New Approaches to the Temple of Zeus at Olympia: Proceedings of the First Olympia-Seminar 8th-10th May 2014 [1 ed.] 1443878162, 9781443878166

Citation preview

New Approaches to the Temple of Zeus at Olympia

New Approaches to the Temple of Zeus at Olympia Proceedings of the First Olympia-Seminar 8th-10th May 2014 Edited by

András Patay-Horváth

New Approaches to the Temple of Zeus at Olympia: Proceedings of the First Olympia-Seminar 8th-10th May 2014 Edited by András Patay-Horváth This book first published 2015 Cambridge Scholars Publishing Lady Stephenson Library, Newcastle upon Tyne, NE6 2PA, UK British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Copyright © 2015 by András Patay-Horváth and contributors All rights for this book reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the copyright owner. ISBN (10): 1-4438-7816-2 ISBN (13): 978-1-4438-7816-6

FOR J. GY. SZILÁGYI

TABLE OF CONTENTS

Preface ......................................................................................................... x List of Illustrations and Tables .................................................................. xii Abbreviations ............................................................................................ xx Introduction ................................................................................................ 1 Adopting a New Approach to the Temple and its Sculptural Decoration András Patay-Horváth Part I: Architecture Chapter One ............................................................................................... 16 The Temple Architecture and its Modifications during the 5th Century BCE Arnd Hennemeyer Chapter Two .............................................................................................. 39 Versatzmarken auf Baugliedern des Zeustempels von Olympia Ulf Weber Chapter Three ............................................................................................ 56 Metrologisches in Olympia Wolfgang Sonntagbauer Part II: Sculpture Chapter Four .............................................................................................. 74 Dating the Corner Figures of the West Pediment and Questions Arising from the Use of Parian and Pentelic Marbles in the Sanctuary Olga Palagia Chapter Five .............................................................................................. 90 Noch einmal rechts und links am Zeus-Tempel von Olympia Tonio Hölscher

viii

Table of Contents

Chapter Six ................................................................................................ 94 An der linken Seite des großen Tempels (Paus. 5.26.2) András Patay-Horváth Chapter Seven............................................................................................ 98 The Invisible God: The Representation of Divine Intervention in the Early Classical Period Toshihiro Osada Part III: Cult and History Chapter Eight ........................................................................................... 112 The Great Hecatomb to Zeus Olympios: Some Observations on IvO No. 14 Julia Taita Chapter Nine ........................................................................................... 140 The Justice of Zeus at Olympia Jim Roy Chapter Ten ............................................................................................. 150 Charles Seltman, le temple de Zeus et le monnayage d’argent des Éléens Franck Wojan Chapter Eleven ........................................................................................ 164 Lucian on Herodotus: A Possible Second Century AD View on the West Pediment Jessica Susanne Krause Part IV: Digital Technologies Chapter Twelve ....................................................................................... 188 The East Pediment and the Temple of Zeus Reconstructed in Virtual Reality András Patay-Horváth Chapter Thirteen ...................................................................................... 201 Polykleitos’ Works “From One Model”: New Evidence Obtained from 3D Digital Shape Comparisons Kyoko Sengoku-Haga, Yujin Zhang, Min Lu, Shintaro Ono, Takeshi Oishi, Takeshi Masuda, Katsushi Ikeuchi

New Approaches to the Temple of Zeus at Olympia

ix

Chapter Fourteen ..................................................................................... 223 Virtual Environments and Technological Solutions for an Enriched Viewing of Historical and Archaeological Contexts Francesco Gabellone Chapter Fifteen ........................................................................................ 233 Cloud-based Collaborative Framework for Remote Real-time Interaction with Large-scale 3D Data Yasuhide Okamoto, Gregorij Kurillo, Takeshi Oishi, Katsushi Ikeuchi, Ruzena Bajcsy Bibliography ............................................................................................ 251 The Authors of this Volume .................................................................... 276 Index of Ancient Personal and Place Names ........................................... 280

PREFACE

This volume contains the proceedings of an international symposium held at Budapest between 8th and 10th May 2014. The idea and the title of this conference emerged from a series of projects and studies related to the temple and its sculptures which represent my unconventional approach to this monument. The results concerning the east pediment and the construction of the temple were published during the last decade and because many specialists seemed to be unconvinced, I originally expected that the conference would focus exactly on these controversial topics. Moreover, in order to facilitate the discussion, I invited first of all those colleagues, whom I knew or suspected to be in disagreement with me. Unfortunately, they have all declined or simply did not answer. Their conspicuous absence from the conference and from the proceedings is therefore not due to some predisposition or selectivity on my part. There is one single contribution in this volume (by Prof. Tonio Hölscher), which was not presented as a paper at the symposium, but it is a most welcome addition, because it clearly shows that it was not my intention to exclude any approach, even if it is fundamentally different from my own one. The conference made an attempt to bridge the gap between classical studies and the latest digital technologies applied to archaeology and cultural heritage and included therefore some talks by specialists on computer graphics and remote sensing. These papers have either offered case studies unrelated to the temple of Zeus at Olympia or presented an overview of recent projects in this domain. In spite of this, some of them are included in this volume not only because they demonstrate the interdisciplinary character of the symposium, but also because they may be of interest for readers specialized in classical studies, informing them about new possibilities and applications which might be useful for them in the future. During the conference, there were lively and fruitful discussions after each paper, which are not recorded separately in this volume, because the speakers decided to incorporate these observations and remarks into their contributions. Last but not least, it is a pleasure to acknowledge the financial help of the Hungarian Academy of Sciences, which enabled the symposium to be held. During the preparatory phase, I benefited from the advice and help of H. J. Kienast (München), O. Palagia (Athens), P. Siewert (Wien) and R.

New Approaches to the Temple of Zeus at Olympia

xi

Scopigno (Pisa). They all agreed to chair some of the sessions and took part in the discussions. Colleagues, friends and students from Budapest also contributed substantially to the organization and the success of the conference and helped to enrich the program in various ways. I would like to thank particularly Á. M. Nagy, M. Dági, Á. Bencze (Department of Greek and Roman Antiquities of the Museum of Fine Arts), Zs. Mráv (Hungarian National Museum) and last but not least Gy. Németh (Institute for Ancient History, University Eötvös Loránd). I am greatly indebted to all of them. Cambridge Scholars Publishing has already made a significant contribution to the study of the temple of Zeus at Olympia by publishing a collective volume dedicated to the chryselephantine statue of Pheidias (The Statue of Zeus at Olympia: New Approaches, ed. by J. McWilliam, S. Puttock, T. Stevenson and R. Taraporewalla, 2011) and I am most grateful for the offer to include these proceedings in the programme as well.

LIST OF FIGURES AND TABLES

Chapter One (pages 33-38) Fig. 1. Temple of Zeus, metrological scheme of ground plan, drawing by Wilhelm Dörpfeld 1880, with measurements in ‘Olympic feet’ (later erroneously rejected by Dörpfeld himself). Copyright DAI. Fig. 2. Temple of Zeus, alteration during construction process (marking in light grey: hypothetical original layout; in dark grey: execution). ‘Steinplan’, detail of the north eastern corner. Fig. 3. Modification of the double-storey lines of columns and their stylobate. Marked in dark grey: remains in situ; in light grey: reconstruction of phase II; line drawing: reconstruction of phase I. Fig. 4. Temple of Zeus, cross section, after Dörpfeld 1892 (architecture) and Adler (statue), with essential modifications according to the current state of research. Fig. 5. The floor in the area behind the statue, section drawing. a) remains in situ in the 19th c. (Forbat 1935); Copyright DAI, Grabung Olympia. b) reconstruction, lime floor covering the gap between the base and the paving block; c) when the base settled, the joint couldn't open but was compressed. Fig. 6. White marble rim, block at the south-western corner of the Zeus' base (drawing by Dörpfeld 1880; anathyrosis and legend added) Copyright DAI. Fig. 7. Reconstruction of the base and the shallow basin, both paved with dark grey Eleusinian limestone, and the white marble rim (on the drawing marked in light grey). Fig. 8. Cella, view about from the threshold. Based on a drawing of Friedrich Krischen (in: Schrader 1941, 51, fig. 70), but fundamentally

New Approaches to the Temple of Zeus at Olympia

xiii

altered according to the reconstruction proposed in the text and with translucent roof tiles and high parapet. Fig. 9. The cella after its modification and the erection of Phidias' statue of Zeus.

Chapter Two (pages 50-55) Fig. 1. Zeustempel, Metopen-Triglyphenfries über dem Pronaos. After Curtius – Adler 1892, 138 Fig. 170. Fig. 2. Pronaos des Zeustempels, linke obere Ecke der Augiasmetope und Oberteil der Atlasmetope mit den Versatzmarken. After Curtius – Adler 1892, 174 Fig. 205; 177 Fig. 207. Fig. 3. Zeustempel, Atlasmetope in Aufriss und Schnitt, Ausklinkungen zum Heben markiert. After Curtius – Adler 1892, 151 Fig. 173. Fig. 4. Zeustempel, Versatzmarken auf drei Löwenkopfsimen, und auf der Oberseite sowie E auf der Rückseite. After Willemsen 1959, Pl. 123. Fig. 5. Zeustempel, Versatzmarken auf Dachziegeln aus parischem Marmor. After Dittenberger – Purgold 1896, 695. Fig. 6. Zeustempel, Versatzmarken auf Dachziegeln aus pentelischem Marmor (die Nr. 18 und 75–88 sind Unternehmermarken) After Dittenberger – Purgold 1896, 706–708. Fig. 7. Zeustempel, Rekonstruktionszeichnung einer Giebelecke mit Marmordach After Curtius – Adler 1892, Pl. 16. Fig. 8. Zeustempel, Kalypterfragment aus parischem Marmor mit Versatzmarken auf der Oberseite

Chapter Three (page 71) Fig. 1. Proportionale Zusammenhänge zwischen den Laufbahnen der Olympien, der Heräen, dem olympischen Zeustempel, dem Poseidontempel in Isthmia und der Cheopspyramide

xiv

List of Figures and Tables

Chapter Four (pages 83-89) Fig. 1. West pediment A. Pentelic marble. Olympia Museum. Photo: Olga Palagia. Fig. 2. West pediment V. Parian marble. Olympia Museum. Photo: Olga Palagia. Fig. 3. West pediment B. Pentelic marble. Olympia Museum. Photo: American School of Classical Studies at Athens: Alison Frantz Photographic Collection PE 124. Fig. 4. West pediment U. Pentelic marble. Olympia Museum. Photo: American School of Classical Studies at Athens: Alison Frantz Photographic Collection PE 126. Fig. 5. West pediment B. Olympia Museum. Photo: American School of Classical Studies at Athens: Alison Frantz Photographic Collection PE 128. Fig. 6. East pediment O. Parian marble. Olympia Museum. Photo: American School of Classical Studies at Athens: Alison Frantz Photographic Collection PE 85. Fig. 7. West pediment B. Olympia Museum. Photo: American School of Classical Studies at Athens: Alison Frantz Photographic Collection PE 18. Fig. 8. West pediment Q. Parian marble. Olympia Museum. Photo: American School of Classical Studies at Athens: Alison Frantz Photographic Collection PE 147. Fig. 9. Portrait statue from the Heraion. Parian marble. Olympia Museum 144. Photo: Olga Palagia. Fig. 10. Head of Agrippina the Elder. From the west slope of the Acropolis of Athens. Athens National Museum 3554. Photo: Olga Palagia.

Chapter Seven (pages 107-109) Fig. 1: Olympia, Zeus Temple, Atlas metope. Photo: © Hirmer Photoarchiv.

New Approaches to the Temple of Zeus at Olympia

xv

Fig. 2: Vienna, Kunsthistorisches Museum 3695, Attic red-figure Kylix by Douris. Photo: © Kunsthistorisches Museum, Vienna. Fig. 3a, b: Cambridge, Corpus Christi College, Attic red-figure Amphora attributed to the Polygnotos-group. Lewis Collection. © Fitzwilliam Museum, University of Cambridge. By permission of the Master and Fellows of Coprus Christi College, Cambridge. Fig. 4: Basel, Skulpturhalle, model of north metopes 24 and 25, Parthenon. Photo: © Dieter Widmer (E. Berger, Der Parthenon in Basel. Dokumentation zu den Metopen, Basel 1986, pl. 3,3). Fig. 5: London, the British Museum E73, Attic red-figure Kylix by the Kleophrades-painter. Photo: © Trustees of the British Museum.

Chapter Ten (page 159) Fig. 1: Les rythmes de la production monétaire éléenne, d’après Ch. Seltman

Chapter Eleven (pages 184-185) Fig. 1: Scheme of Proceeding for the use of written sources in Classical Archaeology (by the author) Fig. 2: Portion of the Plan (to the north) of the sanctuary of Zeus. The space marked by the letter “H” gives an idea of the possible place of the Bomos of the Heralds mentioned by Pausanias.

Chapter Twelve (pages 196-200) Fig. 1. Traditional and virtual 3D reconstruction model of the temple. Fig. 2. Illumination of the chryselephantine statue through the door: (a) reconstruction drawing after Hennemeyer 2011, Fig. 1; (b) virtual 3D model Fig. 3. Illumination of the chryselephantine statue through a hypothetical window in the roof: (a) reconstruction drawing after Hennemeyer 2011, Fig. 2; (b) virtual 3D model

xvi

List of Figures and Tables

Fig. 4. Illumination of the chryselephantine statue through the translucent marble tiles Fig. 5. Conceivable reconstructions of the central group of the east pediment Fig. 6 Open arrangement Type „A”: reconstruction drawing after Herrmann 1972. Fig. 7. Figure G: original fragments (left), virtual 3D model of the fragments (centre), complete virtual 3D reconstruction of the figure (right) Fig. 8. Virtual 3D models according to the closed arrangement Type „A” Fig. 9. Figures G and K according to the open arrangement Type „A”.

Chapter Thirteen (pages 215-222) Fig. 1. Polykleitos’ Doryphoros found in the Palaestra of Pompei (here called “Pompei Statue”) Naples, National Archaeological Museum, inv. 6011 Fig. 2. Polykleitos’ Diadoumenos found in the House of Diadoumenos on Delos. Athens, National Museum, inv. 1826 Fig. 3. Bronze group of the so-called “Dancers” found in the Villa dei Papiri, Herculaneum. Naples, National Archaeological Museum, inv. 5604-05, 5619-21 Fig. 4. 3D scan in the National Archaeological Museum in Naples in 2007 Fig. 5. Alignment of two 3D foot models Fig. 6. Visualization of the distances with color-mapping. Here the threshold is 2mm. Fig. 7. “Mechanical copy” in marble (Pfanner 1989) Fig. 8. Polykleitos’ Doryphoros copied by Apollonios as a herm, found in the Villa dei Papiri in Herculaneum (here called “Bronze Head”) Naples, National Archaeological Museum, inv. 4885

New Approaches to the Temple of Zeus at Olympia

xvii

Fig. 9. 3D shape comparison of the Doryphoros (Bronze Head, left) and the Diadoumenos (right). The distances to the latter are plotted on the former. The scale of the Doryphoros and the Diadoumenos is 1 : 0.978. Threshold 4mm. Fig. 10. A comparison of the right feet of the Doryphoros (left) and the Diadoumenos (right). The scale of the Doryphoros and the Diadoumenos is 1 : 0.978. Threshold 4mm. Fig. 11. A comparison of the left feet of the Doryphoros (left) and the Diadoumenos (right). The scale of the Doryphoros and the Diadoumenos is 1 : 0.978. Threshold 4mm. Fig. 12. Amazon types. From left to right: Sciarra, Sosikles, Mattei. (after Boardman 1985 fig. 195) Fig. 13. Bronze herm of the Mattei type Amazon found in the Villa dei Papiri in Herculaneum. Naples, National Archaeological Museum, inv. 4889 Fig. 14. Amazon head of the Sosikles type. Rome, Musei Capitolini, Palazzo dei Conservatori, inv. 1091 Table 1. A comparison of the left feet of the three Amazon types with the Doryphoros and the Diadoumenos. Threshold 4mm. Table 2. A comparison of the right feet of the three Amazon types with the Doryphoros and the Diadoumenos. Threshold 4mm. Table 3. Face comparison of the three Amazon types with the Doryphoros and the Diadoumenos. Threshold 4mm.

Chapter Fourteen (pages 229-232) Fig. 1. Sites of the World Heritage List Fig. 2. A mobile device showing the actual state and original state of an inaccessible house in Pompeii (Casa dei Ceii). Fig. 3. Reconstruction of a residential area of ancient Syracuse, in Greek period

xviii

List of Figures and Tables

Fig. 4. The reconstruction of ancient Metapontum sanctuary Fig. 5. The reconstruction of Temple of Apollo in Syracuse Fig. 6. An inaccessible monument: The Hypogeum of Festoons in Taranto. Fig. 7. Virtual 3D reconstruction of the same monument. Fig. 8. The objects found in the Hypogeum and used in the virtual reconstruction in the current exhibition of the Nat. Arch. Mus. of Taranto.

Chapter Fifteen (pages 246-250) Fig. 1. The overview of the collaborative framework for real-time large 3D model interaction. Each client is connected to all three servers. Fig. 2. The client application window. On this system users can share, manipulate 3D objects, and communicate with other users. Fig. 3. Collaborative manipulation. Each user can manipulate 3D objects in the shared 3D space with the following operations: translation, rotation, change of scale, and change of color/material properties. Fig. 4. The annotation tool. Users can add annotation markers on the surface of the 3D object and create information pages formatted in HTML that can be shared among remote users. Fig. 5. Pointer tool. User A is using the pointer tool (in the first person perspective) as a host user while remote users, such as User B, can see the avatar of User A pointing at the specific feature with an arrow. Fig. 6. Network protocol for collaboration. Each client sends their viewpoint, the states of active objects, the usage of collaborative tools, and chat messages in each display frame. The server sends all connected users’ states, the changed states of manipulated objects, the usage of tools by other users, and messages. Fig. 7. Hierarchical 3D mesh. The left image represents the process of recursive decomposition of meshes based on the voxel space. The input model is decomposed as a tree structure, simplified by each tree node, and

New Approaches to the Temple of Zeus at Olympia

xix

converted to small mesh-patch based LOD representation such as shown on the right. Fig. 8. Hybrid rendering using geometry and image. We pre-render the appearance of the 3D object from grid points around that object. In the online display process, the images rendered at the closest points to the user’s active viewpoint can be used to improve the rendering quality by projecting it onto the sparse geometric data as a texture. Fig. 9. Sharing process of user’s 3D images. We clip, quantize, and compress the depth and color images from Kinect capture data. That data is transferred via the cloud and shared with other users. Fig. 10. The case studies of the collaboration work in the Roman sculptures’ research project. Top left: the comparison of ancient Roman sculptures to determine whether they originate from the same original statue. Top right, Bottom left and right: the scenes of demonstration and presentation of collaborative discussion between remote locations.

ABBREVIATIONS

Abbreviations generally follow the current usage of the German Archaeological Institute (www.dainst.org). In addition, the following ones are used: Asylia Choix Delphes

CID

Curtius – Adler 1892

DNO

FD

Furtwängler 1897

IED / Minon 2007

IK Ilion IvM

K. J. Rigsby, Asylia. Territorial Inviolability in the Hellenistic World, Berkeley 1996. A. Jacquemin – D. Mulliez – G. Rougemont (ed.), Choix d’inscriptions de Delphes, traduits et commentées, Paris 2012. J. Bousquet, Corpus des inscriptions de Delphes. 2. Les comptes du quatrième et du troisième siècle, Paris 1989. E. Curtius – F. Adler (Hg.), Olympia. Die Ergebnisse der von dem Deutschen Reich veranstalteten Ausgrabung. II.1. Die Baudenkmäler, Berlin 1892. Der Neue Overbeck. Die antiken Schriftquellen zu den Bildenden Künsten dr Griechen, 5 vols. Berlin/Boston 2014. É. Bourguet, Fouilles de Delphes. III, 1. Inscriptions de l'entrée du sanctuaire au trésor des Athéniens, Paris 1929. A. Furtwängler, Olympia. Die Ergebnisse der von dem Deutschen Reich veranstalteten Ausgrabung. IV. Die Bronzen und die übrigen kleineren Funde von Olympia, Berlin 1897. Minon, S. 2007. Les inscriptions éléennes dialectales (VIe-IIe siècle avant J.-C.). I. Textes. II. Grammaire et vocabulaire institutionnel, Genève 2007. P. Frisch (Hg.), Die Inschriften von Ilion, Bonn 1975. O. Kern (Hg.), Die Inschriften von Magnesia am Maeander, Berlin 1900.

New Approaches to the Temple of Zeus at Olympia

xxi

IvO / Dittenberger 1896 W. Dittenberger – K. Purgold (Hg.), Olympia. Die Ergebnisse der von dem Deutschen Reich veranstalteten Ausgrabung. V. Die Inschriften von Olympia, Berlin 1896. Mythos Olympia 2012 W.-D. Heilmeyer – N. Kaltsas – H.-J. Gehrke – G. E. Hatzi – S. Bocher (Hg.), Mythos Olympia. Kult und Spiele, Berlin 2012. NIO P. Siewert – H. Taeuber (Hg.), Neue Inschriften von Olympia. Die ab 1896 veröffentlichten Texte (Tyche, Sonderband 7), Wien 2013. Olympia 1875 – 2000 H. Kyrieleis (Hg.), Olympia 1875 – 2000. 125 Jahre Deutsche Ausgrabungen. Internationales Symposion, Berlin 9.-11. November 2000, Mainz 2002. PMG D. L. Page – M. Davies (eds.), Poetarum Melicorum Graecorum Fragmenta. I. Alcman. Stesichorus. Ibycus, Oxford 1991. Treu 1897 G. Treu, Olympia. Die Ergebnisse der von dem Deutschen Reich veranstalteten Ausgrabung. III. Die Bildwerke von Olympia in Stein und Ton, Berlin 1897. XIII. Olympia-Bericht H. Kyrieleis (Hg.), XIII. Bericht über die Ausgrabungen in Olympia. 2000 bis 2005, Berlin 2013.

