of the desert. "32 The route from the Red Sea ports to the Nile did need considerable organization, and the state may have been involved, both to ensure the collection of its tax revenue and to control raiding nomads. But there were important quarries in the area, for example at Mons Claudianus, which provided materials (including the columns of the Pantheon) for major building projects in Rome and elsewhere. The emperors were deeply concerned with such prestige projects, and that may help to account for a significant part of their interest in the security and stability of this area. In direct control of the ports themselves, whether on the Red Sea or in the Mediterranean at large, it is difficult to find officials, who were appointed by the state-unless the prae/ectusmontis Berenicidisin Egypt, the military commander of the whole Eastern Desert, qualifies for this honor. On this topic in general we hardly seem to have progressed much beyond the scrupulous paper given by George Houston here in the Academy in 1979.33 He pointed out that, when we look outside Ostia and Puteoli, where state appointees are to be expected because of the feeding of the capital, we find no trace of such officials. He concluded that it was the local municipalities themselves who saw to such appointments from among their own local elite. Emperors were, of course, interested in harbors and their provision, even the emperor Gaius, concerned to improve the harbor at Rhegium, though that was on the grain route to Rome. Trajan rebuilt the harbor at Ancona, Vespasian at Seleucia-in-Pieria, and Septimius Severus at his hometown of Lepcis Magna, about which Stephen Tuck is to speak. These actions, however, when they occur, seem to have either specific personal or some special-perhaps strategic-motivation; or they are part of the general desire of emperors to be seen as the greatest of all municipal benefactors, capable of spectacular acts of euergetism.Even Hadrian's particular generosity in harbor provision, commented on by Dio, was seen simply as part of a pattern of benefactions to cities, along with water supplies and other civic amenities. It was not an intervention to benefit the economic life of the empire, even though that may have been partly the result. Imperial interest in harbors did not, therefore, extend to central control of them, though provincial governors, as in the famous inscription from Ephesus, could always intervene. As in other periods of history, there was a mixture of public and private involvement in ports, with local groups shouldering much of the burden of repair, renewal, and administration in normal times and with state intervention only in special circumstances. 34 31 Casson

32 Young

1989; Sidebotham 1986. 2001, 54-66, 170-173; Millar 1998, 131-137.

33 Houston

1980.

(Rhegium) Joseph. A] 19, 205; (Ancona) CIL 9.5894; (Seleucia-in-Pieria) Berchem 1985; (Lepcis Magna) Bartoccini 1960; (Hadrian) Dio 69.5.3; (Ephesus) SEC 19.684 L. Antonius Albus A.D. 147.

34

GEOFFREYRICKMAN

12

The social and economic structures by which the seaborne trade itself was carried on in the Roman world have also been seen as problematic. There is something of a tension between the fact that inscriptions recording traders and merchants seem to reveal people of rather low status of only local significance-and Nick Rauhis to talk about some social aspects of the Roman maritime world-and our belief, derived not least from the fineness of the quality of the shipbuilding (and known comparative costs of such work from other historical periods), that great wealth was needed to indulge in sea-trading on any scale. But the wealthy senatorial upper class in Rome had from the third century B.C. been forbidden to own ships above a certain size, and senatorial literature in both Republic and Empire is permeated with a snobbish and patronizing attitude to trading and sea ventures generally. The resolution of the puzzle seems to lie in recognizing that, although trading "companies" as such, and as known in other periods of history, did not exist in antiquity, individuals of moderate wealth and only local importance could come together, pool their resources, and act in one another's interests in ports and other key locations around the Mediterranean world. The links are half hidden under words like societas,familia,or even simply amid, "friends." I believe also, though the evidence is never quite as clear-cut as one might hope, that, as John D' Arms argued, senatorial families, and even emperors, the wealthiest of all, could and did use their slaves, their freedmen (who were still bound to them in duty), or other intermediaries (negotiatores) to act commercially, while their own involvement was concealed. 35 Certainly funding for some trading ventures by sea from the great ports mentioned earlier demanded large sums of money, which were available, by whatever means, when needed. But not all Roman ports were like this,as Robert Hohlfelder and Lindley Vann have reminded us with their recent work at Aperlae in 4'cia on the south coast of Turkey.36 Here a small provincial city survived, and sometimes prospered, over a millennium during the Hellenistic, Roman, and Byzantine periods in a most unpromising environment. It was cut off by harsh terrain rising abruptly from the sea, both immediately behind it and around it. Without elaborate harbor or docking facilities, it was linked to the outside world by means of cabotage,that coastwise movement of shipping which was typical of the ancient, and even of the mediaeval, Mediterranean. The nature of ancient navigation, about which John Oleson is to speak, dependent on wind and human muscle, and "picturing" the sea world by means of coastal profiles and landmarks, dictated a need for harbors of refuge, if nothing more, every 50 to 70 km. Cabotage,by far the most common form of sea-trading in antiquity, linked to the immemorial tradition of the coastal periplous,needed regular stops of an unpretentious kind but where a cargo might be bought or sold. The two fitted together well-physical need and socioeconomic activity. What made the difference in Aperlae's case was its fishing activity, the harvesting of murex, the marine mollusk from which purple dye was made. This commodity was valuable and could be moved by boat along the coast to Andriake, the international emporium of the town of Myra just inland. From there it was transshipped to textile centers throughout the Mediterranean. Aperlae, in other words, though in so many ways ill-favored, fitted into a bigger world, and it was that larger world that made her survival possible. It was the way short-range, middle-distance, and long-distance trade all meshed and fed into one another that fueled the economic life of the empire. Aperlae, with its niche market in murex purple, reminds me of some of the smaller ports in the Adriatic, up in the lstrian peninsula, with particular products that were ultimately destined for a wider world. But it reminds me also of the Italian port of Amalfiin the early medieval period, on which Barbara Kreutz has done remarkable work. It was similarly cut off, backed up against mountainous 35 D'Arms

1981.

36 Hohlfdder

and Vann 2000.

PORTS,SHIPS,AND POWER IN Tiffi ROMANWORLD

13

terrain, with minimal facilities, but it too became a maritime success in what Kreutz calls "a world of maritime peddlers." What she stresses is that its success is an argument against geographical determinism; it was not geography that determined Amalfi's--or, we might add, Aperlae's-success or failure but human factors, the "worlds" into which they must fit.37 This leads us on to a wider geographical perspective of the Mediterranean as a whole and to a wider temporal sweep, namely its history beyond the end of antiquity, both of which are characteristic of Horden and Purcell's new book, The Co"upting Sea. Their work has a multiplicity of themes, which it will take a long time to absorb properly, but a major focus is on what they call "connectivity," and their contention that "connectivity," particularly by sea, was the major characteristic of the fragmented Mediterranean world. This "connectivity," if I understand their view correctly, is not, or not simply, a geographical concept; it is, on the contrary, very much a social concept. It is not geography that determines connectedness but complex human endeavors over time. This is a Mediterranean world of palpitating, quivering connections between regions and micro-regions, a living matrix. For them, therefore, ports, from the most elaborate to the most basic, are simply "nodes of density in a matrix of connectivity." Such nodes may, over time, intensify, or weaken, or even shift a little in location, as, for example, with Aquileia, Grado, and Venice, or Carthage, Tunis, Utica, and La Goleta. 38 A phrase such as "nodes of density" may seem like jargon, but it has the great advantage that we avoid becoming entangled in any artificial and overly rigid system of classifica39 It also encourages us to think flexibly and tion of ports and landing places, a typologiedes escales. subtly about the relation of the parts, particularly the world of cities and their territories-the live transformers, as it were, in the system of production, consumption, and exchange of the empire-to the dynamic whole, which was the throbbing maritime world of Rome. 40 But what about the more lowly organization of port areas, rather shirked by these high-level ways of talking and thinking? Marseilles is of interest here because of the recent discovery, right up against one of the quays of the Roman port, of numerous small wooden wax tablets, each perforated with a hole, and with marks that connect them to the 4 percent portoriatax collected in the Gallic region. Hesnard and France widen out their discussion of these finds, interestingly, to the whole question of the counting and control of merchandise as it was loaded and unloaded from ships and stored in quayside warehouses. 41 Scholars have always been puzzled by the strange object, shaped like the backbone of a fish, held by an official pictured in a mosaic at Ostia and in the famous wall painting of the river boat Isis Giminianain the Vatican. In each case he seems to be counting, by some method, the measuring of grain in a modius. Hesnard and France raise the possibility that the object is in fact not rigid but a knotted string, on which counters such as the wooden tablets at Marseilles were collected, and hung downward, as the porters of goods handed them in. In support of such "handing in" they cite the famous relief of carriers of wine amphoras at Portus handing in some sort of check or counter to tabulariiat a desk, as they come down the gangplank onto shore. Further recent evidence of such detailed counting, for receipt, storage, and export, comes from Carthage, from an archive of ostraka of the late fourth century A.D., newly restudied by Pefia, containing elaborate records of oil measurers, mensoresolei, at the port. 42 37 Kreutz

1988, 109.

38

Horden and Purcell 2000, 123-172, 393 ("nodes of density").

39

Gilissen 1974.

40 Wilson

2002.

41

France and Hesnard 1995.

42

Pefia 1998.

14

GEOFFREY RICKMAN

I have argued elsewhere that evidence such as this, and the voluminous Egyptian material, implies that there was a vast supporting structure for trading and the movement of goods in the Roman world, which depended, more than we have realized, for its proper functioning on having full and accurate records at every level.43 Admittedly in these cases we are dealing with state activity, and I do not know whether Hesnard and France's particular hypothesis is correct. But careful counting must have been a general, and crucial, requirement in maintaining an ordered use of any substantial port, whether for state or private goods. I have always hoped for greater illumination on these matters from the Red Sea ports. The dry, desert conditions should be right for the preservation of all kinds of documents, as indeed they have been not far away at Mons Claudianus. In interpreting any such documents, we might have to make allowances for the special peculiarities of the trade in the Red Sea and Indian Ocean, vividly known to us from the ancient merchant guide, the Periplusof the ErythraeanSea.But I hoped that they would help to inform us about the ordinary rhythms of port life, port captaincies, the berthing and unloading of ships, and so on. So far, unless I have missed something, despite the fascination of the work at Myos Hormos (firmly located at last at Quseir-al-Quadim) and Berenike, it still looks like a "black hole" so far as administrative evidence goes. Yet voluminous records there must have been, everything from ships' papers that needed checking, through agreements for credit loans, to receipts of all kinds. Their existence, once, in port areas, even if they are now lost, has profound implications, I think, for the sophistication of the Roman maritime world and for the sheer pervasiveness of literacy-sub-elite literacy, if you like-within it. About port construction, building methods, materials, and techniques, as opposed to administration, we are fast becoming much better informed. The invention of the Aqua-Lung has allowed close analysis of submerged structures for the first time. The result of this has been a rush of new evidence from a variety of sites in Turkey and North Africa, as well as in France and Italy, with interesting work at Portus and Antium, about which Robert Hohlfelder, John Oleson, and Christopher Brandon are to talk. Elaine Gazda, who made the original study of the concrete structures at Cosa-and will talk about their date-has put us all in her debt by her recent publication of a historiographical survey, setting this work in context. 44 But Caesarea in present-day Israel, the focus of the most sustained harbor exploration project ever mounted, remains unique in the detail it offers concerning harbor construction. Fortunately its building methods have been the focus of attention by one of the great experts on ancient technology, John Oleson, and now of the architect Christopher Brandon, whose particular attention to the timber formwork for concrete has shown how significant the overlap is with the practices of ship carpentry. 45 The use of concrete was the great Roman contribution to architecture in general, and, to harbor building in particular, the use of a pozzolana mortar, which created a hydraulic concrete that would set underwater. It was an achievement not to be matched again until John Smeaton in the late eighteenth century, and it allowed the Romans to impose artificial harbor basins on coasts where nature provided little or no help but where for other reasons, political, administrative, or economic, it might be desirable that a harbor should exist. It was a further step in the emancipation from geographical determinism. Caesarea shows with what confidence, flair, and sophistication

Rickman 1998; cf. Humphrey 1991, especially contributions by N. Horsfall and K. Hopkins.

43

44 Gazda

2001.

45

Oleson and Branton 1992; Brandon, Kemp, and Grove 1999.

PORTS,SHIPS,AND POWER IN THE ROMANWORLD

15

hydraulic concrete was being used by the time Herod came to build his harbor during the reign of Augustus. It also shows how an international world was already in being. The pozzolana needed was imported from Italy, from the area of Puteoli, as tests have shown, and also perhaps the expert engineers skilled in its use, in a kind of complete "technology transfer." This use of concrete, however, grew within established earlier traditions of harbor building by Phoenicians and Greeks. Piers, or splash walls, therefore, made of concrete could be based on foundations that were rubble mounds, walls made of ashlar blocks could be built on concrete foundations, and quays made of wood and filled with earth continued to appear in river harbors and other contexts in the Imperial period. As John Oleson has pointed out, ancient harbor technology never followed a simple linear development; too much depended on topography, the materials available, and the economic conditions prevailing locally.46 What we find, as he says, is the development of a repertoire of techniques, which increased the flexibility of harbor design and the prospect of success. Other possibilities than basins jutting out into the sea, embraced by projecting concrete moles, continued to be exploited: canal ports, such as Metapontum; or lagoon ports, as at Tarentum in southern Italy; and of course natural harbors, as at Syracuse or at Zancle in Sicily.47 But Caesarea, and Herod's claims for it, was about more than technology. The descriptions of it given by Josephus, the Jewish historian writing in the first century A.O., show how the harbor gleamed with elegant buildings, towers, and columns and a great temple, which overlooked it all.48 They are reminders of the enduring importance of religion in sea-going life and of the striking monumentality and image-making possible in harbor areas. But they also remind us that artificial harbor structures on this scale depended on political decision, and political will, at the highest level. The temple was dedicated to Augustus and Rome, and Herod's harbor was named Sebastos (the Greek version of Augustus) in honor of the emperor himself. We are not just in an international world here but the world of power politics. So where do we go now with the study of ports and their maritime world? I suppose our attack must be multilayered. There are new techniques to be used: obvious in the case of underwater archaeology, with its ever increasing refinement of methodology and analysis and the potential now of deep-water survey; but the picture on land is also changing. Resistivity surveys and other geophysical methods, in conjunction with air photographs, used by Keay and Millet at Portus, and Heinzelmann and Martin at Ostia, 49 raise the possibility of detecting and understanding harbor areas better without massive and expensive excavation or the problem of conservation, about which Pamela Gambogi is to talk. We could be at the beginning of a new era here. Various quite different disciplines-study of texts, archaeological expertise, geography, geomorphology-have to be brought together, to allow us to find more sites and not to be misled by old ones. In our quest small details must be studied but big questions asked-particularly how sites, of all kinds, meshed with their hinterlands and, even more important, meshed with each other-by means of the sea. The Mediterranean was, after all, the Romans' mare internum, that sea which provided them with the short internal routes of empire. Interestingly, the name mare mediterraneumwas not used until very late. The first extant reference, I think, is in Isidore of Seville in the sixth/seventh century A.O., and the name was not apparently in common usage until the twelfth and thirteenth centuries 46

0leson 1988, 147-157.

47 Schmiedt

1970.

48 Joseph.

A] 336-339.

49 Keay et

al. 2006; Heinzelmann and Martin 2002.

16

GEOFFREYRICKMAN

That may reflect a change in perspective, with the surrounding land-for that was the emphasis originally of the adjective medite"aneus-more important than the sea.' 0 Whether that is true or not, for the Romans it was always mare magnum, mare internum, mare nostrum-the large, internal sea base of Rome's fully possessed power. In this Roman sea world-to parody George Orwell-all sites were equal-simply "nodes of density," if you like (cabotageand the conditions of navigation saw to that)-but some sites were more equal than others. If I were, invidiously, to pick out one site that I think was the key stimulus to Rome in this sphere, it would have to be Alexandria. Of course, I know that Carthage had been the great maritime power in the western Mediterranean, the great enemy of Rome, and the original great stimulus to Rome's naval efforts. But Carthage had been defeated and razed to the ground in 146 B.C. When it revived, as it certainly did and became a great city and great port again-an important center of early Christianity, of Augustine and others-it was within a Roman context, the world of Rome, shaped by it and inspired by it. The final struggle of the Republic, however, between Mark Antony and Octavian, may have left Octavian and the West the winners but Alexandria, Antony's base, as the most sophisticated city, with the most complex and elegant waterfront-not forgetting its wondrous lighthouse-in the Mediterranean world. The Rome of Octavian/ Augustus, and all the early emperors, was left hurrying to catch up-to create something equally impressive, with a lighthouse of its own on a similarly low-lying coast, able to accommodate the mass of shipping, of which Rome was now to become the focus, as the unique center of wealth and political power. Rome as an imperial city had to have a dignified and elegant maritime approach to match the monuments and great tombs that lined the approaches by road. And the great exemplum was surely Alexandria. Alexandria set a standard-and the greater understanding of Alexandria as a port, through the work of Jean-Yvres Empereur and others, has helped to restore a key piece of the jigsaw to its rightfulplace in our perception of Rome and its ports.' 1 But it was Rome's greater rhythms as an imperial power-greater by far than Alexandria of the Ptolemies-the pulsing movement of masses of grain from the southern Mediterranean northwards for the city itself and the armies, the organization of the payment of its dues and the collection of its taxes, that created the whole structure of harbors, quayside arrangements, credit facilities, and ship sailings that, once they existed, could be used for a wide range of other trade as well. It was the state's imperatives, the imperatives of power, that ultimately mattered in the maritime world of Rome. Within this grand network of course many other lesser imperatives could also shelter and find support at the height of the Empire: local initiatives in harbor building or repair, short-range and middle-distance circulations of goods, the comings and goings of people and objects of all kinds widely dispersed. But it was the overarching power of the state that guaranteed and underwrote the whole.' 2 As the state structure weakened and lost power, particularly in the West, and an enfeebled Rome, greatly reduced in population by the fifth/sixth century A.O., neither needed nor could afford such shipments of grain, oil, or wine, so general trade, which had ridden on the back of such staples from Africa and elsewhere, fell away entirely or fractured into more and more localized circulations. 53 That affected everything-willingness and ability to maintain harbors, ship sizes, routes, and safety. A.O.

,oShaw2001, n. 13;Isid. Etym. 13.16.1.

' 2 Hopkins

' 1 Purcell 1996,275-276.

' 3 Wickham

1988. 1988.

PORTS,SHIPS,AND POWERIN THE ROMANWORLD

17

It is all this that we have to try to analyze and to understand-this wonderful, interlocking complexity and interdependence. It may be difficult, but, by heavens, it is exciting. No single power, before or since, has controlled all the lands around the Mediterranean Sea or has "possessed" that sea so completely as the Romans did. It will take all our combined efforts to unravel and explain that "possession" properly.

18

GEOFFREYRICKMAN

Works Cited Barker, G., and D. Mattingly, eds., The Archaeologyof MediterraneanLandscapes,5 vols. (Oxford 1999). Bartoccini, R., Il porto romano di Leptis Magna (Rome 1960). Berchem, D. van, "Le port de Seleucie de Pierie et !'infrastructure logistique des guerres parthiques," Bonner Jahrbiicher185 (1985) 47-87. Bezeczky, T., "Amphorae and Amphora Stamps from the Laecanius Workshop," Journalof Roman Archaeology 8 (1995) 41-65. Blackman, D. J., "Ancient Harbours in the Mediterranean," InternationalJournalof NauticalArchaeology11.2 (1982) 79-104, 11.3: 185-211. Brandon, C., S. Kemp, and M. Grove, "Pozzolana, Lime, and Single-mission Barges (Area K)," in Caesarea Papers2, ed. K. G. Hoium, A. Raban, andJ. Patrich (Portsmouth, RI 1999) 169-178.Journal of Roman Archaeology suppl. 35. Camodeca, G., "Puteoli porto annonario e il commercio del grano in eta imperiale," in Le ravitaillementen ble de Rome (Naples and Rome 1994) 103-128. Casson, L., The PeriplusMaris Erythraei (Princeton 1989). Chevallier, R., "Navigation et ports antiques," in Atti del Convegno internazionaledi studi sull'antichita di Classe(Rome 1967) 219-247. Coarelli, F.,Il CampoMarzio (Rome 1997). Columba, G. M., I porti antichi delta Sicilia (Rome 1906). Cornell, T.J., The Beginningsof Rome (London 1995). Curtis, R. I., "Spanish Trade in Salted Fish Products in the 1st and 2nd Centuries A.o.,"InternationalJournal of NauticalArchaeology17 (1988) 205-210. ---, Garum and Salsamenta,Productionand Commercein MateriaMedica (Leiden 1991). D'Arms,J. H., Commerceand SocialStanding in Ancient Rome (Cambridge, MA 1981). Davidson, D. A., and M. L. Shackley, eds., Geoarchaeology: Earth Sciencesand the Past (London 1976). De Fazio, G., Nuove osservazionisopraipregi architettonicidei porti degli antichi (Naples 1832). Degrassi, A., "I porti romani dell' lstria," in Anthemon, Scritti in onore di C.Anti(Florence 1955) 119-169. Dubois, C., Pouzzoles antique: histoire et topographies (Rome 1907). Bibliotheque des Ecoles franc;aises d' Athenes et de Rome 98. Duncan-Jones, R., Structureand Scalein the Roman Economy (Cambridge 1990). Empereur,J-Y., Alexandria Rediscovered(London 1998). Fea, C., Restabilmento delta citta di Anzio, e suo porto Neroniano (Rome 1835). France, J., and A. Hesnard, "Une statio inedite du Quarantieme des Gaules et les operations commerciales dans le port romain de Marseilles (place Jules-Verne)," Journalof Roman Archaeology8 (1995) 79-93. Frederiksen, M., Campania(London 1984). Gazda, E. K., "Cosa's Contribution to the Study of Roman Hydraulic Concrete: An Historiographic Commentary," in New Light from Ancient Cosa,ed. N. W. Goldman (New York 2001) 145-177. Gilissen,J., "Un typologie des escales: histoire des grandes escales vue sous l'angle institutionnel," Les Grandes Escales,Recueils de la SocieteJean Bodin 34 (1974) 681-732. Harris, W. V., ed., Re-Thinking the Mediterranean(Oxford 2005). Heinzelmann, M., and A. Martin, "River Port, Navalia and Harbour Temple at Ostia: New Results of a DAI-AAR Project," Journalof Roman Archaeology15 (2002) 5-19. Hohlfelder, R. L., and R. L. Vann, "Cabotage at Aperlae in Ancient Lycia," InternationalJournal of Nautical Archaeology29 (2000) 126-135. Hopkins, K., "Taxes and Trade in the Roman Empire (200 B.C.-A.D. 400)," Journalof Roman Studies70 (1980) 101-125. ---, "Models, Ships, and Staples," in Tradeand Famine in ClassicalAntiquity, ed. P. Garnsey and C. R. Whittaker (Cambridge 1983a) 84-105.

PORTS,SHIPS,AND POWER IN THE ROMANWORLD

19

"Introduction," in Tradein the Ancient Economy,ed. P. Garnsey, K. Hopkins, and C.R. Whittaker (London 1983b) ix-xxv. ---, "Roman Trade, Industry and Labor," in Civilizationof the Mediterranean,vol. 2, ed. M. Grant and R. Kitzinger (New York 1988) 753-777. Horden, P., and N. Purcell, The CorruptingSea:A Study of MediterraneanHistory (London 2000). Houston, G. W., "The Administration of Italian Seaports during the First Three Centuries of the Roman Empire," Memoirs of the AmericanAcademy in Rome 36 (1980) 157-171. ---, "Ports in Perspective: Some Comparative Materials on Roman Merchant Ships and Ports," American Journalof Archaeology92 (1988) 553-564. Humphrey, J. H., ed., Literacyin the Roman World (Ann Arbor 1991). Journal of Roman Archaeology suppl.

---,

3. Hurst, H., unpublished lecture (Glasgow 2002). Jondet, G., "Les ports submerges de l'ancienne ile de Pharos," Memoiresde l'Institut Egyptien 9 (1916). Keay, S., M. Millett, L. Paroli, and K. Strutt, Portus Romae:An ArchaeologicalSurvey of the Port of Imperial Rome (London 2006). Kreutz, B. M., "The Ecology of Maritime Success: The Puzzling Case of Amalfi,"MediterraneanHistorical Review 3 (1988) 103-113. Lance!, S., Carthage(London 1995). Lehmann-Hartleben, K., Die antiken Ha/enanlagendes Mittelmeeres.Beitriigezur Geschichtedes Stiidtebaues im Altertum (Leipzig 1923). Klio 14. Malkin, I., ed., MediterraneanParadigmsand ClassicalAntiquity (London 2003). Mediterranean Historical Review 18.2, devoted to discussion of Horden and Purcell 2000. McCann, A. M., The Roman Port and Fisheryof Cosa(Princeton 1987). Millar, F.G. B., "Caravan Cities: The Roman Near East and Long-distance Trade by Land," in Modus Operandi:· Essaysin Honouro/Geof/rey Rickman, ed. M. Austin,}. Harries, and C. Smith (London 1998) 119-137. Muckelroy, K., ed., Archaeologyunder Water:An Atlas of the World'sSubmergedSites (New York and London 1980). Oleson, J. P., "The Technology of Roman Harbours," InternationalJournalof NauticalArchaeology 17 ( 1980) 147-157. Oleson,}. P., and G. Branton, "The Technology of King Herod's Harbour," in CaesareaPapers,ed. R. L. Vann (Ann Arbor 1992) 49-67.Joumal of Roman Archaeology suppl. 5. Parker, A. J., "Shipwrecks and Ancient Trade in the Mediterranean," ArchaeologicalReview from Cambridge 3 (1984) 99-113. ---, Ancient Shipwrecksof the Mediterraneanand Roman Provinces(Oxford 1992a). BAR International Series 580. ---, "Cargoes, Containers and Storage in the Ancient Mediterranean," InternationalJournalof Nautical Archaeology21 (1992b) 89-100. ---, "Sea Transport and Trade in the Ancient Mediterranean," in The Sea and History, ed. E. E. Rice (Stroud 1996) 97-110. Pefia, J. T., "The Mobilisation of State Olive Oil in Roman Africa: The Evidence of Late 4th Century Ostraka from Carthage," inJ. T. Pefia et al., CarthagePapers(Portsmouth, RI 1998) 117-238.Joumal of Roman Archaeology suppl. 28. Poidebard, A., Un grandport disparu:Tyr.Recherchesaerienneset sous-marines,1934-35 (Paris 1939). Bibliotheque archeologique et historique 29. Poidebard, A., andJ. Lauffray, Sidon:amenagementsantiquesdu port de Saida (Beirut 1951). Pomey, P., and A. Tchernia, "Le tonnage maximum des navires de commerce romains," Archaeonautica2 (1978) 233-251. Purcell, N., "The Ports of Rome: Evolution of a FafadeMaritime," in "Roman Ostia"Revisited, ed. A. Gallina Zevi and A. Claridge (London 1996) 267-279.

20

GEOFFREYRICKMAN

Rickman, G. E., "The Grain Trade under the Roman Empire," Memoirsof the AmericanAcademyin Rome 36 (1980a) 261-275. --, The CornSupplyof Ancient Rome (Oxford 19806). --, "Towards a Study of Roman Ports," in HarbourArchaeology,ed. A. Raban (Oxford 1985) 105-114. BAR International Series 257. --, "Problems of Transport and Development of Ports," in Nourrir la plebe, ed. A. Giovannini (Basel 1991) 103-118. Schweizerische Beitriige zur Altertumswissenschaft 22. --, "Portus in Perspective," in "RomanOstia"Revisited, ed. A. Gallina Zevi and A. Claridge (London 1996) 281-291. --, "Problems of Transport and Storage of Goods for Distribution: 'les traces oubliees'," in La Memoire Perdue,ed. C. Moatti (Rome 1998) 317-324. Rouge, J., Recherchessur /'organisationdu commercemaritime en Mediterraneesous /'Empire romain (Paris 1966). --, "Routes et ports de la Mediterranee antique," Rivista di studi liguri53 (1987) 151-170. Schmiedt, G., Atlante AerofotograficodelleSedi Umanein Italia (Florence 1970). Shaw, B. D., "Challenging Braudel: A New Vision of the Mediterranean" (review of Horden and Purcell's The CorruptingSea),Journalof Roman Archaeology14 (2001) 419-453. Sidebotham, S., Roman EconomicPolicyin the ErythraThalassa30 B.C.-A.D. 217 (Leiden 1986). Starr, C. G., The Influenceof Sea Poweron Ancient History (Oxford 1989). Steffy,J. R, WoodenShip Buildingand the Interpretationof Shipwrecks(College Station, TIC 1994). Temin, P., "A Market Economy in the Early Roman Empire," Journalof Roman Studies91 (2001) 169-181. Throckmorton, P., and A.J. Parker, "The Amphora: Jerrycan of Antiquity," in Historyfrom the Sea:Shipwrecks and Archaeology,ed. P. Throckmorton (London 1987) 64-71. Wickham, C., "Marx, Sherlock Holmes, and Late Roman Commerce," Journalof Roman Studies 78 (1988) 183-193. Will, E. L., "Defining the Regna Vini of the Sestii," in New Lightfrom Ancient Cosa,ed. N. W. Goldman (New York 2001) 35-47. Wilson, A., "Urban Production in the Roman World: The View from North Africa," Papersof the British Schoolat Rome 70 (2002) 231-273. Young, G. K., Rome's EasternTrade(London 2001).

ROMAN SHIPSHEDS David Blackman

T

he covered slipways or "shipsheds" housing warships were the most distinctive feature of the military harbors of the Mediterranean in the Classical and Hellenistic periods, and many remains of them have now been identified and investigated. Warships were kept out of the water when not in use, to avoid rot and the shipworm, and protected from sun and rain by the roofs of the shipsheds that lined the harbor basins. The monumental "arsenali" of the great maritime cities of the medieval period, with their vaulted sheds, continue the same tradition. There is still, however, a "missing link": we have no certain remains of the shipsheds, or navalia,of the Roman period. There is in my view no doubt that shipsheds were used in the Roman period: the pictorial evidence seems to me clear. Coins depicting entire harbors show colonnades that admittedly may simply be quayside colonnades or arched moles, rather than roofed slipways: for example, the coin of Gallienus showing the harbor of Side as a perfect circle, when in fact the main harbor was triangular (more or less); the arcade probably represents porticos rather than slipway entrances. A coin of Antoninus Pius shows the harbor of Cenchreae with a statue and three ships and a quayside colonnade. But Lehmann-Hartleben thought that the Neronian sestertii of Portus portrayed shipsheds rather than arched moles, and we must not forget Appian's description of the shipsheds at Carthage in 146 B.C.: "In front of every shipshed stood two Ionic columns, so that both harbor and island appeared to be lined with a colonnade" (Pun. 96). 1 Furthermore, there is no doubt about the coins showing a detail on a larger scale-clearly of roofed slipways: notably the denarius of Palikanus, a mint official in 47 B.C. (fig. 1), which shows three ships' prows peeping out of arched buildings, clearly warships in shipsheds, with a "bisellium" above whose interpretation I leave to others. Coarelli convincingly refuted the previous theory that the coin depicted the rostra of the time of Caesar. We clearly have here a depiction of the navalia of the city of Rome, destroyed in 44 B.C. and apparently not rebuilt. We may compare the asses of Ancus Marcius Censorinus, clearly of a warship, which in Bartoccini's view depicted the navalia of Ostia. 2 There are a number of similar scenes in a larger context in mosaics and wall paintings. A striking example is the arcaded border of a mosaic from a Roman villa in Gaul at Grange-du-Bief. Guey dates the mosaic to the early second century A.D., but Redde thinks that it may be earlier. A late Republican mosaic now in the Vatican, but found in the Via Ardeatina, gives a clear idea of the roofing-a double-pitched roof over each shed. A mosaic from Lanuvium, now in the church 1 Lehmann-Hartleben 1923, 238 and coin plate, nos. 9-11; Blackman 1982, 82-85, 206 and fig.I; Redde 1986, 202 and fig. 70. For a recent discussion of the iconography in general with bibliography, presented since this paper was given, see Baika 2003 b, 423-427.

2

Bartoccini 1913; Coarelli 1968, 27-33 and figs. 1, 6; Le Gall 1953, 103-ll0;Redde 1986, 162 and fig. 69. Cf. also possibly a coin issued by L. Rubirus Dossenus in 87 B.c.: Morrison 1996, 225, no. 24 III.

24

DAVID BLACKMAN

Fig. I. The denarius of Palikanus(47 B.c.).

Fig.2. Mosaicfrom the Casadel Marinaio, Pompeii.

of Santa Maria Maggiore, is poorly preserved, but we do have the roofs, and above and behind is visible a wall with towers and battlements which , as Redde rightly argued, are good evidence that military naval installations (at the very least) were included within the fortifications/ A mosaic from the Casa del Marinaio at Pompeii (Reg. VII 15,2) shows three ships' prows framed by columns in an arcade (fig. 2): Pernice dated the mosaic to the borderline between the Second and Third Styles (early Augustan period). He notes that Fiorelli spoke of "sei prore di navi," and that Fiorelli reported "nove prore di navi" from the Casa del Citarista (Reg. I 4,5).4 Maritime scenes on wall paintings mostly show ships by villae maritimae or in naumachiae,but there are exceptions, one of which from Pompeii (Second Style) shows ships in shipsheds. Basch has rightly interpreted the timbers apparently projecting from the oarbox of one of the ships as shores, not oars, since the ship was slipped-unless we are dealing with a mistake by the artist? 5 Trajan's Column (Scene LXXIX) appears to show vaulted shipsheds in the background (with fortifications as in the Lanuvium mosaic) as Trajan departs from an Adriatic port for the Second Dacian War (from Ancona? Ravenna? Brindisi? Misenum?-this question has long been a matter of dispute and is still unresolved). 6 Vaults were clearly a distinctive feature of Roman, as opposed to earlier, shipsheds ; and vaulted sheds continued to be the model in the medieval period. In my 1982 survey of ancient harbors I was brief: "The Romans also used shipsheds. Though no certain remains have been found, a number of representations of them survive on coins, mosaics 3 Grange-du-Bief:

Guey and Duval 1960, doubting the interpretation as navalia; the other mosaics : Coarelli 1968, figs. 3-4; Redde 1986, 162 and figs. 71-72. The scene lower right in the (heavily restored) Palestrina mosaic may represent the royal harbor at Alexandria (as Coarelli 1990, 239, 247) , but the low arcaded structure looks more like an arcaded quay than a row of shipsheds (Coarelli's suggestion); there is no indication of ships within the arches.

the arches sometimes frame very different subjects, so it must have been a popular artistic motif. 5 Basch 1979; 1987, 437-438 and figs. 955-956 : possibly a copy of a Hellenistic painting(?); Coarelli 1968, fig. 5; Blackman 1995b, 74 and n. 4; Morrison 1996, 245-246, no. 43, 312-317 . 6

Pernice 1938, 64 and pl. 27.3, 69; Redde 1986, 162 and n. 80 notes that the arcade motif was not confined to shipsheds :

4

Cichorius 1900, 11-26 and pls. LVIII-LIX; LehmannHartleben 1923, 228-230 ; Redde 1986, 218-220 and n. 231; Lepper and Frere 1988, 129-132 and pls. LVIII-LIX; Morrison 1996, 249-251, no. 47; Baika 2003b , 424-425 .

ROMAN SHIPSHEDS

25

and paintings, with ships showing just inside the arcade. At Rome they lined the south side of the Campus Martius in the Republican period." My note referred to Lehmann-Hartleben (1923), Le Gall (1953), and Coarelli (1968); I did not think that there was anything to add to their accounts. I followed Meiggs's rejection of the supposed shipsheds at Ostia-but to them we must retum. 7 I pointed out that Vitruvius deals with shipsheds (briefly, as I would now add), recommending vaulted roofs to minimize the use of timber in their construction (De arch.5.12.7). (Some feel that he may have lifted the short passage from an earlier text, but I still think that this was a contemporary reference, showing that shipsheds were a standard feature of military harbors in the Roman period.) I noted the possible remains of Agrippa's navaleat Lake Avemus, described by Jacono in 1941. (Redde has since expressed doubts.) 8 In a Haifa conference paper, published in 1988, I noted that the most distinctive feature of a military harbor was the presence of shipsheds, "which do seem to have been used only for warships," though "this does not mean, as is often assumed (explicitly or implicitly), that merchantmen were never hauled ashore: note Theophrastus HP 5 .7.2 [which shows clearly that they were]." I commented: "Oddly, those [shipsheds] found seem to be almost all classical or Hellenistic in date, and form a large part of the surviving harbour structures which we can definitely establish as pre-Roman." For the Roman period, the reverse seemed to be true: "Few shipsheds have yet been found which are definitely of Roman date, but they were doubtless widely used. "9 Where, however, are the remains? In 1988 I could refer to the definite remains of fifteen shipsheds at Minturnae on the river Garigliano, "never studied and probably now destroyed," and I suggested that there may have been remains at Centumcellae, now lost. I concluded: "The pictorial and literary evidence is in any case clear." 10 In discussing shipbuilding yards, I suggested that Ventotene (ancient Pandateria)might have offered some evidence if it had been thoroughly studied before so much of the site was built over. I compared ("a dismal experience") the plans and views inJacono's paper of 1933 with Schmiedt's in his 1972 book on Il live/loanticodel mar Tirreno.I could equally have pointed to J acono's suggestion of shipsheds: here is a matter that deserves on-site checking.11 In a chapter published in 1995 in The Age of the GalleyI could not add much. I referred to reports of possible shipsheds on the outer side of the southeast mole at Leptis Magna. I suggested that for the Roman period, "We must certainly assume a different pattern of distribution [of shipsheds]: no longer did almost every harbour city need to accommodate warships and therefore need shipsheds-they must have been concentrated in the naval bases of the Roman fleet." (If we add the possibility of Roman use of fortified dockyards surviving at maritime cities from the Classical/Hellenistic periods, then this remains a reasonable working hypothesis.) Blackman 1982, 206 and n. 112, referring to LehmannHartleben 1923, 183, n. 4; Le Gall 1953, 103-110; Coarelli 1968; add now Redde 1986, 202-203. Ostia: Meiggs 1960, 126; general discussion on Ostia and Portus in LehmannHartleben 1923, 183-184, 187.

7

8 Jacono 1941, 665 and fig. 3; Redde 1986, 150, 164-171, esp. 167-168; Redde in Pagano, Redde, and Roddaz 1982, 279-294. 9

Blackman 1988, 14-15 and nn. 30-31.

10

Blackman 1988, 19, n. 32, referring to Ruegg 1983, 218;

Bastianelli 1954, 40. Redde 1986, 197-201 does not believe that remains have been found at Centumcellae of the navalia mentioned by Rutilius Namatianus, De reditu l.245. 11 Blackman 1988, 15 and n. 36;Jacono 1933, pls. XLVI. fig.2, XLVII. fig.3 (reproduced by Schmiedt 1964,fig.40); Schmiedt 1972, 176--181, esp. 176, fig. 185. I have visited Ventotene for the first time since the workshop, in May 2003, by kind arrangement with Dr A. Zarratelli. A newly published guide to Ventotene (De Rossi 1999, 18--19, fig. 5) illustrates a late eighteenth-century plan of the port, showing magazines cut in the rock along the western shore; this appears on inspection to excludeJacono's suggestion of shipsheds here.

DAVIDBLACKMAN

26

I noted the epigraphic evidence from the naval base at Moguntiacum (Mainz), on the Rhine at the confluence with the Main-three references to navaliabut no actual remains: the remains found and published by Olaf Hackmann and colleagues were of ships apparently at a breaker's yard, with associated finds including woodworking tools and countless oak wood shavings. This is still, according to Hackmann, the type of material being found at Mainz, and also at Cologne. 12 Other finds were being made along the Rhine-Danube frontier, but information took some time to cross the (regrettable) "north-south divide" in our subject. For some of us the information came in Hackmann's paper at the Venice conference in December 2000. 13 That evidence is presented and discussed by Boris Rankov in his paper. Among works on the Roman navy, our subject is discussed best in Michel Redde's magisterial Mare Nostrum(1986), which concentrates on the Imperial period but refers to earlier evidence throughout (both Greek and Roman Republican). He discusses the importance of naval castrain the Imperial period, not only at Ravenna and Frejus but also at Boulogne, Dover, and Cologne; and he concludes that they were a feature of all Imperial-period naval bases. He then stresses how little we know from archaeological evidence of the naval installations. He notes that Latin preserves a distinction (at least until Isidore's Etymologyin the early seventh century) between navaliaand textrina,on which Servius in a comment on Aeneid 11.326-329 had Redde been very firm. Greek had the same distinction between vswamxm/vsweuxand vocun11ysioc. wonders whether navaliacould also have been used for merchant ships (but leaves the question open) and notes the epigraphic evidence for the existence of navaliaunder the Republic (discussed by Coarelli 1968) and under the Empire: from Aleria (of the Misenum fleet), from Ostia, and from Mainz (of the classisGermanica).He also notes the evidence for vswamxm "still" at Cyzicus ("over 200," according to Strabo 12.8.11) and the Piraeus (in the second century A.D.: Pausanias 1.1). He regards it as a plausible hypothesis that there was little architectural difference between Greek vswetocand Roman navalia,"mais il faudrait (la) confirmer par des fouilles." Redde accepts that there were navaliaon the left bank of the Tiber but that we know nothing of the detail; the best evidence so far is iconographic, which he describes and I have already presented, and almost all of late Republican or Augustan date. He thinksthat we can assume (subject to futurefinds from excavations) that the shipsheds remained similar from Greek to Roman Republican and Imperial times "in view of the similarity of methods of ship construction and the problems of protecting and conserving ships, particularly in winter." 14 This seems a sensible working hypothesis for us to consider, while remembering always the differences in possible locations: the seashore (with tides?), or a river bank (with currents and floods), or a lakeshore (?), and also differences in climate and available building materials in different parts of the empire. First, a word about the Greek prototypes, which the Romans knew from their early days of fleet construction (at least, they would have known those in the West and, let us say, Piraeus and Rhodes and Alexandria): 15 12 Blackman 1995a,232-233; Leptis: Laronde 1988,347-348 and fig. 9; Mainz: Hockmann 1993, 125-126 and personal communication.

historical evolution of Roman naval bases under the Empire. For Piraeus and Cyzicus, see bdow nn. 16, 22. 15

13 Heckmann

2003, 109.

14 Redde 1986, 27-37 on shipsheds and ship dimensions;

158-163 on military harbor installations; 309-319 on the

The literature can be found in Blackman 1968; 1982; 1990 (use of timber); 1995a; 2003; Baika 2003b. Early basic discussion in Lehmann-Hartleben 1923, 113-121; see also Redde 1986, 30--32, 161.

27

ROMAN SHIPSHEDS

Fig.3. Zea harbor,Piraeus:plan of shipshedsexcavation, 1885.

Fig.4. Model of shipsheds(based on the Zea remains), Nautical Museum, Piraeus.

The standard trireme shipshed that occupied the shoreline of the military harbors of Piraeus, Rhodes , or Carthage had (on the basis of evidence available to earlier researchers) a dry length of ea. 40 m and clear width of 5 .5-6 m, with a central ramp 3 m wide; recent work has tended to show a longer dry length and a narrower clear width. The basic Piraeus evidence goes back to an 1885 excavation, well published for its time (figs. 3-4); the site is now being restudied by a Danish researcher , Bj0rn Loven, directing a survey and excavation that is finding the lower ends of slips under water, 50.5 m from the back wall! Current excavations by the Greek archaeological service in the two small ports are not yet published , but it is reported that rather narrowshipsheds(5.20-5.40 m) have been found . Remains elsewhere show a similar gradient at many (ea. 1 in 10, but the slipways were timberclad, so that the original upper surface, now lost, is more difficult to establish); some were shorter, steeper , and narrower, mainly rock-cut shipsheds for guardships .16 16 Piraeus in the Roman period : Redde 1986, 161, 227-230, agreeing with Kienast 1966, 95-96 that it served as a naval relay station ; contraStarr 1960, 19-20 . For the Greek period , see Dragatses and Dorpfeld 1885; and references in Black-

man 1995a; 2003. Loven's work : TropisVIII forthcoming. Rock-cut shipsheds : Baika 2003a ; 2003b . Both Baika and Loven are now members of the team for the shipsheds project mentioned below.

28

DAVID BLACKMAN

Fig.5. Hellenisticshipshed, Rhodes.A rampof the last phaseoverliespiersof the precedingphase.

Henry Hurst's excavations at Carthage confirmed the model-similar widths, lengths, and gradients-while providing evidence for two wider slipways (7.1 to 8 m), clearly reflecting the larger ship types of the Hellenistic period. Wider slipways were reported from Piraeus also by a nineteenth-century investigator, Graser. My work on Rhodes showed that the Rhodian dockyard had to accommodate triremes and tetrereis(in shipsheds with a full 6 m width) and smaller warships, probably ,Qtl'}µto>..fcxt (in shipsheds with a 4.2~.40 m clear width), in the third-second century B.C. (figs. 5-6). Recent excavations in the port of Massalia , which must have been well known to the Romans well before the final siege of 49 B.C., have revealed remains of Hellenistic and earlier shipsheds: lines of timber rollers (phalangae)or fixed cross-timbers (traverses)bearing traces of use; and on one site side-walls over 40 m long and 5 .50-6 m apart . The site has also produced good evidence of shipbuilding activity-not in the shipsheds. This is a point to emphasize: shipsheds were normally not shipyards. 17 At Kition in Cyprus six shipsheds have been found, with several phases of the later fifth/early fourth century B.c.; the ramps swing up at the upper end (for which there are parallels at Oeniadae and perhaps Carthage and Rhodes, where we have only found the top end, with a 1:4.6 gradient), and there is evidence for roofing and possibly for hauling methods. Finally, at Naxos in Sicily M. C. Lentini and I are excavating shipsheds of the fifth century B.C. (with possibly an earlier phase), in a small dockyard (4-6 shipsheds) put out of use in 403 B.C. (Naxos was succeeded by Tauromenium, across the bay). We have a standard gradient of 1:11, narrow clear width of 5.45 m, and hints of stepped roofing like Kition. Postholes in side doorways may have held rubbing posts to support ships during launching and hauling up . My conclusions from recent research are: gradient is a crucial criterion-normally about 1:9 to 1:11 (much steeper than the modern slipway gradients, which are normally ea. 1:20), but there are variations. Even where stone was available for structures, timber played a key role. 17 Built shipsheds used for shipbuilding, besides Massalia: possibly Thurii and Carthage. According to Raban (2003, 95) Dor has now provided good evidence of rock-cut slipways used for shipbuilding; possibly also Apollonia. Discussion

and references in Blackman 1990, 45-48; 2003, 86; Baika 2003a; 2003b, 428-438. Roman shipyard at Mainz (not a shipshed): Hockmann 1993, 125-126; 2003. Recent work at Naxos: see now Blackman and Lentini 2003.

LEGENDE

0 RHODOS SCHFFSHAUSER

6

m

ITi.JESTR!CH §

N0ROL . DES ROMISCHEN TETRAPYLON

REKONSTRUKTIONSVERSUCH

SCHNITT

lml! M0RTa-BRUCHSTEIN

BEi S2 . • Z "

0 0,S 1 \5 2

PERIODE 1

AUSGLEICHSESTRICH

SCHNI TT Z'

-

F0LLUNG

Sm

RHO DOS SCHIFFSHAUSER NORDL. DES R0MISCHEN TETRAPYLON REKONSTRUKTIONSVERSUCH LANGSSCHNITTE

Fig. 6. Reconstructionof Hellenistic shipshed, Rhodes,final phase:cross-sectionand longitudinal section (Paul Knoblauch).

30

DAVIDBLACKMAN

The minimum clear width for triremes turns out to be a little less than we expected: 5.50 m (?), but perhaps 6 m if used also for tetrereis(?); with a narrower shipshed category (4.20-4.40 m) and perhaps a wider category (7-8 m). The dry length is 40+ m for triremes, the wet length up to 50 m; for the other ship types, the answer is not yet clear. Narrower ships only needed narrower and shorter slipways, but harbor planning may have imposed longer slipways than were actually needed for the ships (e.g., at Rhodes). Looking back from the Roman vantage point, there are obvious considerations to be borne in mind: how many Greek dockyards were available for use by Roman ships, and how many were actually used? Piraeus we have mentioned: probably a "relay station" for Roman fleets, as Redde argues (above n. 16). Pausanias in the second century A.D. reflects (1.1) what must have been the result of repairs to the shipsheds, while Strabo (9.1.15) reflects an earlier period of decline, after the Sullan destruction of 86 B.C. He burnt the shipsheds and the naval arsenal (App. Mith. 41). There may prove to be archaeological indications of a Roman phase of shipsheds. Corinth was important in the Roman period, but apparently not for the Roman navy.18 Cenchreae, its eastern harbor, was a flourishing commercial port, but no traces of shipsheds have been located, from either the Roman or earlier periods. Shipsheds at Lechaion, its western harbor, are attested in historical sources for the Classical period (Xen. Hell. 4.4.12 for 392 B.c.), but no remains have been found from that or later periods. Rhodes remained a major port and, as ally of Rome, still had warships in the early Imperial period; the remains of the military harbor studied by us have phases rather of the fourth/third and second centuries B.c.:I wonder whether the later phase could still have served the Rhodian triemioliai of the classisRhodiain the Augustan to Flavian period. The recently excavated harborworks of the Roman period belong to the commercial harbor. 19 At Alimnia, a small island port off the west coast of Rhodes, we have a collection of twentyone rock-cut shipsheds on two bays, both with fine natural protection (figs. 7-8). The shipsheds in Emporia are mostly rather wide: 8.50-9 m or 9.50-10+ m, and one is very wide (18.20 m), clearly for a double or triple shed. The dry lengths are only 14-18 m; they continue into the water but break off after about 5 m. The gradient appears to be ea. 1:7. This gradient seems very steep for large ships such as triremes and larger, but Oeniadae has 1:6 (fig.9). Were these double shipsheds for two ships of ea. 4 m beam and 20 m length? Coates suggests that they were for pentecontorsor hemioliai.These would be appropriate for the purpose that I suggest. Dating rock-cut slipways is very difficult. I have suggested that these slipways were for an outstation of the Rhodian fleet in the Hellenistic period, providing an early warning system against pirates. But evidence from the seabed included pottery of many later periods (not yet fully studied)-this port was on the main trade route from, for instance, Egypt to Constantinople; we have not yet excavated any of the slipways (which should give us an idea of date of use, if not so clearly of date of construction). 20 18 Redde 1986,230; for Lechaion, see most recentlyRothaus

1995. Blackman in Blackman, Knoblauch, and Yiannikouri 1996,404;Herod the Great had ships built for his fleet in the shipyards at Rhodes and financed their restoration Ooseph. AJ16.147);Rhodes was a "naval ally" of Rome at least well 19

into the first centuryA.D., but its fleethad gone by A.D. 79--82: Redde 1986,314, 498-499. 20 Blackman1996;1999;Blackmanand Simosi2003.Another

likely out-station of the Rhodian fleet has now been identified at Loryma in the Rhodian Peraea, with evidence for shipsheds: Held 2002, 291-293 and figs.7-9 (p. 300).

Fig. 7. EmporioBay,Alimnia: viewfrom west, with Rhodes in the background.

I

CJ

vnov•r • ~IA

•

IO

_,..AA,WN

Y~NIA

• TO O•MO T1'4& N.AA-IA

•

•

nOAITl

~

•..,.._...•

•

NAJl',A-

-...·...._

~

...... Fig.8. EmporioBay,Alimnia:plan (Aik. Tagonidou , Ephorateof MaritimeAntiquities).

MOY

A•NAIOTHTeN

32

DAVID BLACKMAN

Fig. 9. Oeniadae:view of rock-cutshipsheds.

The difficulty of dating rock-cut slipways is illustrated also at Rethymno in Crete: groups of rock-cut slipways, probably Classical/Hellenistic in origin, were certainly used from the medieval period down to the last century. 21 One big harbor of which we know little is that of Cyzicus in the Propontis. Strabo (12.8.11) says that it had "over 200" shipsheds. We have no archaeological evidence for this, and I have expressed surprise at this figure, following Lehmann-Hartleben's comment "Riesenzahl." 22 But I tend to see things from the Greek point of view: Cyzicus may have provided a very useful fleet base for the Romans: it was the first port on the southern coast of the Propontis for ships that had sailed up the Hellespont . One still, however, has to explain why it had "over 200" shipsheds as early as Strabo's time-or his source . The main period of Roman use seems to be when it became the base of the classzsPonticain the 170s. It last appears in A.D. 365. Another major port where the literary testimonia are clear is Alexandria, base of the classisAlexandrina.2} In the recent intensive investigations the shipsheds have not been found, as far as I am aware. At Caesarea in Palestine Avner Raban has found remains in Area SILL, the northwest corner of the Inner Basin: elongated parallel compartments, which he believes originally functioned as shipsheds for naval units, probably Herod's royal fleet or perhaps a squadron of the Roman classis Syriaca.He presented these finds, along with two other possible shipshed sites, at the Venice conference in 2000. 24 21 Rethymno:

Baika 2003b, 521-524 and references there.

22 Lehmann

-Hartleben 1923, 101, n. l; 63-64 on the Greek port; Blackman 2003, 82; Redde 1986, 147, 161, 254-255, 388,642 . 2}

Redde 1986, 241-243, 493-494; McKenzie 2003, 36-41; Baika 2003b , 248-252. See also n. 3 above.

24

Now published : Raban 2003. The site at Area SILL remains the most plausible. Some doubts have been expressed since l)is then, but Raban has confirmed to me th111he mi1inrnins interpretation. The other candidates are: vaults in the western

fa\ade of the Temple Platform (Area I) on the southeast side of the Inner Basin, which deserve perhaps to be compared with the Ostia site discussed by Rankov below, and vaults in Area CC to the south , in the northwest part of the South Bay. All have long, narrow compartments, apparently open to the sea in an original phase, before being converted into ho"ea. (The similar appearance of shipsheds and ho"ea isalwayssomething to beware: since giving this paper I have noted rhe remark of Kahler and Guidi 1958, 686, that some ho"ea in the Portus Traiani may be navalia.) Only the compartments in Area SILL have remains surviving of a (slight) seaward slope, making them 1~ ligl~r i;illldWil~ ,Hmmii§ii "ni1vil1 illly"ofRo,rw. =~ckM 1986, 314,498. See now Rankov, in this volume.

ROMAN SH1PSHEDS

33

One site in the West has shipsheds that may have a Roman phase: rock-cut slipwayson the north side of the Little Harbor of Syracuse.First published by Cavallariin 1881, they have recently been studied by BeatriceBasile:she definestwo phases of shipsheds,which seem clear.Then she proposes an early fourth-century B.C. date for the first phase, which forces her to propose tentativelya late Roman date for the second phase. It seems to me much more likelythat phase 1 is of the fifth century B.C. (there must have been shipsheds here then) and that phase 2 represents post-404 reconstruction by DionysiosI. If that is correct, we do not have shipsheds of the Roman period at Syracuse.25 Before passing to the short list of new candidates (or renewed candidates), I should refer to the exceptional case of the Aktion shipsheds (which are distinct from the Aktion victory monument on the hill): Strabo in his description of Aktion (7.7.6) describes a shipshed memorial below the hill on which stood the temple of Apollo Aktios, which had burned down before he wrote his account (that is, it had a very short life): "vsweuxin which Caesar dedicated as first fruits of his victory a set of ten ships (6sxcxvcx(cx), from a 'one' (µovoxeowc;) to a 'ten' (6sxiJel'Jc;);it is said, however, that both the vswomxmand the nwicxhave been destroyed by fire." There is a clear implication here (but no proof) that the shipsheds were of varying widths. I think "dekanaia" means "set of ten" rather than "squadron of ten." One wonders whether any of the substructure might still be traceable.26 It is impossible in a short essayto go into every detail. The subject occupied many pages and enormous bibliography in the great work of Michel Redde. In the project that we are about to start on "Shipsheds of the Ancient Mediterranean," Rankov and I will have the opportunity and the obligation to follow up many points that we can touch on only in passing today: we will have to look again at many of the tantalizing references to navalia,and to horrea(?) close to the shore, and at the evidence of air photographs collected so well by Giulio Schmiedt (1964; 1978). Let me mention as an example the extraordinary gaps in our knowledge of this aspect-the covered slipways-of the main Roman fleet bases in Italy: (i) Ostia /Portus is under close study by colleagueshere present; somewhere lies a navalethat

was restored in the second century A.D. 27 (ii) We shall want to look again at the evidence from Portus Julius: the "navale"(?).28 (iii) Redde describes remains near the water at Misenum as remains of buildings resembling horrea.Was anythingfound before the shorelinewas covered by modem militaryfacilities? Do any remains survive (fig. 10)?29 (iv) At Ravenna, Schmiedt thought he could identify a "bassin de carenage"; Cortesi's soundings have revealed buildings that he identified as barracks, but that Redde suggests were horrea,on the supposed island in the harbor entrance canal.30 25 Basile2002, 170-172 (shipshed phases

1 and 2 are site phases

ill and V); see Lehmann-Hartleben 1923, 106-107, 112-113; Redde 1986, 213: port no longer of military importance. 26 Blackman 1996, 113 and n. 4; Lehmann-Hartleben 1923, 241 described it as a "maritime museum." I gather that Professor lsmene Trianti of Ioannina University is planning an investigation.

27 CIL 14.376. See Starr 1960, 17; Redde 1986, 202, who wonders whether it was really for the navy.

29

Redde 1986, 150, 186-197, esp. 195, n. 123: "bitiments apparemment allonges, separes par des cloisons qui determinent des espaces de 3-4 ma 6-7 m. Ce plan ressemble assez a celui d' horrea.11ssont situes, assez pres de l'eau au pied de la ville et leur appartenance reste peu claire: base militaire? Port? Ville?" I take this to be the area of shoreline on the east side of the "camp?" and between it and the "bourg" on his fig. 12 (my fig. 10). One hopes that it might be possible to have access to the military establishment, to check whether remains have been found and perhaps survive. 30 Redde

28 See

above and n. 8.

1986, 177-186, esp. 181-183 and references there

(nn. 60, 66).

34

DAVID BLACKMAN

Fig. 10. Misenum: plan (after Redde [1986), fig. 12).

(v) At Vada Sabatia (Vado Ligure) the Peutinger Table marks "navalia";31 but I wonder whether the term may have had here the wider sense of "dockyard" or "naval base." (vi) At Frejus, clearly an important base from Octavian to the Flavian period, the existence of navaliais attested. 32 Boris Rankov and I consulted each other during the preparation of our papers and agreed that four sites recently proposed needed detailed discussion now: Velsen, Haltern/Hofestatt, Caesarea, and Ostia. I have tried to provide the context and set the scene; Rankov discusses those sites and whether they are shipsheds. 33 lI

Redde 1986, 204 and references there.

32

Redde 1986, 150, 171-177, 490-492.

JJ G. Ciampoltrini has suggested that remains on the beach at Cosa, by the "villa marittima della Tagliata," may belong to navalia, but we are not convinced : Ciampoltrini et al. 1999, 219-225; cf. Gambogi in this volume. Blackman is intrigued by two early second-century "Hallenbauten" parallel to the river north of the Castell um at Oberstimm, with side aisles of

6 m clear width and a central aisle of 12 m (H. Schonberger in Berichteder Romisch-GermanischenKommission 70 [ 1989] 246-247, referring to K. H. Rieder, Das archiiologischeJahr in Bayern 1982 [1983] 101-103), and thinks of shipsheds; but Rankov is skeptical. Morel (1987, 229, n. 45) notes an oral report from T. Sarnowski at the 1983 Limeskongress at Aalen of possible Roman slipways at Novae, but Sarnowski indicated to Rankov at the 2003 Limeskongress at Pees that both this interpretation and the dating are tentative and cannot be confirmed.

ROMANSHIPSHEDS

35

Works Cited Baika, K., "Operating on Shipsheds and Slipways: Evidence of Underwater Configuration of Slipways from the Neosoikos of 'Trypiti' (Crete)," in Boats,Shipsand Shipyards.Ninth InternationalSymposiumon Boat and ShipArchaeology,Venice2000, Proceedings,ed. C. Beltrame (Oxford 2003a) 103-108. --, "NEO~OIKOI: Installations navales militaires en Mediterranee: les neoria de Corcyre" (diss. Paris I, 2003b). Bartoccini, R., "Le figurazioni degli assi di C. Marcius Censorinus," Rassegna numismatica 10 (1913) 97-105. Basch, L., "Roman Triremes and the Outriggerless Phoenician Trireme," Mariner's Mi"or 65.4 (1979) 292-294. --, Le Musee imaginairede la marineantique (Athens 1987). Basile, B., "I Neosoikoi di Siracusa," in Strumenti per la protezionedel patrimonioculturale:aspetti archeologici,Atti . .. Convegno. .. Palermoe Siracusa8-10/3/2001, ed. V. Livigni and S. Tusa (Milan 2002) 147-175. Bastianelli, S., Centumcellae.ItaliaRomana:Municipie Colonie,ser. 1, vol. 14 (Rome 1954). Bdtrame, C., ed., Boats,Ships and Shipyards.Ninth InternationalSymposiumon Boat and Ship Archaeology, Venice2000, Proceedings(Oxford 2003). Blackman, D. J., "The Shipsheds," in J. S. Morrison and R. T. Williams, Greek OaredShips, 900-322 B.C. (Cambridge 1968) 181-186. --, "Ancient Harbours in the Mediterranean," InternationalJournalof NauticalArchaeology11.2 ( 1982) 79-104, 11.3:185-211. --, "Bollards and Men," in MediterraneanCities:HistoricalPerspectives, ed. I. Malkinand R. L. Hohlfdder (London 1988) 7-20. Mediterranean Historical Review 3.1. --, "Triremes and Shipsheds," in TropisII. 2nd InternationalSymposiumon ShipConstructionin Antiquity, Delphi 1987,Proceedings,ed. H. Tzalas (Ddphi 1990) 35-52. --, "Naval Installations," in The Age of the Galley,ed. R. Gardiner (London 1995a) 224-233. --, "Some Problems of Ship Operation in Harbour," in TropisIII. 3rd InternationalSymposiumon Ship Constructionin Antiquity, Athens 1989,Proceedings,ed. H. Tzalas (Athens 1995b) 73-81. --, "New Evidence for Ancient Ship Dimensions," in TropisIV. 4th InternationalSymposiumon Ship Constructionin Antiquity, Athens 1991,Proceedings,ed. H. Tzalas (Athens 1996) 113-125. --, "Double Shipsheds?," in TropisV. 5th InternationalSymposiumon Ship Constructionin Antiquity, Nauplia1993,Proceedings,ed. H. Tzalas (Athens 1999) 65-78. --, "Progress in the Study of Ancient Shipsheds: A Review," in Boats,Ships and Shipyards.Ninth InternationalSymposiumon Boat and Ship Archaeology,Venice2000, Proceedings,ed. C. Bdtrame (Oxford 2003) 81-90. Blackman, D. J., P. Knoblauch, and A. Yiannikouri, "Die Schiffshiiuser am Mandrakihafen in Rhodos," Archiiologischer Anzeiger (1996) 371-426. Blackman, D. J.,and M. C. Lentini, "The Shipsheds of Sicilian Naxos, Researches 1998-2001: A Prdiminary Report," Annual of the BritishSchoolat Athens 98 (2003) 389-436. Blackman, D. J.,and A. Simosi, "Researches on the Island of Alimnia near Rhodes," in TropisVII. 7th InternationalSymposiumon Ship Constructionin Antiquity, Pylos 1999, Proceedings,ed. H. Tzalas (Athens 2003) 139-149. Ciampoltrini, G., R. Iezzi, and G. Agricoli, "Fra la Tagliata, l'Origlio, ii Chiarone: materiali per l'insediamento etrusco e romano sul litorale capalbiese," Rassegnadi Archeologia16 (1999) 219-258. Cichorius, C., Die Reliefs der Traianssiiule, vol. 3 (Berlin 1900). Coarelli, F., "Navalia, Tarentum e la topografia dd Campo Marzio meridionale," Studi di topogra/iaromana. Quadernidell'Istitutodi Topografia Antica dell'Universitadi Roma 5 (1968) 27-37. --, "Lapompe di Tolomeo Filadelfo e ii mosaico nilotico di Palestrina," Ktema 15 (1990) 225-251. De Rossi, G. M., Ventotenee S. Stefano (Rome 1999).

36

DAVIDBLACKMAN

Dragatses, I. C., and W. Dorpfdd, "Ekthesis peri ton en Peiraiei anaskaphon," Praktika(1885) 63-68. Guey,J., and P.-M. Duval, "Les mosaiques de la Grange-du-Bief," Gallia18 (1960) 83-102. Hdd, W., "Forschungen in Loryma 2001," in 20 Araittrma Sonu,larzToplanttst,27-31 May 2002 (Ankara 2002) 289-300. Hackmann, 0., "Late Roman Rhine Vessds from Mainz, Germany," InternationalJournalof NauticalArchaeology22 (1993) 125-135. --, "An Early Roman Boatyard at Mainz, Germany," in Boats,Ships and Shipyards.Ninth International Symposiumon Boat and Ship Archaeology,Venice2000, Proceedings,ed. C. Bdtrame (Oxford 2003) 109-112. Jacono, L., "Un porto duomillennario," in Atti de/ Ill Congressonazionaledi studi romani, 1933 (Bologna 1934) 1:318-324. "Il porto Giulio," Atti de/la Reale Accademiad'ltalia:Rendiconti de/la Classedi ScienzeMorali e --, Storiche,7th ser., 2.12 (1941) 650-676. Kahler, H., and G. Guidi," Arsenale," in Enciclopediadell'arteantica,vol. 1 (Rome 1958) 683-686. Kienast, D., Untersuchungenzu den Kriegs/lottender romischenKaiserzeit(Bonn 1966). Antiquitas 1.13. Laronde, A., "Le port de Lepcis Magna," ComptesRendus de l'Academiedes Inscriptionset BellesLettres (1988) 337-353. Le Gall,J., Le Tibre,fl,euvede Rome dansl'antiquite(Paris 1953). Lehmann-Hartleben, K., Die antiken HafenlagendesMittelmeeres(Leipzig 1923). Klio suppl. 14. Lepper, F., and S. Frere, Trajan'sColumn:A New Edition of the CichoriusPlates;Introduction,Commentary and Plates(Gloucester 1988). Loven, B., "The Zea Ship-sheds and the Buildings' Utility in Reconstructing the Dimensions of the Athenian Trireme," in TropisVIII. 8th InternationalSymposiumon Ship Constructionin Antiquity, Y dra2002, ed. H. Tzalas (Athens forthcoming). McKenzie, J., "Glimpsing Alexandria from Archaeological Evidence," Journal of Roman Archaeology16 (2003) 35-63. Meiggs, R., Roman Ostia (Oxford 1960, 2nd ed. 1973). Mord,J.-M.A.W., "Friihromische Schiffshiiuser in Haltem, Hofestatt," Ausgrabungenund Fundein West/alenLippe (Munster)5 (1987) 221-249. Morrison,J. S., withJ. F. Coates, Greek and Roman OaredWarships,399-30 B.C. (Oxford 1996). Pagano, M., M. Redde, and J.-M. Roddaz, "Recherches archeologiques et historiques sur la zone du lac d'Aveme," Melangesde !'EcoleFran,aisede Rome, Antiquite94.1 (1982) 271-323. Pernice, E., Die hellenistischeKunst in Pompei"i, vol. 6: PavimenteundfigurlicheMosaiken(Berlin 1938). Raban, A., "Ancient Slipways and Shipsheds on the Israeli Coast of the Mediterranean," in &ats, Shipsand Shipyards.Ninth InternationalSymposiumon BoatandShipArchaeology,Venice2000, Proceedings,ed. C. Bdtrame (Oxford 2003) 91-102. Redde, M., MareNostrum (Rome 1986). Rothaus, R., "Lechaion, Western Port of Corinth: A Preliminary Archaeology and History," OxfordJournal of Archaeology14.3 (1995) 293-306. Ruegg, S. D., "The Underwater Excavation in the Garigliano River: Final Year 1982. The Roman Port and Bridge at Minturnae, Italy," InternationalJournalof NauticalArchaeology12.3 (1983) 203-218. Schmiedt, G., "Contribution of Photo-interpretation to the Reconstruction of the Geographic-topographic Situation of the Ancient Ports in Italy," in Proceedings of the 10th Congressof the InternationalSocietyof Photogrammetry(Lisbon 1964). --, ll livelloanticodel mar Ti"eno (Florence 1972). Starr, C. G., The Roman ImperialNavy,31 B.C.-A.D. 324, 2nd ed. (Cambridge 1960).

COSA AND DEEP SEA EXPLORATION Anna Marguerite McCann

E

ver since the fourth century B.C., when Aristotle wrote of snorkels and diving caldrons, human curiosity has led ever forward to new inventions for exploring the sea. Leonardo da Vinci made drawings for underwater breathing devices in the fifteenth century. True diving bells followed in the sixteenth and seventeenth centuries, enclosed barrels in the eighteenth century, and submarines and hard-hat diving suits in the nineteenth century. With the invention of the Aqua-Lung by Jacques Yves Cousteau and Emile Gagnan in 1943, people could at last move and work freely beneath the water. Such dives, however, are limited to shallow waters. In the first shipwreck site excavated using SCUBA at the Grand Congloue off Marseille in 1952, 1 Cousteau and his team were able to work to a depth of 45 m. The "underwater dig" had been born. In the 1950s Italian archaeologist Professor Nino Lamboglia began the first explorations of ancient shipwrecks by an archaeologist at Spargi (Sardegna) and Albenga with hisunderwater research vessel, the "Daino. "2 He introduced the use of grid frames, allowing for accurate surveying and photography underwater. All who have followed after himowe hima debt as the first professional archaeologist to seriously and systematically recover and study the artifacts found in the sea. He was director of the Istituto lnternazionale di Studi Liguri and Superintendent of Antiquities for Liguria. Recognized for his major contributions to the study of Roman pottery, Lamboglia was also the founder and director of the Centro Sperimentale di Archeologia Sottomarina and the Museo Navale Romana at Albenga. Furthermore, he was the first to organize international congresses in the field of underwater archaeology where scholars, divers, and technicians working in the sea came together from all parts of the Mediterranean. 3 The last congress took place in Lipari (Sicily) in 1976, a year before hisdeath. In 1974, a collaboration was formed between the Istituto lnternazionale di Studi Liguri and the American Academy in Rome. A group of American archaeologists, scientists, and divers, which I directed, and a group ofltalian archaeologists and divers directed by Lamboglia explored together the ancient Etruscan ports of Populonia and Pyrgi. This project represents the first international collaborative effort in Italy for the field of underwater archaeology. 4 The development over the last fifteen years of robotic technology allowing exploration of the seafloor to its maximum depth of 11 km has revolutionized underwater archaeology. Over three-fifths of our planet is covered with water, and approximately 97 percent of the ocean floor still remains unexplored. Now, with remotely operated vehicles (ROVs), archaeologists and scientists can safely search with camera, video, and sonar at unprecedented depths. 1 Delgado 1997, "Grand Congloue," 174-175; Parker 1992, 200-201,nos.472-473. 2

See McCann 1977a. Spargi wreck: Parker 1992, 409-410, no. 1108; Albenga wreck: Parker 1992, 49-51, no. 28.

3

McCann 1967; 1971.

4

See below, nn. 16 and 17.

38

ANNAMARGUERITEMcCANN

Robert D. Ballard understood the potential for exploration in the deeper oceans, and in 1989, with a talented group of engineers from the Deep Submergence Facility at the Woods Hole Oceanographic Institution, he supervised development of a sophisticated ROV named Jason that works in tandem with a towed, unmanned camera sled named Medea.' In the same year, Ballard formed a scientific team to explore part of an ancient trade route in international waters off the northwest tip of Sicily, ea. 60 km north-northeast of Keith Reef at Skerki Bank, a treacherous reef that lies just 1.8 m below the surface. The area where the wrecks were discovered in 1989 as well as in 1995 and 1997 lies about 100 km north of Tunis and about 120 km due west of Trapani on the northwest coast of Sicily.6 This project, named the JASON Project, involved the first archaeological excavation in the deeper seas, at a depth of ea. 800 m. It was also the first occasion to involve a live, interactive television broadcast linked to an ROV at work on the seafloor. This first JASON Project was conceived and directed by Ballard for the education of children in the sciences and archaeology. I served as the archaeological director. Some 225,000 children at science museums in the United States and Canada communicated directly with the team in only a half-second of time. The project won many awards, including the American Association for the Advancement of Science Westinghouse Award and the Computerworld Smithsonian Award. The JASON Projects have continued every year since 1989 and now reach about 1.5 million schoolchildren and 33,000 teachers annually.7Exploration in the deeper seas thus began with an education mission and a desire to share our maritime heritage. The area of the Skerki Bank wrecks was first identified by Ballard in 1988 by means of a towed survey system named Argo using a digital Electronic Camera (ESC) and black-and-white video SIT cameras. 8 His attention was caught by a high concentration of ancient amphoras. Although the amphoras were scattered, they lay roughly in a line running northwest-southeast and may have been jettisoned from ships experiencing difficulties while following an ancient deep-water trade route. The amphoras document use of thistrade route from the later fourth century B.C. through the ninth-twelfth centuries A.O. Also identified in 1988 was a distinct late Roman shipwreck later nicknamed the "Isis" after the Egyptian goddess who protected sailors and promised afterlife. In 1989, using the dual ROV Medea/J ason vehicle system, both the scattered amphoras and the "Isis" shipwreck were investigated. The 1989 Skerki project proved that wide areas of the seafloor could be searched effectivelyin a reason able time. Photomosaicing techniques and acoustically derived bathymetric maps had not as yet been developed, but the site was mapped and documented by both video and still-camera images before objects selected by me and the conservator, Mary-Lou E. Florain, Royal British Columbia Museum, Victoria, B.C., were recovered. They ranged in date from ea. 300 B.C. through the period A.O. 850-1115. The late Roman wreck, the Isis, is dated by its artifacts to the last quarter of the fourth century A.O. Further research expeditions to Skerki Bank took place in 1995 and 1997, using the U.S. Navy's nuclear research submarine NR-1 (fig. 1).9 Designed for use in the cold war by Admiral Hyman Rickover and launched in 1969, its great advantage is its high-powered sonar, which has a forward range of 2,000 yards and can pick out the amphora mounds on the shipwrecks at this distance. With it in 1995 Ballard located three more shipwrecks: a late nineteenth-century wooden sailing ship (Wreck C), a small, probably medieval ship (Wreck A), and another merchantman (Wreck B).In the summer of 1997, using the Medea/Jason vehicle system as well as the U.S. Navy's nuclear submarine ' See the website at http://www.whoi.edu/marine/ndsf/ vehicles/jason/index.html.

project.htm. 8

6

McCann and Freed 1994; McCann and Oleson 2004.

7

See the website at http://www.jasonproject.org/jason_

Ballard et al. 2000.

9 See McCann and Oleson 2004 for full discussion and Ballard et al. 2000.

COSA AND DEEP SEA EXPLORATION

39

43' N

42'

41'

40'

39'

38'

37'

36'

Fig. 1. The Tyrrhenian Sea bounded by the islands of Sicily, Sardinia, and Corsica, the Italian peninsula , and the north coast of Africa . The largest star is the location of the Skerki Bank Research Area. Depth is in meters (Ballard et al. 2000, Institute /or Exploration/Woods Hole Oceanographic Institution) .

NR-1, Ballard, McCann, and D . Yoerger returned to Skerki Bank with a full collaborative team of engineers, archaeologists, conservators, and oceanographers. A major goal was to prove that archaeologists could carry out survey and limited excavation in the deep sea at the highest professional standards. Four additional shipwrecks were found, three being Roman (D, F, and G) and one a late nineteenth -century wooden sailing ship (E).A large portion of a square with sides of 15 x 12 nautical miles was surveyed during these combined projects, and eight shipwrecks in all were discovered at depths of ea. 750-800 m. In total, the project surveyed five Roman merchant ships ranging in date from the early first century B.C. to the late fourth century A.D., one probably early medieval fishing vessel, and two late nineteenth-I early twentieth-century iron and wood sailing ships. Before any artifacts were lifted, each shipwreck was mapped and photographed by ROV Jason. By means of a combination of precisely controlled track-lines and sensor data, detailed photomosaics

•

40

. .·

ANNA MARGUERITE McCANN

·@ SKERKi11s1s0 Q) Cosa · (i) Ostia Q) Puteoli

8 0 0

Arelate Nora Utica

0 0

Luna · Rhodes

@ Londinum

Lilybaeum

0

Gades

Fig. 2. Possibletrade routesof the ships wreckedat Skerki Bank (E. P.Oberlander).

and precise bathymetric maps of each site were made. While precise measurements cannot be taken from the photomosaics, the microbathymetric maps made by scanning sonar techniques are accurate to within a few centimeters. Both the photomosaics and bathymetric maps were printed out on shipboard. Only then did the archaeologists select the artifacts to be lifted. Confirmation of our ability to map precisely and to photograph a complete shipwreck in deep water in a limited amount of time was a major achievement of the 1997 expedition. These tools and techniques may also be applied to many shallow-water sites. The wrecks and artifacts have thrown light on an ancient trade route on the high seas between Carthage and Rome (fig. 2). We knew from literary sources there was a quick route over the open seas between Africa and ports along the west coast of Italy: Cato the Elder in the first half of the second century B.C. showed the Roman senate a fig that "had been picked at Carthage the day before yesterday. " 10 Until now, however, the route was uncertain. The Skerki Bank wrecks also reveal that smaller merchant ships, engaged in cabotage rather than large-scale delivery of goods, also followed or crossed this route. 10 Plin . HN 15.74-75. For length of voyage between Ostia and Africa from two to four days, see Casson 1995, 282-284,

with two days the record time with favorable winds and a speed of 6 knots .

COSA AND DEEP SEA EXPLORATION

41

The 1997 Skerki Bank expedition increases our understanding of shipping patterns in the Mediterranean and the composition of cargoes. The cargoes from the five Roman shipwrecks are surprisingly varied, each including material from both the eastern and western Mediterranean, which emphasizes the role of this sea as an open highway. The varied nature of the cargoes suggests that most, if not all,of the ships wrecked had been engaged to some degree in cabotage as they circled the western Mediterranean. The Skerki ships were allprobably medium-sized merchantmen, varying in size from about 15 m (Isis) to close to about 40 m (Wreck B).They give a new picture of Roman trade and illustrate well P. Temin's conclusion that the early Roman Empire was a market economy based on a number oflocal markets connected together around the Mediterranean: "ancient Rome had an economic system that was an enormous conglomeration of interdependent markets." 11 The Skerki projects have also provided new evidence for the importance of the Roman port of Cosa on the coast of Tuscany for cabotage.12The earliest of the Skerki Roman wrecks (D), dating to ea. 80 to 50 B.C., was probably loaded at Cosa or at least stopped there on its way south before it was wrecked off Skerki Bank (fig. 3). It is also clear from the discovery of the earliest known Sestius stamp found at Skerki on a whole Graeco-italic jar, Will Type ld, that the Skerki route was in use by the Sestii as early as the second quarter of the second century B.C. (figs. 4-5).13 In the light of these discoveries, I was delighted to have the opportunity to return to Cosa and explore ancient trading routes leading to Cosa with some of the newest robotic technology. This opportunity came in 2000 when I was asked by Chryssostomos Chryssostomidis, Doherty Professor of Ocean Science and Engineering, Massachusetts Institute of Technology, and director of MIT's Sea Grant Program, to advise them on an exploration program in Italy. With funds from the U.S. government and MIT the AUV Laboratory at MIT Sea Grant had developed a new AUV, an Autonomous Underwater Vehicle.14 Dott. Giovanni Auletta Armenise of Rome, who wished to bring this newest robotic technology to Italy, funded the cost of the expedition and some of the specialized equipment to be used during the expedition. AUVs are becoming the robot of choice now for underwater exploration. 15 They have already been used extensively by the military and by marine geologists to map the seafloor, but they have not as yet been used very much for archaeology. So this project in Italy was one of the very first to use an AUV for archaeology. The great advantage of an AUV is that it is much less expensive to build and much easier to operate than an ROV, for it works independently and is free of the long cable that links an ROV to a mother ship. An AUV is programmed by computer before it is sent down and then returns with the information. It provides a stable platform for both sonar and visual imaging and can cover wide areas of the seafloor. AUVs in operation today are capable of operating at depths in excess of 2,000 m for extended periods of time. The big disadvantage of an AUV is that there is no real-time connection, and you must wait for information. This has now been developed. See recent bibliography in note 15. I saw this as an opportunity with the MIT engineers both to use an AUV and to return to the Tuscan coastline that I know well to explore ancient trade routes (fig. 6). Of particular interest to me is the port of Cosa since I have directed excavations there for many years. Cosa is the 11 Temin

2001, 181.

12 McCann

et al. 1987; McCann 2002.

13 McCann and Freed 1994, 64-68 (MJ89-29), figs. 46, 47a, and47b.

14 See website for MITs AUVs: http:/ /auvlab.mit.edu/ vehicles.

15 For a list of AUVs in use today, see http://www.transit-port. net/Lists/ AUVs.html; Sibenac et al. 2004; Manley 2005, 3-4; Kocak and Caimi 2005; Curtin 2005; Morash et al. 2005. For ABE, the AUV developed at WHOI, see Yoerger et al. 2007 a; 2007b; Singh et al. 2007. For SeaBED, an AUV also developed at WHOI, and its work in Greece in 2005, see the website: http:/ /www.whoi.edu/sites/ archaeology/Chios 2005.

ANNA MARGUERITE McCANN

42

Fig. 3. Skerki Bank Wreck D, 80-50 B.C. Photomosaic, comprised of about 180 images, with location of individual arti/acts recovered in 1997 as well as arti/acts identified but not recovered (Ballard et al. 2000, Institute /or Exploration/Woods Hole Oceanographic Institution) .

earliest Roman harbor thus far known going back to the third century B.C., and a key harbor for the western Mediterranean during the last two centuries of the Roman Republic. My hope was to link together what we know from shallow-water exploration using SCUBA with research in the deeper waters using the new robotic technology. Furthermore, I have always collaborated with Italians both at the Roman port of Cosa and at the Etruscan ports of Populonia 16 and Pyrgi 17 16 McCann

1972; McCann 1977b.

17 McCann and

Oleson 1974.

r;:=~--·· ,.... r:.~~ '-" ·...\_· I'...::., :·-....J ~ .:_ .___ J} · C-- ...,C...__,_..·· I~r:! I

Ou,

--~

Fig. 5. Sestius stamp on amphora M]89-29 (SES with vertical trident); drawing 1:1, stamp 3.5 x 1.0 cm.

Fig. 4. "Graeco-ltalic"amphora M]89-29 (Will Type ld), second quarter of the second century B.C., Skerki Bank trade route, Amphora Alley I. Scale: 1: 10 (drawing E. P Oberlander).

Fig. 6. TuscanDeep WaterSurvey, areasof exploration 2002 (map R. Damus, MIT, Sea Grant Program,AUV Lab).

44

ANNA MARGUERITEMcCANN

Fig. 7. MIT's AUV Caribou, Odyssey III class (photo A. M. McCann).

Fig. 8. GIB buoys (GPS Intelligence buoys) used for sonar mapping (photo A. M. McCann).

and realize that it is both essential and very valuable to collaborate with archaeologists from the country where we are guests. In today's global world, as the deep sea is opening up, collaboration is even more important. So I formed a collaboration that included: engineers from MIT who provided the technology, the Soprintendenza ai Beni Archeologici per la Toscana with Dott.ssa Pamela Gambogi, director of underwater archaeological research for Tuscany, and the American Academy in Rome with the help of Professor Archer Martin, Archaeology Supervisor. Also included were engineers from a consortium of Italian universities at Genoa, Pisa, and Siena (ISME) as well as the Museum at Arenzano (MUVITA). The Italian navy supplied two support ships in the summer of 2002, when the principal field trip took place from 28 June to 13 July. In all, some nineteen archaeologists and engineers (both professors and students) worked together to explore underwater sites between Populonia to the north and Ansedonia to the south. Over one million square meters of seafloor were explored, including preliminary work covered in 2001 and 2002. Dott. Claudio Mocchegiani Carpano from the Ministero dei Beni Culturali in Rome visited the project, which was filmed by RAI, Italian television and radio. MIT's AUV, named the Caribou, is an Odyssey III class (fig. 7). It was mounted with sonar only-both a sub-bottom profiler and a side-scan sonar that can reach 150 m on a side. It can go to depths of about 1,000 m, and going at about 3 knots it can cover an area of about one square kilometer in one hour . The AUV usually worked about 10 m off the bottom, although it can fly at shallower depths. MUVITA also contributed a remotely operated vehicle, ROV Phantom, with a

COSA AND DEEP SEA EXPLORATION

.. ,

.., ,,

101

,

'! ,(

..,A

45

..,

A

..

A

DePt!f:' 53

~

, ,

·2

"" 11 0

,,

,
~ ,IJ, ~,

~

, 1;... y

~

\(,

y

'

-~

10)

..

y

10 1

,

..

G 10,

in

* wreck - surveyline surveyarea sitearea - AUV.. surveyline 101

120

1:15

•

110

..

,

..

,

Fig. 9. Searchareawith the AUV and ROVs in channelbetween the western coastof the Argentario and the islandof Giglio, from Pte. CalaGrandeand Capod'Uomo (map R. Damus,MIT, Sea Grant Program, AUV Lab).

video camera. Sonar and a ROV (Pluto) were also supplied by the Italian navy ship MHV Viareggio. For sonar mapping four GIB buoys (GPS Intelligence buoys) were used (fig. 8). Each buoy includes a Global Positioning System receiver linked to the underwater vehicle by an acoustic pinger to allow the vehicle to be tracked while it is performing a mission. A particular focus was to explore one of the key trading routes leading into the port of Cosa, the channel between the western coast of the Argentario and the island of Giglio, where depths vary up to 140 m. A systematic search was begun between Pte . Cala Grande and Capo d'Uomo, a distance of about 2.3 nautical miles, using the sonar on board the Viareggio and with the video cameras on ROV Pluto and ROV Phantom (fig. 9). We worked about 1 km from shore, thinking that in strong winds ships would have stayed closer to shore rather than in mid-channel. We used a grid pattern with lines 100 m apart and 1,500 m long working at depths between 60 and 90 m. One east-west course took about one hour, and about twenty-six lines were covered in all. We investigated a number of the sonar targets with the ROV, but only one promised to be of interest. At a depth of about 63 m a large site about 20 m wide was discovered. It turned out to be a large area covered by a fisherman's net. It was difficult to see what was beneath the net, but it could be a ship and bears further investigation. It would be worthwhile to continue with a systematic survey of this key trading route between the Argentario and Giglio and cover the channel itself, where depths between 120 and 140 mare reached. This whole area is rich in ancient shipwrecks found by SCUBA divers, including the unique Etruscan ship found

ANNA MARGUERITE McCANN

46

Fig. 10. Elba South, shipwreck found by Comex in 1996, end of first century B.C., first century A.D. (Comex S. A./Marseille , Archivo Sop. Beni Arch. Toscano 61903/ 1996).

off Giglio. 18 Finds from this area are now on exhibit in the stunning new Underwater Archaeological Museum in the Spanish fortress at Porto Santo Stefano. 19 The majority of amphora finds from this area are Dressel type IB, one of the most popular wine amphoras dating from about 80 to 30 B.C., made by the Sestius family at Cosa as well as elsewhere. With the demise of the port of Cosa at the end of the first century B.C., however, we also found evidence along the Tuscan coast of the shift in the first century A.D. to the dominance of the Spanish import trade in wine and garum. A dramatic example of the extent of this trade was a large Spanish wreck we investigated off the island of Elba, a center for mining and quarrying in antiquity. This Roman wreck is located about 12 km off the southern side of Elba at a depth of 176 m. Spanish wine and garum jars cover an area of about 20 to 30 m (fig. 10). The wreck was actually found earlier by the French and filmed by Comex in 1996. Due to high seas, MIT's AUV Caribou could not get down to it. Unfortunately, the wreck site has been dragged heavily by fishermen, and the pottery is very broken, but a number of the Spanish amphoras are identifiable, including Pascual 1 (fig. 11),20 Haltern 70, 21 Dressel 10,22 and Dressel 7-11. 2 ' These types are also found in the Archaeological Museum in Portoferraio (Elba), reaffirming the dominance of Spanish trade in the area at the end of the first century B.C. and the first century A.D. 24 We located a second Spanish wreck close to the eastern side of the Formiche II island, east of the Argentario, at a depth of about 90 m (fig. 12). About a dozen Dressel 20 25 oil amphoras 18 Delgado

24

20

Sciallano and Sibella 1991, Amphore Pascual I.