INTRODUCTION ADOPTING A NEW APPROACH TO THE TEMPLE AND ITS SCULPTURAL DECORATION ANDRÁS PATAY-HORVÁTH

Abstract Earlier and recent studies by the author concerning the east pediment and the historical circumstances of the temple construction are summarized and updated in the light of recent publications. Iconographical, literary and numismatic evidence is combined to suggest an intimate connection between the Greek victory over the Persians and the genesis of the entire monument.

Introduction For more than a decade, I was investigating the reconstruction and interpretation of the east pediment of the temple of Zeus at Olympia. The resulting view is highly unconventional and is still much debated, but has inevitably led to the conclusion that the sculptural decoration cannot be treated in isolation and any study of the temple has to begin with the analysis of the historical setting of its construction. This principle required an interdisciplinary approach, which emerged gradually during the last years, but I hope that it has yielded some new results which can further the discussion. In this paper, I summarize the results of my studies concerning the east pediment and the historical and economic background of the temple construction; I hope to demonstrate that the two topics are intimately connected to each other. At the same time, I also take the opportunity to reflect on those remarks which have appeared in the meantime.

2

Introduction

The Problems of the East Pediment My first encounter with the temple of Zeus was in 2001, when, in preparing my PhD thesis, I was allowed by the German Archaeological Institute to measure the tiny holes on the pedimental sculptures, which served to fasten the lost metal attachments. Eventually, the discussion of the many holes on figure G of the east pediment (conventionally called Pelops), i.e. the reconstruction and the dating of its metal attachments, became the central part of my work and inevitably led to a detailed investigation of the iconography of the east pediment. The comparison of the fragments with the description and interpretation given by Pausanias resulted in the thesis that the information provided by the text is not to be taken at face value.1 It is generally accepted that the identification of the crouching maid (figure O) or that of the sitting elderly man (figure N) in the east pediment are evidently mistaken, but scholars have usually assumed that the overall interpretation given by Pausanias is – some minor details apart – fundamentally correct and can serve as a secure basis both for the identification of the individual figures and for the reconstruction and interpretation of the whole composition as well.2 However, this is not a matter of fact, but only a hypothesis, which is not absolutely sure, since Pausanias could not rely on authentic written documents concerning the original interpretation and might have been mistaken. Moreover, there are fundamental problems concerning the central figures K, G and I as well. To start with the so-called Pelops, it was already noted by Carl Robert at the beginning of the 20th century, that his armour would be most unusual or even inappropriate, since Pelops is actually never represented in Greek art with shield, helmet and cuirass, and that figure G clearly appears here with all this equipment is therefore a strong argument against identifying him as Pelops. Robert did not suggest any other identification for the figure and interpreted the scene as a general depiction of a warrior’s departure.3 This was of course unconvincing for several reasons: in general, such a pedimental composition was rightly expected to represent a specific and well-known mythological scene and the cuirass in particular might have been a later addition and not part of the original composition. So it is not surprising that Robert’s idea was not favoured by

1

Patay-Horváth 2004. Most recently e.g. Barringer 2008; Kyrieleis 2011, 2013. 3 Robert 1919, 291-296. 2

Adopting a New Approach to the Temple and its Sculptural Decoration

3

anybody.4 But actually the basic iconographical problem concerning the armour of Pelops remained unsolved and the date of the lost metal cuirass was not discussed in detail either. It was exactly this problem which I intended to investigate by collecting and studying the reconstruction and dating of lost metal attachments. After examining ca. 400 pieces of sculpture showing traces of lost metal attachments, I arrived at a reconstruction of the so-called Pelops which is slightly different from the current one and is of considerable importance regarding the date of the attachments and therefore also for the interpretation of the figure. The argument is based on the fastening holes on the temples of the figure: they are not identical in their diameter and some are drilled not in the usual way horizontally into the stone, but sloping downwards. All these features are paralleled on a well-known head from the west pediment of the Aphaia temple and show that the holes were used to fasten both some locks of hair and cheekpieces turned upwards. Fastening metal locks of hair was particularly fashionable during the late archaic and early classical period and is attested only sporadically afterwards, and the same applies to cheekpieces turned upwards. So the reconstructed parts belong most probably to the original rendering of the figure and cannot be attributed to some late restoration or alteration. Since the fastening holes of the cuirass are exactly identical in their diameter with those on the temple of the figure, this attachment is also likely to have been added already in the first half of the 5th century. One can therefore conclude, that the figure was fully equipped with hoplite armour and consequently it is unlikely to have represented Pelops.5 Of course, if this were the only and most serious problem concerning the interpretation, one could accept it as an exceptional rendering of the local hero. But there is another rather more compelling reason against the traditional view, the hairstyle of figure K, which is usually identified either as Hippodameia or Sterope, according to the arrangement adopted for the central group. Trying to find parallels for this hairstyle, I realized that this is a very special one and is typically employed for a rather limited group of female figures: mourners, servants, slaves, courtesans, i.e. for those on the lowest ranks of society. At least this is the general rule during the classical period and does not seem to have been restricted to any geographical area. It is quite inappropriate therefore either for the princess

4

Buschor 1932, 158-160 was the only one to follow Robert in rejecting the PelopsOinomaos interpretation, but he immediately suggested another mythological explanation. 5 For details see Patay-Horváth 2006 and 2008, 55-64.

4

Introduction

Hippodameia or for the queen Sterope.6 It is completely incompatible with the traditional interpretation and strongly suggests that another should be looked for. Before presenting this, however, a recent objection should be discussed briefly. In the latest paper dealing with the iconography of the pediment, H. Kyrieleis thought to refute my argument concerning the hairstyle by referring to some archaic and early classical pieces.7 It is symptomatic, I think, that he questioned the general validity of the observation by referring to a very limited number of exceptions, which would be in any case statistically irrelevant against the bulk of the evidence. Actually all the parallels adduced by Kyrieleis are geographically and chronologically isolated and cannot be regarded as inconvertible proof against the general meaning of the „Pagenfrisur”. In fact, there is only one of them, the Demeter of the large Eleusinian relief, where the interpretation of the figure is beyond doubt and shows a goddess. But in this case, the length of the hair and the hairstyle are both markedly different, and the parallel itself is therefore misleading.8 Kyrieleis also disregards the general observation, made already earlier and independently from me, that this hairstyle is rather rare in classical sculpture,9 and does not offer any explanation for its occurrence on figure K of the pediment; moreover he also completely ignores the testimony of vasepainting, which is of course relevant for interpreting any iconographic element occurring in monumental sculpture. So I see still no reason to disregard the implication of this hairstyle and consider the traditional interpretation of the pediment as fundamentally erroneous. It was most probably borrowed by Pausanias from his local guides, and does not reflect the original intentions of the 5th century, but a deliberate attempt of the Eleans to connect this most important sculptural group of the temple with their own mythical history. From a purely theoretical point of view, it would be desirable to base the interpretation of the pediment exclusively or at least mainly on the fragments themselves and I think it is possible to approach the question in this way and not by accepting or assuming the correctness of the interpretation given by Pausanias. What the east pediment actually depicted will thus emerge from the monument itself. 10 6

Patay-Horváth 2005. Kyrieleis 2013, 67, n. 37. 8 Holloway 1958, 403: „Thus the parallel with the Sterope is superficial and exists only in the regularity of the Demeter’s waving hair.” 9 Bol 1988, 76: „Bei Frauendarstellungen gerät das halblange Haar in archaischer Zeit um die Mitte des 7. Jahrhunderts ausser Mode.” 10 The following summary is based on Patay-Horváth 2007. 7

Adopting a New Approach to the Temple and its Sculptural Decoration

5

First the notion of a chariot race or the preparations for such a contest is to be discarded, even if it may seem at first plausible because of the two chariot teams flanking the central group. Actually, these antithetic chariots and horses are not characteristic elements for the depiction of a chariot race, but can be seen in a variety of cases and are a convenient device filling the middle part of any pediment, and characterize at the same time the protagonists as belonging to the wealthiest and noblest heroes.11 The central figure of Zeus is also of fundamental importance. He turns his head to his right and holds a thunderbolt in his left. This feature is exceptional and is therefore most likely significant as well. It has been supposed that Zeus would punish Oinomaos with this weapon after the chariot race, but in all cases where Zeus holds this attribute in his left hand, it is clearly not his intention to use it against any of the persons represented around him. If he wants to use the thunderbolt, he brandishes it in his right hand. So the peculiar position of his weapon cannot be interpreted in this way, but rather the contrary is to be supposed, that he is happy with and benevolent towards the figures surrounding him. The exact arrangement of these figures has been the subject of long scholarly debates, but is not discussed here in detail, because it will be treated separately.12 Now, I proceed with the iconographical analysis. According to the reconstruction which is in my view most probably the correct one, the two male figures are turned towards each other and both are clearly characterized as warriors. The older one, figure I, is depicted with open mouth and can be most probably interpreted as speaking. What he is saying is directed towards the younger one, figure G, but as Zeus stands between them he is probably also concerned with the speech. It is therefore reasonable to assume that he is invoking the supreme god and uttering an oath for the younger warrior. Last but not least, the female figures should also fit in the evidently military context and figure K must be a servant or slave and at the same time she has to play an obviously important or decisive role for the entire scene. The other one is characterized by her pose as anxious and it is a reasonable suggestion that she is worried about the fate of the young hero standing next to her.

11

A similar composition of antithetic chariots was depicted on the pediment of the Athenian treasury at Delphoi and on the west pediment of the Parthenon. The scheme often occurs in vase-painting as well, and has obviously nothing to do with any kind of chariot race. Examples are collected and discussed in Patay-Horváth 2007, 187-190. 12 See chapter twelve below in this volume.

6

Introduction

There are not, I think, many scenes of Greek mythology meeting all or most of these requirements and the preparations for the chariot race between Pelops and Oinomaos can practically be ruled out. An important, but seldomly represented scene from the Iliad, the reconciliation of Agamemnon and Achilles is far more probable in my view. ”And the son of Atreus made prayer to Zeus; and all the Argives sat thereby in silence, hearkening as was meet unto the king. And he spake in prayer, with a look up to the wide heaven: “Be Zeus my witness first, highest and best of gods, and Earth and Sun, and the Erinyes, that under earth take vengeance on men, whosoever hath sworn a false oath, that never laid I hand upon the girl Briseis either by way of a lover's embrace or anywise else, but she ever abode untouched in my huts. And if aught of this oath be false, may the gods give me woes full many, even all that they are wont to give to him who so sinneth against them in his swearing.”13

The older warrior can thus be identified with Agamemnon, the younger one with Achilles. Agamemnon is swearing an oath addressing Zeus and other deities that he has not touched Briseis and returns the girl to Achilles. She is a captive and therefore, her hairstyle is appropriate, but at the same time, she clearly plays a central role in the scene. It is perhaps also interesting to note, that her real name was according to tradition Hippodameia. Achilles has just declared that he will return to the battlefield, so it is not surprizing that his armour is rendered so elaborately. He has just received these new pieces from his mother Thetis, who can be seen in the worried female figure F standing next to him; mothers are often accompany their warrior sons in Greek art, and there is an especially close parallel, showing the worried Penelope standing next to Telemachos in a similar way. It might be objected that Thetis is not mentioned by Homeros an attendant of the reconciliation scene, but, as the presence of Telemachos in this relief clearly shows, this is not a decisive point. From the remaining figures, the seer N might be Kalchas, the maid O an attendant Nereid and the reclining figures in the corners, if they were really intended to represent river gods, can be the two rivers of Troy, the Skamandros and Simoeis. The exact identification of the other figures is impossible, but they cannot be identified accepting the traditional interpretation either.

13

Hom. Il. 19. 255-265, English translation by A.T. Murray

Adopting a New Approach to the Temple and its Sculptural Decoration

7

The Economic Background and the Historical Setting Now, even though the iconography clearly favours this interpretation, it might seem to be iconologically weaker, since Pelops was according to Pausanias (5.13.1) the most important local hero of the Eleans and the temple was also said by Pausanias (5.10.2) to have been built by the Eleans, from the booty taken from the rebellious Pisatans. But actually the case is not so simple. A few general points deserve to be mentioned first: 1) the reconciliation between Achilles and Agamemnon is practically never depicted in Greek art. This would seem to argue against the new interpretation of the pediment, but actually is rather confirming it. During the early classical period, there were many important works of art created showing scenes of the Trojan cycle which were never chosen before or afterwards. And precisely in Olympia there was the famous sculptural group of Onatas dedicated by the Achaeans, depicting another assembly of the Greeks at Troy.14 Scenes of the Trojan war became obviously especially popular in this period and so the choice of subject seems less strange and certainly more appropriate than the chariot race of Pelops and Oinomaos. 2) Pausanias took his information concerning the monuments of Olympia in most cases from his local guides,15 and these were certainly eager to connect every significant monument to the history of Elis. Their bias is apparent not only for us, but was already felt by Pausanias, who himself detected and refuted such a local-patriotic interpretation concerning a scene on the chest of Kypselos (5. 18. 6-7). That he was misled in the case of the temple and did not realize the problems does not necessarily mean that the pieces of information he was provided with, were correct. In this way, the testimony of Pausanias concerning the construction of the temple can be doubted as well, and the case is even more serious than the problems concerning the interpretation of the east pediment. The probability that Pausanias was misguided is much higher indeed, because he did not even have the possibility to check the information of his local guides. Regarding iconography, he could at least see the monuments and thus check what he was told (and indeed, in the case of the east pediment, he reported an alternative tradition concerning the name of one figure, the

14

Paus. 1. 15. 2-3; 10. 25-27; 5. 25. 8-10. The guides are attested already for the early imperial period (IvO No. 64, 77, 83, 110, 120 and Varro Men. Frg. 34 apud Nonius 676, 8) and are frequently referred to by Pausanias as well (cf. Jones 2001, 33-39). 15

8

Introduction

charioteer of Pelops16) but in the case of „Baugeschichte” he could only report what he was told. And his local guides at Olympia provided demonstrably wrong pieces of information in several cases. Pausanias (5. 16. 1) says e.g. that the Heraion was built in the 8th year of Oxylos by the people of Skillous. The latter statement is simply unreasonable given the size of the building, which finds contemporary parallels only at Argos and Tegea, but the date is definitely and markedly wrong, since the temple was built not in the dark ages implied by Oxylos, but only around 600 B.C.17 Pausanias (5. 15. 1) also says that the Leonidaion was built by a local, i.e. Elean man Leonidas, although it is attested epigraphically that he was a Naxian. The error is however not to be attributed to the periegetes, because the inscription was at the time of his visit already covered under a thick layer of mortar, so must have derived from his local guides.18 They were obviously distorting the evidence in order to glorify Elis. The mentioning of an otherwise unknown local architect for the temple of Zeus (Paus. 5. 10. 4) can thus be seen as yet another example of these local-patriotic tales, but certainty in this matter is of course beyond our reach. The financing of the temple construction from the Pisatan booty is, on the contrary, demonstrably a local myth and has nothing to do with real history. If we accept the ancient testimonia, including Pausanias (6. 22. 24), on the date of this war between Pisa and Elis as historically correct, the war must antedate the temple construction by ca. 100 years and thus cannot be connected with it. But actually, not only the war but even the early existence of a political entity named Pisa has been convincingly questioned recently and so the booty from such a war is ruled out entirely.19 It has also been suggested, that the war mentioned here by Pausanias was identical with the so-called Triphylian war, which is mentioned in passing by Herodotus (4.148). Actually, the phrasing does not conclusive suggest one single war, but even if this had been the case, the temple construction could not have resulted from the hypothetical booty taken from these cities, simply because it was surely insufficient for 16

Paus. 5. 10. 7. For a summary see Mallwitz 1972, 137-143, for details concerning the chronology Mallwitz 1966. 18 IvO No. 651. Cf. Mallwitz 1972, 246-247. 19 Luraghi 2008, 79: „The suspicion becomes certainty at least in the case of Pisatis, which everyone now agrees never existed as an independent political entity before being created by the Arcadians in 365 BC.” K iv 2013 can be regarded as the last and most probably unsuccessful attempt to restore the historicity of archaic Pisa. Cf. Roy, forthcoming, note 14. 17

Adopting a New Approach to the Temple and its Sculptural Decoration

9

financing the architecture, let alone both the temple and the chryselephantine statue, as stated by Pausanias (5.10.2). It is unnecessary to argue these points in detail, since I have presented the arguments elsewhere recently.20 Instead, I give a brief summary on the costs of the temple building first, and discuss the financial resources, which might have been used for the building project. In this way the real history of the temple construction and its historical circumstances will hopefully emerge. The costs of building a Doric temple are attested in some particular cases both in the literary and in the epigraphic record and therefore it is possible to make an estimate in the case of the temple of Zeus as well. The size of the building is of course one of the most important factors and this temple is not only the largest one on the Peloponnese, it is also slightly larger than the late archaic temple of Apollo at Delphoi and a bit smaller than the Parthenon. (Fig. 1) The building costs are known for both approximately and suggest therefore a similar figure for the temple of Zeus as well. Adopting another method one can also calculate the costs of each architectural part on the basis of the building records preserved at Athens, Eleusis and Epidauros and in this way the resulting figure will be roughly the same. So one can, I think, confidently state that the construction cost around 400 Talents and the chryselephantine statue must have been also quite expensive. Comparing this amount with figures known from Elis is revealing and strongly suggests that this polis was not likely to be able to finance the whole project on its own. Contributions in cash from many other Greek states would have most probably left some traces in our testimonia, just as in the case of the Delphic temples and as the Delphic case clearly shows, it usually took much time to gather the amount required to complete the project. In contrast with Delphoi, the huge temple in Olympia was completed very rapidly even though a serious earthquake in 464 BC must have caused serious damage and therefore delay as well. So financing by voluntary contributions offered by numerous sponsors can safely be ruled out and one is still wondering what the source of the huge amount of money was and how it reached the sanctuary. If the sanctuary itself possessed some sort of special income, e.g. taxes, fines connected with the Olympic games or deriving from land lease, this could perhaps have accumulated and been used for financing the costly project. But actually we do not have any evidence for such incomes, at least not on the scale required for the temple building, and the practice of accumulating financial resources for a large project would be also 20

Patay-Horváth 2012a; 2012b.

10

Introduction

exceptional in this period. It is therefore likely that only the sudden appearance of a single large donation can account for the building, which is actually a votive dedication itself. Considering that monuments dedicated in the sanctuary were almost exclusively connected with military successes and athletic victories and the latter motivation for the temple building can be safely ruled out, it is to be assumed that the temple was actually erected as a victory monument and this idea is clearly attested by the Nikai decorating the building and the cult statue as well. The suggestion made by Pausanias, that the construction was financed from a war booty is therefore the most probable one and it is only the war specified by him which has to be modified. The amount of the booty required for building the temple must have been enormous and clearly rules out not only the local conflict mentioned by Pausanias but any other similar war, like the sack of Triphylian cities assumed by modern scholars as well. The pieces of information concerning the scale of the booty deriving from individual battles or military campaigns in the Greek world are sufficient to conclude that the booty from which this temple was financed must have been an exceptional one implying a very prosperous enemy as well. Because the date of the temple construction is sufficiently secure, it is actually quite evident that the war from which the necessary booty derived can only be sought in the great campaign of Xerxes and especially in the battle of Plataia. That the booty taken on this occasion was enormous is related by Herodotus, but he gives no figures. If we try to figure it out from the scattered pieces of evidence, we arrive at some thousands of talents, an estimate which is only to be expected after all. That the battle or the booty is not specifically mentioned in connection with the temple does not have to come as a surprise. Exactly the same can be seen at the same time in Sicily, where the literary evidence is equally abundant, but the temples at Himera and Syracuse, evidently built as victory monuments and financed from the booty or built by the captives taken from the defeated Carthaginians according to modern research, are never mentioned as such by ancient authors. One might perhaps wonder, why Olympia has been chosen, and not the battlefield itself as in the case of Himera. There are, I think, at least two factors which effected this. In general, the selection of a more or less neutral panhellenic sanctuary was most appropriate for celebrating a panhellenic victory in the form of a temple. We have a nice and almost contemporary parallel for this in the case of the great temple’ of Apollo at Delos. The construction of this temple seems to have begun during the second quarter of the 5th century as well, but the project was abandoned around the middle, and this is most probably

Adopting a New Approach to the Temple and its Sculptural Decoration

11

explained by the transfer of the treasury of the Delian league to Athens. Why Olympia and not some other panhellenic sanctuary has been chosen for erecting the victory monument commemorating the battle of Plataia is also quite transparent. Apart from Delphoi, there were practically no other candidates, and Delphoi did not play a glorious role during the campaign of Xerxes. Although the sanctuary could save face and eventually also received costly votive dedications after the victories, it is easy to understand that the main dedication, i.e. the bulk of the booty, did not go to Pythian Apollo but to Olympian Zeus. The Greek victory at Plataia was actually attributed not so much to the bravery of the general Pausanias or to the warriors, but to the seer Teisamenos, who was a native Elean and, as the story related by Herodotus (9.33-35) clearly shows, he was pretty influential in Sparta and could definitely influence the choice for the victory monument. Moreover, his success seems to share an intimate connection with the building of the temple: his first military victory was at Plataia, the fifth and last one at Tanagra, which was commemorated on the apex of the east pediment by a gilded shield and an appropriate and partly preserved inscription. The numismatic evidence can also be adduced as a corroboration and has to be discussed in a bit more detail. The coinage bearing the name of the Eleans was been connected with the temples of Olympia already a century ago by Ch. Seltman.21 He assumed that there were two mints, one attached to the temple of Zeus and the other, later, one with the temple of Hera, and called the whole output temple coins of Olympia implying that each series of coins was specially produced for partivular Olympic games, starting from the late 6th century onwards. Actually, all these assumptions can be refuted and it seems more likely that the coinage was not a series of special festival issues, but a normal civic coinage of Elis, and there were not two mints working parallel to each other, but only one and the production did not start in the late 6th but in the early 5th century, around 480-470 B.C. Furthermore, it is very likely that the issues were not connected with the Olympic games and the output was not constant, as supposed by Seltman. He tried to distribute the series evenly over the centuries, but a fluctuation between extremely intensive minting activity attested by many obverse and reverse dies in the earliest decades followed by a more sporadical minting (clearly less dies) in the third quarter and a minting gap during the last quarter of the 5th century seems much more