21

Sciallano and Sibella 1991, Amphore Haltern 70.

22

Sciallano and Sibella 1991, Amphore Dressel 10.

For examples of other enormous wrecks with Spanish amphoras and dolia, see particularly Diana Marina wreck found off the northwest coast of Italy (Corsi-Sciallano and Liou 1985). It carried at least 1,000 Dressel 2-4 amphoras from Tarraconensis as well as dolia each containing 1,500 liters of wine. The hull is estimated to be 22-25 m in length. See Parker 1992, 163-164, no. 364. AlsoseetheChiessi (Marciana) wreck found off Elba (Rossi 1982) dating probably A.D. 60-85, which carried about 5,000-7,000 Spanish amphoras for wine and fish sauces. See Parker 1992, 140, no . 301.

2'

Sciallano and Sibella 1991, Amphore Dressel 7-11.

25 Sciallano

1997, "Giglio Wreck ," 169-171.

19 See

the wreck from Formica Piccola : Poggesi and Rendini 1998, 86, cat. no. 3.

and Sibella 1991, Amphore Dressel 20 .

COSAAND DEEP SEAEXPLORATION

47

Fig. 11. Amphora Pascual1, scale:1:10 (photoSciallanoand Sibel/a 1991, "AmphorePascual1," Golie de Fas, isolee,Museed'Istres).

were visible, covering an area of about 3 m with a height of about 2 m. Other pottery visible were mortaria, commonware dishes, and cooking vessels. The Dressel 20 amphoras date the wreck from about the mid-first to the mid-third century A.O., further evidence for the shift in early Imperial times to the import of wine, oil, and garum from Spain. It would be valuable to link together the trading history of this active coastline in antiquity with further systematic exploration. The last day of the project, AUV Caribou while mapping the Formiche II wreck site experienced technical difficulties and crashed between two rocks at a depth of about 40 m. But it was rescued by Italian divers on board the Viareggio and was not damaged . Thus, this pioneering collaborative effort in Italy's deeper seas ended on a high note of successful collaboration. The AUV had made twenty-three dives during the whole project, with a total bottom time of 8 hours and 58 minutes . Over seven gigabytes of sonar data were collected. Over 1,000,000 square meters of seafloor were covered in all, including the area previously surveyed by a towed sonar in 2001. The Tuscan expedition provided valuable information for the development of AUV technology, particularly how bathymetry influences an AUV's performance during a mission. Surveys performed over flat seafloor that has a gentle, continuous gradient have a much higher chance of success than those performed over a discontinuous bottom, where the bathymetry varies greatly, like the area around the Formiche Islands. Because of valuable lessons learned from the 2002 Tuscan expedition,

_./"? I -~

......

. - ..... :-.>

....

__

-........ ,

)

NN-

)

--

.,.,,.,. ..., ..

\

,J

_...

I

...........

-- -

.,.

/

NZSOO

......

,.,...

,.,,.

Fig. 12. AUV searchareasoff the FormicheIslands(map MIT Sea Grant Lab).

Fig. 13. Xanthos, OdysseyII class AUV, MIT Sea Grant Lab with 600 kHz side-scansonarand a digitalstill camerawith strobe (photoMIT Sea Grant Lab).

...

,..,

I

COSAAND DEEP SEAEXPLORATION

49

the MIT AUV Lab has now further developed and successfully field-tested an Odyssey II class AUV, Xanthos, for broad area survey and visual inspection (fig. 13). The Odyssey III, Caribou, has since been transferred to the Mechanical Engineering Department at MIT and is no longer used by the AUV laboratory for archaeological missions. The Xanthos has a 600 kHz side-scan sonar and a 1.3 megapixel digital still camera and strobe solution. It was tested in Greece the summer of 2004 off the coast of Kythira. The Xanthos performed side-scan surveys in 85 m of water. After localizing features of interest in the side-scan record during post-mission analysis, the MIT group sent the vehicle down to take pictures of potential targets. This mission was successful, and a seventeenth-century anchor was photographed that had been identified with the sonar by first diving to the location of the anchor. Then a tight survey was performed with track-line spacing of 5 m to get a high-density image coverage and overlap of the expected target field. The forward-looking obstacle-avoidance sonar has now been implemented, and the MIT Lab is working on algorithms to make the AUV avoid obstacles, such as those encountered off the Formiche Islands. This willfurther increase the ability of the AUV to operate in hazardous bathymetric environments. The MIT AUV Lab is currently building and testing a new, general purpose, inexpensive 6,000 m Hovering Autonomous Underwater Vehicle (HAUV), the ODYSSEY IV class. This vehicle is intended for rapid deployments, potentially with minimal navigation, thus supporting episodic dives for exploratory missions. For these situations, the vehicle is capable of fast dive times, short survey on the bottom, and simple navigation. This vehicle is a new direction for hovering vehicles and is based on work from the lab's first HAUV, currently employed by the navy for their Explosive Ordinance Disposal (EOD) to inspect ship hulls with technology developed from a joint research partnership with the Office of Naval Research and Bluefin Robotics. This HAUV will also allow close inspection of archaeological sites, as its high control provides precise maneuvering in proximity to surfaces like a ship hull. Such precision willlead to AUVs performing surveys with the same level of image-mosaic detail that an ROV can currently produce. 26 The plan is to continue to explore systematically ancient trade routes in the future, both along the Tuscan coastline as well as other areas in the Mediterranean. Such collaborative efforts-between Americans and Italians, between archaeologists and engineers, with the generous support of the Italian navy-point the way for the future.

This information has kindly been provided by R. Damus from the MIT Sea Grant AUV Lab.

26

50

ANNAMARGUERITE McCANN

WorksCited Ballard, R. D., A. M. McCann, D. Yoerger, et al., "The Discovery of Ancient History in the Deep Sea Using Advanced Deep Submergence Technology," DeepSeaResearch,pt. 1, 47 .9 (2000) 1592-1595. Casson, L., Shipsand Seamanshipin the Ancient World(Princeton 1995). Corsi-Sciallano, M., and B. Liou, "Les epaves de Tarraconaise a chargement d'amphores Dressd 2-4," Archaeonautica5 ( 1985) 95-107. Curtin, T. B., D. M. Crimmins, J. Curcio, M. Benjamin, and C. Roper, "Autonomous Underwater Vehicles: Trends and Transformations," MarineTechnologyJournal39.3 (2005) 65-75. Ddgado, J.P., ed., Encyclopediaof Underwaterand MaritimeArchaeology(New Haven 1997). Kocak, D. M., and F. M. Caimi, "The Current Art of Underwater Imaging-With a Glimpse of the Past and Vision of the Future," MarineTechnologyJournal39.3 (2005) 5-26. Manley, J.E., "State of Technology Report: Marine Technology in 2005," Marine TechnologyJournal39.3 (2005) 3-4. McCann, A. M., "The Underwater Conference in Miami," Archaeology20 (1967) 302-303. --, "The Fourth International Underwater Archaeology Congress in Nice, October 1970," Archaeology 24 (1971) 20-22. --, "Underwater Excavations at Populonia, 1970," InternationalJournalofNauticalArchaeology1 (1972) 398-402. --, "Necrology: Professor Nino Lamboglia," Journalof Roman Archaeology4 (1977a) 264-265. --, "Underwater Excavations at the Etruscan Port of Populonia," Journalof FieldArchaeology4 (1977b) 275-296. --, The Roman Portand Fisheryof Cosa:A Short Guide (Rome 2002). McCann, A. M.,J. Bourgeois, E. K. Gazda,J. P. Oleson, and E. L. Will, The Roman Portand Fisheryof Cosa: A Centerof Ancient Trade(Princeton 1987). McCann, A. M., and J. Freed, Deep WaterArchaeology:A Late Roman Shipfrom Carthageand an Ancient TradeRoute nearSkerkiBank off NorthwestSicily(Portsmouth, RI 1994). J oumal of Roman Archaeology suppl. 13. McCann, A. M., and J. P. Oleson, "Underwater Excavations at the Etruscan Ports of Populonia and Pyrgi," Journalof FieldArchaeology1 (1974) 398-402. --, Deep-WaterShipwrecks0/fSkerki Bank: The 1997Survey(Portsmouth, RI 2004).Joumal of Roman Archaeology suppl. 58. Morash, J., R. Damus, S. Desset, V. Polidoro, and C. Chryssostomidis, "Adapting a Survey-Class AUV for High Resolution Seafloor Imaging," in Proceedingsof the 14th InternationalSymposiumon Unmanned UntetheredSubmersibleTechnology(UUST) (Durham, NH 2005). Parker, A. J., Ancient Shipwrecksof the Mediterraneanand the Roman Provinces(Oxford 1992). British Archaeological Reports S580. Poggesi, G., and P. Rendini, eds., MemorieSommerse.ArcheologiaSubacqueain Toscana(Pitigliano, Grosseto 1998). Rossi, D., "Rditto di Chiessi," in Relitti romanidell'isolad'Elba,ed. M. Zecchini (Lucca 1982) 128-133. Sciallano, M., and P. Sibella, Amphores:Commentles Identifier?(Aix-en-Provence 1991). Sibenac, M., T. Podder, W. Kirkwood, and H. Thomas, "Autonomous Underwater Vehicles for Ocean Research: Journal38.2 (2004) 63-72. Current Needs and State of the Art Technologies," Marine Technology Singh, H., C. Roman, 0. Pizarro, R. Eustice, and A. Can, "Towards High-resolution Imaging from Underwater Vehicles," InternationalJournalof RoboticsResearch26.1 (2007) 55-74. Temin, P., "A Market Economy in the Early Roman Empire," Journalof Roman Studies91 (2001) 169-181. Yoerger, D.R., A. M. Bradley, M. Jakuba, C.R. German, T. Shank, and M. Tivey, "Autonomous and Remotdy Operated Vehicle Technology for Hydrothermal Vent Discovery, Exploration and Sampling," Oceanography20.1 (2007a) 86-95. Yoerger, D.R., M.Jakuba, A. M. Bradley, and B. Bingham, "Techniques for Deep Sea Near Bottom Survey Using an Autonomous Underwater Vehicle," InternationalJournalof RoboticsResearch26.1 (2007b) 41-54.

ROMANSHIPSHEDSAND ROMANSHIPS Boris Rankov

I

t is striking that the survey of Roman shipsheds by David Blackman, which appears elsewhere . in this volume, notes what appears to be considerable numismatic, iconographic, and literary evidence of their existence while identifying very few such buildings in the archaeological record. To date, there are only four sites where structures that have been identified as Roman-period shipsheds have been published in detail: Velsen in the Netherlands, Haltern in Germany, Caesarea Maritima in Israel, and the river port at Ostia. This paper considers these identifications and their plausibility in the light of what we know about the ships they might have housed and the general practicalities of the operation of shipsheds. The suggested sheds at Velsen and Haltern need to be considered together. Velsen was a Roman fort of ea. A.D. 15 to 30, situated on a tidal inlet of the North Sea at the mouth of the most northerly branch of the Rhine (today the Oer-lJ) on the coast of the Netherlands. The banks of the river formed two sides of the fort, and J.-M. A. W. Morel has argued that two hall-like buildings excavated within the area of the fort between 1978 and 1982, each 20 m long by 6 m wide (fig. 1, buildings F and G), were probably shipsheds. 1 One, the Western Boat House (building F), apparently lay near the riverbank, while the other, the Eastern Boat House (building G), lay farther to the south, some 32 m from the bank, and may have superseded it as a result of remodeling of the harbor. Morel bases his interpretation on comparisons with the ground plans of shipsheds in the Mediterranean, which are of similar widths and either the same length or twice as long. He accepts, however, that no direct comparison can be made between Greek sheds of the Classical and Hellenistic periods located on the Mediterranean and Roman sheds of the early Imperial period located on a tidal river on the northern frontier. The suggested shipsheds at the so-called Hofestatt site on the river Lippe at Haltem in northern Germany, occupied by the Roman army between 715B.C. and A.D. 9, were excavated by E. Kriiger between 1901 and 1904.2 Kriiger identified a large building measuring 45.5 by 25 m (fig. 2) from the last phase of the Hofestatt enclosure (the so-called blue fort, named for the color in which it appeared on the excavation plan), as a series of barracks. He also suggested as an alternative that they might be shipsheds, comparing their ground plan to the shipsheds that had just been found at Oeniadae in western Greece in 1904.3 Both suggestions were dismissed by C. M. Wells, who argued instead that they were granaries,4 but the shipshed theory was revived by J.-M. A. W. Morel in 1987, five years after the Velsen excavations.' 1 Morel

1986,205.

2

Dragendorff, Koepp, and Kriiger 1905.

3

Sears 1904.

4

Wells 1972,203-204.

' Morel 1987.

Fig. 1. The location of the "Boat Houses" at Velsen: G = Eastern Boat House; F = Western Boat House (after Morel 1986, 204, fig. 3B).

I

I Fig. 2. Haltern-Hofestatt: Morel's reconstruction of the "Schi/lshauser" (from Morel 1987, 227, fig. 4).

:

", ....

ROMANSHIPSHEDSANDROMANSHIPS

53

Kriiger had seen the distance of the building from the bank of the Lippe as a problem for the shipshed interpretation, but Morel argued that while Mediterranean shipsheds did not need to be positioned away from the water because of the negligible tides, those sited on a riverbank needed to be positioned above the floodplain. This, he suggests, is why the Hofestatt building was located 30 to 32 m away from the riverbank. What matters, of course, is not how far away from the riverbank the building lay, but how high above water level. Unfortunately, Kriiger's report does not give levels, so that it is impossible to determine the validity of Morel's argument. By associating further postholes with the building, Morel reinterpreted it as measuring 51.8 m by 33.05 m, and a series of parallel long trenches 25 m long and 6 m apart, each with a number of short cross-trenches 1.8 m apart (aptly described by Wells as "fish-bones") as the sleeper trenches for wooden groundways to support ships' keels. Evidence for the use of such groundways can be found in rock-cuttings at several Mediterranean shipshed sites,6 and in fact they would have been essential to the easy slipping and launching of vessels. While these sleeper trenches do indeed look like groundways, it is highly problematic that they are only 25 m long and do not even stretch as far as the southern end of the building. The westernmost "groundway" stops ea. 3 7 m from the water's edge and the easternmost ea. 36 m. Morel appears to show conjectural shallow trenches continuing from each "groundway" to the riverbank in his new reconstruction plan but offers no evidence for these features, which certainly do not appear in Kriiger's report. Morel also underestimates the difficulties of moving, say, a libumian of around 7 tonnes,7 such as might have been housed in such sheds, even along a groundway. The vessel would have to be kept upright with "walking shores" as it was hauled down; that is, as each shore moved forward and fell over with each heave of the ropes, it would be replaced by another, which would itself be replaced on the next heave, and so on. This would be a skilled operation even on a greased wooden groundway but significantly more difficult if the ship were merely being moved along a trench. It is very hard to believe that if the "fish-bone" trenches did represent wooden groundways, rather than, say, internal wall foundations, they would not have been continued down to the riverbank. What really makes unacceptable the interpretation of the Hofestatt building as shipsheds, however, is the near impossibility of launching and slipping vessels from here. The Lippe flows westward toward the Rhine and was perhaps 50 m wide at this point. Launching a long ship directly across a significant stream, as would have been the case from the Hofestatt site, would have been very difficult indeed. A ship 20 m long, such as could have been housed here, would have stretched well into the main stream of the river. As it moved out, the ship would have been increasingly difficult to control. Even if the bows were tethered to the bank, the river would have put a dangerous strain on the ropes and on the hull. Once the whole hull was afloat, it would have been in serious danger of being swept away altogether. Slipping would, if anything, have been even more difficult than launching, since on a flowing stream it would have been quite impossible to line up the stem on the groundway. End launching across a river is accomplished in many modem shipyards but only on tidal rivers, where the flow of the stream is reversed twice a day and where launching is carried out at slack water at the top of the tide. Launching into a nontidal river in antiquity would normally have required either side-launching, such as is practiced today in many parts of the world including the Great Lakes in North America, or end-launching downstream, either at a narrow angle to the bank or, preferably, parallel to the stream into a basin cut into the side of the river. Significantly, the 6 E.g.,

at Kos (Kantzia 1987), Carthage (Hurst 1994, 33-39, esp. 32), Rhodes (Blackman, Knoblauch, and Yiannikouri 1996). See Coates and Shaw 1993; Blackman 1995, 232.

7

Morrison 1996, 317; cf. appendix Don p. 345.

••

II .;. '

•

---

DWELLINGS

-

.'

D

~

~

OUTERHARBOR

I I

\---// •

INNER HARBOR

PALACE

Fig.3. Caesarea,the RoyalHarborand South Bay,suggestedshipshedsitesat LLIS, TP,and CC (fromRaban2003,96,fig. 15.5).

ROMAN SHIPSHEDS AND ROMAN SHIPS

55

two late first-century ships found on the banks of the Brautlach River at Oberstimm 8 and the five late Roman ships abandoned on the banks of the Rhine at Mainz 9 were all found aligned more or less parallel to the river. For these reasons, the identification of the Hofestatt building as shipsheds can effectively be ruled out. The objection about launching and slipping across the stream cannot of course be applied at Velsen, where the river is indeed tidal and where launching could have been accomplished safely at slack water either at high or at low tide. Being able to launch or slip only at four specific times during a twenty-four period would have been restrictive but might have been acceptable. Some other objections do apply, however. The gradient within the "Eastern Boat House" is given as 1 in 150, which is impossibly shallow. We may compare the 1 in 20 gradient at Munychia, which is one of the shallowest among those buildings that can be certainly identified as shipsheds. 10 A significant gradient is essential because of the great difficulty of launching a heavy ship without it and the danger of its sticking even on a greased wooden groundway. Moreover, during launching and slipping the ship must descend to or be drawn up out of the water, and it is not feasible for a "long" ship (navis longa,an oared warship) to move between an essentially flat and an inclined groundway, since the ship would "tip" at the point of transition and an intolerable and possibly fatal strain would be placed upon its straight keel. There is, in any case, no evidence of such a groundway, either leading to the water's edge or within the area of the "Boat House" itself, where it should certainly have been picked up since the building itself is defined by finds of postholes. There are thus very strong reasons for rejecting the identification of this building as a shipshed. But if the "Eastern Boat House" cannot be a shipshed, that greatly weakens the case also for the "Western Boat House," which is of almost identical ground plan and which it may have superseded. It too has no evidence for a groundway, and even though its gradient is steeper at 1 in 30, this remains outside the range known at securely identified shipsheds. 11 Cumulatively, the evidence suggests that the Velsen buildings were not shipsheds. The possible existence of shipsheds at Caesarea Maritima in Israel has recently been suggested by the late Avner Raban. He notes three different sets of structures that might have been sheds in or near the royal harbor of Herod the Great, which was begun around 20 B.c. These all consisted of long, narrow structures of roughly the right dimensions for small warships, which were, unusually, open at their seaward ends. 12 The first is a site in the northwest corner of the Inner Basin (area S, itself the southern part of area LL; see fig. 3). The lower, probably Herodian, level consists of "three elongated, parallel confined spaces over 40 m long and just over 5 m wide, sloping gently towards the waterfront at their south end" (fig. 4).13Raban suggests that these buildings could have housed triremes. The structures would, however, have been broad enough only for triremes without outriggers or the oarboxes that were the norm by the Roman period, while they would be rather long for the 20-m liburnians whose breadth they could have accommodated (see below). The published plan (see fig. 4) furthermore shows that the central chamber, unlike the others, narrows to around 4 m toward the upper end. Raban notes a stone platform at the southern end of the western chamber, which he suggests could have acted as a "sliding device." This platform would in fact have been wholly unsuitable for such a purpose, for which a wooden groundway running along 8

Bockius 2002.

9

Hiickmann 1993.

11 See the tables conveniently collated by Morel 1987, 244-247. 12

10

Raban 2003.

Blackman 1973, 128 with fig. 25. 13 Raban

2003, 97.

BORISRANKOV

56

Fig. 4. Caesarea,the structuresof the Herodianphase in areaLL/S (from Raban 2003, 97, fig. 15.6).

\

I I

I I

-f

l

CAESAREAMARITIMA HERODIANPHASE IN AREA LL End of 1999 Season

Orn

INNERHARBOR

10

20

30m

the whole length of the shed and down to the water would have been required. 14 Moreover, a shed that sloped only at the lower end and not throughout would have been wholly impractical , since the ship would have "tipped" as it rode over the point where the slipway began to slope, putting an intolerable load on the keel at that point. For a number of reasons, therefore, these structures are extremely unlikely to have been shipsheds. Raban's second suggested site is the vaults of the third century A.D. overlying earlier chambers, at the western end of the platform of the Herodian temple (area I; marked as TP on fig. 3). There are two groups of four of these vaults, one group either side of the staircase that runs down to the waterfront of the Inner Harbor (fig. 5). At 21 m long, 5 m wide, and 8 m high, the dimensions of these vaults are entirely suitable for a liburnian (see below), and as Raban notes are very similar to those of the known slipways at Tel Dor, some 13 km to the north. Raban indicates that the vaults, which were open -fronted at their western (seaward) ends , had a low retaining wall there, raised to 2 m over Mean Sea Level, to 1 m over the floor of the earlier chambers, and almost level with 14 Coates

and Shaw 1993.

... .,. 00

00

CAESAREA MARITIMA HERODIAN STRUC TURES IN AREASTP, Z, AND I End of 1998 Seoson

Fig. 5. Caesarea, schematic plan of area TP in the time of Herod the Great (from Raban 2003, 99, fig. 15.8).

BORISRANKOV

58

V.\I.U

I

IAJl7

2

•

VAUl1

11

-==b ::

::J

V& Ull ' 2

C

p

.•~ I

n

•

~

"' 0 0

LUIIH[

D( CUMA US

t

I

o

•

t

..

'

Fig. 6. Caesarea, plan of the Herodianphase of the western vaults in area CC (from Raban 2003, 100, fig. 15.9).

the floor of the third-century vaults . He suggests that the ramp to take the vessels within the sheds could have been built up in timber . The absence of either a natural slope or of one built up in sand or masonry, however, argues against these being shipsheds, since a timber ramp rising, as would be required, to around 2 m at the rear of each vault would be an unsuitable structure to support a vessel of around 7 tonnes. 15 The western ends of the vaults, moreover, lie some 20 m from the water's edge according to Raban's plan (see fig. 5), an unusual situation for shipsheds on the virtually tideless Mediterranean, and one that, as with the suggested sheds at Haltern, would have increased considerably the work involved in launching and slipping. Raban's final site is the vaults in the northwestern part of the South Bay (area CV and/or CC; see fig. 3 ), outside the enclosed harbor areas. These were constructed in the late first century A.D. overlying Herodian buildings, were open at their western (seaward) ends, and were around 32 m long, 6 m high, and 4.85 m, 5.45 m, 5.65 m, and 5.30 m wide respectively (fig. 6). Like the vaults of the temple platform, their lack of gradient and distance from the water's edge (ea. 50 m on Raban's plan, shown in fig. 3) make them very unlikely to be shipsheds, as does their location outside the main harbors . The case for sheds at Caesarea Maritima thus appears to be doubtful . We can now turn to the suggested shipsheds at Ostia . Michael Heinzelmann of the Deutsches Archiiologisches lnstitut and Archer Martin of the American Academy in Rome have recently identified a large, terraced structure there of the early first century A.D., which was investigated in 2000 and 2001, as the platform of a temple complex (figs. 7-8). 16 They have further suggested that the vaulting of the substructure on the west and north sides of the terrace may have doubled as navalia, or shipsheds . This revives a suggestion made by C. L. Visconti, the nineteenth-century excavator 15 Seen

. 7.

16 Heinzelmann

and Martin 2002.

59

ROMAN SHIPSHEDS AND ROMAN SHIPS

,•/ ..4;;:: ..-:~·-·..-................

,,:,· ,,,,

,,1,

''

' ,,

.., :......................................... .......... .. ...................... _,._.

_

I\

I

..

"

Fig. 7. Ostia, the suggested navaliatemple complex, with the northern vaults at the top and the western vaults to the left (from Heinzelmann and Martin 2002, I 0, fig. 5, copyright

]RA) .

.,

"" .. "

..

...

Et ·· 1! i:

11

-

111thc. excavation

-

2000/01

0

cavallon

-

10

-

p,HefY9d

D

mert>lefragments

vautt.

20

50m

of the vaulting on the north side. 17 In the early twentieth century, both Paschetto and Carcopino associated the vaulting with an inscription recording the benefactions of P. Lucilius Gamala, an Ostian grandee of the late second century A.D. 18 Gamala's inscription records that he restored a number of buildings, including a navalea L. Coilio aedificatumI extru[e]ntibus/ere collapsum(lines 25-27). As a result of their recent investigation, Heinzelmann and Martin have now gone further and connected the restoration of the navale with the restoration of an aedem Castoriset Pollucis mentioned earlier in the inscription (line 13).19 The text itself does not associate the two buildings, and although the identification of the temple as that of Castor and Pollux is entirely plausible, the identification of the vaults as the navale 17 Visconti 18

1857, 337-338; cf. Lanciani 1868, 148.

Paschetto 1912, 346-350; Carcopino 1911, 214-224.

Inscription of P. Lucilius Gamala: CIL 14.376. 19 HeinzelmannandMartin2002,

17-18.

BORISRANKOV

60

Fig.8. Ostia, reconstructed plan of the suggestednavaliatemple complex,harbor,and the mouth of the Tiber(from Heinxelmannand Martin 2002, 9,fig. 4, copyright]RA).

.,,.··

....--···

Harbour

.. N

t mentioned is less so. What is referred to in the inscription is not shipsheds-that is, a complex for housing ships, which one would expect to be in the plural (navalia)-but a yard or shed for shipbuilding (extruentibus).Moreover, as Meiggs points out, the archaic spelling of Coilius suggests that Gamala is here referring to a surviving inscription on the navaleitself.2°If so, the spelling-the oe diphthong supplants oi by the middle of the first century B.C.-and the absence of a cognomen would point to the late Republican period rather than the Tiberian date ascribed to the temple complex . This does not, however, preclude the identification of the vaulting as shipsheds, even though the idea was rejected by Meiggs and other recent scholars. 21 Heinzelmann and Martin base their suggestion on the ground plan and dimensions of the vaulting on the west side, and the fact that they found the lower end of a ramp in opus caementicium descending to water level inside one of the western vaults. They argue that the clear width of the vaults of ea. 5 m and height of ea. 4.5 m (they do not mention their length, which appears to have been ea. 22 m to judge by the published plan: see fig. 7) would have been unsuitable for large ships but could have been used for "small warships responsible for protecting the mouth of the Tiber. "22 The northern vaults, which are oriented at right angles to the stream of the Tiber (see fig. 8), should be discounted as shipsheds for the same reasons as those for rejecting the building at Haltern. Even though the temple complex was situated only some 300 m from the mouth of the Tiber when it was constructed (there has been extensive silting during and since antiquity, resulting in a retreat of the coastline by more than 4 km since A.D. 200),23 the stream would still have been strong enough to cause problems . The western vaults, however, are oriented parallel to the stream facing a harbor basin cut into the bank of the river. They are thus in an ideal position for river-based shipsheds, since as long as vessels were firmly tethered to the shore as they were launched, the stream would keep them in position . The use of the western vaults as shipsheds is therefore wholly feasible on this count. The warships epigraphically attested for the Mediterranean fleets of the Imperial Roman navy 20

Meiggs 1973, 126,494.

22

Heinzelmann and Martin 2002, 12.

21

Meiggs 1973, 126.

23

Le Gall 1953, 21-25 with figs.4-5.

ROMAN SHIPSHEDS AND ROMAN SHIPS

61

include sixes (hexereis),fives (pentereisor quinquremes),fours (tetrereisor quadriremes),threes (triereisor triremes),and two-level libumians (liburnae,also known as biremes).24 A reconstructed design for a trireme of the fourth century B.C. (revised by John Coates from his design of t]:ieOlympias reconstruction launched in 1987) is 5.6 m in width and ea. 40 m long overall. This is based in part on the space required for the operation of the 170 single-manned oars in an Athenian trireme of that period, and in part on the dimensions of the fourth-century B.c.sheds at Zea harbor in the Piraeus. 25 This reconstructed design has in tum allowed estimates to be made of the dimensions of other oared warships, similarly based on the space required within a hull for the operation of single- and double-manned oars arranged at two or three levels.26 Individual vessels will certainly have varied from these estimates, but they can only have done so within relatively narrow limits. Variations caused by developments in oar systems between the Classical, Hellenistic, Republican Roman, and Imperial Roman periods can also be estimated. These estimates suggest that fours of the fourth century B.C. would have had the same breadth, 5 .6 m, and been slightly shorter than the revised Olympiasdesign at around 37 m, while fives would have been around 6.4 m broad and 45 m long overall. Roman triremes of the first century A.D., which utilized an oarbox for the rowing system rather than outriggers, would have been of similar length to their Classical antecedents but up to 7 m broad, as would fours, while fives of the period would have been around 8.2 m broad. 27 Even allowing for the uncertainty of such estimates, we may be confident that housing the larger vessels from triremes upward in the Ostia vaults would have been out of the question. There is no direct evidence for the presence of small warships at Ostia. The epigraphic evidence cited by Meiggs consists almost entirely of funerary monuments, which merely record the presence at Ostia of members of the Misenum and Ravenna fleets.28 There is no indication of what role they performed here, which may or may not have involved the crewing of vessels. One may compare the fleet personnel based in Rome, whose recorded duties involved working the awnings in the Colosseum. 29 Some may simply have acted as ordinary troops, which is what most members of the fleet called themselves (milites).Of the thirteen men of the Misenum fleet commemorated at Ostia, three served on a trt'eres,two on a tetreres,and one on the hexeresOps,which appears to have been the flagship of the Misenum fleet.30 If they were at Ostia with their own ships, the latter were certainly not housed in the vaults of the temple complex. The men may nevertheless have been detached from their ships to serve here in the sorts of smaller vessels that could have been housed below the temple complex. Two-level, 50-oared liburnians (liburnae)would have had an overall breadth of around 3.9 m and length of around 20 m (fig. 9).31 This would be about 0.6 m narrower than the mouths of the western vaults at Ostia and 2 m shorter than their overall length on the published plans. Liburnians might therefore have fitted well within a shed with the ground plan of these vaults. The overall height 24

See Morrison 1996, 172-175 based on Miltner 1931 and Casson 1971, 356, n. 357.

27 Morrison

1996, 345 appendix D (by J. F. Coates).

Meiggs 1973, 304, citing CIL 14.110; 232-243; 44964498.

28

25

Morrison, Coates, and Rankov 2000, 132-133, 136-150, 245-246, 267-273; the clear width between the columns of the Zea shipsheds is 5.94 m, and very recent survey and excavation work by Bj0m Loven has indicated that the dry length is significantly longer than the 37 m that was taken by Morrison and Coates to have been indicated by Dragatses and Dorpfeld 1886. 26

Morrison 1996, 345 appendix D (by J. F. Coates).

29

Kienast 1966, 74-75; Redde 1986, 451-452; Starr 1993, 20-21.

Trieres:CIL 14.233; 239; 242; tetreres:CIL 14.237; 4498; hexeres(Ops): CIL 14.232; cf. CIL 10.3560; 3611.

30

31 Morrison

1996, 345 appendix D (by J. F. Coates).

BORIS RANKOV

62

..!..., I

I

CJfCOOJC$199i

Fig. 9. Reconstructionby f. F. Coatesof a two-level liburnian of the first/second centuriesA.D. (from Morrison 1996, 316, fig. 72).

of the vaults is reported as ea. 4.5 m, presumably above ground level at the water's edge. It would have been necessary, however, for the ship's keel to be slid in along a greased wooden rail, say 30 cm high, fixed along the centerline of the vault, effectively reducing the height of the opening for ships to ea. 4.2 m. All the iconographic evidence we have shows the bows of ships at the mouths of shipsheds, which in turn indicates that the ships were drawn up into the sheds stern first.n The height of the stern ornament (aphlaston)of the Olympiastrireme reconstruction is some 6.5 m above the underside of the keel, and that of a liburnian would have been around 4.5 m.33 It is thus open to question whether a liburnian fitted with an aphlastoncould even have entered the Ostia vaults . Moreover, the slipway within the vault would of course have been inclined. It is the traces of the lower end of such a ramp in the second chamber from the south that is one of the strongest indications that the vaults may have been used as shipsheds. In most shipsheds, the roofline will have risen either parallel to the slipway or in steps, but at Ostia the vaults supported a level temple platform and so could not rise toward the rear. Even if the aphlastonwere slightly lower than 4.5 m high and could have entered the mouth of the shed with its 30 cm wooden rail, it could not have been drawn up within it. If the bows came to rest at the mouth of the vault, and the gradient of the ramp inside the vault was about 1 in 20 as at, say, Munychia or the north side of the circular harbor at Carthage, 34 the stern would have been raised to ea. 5 .8 m above water level. On any estimate, the height of the aphlastonwould have made it quite impossible to slip a liburnian here. It is conceivable, however, that the aphlastonmay have been detachable, 35 especially on smaller warships such as liburnians. This 32 For the iconographic evidence, see the essay by Blackman in this volume . 33 Morrison

1996, 345 appendix D (by J. F. Coates).

34 Munychia: Blackman 1973, 128 with fig. 25. Carthage : Hurst 1994, 35 on Ramp F762. 35 Rankov

forthcoming b.

ROMAN SHIPSHEDS AND ROMAN SHIPS

63

would have reduced the overall height of such a vessel to a much more manageable 2.8 m. Without the aphlastonand on a ramp with a 1 in 20 gradient surmounted by a 30 cm wooden rail, the stern would have been ea. 4.1 m above sea level and might thus have fitted within these vaults. Without knowing the precise dimensions of a liburnian, whether its aphlaston could be detached, or the gradient of the ramp, it is impossible to say whether such a ship could have been accommodated here. We may note in this context, however, that not one of the members of the Misenum and Ravenna fleets epigraphically attested at Ostia is recorded as attached to a liburnian. A single-level ship of similar length and breadth-that is, one with 20 to 30 oars-would undoubtedly have fitted much more readily. No such ship is directly attested either at Ostia or for the Mediterranean fleets in general, but there is no need to doubt their existence. Certainly such vessels were used in the riverine fleets on the northern frontiers of the empire. Wrecks of two such vessels of the late first/early second century have been found at Oberstimm in Upper Germany, and the wrecks of five more of the late fourth century at Mainz. 36 The thirty-oared reconstruction of the Mainz vessels in the Museum fiir Antike Schiffahrt at Mainz is 21.6 m long, 2.79 m broad, and ea. 1.9 m high with its aphlastonand 1.1 m without. 37 Although vessels built in Italy for possible coastal use would have differed significantly in detail from the Mainz ships, the overall dimensions are likely to have been similar. With an aphlaston,the stern of the Mainz reconstruction would have stood ea. 4.8 m above water level at the rear of the vault on a steep 1 in 7 gradient 38 surmounted by a 30 cm wooden rail, and ea. 3.5 m above water level without. For a 1 in 20 gradient, the figures would have been ea. 3.2 m and ea. 2.4 m respectively. The Oberstimm and Mainz vessels were too low and narrow to have been used as fighting platforms but may well have been used for scouting up and down rivers and for ferrying troops. 39 Single-level oared vessels that could have fitted into the Ostia vaults would similarly have been of little use for defense from pirates, say, unlike liburnians, but could certainly have been used for policing or for ferrying dignitaries on the Tiber or along the coast. When Cn. Calpurnius Piso was summoned back to Rome from Syria in A.D. 20 to face charges of sedition and having murdered the heir to the throne, Germanicus, he made the last part of his journey by boat from Narnia down the rivers Nera (Nar) and Tiber, alighting in Rome near the Mausoleum of Augustus. 40 Despite these findings, a number of problems remain for our acceptance of the Ostia vaults as shipsheds. Even if two-level, or more likely single-level, ships could have fitted within the vaults, there are several features that might have made these less than ideal for such use. In the Zea sheds in the Piraeus, the individual columns in alternate colonnades have an interaxial spacing of 3 .38 m and 2.16 m respectively,41 suggesting that alternate colonnades had to support a much greater weight than the others. This in turn suggests that the Zea sheds were roofed in pairs, and that more closely spaced colonnades were intended to support the valleys with their guttering probably lined with lead. One of the prime functions of a shipshed would have been to protect the ships within both from the sun and from heavy downpours of rain, which would have to be cleared by the guttering. Performing the latter function would clearly have been difficult at Ostia with what was in effect a large flat roof, even though the vaulting was protected by a layer of impermeable mortar. Certainly, the platform would have had to be very well supplied with drainage systems. 36

Oberstimm: Bockius 2002; Mainz: Hockmann 1993.

39 Hockmann

37

Pferdehirt 1995, 7; 19, fig.17.

40

Tac. Ann. 3.95.

41

Dragatses and Dorpfeld 1886; Blackman 1968.

38 As, for

instance, at the small slipway lying north of the main slipway complex at Sounion: Kenny 1947, 200, fig. 2.

1989; 1993; Rankov forthcoming a.

BORISRANKOV

64

Conversely, shipsheds had to be very well ventilated, to ensure that the hulls dried out and to prevent fungal rot such as that which has previously damaged the planking of the Olympias reconstruction. This would have been much easier with a tiled roof, where louver tiles could have been employed or where the sloping roof could have been stepped to let in light and air, as may have been the case at Carthage. 42 In most shipsheds, internal circulation of air would have been facilitated by the use of colonnades or pillars, which would also have improved lighting and allowed easy movement of personnel within the sheds. The plans of the northern vaults at Ostia show unbroken walls, however, and they must have been rather damp, dark, inaccessible places. The situation on the west is unclear in the current state of investigation, but the walls of the vaults do not appear to be any thicker than on the north, which one might have expected if they were pierced by passageways or actually consisted of a line of pillars as at Carthage or Kition. Wholly enclosed, barrel-vaulted shipsheds cannot be ruled out as impracticable, and ancient workers were probably used to operating in lighting conditions that we would regard as unacceptable, but the conditions they would provide for the housing of wooden ships would certainly have been inferior to any known ancient shipshed. There is one further problem with the identification of the Ostia vaults as shipsheds, concerning the practicality of slipping and of launching ships from them. Ancient shipsheds on the coastlines of the Mediterranean could rely on relatively small fluctuations in tide levels, mostly around 30 cm or so. This meant that sheds could be built at the water's edge in complete confidence that they would not normally be flooded or alternatively be left high and dry. River levels, however, as Morel recognized at Haltem, rise and fall considerably with the seasons. The flooding of the Tiber was a frequent and often disastrous occurrence in ancient Rome, and in the Imperial period officials were regularly appointed to deal with flood defenses in the city. As already noted, when it was built, the temple platform was situated only about 300 m from the mouth of the river,43 where the effects of flooding-normally in the winter months-would have been very much less severe than in the region of the city of Rome but might in extreme cases have been destructive of any ships housed within the vaults. Evidence of weathering by waves on the southwest comer block of the complex suggests that normal water level was around the ground level at the mouth of the vaults, so the water did not normally extend far into them. But if the flow of water in the Tiber became such that the height of the river was raised by about a meter at the mouth of the vaults, the bows of any ships within would have begun to float off, with the danger of their being damaged against the sides of the vaults. Silting at the mouth of the river would, furthermore, have over time greatly increased the chances of this happening. Such damage could have been prevented either by removal of the ships from the vault before the water rose too high or by ballasting the bows to hold them firmly in place. In the latter case, launching would probably have been impossible while the river remained in flood. Thus, although the interpretation of the vaults as shipsheds remains possible, there are a number of potential practical objections that must be taken into account. 44 The conclusions of this paper may appear to be disappointingly negative: Haltem, the Western, and probably also the Eastern Boat Houses at Velsen rejected, all three of the Caesarea sites unlikely, and Ostia possible but problematic. The real purpose of the paper is, however, not to reject or accept particular identifications but to demonstrate that in dealing with potential shipsheds, the normal archaeological approach of comparing ground plans is too simplistic. Unlike buildings designed for 42

Hurst 1979, 28-32; 1994, 33.

43 See fig. 8

and n. 23.

44 None

of the above of course precludes the possibility of storing small fishing boats or the like in either the western or the northern vaults.

ROMAN SHIPSHEDS AND ROMAN SHIPS

65

people, in which the emphasis is on the provision of appropriate space, and within which activity is relatively small-scale and static, shipsheds may be characterized as highly "dynamic" buildings. Not only did they house and preserve some of the largest mobile objects in the ancient world, but they had to be designed to enable those objects to be moved into and out of them on a r~gular basis. They also had to facilitate the transition of those objects from water to dry land and back again, in the course of which, because of the length of the objects involved, one end might be subject to wholly different forces and stresses than the other. Because of this, shipsheds cannot be investigated or identified without reference to the ships they housed or to the fairly complex practicalities of launching, slipping, and housing them. 45

These considerations underpin the methodology of a current researchproject on Shipshedsin the Ancient Mediterranean, which is funded by the Leverhulme Trust. The project is being undertaken by the author in conjunction

45

with David Blackmanof the Universityof Oxford, Kalliopi Baika,Judith McKenzie,Bj0m Loven, andJari Pakkanen of RoyalHolloway,Universityof London, and Henrik Gerding of Lund University,Sweden.

66

BORISRANKOV

WorksCited Blackman, D. J., "The Shipsheds," in Greek OaredShips 900-322 B.C, ed. J. S. Morrison and R T. Williams (Cambridge 1968) 181-186. --, "Evidence of Sea Level Change in Ancient Harbours and Coastal Installations," in MarineArchaeology,ed. D.J. Blackman (Bristol and London 1973) 115-137. Colston Papers 23. --, "Naval Installations," in The Age of the Galley:Medite"anean OaredVesselssincePre-Classical Times, ed. R Gardiner andJ. S. Morrison (London 1995) 224-233. Conway's History of the Ship. Blackman, D. J., P. Knoblauch, and A. Yiannikouri, "Die Schiffshiiuser am Mandrakihafen in Rhodos," Archiiologischer Anzeiger (1996) 371-476. Bockius, R, Die romerzeitlichenSchi/fs/undevon Oberstimm(Bayern)(Mainz 2002). Carcopino, J., "Ostiensia," Melanges d'Archeologieet d'Histoire de /'Ecole Franfaisede Rome 31 (1911) 144-230. Casson, L., Shipsand Seamanshipin the Ancient World(Princeton 1971). Coates,}. F., andJ. T. Shaw, "Hauling a Trireme up a Slipway and up a Beach," in The TriremeProject.Operational Experience1987-90. LessonsLearnt, ed. T. Shaw (Oxford 1993) 87-90. Oxbow Monograph 31. Dragatses, I. C., and W. Dorpfeld, "Ekthesis peri ton en Peiraiei anaskaphon (Report of the Excavations in Peiraeus)," Praktikaof 1885 (1886) 63-68. Dragendorff, H., F. Koepp, and E. Kruger," Ausgrabungen bei Haltem. Das Uferkastell 1903 u. 1904," MitteilungenderAltertumskommision/ur West/a/en4 (Munster 1905) 33-79. Heinzelmann, M., and A. Martin, "River Port, Navaliaand Harbour Temple at Ostia: New Results of a DAIAAR Project," Journalof Roman Archaeology15 (2002)5-19. Hackmann, 0., "Romische Schiffsfunde westlich des Kastells Oberstimm," BerichtderRomisch-Germanischen Kommission70 (1989) 321-350. --, "Late Roman Rhine Vessels from Mainz," InternationalJournalof NauticalArchaeology22.2 (1993) 125-135. Hurst, H. R., "Excavations at Carthage 1977-8: Fourth Interim Report," Antiquaries Journal 59 (1979) 19-49. --, Excavationsat Carthage.The BritishMission.Vol. 2.1: The CircularHarbour,North Side. The Site and Finds Other than Pottery(Oxford 1994). British Monographs in Archaeology 4. Kantzia, K., "}\vaaxacp71Kw (Excavation at Kos)," ArkhaiologikonDeltion 42 (1987) 632-635. Chronika B2. Kenny, E., "The Ancient Docks on the Promontory of Sounion," Annual of the BritishSchoolat Athens 42 (1947) 194-200. Kienast, D., Untersuchungenzu den Kriegs/lottender romischenKaiserzeit(Bonn 1966). Antiquitas Reihe 1, vol. 13. Lanciani, R A., "Ricerche topografiche sulla Citta di Porto," Annali dell' Istituto di Co"espondenzaArcheologica40 (1868) 144-195. Le Gall, J., Le Tibre,/leuve de Rome dans l'antiquite (Paris 1953). Meiggs, R., Roman Ostia,2nd ed. (Oxford 1973). Miltner, F., "Seewesen," RE suppl. 5 (Stuttgart 1931) 906--962. Morel, J.-M. A. W., "The Early Roman Defended Harbours at Velsen, North-Holland," in Studien zu den MilitiirgrenzenRoms III. 13 InternationalerLimeskongre/SAalen 1983, ed. C. Unz (Stuttgart 1986) 200-212. --, "Friihromische Schiffshaiiser in Haltem, Hofestatt," Ausgrabungenund Funde in Westfalen-Lippe (Munster)5 (1987) 221-249. Morrison, J. S., with contributions by J. F. Coates, Greek and Roman Warships399-30 B.C. (Oxford 1996). Morrison, J. S., J. F. Coates, and N. B. Rankov, The Athenian Trireme:The History and Reconstructionof an Ancient Greek Warship,2nd ed. (Cambridge 2000). Morrison,]. S., and RT. Williams, eds., Greek OaredShips 900-322 B.C. (Cambridge 1968).

ROMANSHIPSHEDSANDROMANSHIPS

67

Paschetto, L., Ostia ColoniaRomana,storiae monumenti (Rome 1912). Fferdehirt, B., DasMuseumfur Antike Schi/fahrt,vol. 1 (Mainz 1995). Raban, A.," Ancient Slipways and Shipsheds on the Israeli Coast of the Mediterranean," in Boats,Ships and

Shipyards.Proceedingsof the Ninth InternationalSymposiumon Boat and ShipArchaeology,Venice2000, ed. C. Beltrame (Oxford 2003) 91-102. Rankov, N. B., "Ancient Oar-rigs: The Evidence of the Mainz Roman Ships," in TropisVIII. 8th International Symposiumon Ship-Constructionin Antiquity, Hydra2002 Proceedings,ed. H. Tzalas (Athens forthcoming a). --, "Ship Sheds and Stern Ornaments," TropisIX. 9th InternationalSymposiumon Ship-Constructionin Antiquity, Ayia Napa2005 Proceedings,ed. 1:J.Tzalas (Athens forthcoming b). Redde, M., MareNostrum (Rome 1986). Sears, J.M., "Oeniadae. The Ship-Sheds," AmericanJoumal of Archaeology7 (1904) 227-23 7. Starr, C. G., The Roman ImperialNavy 31 B.C.-A.D. 324, 3rd ed. (Chicago 1993). Archeologica Visconti, L., "Escavazioni di Ostia dall' anno 1855 al 1858," Annali dell'Istituto di Correspondenza 29 (1857) 281-340. Wells, C. M., The GermanPolicyof Augustus (Oxford 1972).

ROME'S CONTRIBUTION TO THE DEVELOPMENT OF PIRACY

Philip de Souza

1. Introduction

T

he image of pirates as violent, flamboyant, seafaring outlaws is now a very familiar one. It is most obviously associated by modern audiences with the Barbary Corsairs, or the eighteenth-century Anglo-American pirates. 1 This paper argues that this familiar image of piracy is, to a considerable extent, a Roman creation, having initially been developed in the period ea. 100 B.C.-A.D. 200. This is not to say that the Romans were leading practitioners of piracy, nor that they were particularly zealous in their suppression of pirates. The argument is rather that Romans took the idea of piracy and fashioned it into a flexible, pejorative label, which they used for political purposes. Roman campaigns against maritime enemies were presented as the suppression of piracy because that suited contemporary political needs, especially when the Roman aristocracy wanted to convince reluctant allies that they should fight with or for the Roman cause. I also argue that this politically inspired image took root in literary accounts because it suited a view of Roman history and of the Roman world that envisaged Rome as the suppressor of piracy and guarantor of maritime security for the civilized Mediterranean world. I define piracy very broadly as "armed robbery normally involving the use of ships." Pirates are sea-borne raiders who rob, kidnap, murder, and destroy. Piracy is similar to banditry, but the maritime aspect makes piracy distinctive in several ways. The use of ships differentiates the modus operandiof pirates from that of bandits by giving pirates the ability to attack targets over long dis. tances, often with the element of surprise. Maritime armed robbery requires a considerable degree of economic organization, and the potential impact of piracy can also be much greater than that of mere banditry. There is an obvious similarity between piracy, thus defined, and warfare, particularly the kind of raiding and plundering warfare that, as Vincent Gabrielsen has recently argued, was a common practice in the ancient world. 2 He argues the case for the existence of what he calls the "raid mentality," a widespread assumption of the right to practice violent acquisition of persons, property, and The first version of this article was presented at the "Maritime World of Ancient Rome" conference, held at the American Academy in Rome in March 2003. I am grateful to Bob Hohlfelder for the invitation and organization, to the Academy for its excellent hospitality, and to my fellow participants for their helpful comments, especially Anna McCann, Boris Rankov, Nick Rauh,and Geoffrey Rickman. A second version was given at the University of Southern California Classics Department's "Admit the Outlaw" conference in February 2004. Thanks to everyone for outstanding hospitality and stimulating discussions, especially Clifford Ando, Del Chrol,

Vincent Farrenga, and Amy Richlin. As always my deepest debt of gratitude is to my wife, Debra Birch. 1 See Heers 2

2003; Earle 2003; Rediker 1987.

Gabrielsen 2001. See also Gabrielsen 2003. Piracy is accorded some prominence in the recent study of Mediterranean history by Peregrine Horden and Nicholas Purcell, who note that piracy and warfare can be hard to distinguish from each other and also consider piracy as a form of trade: Horden and Purcell 2000, 153-160, 387-388.

PHILIP DE SOUZA

72

territory. He sees this as being increasingly disputed in the Classical and Hellenistic periods, with some political communities asserting their right to do so and denying that right to others who clung to "the old ways." These different attitudes to violent acquisition contributed to the development of a distinction between legitimate warfare and illegitimate piracy. It was a gradual development, however, and in a world where most states routinely used raiding tactics in warfare, the difference between piracy and warfare remained to a considerable extent a matter of subjective attribution. 3 What one observer may have chosen to call warfare another could label as piracy. Gabrielsen focuses in particular on the Aegean in the Hellenistic period, but his arguments have a general application in the ancient Mediterranean. Gabrielsen also suggests that raiding and piracy were less destructive to maritime trade and the maritime economy than might at first be thought. Nevertheless, the actual or perceived danger was a significant threat to the security of coastal communities. It could also hamper maritime trade and thus harm the prosperity of all communities that depend on such trade. This threat of piracy, whether it was openly or only potentially manifest, could in tum be used as a justification for the use of violence to suppress those held to be responsible. For modem historians of antiquity, however, it is very difficult to determine whether acts of piracy by one side and the suppression of piracy by another were the declared, or even undeclared but still significant, reasons for military campaigns at the time. This paper argues that most of the historical narratives covering the late Republican period that are available to us, written in many cases long after the events they recount, are heavily influenced by an implicit rejection of the raid mentality and the adoption of a moral and political justification of warfare as a legitimate means of suppressing illegitimate acts of piracy.

2. The EarlyHistoryof Piracy The earliest use of a word for a sea-borne raider that may be translated as "pirate" occurs in the Homeric poems. That word is leistes,used in several variations in both the Iliad and the Odyssey. The noun is used for one of the likely activities that seafarers might be engaged in, according to the formulaic greeting employed twice in the Odysseyand once in the Hymn to Apollo: 0 strangers, who are you? From where have you come along the sea lanes? Are you traveling for trade, or are you just roaming about like pirates (leisteres), who risk body and soul bringing harm to other people? (Od. 3.71-74; 9.252-255; Hom. Hym. Apollo 452-455)

The word clearly has connotations of evil and moral disapproval, but piracy in a general sense is not necessarily an unacceptable or illicit activity in the sociopolitical world of the Homeric epics. It is indicative of the moral uncertainty surrounding piracy that the term leistesis never directly used to describe any of the heroic basileeswho are the protagonists in the Homeric poems. The concept of piracy as something different to warfare had not really emerged at this time. The type of raids conducted by the warrior chieftains of Homeric society are best exemplified by descriptions that Odysseus gives of his activities when he assumes the persona of a Cretan aristocrat, the son of Kastor, on his return to Ithaca: Farming I never cared for, nor life at home, nor fathering fair children. I revded in long ships with oars; I loved polished lances, arrows in the skirmish, the shapes of doom that others shake 3

See de Souza 1999, 11-12, 26-36, 80-82.

ROME'S CONTRIBUTION TO THE DEVELOPMENT OF PIRACY

73

to see. Carnage suited me; heaven put those things in me somehow. Each to his own pleasure! Before we young Achaians shipped for Troy I led men on nine cruises in ships to raid strange coasts, and had great luck, taking rich spoils on the spot, and even more in the division. So my house grew prosperous, my standing therefore high among the Cretans. (Od. 14.222-234) But Zeus the son of Kronos brought me down. No telling why he would have it, but he made me go to Egypt with a company of pirates (/eisteres)-a long sail to the South-for my undoing. Up the broad Nile and into the riverbank I brought my dipping squadron. There, indeed, I told the men to stand guard at the ships; I sent patrols out--out to rising ground; but reckless greed carried my crews away to plunder the Egyptian farms; they bore off wives and children, killed what men they found. (Od. 17.424-433)

It is likely that this kind of roving band of armed seafarers was commonplace in the Archaic world. For example, Thucydides suggests that the site of the city of Zankle (later called Messana) in northeastern Sicily, founded around 730 B.C., was chosen for the opportunities it afforded the settlers from nearby Kyme to practice piracy (Thuc. 6.4.5). Herodotus says that the Phokaians who had settled in Corsica in the mid-sixth century plundered their Etruscan and Carthaginian neighbors, provoking them to war (Hdt. 1.66).Similarly the earliest treaty between Rome and Carthage, dating, probably, to 507 B.c., envisages piracy as a normal activity for seafarers traveling between Italy and North Africa (Polyb. 3.22).4 It seems to have become commonplace to refer to rival sea-borne raiders as pirates, even if their practices did not differ from those of the people who were using the term. Polykrates of Samas acquired a reputation as a pirate leader as the result of his sea-borne raids in the latter part of the sixth century B.C. (Hdt. 3.39). The motif of Etruscan or Tyrrhenian pirates is a familiar one in Archaic Greek literature and art.5

3. Piracyand Politics Piracy continued to be a regular maritime activity and a significant hazard for seafarers and coastal dwellers in the Classical and Hellenistic periods. Literary and epigraphic sources indicate that, as political entities became larger and their relationships more complex, a distinction began to emerge between what might be categorized as warfare and what might be deemed piracy.6 Nevertheless, especially in the eastern Mediterranean, accusations of piracy or supporting pirates were often used in political invective. The aim of these accusations was to delegitimize one's opponents by labeling them as pirates, or the supporters of piracy, and to legitimize or justify one's own actions by claiming to be suppressing such pirates. For example, the Athenians and King Philip II of Macedon frequently accused each other of practicing or encouraging piracy in the mid-fourth century B.C. The lack of a clear distinction between warfare and piracy, which made such accusations so effective, can be seen in the indignant letter of 340 B.C. that Philip addressed to the Athenians: He [Kallias of Euboia] captured and sold as slaves all those sailing to Macedonia, treating them as enemies. And you decreed him a vote of thanks! So I find it hard to imagine how things could be worse if you were actually to declare war on me. For when we clearly had our differences

4

On the dating, see Cornell 1995, 210-214.

De Souza 1999, 17-26; on Etruscans/Tyrrhenians, see Giuffrida Ientile 1983 and the discussion of the myth of

5

Tyrrhenian piracy in Bakhuizen 1988. 6

For details, see de Souza 1999, chap. 2.

74

PHILIP DE SOUZA

you also used to send out pirates and makes slaves of those sailing to us, you helped my enemies and harmed my territory. ([Demos.] 12.5)

While one side claimed such activities were acts of war, the other side labeled them piracy. To a considerable extent it was simply a matter of the dominant perspective. That is not to suggest that there were not those to whom the term "pirates" could be applied without any political overtones, but pejorative use of the terms "pirate" and "piracy" was widespread in the unstable, volatile, and political context of the ancient Mediterranean. Hellenistic states and rulers used the problem of piracy as a means of exerting political influence and achieving both prestige and economic gains. One example is Eumelos, king of the Kimmerian Bosporos, of whom Diodorus Siculus records: On behalf of those who sailed to the Pontos he waged war on the barbarian tribes who were accustomed to plunder them, the Heniochoi and the Tauroi as well as the Achaians. Thus he cleared the sea of pirates, so that not only within his own kingdom, but throughout nearly the whole world, his magnanimity was proclaimed by the merchants, enabling him to receive the very finest of rewards for his good deeds-praise. (Diod. 20.25) 7

The political value of the claim to be suppressing piracy is clearly demonstrated in the history of relations between the Rhodians and the Cretans toward the end of the third century B.C. The relationship between Rhodes and the Cretan cities had been a complex one since the end of the fourth century B.C., when the Knossians lent assistance to the Rhodians in their struggle against the Antigonids (Diod. 20.88.9). It made good economic and strategic sense for the Rhodians to make allies among the numerous small poleisof Crete, but, from time to time, such alliances also caused the Rhodians to become involved in the vicious interstate rivalries that were characteristic of Cretan politics in the Hellenistic period. 8 In 220 B.C. a strong naval force from Rhodes intervened in the conflict between Knossos and ¼ttos on the side of the Knossians. Its commander, Polemokles, was accused of ordering the death of a citizen of Eleuthernai in order to please his allies. This prompted the Eleuthernaians to declare war against the Rhodians and launch attacks ~ reprisal, encouraging their allies to do likewise (Polyb. 4.53.1-2). In 205 B.C. the Rhodians went on the offensive again, embarking on a war that involved several Cretan cities, including Eleuthernai, Gortyn, and Hierapytna, and, through the agency of the Aitolian mercenary commander Dikaiarchos, King Philip V of Macedon. This First Cretan War was mainly a long series of maritime raids and counter-raids, with the Cycladic Islands and other vulnerable coastal areas bearing the brunt of the conflict. Dikaiarchos's activities, as described by the somewhat hostile source Polybius, highlight again the lack of a clear distinction between warfare and piracy at this time. The Aitolian "pirate," as Polybius calls him (Polyb. 18.54), used the twenty warships that Philip V had furnished him with to attack merchant shipping and raid the coasts and islands. Ostensibly he was assisting the Cretans against Rhodes by targeting Rhodian interests and allies, but his plundering activities may have been primarily directed at extorting funds for Philip V's military treasury. Whether his forces were viewed as a flotilla of the Macedonian royal navy or that of a pirate chief was probably a moot point. The political dimension to his actions might make it reasonable to designate them as naval warfare, but from the perspective of his victims, piracy must surely have seemed an appropriate label. 7

Diodoros goes on to say that Eumdos conquered many neighboring territories and would have attempted to rule the entire Pontic region but for his untimdy death in a chariot accident. For further examples of Hellenistic claims to sup-

press piracy, see de Souza 1999, chap. 3. 8

See Willetts 1955, 234-241 on the "social wars" of the Cretans.

ROME'S CONTRIBUTION TO THE DEVELOPMENT OF PIRACY

75

The course of the war is impossible to establish in any detail, but Vincent Gabrielsen has highlighted the extent to which it had positive results from the Rhodian perspective. 9 The main evidence of the Rhodian gains lies in the text of an inscription recording a treaty concluded around 200 B.C. between Rhodes and Hierapytna, one of the dominant cities of eastern Crete. Similar terms are found in two less complete inscriptions recording treaties concluded between Rhodes and the smaller cities of Olous and Chersonisos. 10 By the terms of these treaties the Cretans acknowledged the political influence of the Rhodians and agreed to render extensive military assistance to the Rhodians, including use of their naval facilities by the Rhodians. 11 Some important clauses in the treaty with Hierapytna refer to piracy: And if pirates establish bases in Crete and the Rhodians wage war at sea against the pirates or those who provide shelter or assistance to them, the Hierapytnians shall take part in the operations by land and sea with all possible strength and at their own expense. The pirates who are captured shall be handed over to the Rhodians together with their ships, while each of the allies shall take half of the rest (of the booty).12 And if during a campaign which the Hierapytnians are waging with the Rhodians to destroy a pirate base, any of those who provided shelter or assistance to the pirates wage war on the Hierapytnians because of this campaign, the Rhodians shall come to the help of the Hierapytnians with all possible strength, and anyone who acts in this way shall be an enemy of the Rhodians. 13

In these clauses we can see that the Rhodians are framing their new political relationship in terms of the recruitment of the Hierapytnians (and presumably other Cretans) as allies in a war against pirates. The identity of these "pirates" can, of course, be determined by the Rhodians, who are effectively dividing the communities of Crete into two groups, those who are their allies and those who are not, classifying the latter as pirates. In another clause of the treaty with Hierapytna the Rhodians agree to support their allies against attempts to overthrow their democratic polity or to "deprive them of their legitimate revenues from the sea." 14 The implication would seem to be that it is the Rhodians' prerogative to determine which revenues are legitimate and which are derived from piracy. As the other clauses quoted above show, plundering of defeated enemies by the Rhodians and their allies was a legitimate method of obtaining revenues. There was a wider audience for such declarations than just the immediate parties to the agreements. By the mid-third century B.C. the Rhodians had assumed a leading role in maritime matters. This naval preeminence was what caused other states to invite them to take the lead in a war against the city of Byzantion in 220 B.C., the aim of which was to force the Byzantines to stop imposing tolls on all exports from the Black Sea (Polyb. 4.47-53 ). The Rhodians mobilized forces of their own and various allies but managed to bring the Byzantines into line without actual conflict. 15 The following year they sent ships to lliyria in response to a piratical expedition mounted against the Cyclades and places on the Greek mainland by Demetrios of Pharos (Polyb. 4.16.6-8; 19.7-9). According 9

Gabrielsen 1997, 53-56.

10 Hierapytna and Olous: SdA 3.551-552; Chersonesos: Chaniotis 1991, 258-260.

There were, of course, some obligations on the part of the Rhodians, but it is clear that Rhodes was the dominant party in these treaties. See further Gabrielsen 1997, 53-55. 11

12 SdA 3.551.51-58;

trans. Austin 1981, 165-168.

13 SdA 3.551.80-82;

trans. Austin 1981, 165-168.

14 SdA 3.551.67-(,8; trans. Austin 1981, 165-168. The Greek text reads: ei de tis ka tasapotou dikaiouginomenaspothodous ek thalassasparairetai. 15 See Gabrielsen

1997,44-46 for discussion of this episode.

PHILIP DE SOUZA

76

to Polybius these events occurred about the same time as the dispatch of the squadron of Rhodian ships to Crete at the invitation of the Knossians, as mentioned above. The Rhodians encouraged other Greeks to see Rhodes as their naval protector. By virtue of this role they claimed the right to intervene with their naval forces in order to suppress what they deemed to be acts of piracy. In effect they were using the suppression of piracy as a justification for making war. They lacked the strength to do so unilaterally and were dependent upon the resources of a network of alliances in which they were often the senior partners, and which might be maintained by the threat or application of force against junior partners but which certainly did not amount to a maritime empire. These alliances included the Nesiotic League, leadership of which the Rhodians exercised in the first half of the second century B.C., but extended further, to include places like Hierapytna and Olous. 16 There was, understandably, an economic imperative at work behind the Rhodians' policy of maritime hegemony. As islanders dependent upon maritime trade they needed secure trade routes and access to the harbors of their main trading partners. The capacity to project their military power overseas also gave them an advantage in the commercial sphere. Rhodes's continued prominence as a commercial entrepot went hand in hand with its political status. Gabrielsen has characterized the Rhodian hegemony as a "naval protection racket," similar to that operated on a smaller scale by Athenian naval commanders in the fifth and fourth centuries B.C. 17 In such circumstances it worked to their advantage to be able to present themselves as the selfless leaders in an enterprise of mutual benefit, combating piracy on behalf of all seafarers and coastal communities. The image of this selfappointed hegemony is exemplified by three Rhodian trierarchs and a navarch who were honored in the text of a mid-third-century B.C. inscription from the island of Delos. The inscription states that the men have been "appointed by the people of the Rhodians for the protection of the islands and the safety of the Greeks. " 18 Later historians such as Diodorus Siculus and Strabo, who may well be drawing upon Rhodian sources, praised the Rhodians for their policing actions and for clearing the seas of pirates to the general benefit of the Greeks (Diod. 20.81.3; Str. 14.2.5). 19 The maritime situation began to change for the Rhodians in the second century B.C., with the increasing prominence of Rome in the eastern Mediterranean. Initially the Rhodians benefited from their status as allies of the Romans in a series of campaigns that culminated in the defeat of Antiochos III in 189 B.C. They were rewarded with increased possessions on the Anatolian mainland and substantial trading privileges. Controlling these possessions proved difficult, however, and an ill-timed attempt to arbitrate between Rome and the Antigonid monarch Perseus in 168 B.C. resulted in a stark reversal in fortunes for the island state. Roman favor was withdrawn, mainland possessions seceded, maritime trading partners became less cooperative, and Rhodian revenues declined. This decline was hastened by the declaration of Delos as a free port in 166 B.c.20 A diminution of political influence went hand in hand with the loss of economic power. The Rhodians' diminished status was humiliatingly demonstrated in the Second Cretan War of 155-153 B.C., in which Cretan forces sacked the islands of Siphnos and Karpathos (Diod. 31.43-45; Polyb. 33.17; JG 11.1.1033). Roman intervention to end the conflict aptly illustrated the extent to which Rome had become the dominant maritime power.21 16 For detailed discussion of the evidence relating to Rhodian maritime hegemony, see Gabrielsen 1997, chap. 2. On the enforcement methods of the Rhodians, see Gabrielsen 2001, 228-237.

17 Gabrielsen 2001. On the Athenians, see alsode Souza 1999, 31-36. For a suggestive comparison, see Katele 1988. 18 JG 11.4.596; Durrbach

1921, no. 39.

19 See

de Souza 1999, 48-53; Gabrielsen 1997, 42-63.

See Rauh 2003, 65-75. For arguments against too simplistic an interpretation of Rhodian commercial decline, see Gabrielsen 1997, 64-71.

20

21

See Brule 1978, 61-66.

ROME'S CONTRIBUTION TO THE DEVELOPMENT OF PIRACY

77

4. The Rise of Roman MaritimePower The Romans may have justified their conquest and colonization of coastal areas in central Italy in a similar fashion, by claiming that such actions were prompted by the need to suppress pirates or to establish protection from piratical attacks. The Roman historical tradition presented the imposition of a Roman citizen colony on the Latin city of Antium in 338 B.C. as an antipiratical measure (Livy 8.14; Dion. Hal. 7.37.3; 9.56.5; Str. 5.3.5). Rome's rise as a maritime power was, however, principally the result of the struggle with Carthage for dominance in the western Mediterranean in the third century B.C. 22 Like the Athenians and the Rhodians, they were not averse to using their naval forces to extort money, as was done at Kerkina in 217 B.C. by the naval prefect Gaius Servilius, who was operating between Sicily and the coast of North Africa with a fleet of 120 warships (Polyb. 3.96; Livy 22.31). In 229 B.C. the Romans projected their military power against the Greek mainland for the first time, defeating the illyrians in a swift campaign. Roman intervention in illyria seems to have been prompted by concerns for the security and freedom of movement for Italian traders and for the stability of a strategically significant area. It was encouraged by requests from the Greek cities and tribes in the region, who were alarmed by what one modern scholar has characterized as "large-scale raids for booty and incipient imperialism, "23 This First illyrian War furnishes the earliest example of an historian portraying the Romans as suppressors of piracy for the benefit of other states. That historian is Polybius, writing in the second half of the second century B.C. His account of the war may in part have been influenced by later lliyrian raids against the Achaians. 24 While he says little about the strategic implications of the war, Polybius makes it clear that in his view the Romans were protecting the interests of traders and the coastal cities of western and southern Greece by tackling the piratical illyrians, whom he describes as "the common enemies of all peoples" (Polyb. 2.12.6). 25 It is possible that Romans were already using the justification of a response to piracy for their military campaigns by the time that Polybius wrote in the second half of the second century B.C. Certainly Livy presents campaigns against Liguria and Histria in 181 B.C. as antipiracy measures, responding to complaints from the cities of Tarentum, Brundisium, and Massilia (Livy 40.18). 26 A similar reason seems to have been put forward for Roman aggression against the late second-century B.C. inhabitants of the Balearic Islands, whom the Romans attacked in 123 B.C. under the command of the consul Quintus Caecilius Metellus. The Baleares, mostly armed only with slings, were driven into the hills "like bellowing cattle," according to the Livian version (Livy Per.60), and were virtually eradicated in a search and destroy operation. Metellus then settled 3,000 Roman veterans on Mallorca and went back to Spain. On his return to Rome in 121 B.C. he celebrated a triumph and took the surname Balearicus in commemoration of his conquest. The justification for this ruthless campaign is given as the suppression of piracy by the writers in the Livian tradition (Livy Per. 60; Flor. 1.43.1-2; Oros. 5.13.1), but Strabo, writing in the early first century A.D., hints that the accusation may have been simply an excuse to justify an act of aggressive imperialism: 22 On the growth of Roman naval power, see de Souza forthcoming.

Dell 1967, 358. For more detailed discussion of the sources and scholarship on the lliyrian War, see de Souza 1999, 76-80.

25 On the Romans' deliberate promotion of an image of themselves as benefactors, see Erskine 1994; Avidov 1997, 41-43.

23

24

On Polybius's presentation of Achaian enemies, see de Souza 1999, 73-80; Eckstein 1995.

26 The

conquest of these regions only began in earnest in 181 an imperialist drive that, in the case ofLiguria, is linked to greed for land by Harris 1989, 114-118. B.C.,

78

PHILIP DE SOUZA

On account of the dealings between a few of their worst dements and the pirates, they were all falsdy accused, and Metellus, who was surnamed Balearicus, came against them. He it was who founded the cities. (Str. 3.5.1)

It may be that Metellus at least wanted to present his expedition as an act of suppression of piracy, although the consequences were, perhaps, far worse than the insignificant islanders deserved. 27

5. Rome and the Problemof Piracyin the FirstCenturyB.C. It remains unclear to what extent the suppression of piracy was the declared aim at the time of the Roman operations discussed above. But there can be no doubt that piracy became the principal theme of relations between Rome and the maritime communities of the eastern Mediterranean at the end of the second century B.C. In 102 B.C. Marcus Antonius the Orator, a praetor in 103 B.C., attacked some of the coastal communities of southern Anatolia. 28 The forces at his disposal were mainly drawn from Rome's eastern Mediterranean allies, including Rhodes and Byzantion.29 His campaign is presented in the Roman annalistic tradition as a measure to suppress piracy (Livy Per. 68; Obseq. Prodig.44), but it is difficult to establish from the meager sources exactly what he did. His naval squadrons may have conducted some raids against towns and cities on the coast of Cilicia from a base at Side in Pamphylia. At least one of his subordinate commanders, the prefect Marcus Gratidius, an uncle of Cicero, was killed in action (Cic. Brut. 168; Leg. 3.36). The scope of the expedition seems to have been fairly limited because Antonius withdrew from the area within the year and returned to Rome, where he celebrated a triumph and was eventually elected consul for 99 B.C. The clearest articulation of the Romans' policy of opposing pirates comes in the celebrated lex de provinciispraetoriisof 100 B.C. (formerly, but incorrectly, known as the "piracy law"), fragments of which were discovered at Delphi and Knidos. A key section of this law states: The senior consul is to send letters to the peoples and states to whom he may think fit, to say that the Roman people will have care, that the citizens of Rome and the allies and the Latins, and those of the foreign nations who are in a rdationship of friendship with the Roman people may sail in safety, and that on account of this matter and according to this statute they have made Cilicia a praetorian province .... [H]e is to send letters to the effect that it is also right for them to see that no pirate (peirates)use as a base of operations their kingdom or land or territories and that no officials or garrison commanders whom they shall appoint harbour the pirates (peiratas)and to see that, insofar as it shall be possible, the Roman people have them as contributors to the safety of all ... 30

Through this statute the Romans justify taking military action against those whom they have identified, or will identify, as pirates. The intention to carry out military action is clear from 27 Morgan 1969 suggests that the whole episode may have

been linked to political maneuvers to do with the tribunate of C. Gracchus.

under Antonius. Tac. Ann. 12.62 refers to Byzantine involvement; for further discussion, see de Souza 1999, 102-106. 30 Lex deprovinciispraetoriis,Knidos

28

For detailed discussion of the sources and date of this expedition, see de Souza 1999, 102-108.

29

IGRR 4.1116 honors a Rhodian naval officer who served

III, lines 28-3 7; Delphi B 10--12; 100 B.C. For full text, translation, and commentary, see Crawford 1996, 231-270. On the Hellenistic Greek word peirates,which was a synonym for leistes,and the origin of the Latin word pirata,see de Souza 1999, 2-13.

ROME'S CONTRIBUTION TO THE DEVELOPMENT OF PIRACY

79

the establishment of Cilicia as a praetorian province, meaning in this case a sphere of action for a magistrate of praetorian or pro-praetorian rank, who would, like Marcus Antonius, have naval and military forces under his command. The Romans are inviting anyone who sees themselves as friends or allies of the Roman people to join them, with the clear implication that those who do not are either pirates or the supporters of pirates. The message they are broadcasting to the maritime peoples of the eastern Mediterranean is clear: join us, or be classed as pirates and, therefore, as legitimate targets for military action. 31 It is noteworthy that the Rhodians seem to have played a prominent role in this Roman initiative. The Rhodian contribution to the campaign of Marcus Antonius the Orator has already been noted. Rhodian ambassadors were granted a special hearing in the senate by the lex de provinciispraetoriis and were given their own copies of the letters urging the kings in the eastern Mediterranean to deny harbors to pirates. 32 It may even be the case that the Romans borrowed the idea of requiring others to deny bases to those whom they designated as pirates directly from the Rhodians, who used this approach in their treaties with some of the Cretan cities. Now it seems that the Romans were assuming the mantle of maritime protectors, which the Rhodians had successfully adopted in the late third and early second centuries B.C. Why might the Romans have adopted such an approach at this time? An explanation may be sought in the nature of Roman imperial expansion, especially in the eastern Mediterranean. Traditionally Roman expansion has been portrayed by scholars in terms of reluctant responses to the requests of allies.33 The nature of Roman imperialism in the Republican period has been the subject of several recent studies, which have stressed the manner in which Rome's competitive political system, combined with her traditionally militaristic culture, encouraged the conquest of new territory under the leadership of the senatorial elite. Earlier preoccupations with concepts of defensive imperialism and "just wars" have given way to a view of the Romans as essentially an aggressive, acquisitive people whose political leaders depended heavily on the fruits of war to maintain their positions. 34 The senatorial aristocracy has often been characterized as indifferent to economic gains and slow to conquer and exploit territory for commercial reasons. While the image of the reluctant imperialists has been considerably modified by recent scholarship, the idea persists that, in spite of their military successes and political gains, the Romans did not actively seek economic domination over the Greek world. 35 Nicholas Rauh has recently put forward an interpretation of the spread of Roman influence in the Mediterranean that emphasizes the predatory, commercial nature of Roman expansion. 36 For Rauh the Romans were commercial imperialists whose military might was deployed to achieve and maintain economic dominance. He stresses the fact that their second-century B.c. expansion was at the expense of major commercial maritime rivals like Carthage and Corinth. Recent studies of the amphora evidence are interpreted by Rauh to show that Roman producers and traders usurped local commercial interests. He points out that by the early first century B.C., "Roman mercantile elements firmly controlled Aegean maritime

31 For possible contemporary evidence of maritime communities in southern Anatolia wrestling with this issue, see de Souza 1997.

34

32 See

35 For

de Souza 1999, 112-114.

33 E.g.,

Foremost in the revision of Roman imperialism in the Republican period has been Harris 1979; see also North 1981; Richardson 1986; and Rich 1993. cautious views of economic aspects of Roman imperialism, see Harris 1971 and Crawford 1977.

Frank 1914; Badian 1968. 36

Rauh 2003.

80

PHILIP DE SOUZA

centres. "37 This control forced other maritime communities either to submit to Roman dominance or find ways to challenge it, both politically and economically. Late Republican Roman expansion was driven by the competitive and acquisitive culture of the Roman aristocracy, but their voracious appetite for wealth, power, and prestige depended on the willingness of ordinary Romans and Italians to participate in overseas campaigns lasting many years. That willingness was severely tested in the late Republic. Aristocratic commanders had to promise ever greater rewards to the citizen-soldiers in order to persuade them to join their armies. The resources of Roman Italy were insufficient, so allies were needed from the provinces and neighboring territories, where ambitious local elites and monarchs were encouraged to provide a substantial part of the necessary manpower and money. The burden that fell on these allies was often hard to sustain, especially in the face of military setbacks and limited direct benefits.38 The situation was exacerbated by the possibility of alternative allegiances, offered by leaders like Mithridates VI of Pontus, who deliberately presented themselves up as opponents of Roman imperialism. 39 The Romans tolerated no competitors, however, and those who refused to join the Roman imperium, either as allies or official friends, were branded enemies and forced into submission. Many Roman military campaigns of the first half of the first century B.C., which are presented in our sources as being undertaken against pirates, are more likely to have been deliberate attempts to overcome politically and economically autonomous communities that refused to submit to Roman domination. Some of these communities cooperated in alliance with one another, as was the case in Crete, or with more powerful states like the kingdom of Pontus under Mithridates VI, or even with renegade Roman leaders like Sertorius. In order to cultivate the support of the communities whose men, money, and other resources were so vital to them, the Romans played upon their fears. 40 Prominent among these was the threat of pirates plundering their coastal towns and cities, seizing their property, their slaves, and their children. What the surviving historical sources portray as the suppression of piracy can be interpreted as overcoming the resistance of Rome's rivals. By deliberately blurring the distinction between pirates and enemies, the Romans were able, in a similar fashion to the Rhodians, to maintain a stance as benefactors of the Greek world in general, through their apparent suppression of piracy.41 In many cases the leaders of these enemies were chieftains or local ruling dynasts. Few of their names have survived into our narrative sources, apart from powerful kings like Mithridates or long-term rebels like Sertorius and Viriathus. 42 This is, in part, a consequence of the pirate label, which encourages writers to subsume all pirate-leaders under one general term, severing them from their more legitimate titles or local identities. But a few have made it 37

Rauh 2003, 33.

38 On Roman manpower problems, see Brunt 1971, chaps. 23-25 and Linton 1994, 36-39. Cassius Dio summarizes Roman attempts to overcome their maritime opponents prior to 67 B.C. in these words: "Nothing was achieved, except that the allies had to suffer even greater hardship as a result of these attempts, until their situation became quite desperate" (Dio 3623.2). 39

See McGing 1986.

Mith. 46-48. 41

The idea that piracy in the late Republican period is a form of resistance to Roman imperialism has been put forward before-e.g., by Pohl 1993. What such scholars have failed to do is to examine how the labd "piracy" is being used (exceptions to this are the excdlent articles by Avidov 1997 and Gabridsen 2003). It is insufficient to say that faced with Roman power and domination, maritime communities "turned to piracy." Rather they fought back as best they could, often adopting the raiding mentality outlined by Gabridsen 2001.

40

See, for example, the portrayal of Mithridates as a barbaric, destructive tyrant, intent on overthrowing the social and political stability of the civilized Greek world in App.

a bandit rather than a pirate (App. Iber. 60-75; Str. 3.4.5); see further Grunewald 1999, 49-f8.

42 Vlriathus is labded

ROME'S CONTRIBUTION TO THE DEVELOPMENT OF PIRACY

81

into the historical record, such as Zeniketos the Pamphylian dynast (Str. 14.5.7) or Lasthenes and Panares (App. Sic. 6.2). Even Roman aristocrats could be tarred with the pirate brush if they associated with Rome's maritime enemies. Quintus Sertorius had been an ally of the ambitious aristocrat Lucius Cornelius Cinna and was an avowed opponent of the ruling senatorial oligarchy established by Sulla. He went to Spain in 83 B.C. as a praetor and by reducing the military demands on the local population made himself a popular, if not wholly legitimate, governor. When new magistrates were sent by Sulla to replace him in 81 B.C., he fled to Mauretania, from where he tried to reestablish himself in Spain. He allied with a Cilician naval force, whom Plutarch labels pirates (Plut. Sert. 7.3), and attacked the Balearic Island of Pityussa, modern Ibiza (Plut. Sert. 7-12). After an initial success against the small Roman garrison, Sertorius and the Cilicians were driven off by a large fleet under the command of the Roman praetor Gaius Annius in 80 B.C. and returned to Mauretania, where they became involved in a civil war. Sertorius continued to oppose "legitimate" Roman forces by land and sea, entering into negotiations with Mithridates in 75 B.C. for further naval support, promising in return to secure Roman recognition of the Pontic king's rights over territories in Anatolia, once he and his supporters had gained control in Rome (Plut. Sert. 23-24; App. Mith. 68; Cic. Leg. Man. 46).43What these two opponents of the ruling Roman aristocracy had in common was the maritime dimension to their struggle and Cilician involvement, which meant they could be accused of consorting with pirates.

6. Rome and the "CilicianPirates" By designating Cilicia as a praetorian province, the lex de provinciispraetoriisof 100 B.C. had provided for further campaigns in the southeastern part of Anatolia by Roman magistrates, ostensibly as part of a general policy of eradicating piracy.44 Cilicians often featured among the naval forces of the Persian and Seleukid monarchs, 45 and these areas had put up stubborn resistance to many powerful armies over the previous centuries, including those of Alexander the Great and his successors. Nominally the Seleukid, Ptolemaic, or Attalid kings claimed suzerainty over the region in the third and second centuries B.C. In practice, however, the local communities retained a considerable degree of autonomy in the Hellenistic period, although they were insufficiently numerous or wealthy to extend their own political influence outside the region. In the 140s B.C. Korakesion on the eastern side of the Bay of Pamphylia was used as a base by Diodotus Tryphon. This pretender to the Seleukid kingdom employed Cilician naval allies to raid the coastal cities of Syria and the Levant (Str. 14.5.2; 16.2.19). The surviving ancient narrative sources tend to present this as the beginning of a long period of lawlessness and piracy on the part of the inhabitants of the region, but, as will be argued below, this view is probably the result of a retrospective historical analysis. Recent research has begun to demonstrate that much of this region was well developed and fairly prosperous, producing grapes, sweet wine, timber, and slaves and playing an important role in the maritime economy of the Hellenistic Mediterranean, 46 making it a desirable 43 On

Sertorius and his maritime allies, see de Souza 1999, 132-134, 142-143.

1980, 1230-1234. 45 See Wallinga

44

Cilicia is often an imprecise designation in the sources for this period; e.g., App. Mith. 92. At times it covers most of southern Anatolia, including Rough Cilicia, Pamphylia, and even parts of Lyda; see Sherwin-White 1976; Mitford

1987; Billows 1990, 357-360; de Souza 1999, 98-99; forthcoming. 46

See Rauh 1999; 2003; website.

82

PHILIP DE SOUZA

object of Roman imperialism. It is hardly surprising, therefore, that Pamphylians and Cilicians feature as allies of Rome's enemies and the object of Roman military campaigns. 47 In the first quarter of the first century B.C. there were several major operations in this region by Roman commanders, culminating in the four-year campaign of Publius Servilius Vatia lsauricus (78-7 4 B.c.).48 His conquests in the mountainous regions of eastern Lycia, Pamphylia, and lsauria were also celebrated for their success in destroying pirates and bandits in their strongholds (Flor. 1.41.3-4; Str. 14.5.7; Eutr. 6.3 ). Some of the Roman commanders were involved in wars with Mithridates of Pontus, but in many cases the objects of these military operations are described as pirates or bandits, deserving of their defeat and destruction at the hands of the Roman commanders, such as Lucius Licinius Lucullus (App. Mith. 56; Plut. Luc. 2), Lucius Licinius Murena (App. Mith. 64 and 93; Str. 13.4.17), and Servilius. Mithridates is also presented as the friend and ally of pirates and bandits (App. Mith. 63 and 92).49 Servilius, like Marcus Antonius before him, celebrated a triumph, the pinnacle of Roman military glory. The triumph required the return of the general to the city of Rome, which in turn required him to travel through many of the cities and territories of Rome's friends and allies. On his way Servilius took every opportunity to exhibit his prisoners, in chains and suitably labeled, to demonstrate that the Romans were as good as their word when it came to hunting down bandits and pirates. Thus in these public displays of the effectiveness of Roman policy he was able to show that his campaign was justified: One man, Publius Servilius, captured alive more pirates than all the previous commanders put together. And when did he ever deny to anyone the pleasure of seeing a captured pirate? On the contrary he displayed the most enjoyable spectacle of captive enemies in chains to all and sundry. And so they came from all around, not just from the towns through which the pirates were being led, to behold the sight. (Cic. II Ven: 5.66)

While it may be that some of the campaigns in Anatolia in thisperiod did make a contribution to the suppression of piracy (and banditry), it also seems clear that aggressive imperialism was the major driving force behind the activities of the Roman magistrates. The period 102-74 B.C. saw the gradual extension of Rome's empire to include most of southern Anatolia, until only parts of Pamphylia and Rough Cilicia remained outside of their control. 50 Things did not always go smoothly for the Romans. Within a few years of Servilius's triumph a Roman commander in the eastern Mediterranean was forced to come to terms and surrender to the enemy after a disastrous attempt at further conquest. This commander was another Marcus Antonius, the son of Marcus Antonius the Orator, who had been elected a praetor for 74 B.C. and given a special maritime command with authority to act in any province up to 50 miles inland. The initial purpose of this wide-ranging commission seems to have been to secure the maritime supply routes in the western Mediterranean to Roman forces fighting Sertorius in Spain. He achieved some success at this, and his command was prorogued. 51 In 72 B.C. he launched an attack on the Cretans, using forces that were mostly recruited from and financed by Rome's allies. The extant sources once again indicate an attempt 47 For

a useful summary, see Rauh 2003, 169-172.

48

See de Souza 1999, 116-131.

49

For detailed discussion, see de Souza 1999, 125-128.

account of the campaigns. Pohl 1993 takes an alternative view, stressing the Romans' economic motivations, but see de Souza 1995. On the extent of banditry in the region before and after the Roman conquest, see Hopwood 1983. 51

50 See Sherwin-White

1984, 93-131, 149-158 for a detailed

See de Souza 1999, 141-148 for detailed discussion.

ROME'S CONTRIBUTION TO THE DEVELOPMENT OF PIRACY

83

to justify this act of aggressive imperialism by accusing the Roman magistrate's opponents of being pirates and allies of Mithridates, although they were formally allies of Rome. Antonius's expectations of a quick, easy victory were not realized. In spite of their long history of internal rivalries the Cretan cities successfully resisted his invasion under the leadership of a certain Lasthenes, defeating Antonius in a sea battle of Kydonia (App. Sic.6; Flor. 3.7; Diod. 40.1; Dio fr. 108).52 This reverse, coinciding with a difficult stage of their struggle against Mithridates, appears to have caused considerable embarrassment to the senatorial leadership at Rome. Realizing that they were unlikely to be able to resist a determined attack, the Cretans tried to forestall further action and renew their alliance with Rome. They won over a majority in the senate, who voted to absolve them of all charges of treachery and declare them friends and allies of Rome, but the populace wanted revenge, and the tribune Lentulus Spinther vetoed the senatorial decree. A demand for the surrender of ships and hostages (including Lasthenes) was made, which the Cretans rejected, and war followed, justified to the wider Mediterranean world by renewed accusations of supplying Mithridates with mercenaries and sheltering pirates (Diod. 40.1; App. Sic.6.2). Roman and allied forces, commanded by Quintus Caecilius Metellus, invaded Crete in 69 B.C. and, after a long, bitter struggle involving several sieges, reduced the whole island to submission (Dio. 36.18-19; LivyPer. 98-99; App. Sic. 6.2). The island was incorporated into a new province with Cyrene in 66 B.C. By far the greatest exploiter of the policy of conducting major campaigns in the name of the suppression of piracy was Gnaeus Pompeius Magnus, Pompey the Great. While Metellus was completing his conquest of Crete, Pompey was commissioned to clear the seas of pirates and given substantial resources of men, money, and ships under a law proposed by the tribune Aulus Gabinius in 67 B.C. Pompey's command gave him exceptional powers over Roman forces, allies, and even other senatorial magistrates. It was conferred on him by a law of the popular assembly. In order to convince the Roman citizen body to do this, his advocates seem to have emphasized the threat to the city's food supply, as well as the general vulnerability of trade and Roman interests to maritime enemies: Their [the pirates'] power was felt in all parts of the Mediterranean, so that it was impossible to sail anywhere and all trade was brought to a halt. It was this which really made the Romans sit up and take notice. With their markets short of food and a great famine looming, they commissioned Pompey to clear the pirates from the seas. (Plut. Pomp.25.1)

Nevertheless, to head off objections Pompey had to secure political support from the aristocracy, who were extremely reluctant to grant the widespread powers demanded, but, by playing on fears of pirates devastating commerce in the eastern Mediterranean and even raiding the coasts of Italy, Gabinius and Pompey's supporters were able to convince the equestrians and some influential senators to back them, isolating or reducing the backing for Pompey's opponents. In the end it was overwhelming popular support that carried the day.53 Pompey's campaign was brief but remarkably successful. In three months he secured the seas around Italy and then invaded and conquered his principal targets, the areas of Pamphylia and Cilicia that still resisted Roman domination, receiving the surrender of their so-called pirate-leaders. The secret of his success seems to lie in a remarkable willingness to come to terms with the enemy. Unlike 52 On

the economic prosperity and autonomy of Crete at this time, see de Souza 1998.