21 Seltman 1921. On the history of research see chapter ten in this volume by F. Wojan.

12

Introduction

likely. A new start around 400 BC was marked by the introduction of the Hera coinage. The evidence for this chronology derives mainly from the coin hoards and is therefore quite compelling, but of course I cannot argue all points in detail here.22 What seems to be important for the present purpose is that the highest minting activity conspicuously coincides with the temple construction and this suggests that the two phenomena were directly connected. Actually, this suggestion was made earlier as well,23 and it is quite reasonable to assume that the Elean minting was necessitated by the construction of the temple or more generally by the „Großbaustelle Olympia” (involving the construction of a new stadion as well), since early Greek coinages were usually not issued for a commercial purpose, but for uniform payments made to a large number of people, i.e. to soldiers or to workmen employed on a large-scale.24 Looking at the amount of money, which was most probably produced during the second quarter of the 5th century (ca. 400-500 talents) and comparing this amount with the estimated costs of the temple construction referred to above (ca. 400 talents), one can immediately recognize that they are roughly comparable. I would also add the evidence provided by the coin types, especially the reverse types showing Nikai. Following Seltman, these were most often connected with athletic victories but in that case one would not understand why this motive was abandoned afterwards. Since these Nikai are restricted, accepting the above chronology, to the second quarter of the 5th century, it is likely that they were used only due to a special reason, which was not connected with Olympia or the games in general, but with the temple and its construction. As the temple was most probably a victory monument financed from the booty of the celebrated campaign and was adorned with different Nikai symbolizing this victory, it is most reasonable to suppose that the Nikai on the coins have the same meaning: they show that the coins were struck from the booty or rather from the dekate deriving from the booty, which formed the financial background of the temple construction.25 In this way, both the chronology, the size and the iconography of the coinage point to the same conclusion, which was reached independently, by analyzing the report of Pausanias, the construction costs of the temple and its implications, and it can be reasonably concluded that the temple was celebrating the Greek success at Plataia and was financed from the dekate deriving from this booty. This 22

For details see Patay-Horváth 2013a. Chantraine 1958, 15; Kyrieleis 2011, 36-38. 24 Patay-Horváth 2013b. 25 Patay-Horváth 2013c. 23

Adopting a New Approach to the Temple and its Sculptural Decoration

13

conclusion is admittedly hypothetical, but is actually consistent with all the available evidence and is far more probable than any other one. By way of conclusion, I would like to return to the iconology of the east pediment and ask in the light of the above considerations, which one of the two interpretations is more likely: the one involving Pelops or Achilles? Pelops can be connected to the Persian wars because he was the ancestor of Agamemnon, the leader of the Greeks fighting at Troy, and even his shoulder blade was said to be essential for conquering Troy, but his chariot race has obviously nothing to do with this.26 The Homeric reconciliation scene, on the other hand, involves Agamemnon himself and would therefore appear as a more direct reference to the war. Moreover the scene depicts the moment in which Achilles acknowledges Agamemnon as a leader of the allied forces and is about to fight again. As Agamemnon was clearly considered as a Spartan, at least by the Lacedaemonians themselves,27 his representation on the east pediment would be perfectly appropriate for a monument celebrating a decisive victory under Spartan leadership. This leadership is moreover characterized by the selected scene not as a tyrannical but as a fair one, since Agamemnon is depicted in the act of returning what he has taken from Achilles previously. Regarding Achilles, his paramount importance for the Greek success at Plataia is clearly shown by the recently identified elegy of Simonides. This poem, which starts with an invocation of Achilles and in which he „serves as a paradigm for the collective Greeks at Plataia”28 also contains a passing reference to Pelops, but this can only show, that the importance and honours attached to Pelops at the time of the temple construction were much inferior to that of Achilles. So, in general terms, the east pediment showed the ending of the quarrels between the allied Greeks, which was of central importance and an indispensable precondition for victory not only in the Trojan but also in the Persian wars, and the west pediment depicted the joint struggle against the enemy. So the Homeric interpretation of the east pediment fits in my 26 Pelops as an ancestor of Agamemnon is attested already by Hom. Il. 2, 104-5, the story of his shoulder-blade is described by Paus. 5.13. 4-6. Herodotus mentioned Pelops most probably only twice (7.8 and 11): Xerxes was supposed to describe Pelops as a subject of his own ancestors, who already conquered the Greeks. Concerning Herodotus 7. 159, Griffiths 1976, 23-24 has clearly and most convincingly demonstrated that it did not contain a reference to Pelops. 27 Herod. 7. 159 is the clearest indication of this tradition (cf. most recently Osmers 2013, 143-149), already attested by archaic poets, e.g. Stesichorus and Simonides (Schol. Eurip. Or. 46). Cf. also Pind. Pyth. 11. 32; Nem. 8. 12 and Paus. 3. 19. 6. 28 Boedeker 2001, 158.

14

Introduction

view the intellectual climate of its time29 much better than the traditional one and it is also in perfect accord with the historical circumstances of the temple building outlined above. To put it in another way, the sculptural decoration of the temple in general, and specifically the pediments, have often been connected to the experience of the Persian wars,30 and the above reconstruction of the events and the interpretation of the sculptures make this connection only closer and immediately recognisable and is therefore perhaps not so radically opposed to the traditional view as it would appear at first.

29

Barringer 2009, 243: „The Trojan conflict offers a parallel to the recent conquest of a great foreign enemy by an assembly of Hellenes from various poleis and was intended to liken contemporary Hellenes fighting against foreigners … to the noble Achaiains of the distant Homeric past. This link is made explicit by poetic fragments of c. 479 B.C. by Simonides, who associated recent battles of the Persian wars – specifically the Battle of Plataia – with Tojan War heroes, especially Achilles.” 30 In addition to the essays collected by Herrmann 1987, see more recently e.g. Sinn 1994, Gauer 2012a, Kyrieleis 2013.

PART I ARCHITECTURE

CHAPTER ONE THE TEMPLE ARCHITECTURE AND ITS MODIFICATIONS DURING THE FIFTH CENTURY BCE ARND HENNEMEYER

Abstract Around 430 BC, Pheidias created a new colossal cult statue in ivory and gold for the temple of Zeus at Olympia. With the installation of the statue, substantial modifications of the cella’s original design from ca. 472– 456 BC were carried out. While the statue of Zeus is today entirely lost, a recent architectural documentation of the temple’s remains provides new insights into the structural changes made to the cella at the time of the statue’s installation. These changes were much more substantial than previously thought: the cella floor was entirely remade, including the construction of a shallow basin in front of the cult statue; the interior colonnade was dismantled and re-erected with different axis widths; and the roofing was probably also renewed. These interventions have been at least partially known for over a hundred years, but their design has usually been misinterpreted as mediocre and backward-looking. The new data from fieldwork now paints a completely different picture. The design of the cella cleverly employed different materials and followed advanced design principles comparable to contemporary Attic architecture and sculpture. Perhaps most importantly, the refashioning of the cella also introduced a new spatial hierarchy, which regulated the approach to the cult image and the perception of it. Therefore, the new reconstruction of the temple’s interior also provides a framework for raising novel questions concerning the ritual practices performed in one of the most important cultic places of Antiquity.

Research History The temple of Zeus at Olympia, built in the second quarter of the 5th century BC, is the largest sacred building in the Peloponnese and

Architecture and Modifications during the 5th Century BCE

17

undisputedly the highest-quality building of its time. Its architecture is regarded as the purest implementation of the Doric order – especially in the German research literature, but also in English research. Despite the significance of the building’s architecture, our judgements concerning it are still based mainly on the first publication by Wilhelm Dörpfeld in 1892.1 This brilliant publication represented best practice at its time and was a milestone in developing the methodology in the field of ancient architecture. Dörpfeld focused on regaining an image of the building as a whole, a characteristic approach of the late nineteenth century. The purpose of studying ancient buildings was already mainly scientific, but there was still also a more general interest: detecting prominent architectural patterns for contemporary reuse. Thanks to the clear engravings, the reconstruction was promptly popularised in handbooks of architectural history and even in many school books. In addition, Dörpfeld also considered modifications and repairs of the temple. The question whether the colossal statue of Zeus was originally intended or added later was of special interest, of course, because of the prominence of Pheidias and the Zeus statue – one of the Seven Wonders of the World, as described in the written sources. Among the architectural sculpture of the temple, Georg Treu detected so-called Ersatzstücke, three figures in the west pediment and many lion heads of different repairs of the sima.2 Some architectural members reused in later buildings proved that even the main building structure had been repaired. In addition, some architectural members reused in later buildings also documented repairs on the building structure.3 This view of the building was notably forward-looking, respecting as it did the fact that a building is not a constant but undergoes a process of change. In the reconstruction, however, these thoughts were not integrated. Dörpfeld’s publication is still the only one covering the architecture of the temple as a whole and it remains the most important account, not least because of its clear drawings. But despite its high quality, the publication cannot be used as a basis to address new questions. It is generally not very attentive to detail, and was formed with a certain schematism – typical for the late nineteenth century. Repairs as well as later added furnishings were This project was supervised by Wolf Koenigs and Manfred Schuller at the Technische Universität München and financed by matching funds of the FritzThyssen-Stiftung and the German Archaeological Institute. I wish to express my gratitude to all of them. 1 Dörpfeld 1892. 2 Treu 1892, Treu 1897. 3 Dörpfeld 1892, 22.

18

Chapter One

only marginally documented. About thirty years later, in the 1920s, his publication on the temple of Zeus was already out of date when Armin von Gerkan postulated methodical principles of building archaeology (or Bauforschung). When studying a building and publishing research about it, we should strictly separate the documentation of the remains from the reconstruction in different drawings and chapters and, above all, the reconstruction has to be entirely traceable from the whole to the details.4 From the present point of view, more methodological problems are apparent. Dörpfeld and the other members of the ‘Old Excavation’ (1875– 81) first made sketches in the field; only later were they compiled into final drawings. Due to this separation of the documentation process, errors inevitably crept in. Most significantly, Dörpfeld and the others measured only few selected samples of the numerous scattered blocks, complemented by some non-standard architectural blocks – for example, of the corners. That was appropriate for the target of an idealised reconstruction. However, small variations remained unperceived. But as we know today, such variations are often the results of intended refinements. Later, Dörpfeld himself considered the publication as not detailed enough and as partly preliminary.5 We also have to remind ourselves of other published investigations and papers of primary research – primary in the sense that the work is based on field research and a thorough knowledge of the ruin. Unfortunately, in recent papers and handbooks, they are often neglected.6 In 1935, Dörpfeld restudied two topics with the participation of the famous Hungarian architect Fred Forbat. The first topic was the extensive redesign of the cella floor which was carried out when Pheidias erected the colossal statue of Zeus; based on the redesign, Forbat reconstructed in reverse how the floor of the cella was originally designed.7 For the second topic Dörpfeld identified the damage to and repair of the western front side in the Hellenistic period.8 He assumed that it was of a quite limited extent: some members of the horizontal as well as the raking geison were replaced and, on the western corners, the entablature was reinforced with cramps that were inserted on the visible faces. In the 1930s, the Bauforscher Hans Schleif began a detailed investigation at Olympia and published a preliminary report in 1943. Concerning the temple, the main topics were architectural refinements, such as differences between the interaxes and a 4

Gerkan 1930/1959. Dörpfeld 1935, 222–26. 6 e.g. Younger – Rehak 2009. 7 Forbat 1935. 8 Dörpfeld 1935, 256–61. 5

Architecture and Modifications during the 5th Century BCE

19

possible inclination of the columns, as well as the identification of an abnormal solution at the corner triglyphs of the pronaos and opisthodomos.9 In his paper An Archaeological Earthquake, printed in 1941, William Dinsmoor proved that the temple had been repaired to a much greater extent than had been assumed until then.10 This idea shattered the firm conviction of the excavators that the monument was almost entirely a remnant of the original early classical temple of the architect Libon, mentioned by Pausanias. In the details, Dinsmoor misplaced quite a lot of the attributed architectural members and was erroneous in the dating. But nevertheless he raised the question of extensive repairs to the temple and indicated that it was necessary to investigate all the remains, stone by stone. From then on, the repairs remained a central topic of research. After World War II and the suicide of Schleif, the Bauforscher Friedrich Krauss restarted the investigation in two field campaigns in 1952 and 1956. His observations on the upper parts of the pronaos convinced the last doubters of an extensive repair.11 However, he interpreted other differences in detail, not as a result of a different date of origin, but as intended variations. On the column fragments, he recognized significant differences in the shape between the front sides and the long sides concerning the tapering, fluting, and profile of the echinus.12 The example of the columns should represent the potential for future work on the temple. The archaeologist Franz Willemsen published a detailed study on the lion heads of the sima in 1958.13 He could identify at least 144 heads, whereas originally the roof comprised only 100. According to Willemsen, about half of the preserved heads are later replacements referring altogether to nine different repairs – the first in the fourth century BCE, some decades after the erection, and the last in the fourth century AD, probably under the reign of Diocletian. Actually, the separation of the repair phases as well as the dating are arguable. But in any case, the extent of repeated repairs was thus proved, albeit not directly linked with repairs to the architectural structure. As Friedrich Krauss was fully engaged in studying the temples of Paestum, he entrusted his assistant Peter Grunauer with the documentation and publication concerning the temple of Zeus. Unfortunately, Grunauer never finished the work. But he reported important preliminary results in four papers written between 1969 and 1979. By regaining almost all the 9

Schleif 1943, 29–30. Dinsmoor 1941. 11 Krauss 1957 a. 12 Krauss 1957 b. 13 Willemsen 1959. 10

20

Chapter One

measurements in the blocks of the geison, Grunauer could eliminate an inclination of the columns. Concerning the reconstruction, this change entails slightly wider pediments – an important condition for the arrangement of the sculptures. Furthermore, Grunauer detected that the pediments had been composed in strict correlation to the triglyph frieze.14 All four sides of the peristasis are constructed with a slight curvature that was arranged from the third foundation level on – about one and a half metres below ground level; this was proved by precise levelling of the crepidoma and by sondages at the four corners as well as in the middle of each of the four sides. In the same course, Grunauer could determine that the joints between the ashlar keep the concordance not only where visible but from the bottom layer on, even respecting the corner contraction: before the construction of the first layer, the layout plan and the structure were entirely preordained.15 Last but not least, Grunauer amended Dinsmoor’s study of the repairs through many more observations.16 At that time, Alfred Mallwitz excavated an area next to the so-called South-Eastbuilding. A layer in front of it held a huge number of small fragments belonging to the temple of Zeus – most of them cannelures of column drums. So Grunauer and Mallwitz laid down the convincing hypothesis that the South-East building was entirely built from reworked blocks of the temple’s east façade.17 In addition, Klaus Herrmann studied the fragments of the anta capitals that were complemented by new excavated fragments.18 He could differentiate between a probable original type and a much later repair type, and could partly reconstruct the original one, including its painting.19

Topics of the Current Project The current project, which was initiated by Wolf Koenigs after the death of Grunauer and is reported upon in this chapter, started in 2006. The aim is to produce a comprehensive monograph about the architecture of the temple. We focus on the original shape of the building, the analysis of the design of the temple itself as well as in the context of temple architecture in the fifth century BCE, and the question of which function 14

Grunauer 1974, 17–25, 39–40, 43. Grunauer 1980. 16 Grunauer 1981, 275–80. 17 Grunauer 1981, 280 esp. n. 129a; Mallwitz 1981, pp. 108–10. 18 Herrmann 1981. 19 Ibid.; Herrmann 1983, 11; fig. 18 with a minor correction of the proposed reconstruction. 15

Architecture and Modifications during the 5th Century BCE

21

the architecture of the temple of Zeus served in the cult at Olympia. Until now, the image of the temple has been more or less a snapshot at a certain moment, but it is unclear when this snapshot was taken. We complement this image by detailing the processes of rescheduling during the construction, repairs, later modifications, furnishings, and, finally, the destruction history, to give the building its proper temporal depth in the history of the sanctuary. To prepare for the project, in 1998 and 1999 we documented the remains in situ in two detailed plans – in German, called Maßplan and Steinplan – as is the common standard. According to the basic principles of the discipline Bauforschung, the measurement has to be based on an independent metrological network to document exactly the deformation of the ruin. In the first plan (the Maßplan), all the measurements accurate to a millimetre are inscribed – the length of each stone, the width of gapping joints, and cracks, etc. This enables us to mathematically eliminate the deformation. In the second plan (the Steinplan), technical details such as mortises and cramp holes and the state of conservation are portrayed. The second main pillar of the documentation is the inventory of all preserved fragments and the exact measurement of all important ones, as already demanded by Dinsmoor;20 this work was finished in 2010. Grunauer’s mostly unpublished Trümmerpläne showing all the scattered fragments in their exact positions served as a very valuable support for this work. The inventory contains about 1,000 pieces, including many newly identified ones. For example, almost five times more epistyle fragments are now known. For different important reasons, we could neither turn nor excavate the stones. Of course, this comprehensive and time-consuming documentation was not done for its own sake: it is the methodological precondition to answering the questions mentioned above and will serve as a database for future questions as well. While the monograph is still in preparation, we have published extracts of the research results in eight papers up to now. The papers study different topics as follows: – The research history of the temple of Zeus in the 19th and 20th centuries21 – The observation that the two lines of columns inside the cella differ greatly from the original layout22 – The repair history over a time span of eight centuries23 20

Dinsmoor 1941, 406. Koenigs 2002. 22 Hennemeyer 2006. 21

22

Chapter One

– When the front sides of the peristasis were re-erected in the fourth or third century, the details of the columns were optimized to use the play of sunlight and shadow24 – The illumination of the cella25 – The modification of the architecture inside the cella when Pheidias created the statue of Zeus26 Finally, a paper for the Olympia exhibition catalogue in 2012 provides a short summary of the architecture and some important new results.27 In the following examination, we attend to the architecture of the temple, especially in the fifth century BCE. The general characteristics of its architectural structure as well as its members have been published widely. I assume that they are known. However, in the face of the long history of extensive repairs, the question arises: Which parts of the ruin actually belong to the fifth century? In the case of the marble sculptures, the answer is quite evident, though nevertheless disputed.28 In the case of architectural membra disiecta, however, it is much more difficult because the dating signs are much less evident. As we must conclude by classifying all fragments according to their technical characters and the varieties of the local shell limestone used at the temple, the repairs of the peristasis were drastic. Both façades were entirely rebuilt from the lowermost column drums, the eastern one with new material and the western one by reuse of some original pieces. On the northern long side, only one original capital is preserved; it is the capital of the recently reerected column.29 Only on the southern side do most members belong to the original structure. Consequently, the marble sculptures of both pediments were positioned at least twice. Furthermore, the entablatures of the pronaos and the opisthodomos have been repaired with new pieces, as well as the upper parts of at least two antae. However, concerning the walls, it is almost impossible to arrive at further conclusions: because the ashlar of the walls were very reusable, they were taken away from the site; therefore, unlike all the other classes of architectural members, less than one percent of the ashlar is preserved. 23

Hennemeyer 2013 a; Hennemeyer 2013 c; (Hennemeyer 2010, partly outdated). Hennemeyer 2011, 104–9. 25 Hennemeyer 2011, 101–4; Hennemeyer 2013 b, 4–5. 26 Hennemeyer 2013 b. 27 Hennemeyer 2012. 28 In opposition to the communis opinio, Brinkmann 2012, 40–44 suggests that the sculptures were created only around 400 BC. 29 About the anast losis of the column (N12) cf. Herrmann 2013a. 24

Architecture and Modifications during the 5th Century BCE

23

The Design of the Temple and Metrological Problems When talking about the architecture of the temple of Zeus, we have to allude to its design and metrological problems. The measuring unit that was used for the temple was 32.04 cm long, the so-called ‘Olympic Foot’. At Olympia, it was also identified at the stadium and recently at the retaining wall behind the treasuries,30 but at no other sites. At the beginning, Dörpfeld had identified this foot unit for the temple (fig. 1).31 But later he assumed another slightly longer foot unit, called the ‘Pheidonic foot’, was used because he followed the then common sense that there existed only a few different foot units – not including the Olympic one.32 To analyse the large number of measurements taken at the ruin and on its members, I have used the method from Hans-Georg Bankel33 in many different data sheets. This method is, I believe, still the best one – if not the only usable one – for detecting the applied measuring unit of a building.34 Concerning the lower part of the building – in detail the crepidoma, foundation, and the toichobate – the diagrams argue clearly for the Olympic foot.35 As I have mentioned previously, the temple of Zeus is often regarded as the purest implementation of the Doric order. For this reason, some inconsistencies in the design are usually ignored. A few main dimensions with simple figures in feet determined the further design process.36 Perhaps they were defined in the work order, as Dörpfeld already assumed.37 Between the outer edges of the walls, the cella was planned to be exactly 100 feet long and 50 feet wide, and exactly centred in the 200foot-long peristasis that was twice the cella length. Such definitions of proportions that are not related to the structure of the building and result in conflicts were avoided in later temple layouts. At the temple of Zeus, however, it was only after these definitions that the peristasis was designed, in a way that was completely consistent. According to a new ideal, the builder planned the interaxes between the columns of the peristasis with exactly the same width on all four sides of the temple: 16 1/4 30