53 The

principal sources for the campaign are Dio 36.20-36;

App. Mith. 91-96; and Plut. Pomp.24-29. For further sources and detailed discussion of the background to Pompey's campaign, see de Souza 1999, 161-167. On the supporters and opponents of the law, see Seager 1979, 32-36.

PHILIP DE SOUZA

84

Servilius and Metellus, who eventually reduced their opponents to submission by fighting and hard sieges, Pompey, wary of the demands such a campaign would make on Roman and allied resources yet anxious to obtain a quick victory to further his own political career, offered a general amnesty in return for immediate surrender. So attractive did his terms prove that even some of the cities and communities in Crete, which were under attack from Metellus, tried to surrender to Pompey instead. 54 Meanwhile Pompey himself, through a law proposed by the tribune Manilius, obtained the command against Rome's principal enemy in the East, Mithridates. Pompey's campaign of 67 B.C. is presented by the ancient sources, and by most modern scholars, as a successful action to clear the seas of pirates. Yet there is another way of looking at it.

7. The "PirateWar"as a Genuine War Cassius Dio, in describing the background to the lex Gahinia,writes about the expansion of the power of the "pirates" (he deliberately uses the Greek word katapontistai,to distinguish these pirates from land-bound bandits, whom he calls leistat)from small forces into large fleets with their own generals, naval stations, and networks of alliances. He describes the growing threat from the pirates in terms of plundering and ravaging (Dio 36.20-22), hut those were, as has been outlined above, standard tactics of ancient warfare. The fact that they were able to mount a successful raid on the harbor at Ostia and, so Dio claims, to devastate other cities in Italy indicates that they were an effective naval force. A series of Roman military commanders had already campaigned in the region since Marcus Antonius in 102 B.c." Even in 67 B.C. the proconsul Quintus Marcius Rex was there with a substantial army.56 Contemporary and later writers certainly made use of the vocabulary of warfare when referring to this campaign. The idea that it was just like a regular war goes all the way back to Cicero, who calls it a maritime or naval war in his speech on the lex Maniliain 66 B.C (Cic. Leg.Man. 28, 44, 58); he also refers to the pirates as hostes (enemies), the Latin term for opponents in warfare (Cic. Leg. Man. 33 and 46). Significantly, his phraseology is echoed by Caesar in the Civil Wars,who refers to one of the killers of Pompey as a former centurion who served under him in the "pirate war" (hellopraedonum;Caes. Bell civ. 3.104.3), which indicates that Cicero's was the accepted way of describing the campaign among the Roman aristocracy. Velleius Paterculus calls it a helium (Veil. Pat. 2.32.4). Appian, probably echoing Cicero, referred to it as the maritime campaign (tesstrateias ta peri ten thalassan;App. Mith. 91) and chose to write about Pompey's antipiracy campaign of 67 B.C. within his account of the Mithridatic Wars (App. Mith. 91-96). Pompey's supporters used a warship's stern ornament, or aplustre, as a symbol of this war on coins.' 7 Plutarch indicates that the cities and ships captured from the pirates were a key part of Pompey's triumphal lists (Plut. Pomp. 45), and Cassius Dio has the Roman politicians in 67 B.C. talk of a campaign or a war against the pirates (Dio 36.23.3; 35.2). The allocation of vast military forces and numerous suhcommanders was certainly appropriate to a major war (Dio 36.37).' 8 Nicholas Rauh has argued, based on the analogy of the Anglo-American pirates of the early eighteenth century, that the manpower the pirate communities depended upon was in part made up 54 On

the details of Pompey'scampaign,see de Souza 1999,

56 See Sherwin-White1984,

185-187.

167-178. 57

"See Sherwin-White1984, chaps. 5, 7-8; de Souza 1999, chap. 4.

Crawford 1974, 449/4b; de Souza 1999, 173-174.

58 See Breglia 1970-1971

for details.

ROME'S CONTRIBUTION TO THE DEVELOPMENT OF PIRACY

85

of a disaffected "maritime mob," drawn from communities that had suffered economic and social decline as a result of Roman political and commercial expansion. 59 Whatever the validity of this argument might be, it is clear that the leadership and initiative for attacks on places in the Aegean and the coasts of Italy, and for alliances with Rome's enemies like Mithridates or rebel Romans like Sertorius, must have come from people with a sophisticated political awareness. If we were to substitute the more neutral term "enemies" for the emotive label "pirates," then the nature of the conflict that Pompey and his forces were engaged in would surely seem no different to the foreign wars fought by previous Roman commanders. It is not difficult to view the struggle to overcome the Cilicians in the same light as the struggle to overcome the Carthaginians or the Iberians and Lusitanians. How, then, did the idea that this war was fought against "pirates" gain acceptance?

8. Developingthe Imageof Piracy There seems no doubt that the transformation of the Cilicians from powerful maritime enemies into "pirates" owes a great deal to the rhetoric of men like Marcus Tullius Cicero, who skillfully presented the Cilicians' growing maritime threat to Roman power as a manifestation of the ubiquitous problem of piracy. The need for military action to suppress piracy had become a useful political theme in the late Republican period, but how genuine was that threat? While the capacity of piracy, broadly defined, to disrupt the peace and harmony of the Mediterranean was regularly appealed to in deliberately vague terms, creating the impression of impending maritime anarchy, it is difficult to determine the precise nature of the threat because of the political contexts in which it is invoked. It can be reasonably inferred that there were some mundane pirates operating in the Mediterranean in this period. That is, there were armed robbers with ships who owed no particular political allegiance and whose actions were motivated only by thoughts of immediate material gain. Their existence is demonstrated by occasional references to pirates in inscribed decrees of maritime communities, like the late second-century decree of Ephesos honoring men of Astypalaia who rescued Ephesian citizens from pirates (IG 12.3.171). Small bands of pirates were probably responsible for this and similar episodes like the killing of Atyanas, the Olympic boxer (Cic. Place.31), or even the capture and ransom of Julius Caesar in 74 B.C. (Suet. Iul. 4; Plut. Caes.2). What, then, should we make of the celebrated incidents of piracy around Sicily in the late 70s B.C., as detailed in Cicero's Verrine 60 Cicero's accounts of these instances seem to indicate that there were no major political Orations? motives for these attacks. The pirates do not seem to be involved in a war with Rome but are portrayed as stateless outlaws, acting on their own initiative. Nevertheless, the scale of the devastation seems more appropriate to warfare. For example, in his summary of Verres's failings Cicero says: [During Verres's governorship] ... well fortified harbors and the securest of cities lay open to pirates and bandits; Sicilian sailors and troops, our allies and friends were starved to death; the finest and most excellently turned out fleets were lost and destroyed, bringing great disgrace to the Roman people. (I Ve". 1.13)

If these losses were said to have been inflicted by the forces of a well-known enemy, such as Mithridates, then there would be no problem, but because a political allegiance is not specified for the 59 Rauh2003, 162-200, relyingheavilyon the model provided

by Rediker 1987.

60 See esp. II

Verr.5 .60-108.

PHILIP DE SOUZA

86

perpetrators, who were probably based on the island of Malta (Cic. II Ve". 4.103-104), we are left with a puzzle. Why did the Romans, at a time when they were proclaiming themselves as the suppressors of piracy, allow the presence of sea-borne raiders who could apparently cause such devastation so close to Italy? Cicero, of course, blames the incompetence and corruption of Verres for his failure to maintain a suitable fleet of warships, although he has to admit that some pirates were captured by men under Verres's command (Cic. II Ve". 5.63 ). He dismisses this as pure luck, however, and concentrates on what he claims were serious losses in humiliating defeats at the hands of the pirates (Cic. II Ve". 80-101). When the exaggerated rhetoric is stripped away from Cicero's account, however, what seems to have occurred is that Verres's naval commander, Cleomenes, allowed himself to be surprised while in harbor at Pachynus, losing two small, undecked vessels (Cic. II Ve". 5.87-90). Moving on to Elorus, some of his ships were set on fire in a daring nocturnal raid, while the pirate captain Heracleo sailed on to the harbor of Syracuse. Finding this harbor well defended, he fled (Cic. II Ve". 5.95-101). Was this force the remnants, or even the main part, of the Cilician fleet that had made an alliance with Sertorius some years earlier? It is possible, especially if a reference to pirates exacting protection money from the Lipari Islanders over several years is taken to refer to the same group (Cic. II Ve". 3.85), but uncertainty remains, as Cicero does not make the connection, although he does refer to other fugitives from Sertorius's army (Cic. II Ve". 5.72). Nor do later references to the achievements of Verres's successor, Lucius Metellus, in suppressing piracy around Sicily make any mention of Cilicians (Livy Per.98; Oros. 6.3 .5). Cicero also emphasizes the apparent failure of Verres to execute publicly the captain of the captured pirate vessel. He is contemptuous of Verres's explanation, that he needed to hold the man as a potential witness in his own trial, and uses the contrast with Verres's alleged execution of Roman citizens to make more rhetorical points (Cic. II Ve". 5.74-78). Indeed, Cicero litters the published version of his speeches with references to pirates in order to convey the image of Verres as a rapacious, incompetent, and irresponsible failure. He plays on his audience's perception of pirates as dangerous marauders and exaggerates the impact of their actions, while at the same time characterizing them as unworthy opponents whose success brings shame to the Romans. 61 Cicero takes a similar approach in his speech in support of Manilius's proposal to transfer the Mithridatic command to Pompey, but here he makes the threat posed by pirates out to be even greater because he needs to magnify the achievement of Pompey. 62 What is noticeable in this speech is the extent to which Cicero conjures up, in vague but evocative terms, the specter of pirates "ruling the seas" to the detriment of Rome and her allies: What province did you keep free from pirates in those years? What revenue of yours was safe? Which allies did you defend? Whom did you protect with your fleets? How many islands do you think were deserted, and how many allied cities were abandoned in fear or captured by the pirates? (Cic. Leg.Man. 32)

Cicero offers the Roman people a vision of a recent past in which they have conspicuously failed in their obligation to protect not only themselves but their friends and allies as well: We, who used to guarantee not just the safety of Italy, but were able, through the prestige of our imperial power, to preserve unharmed all our far-flung allies ... were the same ones who were then not only kept out of our provinces, away from the coasts of Italy and our harbors, but were even driven off the Appian Way! (Cic. Leg. Man. 55) 61

See de Souza 1999, 150-157.

62 See

de Souza 1999, 172-175.

ROME'S CONTRIBUTION TO THE DEVELOPMENT OF PIRACY

87

He then makes the most of the contrast provided by what he claims was the result of Pompey's almost miraculous victory against the pirates in his campaign of the previous year: All pirates wherever they were suffered capture and death, or handed themselves over to this singularly powerful commander. Even the Cretans, when they sent emissaries to him in Pamphylia to plead their case, learned that there was hope for their surrender and were ordered to give hostages. (Cic. Leg. Man. 35)

Cicero effectively leaves his audience with the impression that the seas were overrun with piratical fleets, and that it was only the divinely inspired genius of Pompey the Great that saved them from utter ruin (Cic. Leg. Man. 32-33 ). Cicero shows how the appeal of successful suppression of piracy could be used to further a Roman political career. It is clear that Pompey did not, as Cicero claimed, rid the Mediterranean of pirates, but he did enable the Romans to claim that they had done so while incorporating more territory into their rapidly expanding empire through his annexation of Cilicia.63 In a speech delivered just a few years later in defense of Gaius Valerius Flaccus in the repetundae court, Cicero once more invoked the image of Pompey ridding the Mediterranean of fleets of marauding pirates (Cic. Place.27-33 ). But this time he had to admit that just one campaign, however spectacular, had not been enough to end the pirate menace entirely; otherwise there would be no reason for his client to have demanded money from the provincial cities of Asia in order, he said, to finance a fleet of warships to guard against pirates. Here we see Cicero presenting to the jury of Roman aristocrats the image of pirates operating from bases in Syria, Cilicia, Cyprus, and Crete as a continual threat to the maritime security and commercial life of the eastern Mediterranean. Nevertheless, Cicero struggles to come up with specific examples of the danger that piracy represents, citing only one instance, the killing by pirates of an Olympic boxing champion called Atyanas from Adramyttion on the western coast of Anatolia (Cic. Place.31). Cicero has to admit that Flaccus undertook no successful naval actions against pirates, which meant that he had no prisoners to parade through the province to justify his exactions. Cicero also has to admit that his own brother, who succeeded Flaccus as governor of Asia, did not continue to demand money for this fleet. Nevertheless, he maintains that the threat is there, and he uses the pirate menace to justify recent expenditure on fleets and coastal defenses for Italy (Cic. Place.30). A further example of a Roman provincial commander exploiting the suppression of piracy to justify acts of imperialism comes from the career of Gabinius, the politician who as tribune proposed Pompey's extraordinary command in 67 B.C. Gabinius was proconsul of Syria from 58 to 55 B.C. His proconsulship was most notorious for the restoration of King Ptolemy XII Auletes to the throne of Egypt, an operation that necessitated himleaving his province with an army in order to drive out the usurper Archelaus. In a speech defending Gaius Rabirius Postumus, Cicero tells us how Gabinius defended his actions: Gabinius said that he had done it for the sake of the Republic, since he was worried about the fleet of Archelaus, because he thought he might fill the seas with pirates. (Cic. Rab. Post.20)

Few of Gabinius's fellow senators would have believed this excuse, especially as they knew how much money Ptolemy had been offering various politicians to induce them to aid hisreturn, but it would have

63 Cic. Leg.Man.28-33; in this speech Cicero portrays Pompey's achievements in Cilicia as comparable to the conquests of Republican heroes like Scipio and Marius, who defeated powerful enemies and added new provinces to the Republican

empire. For the continuation of piracy soon after 67 B.C. in the regions where Pompey had supposedly eradicated it, see Cic. Place.28-33; Dio 39.56.1 and 59.2.

88

PHILIP DE SOUZA

been a useful way of convincing allies and subjects that there was something more than imperialistic greed behind the proconsul's expedition. 64 Ten years later accusations of banditry and piracy were still being leveled by Roman politicians against their Egyptian opponents. Toward the end of the third book of his account of the Civil War Julius Caesar describes the army arrayed against him at Alexandria under the command of Achillas in highly derogatory terms. He dismisses its contingent of Roman soldiers from among those brought to Egypt by Gabinius in 55 B.C. as no longer deserving of the Roman name, and he labels the troops from Syria and Cilicia as an assortment of pirates and bandits (Caes. Bell.civ.3.110). 6' What can be observed here is a process of shaping and exploiting the image of piracy for specific political ends. Cicero would certainly not have been the only one to present maritime enemies as pirates in this way, although it is difficult to demonstrate that piracy was a widespread theme in contemporary political discourse because we have very little non-Ciceronian material. Caesar's Civil Wars,already mentioned above, does contain examples of the casual and dismissive way in which the Roman aristocracy deployed the term "pirate" to describe those whom they did not consider worthy of respect. It therefore seems reasonable to use Cicero's writings as evidence of a generally accepted image of piracy in the late Republic. In his De republica,written in the late 50s B.C., Cicero deploys the accusation of piratical practices by Rome's enemies in a historical context: For, indeed, among the barbarians there were in former times none who were seafarers, except the Etruscans and the Phoenicians, the one on account of trade, the other for the sake of piracy. (Cic. Rep. 2.9)

The claimis made to highlight the immoral tendencies of coastal cities and the superiority of Rome's location: inland but on a navigable river for access to the coast. In his famous work On Duties, composed in 44 B.C., Cicero argues that there is an obligation to keep one's sworn word even to enemies because warfare is governed by legal principles and good intentions should not be set aside when dealing with foes. To illustrate the point that the validity of a sworn oath is dependent on one's intention to keep it, he offers an instructive counterexample:

ff, for example, you do not hand over to pirates the amount agreed upon as the price for your life, this is not perjury, not even if you have sworn an oath and do not do so, for a pirate is not included in the category of lawfulenemies, but they are the enemies of all mankind. (Cic. Off. 3.107) Here Cicero expands on the label "pirate," defining the pirate as the enemy of all mankind, presenting piracy as an all-pervading maritime evil, outside the laws and conventions of civilized peoples. The moral and sociopolitical dynamic at work here is the denial of equal status, thus effecting a transformation of the so-called pirates into primitive, uncivilized barbarians who are much lower in status than legitimate, wartime enemies and to whom no social or moral obligations need be felt.66 64

On Gabinius's expedition, see Sherwin-White 1984, 271-275; de Souza 1999, 18.3-184.

65Hueaccedebant collectiexpraedonibuslatronibusque Syriae

Ciliciaequeprovinciae/initimarumqueregionum.(Added to these were men gathered from the pirates and bandits of Syria and the province of Cilicia and neighboring regions.) A degree of self-awareness among the Roman elite regarding the way that they were presenting their own rapacious violence as a morally justified response to the depredations

of others can be seen in Sallust's Histories.In a letter attributed to Mithridates, Sallust, writing in the early 30s B.C., has Mithridates refer to the Romans as "bandits who have plundered the whole world" (Sall. Hist. fr. 4 69M). 66

This echoes the language Cicero employed in 70 B.C. to describe a pirate captain in the Verrines as "the most dangerous enemy of the Roman people, or, rather, the common enemy of all peoples" (Cic. I Vm. 5.76). Similar views are expressed elsewhere in this speech: pirates should be led in

ROME'S CONTRIBUTION TO THE DEVELOPMENT OF PIRACY

89

By the latter part of the first century B.c.an accepted image of piracy had emerged in political, historical, and moral literature. Piracy was an evil that afflicted the civilized world, and pirates were outlaws whom the Romans and those who shared their political and moral values should suppress, for the good of all mankind. This moral and political imperative is picked up on by other first-century B.C. writers besides Cicero. Cornelius Nepos has Themistocles suppressing pirates in the early fifth century B.C. (Nep. Them. 2.3 ), and Diodorus Siculus presents numerous Hellenistic kings and city-states as the suppressors of piracy, including the Rhodians: Indeed, she [Rhodes] attained such a position of power that she took up the war against the pirates by herself, and cleared the sea of their evil manifestation. (Diod. 20.82.3)

Generalizations about piracy also seem to have made their way into Jewish historiography around the same time. Josephus has the Jewish leader Hyrcanos accuse his rival Aristoboulos of instigating piratical raids on his neighbors in a speech delivered before Pompey in 63 B.C. (Joseph. AJ 14.43).67 The speech occurs in a section of his Antiquities of the Jews that is clearly based upon the firstcentury B.C. UniversalHistoryof Nicolaus of Damascus. By the end of the first century B.C, then, it was clearly established that pirates were beyond the conventions of law and civilized society, that they were the enemies of all, and that action against them was not only appropriate but necessary. The manipulation of the menace of piracy for political reasons is particularly prominent in the works of Cicero, a writer who had a tremendous influence on the elite culture of the Roman world and is our main source, but it can also be detected elsewhere.

9. The Imageof Piracyin the Principate As with so many aspects of late Republican political culture, full realization of the political capital inherent in presenting oneself as the suppressor of pirates came with Rome's first emperor, Augustus. He initially used this tactic in his conflict with Pompey's younger son, Sextus Pompeius, in the 40s and 30s B.C. and then as part of the general presentation of the image of the Principate-Rome's vision of a new political order for the civilized world: I made the sea peaceful and freed it of pirates. In that war I captured about 30,000 slaves who had escaped from their masters and taken up arms against the Republic, and I handed them over to their masters for punishment. (Aug. RG 25)

In the Res Gestae,Augustus's posthumous manifesto for theJulio-Claudian dynasty, the suppression of pirates, here neatly equated with another threat to the Roman order, runaway slaves, has become an integral aspect of what modern historians often call the Pax Romana. Contemporary expression of this concept can be found in the work of the historian and geographer Strabo, writing in the Augustan period, who says in his description of the civilized, romanized world of ea. A.O. 20: On top of that there is the current state of peace, for piracy has been suppressed, so that those sailing [between Spain and Italy] are beginning to relax. (Str. 3.2.5)

chains, as was done by Servilius (Cic. II Ve". 5.66); pirate captains should be publicly executed (Cic. II Ve". 5.64 and 67---08);pirate ships should be no match for a Roman naval

force (Cic. II Ve". 5.83-100). 67 See Avidov

1997, 30-31.

90

PHILIP DE SOUZA

Strabo associates piracy with poverty, the absence of strong rule, and lack of good moral examples (Str. 1.3.2; 5.1-2; 6.2.2; 7.3.7; 7.4.2). He points out that the Romans have brought prosperity and good government to many regions where poverty and piracy were found (Str. 2.5.26). He claims that their conquests of Crete and Cilicia have brought an end to piracy everywhere (Str. 10.4.9). A clear message that can be taken from Strabo's Geographyis that in the political and moral landscape of the Roman imperiumthere are "insiders," who have entered into a partnership with the ruling power, either willingly or though conquest, and there are those who have rejected or avoided Roman domination. These "outsiders" are portrayed by Strabo to the intellectual elite of the Mediterranean world as barbarians, bandits, and pirates (Str. 17.3.25).68 There is, of course, a genuine truth here, beyond the political spin. It cannot be doubted that the political power and stability of the Augustan settlement, backed up by the armies and navy of the Principate, did make a considerable difference to maritime security. In one sense it was a realization of the policy inherent in the claims articulated in the lex de provinciispraetoriis,that the suppression of the conditions for widespread piracy, as broadly defined above, could be achieved by the conquest and control of territory. Thus there was nowhere left for pirates (or any other maritime enemies of Rome) to base themselves for their nefarious activities, whether they were ordinary pirates or political enemies whom the Roman authorities found it convenient to label as pirates. 69

10. Rewritingthe Historyof Piracy As piracy was being reinvented in the early Principate to suit a Roman worldview, historical writers began to fashion the story of the suppression of piracy into a long-term mission for the just, morally upright Roman aristocracy who pledged to clear the seas of the unjust, immoral pirates. An early example of this is seen in Velleius Paterculus's pocket history of Rome, when he presents Julius Caesar, captured and ransomed by pirates while on his way to study rhetoric in Rhodes, acting as a precursor of Pompey and Augustus: It would take too long to tell of his many bold plans for the punishment [of the pirates], or how obstinately the timid governor of Asia refused to second them ... although he was but a private citizen without authority, and his fleet had been collected on the spur of the moment, he directed his course to the place where the pirates themselves were, put to flight part of their fleet, sank part, and captured several ships and many men ... (Vell. Pat. 2.42)

Similarly, the more recent past is written up by Velleius to stress the achievements of the good emperor Augustus Caesar against the wicked pirate leader Sextus Pompeius: Through his father's freedmen, Menas and Menekrates, the commanders of his fleet, Pompeius infested the seas with raiding and piracy, using plunder to support himself and his army. Nor was he ashamed to plague with the wickedness of piracy the very sea which had been cleared of it by his father's arms and leadership. (Vell. Pat. 2.73.3)

Strabo, for whom the current absence of piracy is one of the major benefits of the Roman Empire, presents the suppression of pirates as an integral aspect of Rome's conquest of Italy (Str. 5.3.5). 68 For more on Strabo and piracy, see de Souza 1999, 200-204. On Strabo and the Roman world in general, see the excellent study by Clarke 1999, chaps.~-

69

On piracy and the Pax Romana, see de Souza 1999, chap. 6.

ROME'S CONTRIBUTION TO TIIE DEVELOPMENT OF PIRACY

91

He also commends Rome's naval allies the Rhodians for their success in clearing the seas of pirates (Str. 14.2.5). An important effect of the acceptance of this model of Rome as the suppressor of piracy is a marked tendency among historical writers of the Principate to portray Rome's maritime enemies as piratical. The Antonine historian Appian of Alexandria is a good example. He chooses to emphasize the suppression of piracy at several points in his account of Roman history.70 Piracy is presented as a regular practice of the illyrian tribes and provides the cause for Roman invasions in the First and Second illyrian Wars of 229 B.c. and 222 B.C. (App. Illyr. 3, 7, 8). Appian highlights the piratical tendencies of the Carthaginians in his account of the Punic Wars. In his description of the response of the cities of Utica and Carthage to Roman blockades in 203 B.C. he says that they resorted to plundering merchantmen, until Roman naval forces arrived to intercept what Appian calls "the pirate ships" (App. Pun. 25). The plundering of merchants is included in the catalogue of past Carthaginian crimes that he puts into a speech by the Roman consul Censorious in 149 B.C. at the start of the Third Punic War (App. Pun. 86). Appian also says that suppression of piracy was the justification for Roman attacks on Crete in 72 B.C. and again in 69 B.C. (App. Sic.6). Mithridates is accused of both practicing and promoting piracy (App. Mith. 62, 63, 92). Pompey's campaign of 67 B.C. against pirates is a feature within Appian's account of the Mithridatic Wars (App. Mith. 91-96). 71 The way that Appian describes the Cilician pirates may even betray hints of the tension between the earlier and revised versions: Holding the name of pirates in no esteem, from this time onwards they called their gains "profits of war." They kept craftsmen in chains to work on their never-ending projects, supplying them with wood, copper, and iron. Puffed up by the fruits of their success, yet being not at all inclined to give up piratical ways, they envisaged themselves as kings, tyrants, and generals, thinking that if they banded together they would be invincible. (App. Mith. 92)

Appian seems to find it difficult to make his material on the Cilicians fit into the pirate mold, so he is forced to present the subjects of his narrative as pirates who would be kings. 72 A similar dissonance can be detected in Cassius Dio's description of Pompey's piratical opponents (Dio 36.20-23). The suppression of piracy is a theme in Appian's account of the Sicilian War between the forces of Octavian and Sextus Pompeius (App. Bell.civ.5), 73 and Appian justifies the future emperor Augustus's campaigns against the Libumians and the Ionian Islands in 35 B.C. as antipiratical measures (App. Illyr. 16). Thus he is able to present the founder of the Principate as a suppressor of pirates. 74 The historiographical trend outlined above has, I would argue, produced a serious distortion in the portrayal of Roman Republican history. The distorting effect of the widespread use of the label "pirates" can be seen in the way that one historical episode from 67 B.c. has come down to us from the surviving sources. This intriguing case is the kidnap and eventual ransom of a Roman aristocrat, Publius Claudius (later called Clodius) Pulcher. In 69-68 B.C. the young Pulcher had 70 While

it is clear that Appian is dependent upon his sources for information about incidents that are presented as the suppression of piracy, it was his choice to include and emphasize this aspect. On Appian's distinctive historical viewpoint and the extent to which he selected, arranged, and elaborated on his material, see Goldmann 1988.

71 Note

also the way that the activities of Mithridates's general Archelaus are described as "more like piracy than warfare" (App. Mith. 45).

72 For the possible relationship of this description Posidonius's Histories,see Avidov 1997, 44-45. 73

74

to

See de Souza 1999, 185-195.

On the widespread consensus that developed during the Principate regarding the beneficial nature of Roman rule, see Ando 2000.

PHILIP DE SOUZA

92

been attached to the staff of the proconsul Lucius Licinius Lucullus, who was in charge of the war with Mithridates, but he had encouraged some of Lucullus's troops to mutiny and left his command in 67 B.C.to join the proconsul Quintus Marcius Rex, to whom the province of Cilicia had been assigned as a military command with three legions (Dio 36.2.2; 36.15.1; 36.16.2; Sall. Hist. fr. V 13-14M). We do not know exactly what Pulcher's duties were, although it has been conjectured that he was Quintus Marcius's fleet commander, but at some point he was held captive by some Cilicians, who demanded a ransom for him. 7' It may be that he was detained while negotiating with some of the local Cilician dynasts, or it may be that he was captured in battle. Either way, he had to call upon Ptolemy, the king of Cyprus, to finance his ransom. Our only contemporary source for this episode, Cicero's speech On the responseof the Haruspices,written in 56 B.C.,claims that Pulcher was sexually debased by the Cilician pirates (Cic. Harusp.42), but this is probably an exaggeration. Cicero's main aim in his speech was to discredit Pulcher, and this section is replete with accusations of moral degeneracy. Nor do any of the later sources make reference to the circumstances of Pulcher's capture, saying only that he was taken by powerful Cilician pirates, who disdained the small ransom that Ptolemy offered them but released him because they feared the approach of Pompey (Str. 14.6.6; App. Bell. civ. 2.23; Dio 36.17.3). 76 The politics of this episode must have been rather complex. Was Pulcher taken hostage in an attempt to forestall Roman attacks on Cilicia? Was he trying to negotiate a deal with the Cilicians that went wrong, either on behalf of Quintus Marcius, or even Pompey, whose staff he may have joined in 67 B.c.?77 Was King Ptolemy's ransom offer really too small? None of these questions can be answered because the surviving ancient sources are not interested in the full picture, preferring to tum the episode into a colorful anecdote about pirates in order to make their own rhetorical, moral, and political points. We can get a sense of what the labels "pirate" and "piracy" implied to writers of the Principate from the Controversiaeof the Elder Seneca. In these rhetorical exercises pirates are agents who create legal and moral problems on which the speakers can exercise their skills in coming up with persuasive arguments. It is precisely because they live outside the conventions of law-abiding civilized society that pirates are such a useful device. They are the very embodiment of greed, lust, violence, and impiety (Sen. Contr. 1.2; 1.6; 1.7; 3.3; 7.1; 7.4). A virgin captured by pirates has no chance of retaining her chastity (Sen. Contr. 1.2). For a son to marry a pirate chiefs daughter is a disgrace to the family, however noble the girl's intentions might have been (Sen. Contr. 1.6). The worst fate for a daughter, son, or father is to be captured by pirates (Sen. Contr. 1.2; 1.6; 7.4 ). The only place where a parricide can gain acceptance is among pirates (Sen. Contr.7.1). They are also employed as exemplars of moral degeneracy to show up the wickedness of others. Actions that even pirates find cruel and abhorrent must be the ultimate in evil (Sen. Contr. 1.7; 7 .1). Gradually all those who were labeled pirates in antiquity assumed the attributes of romantic but scary outlaws. Hence, by the time Plutarch came to write about them in the early second century A.O., all Cilicians had been transformed into despicable pirates: His [Sertorius's] allies, the Cilician pirates, had no desire for peace or leisure; their interest was all in spoils and riches. (Plut. Sert. 9)

See the imaginative reconstruction of the events in Tatum 1999, 49-50, 265.

two talents offered by Ptolemy the reason for Pulcher's later tribunician law to annex Cyprus.

76 Strabo

77

7'

and Appian make the insult of the small ransom of

See Tatum 1999, 265, n. 99.

ROME'SCONTRIBUTIONTO THE DEVELOPMENTOF PIRACY

93

... but what was most offensive of all about them was their hateful arrogance-their gilded sails, purple coverings, silver oars-the general image that they projected of delighting in their way of life and taking pride in their malicious acts. Roman power was ridiculed by their flute-playing, cithara-plucking, and drunken debauchery, by their captures of prominent Roman magistrates and their demands for ransom from captive cities. (Plut. Pomp.24)

Plutarch's colorful pirates had their counterparts in the fictional works of the early Empire. Pirates are a ubiquitous menace for the heroes and heroines of the Greek and Latin novels. The Satyrica,probably written by Petronius Arbiter in the mid-first century A.D., opens with the ironic statement that "pirates in chains standing on a beach" are among the sights that overeducated young men no longer have the chance to see (Petron. Sat. 1), but pirates play a major part in the plots of the novels of Chariton, Xenophon of Ephesos, Achilles Tatius, Longus, and Heliodoros. These violent and rapacious villains appear on the margins of civilization to kidnap, enslave, plunder, and kill in a fashion familiar to modern audiences of Hollywood films or readers of romantic fiction. They endanger the lives and virtue of the hero and heroine and stretch their devotion to each other to its romantic limits.78

11. Conclusions In the last century of the Republic the Roman aristocracy regularly sought to deploy the military and naval resources at their command in acquisitive, imperialistic campaigns, both for short-term gains-booty, prestige, political prominence-and long-term advantages-commercial influence, tribute, extended political dominion. In order to persuade their citizens, subjects, and allies to contribute to these endeavors they sought justifications for their wars that had a wider, more effective appeal than their own selfish interests. In the suppression of piracy they found and developed an excellent justification for the application of military might against maritime communities. To some extent they were imitating the practice of Greek maritime powers in the Classical and Hellenistic periods, particularly the Rhodians, who may well have been their main inspiration. The Romans made considerable use of this strategy in their conflicts with maritime communities in the late second and early first centuries B.C., when their blatant imperialism was often justified by a fa~ade of suppression of piracy. A problem that this development in the history of piracy confronts us with is that the "genuine" pirates, as opposed to the political opponents of the Romans, are hard to discern. It is in the nature of our surviving sources that they concentrate upon the larger, politically significant groups and their conflicts, paying little heed to the smaller-scale, mundane piracy. My argument is not that there never were any pirates who preyed on the maritime communities and traders of the Mediterranean for immediate gain without grander political aspirations. Nor would I argue that the suppression of piracy by powerful states did not bring tangible benefits to seafarers and coastal dwellers. Rather, it is my contention that the application of the "pirate" label to certain maritime communities was often a deliberate misrepresentation of the nature of these communities. It was intended to demonize them in contemporary eyes in order to justify imperialist aggression against them for far less noble reasons. This rhetorical strategy was successful in both its contemporary political context and as an historiographical theme. Its effectiveness can be seen in the way that the surviving narratives of Roman (and Greek) history are liberally sprinkled with stories of the suppression of piracy. We should read 78

For further details, see de Souza 1999, 214-218.

94

PHILIP DE SOUZA

these stories with a skeptical eye, wary of the way that contemporary accounts, represented for us principally by the works of Cicero, have influenced later writers. These writers may be reproducing what they have read in an uncritical manner, or they may be embracing and elaborating on the idea that the increasing dominance of the Mediterranean was the result of a series of campaigns against pirates. They were comfortable with the notion of Rome as the suppressor of pirates (and other enemies) for the greater good. The key point is to look beyond the familiar images conjured up by the terms "pirate" and "piracy" to see the pejorative labels applied by politicians and historical writers to delegitimize opponents of Roman imperialism and justify their suppression. 79

79 The

use and abuse of the labels "pirate" and "piracy" have a long history beyond the Roman Empire. For some more modem examples, see de Souza 1996, 125-126. I do not think

that it is necessary to spell out the contemporary resonance of this concept to a post-9/11 readership.

ROME'S CONTRIBUTION TO THE DEVELOPMENT OF PIRACY

95

Works Cited Ando, C., ImperialIdeologyand ProvincialLoyalty in the Roman Empire (Berkeley 2000). Avidov, A., "Were the Cilicians a Nation of Pirates?," MediterraneanHistoricalReview 12 (1997) 5-55. Badian, E., Roman Imperialismin the Late Republic,2nd ed. (Oxford 1968). Bakhuizen, S. C., "The Tyrrhenian Pirates: Prologomena to the Study of the Tyrrhenian Sea," in Navies and Commerceof the Greeks, the Carthaginiansand the Etruscansin the TyrrhenianSea, PACT 20, ed. T. Hackens (Strasbourg and Ravello 1988) 25-32. Billows, R. A., Antigonos the One-eyedand the Creationof the HellenisticState (Berkeley 1990). Breglia, L., "I legati di Pompeo durante la guerra piratica," Annali dellaFacoltiidi Lettere e FilosofiaUniversitii di Napoli 13 (1970-1971) 47-66. Brule, P., La pirateriecretoisehellenistique(Paris 1978). Brunt, P.A., Italian Manpower225 B.C.-A.D. 14 (Oxford 1971). Chaniotis, A., "Vier kretische Staatsvertriige," Chiron21 (1991) 241-264. Clarke, K., Between Geographyand History:HellenisticConstructionsof the Roman World (Oxford 1999). Cornell, T. J., The Beginningsof Rome: Italy and Rome from the Bronze Age to the Punic Wars c. 1000-264 B.C. (London 1995). Crawford, M. H., Roman RepublicanCoinage,2 vols. (Cambridge 1974). ---, "Rome and the Greek World: Economic Relationships," EconomicHistory Review 30 (1977) 42-52. --, ed., Roman Statutes (London 1996). Dell, H.J., "The Origin and Nature of illyrian Piracy," Historia 16 (1967) 344-358. De Souza, P., "Piracy and Republican Politics" (review of Pohl 1993), The ClassicalReview 45 (1995) 99-101. ---, '"They are the enemies of all mankind': Justifying Roman Imperialism in the Late Republic," in Roman Imperialism:Post-colonialPerspectives,ed. J. Webster and N. Cooper (Leicester 1996) 125-133. ---, "Romans and Pirates in a Late Hellenistic Oracle from Pamphylia," ClassicalQuarterly47 (1997) 477-481. ---, "Late Hellenistic Crete and the Roman Conquest," in Post-MinoanCrete, ed. W. Cavanagh and M. Curtis (London 1998) 112-116. British School at Athens Studies 2. ---, Piracyin the Graeco-RomanWorld (Cambridge 1999). ---, "Hellenistic and Roman Republican Naval and Siege Warfare," in The CambridgeHistory of Greek and Roman Warfare,vol. 1, ed. P. Sabin, H. van Wees, and M. Whitby (Cambridge forthcoming). Durrbach, F., Chaix d'inscriptionsde Delos (Paris 1921). Earle, P., The Pirate Wars(London 2003 ). Eckstein, A. M., Moral Visionin the Historiesof Polybius (Berkeley 1995). Erskine, A., "The Romans as Common Benefactors," Historia43 (1994) 70-87. Frank, T., Roman Imperialism(New York 1914). Gabrielsen, V., The NavalAristocracyof HellenisticRhodes (Aarhus 1997). ---, "Economic Activity, Maritime Trade and Piracy in the Hellenistic Aegean," Revue des Etudes Anciennes 103 (2001) 219-240. ---, "Piracy and the Slave Trade," in A Companionto the HellenisticWorld,ed. A. Erskine (Oxford 2003) 389-404. Giuffrida I entile, M., La pirateriatirrenica.Momenti e Jortuna (Rome 1983). Goldmann, B., Einheitlichkeit und Eigenstiindigkeitder HistoriaRomana des Appian (Hildesheim 1988). Grunewald, T., Riiuber,Rebel/en, Rival/en, Racher.Studien zu Latrones im romischenReich (Stuttgart 1999). Harris, W. V., "On War and Greed in the Second Century B.c.," American Historical Review 76 (1971) 1371-1385. ---, War and Imperialismin RepublicanRome (Oxford 1979).

96

PHILIP DE SOUZA

"Roman Expansion in the West," in The CambridgeAncient History,vol. 8, 2nd ed., ed. A. E. Astin and F. W. Walbank (Cambridge 1989) 107-162. Heers,}., The BarbaryCorsairs:War/arein the Mediterranean1480-1580(London 2003). Hopwood, K., "Policing the Hinterland: Rough Cilicia and Isauria," in Armies and Frontiersin Roman and ByzantineAnatolia,ed. S. Mitchell (London 1983). Horden, P., and N. Purcell, The CorruptingSea:A Study of MediterraneanHistory (Oxford 2000). Katde, I. B., "Piracy and the Venetian State: The Dilemma of Maritime Defence in the 14th Century," Speculum 63 (1988) 865-889. Lintott, A., "The Roman Empire and Its Problems in the Late Second Century," in The CambridgeAncient History,vol. 9, 2nd ed., ed. J. A. Crook, A. Lintott, and E. Rawson (Cambridge 1994) 16-39. McGing, B. C., The ForeignPolicyo/Mithridates VI EupatorKing o/Pontus (Leiden 1986). Mitford, T. B., "Roman Rough Cilicia,"inAu/stieg und NiedergangderromischenWelt,vol. 2, ed. H. Temporini (Berlin 1980) 7.2:1230-1261. Morgan, M. G., "The Roman Conquest of the Balearic Islands," CalifoniaStudies in ClassicalAntiquity 2 (1969) 217-231. North,}., "The Devdopment of Roman Imperialism," Journalof Roman Studies71 (1981) 1-9. Pohl, H., Die romischePolitik und die Piraterieim ostlichenMittelmeervom 3. bis zum 1. Jh. v. Chr. (Berlin 1993). Rauh, N ., "Rhodes, Rome and the Eastern Mediterranean Wine Trade," in HellenisticRhodes:Politics,Culture and Society,ed. V. Gabridsen, P. Bilde, T. Engberg-Pedersen, L. Hannestad, andJ. Sahle (Aarhus 1999) 162-186. --, Merchants,Sailors& Piratesin the Roman World(Stroud 2003 ). --, The Rough Cilicia Archaeological Survey Project website, http:/ /pasture.ecn.purdue.edu/ ~rauhn/. Rediker, M., Between the Devil and the Deep Blue Sea:MerchantSeamen,Piratesand the Anglo-American MaritimeWorld1700-1750(Cambridge 1987). Rich, J., "Fear, Greed and Glory: The Causes of Roman War-making in the Middle Republic," in Warand Societyin the Roman World,ed.J. Rich and G. Shipley (London 1993) 38--68. Richardson,}. S., Hispaniae(Cambridge 1986). Seager, R, Pompey:A PoliticalBiography(Oxford 1979). Sherwin-White, A. N., "Rome, Pamphylia and Cilicia 133-70 B.c.," Journalof Roman Studies 66 (1976) 1-14. --, Roman ForeignPolicyin the East 168B.C. to A.D. 1 (London 1984). Tatum, W.J.,The PatricianTribune:PubliusClodiusPulcher(Chapd Hill 1999). Wallinga, H. T., "The Ancient Persian Navy and Its Predecessors," in AchaemenidHistory1:Sources,Structures, Synthesis,ed. H. Sancisi-Weerdenburg (Leiden 1987) 47-76. Willetts, R F.,AristocraticSocietyin Ancient Crete(London 1955).

--,

RECENT ARCHAEOLOGICALSURVEYAT PORTUS Simon Keay, Martin Millett, and Kristian Strutt

1. Introduction

'T1heRoman harbors at Portus are of central importance for any understanding of the maritime .1 economy of the Roman world. Although they have been the subject of antiquarian research since the sixteenth century, the enormous scale of the ancient structures and limitations on accessibility have meant that knowledge of the site has remained comparatively limited. The first published survey of the site was produced by Rodolfo Lanciani in the mid-nineteenth century. 1 His work was complemented by the studies of Lugli undertaken in the 1920s and 1930s. Although Lugli's results were comprehensively published, the volume was produced in a limited edition and did not reach a wide audience. 2 Extensive work on the Claudian harbor was undertaken during the construction of Fiumicino airport in the 1960s, and a general account of this work was produced,' together with a full report on the excavated ships.4 The Soprintendenza per i Beni Archeologici di Ostia has continued with a program of excavation since this date and most recently has been undertaking extensive topographic survey and on-site conservation in preparation for the opening of part of the site to the public.' Since 1998, at the invitation of the Soprintendenza per i Beni Archeologici di Ostia, we have undertaken an extensive geophysical survey covering all available areas of the site. The results of this work complement previous studies and provide new evidence for the development and organization of the harbor complex. In the present paper we summarize our approach and the principal conclusions. A full report on the project has recently been published, and this provides the detailed evidence on which the present account is based. 6 The survey was undertaken in collaboration with the Soprintendenza per i Beni Archeologici di Ostia and the British School at Rome, with financial support from them and from the Arts and Humanities Research Board of the United Kingdom. It has also enjoyed the support of the Universities of Southampton, Cambridge, and Durham. We also wish to acknowledge Duke Sforza Cesarini for allowing access to his property.

2. HistoricalContext Portus lies on the Tyrrhenian coast of Italy, just north of the mouth of the river Tiber and ea. 4 km north of Ostia (fig. 1). Ancient literary sources and an inscription dated to A.O. 46 (CIL 14.85) 1 Lanciani 1868. 2

Lugli and Filibeck 1935.

3 Testaguzza 1970.

4

Scrinari 1979.

5

Mannucci 1996.

6

Keay et al. 2006.

98

SIMON KEAY,MARTINMILLLEIT, AND KRISTIANSTRUIT

Fig. 1. Map showing the location of Portus on the Ty"h enian coast of the Italian peninsula.

-

0

.

20km

t

N

200

t

record that the complex was planned by the emperor Claudius in order to provide a harbor for seagoing vessels at the mouth of the Tiber. The project was intended to improve the supply of the City of Rome, which had hitherto been served only by the river port at Ostia , with many goods thus having to be carried over land from the harbors in the Bay of Naples. However , construction took a considerable period, and the inauguration of the complex was celebrated with a coin issue struck by Nero in A.D. 64 (RIC 1, nos. 178-183 ). Subsequently the complex was enlarged under the emperor Trajan as one of his actions to improve the infrastructure of Italy and to secure Rome's supply of grain . It has long been recognized that the harbor constructed by Claudius (the Porto di Claudio) can be identified with the remains of moles and quays in the northern part of the site, while the later extension (the Porto di Traiano) is represented by the hexagonal basin to the southeast . The problem in reconstructing the topographic development of the harbors in greater detail has always centered on the identification of the different canals known from inscriptions and establishing how they worked in relation to the harbor basins .

3. Methodsof Study Our work comprised a geophysical and surface survey that was undertaken at Portus between 1998 and 2004. The survey, commissioned by the Soprintendenza per i Beni Archeologici di Ostia, formed part of a broader research project, "Roman Towns in the Middle and Lower Tiber Valley," and included topographic survey, large-scale magnetometer survey, and the systematic collection of surface materials. The techniques deployed in the project have been reviewed elsewhere .7 They 7

Keay et al. 2004.

RECENT ARCHAEOLOGICALSURVEYAT PORTUS

99

Anoporto lnternnional~ dJ leourdo cb Vind

Porto di Oaudio

Porto di Tn.lano

-;

FON.lTr.t.laN

lsolas.cni

Portus

-

0.5

1km

Fig. 2. Overallplan showing the areasof Portus coveredby the geophysicalsurvey (magnetometer)discussedin this paper. Gray indicatesthe areaof geophysicalsurvey, with the darker shade indicatingthe presenceof high magneticanomalies.

enable large areas of ground to be covered very rapidly and are well suited to research on extensive archaeological sites like Portus. The magnetometry produced a plan of localized unidimensional magnetic anomalies generated by buried features, including construction material (fig. 2). These geophysical features were then interpreted in the light of knowledge of ground plans of known Roman building types. The results were incorporated into an existing cartographic base, enhanced with a new topographic survey. The geophysical survey does not provide any information about the chronology of the buried features, but dating evidence can be established by analyzing the spatial relationships among the features revealed. Surface materials were also collected systematically wherever the ground conditions were suitable in order to provide further information about occupation and abandonment dates as well as about the character of the buried deposits . The total survey area covered an area of ea. 173 ha and encompassed the southern and eastern parts of the Porto di Claudio, the areas around the hexagon of the Porto di Traiano and between it and the ancient coast to the west, as well as the extensive flatlands lying between the hexagonal harbor and the Tiber.

4. The Topographyof Portus: The ClaudianComplex Much of the Claudian harbor was unsuitable for geophysical survey. To the north it is partially obscured by the modern Aeroporto di Leonardo da Vinci di Fiumicino, while buildings of later phases in the Roman development of the site are superimposed over the area to the south, creating a complex building sequence around the Darsena. Furthermore, the construction of the Trajanic hexagonal basin seems likely to have destroyed parts of the earlier complex. Nevertheless, limited geophysical

SIMONKEAY,MARTINMILLLETI,AND KRISTIANSTRUTI

100

30

60

90

120m

12

~T

.12

Fig. 3. Geophysicalsurvey resultsof the areaimmediately south of the Darsena.The building identified as the "ForoOlitorio" is visible in the centralarea. Gray indicatesthe areaof geophysicalsurvey, with the darker shade indicatingthe presenceof high magneticanomalies.

survey, combined with observations of standing buildings and a reconsideration of earlier published accounts , allowed us to put forward a new reconstruction of this initial phase of the port. Key to understanding the Claudian harbor is the identification of the canals recorded as having been constructed in A.D. 46 in order to relieve flooding at Rome and to facilitate the construction of the port. Previous authors have suggested a variety of possible identifications . Our work leads to several key conclusions based on the premise that there were only two Claudian canals. One, in the northern part of our survey area and also visible on aerial photographs, lay immediately north of the Claudian harbor basin and would have carried water westward to the coast directly from the point where the Tiber bends sharply southward. The other links the Tiber with the sea immediately to the south of the Claudian harbor and is to be identified with the modern Canale Navigabile di Fiumicino-confusingly and erroneously called the Fossa Traiana by earlier commentators. 8 The presence of a statio marmorum dated to the later first century A.D. on its southern side confirms its pre-Trajanic date. 9 These two canals defined the area to be developed for the harbor , which was constructed with enormous concrete moles reaching out from the shore and an area of the ancient coast also being excavated to extend the basin inland . Our survey work revealed new details of the quays on the landward side of this harbor basin as well as new information about the structures on its southeastern side . To the south of the Darsena, a building identified by Lanciani 10 as the Foro Olitorio and excavated features in the 8

E.g., Lugli and Filibeck 1935; Testaguzza 1970.

9

Pensabene 2002.

10 Lanciani

1868.

RECENTARCHAEOLOGICALSURVEYAT PORTUS

101

area of the Portico di Claudio suggest that an orthogonally planned harbor town was developed during the second half of the first century A.D. (fig. 3). This flanked the so-called Fossa Traiana at the south and, at least by the later first century, was connected to Ostia by a road, the Via Flavia. These conclusions have allowed us to suggest a new hypothesis about the development of Portus in the first century A.D. The port began with the construction of two canals. This kept the area of the future harbor free of Tiber floods during construction. Next the great harbor basin was built. Its primary function was to act as an anchorage for sea-going ships while they unloaded their cargoes onto river barges for transport up the Tiber to Rome; so the port could have started to function in relation to Ostia even before other local shore facilities were fully developed. The inner harbor, known as the Darsena, and the canal connecting it to the outer basin and to the so-called Fossa Traiana, were then built, and there is some recent evidence to suggest that their construction took place in the Neronian period. Once this infrastructure was in place, the magazzini(warehouses) and other buildings could be built within the orthogonal grid. With these came the construction of the road link to Ostia. The survey has also located the line of the first-century aqueduct supplying this area of the settlement, as well as a series of funerary and other structures presumably facing onto a road that we believe ran beside the Tiber to the east of the new settlement. This activity illustrates the importance of the river as a focus for activity from this period. In trying to understand the operation of the first-century harbor, we should bear in mind first that Portus was subsidiary to Ostia at this stage (and indeed remained administratively so until the fourth century), and second that although the new construction was monumental in scale, it seems to have lacked much in the way of architectural elaboration or display. We thus envisage Portus growing first as an enormous haven connected by sea to Ostia, then with the development of harbor and warehousing facilities increasingly becoming an independent entity with goods being transhipped, stored, and transported by barge via the canals to the Tiber and upriver to Rome. Thus, although still linked to Ostia, it was by the end of the first century increasingly operating in its own right as the principal harbor for Rome.

5. The Topographyof Portus:The TrajanicComplex Previous work shows how the 33.25 ha hexagonal harbor constructed under Trajan, each side of which is ea. 358 m in length, created a major new focus for the complex. Our work adds detail to the understanding of the hexagon, but in addition it has provided key new information from the area of flat land between the hexagon and the Tiber. This provides a clearer picture of the scale of the Trajanic enterprise and clarifies how the complex may have functioned. The new hexagonal basin to the southeast of the Claudian harbor was probably designed as an inner harbor, perhaps a larger-scale replacement for the earlier Darsena. It also accommodated far more extensive storage facilities than before, with the construction of large magazzinialong at least five of its six sides. Its position with respect to the Claudian basin and Darsena also ensured that the earlier harbor could continue to function during the years of its construction. Once completed, the integrated complex of the Claudian outer basin and the Trajanic hexagon functioned as a unified whole. Its focus lay on side II of the hexagon, directly opposite the entrance via the Canale di Imbocco al Porto di Traiano. At the center of this side, directly on axis to the entrance, the geophysical survey revealed a rectangular temple within a temenos that is probably

SIMON KEAY,MARTINMILLLETI, AND KRISTIANSTRUTI

102

Fig. 4. Geophysicalsurvey resultsof the centralareaof the Trajaniccanalas it approachesthe Tiber to the east of the Trajanichexagon. The road runs parallelto its northern side, while the path of the aqueduct is marked by the line of dark dots beyond. Buildings can be identified between the canaland the road. Gray indicatesthe area of geophysicalsurvey, with the darker shade indicatingthe presenceof high magneticanomalies.

to be identified with the building that Lanciani describes as a temple to Liber Pater Commodiana, in front of which was supposed to have been a colossal statue to the emperor Trajan. 11 Most of the quays around the rest of the hexagon and beside the Darsena were occupied by huge warehouses. Our survey suggests that the entire ground-floor area occupied by warehouses of this period totaled ea. 92,000 m 2 • The form of the warehouses varies, but the majority are of the corridor type. The principal exception is on side VI of the hexagon in the area of the so-called Palazzo lmperiale. The survey provided no evidence to support this interpretation, although in the western part of the area major structures survive, including a cryptoporticus and the so-called Terrazza di Traiano. To the east there is also evidence for a monumental structure looking over the Claudian harbor basin and beside this a row of substantial courtyard warehouses with access onto both harbor quays. The character of the warehouses at Portus contrasts with those at Ostia, supporting the idea that they were principally associated with state-controlled supply. It can be no coincidence that the development of the harbor by Trajan coincides both with his reforms of the annona and also with Portus replacing Puteoli as the destination of the Alexandrian corn fleet. The survey also provided new evidence of the canal that is recorded to have been constructed in the Trajanic period (Plin. Ep. 8.17.1-2; CIL 14.88). Features had previously been noted in the area between the Tiber and the hexagonal basin, but the geophysical survey revealed a comprehensive replanning of this area with the construction of a new canal, ea. 40 m wide and ea. 1.4 km long, which connected the Fossa Traiana near the Episcopio with the Tiber. In its western stretch, this canal was constructed to run parallel with side III of the hexagon, so that goods in the warehouses here could be loaded directly onto canal barges. This arrangement ensured that the transshipment of cargoes and their movement upriver to Rome would have been much more efficient. Farther east the canal curved and then cut straight toward its intersection with the Tiber. Here it ran parallel with a major new road and the rerouted aqueduct (fig. 4). The overall effect of this must have been dramatic for visitors approaching the port from Rome and would have been 11

Lanciani 1868, 179-182.

RECENTARCHAEOLOGICALSURVEYAT PORTUS

103

an eloquent symbol of the imperial power represented by the port as a whole. The survey shows that the area between the road and the canal developed as a district of commercial activity. Major warehouses were found facing onto the road and extending back to the canal, while the surface collection yielded very high densities of imported African, Tripolitanian, and east Mediterranean amphorae associated with them. To the north of the aqueduct, the geophysical survey and surface collection also revealed the presence of extensive cemeteries of middle to late Imperial date. This implies that Portus was becoming a major center of population, although it has proved difficult to identify houses or apartments in the areas surveyed. It seems probable that the main focus of occupation lay in the area to the south and southwest of the hexagonal harbor, flanking the Fossa Traiana. However, the deposits in this area are known to be extremely deep and complex, and only a limited area was available for survey, so our evidence for the nature of this part of the settlement remains limited. There was also a zone of domestic occupation on the Isola Sacra, around the bridgehead where the Via Flavia crossed the Fossa Traiana. However, this seems to have been rather limited in extent. Flanking the Tiber, to the north of the canal entrance, a major complex of buildings subsequently developed. Important new evidence from aerial photographs provides details of the full layout of this complex. It comprises a series of temples and a series of magazzinithat opened onto the Tiber; elsewhere there are extensive mausolea. The position of this site close to the intersection of the canal and road with the Tiber suggests that it acted as a river port. However, the presence of large temples suggests that it may also have acted as a sanctuary, possibly marking a transition from the terrestrial to maritime spheres. It is clear that the Trajanic development significantly altered the character of Portus. The capacity of the harbor and storage facilities was greatly increased, and there is evidence for greater monumentalization. The epigraphic and sculptural evidence confirms the growth of a more cosmopolitan community and a much more urban landscape. We would interpret this as a culmination of the process by which Portus had started to become increasingly independent from Ostia during the first century A.O., but it is equally evident that the impetus for the second-century development comes from imperial initiative and is a direct consequence of the reorganization of Rome's food supply. Although Portus only became an autonomous municipality in the fourth century, the Trajanic development certainly established it as an independent entity and as the formal gateway to the imperial capital. The survey has provided evidence for the later development of the site and its continued use into the seventh century A.O., but the layout and functions seem to have been altered little after the second century.

104

SIMONKEAY,MAIITINMILLLETI',AND KRISTIANSTRUTI'

WorksCited Keay, S., M. Millett, L. Paroli, and K. Stnltt, PortusRomae:An Archaeological Surveyof the Port of Imperial Rome (London 2006). Keay, S., M. Millett, S. Poppy, J. Robinson, J. Taylor, and N. Terrenato, "New Approaches to Roman Urbanism in the Tiber Valley," in Bridgingthe Tiber:Approachesto RegionalArchaeologyin the Middle Tiber Valley,ed. H. Patterson (London 2004) 223-236. Archaeological Monograph of the British School at Rome 13. Lanciani, R, Ricerchetopografiche sullacittadi Porto(Rome 1868). Lugli, G., and G. Filibeck, It Portodi Roma imperialee I:Agro Portue~se(Bergamo 1935). Mannucci, V.,ed., It ParcoArcheologicoNaturalisticodel Portodi Traiano:metodoe progetto(Rome 1996). Pensabene, P., "Il fenomeno dd marmo nd mondo romano," in I marmi coloratideltaRoman imperiale,ed. M. De Nuccio and L. Ungaro (Rome 2002) 3-67. Scrinari, V. S. M., Le navi del Portodi Claudio(Rome 1979). Testaguzza, 0., Portus:illustra1.ione dei Porti di Claudioe Traianoe delta citta di Portoa Fiumicino(Rome 1970).

IMPORTS AT OSTIA IN THE IMPERIAL PERIOD AND LATE ANTIQUITY: THE AMPHORA EVIDENCE FROM THE DAI-AAR EXCAVATIONS Archer Martin

1. Introduction

I

n this paper I wish to examine trends in the importation of amphora-home foodstuffs (essentially wine, oil, and fish sauces) to Ostia from the first century A.D. to the fifth. I should stress at the outset that I willbe discussing the sum of these goods imported in amphorae and not necessarily the entire supply of them to Ostia. That is, appropriately for a conference dedicated to ports, I will be concerned only with that part of the supply imported through the port facilities of Ostia. This discussion is based on the preliminary classification (under my direction and with the help of Eric De Sena of the AAR's Archaeology Laboratory and a team of Italian, German, and American students) of the material from selected contexts excavated in an DAI-AAR project between 1998 and 2001. 1 The 37 sondages excavated were designed to provide a maximum of information, particularly with regard to dating, for elements identified through non-invasive surveying in the previously unexcavated areas of the city. Because of their purpose, they tended more to depth than breadth, which means that often the single stratigraphic units were of limited size. However, by putting together several units of similar date, we were able to create cumulative groups for a range of chronological horizons. 2 Ideally one would want to have a continuous series of horizons of the same length of time with similar amounts of material in each one. In practice we have three groups of contexts of roughly similar size (one for the second half of the first century A.D. and two less closely defined groups for the years from 280 to 350 and from 350 to 475) and a larger one (for the first half of the second century), with a gap between 150 and 280. Of course, as the study of the pottery progresses, we hope to approximate more closely the ideal, but even now we feel that our results are worthy of note. A particular strength of the present study is its methodological basis. All the sherds were retained and examined for attribution to fabric groups. This has allowed us to proceed with quantification in various ways: total sherd count and total sherd weight; the minimum number of vessels represented by the rim sherds and a maximum represented by the number of all diagnostic sherds minus joins; and finally the estimated vessel equivalent calculated on the basis of the percentage of the rims preserved. Indeed, the comparison of the results of the different quantifications forms one of the features of a paper that was presented at a conference at Barcelona on late antique pottery. 3 Here we have used an average of results expressed in percentages of all the quantifications. What is significant is that this material was collected and processed with the intention of undertaking statistical analysis 1

For the project and previous bibliography, see Martin et

al. 2002. 2 At the time the selection was made the excavators had not yet supplied the phasing of the stratigraphies. Therefore, we

were able to base it only upon the ceramic date of the single stratigraphic units, which will inevitably be superseded in some cases by other dating evidence. 3

Martin and De Sena 2005.

ARCHERMARTIN

106

Wine 46%

OiVF1sh Sauce 43%

Fig. I . Provenience of imports at Ostia: horizon I (50-100).

Fig. 2. Products imported at Ostia: horizon I (50-100). Eastern Mediterranean 0% Italy 0%

Eastern Mediterranean 7%

Southern Gaul 28%

Italy 58%

Southern Gaul 0%

lbena 73%

Fig. 3. Provenience of imported wine at Ostia: horizon I (50-100).

Fig. 4. Provenience of imported oil/fish sauce at Ostia: horizon I (50-100).

of entire assemblages, unlike the material from older excavations that are still reference points in the literature but were subjected to statistical analysis only as an afterthought once most of the body sherds and possibly others had been discarded without recording. We expect the DAI-AAR contexts to constitute the first major publication of material at Ostia to be processed in this way. For each horizon I present four graphs. First I show the percentages of amphorae of any sort arriving from the various regions in order to give an idea of their relative commercial weight at Ostia at the time in question. The second graph is intended to point out what the Ostian market was interested in importing by breaking down the amphorae according to their contents: wine; oil and fish sauces grouped together (because one cannot always be certain of the content, as some regions exported both); unknown. For the purposes of these graphs I take into consideration the primary contents of the amphorae, as there is no immediate way to identify possible secondary contents deriving from the reuse of vessels, which may not have been important in any case at such a major port as Ostia. The third graph illustrates the proveniences of the wine amphorae, and the fourth does the same for the containers for oil and fish sauce.

IMPORTSATOSTIAIN THE IMPERIALPERIODAND LATEANTIQUITY

Unknown 8%

Unknown 8%

Eastern Mediterranean 5%

. :-·· ·.·

107

-:-:- ·-:-· OiVFish Sauce 59% Iberia

44%

Fig.5. Provenienceof importsat Ostia: hon"zon2 (100-150).

Fig. 6. Productsimported at Ostia: horizon 2 (100-150). Eastern Mediterranean 0%

Eastern Mediterranean 15%

Southern Gaul 21%

. ... . . .... ... . . .. ....

.

.. . . . . ... .... . . ..... .. . ... .

.

. . . . . ...... .. ..... Iberia

....... Fig. 7. Provenienceof imported wine at Ostia: horizon 2 (100-150).

74%

Fig. 8. Provenienceof imported oil/fish sauceat Ostia: horizon 2 (100-150).

2. The Horizons The first horizon (figs. 1-4), attested by 3,367 sherds weighing 91.7 kg, covers the half century from A.O. 50 to 100. Then Italy provides just over a quarter of the amphora-borne products at Ostia and Iberia something more than a third, with southern Gaul and North Africa exporting lesser amounts and the eastern Mediterranean less yet (fig. 1). When we turn to the products, we see that there is a somewhat stronger interest in importing wine than oil or fish sauce (fig. 2). Italy is the provenience of the largest percentage of wine amphorae (fig. 3 ). Within Italy, Campania (represented by the typical black-sand fabric) and the rest of Tyrrhenian central Italy each account for about half, while a slight contribution comes from the Adriatic. Southern Gaul supplies a significant proportion of the wine amphorae, with Iberia and the eastern Mediterranean/Black Sea area exporting less and North Africa next to nothing. Oil and fish sauce is much less variegated in its provenience: about three-quarters from the Iberian peninsula and one-quarter from North Africa (fig. 4). The second horizon (figs. 5-8), with 7,991 sherds weighing 404.8 kg, dates to the first half of the

ARCHERMARTIN

108

Unknown

Italy

Unknown

3%

7%

3%

Southern Gaul

4%

OiVFish Sauce 66%

North Afnca

50%

Fig. 9. Provenienceof importsat Ostia:horizon 3 (280-350).