Herrmann 1999, 371 f. Curtius et al. 1879, 26. 32 Dörfeld 1892, 18 f. 33 Bankel 1993, 117–19, 141–44. 34 The method has drawn criticism unjustifiedly, e.g. by Pakkanen 2013, 3. 35 The reasoning including the diagrams will be presented in the monograph under way. 36 Hennemeyer 2012. 37 Dörpfeld 1892, 19. 31

24

Chapter One

feet (i.e. 260 dactyls – the subdivision of the Greek foot). The perfect order of all elements ranged from the lowermost layer of the foundation to the roof tiles. Even the sculptures in the tympana corresponded with the triglyph frieze. For the sake of symmetry, the opisthodomos on the backside of the cella was exactly as long as the pronaos. But concerning the relationship between the peristasis and the naos (that is, the cella, pronaos, and opisthodomos), the primary definitions resulted in problems. On the long sides of the peristasis, the front sides of the naos foundation align exactly in the middle of the second intercolumnium. This problem could easily be solved by trimming the pronaos and opisthodomos as required. But on the front sides of the peristasis, the case is much more difficult: according to a generally assumed rule, the side wall should align in the centre of the second columns. Indeed, the naos was 50 feet wide, but the three theoretically corresponding interaxes of the peristasis were just 48 3/4 feet. This is the original planned layout as it can be reconstructed on the basis of the regular block lengths in the foundation, crepidoma, and toichobate. But between the crepidoma and toichobate, on the one hand, and the upper parts of the building, on the other, we observe certain discrepancies (fig. 2). The cella walls are arranged slightly off-centre to their foundations. And the columns are excentric to the blocks of the stylobate – this effect is dependent on the position of the columns: the nearer they are to a corner of the temple the more they are excentric. The discrepancies are small but quite systematic and much bigger than the inaccuracy during construction. Obviously, the layout and all the upper parts of the temple were slightly redesigned when the crepidoma and toichobate had been finished. By the analysis of all observed discrepancies and thanks to the remarkable accuracy of construction, we can identify the particular alterations, their degree, and – thinking backwards – reconstruct the original layout. We have to conclude that the columns were considerably diminished by about 16 cm in diameter. The distance edge of the stylobate to the outside of columns was kept; therefore, the centre of the columns was arranged nearer to the edge. Furthermore, the interaxes were widened by the minimum amount of about 1 cm on the long sides and 2.5 cm on the front sides. On the contrary, the naos was narrowed by around 16 cm. Compared with the total length of the stylobate (64.08 m), the alterations appear insignificant. What could have been the goals of the redesign? There are three aspects. First, the column proportions underwent a clearly perceptible change. Second, the discrepancy between the width of the cella and the three interaxes in the middle of the front sides decreased considerably, from 1 1/4 feet to just about 1/4 foot, or not more than 4 cm on

Architecture and Modifications during the 5th Century BCE

25

each side. Third, the proportions of the entire structure were brought to mathematical perfection. For example, the rectangle of the entablature received a width-to-length ratio of exactly 3:7 (measured at the taenia on the epistyle).38 But after the redesign, even the quite new achievement of a standard interaxis was slightly dismissed. With regard to the metrology, the alteration had a shocking effect. Simplified, the measurements in the upper parts of the building differ from the foot unit observed in the crepidoma. After these observations of metrological problems and design, we turn to the interior of the temple.

The Modification of the Cella When Pheidias created the statue of Zeus the cella was modified as well. New observations place this well-known fact in a very different light.39 The dating of the statue to at least twenty years after the temple was more or less completed is non-controversial today.40 As already noted in antiquity, the statue gave the impression that it was too big in relation to the dimensions of the cella.41 Obviously, the objective was to exceed the dimensions of the Athena Parthenos which had previously been made by Pheidias.42 The only remains of the statue of Zeus are some scant fragments of the base. But thanks to the design of some coins of the Roman Imperial era and the detailed description by Pausanias, we have a certain idea of the statue. It was visualized in reconstructions by Quatremere de Quincy and Friedrich Adler, to mention the most influential drawings.43 Furthermore, some main measurements of the statue are provided in a fragment from the early Hellenistic writer and librarian Callimachus of Cyrene.44 The expense of the colossal chryselephantine statue must have been enormous; at the sculpture and its throne, there was a concomitant program of about 100 figures specifying a whole cosmos of its own. In view of these costs, it is not surprising that 38

It is not evident why the integer ratio at the taenia of the architrave was of particular interest. But similarly, at the early classical temple of Hera at Paestum, there is also an integer ratio, of exactly 2:5, according to Mertens 2006, 286. 39 Hennemeyer 2013b. 40 Mallwitz – Schiering 1964, 169–80. Cf. Lapatin 2001, 79–86. 41 Strabon, Geographika, 8, 3, 30. 42 Concerning the competitive aspects between the temple of Zeus at Olympia and the Parthenon at Athens, and respectively their statues, see Hurwit 2005, Taraporewalla 2011. 43 Quatremère de Quincy 1815; Friedrich Adler, in Dörpfeld 1892, 15–16. 44 Callimachus, Iambus 6; Pfeiffer 1941; Pfeiffer 1949, 188–91.

26

Chapter One

the architecture of the cella was thoroughly modified as well. This modification had been already an important issue for the excavators and was later studied in detail by Dörpfeld and Forbat.45 The subject was regarded as solved, and therefore for the following seventy years until our investigation, the cella was never re-examined. Originally, the cella was clearly lengthwise oriented. According to the Doric design tradition, two double-storey lines of columns divided the cella into a wider central nave with a homogenous pavement and smaller side aisles. When the statue was erected, the middle nave was divided into four sections by the statue itself and its base by erecting parapet walls, deepening a field in front of the statue as a shallow basin and paving it with dark grey Eleusinian limestone, and, finally, by constructing a rim of Pentelic marble bordering the dark grey field and the statue base. According to Dörpfeld and Forbat, other measures have to be dated much later. Compared to this current state of research, the recent documentation brought important changes. Indeed, the modification was much more profound. Even the two double-storey lines of columns and their stylobates were dismantled and re-erected (fig. 3).46 To bring the cella architecture up to date, wall pillars were newly integrated as a connecting element between the line of columns and the flat wall. In return, it was necessary to reduce the interaxis between the columns. Some observations prove that this change was not a part of the previously described redesign during the original construction. On the other hand, it cannot be dated later than the erection of the statue. Obviously, both measures were linked. Most probably, when the columns were dismantled it was also necessary to untile the roof. At this point, we have to focus on the question of light. Dörpfeld argued convincingly against the idea that the temple of Zeus was a hypæthros and accepted the door as the only natural source of light inside the cella.47 But different authors saw the illumination of the colossal statue in the back of the cella as problematic,48 especially as the head of Zeus was just under the ceiling (i.e. in the darkest part of the cella). Furthermore, light from below gives a dæmonic impression to a face, which seems to be totally inappropriate for Zeus. In this context, a new observation on a sample of roof tile fragments is meaningful. Until now, the known marble roof tiles supported Dörpfeld’s reconstruction. But in truth these tiles are from later repairs. Some of the original tiles, however, 45

Dörpfeld 1892, 11–16; Dörpfeld 1935, 222–26; Forbat 1935. Hennemeyer 2006; Hennemeyer 2013 b, 6–7. 47 Dörpfeld 1892. Later Dörpfeld’s position was less strict, cf. Dörpfeld 1913, 8– 12. 48 For example, Wölfel 1990. 46

Architecture and Modifications during the 5th Century BCE

27

were much thinner.49 As is well known, they consist of Cycladic marble, which is exceptionally translucent.50 Therefore, the daylight could shine through them with considerable intensity. On a summer’s day, light comparable to a conventional 20-watt bulb would have shone through each of the 1,000 tiles, altogether equivalent to about 20,000 watts (resulting in an illuminance at ground level of at least 200 lux). And this striking effect might be part of the modifications made to illuminate the statue, especially its upper parts. Such translucent roof tiles were already in use on the Cyclades in late archaic times, as Aenne Ohnesorg has shown in detail.51 In the temple of Zeus, this entails another change in the reconstruction (fig. 4): there cannot have been a wooden ceiling above the cella, but the marble roof itself must have covered the room. With respect to other details, the modification of the cella is much more complicated than has been thought, and has its own history. I mention only the uncanny fact that the statue base, the dark grey field, and the rim of Pentelic marble were apparently planned before the lines of columns were dismantled and re-erected.52 In the dating of other measures mentioned before, Dörpfeld was probably wrong. The interpretation and evaluation of these measures, however, is essentially linked to their dating. A lime floor was applied almost everywhere in the cella, according to Forbat, as a later repair.53 But a closer look refutes his dating argument (fig. 5). Behind the statue the lime floor covered a gap of about 4 cm between the marble rim and the original pavement. Forbat supposed the gap was damage from an earthquake. But, in truth, the gap must be from the time when the statue and the marble rim were added. The foundations around this area (i.e. the foundations of the cella walls and under the stylobates of the columns) are perfectly preserved and all joints are still closed. This would not be the case if the mentioned gap had opened later. But, above all, the gap cannot be the result of a settling, for the joint would not have been opened but rather would have been compressed. This forces the following conclusion: the marble rim was never tightly joined to the older pavement. Thus, the lime floor was made more or less immediately after the erection of the statue. The question then arises: Why was the lime floor added? It is not very convincing to posit that its surface was 49 Until this observation, there was no reason to mistrust the reconstruction by Dörpfeld. Under the new circumstances, Hennemeyer 2011, 102–4, is now partly outdated. 50 Herrmann 2000. 51 Ohnesorg 1993; Ohnesorg 2011. 52 Hennemeyer 2013 b, 7. 53 Forbat 1935, 236, 247; Dörpfeld 1935, 260.

28

Chapter One

considered more precious than the original pavement of large blocks of shell stone, although the surface was formed by a smoother, thin white layer.54 In regard to the purpose, let us remember that the lime floor contains volcanic trass. Such mortar hardens by a hydration reaction and is therefore water-resistant. Usually, we associate it with later times; however, for cisterns it was already in use at the time when Pheidias created the statue of Zeus.55 A similar purpose could have been intended: the water-resistant floor was perhaps laid to preserve the ivory of the statue against the humidity of the marshy environment that was described by Pausanias.56 If this was the case, the measures were successful, for the statue was conserved for more than 800 years. Together with the lime floor, the intercolumniations along the statue and behind it were closed, probably with a wooden lattice, as we can conclude by the holes at the bottom of the columns and a long dowel hole behind the statue.57 The apparently hidden area behind the statue was carefully designed: during the modification, its design was twice changed. At first this area was understood to be a part of the middle nave. The marble rim should follow the stylobate until the rear wall: the block under the south-eastern corner of the base was prepared with a protrusion as a joint to the next block in the direction of the rear wall (fig. 6). The front of the marble rim was perfectly smooth; obviously, at first, it was planned to lower the floor level to the lower edge of the marble rim. But then, instead of that, the area was heightened by a 6-cm-thick layer of lime floor up to the level of the floor in the side aisles. And finally, it was heightened again by a second layer of about 3 cm to the level of the stylobate. We can virtually look over the workman’s shoulder to see how he was struggling to arrive at the best solution. The final version presents itself as a reduced form of the groundbreaking design idea in the cella of the Parthenon, presumably by Pheidias. There, as a figurative architectonic pattern, the lines of columns 54 The smooth layer is not preserved but it was clearly documented by Dörpfeld in his sketch, temple of Zeus, fol. 3. 55 According to Brinker – Garbrecht 2007, 100, the composition of plaster to build cisterns had clearly changed at the time when Rome expanded to the Mediterranean East; volcanic trass was substituted by crushed ceramic tiles. According to Frerich Schön, coordinating member of the research project “Technologietransfer in der Antike – Untersuchungen antiker hydraulischer Mörtel mit analytischen und numerischen Methoden,” there is clear evidence that hydraulic lime was already used in the western and central Mediterranean area in the first half of the fifth century. 56 Paus. 5.11.10. 57 Forbat 1935, 236; pl. 19.

Architecture and Modifications during the 5th Century BCE

29

embrace the statue of Athena on three sides.58 It is undisputed that the structure of the cella substantially contributed to the perception of the statue of Athena Parthenos. Thenceforth, the interior design of the cella was a central issue in the construction of Doric temples.59 But in the temple at Olympia, it was necessary to respect the existing structure. Therefore, it is not surprising that the modification used more visual and sublime effects, to correlate with the new comprehension of interior design. Nevertheless, this was a lot of effort as regards alterations and new integrated elements. Wrongly, the design of the modification has usually been misinterpreted as mediocre and backward-looking. In truth, its particular elements re-interpreted the cella and its space in a sophisticated way. The inner part (i.e. the base of the colossal statue and the shallow basin which was as wide as the entire middle nave) was distinguished by superior stones – dark grey limestone and, in contrast, white marble.60 As already mentioned, according to Pausanias, there was olive oil on the slightly recessed field of dark grey limestone.61 The intention might have been to lay a reflecting surface in front of the statue in order to mirror the statue and reflect the light from the door62 and the translucent roof tiles. But, above all, the liquid could have been a perfect means to prohibit an improper approach to the statue. A closer examination shows that the dark grey pavement, the statue base, and the rim of Pentelic marble form a single entity. The pavement and the base are not only linked by the marble rim and the use of the same dark-grey Eleusinian limestone, but they are also disposed by the same specific raster of joints, which are different from the architecture around them (fig. 7). Creating a dynamic ambiguity, the marble rim merges two functions. On the one hand, it borders the dark-grey paved field as well as the statue base; on the other hand, it serves as the stylobate of the lines of columns. The base, the paved field, and the stylobate interpenetrated almost as in a painting by M. C. Escher. This principle of combining antithetic characteristics is far beyond the original design of the temple. It is similar to some characteristics of buildings in Athens contemporaneous to the statue of Pheidias; for example, the hypercontraction at the corners of the Parthenon. In the field of visual arts, the contrapposto is related.63 58

Gruben 2001, 182 (1st ed. 1966). Leypold 2005, 101–3. 60 cf. Shoe 349. 61 Paus. 5.11.10. 62 Wölfel 1990, 44. 63 Gruben 2001, 181–82. 59

30

Chapter One

In the temple of Zeus at Olympia, the nucleus we have just described was enclosed, as already mentioned, by a kind of lattice along the statue and behind it: parapet walls that had been painted on their inner sides with mythological scenes surrounded the dark-grey paved field, three on each side in the intercolumniations alongside it and three in front of it. Only scant remains of the parapet walls have survived. Therefore, their height can only be hypothetically reconstructed. Until now, it was assumed that they had a maximum height equal to the shoulders, according to the premise that the visitors must be able to see the paintings on the opposite parapets.64 Correspondingly, the main function of the parapet was accepted as being to prevent people from entering the inner area.65 But the parapet also fulfilled an architectural, spatial function – in a sense, to articulate the nucleus within the cella. And it would have achieved this function much better if it was taller than a man. Perhaps that is what Pausanias had in mind when he emphasized with a not very comprehensible comment that the parapets had been made “in the kind of walls.”66 Even if the parapet was about 2.10 m high, the statue would have been very visible everywhere in the cella. In the view of a visitor entering the cella, the parapet would have covered no more than the ankles of the statue (fig. 8). Despite the destroyed state of the ruin, the appearance and furnishing of the cella has been comparatively well recovered. Because the architecture of Greek temples was generally very regular, we have to postulate functional reasons where we find that the architecture or interior elements differ from the standard. I totally agree with this methodological approach as established by Dieter Mertens concerning the specifics of the architecture of Magna Graecia.67 In this sense, we have to interpret the measures of the modification (fig. 9). In addition, the appearance of the elements indicates an architectural semiotic sense. Obviously, the front area until the parapet wall was open to the public; otherwise, the barrier would not have been necessary. According to Pausanias, there was an altar inside the cella,68 which is very unusual for ritual praxis in ancient Greece. On a monthly tour, the Eleans sacrificed on a huge number of altars; the one in the temple had the prominent second position. Perhaps we have to

64

Dörpfeld 1935, 253. About visual functions of parapet walls in Greek temples cf. Zinserling 1957, 18–20; Mattern 2007, 156–57; Mylonopoulos 2011a, 284. 66 Paus. 5. 11. 4: . 67 Mertens 2006, e.g. 121–22. 68 Paus. 5. 14. 4. 65

Architecture and Modifications during the 5th Century BCE

31

link a square bedding in front of the parapet with this altar.69 Of course, from Pausanias’ text we know only that the altar existed in his period – that is, in the second century AD. But there are indications that it was a part of the fifth-century modification.70 In alignment with the front parapet wall, a door in each side aisle limited access. The question is: Who was allowed to enter? According to some indications, a visit to the temple did not generally end at the parapet wall. I have already emphasized that the area behind the statue was carefully designed and even twice altered. If only the priest had been allowed to enter, why was this area closed to the side aisles and rearward by the parapet walls and lattices? Besides, a part of the statue’s iconographic programme could not be seen face on.71 The side aisles, and the back area behind the statue that was as wide as the aisles, formed a Pishaped ambulatory. Inevitably, we associate this architectural composition with processions. One could enter in one aisle and leave from the other. What about access to the innermost area? The parapet wall across the middle nave was only about 6.50 m long but nevertheless had two openings, repeating the scheme in the side aisles. Therefore, it was demonstratively penetrable. Apparently the barrier could be crossed much more often than was generally assumed. In this sense, another characteristic is also important: the parapet wall was not adjacent to the shallow basin, but at a distance of a formidable 3 ½ m it formed a kind of platform that could be used by certain visitors. Here they faced the god and were encircled by the mythological figures painted on the parapet walls. About the lost statue, we can only speculate. Perhaps it was optimized to be seen from this platform?

Conclusion As we have seen, the modification articulated the space inside the cella, created an extraordinary atmosphere, and organized different zones inside the cella with different grades of approach to the goddess. We have to interpret this as a means to control and strengthen the perception of the Zeus statue. In this context, we can cite a didactic poem written by Ioannes Tzetzes for the German-born Byzantine Empress Irene in order to make

69 Mallwitz 1972, 230. Differently interpreted by Forbat 1935, 228, 238 as a “Roman pedestal” or “large base.” 70 The details will be presented in the monograph under way. 71 According to Paus. 5.11.4, namely an amazonomachy with Theseus and Heracles.

32

Chapter One

her fitting for the imperial family:72 at an art competition at Athens, both Pheidias and his follower and rival Alcamenes had been asked to design a column monument with the goddess Athena standing on top. Alcamenes created a perfectly charming and womanlike statue in the shape of a young girl; but Pheidias’ statue had a distorted face and other deformations. Therefore, the decision was made in favour of Alcamenes’ statue. However, when the statues had been lifted and placed atop the columns, Pheidias’ statue revealed his mastery of his art. Having mastered the rules of both optics and geometry, he knew that an object looks different when set farther away and higher up. Now Pheidias was highly praised while Alcamenes was mocked. Of course, we do not know if this anecdote, written 1,600 years after the death of Pheidias, tells of a real incident or is a later fiction. But there are also much earlier and more credible sources which praised Pheidias for the masterly presentation of his sculptures.73 According to these sources, Pheidias respected the impression of a sculpture on the beholder. At the same time – in the second half of the fifth century BCE – the first perspectives had been created for theatre paintings; Agatharchos, Anaxagoras, and Democritus wrote theoretical texts about this. Relief pictures used elaborate foreshortening and bottom view.74 Obviously, the difference between the real shape and its visual perception was a central topic of artistic discourse. Against this background, we should assess the modification inside the temple of Zeus: it follows the artistic ideals of the High Classical period. The colossal statue and the cella encircling it are a veritable Gesamtkunstwerk: a synthesis of sculpture, architecture, and painting in appearance and function. Rather than just setting a statue into the cella, Pheidias also organized and designed the cella for the most awesome perception of the colossal statue of Zeus.

72

John Tzetzes, Chiliades VIII, l. 333–62; Davison – Lundgreen 2009, 991–92. Davison – Lundgreen 2009, thoroughly compiling all sources and passages about Phidias. 74 Grüner 2005. 73

Architecture and Modifications during the 5th Century BCE

33

Fig. 1. Temple of Zeus, metrological scheme of ground plan, drawing by Wilhelm Dörpfeld 1880, with measurements in ‘Olympic feet’ (later erroneously rejected by Dörpfeld himself). Copyright DAI.

Fig. 2. Temple of Zeus, alteration during construction process (marking in light grey: hypothetical original layout; in dark grey: execution). ‘Steinplan’, detail of the north eastern corner.

34

Chapter One

Fig. 3. Modification of the double-storey lines of columns and their stylobate. Marked in dark grey: remains in situ; in light grey: reconstruction of phase II; line drawing: reconstruction of phase I.

Fig. 4. Temple of Zeus, cross section, after Dörpfeld 1892 (architecture) and Adler (statue), with essential modifications according to the current state of research.

A Architecture andd Modifications during the 5th Century BCE

35

Fig. 5. The floor in the arrea behind thee statue, sectioon drawing. a) remains C DA AI, Grabung Olympia; in situ in thhe 19th c. (Foorbat 1935) Copyright b) reconstruuction, lime flloor covering the gap betw ween the basee and the paving blocck; c) when thhe base settleed, the joint ccouldn't open n but was compressed..

Chapterr One

36

S

W

E

N m, block at thee south-westeern corner of the Zeus' Fig. 6. Whiite marble rim base (drawinng by Dörpfelld 1880; anath hyrosis and leggend added) Copyright C DAI.

d the shallow bbasin, both paaved with Fig. 7. Recoonstruction off the base and dark grey Elleusinian limeestone, and th he white marb le rim (on thee drawing marked in liight grey).

Architecture and Modifications during the 5th Century BCE

37

Fig. 8. Cella, view about from the threshold. Based on a drawing of Friedrich Krischen (in: Schrader 1941, 51, fig. 70), but fundamentally altered according to the reconstruction proposed in the text and with translucent roof tiles and high parapet.