Fig. 10. Productsimported at Ostia:horizon 3 (280-350). Eastern Mediterranean 0%

Iberia

0%

::::'::::': ~ ~~ Southern Gaul Eastern Mediterranean 64%

Fig. 11. Provenience of imported wine at Ostia: horizon 3 (280-350).

~~~

Southern

'TIGaul

0%

14%

Fig. 12. Provenienceof imported oil/fish sauce at Ostia: horizon 3 (280-350).

second century. By now Iberia has gained significantly and North Africa to a lesser extent as suppliers of amphora-borne products , while Italy and southern Gaul both have declined (fig. 5). The change in the products imported is striking: wine stands at only a third , and oil and fish sauce reach nearly 60 percent (fig. 6). Looking at the proveniences of wine amphorae, we see little change, however, essentially only a decline in wine from Iberia and an increase in that from the eastern Mediterranean/Black Sea area (fig. 7) . The data for Italy hide an important shift: now Campanian containers are rare at Ostia, undoubtedly because of the destruction caused by the eruption of Vesuvius. The provenience of the oil and fish-sauce containers remains the same as in the previous horizon: ea. three-quarters from the Iberian peninsula and one-quarter from North Africa (fig. 8). The third horizon (figs. 9-12) picks up the story between 280 and 350, after a gap of more than a century. Here 2,610 sherds weighing 169 kg were recovered . The proveniences of the amphoraborne products are quite different (fig. 9). Now North Africa accounts for half of them , while the formerly dominant Iberian peninsula declines to 16 percent. The percentages for Italy and southern Gaul continue to decrease. Those for the eastern Mediterranean, however, rise significantly. What

IMPORTSAT OSTIA IN THE IMPERIALPERIOD AND LATEANTIQUITY

109

01VF1shSauce

68%

North Africa 61%

Fig. 13. Provenienceof importsat Ostia:horizon4 (350-475).

Fig. 14. Productsimported at Ostia: horizon 4 (350-475). Eastern Mediterranean Italy 0% 0%

~::=o---1""----

lbena

9%

Eastern

,ft

Italy 44% :.::_,=:_,.,=,:,:_:,,_:.

'::;:}'.-

·=t=·=·

=.,:=_,·_,i,=_,_:',',. .:

;:::=:;-

·t::.·

91% 0%

Fig. 15. Provenienceof imported wine at Ostia: horizon 4 (350-475).

Fig. 16. Provenienceof imported oil/fish sauceat Ostia: horizon 4 (350-475).

does not change with respect to the first half of the second century is the strong interest in importing oil and fish sauce, containers for which make up two-thirds of the sample (fig. 10). Now the eastern Mediterranean/Black Sea area is by far the most important supplier of wine in amphorae, largely with Kapitiin I and II amphorae of unknown but probably Aegean or Pontic origin (fig. 11). Italy itself accounts for the second largest percentage of wine containers, mostly small flat-bottomed amphorae from Tyrrhenian central Italy. Amphorae from southern Gaul still make up a significant part of the sample, although they are essentially residual, as their production ends at the beginning of this horizon. The supply of oil and fish sauce is inverted with respect to the previous horizon: now three-quarters of the containers come from North Africa and one-quarter from the Iberian peninsula (fig. 12). Horizon 4 (figs. 13-16) goes from 350 to 475, with 2,165 sherds weighing 90 kg. Here North Africa is even more important as a provenience than in the previous horizon (fig. 13). The eastern Mediterranean/Black Sea area is the next most important, although it has lost ground with respect to horizon 3. Italy, on the other hand, has increased somewhat . The Iberian peninsula has declined

110

ARCHERMARTIN

and southern Gaul disappeared from the picture. Oil and fish sauce still account for by far the largest percentage of the amphora-home products (fig. 14). Somewhat more than half the wine imported in amphorae comes from the eastern Mediterranean and rather less than half from Italy, with a minimal amount from the Iberian peninsula (fig. 15). The eastern Mediterranean/Black Sea area no longer has one major supplier hut rather a range of amphorae of different origins, of which Palestinian Carthage LRA 5-6 are the most important, followed by Kapitan I and II amphorae and Carthage LRA 3 from western Asia Minor. In this last horizon North Africa supplies almost all the oil and fish sauce imported in amphorae (fig. 16). The only exception is a small quantity from Lusitania.

3. Comparisons The publications of the amphorae from the Terme del Nuotatore form the standard point of reference for the class at Ostia. 4 Although the Terme del Nuotatore excavations were not designed for quantification of the ceramic material, subsequent quantification of the diagnostic pieces has provided insight into the trends in the supply of amphora-home goods at Ostia. This evidence is particularly welcome because it sheds light on the period that is lacking in our evidence. For layers between the Flavian and Severan periods the Terme del Nuotatore material has been presented in various syntheses. There is no equivalent synthetic treatment for the amphorae from the main layers documenting the second half of the third and the fourth centuries at the Terme del Nuotatore, although full details have appeared. The provenience of the amphorae attested at the Terme del Nuotatore has been calculated by geographic regions similar to this study with similar results.5 For the Flavian period the three most important sources of amphorae are in rising importance Italy, Gaul, and the Iberian peninsula, each with about one-quarter, with the North African sources in a decidedly secondary position. In the Hadrianic period there are two important proveniences: Gaul and the Iberian peninsula, while the African sources rise to the third position and Italy falls back to fourth, with Aegean amphorae making their first significant appearance. By the late Antonine period the Iberian peninsula and the North African sources are the two major groups, with Italy and Gaul still important secondary ones and the Aegean remaining as a lesser presence. The late Severan period shows the North African dominating by far, with the Aegean in a clear second place and all other proveniences of minor importance. In the fourth century the North African sources are even more important, while the Aegean continues as a decidedly less important second and the other proveniences are insignificant. The major difference between the Terme del Nuotatore evidence and the comparable horizons in our series is a greater importance of Gaulish amphorae at the Terme in the Flavian and Hadrianic periods. The fourth-century picture for the Terme del Nuotatore overlaps in part with our horizon 3, to which it is more similar, and to a lesser extent with our horizon 4. The latter shows a diminished importance for the Iberian peninsula and an increased one for Italy, essentially accounted for by Keay LII amphorae, which begin to be distributed during the fourth century. The Terme del Nuotatore material agrees with ours in showing an increasing interest in importing 4 The

basic bibliography is: Palma and Panella 1968; Panella 1970; 1973a; 1973b; 1977a; 1977b; Pavolini 1977a; 1977b. For Strato I in Area XVI, dating to the fourth century, the most useful for comparison for this study in Ostia IV, see the table appended to this article with the reelaborated data from the table at Manacorda 1977a, 241-244, with the addi-

tion of the data from Pavolini 1977a. Other considerations of the amphora material from the Terme de! Nuotatore will be cited where relevant. 5

See Panella 1986, fig. 1, 610, for a synthesis.

IMPORTSATOSTIAIN THE IMPERIALPERIODANDLATEANTIQUITY

111

oil.6 The value for definite oil containers rose from less than 10 percent in the Flavian period, to more than 20 percent in the Hadrianic period, to almost 30 percent in the Antonine period, to more than 40 percent in the late Severan period, and to somewhat more than half at the end of the fourth century. The Terme del Nuotatore evidence for the provenience of the wine imported to Ostia corresponds to what our evidence shows. Gaul was the major exporter of wine to Ostia in the Hadrianic and Antonine periods, with some coming from the Aegean. 7 Italian wine was of relatively minor importance and derived from regions previously not concerned with the trade (the Tiber valley and Emilia) rather than Tyrrhenian central Italy.8 In the late Severan period Gaulish wine was declining in importance but still well represented, and eastern wine containers continued to be attested. 9 In fourth-century contexts wine amphorae from the East make up the bulk of those recognized. Otherwise North Africa supplied some wine in containers from Mauretania and possibly some of the series of small, flat-bottomed vessels. The situation with oil and fish sauce containers from the Terme del Nuotatore also agrees in general with our material. The Iberian peninsula is the main supplier of imported oil and fish sauce until toward the end of the second century, although African products constituted a strong secondary presence, with North Africa taking over the dominant position after that. 10 In the fourth-century contexts North Africa is practically unrivaled. The early fifthcentury layers from below the Casone del Sale (current museum and office building) at the center of the city belong to part of the range covered by our horizon 4.11North Africa is the provenience of some 45 percent, while the East supplies nearly one-quarter. The only other significant contemporary source of amphorae is the area of the Strait of Messina, with somewhat less than 3 percent. As the African amphorae probably all contained oil or fish products and most of the eastern amphorae and those from the Strait of Messina served for wine, the relative interest in importing oil or fish products remained constant at about 60 percent. The contexts dating from the first half of the fifthcentury until ea. 450/470 from the extramural basilica at Pianabella correspond to the latter part of our horizon 4 .12 The three important groups are the same. Africa leads by far with 57 percent, but the containers from the Strait of Messina are in second position at 12 percent, while the eastern amphorae come to 4.5 percent. Here there appears to be an even greater interest in importing oil and fish products than at the other late contexts examined.

4. Conclusion The shifts in the proveniences of the amphorae in these horizons correspond to trends that have become well known through various studies documenting the origin of amphorae imported to Ostia. 6 See the data presented by Panella 1983, figs. 49-50, 260-261, for selected oil containers (those most securely attributed to that function, such as Dressel 20) against containers of other contents (i.e., wine and fish sauces, as well as less-known ones, including possible oil).

7 Carandini and Panella 1981, 491-492, 494; Panella 1991, 290.

10 Carandini and Panella 1981, 492-494, 496-497, 498-499 seem to indicate that African amphorae for these products were increasing from the Antonine period through the early Severan period to become dominant in the late Severan or Gordian period; Panella 1983, figs. 49-50, 260-261 shows Baetican oil as second to African in the Antonine period; Panella 1991, 290-291 also places Baetican oil in the second position with respect to African in the Antonine period.

8 Panella

11 Martin

9

1991, 290.

Carandini and Panella 1981, 498,500.

1993, 207-208; Martin 2005, 62--63.

12 Ciarrocchi

1993.

112

ARCHERMARTIN

Amphorae from Tyrrhenian central Italy give way first to ones from the Iberian peninsula and then to North African ones, with a moderate rise in eastern containers in late antiquity. Thus, our data offer merely a confirmation with the comfort of more solid quantification. The change in emphasis from importing wine to importing oil and fish sauce has perhaps not been brought so sharply into focus as in this study. It is, however, a phenomenon with interesting implications. The time of the change according to our data, between the second half of the first century and the first half of the second, corresponds approximately to the end of the production of the large wine amphorae of Tyrrhenian central Italy associated with the slave-run villas that had flourished there for several centuries, which still dominate our first horizon. Discarding the highly unlikely hypothesis of a sudden drop in demand for wine with respect to oil and fish sauce but rather assuming it to have remained relatively constant throughout the period from the first to the fifth centuries, one is faced with accounting for the origin of the wine to replace what previously was supplied by the villas of Tyrrhenian central Italy. Here one must remember, as was said at the beginning, that this paper deals only with produce carried in amphorae. The difference must have been taken up by wine brought to market in some other fashion. 13 In the last few decades it has often been recalled as a general note of caution that wine and oil could be and were undoubtedly produced in most parts of the Mediterranean world and that not all ended up transported in amphorae. There is literary and iconographic evidence for barrels and skin containers, and wine and oil moving a short distance within a local market could simply have been placed in miscellaneous containers, much as vino s/usostill is today. Our statistics suggest that from about A.D. 100 a greater percentage of Ostia's wine supply reached it in some such way.

13 See De Sena 2003 for methodological considerations and an attempt to estimate how much of the need for oil the

hinterland of Rome and Ostia could supply (about one-fifth on average).

IMPORTS AT OSTIA IN THE IMPERIAL PERIOD AND LATE ANTIQUITY

113

Appendix : Strata I of Area XVI at the Terme de/ Nuotatore

The table below presents the data on the amphorae from Strato I of Area XVI at the Terme del Nuotatore according to current denominations and grouped geographically but without any checking of the material itself. 14 Identification in Os tia IV

Current Denomination

Rims

Handles

Bases Bodies

Italian Peninsula/Sicily

Totals

Ostia III, 380

Dressel 1

2

3

Os tia Il , 554

Dressel 2-4

4

4

0 tia I, 451

Forlimpopoli Amphora/Ostia I, 45 l

Ostia I, 455

Ostia I, 455

2

11

11

48

50

Iberian Peninsula

0 tia IV, 109

otes

(,8

70

Dressel 2-4 tarraconen is

Ostia I, 555

Beltran IIB

Ostia III , 644-645

Dres el 14 simili /Beltran IVB

0 tia I , 513-5 15

Dressel 20

0 tia I, 516

Dr essel 20

Os tia I, 565

Dressel 23

Ostia I, 546

Almagr o 50

stia IV, 255

Almagro51C

Ostia IV, 257

Almagro51C

Ostia I , 549

late Dressel 28

Ostia IV, 291

late Dressel 28

3

3

2 17

26

43 2

2

6

12

6

2

2 2

80

Gaul

0 tia I, 535/Ostia III, 528

Pelichet 47 / Gauloise 4

7

0 tia I, 538

P elich et 47/ Gauloi se 4

3

3

0 tia I , 537/ Ostia Ill , 121

Pelichet 47 / Gauloise 4

2

2

Ostia I, 542/ Ostia III, 264

P elichet 47 / Gauloise 4

Ostia l , 535-541

Pelichet 47 / Gauloise 4

Ostia Il , 552

Nijme gen 132B/ Gauloi se 5

7

LO

10 55

55 3

3

..[m] they have thrown in, by which sign they also sound different places in the seas [but in Pontos the springs wash the sand off the sounding weights].

The relevant passage in Aristotle (Mete. l.13.351a) does not mention collection of a sample of the bottom . . . . the so-called Deeps of Fontus. This is a certain part of the sea the depth of which is immeasurable. No one, at any rate, letting down a line has been able to find the bottom.

Pliny the Elder (HN 1.224) echoes Aristotle. 23 GreatBritain,Admiralty 1922, 130; cf.Waters 1958, 19-20. May and Holder 1973, 203-204. 24

Motzo 1947, 2:29-30; Taylor 1956, 107.

2'

On diving in antiquity, see Oleson 1976, 22-29; Nardi 1984--1985, 51--63; Tchernia 1988, 489-499. E. de Bruijn of the University of British Columbia is preparing a Ph.D. dissertation concerning diving in antiquity.

THE ROLE OF SOUNDING WEIGHTS IN ANCIENT MEDITERRANEANNAVIGATION

127

Others report an immense depth of water, called the Deeps of Pontus, off the coast of the Coraxi tribe on the Pontus, nearly 300 stades (54 km) from land, where soundings have never reached bottom.

The greatest depth in this part of the Black Sea has now been measured at 2,155 m.26 The passage in the Meteorologica is interesting because it implies that deep sounding was a routine activity and that there was already a significant corpus of depth soundings available for the Mediterranean and Black Sea in the fourth century B.C. But why would one take soundings in deep water 54 km from land unless the activity was part of a larger program of scientific research? Herodotus (Hist.2.28) mentions an attempt by the pharaoh Psammetichus I to sound the sources of the Nile with a line carrying a sounding lead. Given the date of his reign--656-610 B.c.-this is the earliest testimony for the use of sounding weights, and in this case, too, the purpose was scientific curiosity rather than practical navigation. King Psammetichus ... had a rope made many thousands of fathoms long which he let down into the water without finding the bottom. It seems to me that if there is any truth in this scribe's story, it indicates the presence of powerful whirlpools and eddies in the water ... which prefrom reaching the bottom. vented the sounding weight [xcx-rcx1tEtQl'}'tl'}Q!l'}v]

Invisible whirlpools and eddies that thwart attempts at sounding appear in most accounts of the Greek and Latin literature that deals with attempts at measuring the depth of mysterious rivers, lakes, and seas, and these hazards represent real conditions (see below). Some Roman emperors showed similar curiosity. Pausanias (2.37.5), for example, reports that the emperor Nero, like Psammetichus I, used a sounding lead tied on "several lengths of rope" to plumb the depths of the Alcyonian Lake near Lema, but that he was unsuccessful. The depth of the Alkyonian Lake has no limit, and I do not know of anyone who has been able to plumb its limits with any device [ou8eµt~ µ11xcxvn]. Even Nero, who had ropes made many stades long and tied end to end, fastening a sounding lead [µ6>..u~8ov]to them and omitting nothing that might hdp his experiment-even he was not able to find any limit to its depth.

Finally, there is a similar tale about the emperor Hadrian in the Talmud (Midrash Psalms 93 .6, 415-416): "Once, Hadrian the emperor wished to measure the depth of the Adriatic [Sea]. He took ropes and lowered them [into the sea] for three and a half years ... "27 Sperber's glossary of Greek and Latin nautical loan words in Talmudic literature, however, does not contain any terms for sounding weight. 28 Herodotus unfortunately does not indicate the material of the weights he mentions, and the more or less "test below" -is unspecific. The term used by OlympioGreek term-xottot1tEtQYj•YjQ!Yj, dorus regarding the passage in Aristotle-~6i..u;;, "missile"-is also unspecific. The earliest clearly identifiable sounding weight in the archaeological record, from a ship that sank at Gela around 500 B.C., about fifty years before Herodotus wrote his history, is a flattened hemisphere with slightly flattened sides and shallow tallow cup, made of lead (cat. no. 002, fig. 1). A rectangular lead weight with a suspension hole at one end, and about the same size as later sounding weights, was found on the Ulu Burun Wreck (cat. no. 001, 1325 B.C.). 29 Given the absence of a tallow cup, however, and 26 Leier 2001, 234-235.

28

27 Quoted

29

in Sperber 1986, 110.

Sperber 1986, 128-158.

Pulak 1988, 33, no. KW 267, fig. 41; Parker 1992, 439-440.

JOHN PETEROLESON

128

the long chronological gap between this wdght and the Gela sounding lead, identification of it as our earliest sounding weight is problematic. The first literary mention of lead as the material of a sounding wdght, a verse by the poet Lucilius quoted in Isidore (Origines19.4.10), belongs to the mid-second century B.C. Catapirateslineacum massaplumbea,qua marisaltitudo temtatur.Lucilius(1191): Rune catapiratempuer eodem devoret3°unctum Plumbipaucillumruduslineiquemataxam. A sounding weight, a line with a lump of lead, with which the depth of the sea is tried. Lucilius (1191):

"A slave boy would swallow down this sounding weight if it were smeared with the same stuff, A small lump of lead and a flax rope."

Although the context is not preserved, Lindsay suggests that "the delicious flavour of some syrup or paste is bdng described. "31 Description of the wdght as "smeared" or "anointed" strongly suggests that Lucilius was familiar with the practice of applying a greasy substance to the tallow cup on a sounding lead. The remaining Greek and Latin literary testimonia provide little further information. Plautus, about a generation before Lucilius, compares the slave of a love-struck master to a lifeboat (Aulularia598): ut toleret,ne pessumabeattamquam "[he is meant] to hold him up, so he doesn't sink like a ."

Unfortunately, the crucial word for sounding wdght is corrupt. The text of a passage in Statius (Silv.3.2.30) concerning tasks on board a boat also seems hopelessly corrupt. One popular reading provides this text: tSint quibusexploretplumbo gravisalta molybdust "Let there be some to make the heavy sounding-lead explore the depths ... "

Other editors, however, come up with solutions that have nothing whatsoever to do with sounding weights. 32 Herodotus, Lucilius, and Plautus all use essentially the same word for sounding weightM, por1bolts or deck beam.J

Fig. 7. Ship graffito from trench BE9 5-4, locus 045, pottery bucket no. 64 (2669-G) from A.O.

Fig. 8. Ship graffito from trench BE95-4, locus 045, pottery bucket no. 64 (2669-G) from A.O. 50-60/70 (drawing D. E. A. van Zi;l).

50-60/70 (photo author).

§ 0

"'II

...

1 Fig. 9 (left above). Teakwood reused as leveling course in late Roman wall in trench BE97/98-16 looking south (photo author). Fig. 10 (left). Detail of fig. 9 (photo author). Fig. 11 (above). Teakwood (over 3 m long) reused in late Roman religious shrine in trench BE98/99-23 looking north (photo author).

BERENIKE

·t

(

.... 'b

Early Roman seawall

In trench 2

,··· Ptolemalc : : _bulldl,ng

Early Roman tiarbor

:~\

2oome1ara

.,

SOUTHHARBOR ,,.---

...-... __ ____.,---,,

.

./.

""'\

possible harbor ,' monument ,/

)

,"

Fig. 12 (above). Plan of Berenike highlightingthe putative Ptolemaicharborand (from north to south at the eastern end of the site) Roman period seawallsand piers in trenches 17, 22, and 2 (drawingA. M. Hense).

Putative harbor area

0

1-------+----

Surveyed •nd dnwn by: H. Barnard and 'Ayu AhlMd N•f •'• Ffi>n,ary2001

Fig. 13. Detail plan of putative Ptolemaic harbor (drawing H. Barnard).

100m .

ARCHAEOLOGICAL EVIDENCE FOR SHIPS AND HARBOR FACILITIES AT BERENIKE, EGYPT

313

3. ArchitecturalEvidencefor HarborFacilities PTOLEMAIC

Fieldwork has documented little architectural evidence for harbor/port facilities at Berenike. In the latest season of excavation at the site in winter 2000-2001 surveying recorded a series of "arms" extending from a semicircular berm that formed a southward-facing enclosure strongly reminiscent of a harbor (figs. 12-13). Atop the berm and projecting "arms" were piles of unworked vesicular basalt, a stone not indigenous to Egypt and one frequently imported for use as grinding stones from the northern or southern portions of the Red Sea or, perhaps, from farther afield. It is likely, in this instance, that they were used as ships' ballast. 30 Perhaps upon arrival, after the ships' contents had been unloaded and before departure, the ballast was emptied to make room for outbound cargoes. The entire area had silted up, and a few later Roman buildings had been erected in the area. Previously no particular notice had been taken of this feature. The current hypothesis is that it might be the Ptolemaic, and/ or possibly the early Roman, harbor. When excavators are allowed once again to work at Berenike, this putative harbor area will be subject to excavation to confirm its function and dates of use. This massive "harbor" feature and the associated basalt/putative ballast lay immediately southeast of the early to mid- (mid-third- to mid-second-century B.c.) Ptolemaic-era industrial area noted above, where ship-related items were made. It would have been logical to place the harbor facility and workshops making items used in ship repair/refurbishment in close proximity to each other. Location of the putative harbor on the southern side of the city would also have protected it somewhat from the strong prevailing northerly winds of the region. EARLY ROMAN

Local geological conditions caused by harbor silting necessitated that the early and late Roman cities of Berenike be situated farther east and somewhat south of the Ptolemaic settlement. 31 Similar silting problems also troubled the port at Quseir al-Qadim. 32 Excavations unearthed two architectural features from the early Roman period at Berenike that are or appear to be harbor-related (fig. 12). From north to south along the eastern edge of the visible ruins are a seawall made of coral heads and ashlars (in trench BE97/98-17, loci 110/116/125) (figs. 12 and 14)33 and another seawall made of giant limestone boulders (in trench BE94/95-2, loci 054a-b) (figs. 12 and 15).34 These features are all approximately contemporary (early Roman, late first century B.c./first century A.D.)and provide a limited glimpse of the configuration of the early Roman port. These combined with the results of Masser and Thomas 2001, 13 (probably thatching for roof and wood doorstop); 15 (fig.4.2), 16 (fig.43). For possible Islamic ship timbers in Islamic graves, see Blue 2002, 149; Macklin2006, 158.

bor at Berenikehas shown that 100percent of those samples derived from the South Arabian port of Qana' on the Indian Ocean coast of Yemen.For this, see Peacock, Williams,and James 2007, 59.

30 Williams-Thorpe1988,286 (followingRouge1981,68-69)

31 See Harrell 1996, 105-106, 112-126 on harbor silting at Berenike.

indicatesthe use of basaltmillstonesas ballastin ancientMediterranean ships. For all manner of ballast:boulders, pebbles, beach rock, clay,and sand (but not metal ballast),usuallycollectedfrom the shore rather than ballastpiles,seeParker 1992, 28. I wishto thank ProfessorJ. A. Harrell and Dr. L. Casson for these references.Further for stone as ballast, see Casson 1971, 176and n. 43 and219; d. Harrell 1996,109. Since this paper was written, geochemicalanalysisof the basalt ballast from the putative Ptolemaic/earlyRoman har-

32 Blue2002, 142, 144-147, 149;2006,43-(,l;Thomas2006a,

90; Peacock and Blue 2006b, 174-176. 33

Sidebotham 2000, 74-75.

34

Sidebotham 1996b,25.

STEVENE. SIDEBOTHAM

314

Fig. 14. Trench BE97/98-17, loci 110/116/125 (sea wall in northern harbor) looking northwest (photo author).

Fig. 15. Trench BE94/95-2, loci 054a-b (limestone sea wall) looking southwest (photo author).

a magnetometric survey indicate that there were at least two bays or inlets, one on the north side of the site 35 and another on the south. Whether there were others and whether the northern inlet accommodated only smaller tender vessels associated with unloading larger ships or small coastalhugging Red Sea "tramp" steamers and the more substantial southern harbor served only larger longer-distance merchantmen can only be suggested and not proven. Parallels for the early Roman harbor facilities at Berenike may be found at the ongoing excavations of Quseir al-Qadim, where an early Roman pier (early first century A.D.) has been found, built, however, on Nile, Dressel 2-4 (Italian), Rhodian, and Dressel 6 (Istrian peninsula) amphoras dating late first century B.C. to early first century A.D. 36 The port at Quseir al-Qadim was very primitive and not at all monumental,3 7 as seems to have been the case at Berenike. The amphoras described above from Quseir al-Qadim were part of a land reclamation project similar to what took place at Berenike all along the eastern edge of the site in the late Roman period (cf. trenches BE95/96/97-5, BE96/97 /98-12, etc., noted below). 38 The literary sources and some of the ostraca that have been found in the early Roman trash dumps at Berenike may provide some insights as to when some of these first-century building 35

36

Herbich 2007.

Bradford 2000, 44; 2001; Whittaker 2002, 38; Blue 2002, 145, 148 (fig. 8); 2006, 47; Blue and Peacock 2006, 70, 74; Thomas 2006a, 90.

37 For the primitiveness of the port at Quseir al-Qadim, see Whitcomb and Johnson 1979; 1982, passim; Sidebotham 1986, 56; Peacock et al. 1999; 2000; 2001; 2002; 2003; Blue 2002; Peacock and Blue 2006a, 66. 38

Sidebotham 1999, 93; 2006b .

ARCHAEOLOGICAL EVIDENCE FOR SHIPS AND HARBOR FACILITIES AT BERENIKE, EGYPT

315

Fig. 16. Vespasianic inscription of A.D . 77178 recording the hydreuma ... praesidium et lacus at Siket, 7.2 km west-northwest of Berenike and just off the main route between Berenike and the Nile (photo author) .

Fig. 17.Squeeze of the text from the inscription in fig. 16 (drawing A. M. Hense).

Fig. 18. Artist 's rendering of placement of the inscription from fig. 16 in the gate of the praesidium at Siket (drawing A. M. Hense).

activities took place. Strabo (Geography 17.1.45) indicates that Berenike had convenient landing places, suggesting that it was an anchorage or roadstead, not a port with a proper harbor, in his day. By Pliny's time (A.O. 50s-70s) (HN 6.26 .103) it seems that there was a proper port at Berenike. Thus, we might extrapolate that sometime in the first three-quarters of the first century A.O. a proper harbor was constructed at Berenike. 39 The Nikanor Ostraca(late first century B.C. to A.O. 60s40 indicate a spurt of trade activity in the reign of Tiberius, and it may be that major renovations/enlargements to augment, or more likely replace, the silted up Ptolemaic facilities took place or were completed sometime during his reign at Berenike. 41 Numerous Tiberian coin finds in India and the reintroduction of the tetradrachm coin in Egypt during the reign of Tiberius all point to increased mercantile activity during his reign,42 as does the presence of a number of Tiberian -era graffiti on the road joining Quseir al-Qadim to the Nile. 43 Numerous ostraca excavated from the early Roman trash dump and inscriptions erected at a number of praesidiaon the road between Berenike and Coptos (figs. 16-18) also suggest that the 39

Sidebotham 1986, 52.

40

Tait 1930; Rostovtzeff 1931; Fuks 1951; Meredith 1952, 104.

41

First suggested by Sidebotham 1986, 52-53.

42

Sidebotham 1986, 27-29, 52-53; Turner 1989, 10, 123

(table 2), 124 (table 3 ). Bernand 1972, nos. 41, 42, 43, 45, 46, 47; nos. 38, 44, 48, 49, and 95 are possibly Tiberian; from the Paneion at el-Boueib (south of the Myos Hormos-Nile road): nos. 143, 144, 145, and possibly 142 and 146. See also Sidebotham 1986,55 . 43

316

1 2 3 4 5 6 7

STEVEN E. SIDEBOTHAM

road to the north and the Nile valley Late Roman cemetery Ptolemaic quay Ptolemaic industrial area demolished Ptolemaic building Palmyrene sanctuary Serapis temple

8 possible lseum 9 street with shops 1O possible temple 11 Christian basilica 12 main street 13 house with monumental entrance 14 warehouse 15 possible lighthouse

Fig. 19. Artist 's view of late Roman Berenike locating/eatures noted in the text (drawing A. M. Hense).

Julio-Claudian (Tiberius-Nero, A.O. 14-68) and Flavian periods activity at Berenike and along the route joining it to the Nile. 44

(A.O .

69-96) were ones of robust

MIDDLE ROMAN

Archaeological evidence, or rather lack of it, points to a decline in operations at Berenike through out some of the second, third, and early fourth centuries A.O. As known from recent excavations at Quseir al-Qadim, there was continued vibrant activity there in the second century, especially in the Trajanic-Antonine periods, 45 suggesting that , for whatever reason, Myos Hormos seems to have played a larger role than Berenike in the "Eastern" commerce in the Red Sea at that time . This situation did not last too long, as archaeological evidence from Myos Hormos indicates that the port 44

See Kennedy 1985; Bagnall, Helms, and Verhoogt 2000a, 7 and from R.S. Bagnall, personal communication, for JulioClaudian and Flavian dates of the Berenike ostraca; Bagnall, Bulow-Jacobsen, and Cuvigny 2001 for Flavian inscriptions from sites on or near the Berenike-Nile road from Siket (23° 55.88' N/3 5° 24.46' E), Khawr ej-Jer (Aphrodito at 25° 36.36' N/ 33° 37.38' E) , and Khashm el Menih/ Zeydun

(Didyme/ Didymoi at 25° 45.30' N/ 33° 23.58' E).Although serving a quarry near Mons Porph yrites, the small station at Umm Balad (at 27° 08.78' N/ 33° 17.43' E) is also Flavian in date: Dr. H . Cuvigny, personal communication. 45

Tomber 2001, 43; 2002 , 57; personal communication.

ARCHAEOLOGICAL EVIDENCE FOR SHIPS AND HARBOR FACILITIES AT BERENIKE, EGYPT

317

there ceased to operate sometime in the third century A.o.46 Thereafter, in the late Roman period, Berenike, Clysma, and Aila were the dominant trade emporia at the northern end of the Red Sea and Adulis toward the southern end. LATE ROMAN

Ironically,while excavations have uncovered more oflate Roman Berenike (mid-fourth-sixth century A.O.) than any other period of the city's history, no positively identified architectural evidence has been found at all for what port facilities might have looked like at that time, aside from a possible quay or pier made of coral heads and ashlars found beneath the eastern end of a fifth-century church in trench BE98/99-22, loci 106/116(figs. 12 and 19, no. 11).47 A late fourth/early fifth-century A.O. warehouse at the extreme southeastern corner of the site in trench BE95/96/97-5 (fig. 19, no. 14) suggests that the southern harbor was still operational at that time. 48 The find of an Axumite aes coin of the late third/early fourth century49 and an Indian silver coin of the second half of the fourth century,50 as well as recovery of quantities of Axumite pottery,51 Indian and Sri Lankan beads, 52 a bead fromJatim on the eastern end of the island ofJava/ 3 Indian textiles, 54 coconut,55 black pepper, 56 teakwood, 57 bamboo, 58 Job's tear,59 abrus precatorius seeds, 60 rice,61 sesame,62 and sorghum, 63 all indicate ongoing commercial contacts with other regions in the Red Sea/Indian Ocean in late antiquity. J. G. Wilkinson's plans of Berenike drawn in 182664 indicate a mound on the eastern end of a peninsula that enclosed Berenike's southern harbor (fig. 19, no. 15). This area is, today, extremely dangerous due to the presence of land mines and cannot be examined, but the mound would have been a good location for a small temple, a large altar, or another edifice associated with coordination or monitoring of ship movements in the southern harbor. Where documented in the Mediterranean, 46

Johnson 1979, 67; Whitcomb and Johnson 1982, 8; Tomber 2001, 43; 2002, 57; 2004, 351; Peacock and Blue 20066, 174.

1999a, 302 (table 16-1); 19996, 54-56;2000, 306 (table 13-1), 309; Sidebotham 20076. 57

47

Sidebotham 20076.

48

Sidebotham 19966, 76-82.

Vermeeren 1998, 334-335 (table 16-2), 337 (table 16-3), 343,347,348; 1999, 311,318 (table 17-11), 321-324; 2000, chart on 315-328, 329,334,335,337 (table 14-10), 339 (table 14-11), 340-341, 342; Cappers 19996, 59.

49

Sidebotham and Wendrich 2001-2002, 41 (fig. 40); 2002, 30 (figs. 9-10); Sidebotham 2007a, cat. no. 114 (C4615).

58 Vermeeren 2000, 318 (no. 49), 321 (no. 96), 325 (no. 153), 328, 335, 340, 342.

50

Sidebotham and Wendrich 2001-2002, 41 (fig. 40); 2002, 30 (figs. 9-10); Sidebotham 2007a, cat. no. 115 (C5649).

59 Cappers

51

Tomber 2007.

60

Cappers 2000, 307.

52

Francis 2000, 221-223; 2007.

61

Cappers 19986, 82; 19996, 56-57; 2000, 306.

1996, 331-332; 1998a, 289,299 (table 15-3), 303 (table 15-6), 311; 19986, 83; 19996, 58--59.

2007; cf. Lankton 2003, 70-72, 79-81.

62

Wild and Wild 2000, 269-273; 2001; Wild 2002, 9-10; 2004, 61--63; Wild and Wild 2005, 13-14.

63

53 Francis

Introduced into Egypt by the third century B.c.: Cappers 1998a, 295 (table 15-3), 311; 2000, 307.

54

55 Cappers

1996,331; 1998a,289,304-305,313,316; 83; 1999a,303; 19996,57.

Cappers 1998a, 290, 295 (table 15-3), 309, 324; 2000, 307.

19986,

56 Cappers 1996, 322 (table 19-2), 327, 330-331; 1998a, 289, 296 (table 15-3), 300 (table 15-4), 311-313; 19986, 81-82;

64 MS G.

Wilkinson XLV D. 6 and 11 from Gardner Wilkinson papers from Calke Abbey, Bodleian Library, Oxford, reproduced in Aldswortb, Sidebotham, and Wendrich 1995, 16 (fig. 4), 18 (fig. 5, though upside down).

318

STEVENE. SIDEBOTI-IAM

Roman harbors had such structures visible to those approaching from the sea on the port (left) side. Frequently, these were paired with a monument, often a lighthouse, on the starboard (right) side as one approached the harbor entrance. 65 In this connection, excavations in 1996 in trench BE96-9 of a high circular mound about two streets back of the southern/southeastem easternmost extant harbor structures aimed to reveal the remnants of a late Roman lighthouse, but the late Roman architectural remains-although clearly indicative of supporting a massive, heavy structure-and associated small finds found there were inconclusive on this point. 66 We tentatively now posit that the feature on the mound in trench BE96-9 may have been paired with the circular feature on the peninsula on the south side of the harbor. If the harbor entrance at Berenike fit into the Roman harbor model that seems to have been typical for the Mediterranean, 67 then such a paired arrangement at or near the harbor entrance would have been likely.

4. Conclusion Eight seasons of excavation at Berenike have barely scratched the surface of the site, and while something of the location and appearance of the possible Ptolemaic and early Roman harbor facilities is known, nothing convincingly identified has come to light of the late Roman port. Documentation of the physical appearance and building materials of ships as well as sails, brailing rings, and possible lifting nets is known, and early Roman texts from the site list other sailing-related equipment. Future work at the site will concentrate more on harbor facilities, and continued excavation in the trash dumps (both early and late) should also provide more written documentation and physical evidence for ships' tackle and other maritime-related objects.

65

Tuck 1997, 102-104, 117, 182.

66

Sidebotham 1998, 79-96.

67 Tuck

1997, passim.

ARCHAEOLOGICAL EVIDENCE FOR SHIPS AND HARBOR FACILITIES AT BERENIKE, EGYPT

319

Works Cited Aldsworth, F. G., S. E. Sidebotham, and W. Z. Wendrich, "The Town Site: Survey, Plan and Description," in

Berenike 1994. PreliminaryReport of the 1994 Excavationsat Berenike (EgyptianRed Sea Coast)and the Survey of the EasternDesert, ed. S. E. Sidebotham and W. Z. Wendrich (Leiden 1995) 13-20. Bagnall, R. S., A. Bulow-Jacobsen, and H. Cuvigny, "Security and Water on the Eastern Desert Roads: The Prefect Iulius Ursus and the Construction of Praesidiaunder Vespasian," Journalof Roman Archaeology 14 (2001) 325-333. Bagnall, R. S., C. C. Helms, and A. M. F. W. Verhoogt, "The Ostraka," in Berenike 1997. Report of the 1997

Excavationsat Berenike and the Survey of the EgyptianEasternDesert,IncludingExcavationsat Shenshef, ed. S. E. Sidebotham and W. Z. Wendrich (Leiden 1999) 201-205. ---, Documentsfrom Berenike, vol. 1: Greek Ostrakafrom the 1996-1998 Seasons (Brussels 2000a). Papyrologica Bruxellensia 31. "Preliminary Report on the Textual Finds," in Berenike 1998. Report of the 1998 Excavations at Berenike and the Survey of the Egyptian Eastern Desert, Including Excavationsin Wadi Kala/at,ed. S. E. Sidebotham and W. Z. Wendrich (Leiden 20006) 179-181. Bagnall, R. S., C. Helms, A. M. F. W. Verhoogt, A. Billow-Jacobsen, H. Cuvigny, M. Dijkstra, and U. KaplonyHeckel, Documentsfrom Berenike,vol. 2: Textsfrom the 1999-2001 Seasons(Brussels 2005). Papyrologica Bruxellensia 33. Belfiore, S., I! Periplode! Mare Eritreadi anonimo de! I sec. d.C. e altri testi sul commercio/ra Roma e l'Oriente attraversol'OceanoIndiana e la Via delta Seta (Rome 2004). Bernand, A., De Koptos aKosseir (Leiden 1972). ---, A., Les portes du desert (Paris 1984). Blue, L., "Excavation of Trench 4," in Myos Hormos-Quseir al-Qadim:A Roman and Islamic Port Site on the Red Sea Coasta/Egypt Interim Report, 1999, ed. D. Peacock et al. (Southampton 1999, unpublished) 27-28. ---, "Myos Hormos/Quseir al-Qadim: A Roman and Islamic Port on the Red Sea Coast of Egypt-A Maritime Perspective," Proceedingsof the Seminarfor Arabian Studies 32 (2002) 139-150. ---, "The Sedimentary History of the Harbour Area," in Myos Hormos-Quseir al-Qadim. Roman and Islamic Ports on the Red Sea, vol. 1: Survey and Excavations 1999-2003, ed. D. Peacock and L. Blue (Oxford 2006) 43-61. Blue, L., and D. Peacock, "Trench 7A," in Myos Hormos-Quseir al-Qadim.Roman and Islamic Ports on the Red Sea, vol. 1: Survey and Excavations1999-2003, ed. D. Peacock and L. Blue (Oxford 2006) 68-74. Bourdon, C., Anciens canaux,ancienssites et portes de Suez (Cairo 1925). Bradford, N., "Harbour Area Excavations," in Myos Hormos-Quseir al-Qadim:A Roman and Islamic Port Site, Interim Report, 2000, ed. D. Peacock, L. Blue, N. Bradford, and S. Moser (Southampton 2000, unpublished) 43-46. ---, "Pit 10,000 Harbour Area," in Myos Hormos-Quseir al-Qadim:A Roman and Islamic Port Site, Interim Report, 2001, ed. D. Peacock, L. Blue, N. Bradford, and S. Moser (Southampton 2001, unpublished) 31-34. Burkhalter-Arce, F., "Le 'Tarif de Coptos'. La Douane de Coptos, les fermiers de l'apostolion et le prefet de desert de Berenice," Topoisuppl. 3 (2002) 199-233. Cappers, R. T. J ., "Archaeobotanical Remains," in Berenike 1995. PreliminaryReport of the 1995 Excavations at Berenike (EgyptianRed Sea Coast)and the Survey of the Eastern Desert, ed. S. E. Sidebotham and W. Z. Wendrich (Leiden 1996) 319-336. ---, "Archaeobotanical Remains," in Berenike 1996. Report of the 1996 Excavationsat Berenike (Egyptian Red Sea Coast)and the Survey of the Eastern Desert, ed. S. E. Sidebotham and W. Z. Wendrich (Leiden 1998a) 280-330. ---,

320

--,

STEVENE. SIDEBOTHAM

"A Botanical Contribution to the Analysis of Subsistence and Trade at Berenike (Red Sea Coast, Egypt)," in Life on the Fringe:Living in the Southern Egyptian Deserts during the Roman and Early

Byzantine Periods.Proceedingsof a ColloquiumHeld on the Occasionof the 25th Anniversary of the NetherlandsInstitute/or Archaeologyand ArabicStudies in Cairo9-12 December1996,ed. 0. E. Kaper (Leiden 1998b) 75-86. "Archaeobotanical Remains," in Berenike 1997.Report of the 1997Excavationsat Berenikeand the Surveyof the EgyptianEasternDesert,IncludingExcavationsat Shenshef,ed. S. E. Sidebotham and W Z. Wendrich (Leiden 1999a) 299-305. --, "Archaeobotanical Evidence of Roman Trade with India," in Archaeologyof Seafaring:The Indian Oceanin the Ancient Period(Delhi 1999b) 51-69. Indian Council of Historical Research Monograph 1. --, "Archaeobotanical Remains," in Berenike 1998.Report of the 1998 Excavationsat Berenikeand the Survey of the EgyptianEasternDesert,IncludingExcavationsin WadiKala/at,ed. S. E. Sidebotham and W Z. Wendrich (Leiden 2000) 305-310. Casson, L., Shipsand Seamanshipin the Ancient World(Princeton 1971). ---, The Periplus Maris Erythraei Text with Introduction, Translation,and Commentary (Princeton 1989). --, "Ptolemy II and the Hunting of African Elephants," Transactionsof the American PhilologicalAssociation123 (1993) 247-260. --, Ships and Seafaringin Ancient Times (Austin 1994). Cataudella, M. R, "Quante vie d'aqua frail Mediterraneo e la Persia," in StuttgarterKolloquiumzur historischenGeographiedesAlterrums 7, 1999.Zu Wasserund zu Land. Verkehrswegein der antiken Welt, ed. E. Olshausen and H. Sonnabend (Stuttgart 2002) 48-59. Chittick, N., "Sewn Boats in the Western Indian Ocean, and a Survival in Somalia," InternationalJournalof NauticalArchaeology9.4 (1980) 297-309. 1t01:0tµoc;. Mediterraneo e Mare Rosso da Traiano a Maometto," in Controllodegli De Romanis, F., "Teoii0tvoc; stretti e insediamentimilitarinel Mediterraneo,ed. R. Villari (Rome 2002) 21-70. Francis, P., Jr., "Human Ornaments," in Berenike 1998.Report of the 1998 Excavationsat Berenikeand the Surveyof the EgyptianEasternDesert,IncludingExcavationsin WadiKala/at,ed. S. E. Sidebotham and W Z. Wendrich (Leiden 2000) 211-225. --, "Personal Adornments at Berenike through the 2000 Excavation," in Berenike 1999-2000.Report --,

of the 1999-2000Excavationsat Berenike (EgyptianRed Sea Coast)and the Surveyof the EasternDesert, IncludingExcavationsin WadiKala/atand at Siket, ed. S. E. Sidebotham and W. Z. Wendrich (Los Angdes 2007) 251-257. Fuks, A., "Notes on the Archive of Nicanor," Journalo/]uristicPapyrology5 (1951) 207-216. Fussman, G., "The Periplusand the Political History of India," in Crossings:Early MediterraneanContacts with India, ed. F. De Romanis and A. Tchernia (New Delhi 1997) 66-71. Ghisotti, A., "The Amphorae at Fury Shoal," in Diving Guide to the Red Sea Wrecks,ed. V. Manferto and L. Accomazzo (Shrewsbury, UK 1996) 92-95. Groom, N., "The Periplus,Pliny and Arabia," ArabianArchaeologyand Epigraphy6 (1995) 180-195. Harrell, J. A., "Geology," in Berenike 1995. PreliminaryReport of the 1995 Excavationsat Berenike(Egyptian Red Sea Coast)and the Surveyof the EasternDesert,ed. S. E. Sidebotham and W Z. Wendrich (Leiden 1996) 99-126. Herbich, T. M., "Magnetic Survey," in Berenike 1999-2000.Reportof the 1999-2000Excavationsat Berenike

(EgyptianRed Sea Coast)and the Surveyof the EasternDesert,IncludingExcavationsin WadiKalalatand at Siket, ed. S. E. Sidebotham and W Z. Wendrich (Los Angdes 2007) 22-29. Hourani, G. F.,Arab Seafaringin the Indian Oceanin Ancient and EarlyMedievalTimes,revised and expanded by J. Carswell (Princeton 1995). Johnson, WR, "Roman Pottery," in Quseiral-Qadim1978PreliminaryReport,ed. D.S. Whitcomb andJ. H. Johnson (Cairo 1979) 67-103.

ARCHAEOLOGICAL EVIDENCEFOR SHIPSAND HARBORFACILITIESATBERENIKE,EGYPT 321

Kennedy, D. L., "The Composition of a Military Work Party in Roman Egypt (ILS 2483: Coptos)," Journalof EgyptianArchaeology71 (1985) 156--160. Lankton,J. W, A Bead Timeline,vol. 1: Prehistoryto 1200 C.E. (Washington, DC 2003). Macklin, A., "Trench lA, with a Note on the Ostrich Egg by Dionisius Agius," in Myos Hormos-Quseir alQadim.RomanandIslamicPortson the Red Sea,vol. 1: Surveyand Excavations1999-2003, ed. D. Peacock and L. Blue (Oxford 2006) 157-160. Maclaren, C., "Account of the Ancient Canal from the Nile to the Red Sea," EdinburghPhilosophical Journal 13.26 (1825) 274-291. Masser, P., and R. Thomas, "Trench 8," in Myos Hormos-Quseir al-Qadim:A Roman and IslamicPort Site, Interim Report,2001, ed. D. Peacock et al. (Southampton 2001, unpublished) 11-18. Meredith, D., "The Roman Remains in the Eastern Desert of Egypt," Journal of EgyptianArchaeology38 (1952) 94-111. Meyer, C., "Large and Small Storerooms of the Roman Villa," in Quseiral-Qadim1980 PreliminaryReport, ed. D.S. Whitcomb and}. H. Johnson (Malibu, CA 1982) 201-213. American Research Center in Egypt Reports 7. Paice, P., "The Punt Relief, the Pithom Stele, and the Periplus of the Erythraean Sea," in Contactsbetween Cultures:WestAsia and North Africa,vol. 1, ed. A. Harrak (Lewiston, NY 1992) 227-235. Parker, A. J., Ancient Shipwrecksof the Mediterraneanand the Roman Provinces(Oxford 1992). Peacock, D., "The Roman Harbour," in Myos Hormos-Quseir al-Qadim.Roman and IslamicPortson the Red Sea,vol. 1: Surveyand Excavations1999-2003, ed. D. Peacock and L. Blue (Oxford 2006) 67-94. Peacock, D., et al., Myos Hormos-Quseir al-Qadim:A Roman and IslamicPort Site on the Red Sea Coastof Egypt,Interim Report, 1999 (Southampton 1999, unpublished). --, Myos Hormos-Quseir al-Qadim:A Roman and IslamicPortSite, Interim Report,2000 (Southampton 2000, unpublished). --, Myos Hormos-Quseir al-Qadim:A Roman and IslamicPortSite, Interim Report,2001 (Southampton 2001, unpublished). --, Myos Hormos-Quseir al-Qadim:A Roman and IslamicPortSite, Interim Report,2002 (Southampton, 2002, unpublished). --, Myos Hormos-Quseir al-Qadim:A Roman and IslamicPortSite, Interim Report,2003 (Southampton 2003, unpublished). Peacock, D., and L. Blue, "Part II: The Excavations Overview," in Myos Hormos-Quseir al-Qadim.Roman and IslamicPortson the Red Sea,vol. 1: Surveyand Excavations1999-2003, ed. D. Peacock and L. Blue (Oxford 2006a) 65-66. --, "Discussion and Conclusion," in MyosHormos-Quseir al-Qadim.Romanand IslamicPortson the Red Sea,vol. 1: Surveyand Excavations1999-2003, ed. D. Peacock and L. Blue (Oxford 2006b) 174-177. D. Peacock, D. Williams, and S. James, "Chapter 3: Basalt as Ships' Ballast and the Roman Incense Trade," in Food/or the Gods:New Light on the Ancient IncenseTrade,ed. D. Peacock and D. Williams (Oxford 2007) 28--70. Raschke, M. G., "New Studies in Roman Commerce with the East," Au/stieg und Niedergangder romishen Welt 2.9.2 (Berlin and New York 1978) 604-1378. Richardson, S., "Basketry, Matting and Cordage," in Myos Hormos-Quseir al-Qadim:A Roman and Islamic PortSite, Interim Report,2002, ed. D. Peacock et al. (Southampton 2002, unpublished) 77-S0. Robin, C., "The Date of the Periplusof the ErythraeanSeain the Light of South Arabian Evidence," in Crossings: EarlyMediterraneanContactswith India, ed. F. De Romanis and A. Tchemia (New Delhi 1997) 41-65. --, "Saba and the Sabaeans," CaravanKingdoms:Yemenand the Ancient IncenseTrade,ed. A. C. Gunter (Washington, DC 2005) 8--19. Rostovtzeff, M. I., "Review of Greek Ostracain the BodleianLibraryat Oxfordand VariousOther Collections, Edited by J. G. Tait, vol. I, London, 1930," Gnomon 7 (1931) 21-26. Rouge,]., Shipsand Fleetsof the Ancient Mediterranean,trans. S. Frazer (Middletown, CT 1981).

322

STEVEN E. SIDEBOTHAM

Sayed, A. M. A. H., "On the Non-existence of the Nile-Red Sea Canal (So Called Canal of Sesostris) during the Pharaonic Times," in The Red Sea in Antiquity: A Collectionof PapersPublishedin the Arabic and EuropeanPeriodicals,ed. A. M.A. H. Sayed (Alexandria 1993) 127-147. Scullard, H. H., The Elephantin the Greek and Roman World (Ithaca, NY 1974). Sidebotham, S. E., Roman EconomicPolicyin the ErythraThalassa30 B.C.-A.D. 217 (Leiden 1986). Mnemosyne suppl. 91. --, "Ports of the Red Sea and the Arabia-India Trade," in Rome and India: The Ancient Sea Trade,ed. V. Begley and RD. DePuma (Madison 1991) 12-38. --, "The Ship Graffito," in Berenike1995. PreliminaryReportof the 1995 Excavationsat Berenike(Egyptian Red Sea Coast)and the Survey of the EasternDesert,ed. S. E. Sidebotham and W Z. Wendrich (Leiden 1996a) 315-317. --, "The Excavations," in Berenike1995.PreliminaryReportof the 1995Excavationsat Berenike(Egyptian Red Sea Coast)and the Surveyof the EasternDesert,ed. S. E. Sidebotham and W. Z. Wendrich (Leiden 1996b) 7-97. --, "The Excavations," in Berenike1996.Reportof the 1996Excavations at Berenike(EgyptianRedSeaCoast) and the Surveyof the EasternDesert,ed. S. E. Sidebotham and W Z. Wendrich (Leiden 1998) 11-120. --, "The Excavations," in Berenike 1997.Report of the 1997 Excavationsat Berenikeand the Surveyof the EgyptianEasternDesert,IncludingExcavationsat Shenshef,ed. S. E. Sidebotham and W Z. Wendrich (Leiden 1999) 3-94. --, "The Excavations," in Berenike1998.Reportof the 1998Excavationsat Berenikeand the Surveyof the EgyptianEasternDesert,IncludingExcavationsin WadiKala/at,ed. S. E. Sidebotham and W Z. Wendrich (Leiden 2000) 3-147. --, "The Roman Empire's Southeastem-most Frontier; Recent Discoveries at Berenike and Environs (Eastern Desert of Egypt) 1998--2000," in Roman FrontierStudies2000. Limes XVIII Proceedingsof the XVIIIth InternationalCongressof Roman FrontierStudiesHeld in Amman, Jordan(September2000), ed. P. Freeman (Oxford 2002) 361-378. BAR International Series 1084.1. --, "The Coins," in Berenike1999-2000.Reportof the 1999-2000Excavationsat Berenike(EgyptianRed Sea Coast)and the Surveyof the EasternDesert,IncludingExcavationsin WadiKala/atand at Siket, ed. S. E. Sidebotham and W Z. Wendrich (Los Angeles 2007a) 200-210. --, "Excavations," in Berenike 1999-2000.Report of the 1999-2000Excavationsat Berenike (Egyptian

Red Sea Coast)and the Survey of the EasternDesert,IncludingExcavationsin WadiKala/atand at Siket, ed. S. E. Sidebotham and W Z. Wendrich (Los Angeles 2007b) 30-165. Sidebotham, S. E., and W Z. Wendrich, "Berenike: Roman Egypt's Maritime Gateway to Arabia and India," EgyptianArchaeology8 (1996b) 15-18. --, "Fieldwork at Berenike (Red Sea Coast), Egypt 1994-1996," Archaeological News 21-22 ( 1996-1997) 34-40, 117-131 (pls. 26-48). --, "Berenike. Archaeological Fieldwork at a Ptolemaic-Roman Port on the Red Sea Coast of Egypt: 1994-1998," Sahara10 (1998b) 85-96. --, "Berenike. Archaeological Fieldwork at a Ptolemaic-Roman Port on the Red Sea Coast of Egypt 1999-2001," Sahara13 (2001-2002) 23-50. --, "Berenike: A Ptolemaic-Roman Port on the Ancient Maritime Spice and Incense Route," Minerva 13.3 (May/June 2002) 28-31. --, eds., Berenike 1994.PreliminaryReport of the 1994Excavationsat Berenike(EgyptianRed Sea Coast) and the Surveyof the EasternDesert (Leiden 1995). --, eds., Berenike 1995.PreliminaryReportof the 1995Excavationsat Berenike(EgyptianRed Sea Coast) and the Surveyof the EasternDesert (Leiden 1996a). --, eds., Berenike 1996. Report of the 1996 Excavationsat Berenike (EgyptianRed Sea Coast)and the Surveyof the EasternDesert (Leiden 1998a). --, eds., Berenike 1997.Reportof the 1997Excavationsat Berenikeand the Surveyof the EgyptianEastern Desert,IncludingExcavationsat Shenshef (Leiden 1999).

ARCHAEOLOGICAL EVIDENCE FOR SHIPS AND HARBOR FACILITIES AT BERENIKE, EGYPT

323

eds., Berenike1998.Reportof the 1998Excavationsat Berenikeand the Surveyof the EgyptianEastern Desert,IncludingExcavationsin WadiKala/at(Leiden 2000). --, eds., Berenike1999-2000.Reportof the 1999-2000Excavationsat Berenike(EgyptianRed SeaCoast)and the Surveyof the EasternDesert,IncludingExcavationsin WadiKala/atand at Siket (Los Angeles 2007). Sijpesteijn, P.J., "Der TTOTAMO:E TPAIANO:E,"Aegyptus43 (1963) 70-83. Stoll, C.-P., U. Kefrig, and C. Miete, eds., Wracktauchen.Die schonstenWracksim Roten Meer (Augsburg

--,

1999). Tait, J. G., Greek Ostracain the BodleianLibraryat Oxford and VariousOther Collections,vol. 1 (London 1930). Thomas, R., "Trench 15," in Myos Hormos-Quseir al-Qadim.Roman and IslamicPortson the Red Sea,vol. 1: Surveyand Excavations1999-2003,ed. D. Peacock and L. Blue (Oxford 2006a) 87-94. --, "Trench 6P," in Myos Hormos-Quseir al-Qadim.Roman and IslamicPorts on the Red Sea, vol. 1: Surveyand Excavations1999-2003,ed. D. Peacock and L. Blue (Oxford 2006b) 151-154. Thomas, R., and P. Masser, "Trench 8," in MyosHormos-Quseir al-Qadim.Romanand IslamicPortson the Red Sea,vol. 1: Surveyand Excavations1999-2003,ed. D. Peacock and L. Blue (Oxford 2006) 127-140. Thomas, R., and J. Whitewright, "Roman Period Maritime Artefacts," in Myos Hormos-Quseir al-Qadim: A Roman and IslamicPortSite, Interim Report,2001,ed. D. Peacock et al. (Southampton 2001, unpublished) 37-40. Thomas, R., J. Whitewright, and L. Blue, "Maritime Artefacts," in Myos Hormos-Quseir al-Qadim:A Roman and IslamicPortSite, Interim Report,2002,ed. D. Peacock et al. (Southampton 2002, unpublished) 81-83. Tomber, R., "Pottery from the Roman Deposits," in Myos Hormos-Quseir al-Qadim:A Roman and Islamic PortSite, Interim Report,2001,ed. D. Peacock et al. (Southampton 2001, unpublished) 43-44. --, "Roman Pottery," in Myos Hormos-Quseir al-Qadim:A Roman and IslamicPortSite Interim Report, 2002, ed. D. Peacock et al. (Southampton 2002, unpublished) 57-58. --, "Rome and South Arabia: New Artefactual Evidence from the Red Sea," Proceedingsof the Seminar for ArabianStudies34 (2004) 351-360. --, "Aksumite Sherds from Berenike 199&-2000," in Berenike 1999-2000.Report of the 1999-2000Ex-

cavationsat Berenike(EgyptianRed Sea Coast)and the Surveyof the EasternDesert,IncludingExcavations in WadiKala/atand at Siket, ed. S. E. Sidebotham and W. Z. Wendrich (Los Angeles 2007) 175-182. Tuck, S. L., "Creating Roman Imperial Identity and Authority: The Role of Roman Imperial Harbor Monuments" (Ph.D. diss., University of Michigan 1997). Tuplin, C., "Darius' Suez Canal and Persian Imperialism," in Asia Minor and Egypt: Old Culturesin a New Empire.Proceedingsof the Groningen1988AchaemenidHistory Workshop,ed. H. Sancisi-Weerdenburg and A. Kuhrt (Leiden 1991) 237-283. Achaemenid History 6. Turner, P. J., Roman Coinsfrom India (London 1989). Institute of Archaeology Occasional Publication 12; Royal Numismatic Society Special Publication 22. Van Hecke,}., B. Bossue, V. de Buck, and E. Carpentier,Acta Sanctorum,OctobrisX (repr. Brussels 1970). Van Rengen, W., and R. Thomas, "The Sebakh Excavations," in Myos Hormos-Quseir al-Qadim.Roman and IslamicPortson the Red Sea,vol. 1: Surveyand Excavations1999-2003,ed. D. Peacock and L. Blue (Oxford 2006) 14&-151. Vasiliev,A. A.,Justin the First.An Introductionto the Epocho/]ustinian the Great (Cambridge, MA 1950). Vermeeren, C. E., "Wood and Charcoal," in Berenike1996.Reportof the 1996Excavationsat Berenike(Egyptian Red Sea Coast)and the Surveyof the EasternDesert,ed. S. E. Sidebotham and W. Z. Wendrich (Leiden 1998) 331-348. --, "Wood and Charcoal," in Berenike1997. Reportof the 1997Excavationsat Berenikeand the Surveyof the EgyptianEasternDesert,IncludingExcavationsat Shenshef,ed. S. E. Sidebotham and W. Z. Wendrich (Leiden 1999) 307-324. --, "Wood and Charcoal," in Berenike 1998.Report of the 1998Excavationsat Berenikeand the Survey of the EgyptianEasternDesert,IncludingExcavationsin Wadi Kala/at,ed. S. E. Sidebotham and W. Z.

324

STEVENE. SIDEBOTHAM