38

Chapter One

Fig. 9. The cella after its modification and the erection of Pheidias' statue of Zeus.

CHAPTER TWO VERSATZMARKEN AUF BAUGLIEDERN DES ZEUSTEMPELS VON OLYMPIA ULF WEBER

Abstract Most of the extant assembly marks on architectural elements of the temple of Zeus were used for repairs. These marks usually consisted of single Greek letters engraved on cut stones to number contiguous elements. The most famous numbers are the letters alpha and gamma on two metopes with depictions of Herakles. They were probably applied during a restoration of the 4th c. BC and helped determine the original position of these two metopes. More assembly marks are found on some lionhead watersprouts. These marks derive from several repair phases including Roman imperial times as well. The same can be concluded about the assembly marks on dozens of marble roof tiles. In contrast to the known researches they derive not from the new construction of the temple, but from different later repair phases. In order to date them better was the investigation of the letter forms of the assembly marks and the kind of the used marble helpful. Hence the assembly marks contribute to date the repair phases of the temple of Zeus more precisely.

Von vielen antiken griechischen Gebäuden, die aus behauenen Steinen bestanden, haben sich Bauteile mit Versatzmarken erhalten1. Dazu gehört auch der Zeustempel von Olympia. Seine Versatzmarken wurden schon von den ersten Ausgräbern dokumentiert und gedeutet, doch dabei kam es teilweise zu Fehlinterpretationen. Durch neue, umfangreiche Forschungen zu Versatzmarken ergaben sich jetzt weitere Deutungsmöglichkeiten. So

1

Ihre Zahl geht in die hunderte, s. z. B. Weber 2013, 370–388.

40

Chapter Two

konnte festgestellt werden, dass die meisten Versatzmarken des Zeustempels im Zuge verschiedener Restaurierungsphasen angebracht wurden2. Doch bevor diese Erkenntnisse dargelegt werden, soll eine kurze Einführung zum Thema „Versatzmarken“ folgen: Versatzmarken bilden einen Teilbereich der im antiken griechischen Bauwesen gebrauchten Markierungen und Zeichen. Die meisten erhaltenen Versatzmarken bestehen aus einfachen griechischen Buchstaben, die der Nummerierung behauener Steine dienten. Diese Nummern, also Ordnungszahlen, wurden angebracht, um die Blöcke als zusammengehörig zu kennzeichnen; wobei man benachbarte Bauteile mit fortlaufenden Nummern markierte. Dabei konnten sie entweder aufgemalt oder eingemeißelt sein. Versatzmarken benutzte man folglich, um das Zusammenfügen von Bauteilen zu erleichtern. Im antiken griechischen Bauwesen hatten Versatzmarken zwei Einsatzgebiete: 1. Neubau (= Erstverwendung) 2. Wiederverwendung, d. h. wenn man Bauteilgruppen oder sogar ganze Bauten wiederverwenden wollte (einschließlich Reparaturen) 3. Die Griechen benutzten Versatzmarken vom Anfang des 6. Jhs. v. Chr. an. Ihr Gebrauch war anschließend in der gesamten Antike üblich und ebenso in der folgenden byzantinischen Zeit. Letztlich sind Versatzmarken überall im Werksteinbau zweckmäßig, wenn es darum geht, Teile vorzufertigen oder wiederzuverwenden; und das ist bis heute so geblieben4. Am Zeustempel von Olympia waren nur wenige Bauglieder aus Marmor gefertigt. Dazu zählen die Giebelskulpturen, die Metopen über Pronaos und Opisthodom sowie die Dachziegel mit der Sima. Der Rest bestand aus Muschelkalk, dessen sichtbare Außenseiten man mit Stuck überzogen hatte5. Auf Bauteilen aus Muschelkalk sind nur wenige Versatzmarken überkommen, die deshalb kaum eine sichere Interpretation erlauben. Ursprünglich gab es wahrscheinlich mehr, aber die meisten von

2

Die Ergebnisse dieses Artikels basieren auf den Forschungen für meine Doktorarbeit (Weber 2013). Ich danke der 7. Ephorie des griechischen Kultusministeriums für die Erteilung der Studiengenehmigung in Olympia. 3 Martin 1965, 221–231; Orlandos 1968, 84–87; Ginouvès – Martin 1985, 123; Hellmann 2002, 88–91; Höcker 2004, 271 s. v. Versatzmarken; Weber 2013, 1–7. 4 Weber 2013, 3. 359–364. 5 Curtius – Adler 1892, 10; Treu 1897, 151; Mallwitz 1972, 212. 219 f.; Herrmann 2000, 381; Gruben 2001, 57. 61; Younger – Rehak 2009, 100; Hennemeyer 2012, 123.

Versatzmarken auf Baugliedern des Zeustempels von Olympia

41

ihnen sind auf der verwitterungsanfälligen Oberfläche des Muschelkalks vergangen6. Besonders wichtig bei der Erforschung des Zeustempels waren die Versatzmarken auf zwei figürlich verzierten Metopen. Sie hatte man aus parischem Marmor gefertigt und über dem Pronaos und dem Opisthodom angeordnet (Fig. 1). Vor dem Pronaos trat bei den Ausgrabungen die sog. Augiasmetope in Sturzlage zutage. Ihr ist ein Alpha auf dem Oberlager eingehauen (Fig. 2). Ebenfalls vor dem Pronaos kam die Metope mit Athena, Herakles und Atlas zum Vorschein, der man auf ihrer Oberseite ein Gamma eingemeißelt hatte. Da ihr Oberlager ganz erhalten ist, kann es als sicher gelten, dass jede Metope nur eine einzige Versatzmarke trug. Folglich zeigten die Buchstaben die Reihenfolge der Metopen an. Die erste, nördlichste Metope war mit einem Alpha versehen und die dritte mit einem Gamma7. Mithilfe dieser zwei Versatzmarken gelang es ferner, den Platz der einzigen Metope zu bestimmen, die Pausanias nicht erwähnt8: Er nennt nämlich nur elf Taten des Herakles bei seiner Beschreibung des Zeustempels, obwohl Herakles doch zwölf Heldentaten vollbracht hatte und es auch insgesamt zwölf Metopen über dem Pronaos und dem Opisthodom gab. Letztlich fand man tatsächlich vor dem Pronaos Fragmente einer zwölften Metope, die die Heraufholung des Kerberos aus dem Hades zeigt. Ihre ursprüngliche Position konnte durch die beiden Versatzmarken Alpha und Gamma bestimmt werden, denn die Kerberosmetope befand sich zwischen beiden und trug wohl ein Beta als Versatzmarke (Fig. 1)9. Das würde allerdings bedeuten, dass die Triglyphen zwischen den Metopen nicht mit nummeriert waren, denn sonst hätte die zweite Metope ein Gamma und die dritte bereits ein Epsilon tragen müssen. Den wichtigsten Hinweis für die Richtigkeit dieser Annahme stellt die nördliche Ecktriglyphe des Pronaos dar, denn sie ist erhalten und weist keine Versatzmarke auf. Allerdings besitzt sie Hebelöcher aus zwei verschiedenen Bauphasen10. Eine Versatzmarke benötigte sie nicht, weil ihre Position auch ohne eine solche leicht zu erkennen war. Die Triglyphen aus Muschelkalk zwischen den Metopen wurden offensichtlich nicht wiederverwendet, d. h. sie waren vielleicht beschädigt und wurden

6

IvO Nr. 669; Curtius – Adler 1892, 20; Weber 2013, 198. Treu 1897, 148–150. 173. 176; 174 Abb. 205; 177 Abb. 207; Younger – Rehak 2009, 94 f. 8 Paus. 5, 10, 9. 9 Treu 1897, 148 Abb. 172; 149 f. 10 Curtius – Adler 1892, 10; Mallwitz 1972, 233 f. 7

42

Chapter Two

durch neue ersetzt. So fand man Reste bereits in der Antike demontierter Triglyphen in einer Wasserrinne beim sog. Südostbau11. Die figürlichen Metopen weisen ebenfalls ein Merkmal auf, welches auf ihre Abnahme und spätere Wiederanbringung hindeutet, nämlich seitliche Vertiefungen für Hebezangen (Fig. 3). G. Treu und A. Furtwängler meinten noch, diese hakenförmigen Ausklinkungen zum Heben stammten vom Neubau und die Metopen seien fertig ausgearbeitet versetzt worden12. Dies ist jedoch unwahrscheinlich: So hat sich heute die Ansicht durchgesetzt, dass die Metopen erst am Bau ausgearbeitet wurden. Dafür sprechen u. a. Details der Figuren, die unfertig blieben, wie an den Haaren oder Mantelfalten. Wären die Metopen schon am Boden fertig gestellt worden, würde man eine Vollendung dieser Details erwarten13. Außerdem waren die Vertiefungen für die Hebezangen auch nach dem Versetzen noch sichtbar. Diese Störung des äußeren Gesamteindrucks hätte man beim Neubau wohl kaum hingenommen. Aber zur Demontage und zum Wiederversetzen der Metopen bei einer Restaurierung waren sie unvermeidbar14. Das gibt ebenso für die Versatzmarken, wenn man die Metopen erneut in der richtigen Reihenfolge anbringen wollte. Einen weiteren Beleg für die Demontage der Metopen stellen die Antenkapitelle über Opisthodom und Pronaos dar, denn sie sind ihrer Form nach nicht frühklassisch, sondern spätklassisch bzw. hellenistischrömisch, d. h. sie wurden bei einer Reparatur ausgetauscht15. Um sie auswechseln zu können, musste man das Gebälk über den Anten abtragen, wozu wahrscheinlich auch der Bereich dazwischen mit den figürlichen Metopen gehörte. Am Schluss stellt sich die Frage, wann diese Restaurierung erfolgte. Einen Hinweis bieten die Buchstabenformen der Versatzmarken, denn das Alpha und Gamma passen nicht mehr zum Neubau in der ersten Hälfte des 5. Jhs. v. Chr., weil ihre Gestalt mit frühklassischen Buchstaben nichts mehr gemeinsam hat. Allerdings lassen sie sich danach nicht genauer datieren16. 11

Mallwitz 1972, 234. Furtwängler 1880, 5; Treu 1897, 151; 153 Abb. 175. 13 Rehak 1998, 202 f.; Younger – Rehak 2009, 62–73. 14 Dinsmoor 1941, 417. 15 Herrmann 1981, 316 f. 16 W. B. Dinsmoor meinte, die Versatzmarken seien vom Tempelneubau, obwohl er anmerkte, dass das Gamma jünger sein könnte: Dinsmoor 1941, 417 Anm. 60; s. dazu auch Jeffery 1990, Taf. 43 mit elischen Inschriften aus der ersten Hälfte des 5. Jhs. v. Chr. Younger – Rehak 2009, 94 schreiben, die Buchstabenformen des Alpha und Gamma seien hellenistisch. Diese Eingrenzung ist nicht ganz korrekt, 12

Versatzmarken auf Baugliedern des Zeustempels von Olympia

43

Die erste größere Reparatur am Zeustempel erfolgte im 4. Jh. v. Chr., vermutlich nach dem starken Erdbeben von 374 v. Chr. Dabei wurden auch die Tempelfronten in Mitleidenschaft gezogen. In den Fundamenten danach errichteter Bauten fand man überdies häufig stuckierte Fragmente von Bauteilen des Zeustempels, d. h. diese Werkstücke waren zuvor schon verbaut gewesen. Erste Umbaumaßnahmen im Tempelinnern hatte man aber wohl schon in der zweiten Hälfte des 5. Jhs. v. Chr. vorgenommen, als die berühmte Kultstatue des Zeus von Pheidias eingebaut wurde. Doch dabei werden kaum die Gebälke des Pronaos und Opisthodom betroffen gewesen sein. Eine weitere Erneuerung, ebenfalls nach einem Erdbeben, wurde in der ersten Hälfte des 2. Jhs. v. Chr. ausgeführt17. Mithin werden die figürlichen Metopen entweder bei der Reparatur im 4. Jh. oder bei der im 2. Jh. v. Chr. abgenommen und wiederangebracht worden sein. Am Zeustempel sind weitere Versatzmarken auf den LöwenkopfWasserspeiern der Sima überkommen. Dieser Dachrand der Langseiten wurde in der Antike mehrmals repariert, zum letzten Mal geschah dies wohl am Anfang des 4. Jhs. n. Chr.18. Folgende Fragmente mit Versatzmarken sind erhalten: Auf einem Simastück hatte man die Buchstaben und auf einem anderen

weil sie durchaus noch klassisch sein können. Die Form des Alpha mit gerader Querhaste spricht sogar gegen eine hellenistische Datierung, weil im Hellenismus (ab dem 2. Jh. v. Chr.) das Alpha mit gebrochener Querhaste üblich war. – Younger – Rehak 2009, 94 betonen in ihrem Artikel besonders die Ausmaße der Erneuerungen im 2. Jh. v. Chr. Dennoch gab es schon im 4. Jh. v. Chr. umfangreiche Reparaturen, die Giebelfiguren, Löwenkopf-Wasserspeier und Dachziegel umfassten (Younger – Rehak 2009, 55. 100; die erneuerten Dachziegel werden in ihrem Artikel überhaupt nicht erwähnt [s. unten]) 17 Curtius – Adler 1892, 18. 22; Treu 1897, 151; 153 Abb. 175; Kunze – Weber 1948, 492; Grunauer 1971, 131; Mallwitz 1972, 233 f.; Grunauer 1981, 276. 280; Buchert 2000, 117. 122–124. 149; Koenigs 2002, 140 f.; Hennemeyer 2006, 104– 108; Younger – Rehak 2009, 57. 93 f.; Hennemeyer 2012, 123–125; Hennemeyer 2013a, 20–22; Beitrag Palagia 77. Den Umbaumaßnahmen für den Einbau der Kultstatue werden auch Reparaturarbeiten zugeordnet, deren genauer Umfang und Ursache bisher jedoch nicht gänzlich geklärt sind (Hennemeyer 2013a, 19 f.; Beitrag Hennemeyer 25–31). 18 Curtius – Adler 1892, 8 f. 26; Dinsmoor 1941, 402 f.; Willemsen 1959, 60–62. 124; Mallwitz 1972, 219. 233; Buchert 2000, 143–147. 150; Younger – Rehak 2009, 55; Hennemeyer 2013a, 20–23. Curtius – Adler 1892 gehen von mindestens vier Restaurierungen aus, die die Sima betrafen. F. Willemsen unterscheidet sogar neun verschiedene Phasen.

44

Chapter Two

eingemeißelt (Fig. 4). Beide Bauteile bestehen aus parischem Marmor. Von dieser Marmorsorte machte man nur beim Neubau und bei den Reparaturen im 4. Jh. v. Chr. Gebrauch, anschließend verwendete man pentelischen Marmor, der z. B. auch für die erneuerten Giebelfiguren benutzt wurde19. Zu dieser Beobachtung passt die Form des Xi, denn es besitzt noch einen senkrechten Strich, der die drei waagerechten miteinander verbindet und ab dem 3. Jh. v. Chr. nach und nach wegfiel20. Interessant ist weiterhin, dass die Löwenkopfwasserspeier jeweils zwei Versatzmarken aufweisen. Eine davon diente sicher der Nummerierung der Bauteile. Die zweite könnte eine Seitenbezeichnung darstellen, denn es gab ja zwei Langseiten, die unterschieden werden mussten. Also könnte beispielsweise das Ny als Abkürzung für NOTO (Süden) gedient haben. Jedoch gibt es so keine sinnvolle Auflösung für eine der Versatzmarken auf dem anderen Werkstück, d. h. für oder . Folglich wäre es auch denkbar, dass eine der beiden Versatzmarken verschiedene Alphabetserien kennzeichnete; denn das griechische Alphabet hatte nur 24 Buchstaben und es gab deutlich mehr Löwenkopf-Wasserspeier zu nummerieren (wohl 98 Stück). Einen anderen Vorschlag machte F. Willemsen, der meinte, eine der Versatzmarken stamme jeweils aus einer jüngeren Reparaturphase21. Da aber nur auf sehr wenigen Wasserspeiern Versatzmarken erhalten sind, ist im Moment keine Klärung dieser Frage möglich. Auf zwei Simablöcken aus pentelischem Marmor, die der hadrianischen Epoche zugeordnet werden, sind einzelne Versatzmarken eingehauen: Ein Block trägt ein Epsilon (Fig. 4) und ein anderer ein Phi. Hier sind allerdings ebenfalls keine weiteren Aussagen möglich, außer dass es kaum Zweifel daran gibt, dass sie von Restaurierungen stammen22. Die marmornen Dachziegel des Zeustempels stellen das einzige Bauglied dar, auf dem sehr viele Versatzmarken überkommen sind. Im Inschriftenband von Olympia werden neunzig Versatzmarken auf parischen Marmorziegeln und achtundachtzig auf pentelischem Marmorziegeln genannt. Die Buchstaben der Versatzmarken sind nicht sehr groß und befinden sich immer auf den Oberseiten der Kalyptere oder 19

Willemsen 1959, 1; Taf. 123; Buchert 2000, 132; Herrmann 2000, 386; Trianti 2002, 292; Younger – Rehak 2009, 91. 100. An anderen Bauten Olympias wurde parischer Marmor bis weit in die hellenistische Zeit hinein verwendet (s. Beitrag Palagia 78). 20 s. z. B. Cook 2004, 14. 21 Willemsen 1959, 1. 22 IvO Nr. 682, 66 a; Curtius – Adler 1892, 26; Willemsen 1959, 1. 93 f. 121; Taf. 123.

Versatzmarken auf Baugliedern des Zeustempels von Olympia

45

Strotere23. Einige Marmorziegel hatte man zusätzlich mit größeren Buchstaben oder Buchstabenkombinationen versehen, und zwar auf den Ober- oder Unterseiten. Da diese großen Buchstaben sich meist auf mehreren Dachteilen wiederholen, wird es sich bei ihnen nicht um Versatzmarken, sondern um Namensabkürzungen der Steinmetze handeln, die sie angefertigt haben (Fig. 6 Nr. 18. 75–88)24. Bevor der Dachziegelbefund des Zeustempels genauer betrachtet werden soll, ist eines vorauszuschicken: Beim Neubau waren bei einem Dach aus genormten, d. h. gleichgroßen Dachziegeln keine Versatzmarken nötig, weil sich die Teile nicht voneinander unterschieden (Fig. 7). Die Werkstücke für die Dächer gelangten in der Regel als Rohlinge, d. h. in Bosse, auf die Baustelle. Dort bereitete man sie für den Versatz vor und glättete sie entsprechend25. Erst während dem Vorgang des Dachdeckens wurde jeweils ein Ziegel an seinen Nachbarn angepasst, wenn es nötig war. Falls man die Dachziegel später, z. B. bei einer Reparatur, wieder abnehmen wollte, mussten sie vorher mit Versatzmarken versehen werden, um sie ohne aufwändiges Nachbearbeiten erneut verlegen zu können26. Somit kann man jetzt schon schlussfolgern, dass die Marmordachziegel des Zeustempels ihre Versatzmarken wohl sämtlich zum Zwecke ihrer zweiten Verlegung erhielten. Die Versatzmarken auf den Stroteren und Kalypteren des Zeustempels bestehen zum großen Teil aus mindestens zwei Buchstaben (Fig. 5. 6). Folglich wird einer von ihnen die Nummer der Ziegelreihe und der andere die Position innerhalb dieser angezeigt haben, denn um die Lage eines Dachziegels eindeutig zu bestimmen, waren zwei Informationen nötig.

23

IvO Nr. 681. 682. IvO Nr. 682, 18. 75–88. 25 Ohnesorg 1993, 47; Herrmann 2000, 384 spricht von „Halbfabrikaten“, die verschifft wurden. 26 Beim römischen Historiker Titus Livius ist dazu eine passende Geschichte überliefert (Liv. 42, 3, 28): Q. Fulvius Flaccus hatte sich verpflichtet, der Fortuna Equestris in Rom einen Tempel bauen zu lassen. Dazu ließ er 173 v. Chr. u. a. marmorne Dachziegel des Juno Lacinia-Tempels von Kroton nach Rom bringen. Nachdem er wegen dieses Frevels aber vom Senat verklagt worden war, schickte man die Bauteile zurück ins Heiligtum nach Kroton. Dort blieben sie jedoch im Vorhof des Tempels liegen, weil sie niemand wiederversetzen konnte. Zu guter Letzt vernebelte Juno dem Q. Fulvius Flaccus die Sinne, sodass er sich erhängte. Was Livius nicht erwähnt, ist, dass man die marmornen Dachziegel wohl deshalb nicht erneut auf dem Tempeldach in Kroton versetzen konnte, weil auf ihnen keine Versatzmarken angebracht worden waren, bevor man sie abgenommen hatte. 24

46

Chapter Two

Bisher ist allerdings nicht feststellbar, welcher der beiden Buchstaben die horizontale und welcher die vertikale „Koordinate“ beinhaltete. Bei den Formen der Buchstaben fällt auf, dass die meisten auf Ziegeln aus parischem Marmor älter erscheinen als die auf pentelischem Marmor. Folglich gehören die auf parischem Marmor wohl zu einer Gruppe, die bereits im 4. Jh. v. Chr. angebracht wurde (Fig. 5 Nr. 1–48)27. Außerdem bemerkte bereits A. Furtwängler, dass ihre Buchstabenformen nicht zu denen des frühklassischen Olympia passten. Er verwies u. a. darauf, dass das als Versatzmarke verwendete Xi im 5. Jh. v. Chr. in Elis noch nicht üblich war. Allerdings zog er daraus den falschen Schluss, es handele sich um ionische Buchstaben, und u. a. deshalb vermutete er, dass man die Versatzmarken bereits in den Steinbrüchen auf Paros angebracht habe28. Stattdessen datieren die Versatzmarken, die A. Furtwängler noch in das 5. Jh. v. Chr. einordnete, in das 4. Jh. v. Chr., als man in Elis das ionische Alphabet schon übernommen hatte29. Auffällig ist noch eine Eigenart der Buchstaben: Sie weisen kaum Rundungen auf. So sind das Theta und das Omikron quadratisch (Fig. 8). Die Buchstaben Beta und Rho, die normalerweise aus Halbkreisen bestehen, haben an ihrer anstelle Dreiecke. Der Grund für das Auftreten dieser „eckigen“ Buchstaben kann jedoch darin liegen, dass sich Buchstaben mit geraden Linien immer leichter einmeißeln lassen als solche mit „Rundungen“; einen Datierungshinweis kann man daraus aber nicht gewinnen. Insgesamt ergeben die Buchstabenformen der Versatzmarken auf parischem Marmorziegeln ein einheitliches Bild, welches ihre Datierung in die frühklassische Zeit ausschließt. Beispielhaft seien nur folgende Merkmale genannt: Beim My und Ny sind die Striche rechts ihrer linken Haste nicht mehr verkürzt. Darüber hinaus sind die drei Querstriche des Epsilon bereits waagerecht und nicht mehr schräg (Fig. 5 Nr. 1–48)30. Neben A. Furtwängler bezeichneten schon W. Dittenberger und K. Purgold im Inschriftenband von Olympia diese Buchstabenformen als „wenig altertümlich“. Dennoch meinten sie, die Versatzmarken stammen 27

Dabei handelt es sich um folgende Stücke: IvO Nr. 681, 1–48. Furtwängler 1893, 76 f. So verweist er u. a. darauf, dass das Xi im 5. Jh. v. Chr. in Elis noch nicht üblich war. Dies trifft zu, wie man bei Jeffery 1990, 206 sehen kann. Mallwitz 1980, 362 Anm. 9 schloss sich der Meinung Furtwänglers an. 29 Larfeld 1902, 452. 455. 30 Zur Form der in Elis gebräuchlichen archaischen und frühklassischen Buchstaben s. Jeffery 1990, 206. 220 f.; Taf. 43. Nicht nur von ihnen unterscheiden sich die Versatzmarken deutlich, sondern auch von den archaisch bzw. frühklassischen der ägäischen Inseln: Jeffery 1990, 289. 28

Versatzmarken auf Baugliedern des Zeustempels von Olympia

47

vom frühklassischen Neubau31. Das ist aber nicht der Fall, wie hier gezeigt werden konnte, denn wahrscheinlich stammen die Versatzmarken von der ersten größeren Reparatur des Tempels in der Mitte des 4. Jhs. v. Chr. Dabei wurden offensichtlich die alten parischen Marmorziegel zum großen Teil wiederverwendet32. Eine Besonderheit zeichnet die spätklassischen Versatzmarken der parischen Dachziegel noch aus: Bei ihnen sind die Buchstaben in der Regel an Stellen eingehauen, die vom nächsten Dachziegel überdeckt wurden (Fig. 7)33. Dies lässt erkennen, dass man die Versatzmarken erst anbrachte, nachdem man den „Nachbarn“ abgenommen hatte. Auf diese Weise waren die Buchstaben nach dem erneuten Verlegen der Dachteile nicht mehr sichtbar, d. h. der alte Zustand war vollkommen wiederhergestellt. Solch eine Vorgehensweise konnte bisher an keinem anderen Bau beobachtet werden. Sie passt aber gut zu den aufwändigen Reparaturen des 4. Jhs. v. Chr.; später in hellenistischer, wie auch in römischer Zeit schlug man die Versatzmarken mitten auf den Dachteilen ein34. Es gibt auch typisch „römische“ Versatzmarken auf den Dachziegeln des Zeustempels, die sich auf der zweiten Gruppe aus parischem Marmor und auf vielen aus pentelischem Marmor befinden (Fig. 5 Nr. 66–87; Fig. 6 Nr. 1–17. 19–74)35. Ein ganz erhaltener Stroter hat auf seiner Oberseite etwa in der Mitte die Buchstaben eingehauen (Fig. 6 Nr. 20)36. Er trägt somit zwei Nummern: eine Sieben ( ) und eine Neunzehn ( ), wobei letztere mithilfe alphabetischer Zahlen ausgedrückt ist. Diese besondere Art griechischer Zahlen wurde für Versatzmarken zum ersten Mal am Ende des 3. Jhs. v. Chr. benutzt, und zwar auf Bauteilen der

31

In diesem Zusammenhang ist außerdem zu lesen, dass A. Furtwängler wegen der nicht frühklassischen Gestalt der Versatzmarken zuerst annahm, das Dach des Zeustempels sei viel später fertig geworden. Diese Meinung revidierte er allerdings kurz darauf und gelangte zu dem wiederum nicht zutreffenden Schluss, die Versatzmarken hätte man in frühklassischer Zeit in parischen Steinbrüchen angebracht (Furtwängler 1880, 5 Anm. 1; Furtwängler 1893, 75–77; IvO 699 f.). 32 Fraglich ist, ob schon in der zweiten Hälfte des 5. Jhs. v. Chr., als die Cella für das Kultbild des Phidias umgebaut wurde, bereits Teile des Daches demontiert wurden (s. dazu Hennemeyer 2006, 104–108; Hennemeyer 2012, 123 f.; Beitrag Hennemeyer 26). Jedenfalls scheinen keine Marmordachziegel mit Versatzmarken des 5. Jhs. vorhanden sein. 33 IvO Nr. 681, 1–48. 34 s. IvO Nr. 681, 66–87; Nr. 682. 35 IvO Nr. 681, 66–87; Nr. 682, 1–74 (ohne 18). 36 IvO Nr. 682, 20.

48

Chapter Two

Philippsstoa von Delos37, d. h. die Versatzmarken mit solchen Nummern können kaum älter sein. Dazu passt auch, dass keine Versatzmarke auf den älteren parischen Marmordachziegeln aus alphabetischen Zahlen gebildet ist. Des Weiteren weist das Theta auf dem Stroter aus pentelischem Marmor einen durchgehenden Querstrich auf, wie er erst häufiger ab der frühen Kaiserzeit gebraucht wurde38. Ferner ist zu bemerken, dass auch auf diesem in römischer Zeit wiederverwendeten Flachziegel eine Nummer zur Kennzeichnung der horizontalen Lage diente und die andere zur Angabe der vertikalen. Wie man feststellte, wurde pentelischer Marmor für Dachziegel zum ersten Mal bei einer hellenistischen Restaurierung verwendet, die vielleicht um 160 v. Chr. erfolgte39. Versatzmarken können diese Werkstücke frühestens bei der nächsten Reparatur erhalten haben, die vermutlich ans Ende des 1. Jhs. v. Chr. datiert (Fig. 6 Nr. 1–17. 19–74)40. Folglich sind die Versatzmarken auf Dachziegeln vom Pentelikon zwischen dem ausgehenden 1. Jh. v. Chr. und der letzten Erneuerung Anfang des 4. Jhs. n. Chr. einzuordnen41. Die „späten“ Versatzmarken auf parischen Marmordachziegeln könnten dagegen schon um 160 v. Chr. angebracht worden sein (Fig. 5 Nr. 66–87)42. Zu der großen Zahl an Marmordachziegeln mit Versatzmarken wären sicher weitere Aussagen möglich. Solche erforderten aber eine detaillierte Untersuchung aller erhaltenen Fragmente. Ihr Arbeitsumfang ist jedoch so groß, dass sie bisher nicht erfolgte. Am Ende lässt sich zu den Versatzmarken des Zeustempels Folgendes zusammenfassen: Auf den marmornen Dachziegeln haben sich – entgegen der bisherigen Forschungsmeinung – keine Versatzmarken erhalten, die vom Tempelneubau in frühklassischer Zeit stammen. Wenn doch beim Neubau schon Versatzmarken eingesetzt worden wären, müssten alle Dachziegel mit solchen markiert gewesen sein. Das ist aber nicht der

37

Vallois 1923, 63–66; Weber 2013, 72–75. Larfeld 1902, 484. 39 Willemsen 1959, 74; Beitrag Palagia 79–81. 40 Willemsen 1959, 88. 41 Willemsen 1959, 124. 42 Eine Untersuchung der einzelnen Buchstabenformen der „späten“ Versatzmarken erfolgte bisher nicht, weil klar ist, dass sie alle von Wiederverwendungen stammen. Als Beleg für ihre tatsächlich späte Entstehung kann man sie mit den Inschriften „Verzeichnisse des Kultpersonals“ vergleichen, die zwischen 36 v. Chr. und 265 n. Chr. entstanden (IvO Nr. 59–141). 38

Versatzmarken auf Baugliedern des Zeustempels von Olympia

49

Fall43. Auch die wenigen Versatzmarken der Löwenkopf-Wasserspeier zeigten, dass sie erst bei Reparaturen eingesetzt wurden. Die Versatzmarken auf den beiden Herakles-Metopen erlaubten es schon kurz nach der Ausgrabung, ihre Position und die der Kerberosmetope über dem Pronaos sicher feststellen zu können. Sowohl diese Versatzmarken als auch alle anderen genannten stammen folglich nicht vom Tempelneubau, sondern frühestens von seiner ersten umfangreichen Reparatur im 4. Jh. v. Chr.

43

Komplett erhaltene Kalyptere und Strotere aus parischem und pentelischem Marmor, die keine Versatzmarken besitzen, sind vorhanden. Sie befinden sich im Magazin von Olympia z. B. in den Kisten Nr. 921. 923. 924. 928. 929.

50

Chapter Two

Fig. 1. Metopen-Triglyphenfries über dem Pronaos. After Curtius – Adler 1892, 138 Fig. 170.

Fig. 2. Linke obere Ecke der Augiasmetope und Oberteil der Atlasmetope mit den Versatzmarken. After Curtius – Adler 1892, 174 Fig. 205; 177 Fig. 207.

Versatzmarken auf Baugliedern des Zeustempels von Olympia

Fig. 3. Atlasmetope in Aufriss und Schnitt, Ausklinkungen zum Heben markiert. After Curtius – Adler 1892, 151 Fig. 173.

Fig. 4. Versatzmarken auf drei Löwenkopfsimen, und auf der Oberseite sowie E auf der Rückseite. After Willemsen 1959, Pl. 123.

51

52

Chapter Two

Fig. 5. Versatzmarken auf Dachziegeln aus parischem Marmor. After Dittenberger – Purgold 1896, 695.

Versatzmarken auf Baugliedern des Zeustempels von Olympia

53

Fig. 6. Versatzmarken auf Dachziegeln aus pentelischem Marmor (die Nr. 18 und 75–88 sind Unternehmermarken) After Dittenberger – Purgold 1896, 706–708.

54

Chapter Two

Fig. 7. Rekonstruktionszeichnung einer Giebelecke mit Marmordach After Curtius – Adler 1892, Pl. 16.

Versatzmarken auf Baugliedern des Zeustempels von Olympia

Fig. 8. Kalypterfragment aus parischem Marmor mit Versatzmarken auf der Oberseite

55

CHAPTER THREE METROLOGISCHES IN OLYMPIA WOLFGANG SONNTAGBAUER

Abstract At the analysis of a metrological system in Olympia we found following results: Only the Pheidonic foot could be found in Olympia. For the often mentioned Olympic foot at the race-course and the temple of Zeus there is no basis. Therefore the race-course is 588 Pheidonic feet long and not 600 Olympic feet. The numbers 49 and 50 are important in the metrological system. The length of the stylobate of the temple of Zeus makes one third of the length of the race-course. As a consequence of these facts the temple of Zeus measures 84 x 196 Pheidonic feet. The old measure of 16 Pheidonic feet of the column-spacing is confirmed. There is also a metrological connection between the old temple of Zeus (= Heraion) and the racing-course, because the length of the stylobate is 49 x 50 Pheidonic dactyls. Also interesting metrological connections between Olympia, the temple of Poseidon at Isthmia and the pyramid of Cheops can be shown.

Einleitung Die Frage stellt sich nach wie vor, welche Bedeutung Maße in der griechischen Architektur, im Besonderen am griechischen Tempel haben, wie die Griechen die Zahlen und Proportionen ihrer Maße gesehen haben, ihrer Maße in ganzen Zahlen und Zahlenbrüchen1 und nicht unserer Maße in Dezimalzahlen. Wenn wir vorgreifend wie schon so oft behaupten, daß sich in Zahl und Proportion das über allem stehende Beziehungsdenken der Griechen zuvorderst zeigt, daß sie die eigentlichen Wegmarken zu einem tieferen Verständnis des griechischen Tempels sind, so ist die Metrologie griechischer Tempel nicht nur 'Beschreibung' der Außenhaut, sondern soll auch als 'Gang in die Tiefe' die kultisch-religiösen Hintergründe

1

Auch Zahlenbrüche (z. B. 1/2; 2/3) haben die Griechen als Proportionen gesehen.

Metrologisches in Olympia

57

dieser Bauten aufzeigen2. L. Weniger bringt es auf den Punkt, wenn er betont, "daß das Verständnis der vielseitigen Einrichtungen des olympischen Heiligtums nur dann sich erschließt, wenn man sich vergegenwärtigt, daß sie letzten Grundes auf dem Gottesdienste beruhen3." Auch Tempel und Laufbahn sind 'Einrichtungen' und wir müssen uns auf die Suche machen, warum und auf welche Weise sie 'eingerichtet' wurden. Vielleicht kann Metrologisches aus Olympia abermals einen kleinen Beitrag leisten, in eine eingefrorene Debatte wieder etwas Bewegung zu bringen. Die vorgestellten Ergebnisse sind zum größten Teil an verschiedenen Stellen publiziert4 und werden hier noch einmal zusammengefaßt, ergänzt und erweitert.

Kurzer Abriß der Fußmaß- und Proportionsdiskussion Um die eigenen Positionen mit früheren vergleichen und damit klarer machen zu können, sei die bisherige Fußmaß- und Proportionsdiskussion kurz gestreift. Seit Dörpfeld 1890, Riemann 1935, von Gerkan 1940, Dinsmoor 19615 ging man von einem kanonischen Fußmaßsystem aus, in dem 3 Fußmaße in festen Proportionen zueinander stehen. 1940 hat v. Gerkan anhand des Jüngeren Didymaions folgendes Schema kanonischer Fußmaße vorgestellt: Attischer Fuß (AF) : dorischer bzw. pheidonischer Fuß (PF) : ionischer Fuß (IF) = 27 : 30 : 32 AF:PF = 9:10 und PF:IF = 15:16

Gruben, Büsing, Berger, Mertens, Knell, Bankel, Wesenberg u. a.6 sind ihm im Wesentlichen darin gefolgt. Freilich war immer wieder strittig, in welchem Fußmaß die einzelnen Tempel entworfen wurden. Angeregt durch den Parthenonkongreß 1982 in Basel gab es in den 80-iger Jahren heftige Kontroversen um das Fußmaß attischer Bauten, im Besonderen des Parthenon7.

2

Vgl. Dinsmoor 1938. Weniger 1920/21, 41. 4 Sonntagbauer 1998; 2003a; 2003b; 2004; 2009; 2013. 5 Dörpfeld 1890; Riemann 1935; von Gerkan 1940; Dinsmoor 1961. 6 Gruben 1966 (2001); Büsing 1982; Berger 1984; Mertens 1984b; Knell 1988; Bankel 1991; Wesenberg 1995. 7 Aus einer großen Zahl von Beiträgen seien Wesenberg 1984 und Bankel 1991 herausgegriffen. 3

58

Chapter Three

Diese Diskussion ist bis heute nicht abgeschlossen, obwohl Dörpfeld schon 1890 den PF am Erechtheion eindeutig nachgewiesen hatte8. Noch 1995 zweifelt Wesenberg daran, ob "der Forschung weitere hundert Jahre zugestanden werden, sich darauf zu einigen, in welchem Fußmaß der Parthenon erbaut wurde." Seit den 90-iger Jahren hat die Diskussion um Fußmaße eine andere Färbung: Nicht mehr die Frage nach AF, PF und IF, also nach kanonischen Fußmaßen, steht im Vordergrund, sondern die Frage, welcher Modulus (der Begriff stammt von Vitruv und heißt eigentlich 'kleines Maß'), der bei jedem Bau neu festgesetzt wird, zu veranschlagen ist. Die 'Kanoniker' geraten zunehmend in die Defensive und die jüngste Forschung neigt zu einem System variabler Moduli, sodaß beinahe jede neue Entwurfsanalyse eines Bauwerkes ein anderes Fußmaß besitzt9. Hier spielt Vitruv herein, zwischen dem und der Spätklassik der gesamte Hellenismus liegt. Vitruv scheint als Römer die frühe, im Besonderen von der Musiktheorie bestimmte Architekturtheorie nicht mehr verstanden zu haben und liefert offensichtlich Versatzstücke vormaliger Architekturtheorien. Die Ablehnung kanonischer Fußmaße ist umso erstaunlicher, als gerade in den letzten Jahren sowohl der PF abermals als auch der AF eindeutig nachgewiesen wurden. 2000 hat Wilson Jones an einem metrologischen Relief aus Salamis den PF eindeutig belegt und 2007 hat Ulf Weber anhand einer Buchstabenfolge von 22 Buchstaben auf der Euthynterie des Jüngeren Didymaions den AF festgestellt10. Eng verbunden mit der Fußmaßfrage ist die Frage nach Proportionen vor allem am klassischen Tempel bzw. ob es zusammenhängende Proportionssysteme gibt. Seit Riemann hat man immer wieder Proportionen festgestellt, vor allem auch in den Handbüchern. Daneben spricht vor allem Mertens immer wieder von "festen und klaren Proportionen", "daß die Proportionen auch am Bau genau und nicht nur annähernd ausgeführt sind" und "daß sie untereinander im Zusammenhang stehen, ein homogenes Gerüst bilden"11. Letztlich sind alle diese Vorstellungen Stückwerk geblieben. 1998 haben wir versucht, unsere Vorstellungen eines zusammenhängenden Proportionssystems am Grundriß des Parthenon zu nachzuzeichnen, 2003 in Ansätzen auch am Zeustempel12. Seither ist die Diskussion darüber mehr oder weniger zum Stillstand gekommen. In der 8

Dörpfeld 1890; vgl. Wesenberg 1995, 205-209. U. a. Höcker 1993; Hennemeyer 2006; kritisch dazu Wesenberg 1995, 216f. 10 Wilson Jones 2000; Weber 2011, 36f. 11 Mertens 1984a, 137f. 12 Sonntagbauer 1998; ders. 2003b. 9

Metrologisches in Olympia

59

gegenwärtigen Forschung ist nicht nur die Frage nach zusammenhängenden Proportionsgefügen kaum ein Thema, sondern auch das Fragen nach proportionalen Zusammenhängen ist weitgehend außer Mode gekommen. Proportionen werden in erster Linie als "praktische Entscheidungshilfen"13 und konkrete Arbeitshilfen für die Architekten gesehen. Fallweise ist zwar von "Streckenverbund", "konsistenter Rhythmisierung der Einzelteile" und von einem "die Einzelheiten festschreibenden Proportionsgerüst" die Rede14, was genau darunter zu verstehen ist, bleibt im Unklaren, vor allen deswegen, weil dies an Beispielen kaum nachvollzogen werden kann, sodaß die Sache mehr vernebelt als geklärt wird15. Wie wir meinen, ist in diesem Durcheinander einiges zu klären; denn das über ein Jahrhundert diskutierte - man war sich der Bedeutung dieser Sache offensichtlich bewußt - und in vielen Teilen noch ungelöste Problem als nicht mehr aktuell und zeitgemäß zu bezeichnen und ad acta zu legen, wie es nicht selten getan wird, würde dem Verständnis des griechischen Tempels keinen Schritt weiterhelfen.

Metrologische Voraussetzungen für das Verständnis des klassischen Tempels und damit des Zeustempels Zum Nachvollzug des hochklassischen Tempelentwurfs, als dessen Höhepunkt der Zeustempel allgemein gilt, halten wir folgende Annahmen bzw. Voraussetzungen für notwendig: 1. Es gibt ein kanonisches System von Fußmaßen, wobei der dorisch-pheidonische Fuß (PF) am dorischen Tempel offensichtlich die 13

Korres 1994, 62f. Schneider-Höcker 2001, 125. 122. 15 Wohin Entwurfsanalysen unter Weglassen fester, kanonischer Fußmaße führen, zeigt eine Entwurfsanalyse des Parthenon (Schneider-Höcker 2001, 122-125): Auf der Grundlage eines "Parthenonmoduls" von "14/16 attischen Fuß", das als gutes Beispiel für die "Modultheorie" angeführt wird, wird ein "Streckenverbund" für die STB hergestellt, der unvollständig ist, da nicht alle Bauteile miteinbezogen sind, der ein künstlich konstruiertes, den Sachverhalt vernebelndes Modul von 28,627cm anwendet und bei dem keine Proportionszusammenhänge dargestellt sind. Zudem hat man nicht wahrgenommen, daß die Ergebnisse mit unseren 1998 dargestellten (Sonntagbauer ÖJh 1998, 133-169; ders. 2003b, 35-42) weitgehend übereinstimmen, dies notwendigerweise, weil das Modul an den PF gebunden ist. Warum nimmt man nicht gleich den PF und führt damit den "Streckenverbund", der nichts anderes ist als ein Proportionszusammenhang, in allen Teilen durch? 14

60

Chapter Three

Hauptrolle spielt16. Rückschlüsse von der Vitruv'schen Architekturtheorie auf den griechischen Tempel des 5. Jh.s. scheinen mehr zu verstellen als sie zur Klärung beitragen17. Die vor allem auf dem vitruv'schen Modulus gründenden Entwurfstheorien spiegeln mehr hellenistisches als klassisches Denken. Mehrere Fußmaße an einem Tempel tauchen erst im 4. Jh. auf18. 2. Es ist von einer bis ins Kleinste durchgehenden Proportionsstruktur auszugehen, die alle Teile des Tempels in einem 'Beziehungskosmos' zu einer Einheit zusammenbindet. Diese halten wir für grundlegend und wesensbestimmend, gleichsam für die 'Partitur' dieser Architektur. Die Proportionsstruktur wird nicht von Dezimalzahlen bestimmt - diese gab es in der Antike nicht - sondern von ganzen Zahlen und Zahlenbrüchen. Die Zerlegbarkeit der Zahlen, d. h. ihre Proportionsfähigkeit ist dabei von entscheidender Bedeutung. 3. Die Riemann'sche Regel, nach der die an archaischen Tempeln übliche einfache Proportionierung des Stylobats aufgegeben und in die Achsmaße verlegt wird19, gilt nicht für den hochklassischen Tempel bzw. es gibt diese nicht, denn die ganzzahlige Stylobatproportion ist zwar nur am Parthenon und mit Abstrichen in Isthmia exakt ausgeführt, ist aber intendiert und wird deshalb mehr oder weniger exakt umspielt20. Die ganzzahlige Stylobatproportion ist schon im 6. Jh. so wichtig, daß es undenkbar ist, daß sie im 5. Jh. beiseitegelassen wird. Die Formel n : 2n + 1 für die Stylobatproportion (z. B. 3 : 7 am Zeustempel) weist auf Musikalisches.