Wendrich (Leiden 2000) 311-342. Whitcomb, D. S., "Trench Summaries, Roman Areas," in Quseir al-Qadim 1980 PreliminaryReport, ed. D. S. Whitcomb and J. H. Johnson (Malibu, CA 1982) 21-49. American Research Center in Egypt Reports 7. Whitcomb, D. S., and J. H. Johnson, "Introduction," in Quseiral-Qadim1980PreliminaryReport,ed. D. S. Whitcomb and}. H. Johnson (Malibu, CA 1982) 1-20. American Research Center in Egypt Reports 7. --, eds., Quseiral-Qadim1979PreliminaryReport (Cairo 1979). --, eds., Quseir al-Qadim 1980 PreliminaryReport (Malibu, CA 1982). American Research Center in Egypt Reports 7. Whitewright, J., "Maritime Artefacts," in Myos Hormos-Quseir al-Qadim:A Roman and IslamicPort Site, Interim Report,2003, ed. D. Peacock, L. Blue, and S. Moser (Southampton 2003, unpublished) 71-73. Whittaker, P., "Trenches lOA, lOB and lOC," in Myos Hormos-Quseir al-Qadim:A Roman and IslamicPort Site, Interim Report, 2002, ed. D. Peacock, L. Blue, and S. Moser (Southampton 2002, unpublished) 35-41. Whittaker, P., M. Walsh, and L. Blue, "Trench lOA, lOB and lOC," in Myos Hormos-Quseir al-Qadim.Roman and IslamicPortson the Red Sea, vol. 1: Surveyand Excavations1999-2003,ed. D. Peacock and L. Blue (Oxford 2006) 74-81. Wilcken, U., Griindzugeund Chrestomathieder Papyruskunde.Erster Band:HistorischeTez1Zweite Hiil/te: Chrestomathie(repr. Hildesheim 1963). Wild, F., "The Webbing from Berenike: A Classification," ArchaeologicalTextilesNewsletter34 (Spring 2002) 9-16. --, "Sails, Sacking and Packing: Textiles from the First Century Rubbish Dump at Berenike," in Purpureae VestesActas del SymposiumInternacionalsobreTextilesy Tintasdel Meditemineoen epocaromana,ed. C. Alfaro,}. P. Wild, and B. Costa (Valencia 2004) 61-67. Wild, J. P., and F. C. Wild, "Textiles," in Berenike 1998.Report of the 1998Excavationsat Berenikeand the Surveyof the EgyptianEasternDesert,IncludingExcavationsin WadiKalalat,ed. S. E. Sidebotham and W. Z. Wendrich (Leiden 2000) 251-274. --, "Sails from the Roman Port at Berenike, Egypt," InternationalJournalof NauticalArchaeology30.2 (2001) 211-220. --, "Rome and India: Early Indian Cotton Textiles from Berenike, Red Sea Coast of Egypt," in Textiles in Indian OceanSocieties,ed. R Barnes (London and New York 2005) 11-16. Williams-Thorpe, 0., "Provenancing and Archaeology of Roman Millstones from the Mediterranean Area," Journalof Archaeological Science15.3 (1988) 253-305. Young, G. K., Rome's Eastern Trade:InternationalCommerceand ImperialPolicy,31 B.C.-A.D. 305 (London and New York 2001).

THE EXPANSION OF TRIUMPHAL IMAGERY BEYOND ROME: IMPERIAL MONUMENTS AT THE HARBORS OF OSTIA AND LEPCIS MAGNA Steven L. Tuck

T

his article analyzes the architectural remains and associated artistic evidence primarily from the imperial harbors of Ostia and Lepcis Magna. This analysis leads to the conclusion that emperors beginning with Claudius utilized harbor monuments for the creation and projection of triumphal imagery-contrary to Roman tradition that kept that imagery sited on and associated with the triumphal route in Rome. Further, this evidence reveals a pattern of emperors rejecting Augustan practice in favor of the model of behavior of Mark Antony with deliberately specific associations of Claudius, Trajan, and Septimius Severus with Dionysos and Antony's Alexandrian triumph initiated in 34 B.C. This material provides tangible evidence of the intentions of emperors and their administrations in monumentalizing the harbors they built, restored, enlarged, and elaborated. The repetition of forms and the effort and expense that went into advertising the construction of such monuments-in addition to that which ensured their foundation-reinforces the conclusion that the role harbor monuments played in imperial propaganda was significant and differed considerably from terrestrial building programs.'

1. Claudius,Lighthouses,and PortusAugusti The earliest evidence for triumphal monuments outside Rome comes from the principate of Claudius and consists of a series of five lighthouses built to mark his reditus,his return from the successful conquest of Britain. These, at Dover, Boulogne, Ravenna, Brundisium, and finally Portus Augusti, are examined below as precedent-setting monuments, but it is important to note that Claudius was also developing a pattern established by a previous emperor. Specifically, the use of lighthouses as markers of imperial movement and presence is first seen explicitly under Tiberius with the lighthouse at Capri. The Capri lighthouse, notably the only Roman imperial one that does not mark a harbor, city, or similar facility,was a working seamark. 2 It also provides the earliest evidence of the association of emperors with lighthouses. It was constructed on Capri on the highest point of the island closest to the mainland and adjacent to the Villa Iovis, the villa in which Tiberius lived.' Probably only lighted 1 Roman attitudes toward and conception of the sea, particularly from an urban perspective, are explored in Purcell 1996. The importance of port monuments in defining Roman territory and their unique contribution as liminal markers are accepted by many modem scholars. Horden and Purcell 2000, 126 make the point, "Port monuments such as lighthouses and moles or the famous colossus of Rhodes allbelong in this context-as only the most developed expressions of

the idea that the identity of a powerful Mediterranean figure depends on how his territory is perceived from its maritime approaches." 2 Stat.

Silv.3.5.100.

'Suet. Tib. 65.15-18.

STEVENL. TUCK

326

during Tiberius's presence at the villa, the lighthouse was so closely associated with Tiberius that its collapse, caused by an earthquake and described by Suetonius in a list of omens of Tiberius's imminent demise, was considered a portent of Tiberius's death, which occurred a few days later.4 The identification with the emperor establishes the relationship between an emperor and the lighthouse building type. It provided a concrete marker of the princeps's location, thus reflecting his movements and presence. While the degree of private versus public actions for elite Romans is a debatable issue, it is clear that the movement and acts of the emperor were never "private" in our understanding of the word. 5 The Claudian lighthouses develop this theme through direct association with the emperor and his achievements as well as his presence. The lighthouse at Ravenna arguably belongs in this category.6 Pliny the Elder mentions the lighthouse at Ravenna together with that at Portus, opening the possibility that they were contemporary foundations. 7 Certainly there is good reason to accept that the two were conceived together since Ravenna was the point at which the Claudian triumphal procession from Britain entered Italy. Claudius, having constructed the Via Claudia Augusta into the city, made his ceremonial exit from the harbor on a vast floating palace: The mouth nearest to Ravenna forms the large basin called the harbor of the Santemo; it was here that Claudius Caesar sailed out into the Adriatic, in what was a vast palace rather than a ship, when celebrating his triumph over Britain. 8

Claudius is performing some imperial ceremonial theater here; the reditusobviously recreates-for the audience of northern Italy-his departure from Britain while explicitly extending, as noted by Pliny, his triumph outside of Rome. This is the first instance we have of this extension of triumph beyond its traditional route within the pomerium of Rome, and the fact that it is so specifically connected with both the actions and constructions of Claudius supports the notion that they were deliberately conceived of in concert. In addition to the twin monuments at Portus and Ravenna, the lighthouses at Dover and Boulogne were probably Claudian foundations, which simultaneously marked the expansion of the empire and the reditusroute of Claudius. 9 These are less certain in their chronology than the Italian examples but fit the pattern of Claudian construction. Based on the testimony of Suetonius, the lighthouse at Boulogne has been traditionally assigned to Caligula.10 This seems unlikely and, although it may have been begun by him, the short term of his principate and apparent lack of sustained need for 4 Suet. Tib. 74.4--6, Et ante paucos quam obiret dies, turris phari te"ae motu Capreis concidit. 5 See D'Arms

1970, 84.

6 Purcell

1991, 22 first drew this conclusion. His brief discussion of this point and the work of Potter 1987 did much to stimulate my thinking on issues of symbolic foundations at harbors and boundaries in general.

Hadriam. Purcell 1991, 22 translates the passage as Claudius's entry into the harbor. Pliny, however, says that Claudius sailed out into the Adriatic, indicating that the "floating palace" was the vehicle for the middle stage of the triumphal passage: the entry into Portus. The factthat it was a ceremonial exit, rather than entrance, does not diminish the prestige given to the harbor at Ravenna; both the adventus and pro/ectio were equally important and worthy of ceremony and monumentalization. In contrast to Pliny, Tac. Ann. 12.36 relates only those parts of the British triumph that occurred in Rome.

7 Plin.

HN 36.83, Quales iam compluribus locisflagrant, sicut Ostiae ac Ravennae. ("Similar beacons now burn brightly in several places, for instance at Ostia and Ravenna. ")

8 HN 3.16.120, Proximum inde ostium magnitudinem portus habet qui Vatreni dicitur, qua Claudius Caesar e Britannia tn·umphans praegrandi ilia domo ven·us quam nave intravit

9 Strabo 4.1.9; 4.3.3 includes the concept of imperial embarkation points. On the construction date of Dover, see Hasenson 1980 and Wheeler 1929. 10 D'Erce

1966.

THE EXPANSION OF TRIUMPHAL IMAGERY BEYOND ROME

327

the lighthouse on the route between Britain and Europe make a Claudian date---or at the very least a Claudian completion-much more likely than a construction by Caligula. To extend the reditus route and the triumphal procession by sea in Claudius's floating palace, at least one additional harbor between Ravenna and Ostia should be involved. A lighthouse at Brundisium is attested by Pomponius Mela; he describes it topographically by its place in the harbor parallel to that of the Pharos of Alexandria.11If the testimony of Pomponius can be accepted, the lighthouse must have been constructed early in the reign of Claudius since his work dates to no later than the 40s. It certainly cannot predate the lighthouse of Claudius at Ostia, which is so often described as the first of its kind in Italy. It is most likely that it was an otherwise unknown component of the Claudian lighthouse system at Portus, Ravenna, Dover, and Boulogne. Brundisium was a well-traveled route, which was repeatedly monumentalized-most notably under Augustus with the arch mentioned by Dio Cassius. 12 No contemporary accounts attest to Claudius's movement through the port city, although he certainly would have utilized it on his return from Britain, sailing between Ravenna and Portus Augusti. Archaeological, literary, and numismatic evidence securely dates the lighthouse at Portus Augusti to the principate of Claudius and directly associates it with the emperor himself. 13 This close association suggests the subject of the statues that adorned the lighthouses. The distant model was the Zeus Soter on the Pharos of Alexandria, where he served as the savior of those who entered the harbor. 14 That role of savior of those who go to sea is transferred as early as the principate of Augustus to the emperors themselves. Suetonius makes it clear when, near the end of Augustus's life, a group of merchants sacrifice to him when entering the harbor at Puteoli rather than to the local temple for they clearly view Augustus as their new god. 15 The Roman imperial lighthouses were probably decorated with statues of the emperors themselves. Because it closely copied the Pharos at Alexandria and was unprecedented among Romans for its size and decoration, it elicited much comment. Suetonius, in his description of the construction of the harbor, says of the lighthouse: Before the entrance a breakwater was built in deep water. In order to provide more secure foundations for this he [Claudius] first sank the ship which had brought the great obelisk from Egypt ... above this he set a very high tower on the model of the Alexandrian Pharos. 16

Dio Cassius's account is briefer: "Next, in the sea itself he [Claudius] built great moles, one on each side, enclosing a large expanse of sea. He formed an island in the sea and built on it a tower with a beacon." 17 Neither of these sources provides a detailed description of the lighthouse except that, by comparison with that at Alexandria, they suggest that it was a multistory tower. For the decoration of the lighthouse we need the evidence of art, which shows the multistory tower associated with a statue often a component of the penultimate story. 11 Mela 2.7.113-114, atque ut Alexandriae ita Brundisio adiacens Pharos.

Dio Cass. 51.19. Notably the arch at Brundisium under Augustus was linked in Dio's narrative with his victory arch in the Forum Romanum. While not explicitly creating an antecedent for Claudius's reditu,, the notion of monumental linkage is already present under Augustus.

15 Suet.

Aug. 98.

16 Suet.

Claud. 20.3.

12

13 Meiggs

14 Fraser

1973, 54ff.

1972, 194-195.

17 Dio Cass. 60.11.4.Juv. 12.75-78 also notes the construction of the moles and lighthouse but provides no further details. The descriptions of Dio and Suetonius have been confirmed by excavation; see Meiggs 1973, 154-156 for an account of the early explorations. The latter excavations are published by Testaguzza 1970 and Scrinari 1970; 1971.

328

STEVEN L. TUCK

Fig. 1. Torloniaharbor relief (after Meiggs 1973, pl. XX; photo FototecaNazionale ICCD).

Fig. 2. Nero's harbor sestertius (photo author).

The many reliefs and mosaics that portray the lighthouse support this conclusion. The Torlonia relief, a Severan-period relief found along the edge of the harbor basin (fig. 1), is one of the largest and most detailed representations, but many smaller reliefs parallel the form on the Torlonia relief. Eleven of the portrayals collected by Red de show the lighthouse at Portus as a four-story tower. 18 Many of the reliefs, despite their small size, exhibit a statue on the uppermost or penultimate stage of 18 Redde 1979, 845-872 has catalogued sixty-four representations of lighthouses on Roman coins, reliefs, and mosaics. The majority of these are multistory towers of three or four stories . The representations are not intended to be exact

copies, as evidenced by the varying number of windows and irregular appearance of the door . However, the number of stories seems to be a defining characteristic and is shown rather consistently.

THE EXPANSIONOF TRIUMPHALIMAGERYBEYONDROME

329

the lighthouse, seen to best advantage on the reverse of the sestertii of Nero (fig.2) and the Torlonia relief. Both show a standing male figure either nude or partially clothed. He stands in a contrapposto pose with a scepter or staff held in an upraised hand and the other hand outstretched. While the identity of the statue is not attested in the literary sources, it is clearly the successor of the Zeus Soter on the Pharos of Alexandria. Who would the emperors display on such a monument? In the context of a harbor Neptune seems the natural choice and is, in fact, known to adopt this pose. Although the Lateran Poseidon type is much more common and replicated across the Mediterranean, a standing Poseidon leaning on a trident was a recognized type. 19 Neptune, however, is already in evidence at Portus on the Torlonia relief, shown as a standing figure next to the lighthouse. More specifically, the figure on the lighthouse shows no sign of a beard or the "briny hair" or other attributes of Neptune, such as dolphins or a rudder. The standing pose varies slightly among representations, particularly in the position of the extended hand. A Hellenistic antecedent is known from Rome in the colossal statue known as the "Hellenistic ruler. "20 The pose is also copied by Roman emperors; for example, it is known from two statues of Claudius. These are almost identical to the pose on Nero's sestertii, with the staff held in the left hand and the right extended. 21 The statues of Claudius are crowned with an oak wreath. Such wreaths adorn, or are held over, the heads of emperors in many imperial statues and reliefs and were an exclusive perquisite of the emperors and deities under the Principate. It is impossible to say whether the figure on the sestertii is wreathed, but that on the lighthouse of the Torlonia relief has a wreath held over his head in the established manner. 22 Since the lighthouse at Portus Augusti was so explicitly held to be in imitation of the Pharos of Alexandria, and Claudius presented himself as the new Jupiter or Zeus Soter, it requires little conjecture to conclude that the statue adorning the lighthouse of Portus was of Claudius. This association develops naturally from that between Tiberius and the lighthouse at Capri, where the lighthouse was accepted as a metaphor for Tiberius. It further serves to mark Claudius's place in the harbor as the terminus of the sea voyage of his triumphal return from Britain while operating as did his other lighthouses as a monument of imperial achievement as well as the personal presence of the emperor. While the literary sources do not refer to this as a triumph, it is clearly conceived of as such and may in fact have been the motivator for a closely parallel spectacle by Titus. On his return to Italy in 70, Titus passed through the eastern provinces, exhibiting prisoners and spectacles in the Syrian cities, according to Josephus. The adventusof the imperator through these cities seems to follow the pattern of Claudius. 23 19 See Pollitt

1986, fig. 290.

wreaths held over their heads include the Gemma Augustea and the Cancelleria reliefs.

20 See Pollitt

1986, fig. 74. The figure has also been identified as a Roman Republican triumphator,specifically as T. Quinctius Flamininus; see Pollitt 1986, 308, n. 23. The statue's identification as a Hellenistic ruler is upheld by Himmelmann 1989, whodatesitto ea. 170B.C.and sees it as a depiction of Attolos II as one of the Dioskouroi. 21 Kleiner 1992, figs. 106, 107. The statues show Oaudius with the attributes of Jupiter, who was closely tied to the ideology of power of theJulio-Claudians; see Fears 1984a. 22 Good

parallels of representations of emperors with closed

23 Ando

2000, 256-257 draws the logical conclusion that Titus conceived of this return as a triumph extended outside of Rome through the cities of the eastern provinces. The major distinction is one of audience. For Oaudius, the audience is found in other Roman communities outside of Rome itself, thus extending the triumph to additional Roman space and spectators. For Titus, the audience seems more complex, for while local Romans in Syria would naturally view the spectacle, the Jewish prisoners are paraded presumably as exemplary models of Roman military success.

STEVEN L. TUCK

330

Fig. 3. Torloniarelief detail (after Meiggs 1973, pl. XXla; photo FototecaNazionale ICCD).

2. Trajan, Triumph, and Portus Traiani

The second harbor basin at Portus founded by Trajan develops from the first and continues the triumphal associations of Claudius while creating more explicit imagery of triumph. Portus Traiani was constructed from 106 to 113 in the interval between the Dacian and Parthian wars.24 The money from the Dacian war was used to pay for the construction, while the monuments of the harbor, I conclude, establish the justification for the Parthian campaign yet with presumptuously triumphal imagery. Portus Traiani is, in fact, the first harbor to contain explicitly military images of a princeps . An examination of the monumental component gives a picture ofTrajan's plan. Portus Traiani enjoys neither systematic excavation nor detailed literary description as we find at Caesarea Maritima and Portus Augusti. Nevertheless, the overall pattern of monumentalization is fairly clear. Sailing through the outer Claudian harbor and into the inner Trajanic Portus, one was presented with a series of monuments dedicated to Trajan. The first was a triumphal arch-seen here on the Torlonia relief found alongside the harbor itself (fig. 1). Perhaps the arch is a marker of Trajan's pro/ectio, which might have taken him through the harbors of Portus on his way to the east in 113.25 The archaeological evidence for an arch along the harbor basin at Portus is largely representational. One of the earlier surveyors of the harbor, Texier, found on both the right and left moles, opposite each other and ea. 950 m from shore, square foundations that would have been appropriate for arches. These were not precisely at the harbor entrance but instead where the moles began to curve to close the harbor. 26 However scant the archaeological evidence, it is supported by some representations of the harbor. One of the variants of Nero's harbor sestertii shows a shorter building along the right mole, which appears to be either an arch or a single-story building with an arched opening (fig. 2). The 24

The dedication of the work is generally taken to be 112/113, corresponding with the Trajanic coin showing the harbor basin . On the date and the coin, see Mannucci and Verduchi 1996.

25 Dio

Cass. 68.17.3; on Trajan's movements in prosecuting his wars, see Halfmann 1986, 184-188.

26 Texier

1858, 31; Meiggs 1973, 159.

THE EXPANSIONOF TRIUMPHALIMAGERYBEYOND ROME

331

Fig.4. Harborrelief, VaticanMuseums, BelvedereGallery(photoauthor).

final arcaded section on the opposite mole does not appear to be architecturally distinguished in any way, although the coin field is a notoriously restricted canvas. The Torlonia relief provides a rendering of a triumphal arch surmounted by a figure in a chariot drawn by four elephants (fig. 3). A number of problems arise from this image, including the identity of the occupant of the chariot and the arch's position within the harbor. The monument is identified as an arch because, although the passages cannot be seen, it is consistent with the flank view of an arch. From the position on the relief, the chariot group faces the lighthouse and thus the entrance to the Claudian harbor, where Texier found the remains he interpreted as foundations of arches. Evidence in addition to the Torlonia relief places such a monument along the harbor basin. A relief in the Vatican (fig. 4), attributed by some scholars to Ostia, shows a harbor scene with an arch surmounted by an elephant quadriga.27 More conclusive are some relief bowls of South Gaulish ware. The central medallion in the bowls shows an elephant quadrigawith a single occupant in the chariot. The relief represents a monument, rather than an actual triumphal procession, because the chariot rests on a platform consistent with the attic of an arch . In the space below the ground line is a reclining harbor god identified by his pose and anchor attribute. 28 The specific harbor is identified by the inscription as Portus Augusti. The artistic evidence clearly supports the notion that there was an elephant quadrigain the harbor . Still unclear, however, are the identity of the occupant of the chariot and the purpose behind the monument. Regrettably, the South Gaulish bowls with elephant quadrigado not completely preserve the body of the chariot occupant. For that it is necessary to rely on the Torlonia relief (fig. 1), which shows a single figure standing in the chariot wearing a cloak and what has been interpreted as a helmet, carrying a scepter in one hand and something unidentifiable in the other. The "helmet" is only a result of the small scale and shallow relief of this figure; it seems instead to be an incompletely carved wreath with unfinished hair above it. A parallel can be found on the, arguably contemporary , 27

Robert 1911, pl. l attributes it to Ostia. I see it as a composite of many harbors, perspectives, scales, and topoi, with no suggestion of any coherent harbor scene.

28 Two

varieties of this bowl relief are illustrated by Alfoldi 1965-1966, pl. 10; for harbor gods and their iconography, see Boyce 1958, 77, pl. 15.

STEVENL. TUCK

332

arch of Titus spoils relief, where the figure immediately to the left of the menorah is in low relief, with the same unfinished head treatment. 29 The figure, if an emperor, could be no later than Trajan as he is unbearded. Meiggs suggests that it is Domitian, the first emperor known to have set up an arch in Rome surmounted by an elephant chariot. 30 In the scepter in his left hand Meiggs also sees an attribute of Domitian's: a scepter terminating in a human head, as seen on the coins of his second consulship in A.D. 73.31 Elephant chariots were used as a symbol by emperors from the time of Augustus. 32 The origin of the symbolism was in the Indian Triumph of Dionysos, which frequently showed the deity in a chariot drawn by elephants. Augustus, and later Trajan, emphasized this iconography in their triumphal programs as the traditional iconography of Hellenistic monarchs establishing dominance over the East. 33 Domitian had a similar need to establish his authority over the East, especially following the reigns of his successful and popular father and brother. 34 There is, however, no archaeological, literary, or epigraphical evidence that shows any building or patronage activity by Domitian at Portus. Except for the evidence of the scepter, difficult to judge at this scale, Domitian does not appear as likely as Trajan to be portrayed in the quadriga. Current scholarly opinion on the Portus Augusti is that it was still functional when Trajan's harbor was built. 35 Certainly it is reasonable to believe that Trajan would have erected personal monuments in the old, now outer harbor, as was done in the new inner harbor. The iconography of the chariot and its traditional ties to Dionysos also more closely match the decorative program of Trajan as the conqueror of the East, as determined from the decoration of the new harbor. In addition, the South Gaulish relief medallions contain a fragmentary and largely illegible inscription that reads, in part, FELICIS. The word was part of the official title of Trajan's new harbor: Portus Traiani Felicis, as attested in inscriptions and coins, but a title not associated with the Claudian harbor. 36 The evidence of the scepter admits another interpretation than that of Meiggs. It appears to be topped with the eagle commonly displayed on coins of the later emperors rather than a human hand. 37 Taken together, the evidence supports the conclusion that Portus Traiani Felicis was marked by at least a single triumphal arch at its entrance, probably serving as a marker of the passage between it and the Portus Augusti, which served as the outer basin for the new construction. Across from the entrance were the two most impressive monuments surrounding and facing the harbor basin: a colossal cuirassed statue of Trajan and immediately behind it a temple to Dionysos. 38 In the center of the shorefront of the Portus Traiani opposite the entrance stood the colossal statue of Trajan in military dress. 39 The placement is notable because the colossus was between the harbor 1992, fig. 155.On the carving style and identification of these figures, see Pollini 2003, 164.

with chariots and triumphal insignia across the city under that emperor.

30 Meiggs

35 Meiggs

29 Kleiner

31 BMC,

1973, 158; Matz 1952, 31.

1973, 166; Testaguzza 1970, 123.

36 Meiggs

1952, 31. Augustus's use of the chariot appears only on coins; there is no evidence that he ever erected such a monument; see Kleiner 1985, pl. 6, 3.

1973, 162; Testaguzza 1970, 122, where he cites a coin of Antoninus Pius with the same legend. On Felix and Felicitas in the ideology of emperors in the second century, see Halfmann 1986, 146--147; Fears 1984b, 910-924, esp. 913 on Trajan.

33 Matz

37 RIC7.88-91.

2:pl. 12, 2.

32 Matz

1952, 21.

Suet. Dom. 2 and 6 discusses Domitian's attempts to orchestrate a military success in the East. He also describes in Dom. 13 the proliferation of arcades and arches adorned

34

38 Meiggs

1973, 165-166.

39 Meiggs

1973, 165-166.

THE EXPANSION OF TRIUMPHAL IMAGERY BEYOND ROME

333

Fig. 5. Plan o/Portus (after Tuck 1997,fig. 22).

entrance and the temple of Bacchus, set just off the edge of the harbor behind the colossus (fig. 5). This temple was effectively screened from view from the harbor entrance. Perhaps the axis from the triumphal arch at the entrance through the colossus to the temple created a deliberate subset of self-referential monuments within the harbor. The colossus is one of only a few explicitly military images of an emperor yet found from a harbor. 40 Although the triumphal arch certainly assumed a military role for the emperor, even it showed the emperor as successful or former general, togated as on the triumphal monuments from Rome, rather than as an active one as Trajan is depicted here. 41 Remains discovered of the statue were so fragmentary that the decoration of the cuirass could not be determined. Another over-lifesized statue of Trajan-this one from his harbor construction at Terracina, where it was found in a similar position along the edge of the harbor-provides a possible parallel and suggests that it may have had a nautical theme. 42 The placement of the colossus is especially remarkable as it fills the position usually taken by a temple to the emperor along the harbor edge. The temple, identified by inscriptions as dedicated to Liber Pater, is in the center of the northeast side of the hexagonal harbor of Trajan.43 The numismatic issues of Trajan (fig. 6), probably minted to celebrate the opening of the harbor, do not explicitly distinguish the temple among the other buildings of the harbor basin, although it is probably the building with the rounded roofline just to the left of the colossus, itself visible on the best-preserved examples of this issue.44 The figure of Liber Pater appears on the Torlonia relief (fig. 1) with the harbor monuments but, like the Neptune in the foreground, not on a base. It cannot, therefore, be considered to portray a statue but is probably a personification of the presence of the god and 40 All of these military images are of the

emperor Trajan; they include the colossus at Portus, a cuirassed over-life-sized statue at Terracina, and the statue on the arch at Ancona.

The inscribed base of another over-life-sized statue of Trajan (CJL 9.37) has been found at Brundisium, where it probably marked the termination of the Via Traiana at the harbor; see Ashby and Gardner 1916, 116, 170-171.

described it as a grand statue of Trajan in armor all carved of Greek marble. It stood, he goes on to say, on a well-built, square pedestal over 6 feet high. None of the details of the decoration or possible inscription were recorded.

41

42

Fea 1802, 36 recorded the discovery of the statue and

The attribution is made by Lanciani 1868, 181 based on epigraphical evidence. The identification is accepted by all subsequent authors on the harborworks. 43

44 Testaguzza

1970, 170.

334

STEVEN L. TUCK

Fig. 6. Trajan's harbor sestertius (after Meiggs 1973, pl. XVIIIb) .

the trade in wine in the harbor. As the figure stands in the background behind the ship off-loading amphoras, this would be a consistent message. Certainly the patron of wine is appropriate to oversee the trade and storage of the vessels, but the temple seems to be part of a larger program of monumentalization in the harbor. Although Liber Pater is occasionally depicted with nautical familiars, he is primarily neither a naval or a trade god nor the object of sailors' prayers. 45 The reason for the presence and presentation of the temple in the harbor can be found in the ideological program of Trajan. In the tradition of past emperors, he associated himself with Dionysos. 46 Trajan was presented iconographically, particularly on numismatic issues, as the Neos Dionysos. He became the patron of the associations of Dionysiac actors, promoted the influx of easterners to Rome and, with his assimilation with Alexander, became the domitor Orientis. 47 The purpose of this was the establishment of a mythological basis and justification for the eastern campaigns against the Parthians. With the assimilation of Dionysos with Trajan, it became a sacred duty to liberate the East. In many ways Dionysos became essentially one aspect of the identity of Trajan. From this perspective it becomes clear why this particular temple was constructed within the program of monumentalization of Trajan's harbor. There was simply no place, programmatically, for another prominent figure in the harbor basin to compete with Trajan. Instead, the temple serves to reinforce the authority of Trajan. Much evidence survives of Trajan's descriptions of himself as the Neos Dionysos-coins, inscriptions , etc. Unlike Augustus or Claudius, Trajan does not present himself explicitly as Jupiter 45 The

nautical association of Liber Pater occurs in Dionysiac iconography on a series of sarcophagi with Nereids and marine centaurs; see Foucher 1984, 697. Dionysos was, along with Hercules , one of the two most common traveling gods in the classical world . His aspect as the traveler could account, by itself, for the prominence of his worship in the harbor; see Bowersock 1990, 41-53 on this aspect of his worship. A possible temple to Hercules has also been suggested along the harbor basin, but the evidence is correctly concluded by Meiggs 1973, 386 to be "not decisive ." In fact, the syncretism of Dionysos the conqueror with a ruler can be traced back to Alexander the Great, whom the

46

Roman emperors may have, once again, been emulating in their ideological programs. Nock 1972 analyzes this relationship at length in his essays on "Alexander and Dionysus" and "Neos Dionysos." All of the emperors, from Augustus on, were listed as the patrons of the acting guilds. More evidence survives that Trajan was proclaimed as such than for any other emperor. On the increasing number of easterners in Rome, see Foucher 1984, 700 as well as Hammond 1957, 77-79 , whose table on p. 77 shows that the number of attested eastern senators under Trajan was double that under his predecessors Vespasian and Domitian . 47

THE EXPANSIONOF TRIUMPHALIMAGERYBEYONDROME

335

or with his attributes but as "the subordinate of the divine father, ruling the earth in his behalf, the warrior vice-regent of Jupiter who holds the barbarians at bay," especially as his son, either Hercules or, here in the harbor, Dionysos. 48 The harbor admits no other deities than Trajan and Dionysos, and the latter appears as merely a component ofTrajan's imperial personality. He serves to reinforce the virtusofTrajan as a successful military commander. The decorative/building program is designed to establish the justification for Trajan's Parthian war: Trajan as the Neas Dionysos had the right to eastern conquests; triumph is preordained. While we do not know for certain the intended audience for these messages, we know that Portus Traiani was designed as the terminus for the grain fleet from Alexandria, providing one specific audience. These military monuments might have been designed to appeal most directly, however, to those soldiers and sailors who we know sailed out of Portus to launch Trajan's Parthian campaign, sending them to war with a triumphal message. This explicitly triumphal message serves as a bridge between the harbor monuments of Claudius and those of Septimius Severns at Lepcis Magna, which mark the development of imperial ideology from Parthicus to Parthicus Maximus.

3. SeptimiusSeverus,Triumph,and the Reconstructiono/Lepcis Magna The Severan constructions at Lepcis Magna show an understanding of Trajan's harbor at Portus, as we see embedded in the architectural forms a message declaring Septimius Severns as the new Trajan-divinely led to rule and to conquer not as Parthicus hut as Parthicus Maximus. Septimius Severus's decision to aggrandize Lepcis Magna operates on many levels. First, one cannot overestimate the importance of announcing control of the resources of the empire through construction of an architectural wonder by a princeps whose position results from civil war. The example of Vespasian and the Flavian amphitheater exists as a marked parallel. We should resist the temptation to think of this as personal interest or an example of ancient "pork" in the lavish building of Lepcis Magna. I think that John D' Arms gave the correct assessment when he stated, "An emperor's presence ... , his local administrative actions, his public works and his benefactions, should not be construed as expressions of his personal liking .... They were also, as with Augustus, acts of imperial policy. "49 And what we see at Lepcis Magna is a step along the route from empire to commonwealth as the monuments and messages previously exclusive to Rome become attached to other cities as triumphal monuments and images are so clearly established here. A massive lighthouse-echoing those at Portus and Alexandria and certainly deliberately invoking them-on the right side to those entering the harbor marks the mouth of the new Severan harbor, with a companion tower across the harhor mouth (fig. 7). The entrance is flanked by two small distyle temples, while on the shore opposite the entrance is a large prostyle podium temple. We will examine the meaning of these buildings, beginning with the latter, most prominent one.' 0 This large podium temple exists only in the podium, foundation, and associated blocks of the cella walls and architrave. The temple was founded in the most prestigious location in the Severan harbor: along the harborside, centrally opposite the harbor entrance. The form of the superstructure is unknown because little survives above the level of the foundation other than scattered blocks of 48 Fears 49

1984a, 114.

D'Arms 1970, 84. His subject is the emperors on the Bay of Naples, but the point can be made anywhere in

imperial lands.

,oThe major source for the understanding of the harbor is Bartoccini 1958, followed by Cafarelli and Caputo 1966.

STEVENL. TUCK

336

Fig. 7. Plan of harborat Leptis Magna (after Tuck 1997,fig. 2).

....•'

...,,..•·

Lcptis Magna harbor temples: 5 - Neronian temple at terminus of first century A.D .

mole; 6 & 8 • Sevcran harbor entrance iemplcs; 7 • Scvcran cencralharborsidc temple; 9 • temple built under Domitianto Vcspasianand Tirus

the architrave. 51 There is little doubt, however, that the temple would have dominated the Severan harbor; it was constructed on a high (4.47 m) podium above the upper terrace of the two-stepped quay. The front of the podium is a broad staircase of twenty-two steps. This central podium temple is dedicated to Jupiter. The altar before the temple, dated between 202 and 204, preserves in an inscription the dedication plan for the entire harbor complex. It is dedicated to Jupiter Optimus Maximus Dolichenus and to the Augusti in triumph: I(ovt) O(ptimio) M(aximo) I Dolicheno/ pro salute et victoria domi/norum nostrorum Augg[g(ustorum trium)] et I [C(ai) Fulvi Plautiani pr(aefecti) pr(aetorio) cl(arissimi) v(iri)] I [necessariAugg(ustorum)] [e]t rediltu [I]mppp(eratorum III) in urbem [s]uam/ T(itus) Flavius Marin[us] c(enturio) leg(ionis) I [tertiae A]ug(ustae) I v(otum) l(ibens) p(osuit) I Right side: D(e)d(icavit) Illidus Apriles52

This declaration of sovereign divinity is consistent with that in the programs of previous emperors, including Claudius and the early ideological propaganda of Augustus.Jupiter is an expected and critical deity promoted early in Septimius's principate as stabilizing the Roman state through consolidating the Severan dynasty. The emperor owed his position as princeps to the support of Jupiter and in his early coinage, before 194, stresses his divine election by Jupiter, who had appointed him monarch. These coins portray Severus in military dress as a "warrior vice-regent of Jupiter." This ideological observance is fairly straightforward. In fact, some assimilation between the roles of 51 Brouquier-Redde

1992, 120.

52 Reynolds

and Ward-Perkins 1952, 292.

THE EXPANSIONOF TRIUMPHALIMAGERYBEYONDROME

337

Jupiter and Septimius Severns might be expected. Certainly Augustus was identified with Jupiter, as were Claudius and Trajan, the latter two addressed as Optimus. What suggests that this temple should be identified with Septimius Severns personally? First, it is a component of a massive building campaign whose climax was a temple to the Genius of the Severans in the new Forum. The position the temple occupies in the harbor is that usually held by monuments to the emperors, generally to the reigning emperor. This combined with the fact that it replaced a previous temple to Titus and Vespasian-which dominated the prior harbor-argues continuity of use. That this place is taken by a temple to a deity so closely connected to the army and to the emperor's family strongly suggests that it was constructed, if not on his orders or as part of a program by his administration, then perhaps in an attempt to promote the worship of a deity favored by the emperor. 53 Certainly the dedicatory inscription attributes the construction to the Victory of the Severans. The construction of the temple implies a community of worshipers and priests to support it, but that community would not have existed if not for the reign of Septimius Severns. Because of Severus's ties to the army, it is unsurprising that the altar (and perhaps the temple) were dedicated by a centurion. The dedication of the major temple in the harbor to Iuppiter Dolichenus suggests the identity of the deities for the other two temples in the Severan-period harbor: the small distyle in antis temples on the ends of the moles. The evidence for them and the possible identities for their deities are addressed below. The evidence suggests that the Severan harbor restoration was constructed with an integrated plan of monumentalization, with the three temples dedicated to related deities: Iuppiter, Hercules, and Dionysos. The temples directly face the harbor entrance, framing it from near the ends of the moles. On their low ( 1.5 m) socles and with small (ea. 6 m) cellas, they would not have dominated the harbor, as did the temple across from the harbor entrance, but would have been clearly visible to those entering or exiting the harbor. With the exception of the lighthouse and the tower opposite it, they represent the terminal monuments on either mole. Both are small Doric order temples (fig. 7, nos. 6 and 8) with the columns arranged distyle in antis, and they are very parallel in type, materials, orientation, and location within the harbor. 54 The archaeological evidence suggests that they were both part of the Severan rebuilding of the harbor and were conceived as part of the overall building program.'' Many deities have been recommended in the past for the anonymous temples: candidates include Hercules and/or Bacchus (the divine patrons of the city), Neptune, Tyche, or the Capitoline triad.' 6 The harbor entrance temples, subsidiary to the great podium temple yet components of the same plan, have been identified by their excavator, Bartoccini, with the worship of Bacchus and Hercules. These deities were the two divine patrons of Lepcis Magna, and certainly their worship would be appropriate here for that reason. Further, Dio Cassius relates (77.16) that Septimius 53 The

worship ofluppiter Dolichenus is strongly connected to that of Sol Invictus, which alsohad close ties to the Severan family: Speidel 1978, 66.

fa~ade and that of the temple of Fortuna Augusta at Musti: Romanelli 1925, 108. Both are in fact rather generic temples with Doric columns distyle in antis, with triglyph metope friezes as the only decoration relieving the plain fa~ades.

54 Bartoccini 1958, pl. 18, 47-48 restored both temples

in this form based on the extensive remains found during excavation. Subsequent authors have all accepted the restorations. "Brouquier-Redde 1992, 118-119, 122-125. It has been suggested that the temple on the east mole dates to the period of Marcus Aurelius based on alleged similarities between the

56 Brouquier-Redde 1992, 125; the suggestion of Hercules

and Bacchus has an attractive symmetry. Both were patrons of the city as well as the most famous and prominent of the traveling deities in the classical world; see, e.g., Bowersock 1990, 41-53 on the tradition of Dionysus the traveler. A temple to Hercules may have existed in a similar location to these temples at Portus according to Meiggs 1973, 386.

STEVENL. TUCK

338

built an enormous temple to Bacchus and Hercules, possibly to be placed in the city of Lepcis Magna. Again, assimilation seems to be taking place between the deities and the Augusti. At Portus Traiani, Trajan was closely associated and assimilated with the worship of Dionysos. Here the temples to Hercules and Bacchus are as certainly associated with the imperial family. They might be the temples of Septimius Severns and Caracalla: the latter was raised to the rank of Augustus prior to their departure for the Parthian campaign, and thus the monuments, which celebrate that triumph, would logically honor the two Augusti, the major protagonists of the campaign. Perhaps we should see the temple of Bacchus as that associated with Septimius while Hercules is assimilated to Caracalla. The reverses of contemporary coins of 204 portray Bacchus and Hercules, thought by some to represent Severns and Caracalla. Certainly from 204 these Severan patron deities play a larger role in dynastic ideology. The associations of the principals are not, however, explicit or exclusive. The inscription on the altar has room for a third G on the abbreviation for Augustorum and, in fact, Imperatorum preserves its three Ps. The damaged condition of the inscription face, largely owing to weathering, does not allow for absolute certainty, but it is likely that the third G was carved and then erased rather than a space left at this point. In fact, Geta as well as Caracalla is found with the title Augustus in Tripolitanian inscriptions as early as 201.57 If the inscription is restored with reference to all three Augusti, then it is more supportable that the Jupiter Dolichenus is the Jupiter from the East-Septimius-aided by his two sons, Hercules and Bacchus, the supporters of his Parthian campaign. Is there additional support for this interpretation? It can, perhaps, he found again in coinage. After 196, Septimius alone issued Jupiter types, while from 204 on Hercules and Bacchus do not appear on his coins.58 Instead, they are prominent on the issues of Caracalla and Geta, who are even shown crowned by Hercules and Bacchus respectively. These Severan patron deities are explicitly linked to the sons of Septimius from the 190s on, with Caracalla deliberately linked to Hercules in an assimilation of their personalities, not just as a patron deity.59 Septimius's special patron continues to he Jupiter. Parallels to this use of Hercules and Bacchus can be found on the Severan monuments from Rome, specifically on the northwest and southwest keystones of the Severan Arch in the Roman Forum, as well as on the Arch of the Argentarii in the Forum Boarium. 60 This interpretation indicates a shift in meaning from that seen at Portus Traiani. In his harbor, Trajan constructed a space in which the ideological message established the religious basis for a future campaign, as it was prior to his eastern campaigns that Trajan promoted himself as the Neos Dionysos. Here at Lepcis Magna, the Severan building program instead commemorates the completed campaign and the subsequent triumph. Up to this point the building program of this harbor and its association with Hercules, Bacchus, and Jupiter through direct assimilation with the members of the imperial house are supported by at least one inscription, but the military triumphal component has not been fully explored. This component arises from the buildings at the very mouth of the harbor: the lighthouse and the tower opposite it. These survive without decoration, although we would expect either or both to have served as platforms for statues of the emperor and imperial family, as did the lighthouses at Portus and Alexandria. The lighthouse at Lepcis Magna was clearly conceived of, however, as a component in the overall scheme of construction and is portrayed as a triumphal monument in honor of the Severan military 57 Reynolds

and Ward-Perkins 1952, 292.

58 McCann 1968.

59 This

point is well established in Kleiner 1992.

60 Brilliant

1967.

THE EXPANSIONOF TRIUMPHALIMAGERYBEYOND ROME

339

Fig. 8. Leptis Magna relieffrom Severan arch (photo author).

campaign in the East . The evidence for this appears on the reliefs of the Severan arch in the Forum at Lepcis Magna. The lighthouse is portrayed on the arch as a three- or four-story tower, with arched openings in each story acting as windows to allow light into the barrel-vaulted chambers. One of the four relief panels depicts the entrance of the Augusti into the city: the central, frontal figure of Septimius Severus sharing the quadrigawith his sons Caracalla and Geta as he shared the religious monuments of the harbor with them (fig. 8). The lighthouse provides the only architectural context in the scene as the Augusti pass through the harbor into the city. The lighthouse establishes the location of the triumphal procession as commencing along the Severan harbor , from whence it will proceed through the arch to the Forum. Visually and ceremonially the two major public spaces in the city constructed under Severus are connected through this monument; the lighthouse acts as a marker for the beginning of the procession and a shorthand for the entire harbor project. The lighthouse fills the position that on previous monuments would have been occupied by the porta triumphalis as the architectural marker of a triumph. Its use here compares to the Arch of Titus, which shows the porta as the architectural context for the triumphal procession, in fact defining the procession as triumphal. The other details of the chariot, patron deities, emphasis on victories, and especially the presence of captives all support the argument that this is a triumph and not just an adventus or reditus. What we see at Lepcis Magna is a programmatic display of building types whose polysemic nature establishes them as both triumphal and religious and marking the triumphal return of the Augusti into Lepcis Magna, supported by their divine patrons . Septimius Severus adopts the dual meaning of military triumph combined with religious sanction . This adoption exists to grant authority to his rule and establishes him as the successor of Trajan, implying the inception of a new period of peace and stability for the empire. The monuments in the harbor transform the pattern of imperial ideology found in Portus Traiani into one appropriate for the founder of a dynasty. The multiple levels of meaning and deities into whose identity the Severans assimilated themselves were appropriate to mark the growth of the princeps from Parthicus to Parthicus Maximus and to greatly expand the notion of triumph outside the traditional boundaries of the pomerium of Rome. As such the constructions at Lepcis Magna are a leading indicator of the future of the Principate . Within a hundred years Rome would become primarily symbolic , and the active capitals of the Augusti and Caesari were to be found at more strategic sites, completing the shift from empire to commonwealth that developed throughout the Principate.

340

STEVENL. TUCK

WorksCited Alfoldi, A., "Die alexandrinischen Gotter und die Vota Publica am Jahresbeginn," Jahrbuchfar Antike und Christen/um8/9 (1965-1966) 53-87. Ando, C., ImperialIdeologyand ProvincialLoyaltyin the Roman Empire (Berkdey 2000). Ashby, T., and R Gardner, "The Via Traiana," Papersof the BritishSchoolat Rome 8 (1916) 104-171. Bartoccini, R., It Porto Romana di Leptis Magna (Rome 1958). Bollettino dd Centro Studi per la storia dell' Architettura suppl. 13. Bowersock, G. W, Hellenismin Late Antiquity (Ann Arbor 1990). Boyce, A. A., "The Harbor of Pompeiopolis," AmericanJournalof Archaeolog,y 62 (1958) 35-54. Brilliant, R., The Arch o/Septimus Severusin the Roman Forum(Rome 1967). Brouquier-Redde, V., Templeset cultesde Tripolitaine(Paris 1992). Cafardli, E., and G. Caputo, The BuriedCity:Excavationsat Leptis Magna(New York 1966). D' Arms, J. H., Romanson the Bay of Naples(Cambridge 1970). D'Erce, F., "La Tour de Caligula a Boulogne-sur-Mer," Revue archeologique1 (1966) 89-96. Fea, C., Relazionedi un viaggioad Ostia e allavilla Plinio detta Laurentio(Rome 1802). Fears, J. R., "The Cult of Jupiter and Roman Imperial Ideology," in Aufstieg und Niedergangder romischen Welt, vol. 2, ed. H. Temporini (Berlin 1984a) 17.1:3-141. --, "The Cult of Virtues and Roman Imperial Ideology," in Aufstieg und Niedergangder romischenWelt, vol. 2, ed. H. Temporini (Berlin 1984b) 17.2:827-948. Foucher, L., "Le culte de Bacchus sous l'empire Romain," in Au/stieg und Niedergangder romischenWelt,vol. 2, ed. H. Temporini (Berlin 1984) 17.2:684-702. Fraser, P. M., PtolemaicAlexandria(Oxford 1972). Halfmann, H., ItineraPrincipum(Stuttgart 1986). Hammond, M., "Composition of the Senate, A.O. 68-235," Journalof Roman Studies47 (1957) 74-81. Hasenson, A., The Historyof DoverHarbour(London 1980). Himmdmann, N., Herrscherund Athlet. Die Bronzenvom Quirinal(Milan 1989). Horden, P., and N. Purcell, The CorruptingSea (Oxford 2000). Kleiner, D. E. E., Roman Sculpture(New Haven 1992). Kleiner, F. S., The Arch of Nero in Rome:A Study of the Roman HonoraryArch beforeand under Nero (Rome 1985). Lanciani, R., Ricerchetopographiche sulla cittadi Porto(Rome 1868). Mannucci, V.,and P. Verduchi, "Il porto imperiale di Roma: le vicende storiche in V. Mannucci," in It parco archeologiconaturalisticode! Portodi Traiano:metodo e progetto,ed. V.Mannucci (Rome 1996) 15-28. Matz, F.,Der Gott au/ dem Elefantenwagen(Mainz 1952). McCann, A. M., The Portraitsof SeptimiusSeverus,A.D. 193-211 (Rome 1968). Meiggs, R., Roman Ostia,2nd ed. (Oxford 1973). Nock, A. D., "Notes on Ruler-Cult I-IV," in Essayson Religionand the Ancient World,ed. Z. Stewart (Oxford 1972) 134-159. Pollini, J.,"Slave-boys for Sexual and Religious Service," in FlavianRome: Culture,Imageand Text, ed. A. J. Boyle and W J. Dominik (Leiden 2003) 149-166. Pollitt, J. J., Art in the HellenisticAge (Cambridge 1986). Potter, D.S., "The TabulaSiarensis,Tiberius, the Senate, and the Eastern Boundary of the Roman Empire," Zeitschriftfur Papyrologieund Epigraphik69 (1987) 269-276. Purcell, N., "The Creation of Provincial Landscape: The Roman Impact on Cisalpine Gaul," in Roman Landscapes:ArchaeologicalSurvey in the MediterraneanRegion, ed. G. W Barker and J. A. Lloyd (London 1991). --, "The Ports of Rome: Evolution of a 'FafadeMaritime'," in "Roman Ostia"Revisited,ed. A. G. Zevi and A. Claridge (London 1996).

TIIE EXPANSIONOF TRIUMPHALIMAGERYBEYONDROME

341

Redde, M., "La representation des phares al'epoque romaine," Melangesde !'Ecole/ranfaise de Rome, Antiquite91.2 (1979) 845--872. Reynolds,}. M., and}. B. Ward-Perkins, The Inscriptionsof Roman Tripolitania(Rome 1952). Robert, C., "Ostia und Portus," Hermes47 (1911) 249-253. Romanelli, P., Leptis Magna(Rome 1925). Scrinari, V., "Strutture portuali rdative al 'port di Claudio' messo in luce durante i lavori dell'Aeroporto lntercontinentale di Fiumicino (Roma)," Rassegnadei lavoripubblici7 (1970) 173-190. --, "Il 'Portus Claudii' e i piu recenti ritrovamenti nella zona di Fiumicino," in Atti de! III Congresso Internazionaledi ArcheologiaSottomarina,Barcellona1961 (Bordighera 1971) 215-224. Speidd, M. P., The Religiono/Iuppiter Dolichenusin the Roman Army (Leiden 1978). Testaguzza, 0., Portus(Rome 1970). Texier, C., Memoiresur lesports antiquessituesa!'embouchuredu Tibre(Paris 1858). Tuck, S. L., "Creating Roman Imperial Identity and Authority: The Role of Roman Imperial Harbor Monuments" (Ph.D. diss., University of Michigan, 1997). Wheder, R. E. M., "The Roman Lighthouses at Dover," AntiquariesJournal86 (1929) 29--46.

"From papers on warship slipways to prostitutes, and from piracy to hydraulic concrete, this volume will be a required source for researchers dealing with maritime life in Roman times. As with all good scholarship, the combined gravitas of the contributions here pushes research forward by discussing new fieldwork, reviewing critically previous conclusions, studying evidence in new patterns and experimental archaeology."

-SHELLEY WACHSMANN, MeadowsProfessor of BiblicalArchaeology, NauticalArchaeology Program,TexasA&MUniversity

"Tl1eMaritimeWorldof A11cie11t Romeprovides both theoretical and descriptive discussions of recent scholarly work devoted to expanding our modern understanding of the role of waterways and seas in Roman life. Drawing upon history and archaeology through cogent and accessible contributions by top scholars, the collection will stimulate discussion and debate for years to come. Readers will, like me, be inspired by the overarching perspective of the maritime network and its influence on so many aspects of life in the ancient Roman world."

-CHERYL WARD,AssistantProfessor of Anthropology, FloridaStateUniversity

"Tl1eMaritimeWorldof A11cie11t Romeis not just of interest to maritime scholars but also to anyone working on the ancient Roman world."

-HECTOR WILLIAMS, Trustee,Vancouver MaritimeMuseum,andProfessor, Department of Classical,NearEasternandReligiousStudies,Universityof BritishColumbia

E UNIVERSITY OF MICHIGANPRESS

Arbor

www.press.umich.edu