16 Die immer wieder eingeforderte Übereinstimmung von Messwerten und Maßinterpretationen im Rahmen einer tolerierbaren Schwankungsbreite setzen wir voraus (vgl. u.a. Mertens 1984a, 137). 17 Vgl. Knell 1985; zuletzt Sonntagbauer 2013, 155-158. 18 Wesenberg 1983, 32-68; ders. 1995, 202f.; ablehnend Büsing 1986; Im Rahmen der Untersuchungen der Buchstaben auf der Euthynterie des Jüngeren Didymaions hat U. Weber herausgefunden, daß neben dem attischen Fuß (AF) für die Dimensionierung der Peristasis, des Stufenbaues und des Aufrisses noch weitere Fußmaße zu veranschlagen sind (Weber 2011, 35-37). In meinem noch unpublizierten Aufsatz "Metrologische Analyse des Jüngeren Apollontempels von Didyma" habe ich den Begriff Modulusfuß verwendet, besser wäre erweiterter Fuß, da die Fußmaße im Vergleich zum AF um 1/3 bzw. 1/4 Daktylos (1AF + 1/4 Daktylos = 29,42cm x 65/64 = 29,87cm = sog. Haselberger'scher Fuß, der im Aufriß vorkommt) erweitert, sodaß sich auch diese Fußmaße in untergeordneter Weise in das kanonische System einfügen lassen. 19 Riemann 1935, 8-9. 20 Sonntagbauer 2009, bes. 45.

Metrologisches in Olympia

61

4. Der Kern des Entwurfes besteht in der Antinomie zwischen Stylobat und Peristasis21. Der Entwurf geht also vom Grundriß aus. Die Schwierigkeit besteht darin, auf dem proportionierten Stylobat die Säulen so anzuordnen, daß sie mit Ausnahme der Eckkontraktion alle den gleichen Abstand haben, der jedoch nicht beliebig, sondern ebenso in Proportionen festgelegt ist. Gleicher Abstand der Säulen von der Stylobatkante ist ebenso kanonisch. Es entsteht ein Konflikt, der zahlhaft umspielt - allerdings sind zusätzlich meist Anpassungen wie beim Zeustempel notwendig - und erst beim Parthenon in höchster Vollkommenheit gelöst wird. Mit dem Ende des 'Umspielens' geht auch die Blüte des dorischen Tempels zu Ende22. 5. Die Kenntnis der antiken Musiktheorie23, die der die antike Architektur tragenden Proportionstheorie völlig entspricht24, halten wir für das Verständnis griechischer Tempelarchitektur für unbedingt notwendig25. Musiktheorie und Proportionstheorie sind Teil des Arithmetik, Geometrie, Musik und Astronomie umfassenden Quadriviums, das das gesamte Kunstschaffen der Zeit prägt und Kern der allumfassenden Musiké ist26. 6. Der griechische Tempel ist nicht 'musikalisch' im engen Sinn unserer Musik, sondern im Sinn eines von Zahlen und Proportionen geprägten allumfassenden Beziehungsdenkens. Die Musiktheorie = Proportionstheorie zeigt uns, wie zahlhaft konkret griechisches Denken und damit griechische Tempelarchitektur sind. Harmonia ist in erster Linie das Verhältnis 1 : 2, die Oktav und nicht eine diffuse Vorstellung von

21

Ebd. 39. Vgl. Gruben 2001, 336f. 23 Die griechische Musik (ein Überblick bei Neubecker 1977) ist geprägt durch die Proportionen zwischen den Zahlen der Zahlenreihe (1 : 2 : 3 : 4…8 : 9 usw.), die die symphonen Intervalle (Oktav, Quint, Quart, Ganzton usw.) ergeben. Es sind dies epimore (überteilige) Verhältnisse, deren besonderes Kennzeichen ein Darunter und Darüber, also ein Pendelrhythmus um eine irrationale Mitte ist. Ebenso wie die Pendelstruktur der pythagoreischen Musik am Ende des 5. und im 4. Jh. erstarrt (Teilung der Oktav nicht mehr in 5 Ganztöne und 2 ungleiche Halbtöne, sondern in schematischer Weise in 6 Ganztöne bzw. 12 gleiche Halbtöne [Philolaos VS 44 B 6; Kaiser 2010 ), so erstarrt auch der dorische Tempel in dieser Zeit im Schematischen. 24 Im Kanon des Argivers Polyklet - die Argiver bezeichnet Herodot als Meister der Tonkunst (3, 131, 3) - verbinden sich Musiktheorie und Proportionstheorie zu einer wunderbaren Einheit. 25 Um den 'musikalischen Vergleich' noch weiter auszureizen: Ein Geiger muß wohl Noten lesen können. 26 Vgl. Sonntagbauer 2013. 22

62

Chapter Three

'Fügung'27. Das übliche Stylobatverhältnis n : 2n +1, beim Zeustempel 3 : 7, ist ebenfalls ein epimores, d. h. muikalisches Intervall, denn es ist als 6/2 : 7 zu verstehen. 7. Abschließend soll noch einmal betont werden, daß der griechische Tempel ein sakrales Bauwerk ist und keine Aufbewahrungshalle für Kultstatuen und sakrales Gerät. Sakralität28 ist wie alles bei den frühen Griechen konkret. Diese Konkretheit ist nie materiell, sondern immer von tiefer Geistigkeit 'beseelt', die im Kultisch-Religiösen fußt. Diese äußert sich am Tempel in einer 'Musik' von Zahlen und Proportionen, sodaß die Teile in ihrem 'Zusammenklang' zum 'Singen' kommen29. Dies hat mit schwärmerischem, unbestimmtem 'Harmonie-Gerede' nichts zu tun, im Gegenteil: Es ist strengste, sichtbar-unsichtbare Ordnung30.

Fußmaß am Zeustempel Seit jeher wurde am Zeustempel von Dörpfeld, Dinsmoor, Gruben, Grunauer und Knell u. a.31 ein PF von 32,6-8cm angenommen. Riemann geht von einem Modulus von 521cm aus32, dessen 1/16 einem PF entspricht. Dies wird auch in den Handbüchern übernommen. Noch ein zweites Fußmaß wird in Olympia und am Zeustempel angeführt, der sog. Olympische Fuß (OF), der sich aus der Teilung der Laufbahn von 192,28m ergibt (192,28m : 600 = 32,04cm33). In einem 2002 aus Anlaß des Jubiläums 125 Jahre Deutsche Ausgrabungen in Olympia herausgegebenen Sammelband geht Koenigs zwar kurz auf Metrologisches ein, u. a. auf die Fußmaße am Zeustempel, doch meint er zusammenfassend, daß der Zeustempel trotz seiner Bedeutung in der Geschichte des dorischen Tempels nach wie vor ein "schwarzes Loch" in

27

In Zusammenhang mit dem klassischen dorischen Tempel kommen immer wieder die Pythagoreer (z. B. Gruben 2001, 336f.) ins Spiel. Ohne genaue Kenntnis der griechischen, d. h. in weiten Teilen pythagoreischen Musiktheorie bleiben solche Andeutungen im Unbestimmten. 28 Auf die Position Wenigers haben wir oben hingewiesen. 29 Vgl. Sonntagbauer 2003a. 30 Wie könnte es sonst einen Kanon des Polyklet geben. Hätten wir nicht die Quellen, würden wir nicht im Entferntesten auf den Gedanken kommen, daß auch Plastik so gestaltet wurde. 31 Dörpfeld 1890, 171; Dinsmoor 1950, 151-153; Gruben 2001, 57; Grunauer 1981, 273; Knell 1988, 43. 32 Riemann 1935 Tab. 2. 33 Der OF steht zum PF im Verhältnis 49 : 50, d. h. vom PF wird 1/50 abgezogen.

Metrologisches in Olympia

63

der Forschung sei34. Zur Frage, ob OF oder PF am Zeustempel, gibt er eine "salomonische" Antwort: Für die STL veranschlagt er wie schon Dörpfeld, Grunauer und Romano35 ein "sakrales" Maß von 200OF; nach den neuen Vermessungen sei auch die Länge des Toichobats 100OF (100OF = 3204cm [Messwert: 3205,7 – 3208,4cm])36. Trotz Ungereimtheiten und Modifikationen in der Säulenstellung, möglicher Planänderungen und Reparaturen, mußte es von Anfang an einen einheitlichen Plan gegeben haben, worauf auch die Präzision des Fundamentes und der Krepis hinweist37. Hennemeyer hat in seinem Vortrag am Symposion den PF vollkommen fallengelassen und den unkanonischen olympischen Fuß für den gesamten Tempel vorgeschlagen, sodaß auch das Normaljoch mit 16 1/4OF veranschlagt wird.

Wie die Teile zueinander stehen: Proportionen am Zeustempel Die Metrologie des Zeustempels wurde seit jeher mit der olympischen Laufbahn verknüpft. Wie schon erwähnt haben Dörpfeld, später Romano und Grunauer festgestellt, daß die Stylobatlänge (STL) des Zeustempels 1/3 der Stadionlänge beträgt (192,28m : 3 = 64,09m ~ 64,12m), woraus sich bei Annahme einer Stadionlänge von 600 Fuß der OF von 32,04cm ergibt. Bei Annahme des PF auch für die Laufbahn ergeben sich 588PF. Die auf den ersten Blick unhandliche Zahl und die Vorstellung, daß ein Stadionlauf immer 600 Fuß betragen muß, haben bisher nicht einmal eine Diskussion über diese Sachverhalte geschweige denn Akzeptanz aufkommen lassen. Ein Anzahl von Indizien für diese These sei abermals vorgelegt und ergänzt:

34

Koenigs 2002, 143; Zwar wurde der Zeustempel seit 1997 von Koenigs und Hennemeyer neu bearbeitet und vermessen, doch sind die Ergebnisse zum größten Teil noch nicht publiziert. Nach Auskunft von Hennemeyer ist in der Monographie auch ein Kapitel über Metrologie, Planungsphasen und Entwurf vorgesehen. Teilergebnisse wurden auf der Tagung in Budapest vorgestellt. 35 Dörpfeld 1966, 19; Romano 1981, 132f. 36 Koenigs 2002, 143. 37 Vgl. ebd. 143; Den Toichobat interpretieren wir anders als Koenigs mit 98PF = 1/6 der Laufbahn. Die Beziehung zwischen 98 und 100, bzw. 49 und 50, worauf wir noch zurückkommen werden, taucht in Mythos und Historie immer wieder auf, so bei Herodot 6, 27, 2-3.

64

Chapter Three

1. Die Übereinstimmung ist so genau (588 x 32,68cm = 192,15m ~192,28m)38, daß ein Zufall nahezu ausgeschlossen ist. 2. Die Wichtigkeit des NJ für den Entwurf des Zeustempels, die immer wieder betont wurde, legt die Ganzzahligkeit desselben nahe39. Es entspricht nicht dem Denken der frühen Klassik, das NJ mit 16 1/4OF statt mit 16PF zu veranschlagen. 3. Das vermutete Fußmaß der Laufbahn wurde bisher immer, chronologisch richtig, auf die STL des Zeustempels übertragen. Warum wird nicht die STL ebenso wie das Joch im PF interpretiert (6412cm : 32,68cm ~196PF; 196 x 32,68cm = 6405,28cm) und diese auf die Laufbahn übertragen? 4. Die Zahl 588 ist eine besondere Zahl. Neben ihrer hervorragenden Teilbarkeit und daraus folgenden Proportionsfähigkeit (588 = 2.2.3.7.7 = 12 x 49) bleibt sie 12PF hinter den scheinbar notwendigen 600 Fuß des üblichen Stadionmaßes zurück, d. h. 588 und 600 stehen im Verhältnis 49 : 50. Es sind dies besondere Zahlen, wie wir sehen werden. 5. Wie in der Forschung schon immer mit Erstaunen bemerkt worden ist40, haben fast alle gemessenen Stadien unterschiedliche Längen: Athen (184,96m), Epidauros (181,30m), Delphoi (177,96m) usw.41. Wenn es nicht für jedes Stadion ein eigenes Fußmaß geben soll, bleibt nur die Annahme von Stadien, die nicht 600 Fuß lang sind. 6. Nach Pausanias 5, 16, 3 gab es in Olympia beim alten Fest der Heräen einen Lauf der Mädchen, bei dem die Stadionlänge um 1/6 verkürzt war, welche folglich, ausgehend von 588PF, 490PF betrug (588PF x 5/6 = 490PF [2.5.7.7]; 192,28m x 5/6 = 160,23m). In üblicher Berechnung ergäbe das den unkanonischen Fuß von 26,70cm (= 160,23cm : 600), sodaß auch hier anderes im Spiel zu sein scheint. 7. Der im Raum stehende kultische Hintergrund für die Stadionlänge hat sein gewichtigstes Argument in der sog. Oktaeteris, einem Schaltzyklus, der Sonne und Mond in Einklang bringt, da der Mondmonat 38

Nimmt man für den PF einen Wert von 32,7cm, der sich am Zeustempel ebenfalls nahelegt, so ist die Übereinstimmung bis auf den cm genau (32,7cm x 588= 192,276m). 39 Nach unserer Erfahrung kommt es höchstens zur Halbierung einer ganzen Zahl wie am älteren Poseidontempel in Sunion (Sonntagbauer 2009, 49 Taf. 1). 40 Dinsmoor (1950, 251) bezweifelt, daß hier unterschiedliche Fußmaße für Stadien von 600 Fuß zu veranschlagen sind: "It is hardly reasonable to assume, that these stadia were erected with reference to so many local varieties of foot units; the true reason still escapes us." Den wahren Grund glauben wir gefunden zu haben. 41 Sonntagbauer 2004, 76. 84f. Anm. 50.

Metrologisches in Olympia

65

hinter dem zwölften Teil des Sonnenjahres zurückbleibt und deshalb in einem Zyklus von 8 Jahren 3 Schaltmonate eingeschaltet werden müssen, sodaß 8 Sonnenjahren 99 Monate entsprechen42. Die Bedeutung der Oktaeteris für die Festzyklen und das religiöse auch die Politik umfassende Leben der Griechen ist in der archäologischen Forschung kaum, in der philologischen zu wenig im Bewußtsein, auch nicht, daß der 4-jährige Festzyklus, die Olympiade, eine halbierte Oktaeteris von 49 bzw. 50 Monaten ist. In zahllosen Mythen kommt die Oktaeteris mehr oder weniger offen vor und mit mythischen Erzählungen muß auch der olympische Stadionlauf, der den Kern des Festes ausmacht, vor allem auch dessen Gründung und Länge, wohl verbunden gewesen sein. All diese Mythen aufzuzählen und zu deuten würde ein eigenes Buch füllen, 2004 haben wir einiges gestreift43. Lediglich der Mythos von Atalante, der eine besondere Verbindung zwischen Olympia und Oktaeteris herstellt, soll zusätzlich erwähnt werden44. Der Zeitpunkt der Einführung der Oktaeteris ist umstritten, doch könnte er durchaus mit der Olympiadenzählung seit 776 v. Chr. zusammenhängen45. Durch den Hesperiden- und AtalanteMythos wird der astronomisch-kultische Hintergrund der Olympien bestätigt. Der Lauf der Gestirne prägt wie im Orient auch bei den Griechen die Religion und die 'Verhältnisse' am Himmel spiegeln sich in den Bauten des Heiligtums in Zahlen und Proportionen46. Die in den Handbüchern wiederholte und nur sehr lückenhafte Entwurfsanalyse, fußend auf einem PF von 32,62cm (522/16 cm), stammt 42 Zur Oktaeteris allgemein Nilsson 1937, in besonderer Beziehung zur Laufbahn Sonntagbauer 2004, 78f. 43 Ebd. 81-83. 44 Im Mythos von Atalante, der Parallelen zum Mythos von Pelops und Hippodameia besitzt, kann der Heros Hippomenes Atalante nur dann zur Frau bekommen, wenn er sie im Wettkampf besiegt. Auf den Unterlegenen wartet der Tod. Deshalb schenkt ihm Aphrodite drei goldene Äpfel aus dem Garten der Hesperiden, - ursprünglich war in Olympia der Apfel der Siegespreis - die Hippomenes beim Lauf vor Atalante hinwirft. Atalante hebt sie auf und schafft es dadurch nicht, Hippomenes zu überholen (Lugauer 1967, 90f.). Den direkten Bezug zu Olympia stellt eine Metope vom Zeustempel her, auf der Atlas dem das Himmelsgewölbe tragenden Herakles die drei Äpfel der Hesperiden überreicht, zwei mit der Rechten, einen mit der Linken. Beschützerin ist hier nicht Aphrodite, sondern Athene. In den drei Äpfeln sind ohne Zweifel die drei Schaltmonate zu erkennen, zwei in der einen, einer in der anderen Olympiade, die den 8-jährigen Sonnenzyklus mit dem 99-monatigen Mondzyklus (8 x 12 + 3) vereinen und zur Hochzeit führen (Hinweis H. Seiler, Zürich). 45 Sonntagbauer 2004, 86-90. 46 Vgl. Platon, leg. 822a-b.

66

Chapter Three

von Mallwitz aus dem Jahr 197247. Im Mittelpunkt steht das Einheitsjoch mit einfacher Eckkontraktion. "Im Aufbau zieht sich diese Einheit in gleichmäßiger Halbierung bis ins Dach hinauf: Joch = 16 Fuß; Triglyphe + Metope = 8 Fuß; Mutulus + Via = 4 Fuß; Ziegelbreite = 2 Fuß. Auch die Höhen scheinen angeglichen: Die Säulenhöhe (10,51m oder etwas größer) übersteigt um Geringes zwei Jochbreiten48." Die Stylobatmaße und die Säulendurchmesser mit leichten Unterschieden an Front und Langseiten werden nicht interpretiert. Nicht zuletzt hat auch die scheinbar nicht in einfachen Zahlen ausdrückbare Stylobatproportion zur Entstehung der Riemann'schen Regel beigetragen. In meinem 2003 erschienen Aufsatz "Einheitsjoch und Stylobatmaß. Zu den Grundrissen des Zeustempels in Olympia und des Parthenon"49 habe ich Grundschritte des Entwurfs des Zeustempels vorgestellt, wobei von folgenden Annahmen, die in der Forschung weitgehend gesichert sind, ausgegangen werden kann: 1. Die Peristasis besitzt ein Einheitsjoch. 2. Das NJ wird in der Forschung weitgehend übereinstimmend mit 16PF angegeben, wobei ein PF von 32,6-7cm (522cm : 16 = 32,62cm) angenommen wird. Folglich ist auch für die STL der PF anzunehmen, woraus sich eine STL von 196PF ergibt (196 x 32,68cm = 6405,28cm = ~6412cm oder 6412cm : 196 = 32,71cm). 3. Daraus läßt sich mit folgender Formel50 der Abst. E errechnen: 10NJ + 2Abst.E = 196PF 2 Abst.E = 196PF - 160PF (16PF x 10) Abst.E = 18PF

4. Daraus ergibt sich die STB: 3NJ +2Abst.E = 48PF + 36PF = 84PF 5. Folglich stehen STB und STL im Verhältnis 3 : 7 (84PF : 196PF).

47

Mallwitz 1972, 230-233. Gruben 2001, 57. 49 Sonntagbauer 2003b. 50 Diese Formel, bei der das Entscheidende das Zusammenziehen des EJ und des Achsabstandes der Ecksäule von der Stylobatkante (Abst.E) ist, kommt schon in der Grundrißanalyse des Parthenon (Sonntagbauer 1998) vor. 48

Metrologisches in Olympia

67

Es läßt sich also bei Annahme des PF für den gesamten Tempel in einer lückenlosen Beweiskette darstellen, daß das Stylobatverhältnis des Zeustempels mit 84PF x 196PF im Verhältnis 3 : 7 geplant war und somit der Riemann'schen Regel widerspricht. Im Aufriß finden sich an runden Maßen bei Grunauer und Gruben u. a. die Maßinterpretationen der Gebälkhöhe (12PF = 392,16 cm [Messwert 391cm]) und die Säulenhöhe der Cellafront (32PF = 1045,76cm [Messwert: 1044cm]), Triglyphe + Metope (8PF) und der Abacus (8PF)51. Die größte Lücke bildet die fehlende Interpretation der Säulen der Peristasis. Nicht zuletzt durch die unterschiedlichen Säulendurchmesser an Front und Seite, durch mögliche Planänderungen und Reparaturen lassen sich bis jetzt keine Maßinterpretationen finden, die zu einem zusammenhängenden Proportionssystem passen. Dies heißt nicht, daß es diese in der ursprünglichen Planung nicht gegeben hat.

Der Zeustempel als Teil übergreifender metrologischer Beziehungen (Fig. 1)? Wir wissen heute, daß das Heraion, wie schon Dörpfeld vermutet hat, als alter Zeustempel beiden Gottheiten geweiht war. Da der Nachfolgetempel des Heraions in derart enger metrologischer Verbindung zur Laufbahn bzw. den kultischen Gegebenheiten steht, so ist auch am Heraion eine Verbindung zu vermuten. Tatsächlich ergibt sich folgender Sachverhalt: STL = 49 x 50/16PF = 49 x 50 pheidonische Daktylen (49 x 50/16 x 32,68cm = 5004,12cm [Messwert: 5001cm] oder 5001cm : 49 x 50/16 = 32,65cm). Es ist kaum ein Zufall, daß hier bei hervorragender Übereinstimmung von Messwert und Maßinterpretation abermals die mit der Laufbahn so eng verbundenen 'Mondzahlen' 49 und 50 auftauchen52. Über das Verhältnis der STL des Zeustempels zur Laufbahn von 1 : 3 hinaus steht das Planungsmaß der STB von 84PF zur Laufbahn im Verhältnis 1 : 7. Auch in Athen ist eine Beziehung des Parthenon zur 51

Grunauer 1981, 273; Gruben 2001, 57. Auch hier lassen sich interessante metrologische Verbindungen zur Laufbahn herstellen, z. B. STB : Laufbahn Olympien = 25 : 16 x 16; STL : Laufbahn Olympien = 25/4 : 24; STL : Laufbahn Heräen = 5 : 16; STL : STL Zeustempel = 25 : 32. 52

68

Chapter Three

Stadionlänge auszumachen: u. a. STB : Stadion Athen = 3088cm : 184,96m = 189/2PF53 : 567PF (= 3.3.3.3.7; 184,94m : 567 = 32,61cm) = 1 : 6. In weiteren Vergleichen stellt sich heraus, daß die metrologischen Beziehungen weit über das olympische Heiligtum hinausreichen. Vergleicht man die Meßwerte bzw. Maßinterpretationen des Stylobats und der Joche bzw. des Abst.E des Poseidontempels in Isthmia mit denen des Zeustempels, so kann man feststellen, daß diese genau um 1/6 reduziert sind, d. h. auch, daß die STL von 490/3PF ein Drittel der Laufbahnlänge der Heräen ausmacht54: STB: 84PF x 5/6 = 70PF = 2287,60cm (Messwert: 2290cm55) STL: 196PF x 5/6 = 490/3PF = 5340,99cm (Messwert: 5350cm) NJ: 16PF x 5/6 = 40/3PF = 435,73cm (Messwert: 436cm) Abst.E: 18PF x 5/6 = 15PF = 490,00cm (Messwert: 491cm) Auch die STB des Parthenon steht zur STB des Zeustempels im Verhältnis 9 : 8 (189/2PF : 84PF)56. Neben den Hinweisen auf ein gemeinsames Fußmaß muß es wohl auch Gemeinsamkeiten in der Planung gegeben haben. Metrologische Zusammenhänge, die weit über einzelne Heiligtümer, geschweige denn einzelne Tempel hinausreichen, tauchen schon bei Herodot (2, 168, 1-2) auf, indem er auf die Entsprechung von ägyptischer Königselle und samischer Elle hinweist. Dies ist auch der Eckpfeiler für die Erstellung des kanonischen Fußmaßsystems. Nähere metrologische Zusammenhänge zwischen der Cheopspyramide57, von der die Königselle abgeleitet wird, und griechischen Heiligtümern lassen sich finden: Die Basisseitenlänge der Cheopspyramide von 441 Königsellen (1E = 52,236cm) = 230,36m steht zur Stadionlänge in Olympia im Verhältnis 6 : 5 (441E x 3/2 x 16/15 : 588PF = 6 : 5; 230,36m x 5/6 = 191,96m [Meßwert: 192,28m])58. All dies wirft ein Schlaglicht auf die Diskussion, 53

Sonntagbauer 1998, 136 Tab. 1. Ob es hier auch kultische Zusammenhänge gibt, können wir nicht sagen. 55 Vgl. Gruben 2001, 107; Sonntagbauer 2009 Anm. 5. 56 Nach Gruben (2001, 129) wird auch der Apollon-Tempel in Bassai als ein in den Außenmaßen um 1/3 verkleinertes "Zitat" des Apollon-Tempels in Delphoi bezeichnet. 57 In einer aus unserer Sicht hervorrragenden Studie hat F. W. Korff 2009 die Cheopspyramide metrologisch analysiert. 58 Die geringe Abweichung liegt auch durch die Schwankungsbreite des PF im tolerierbaren Bereich. Der Stadionlauf der Heräen steht zur Basisseitenlänge der 54

Metrologisches in Olympia

69

wie sehr der Orient in der sog."orientalisierenden Epoche"59 die griechische Kultur beeinflußt hat.

Der Zeustempel als Höhepunkt klassischer Tempelarchitektur? Von Gruben als "gültigste Verwirklichung des dorischen Kanon" und in seiner "klassischen Vereinfachung"60 als Höhepunkt der dorischen Tempelarchitektur bezeichnet, können wir den Zeustempel aufgrund der vorgelegten Ergebnisse zwar als unmittelbare Vorstufe zum Höhepunkt des Parthenon, ihn selbst aber nicht als solchen bezeichnen61. Auch meinen wir, daß die Proportionsstruktur des klassischen Tempels nicht bloß aus dem Einheitsjoch in hälftigem Teilen entwickelt wird, sondern sich aus der Antinomie zwischen Stylobat und Peristasis entwickelt. Der nahezu unlösbare Gegensatz zwischen Stylobat und Peristasis, bei Heraklit als Harmonia62, in der Plastik als Kontrapost bezeichnet, wird dabei von einem im weitesten Sinn musikalischen System von Zahlen und Proportionen umspielt, das in seiner schon oft beschriebenen Pendelstruktur einer starren, von hälftigem Teilen bestimmten Schematik völlig widerspricht. Griechische 'Musik' besitzt eine irrationale, letztlich unfaßbare, 'göttliche' Mitte, die nicht beliebig, sondern streng zahlhaft geordnet umtanzt wird. Immer wieder stellt sich die Frage, warum sich griechische Baumeister, abgesehen vom technischen und materiellen Aufwand, derart komplizierten Konstruktionsprinzipien unterwerfen und sich die Arbeit so schwer machen. Nichts wäre einfacher gewesen, als auf einem proportionierten Stylobat die Säulen im proportionierten Einheitsjoch anzuordnen oder um proportionierte Joche einen unproportionierten Stylobat zu legen. Mit demselben Recht könnte man fragen, warum griechische Tragödien mit derart schwierigen Kompositionsprinzipien geschaffen wurden, daß wir Cheopspyramide im Verhältnis 25 : 36, eine Proportion aus Quadratzahlen, wie sie nicht selten in der Tempelarchitektur anzutreffen ist (z. B. Parthenon: Stylobatproportion 4 : 9; STL : Höhe der Langseiten = 81 : 16 [Gruben 2001, 185]). Um Metrologisches weiterzuspinnen: 1/4 der Basisseitenlänge der Cheopspyramide (57,59m) steht zur STL des Zeustempels im Verhältnis 9 : 10 (192,28m : 3 = 64,09m x 9/10 = 57,68m). 59 Burkert 1984. 60 Gruben 2001, 56f. 41. 61 Auch der Strenge Stil ist eine Vorstufe zur Parthenonzeit. 62 Heraklit VS 22 B 51: Harmonia als "des Wider-Spännstigen Fügung wie bei Bogen und Leier" (Snell) definiert.

70

Chapter Three

diese heute erst in ihren Anfängen verstehen. In der Architekturforschung werden zunehmend praktische Gesichtspunkte in den Vordergrund gerückt und die Proportionen als konkrete Hilfen für den Architekten gesehen. Das sind sie auch, aber Tempel sind Kunstwerke mit kultisch-religiösem Hintergrund zur Verherrlichung der Gottheit. Für die Götter ist kein Aufwand zu groß. Man denke nur an riesige monolithische Säulen, monolithische Schwellen und Wandungen von Erscheinungstüren wie in Didyma. Begriffe wie Einheit, harmonische Geordnetheit oder auch kontrapostische Gespanntheit tauchen in der Diskussion um den Tempel immer wieder auf, doch bleiben diese Begriffe im Unbestimmten stehen. Schwammige Vorstellungen von Harmonie lagen den Griechen völlig fern. Harmonia war für die Griechen des 5. Jh.s. ganz konkret das Verhältnis 1 : 2, die Oktav und ebenso der Urbegriff der Vereinigung der Gegensätze, im Mythos so wunderbar als die Vereinigung von Aphrodite und Ares, von Liebe und Haß, am Tempel in einem kontrapostischen Spiel von Zahlen und Proportionen dargestellt. Harmonia, Kontrapost, Antinomie, Vereinigung der Gegensätze, oder wie immer man das nennt: Wo kann sie am dorischen Tempel besser gefaßt werden als im einerseits freien und doch durch Zahlen fest gebundenen Tanz der Säulen auf dem Stylobat. Die Variationsmöglichkeiten dieser 'Tänze' zeigen sich in der rund zweihundertjährigen Geschichte des dorischen Tempels: Er bleibt gleich und doch verändert er sich ständig.

Metrologisches in Olympia

71

Fig. 1. Proportionale Zusammenhänge zwischen den Laufbahnen der Olympien, der Heräen, dem olympischen Zeustempel, dem Poseidontempel in Isthmia und der Cheopspyramide

PART II SCULPTURE

CHAPTER FOUR DATING THE CORNER FIGURES OF THE WEST PEDIMENT AND QUESTIONS ARISING FROM THE USE OF PARIAN AND PENTELIC MARBLES IN THE SANCTUARY OLGA PALAGIA

Abstract Since Treu’s publication of the architectural sculptures of the temple of Zeus in 1897, it has been known that three out of the four female corner figures of the west pediment are in Pentelic marble, in contrast to the rest of the pedimental statues of both pediments, which are in Parian. Treu suggested that the figures in Pentelic marble were later replacements, thus begging the question of their date. This paper will examine the use of Parian and Pentelic marbles in the sanctuary from the archaic to the Roman periods. It appears that Parian marble predominates in the archaic, classical and Hellenistic periods. Pentelic marble takes over from the period of Augustus with the notable exception of the so-called portrait of Poppaea from the Heraion. The almost exclusive use of Parian marble until the Hellenistic period is not only important for dating the repairs to the west pediment of the temple of Zeus but also for dating the Hermes attributed to Praxiteles. In addition, the fact that the portrait of the so-called Poppaea is unique among all Roman portraits in Olympia in its use of Parian marble, raises the question of its workshop and origin.

The sculptures of the temple of Zeus at Olympia are in Parian marble1 except for three out of the four corner figures of the west pediment, which are in Pentelic marble (Figs. 1-4). This discrepancy has been known since Treu’s publication of the sculptures of Olympia and has been the subject

1

Herrmann 2000, 381-383.

Dating the Corner Figures of the West Pediment

75

of debate for the best part of the 20th century and the early 21st.2 The three statues in Pentelic marble (Figs. 1, 3, 4) are considered as replacement figures, but their date remains a matter of controversy. This paper will reopen the question of their chronology and will also offer a brief survey of the distribution of Parian and Pentelic marbles in the sanctuary of Zeus. In addition, it will discuss the issues raised by the use of these marbles. The marble provenance is based on visual identification, attempted by several generations of scholars including this author. Only the statues from the Nymphaion of Herodes Atticus have been scientifically tested and were all found to be in Pentelic marble.3 We begin by taking a look at the two pairs of reclining corner figures of the west pediment. In each corner we see a young Lapith woman in front, half naked, her lower body covered by a himation. The woman in the left corner, known as figure A, retains her head, covered with a head scarf or kekryphalos (Fig. 1).4 She is in Pentelic marble and appears to be the mirror image of figure V in the right corner, who is in Parian marble (Fig. 2).5 Figure V’s head is missing and her right arm is a replacement in Pentelic marble. Its position differs from that of the corresponding arm of her mirror image, figure A. Given the close similarity of figures A and V, we assume that figure V also wore a kekryphalos. We may also assume that A is in fact a copy of an original in Parian marble that was damaged in unknown circumstances. The style of both A and V conforms to the general style of the Olympia pediments. The suggestion that A is a copy is reinforced by the rough underside of her plinth as opposed to the smooth underside of V’s plinth.6 But the situation changes with the rear figures of old women, presumably the servants or slaves of the Lapith women in the foreground. The rear figures are placed at an elevation. These are known as figures B (Fig. 3)7 and U (Fig. 4)8 and are both in Pentelic marble. U leans on a cushion in Parian marble that is presumed to have survived from the original arrangement,9 while B leans on a modern support. The best part of 2

Treu 1897, 89-95. For a summary discussion of the issues, see Younger and Rehak 2009, 91-98. 3 Kane et al. 1999. 4 Ashmole - Yalouris 1967, figs. 62-63. 5 Ashmole - Yalouris 1967, fig. 65. 6 Trianti 2002, 292, fig. 46. 7 Ashmole - Yalouris 1967, figs. 64, 67. 8 Ashmole - Yalouris 1967, figs. 66 and 68. 9 The association of this cushion with the pediment has been challenged by Younger and Rehak 2009, 93.

76

Chapter Four

U in the right corner is damaged except for her head. We do not know how closely she reflected her counterpart B in the left corner but the position of her left arm is certainly different. Unlike B, she appears to wear an ungirt peplos judging by the folds of its overfold. Her short curls are roughly cut, giving the appearance of a wig. The style and drapery of the best preserved figure B in the left corner are peculiar (Fig. 3). She wears a peplos with an overfold, girded at the waist but also forming a low kolpos that is visible under the overfold (Fig. 5). The texture of her dress exhibits a harsh precision and regularity sharply contrasting with the flowing and rather uneven folds of the draperies of the original pedimental statues. In addition, the type of peplos with girded overfold and low kolpos which will not be formulated precisely until the fourth century, makes a tentative appearance in Olympia in the kneeling O of the east pediment (Fig. 6), where it is revealed only at the figure’s back.10 And what are we to make of B’s hairstyle (Fig. 7), especially when we compare it to Lapith Q in the west pediment (Fig. 8)?11 They both have short curls but the Lapith’s are in high relief with drilled centers, while B’s curls are shallow, cut with a flat chisel. The parting in the middle is not found in any other of the Olympia heads. Her features too, frowning brow and lowered corners of the eyes are unique for Olympia, except, of course, for figure U, who has similar features. We infer from all this that whereas the Pentelic marble Lapith woman A (Fig. 1) is a copy of a missing original, her servant B (Fig. 3) is a later interpretation of an original that may have been too damaged to reproduce. The same holds true for U. The date of the corner figures in Pentelic marble has vexed the ingenuity of scholars. It has been correlated with the gradual replacement of the temple’s original lion’s head water spouts in Parian marble with copies and variants first in Parian and later in Pentelic marble. The replacement lion heads vary from one another thus forming several groups which are thought to date from the fourth century B.C. to the fourth century A.D.12 They carry masons’ marks which allow a tentative dating of the replacements in Parian marble to the late classical and Hellenistic periods, while the water spouts in Pentelic marble appear to be Roman.13 The roof tiles of the temple also began to be gradually replaced from the fourth century onward. They too have been tentatively dated thanks to

10

Ashmole - Yalouris 1967, figs. 22-24. Ashmole - Yalouris 1967, figs. 86-88, 90. 12 Willemsen 1959; Younger and Rehak 2009, 55-56, fig. 6a-b. 13 Weber 2013, 200, fig. 148. 11

Dating the Corner Figures of the West Pediment

77

masons’ marks: the earlier replacement tiles are in Parian marble until Pentelic marble is introduced from the mid-second century B.C.14 The cause of the damage or damages is anyone’s guess. Earthquakes in Olympia are rather frequent but few are recorded in the ancient sources; we are therefore left on our own to conjecture what Dinsmoor has aptly named an archaeological earthquake.15 Several solutions have been proposed. Extensive damage to the temple in the fourth century B.C. is attested by the replacement of lion head water spouts and roof tiles mentioned above, as well as the excavation of architectural debris from the temple in the foundations of fourth century buildings like the Echo Colonnade and the Leonidaion, therefore indirectly attesting a fourth-century earthquake.16 The replacement figures A (Fig. 1), B (Fig. 3) and U (Fig. 4) have accordingly been dated to the fourth century B.C.17 Alternatively, they were placed in the second century B.C. in connection with Damophon’s repairs to the cult statue of Zeus mentioned by Pausanias (4.31.6).18 In 56 B.C. the statue of Zeus was struck by lightning. This implies damage to the roof, which was apparently repaired at some later date by Agrippa, as attested by a Latin inscription in bronze letters sunk into a pavement of Phrygian marble that was inserted into the pronaos of the temple of Zeus.19 The replacement figures B and U were recently tentatively attributed to this repair.20 In 36 B.C. a discarded roof tile in Parian marble was reused for inscribing a list of Olympia officials, but this need not be associated with Agrippa’s restoration.21 Henceforth roof tiles from the temple in both Parian and Pentelic marble will be re-employed for honorary inscriptions or lists of officials until the second century A.D.22 The extent of the damage suggested by the removal of these roof tiles is of course unknown. Some scholars have hedged their bets by proposing different periods for the replacement of individual figures. Yalouris, for example, dated the 14

Weber 2013, 202-204, figs. 149-150. Dinsmoor 1941. 16 On debris from the Zeus temple found in the foundations of fourth-century buildings, see Kunze – Weber 1948. For repairs to the temple after 374 B.C., see Weber 2013, 199. 17 Kunze – Weber 1948, 495 n. 11. 18 Dinsmoor 1941, 421; Ashmole in Ashmole - Yalouris 1967, 21-22. On the temple repairs of the second century B.C., see also Weber 2013, 199. 19 IvO no. 913; Spawforth 2012, 163-164; Williams 2013, 84-85. 20 Williams 2013, 84-85. 21 IvO no. 59. 22 IvO nos. 60-91a. 15

78

Chapter Four

replacement figure A to the fourth century but B and U to the first century B.C.23 Younger and Rehak also accepted a fourth-century date for figure A, but suggested that the original V and its late classical replacement A were irreparably damaged in the second century B.C.; they were therefore permanently discarded and replaced by B and U not later than the first century B.C.24 Before we proceed any further, let us consider the distribution of Parian and Pentelic marbles in the sanctuary of Zeus. Thasian marble is apparently absent from the sanctuary. The only attempt to chart the use of Parian marble in Olympia was by Klaus Herrmann.25 In addition to its extensive use for the architectural sculptures of the temple of Zeus, it appears that Parian marble predominated in the sanctuary from the late archaic to the late Hellenistic periods. It was used for statues, large statue bases, columns of votive monuments and parts of buildings like the Philippeion. Some prominent examples will be considered here. The substantial statue base of the dedication of Praxiteles of Mantinea from the early fifth century B.C.26 and the base of a bronze chariot group attributed to the tyrant Gelon of Gela, who scored an Olympic victory in 488,27 are made of large blocks of Parian marble. Gelon’s chariot is signed by Glaukias of Aigina. Two heads of warriors and an arm carrying a shield from a battle group dated ca. 480 B.C. are, as far as I know, the earliest extant statues in Parian marble from Olympia.28 Paionios’ statue of Victory and its pillar, dating from shortly after 425 B.C., are also of Parian marble.29 The Hermes of Praxiteles is of Parian marble, and this reinforces the attribution to Praxiteles considering the scarcity of Pentelic marble in Olympia before the Roman period as we shall see shortly.30 The stylobate, the Ionic columns and the pedestal of the statuary group of the Philippeion, built by Philip II of Macedon soon after 338 B.C., are in Parian marble.31 Parian marble was also used for Ionic columns carrying dedications, for example the twin columns that carried the statues of Ptolemy II and Arsinoe II dedicated by their admiral Callikrates of Samos 23

Ashmole - Yalouris 1967, 179. Younger and Rehak 2009, 92-96. 25 Herrmann 2000. 26 Herrmann 2000, 380. 27 Walter-Karydi 1987, 35, pl. 5; Herrmann 2000, 380, fig. 3. 28 Walter-Karydi 1987, 41-43, figs. 43-48; Herrmann 2000, 379, fig. 380. 29 Herrmann 2000, 384, fig. 11. 30 Herrmann 384, fig. 12; Pasquier – Martinez 2007, 97-103; Palagia 2010a, 102 with n. 48. 31 Herrmann 2000, 385-386; Palagia 2010c, fig. 4.2 (pedestal). 24

Dating the Corner Figures of the West Pediment

79

in the 270s.32 The use of Parian marble continues into the late Hellenistic period as attested by the head of a statuette of Aphrodite.33 During all this time, only a small amount of artifacts in Pentelic marble have come to light in Olympia, mainly in the form of statue bases, like the base of Lysippos’ statue of the boxer Pulydamas, described by Pausanias (6.5).34 An extensive use of Pentelic marble seems to be introduced with the colossal statue of Augustus,35 which was set up in the Metroon in 27 B.C., when it was converted into a temple of the imperial cult as attested by an inscription on the architrave, dedicating the temple to Augustus.36 Late in the reign of Claudius, between his marriage to Agrippina the Younger in A.D. 49 and his death in A.D. 54, over-life-size portraits of Claudius37 and Agrippina38 in Pentelic marble were added to the cult group. Both statues are signed by Athenian sculptors. Philathenaios and Hegias signed their names on the support of Claudius’s statue,39 while Dionysios, son of Apollonios, signed on the side of Agrippina’s plinth.40 More imperial portraits of Claudius’ family in Pentelic marble though without sculptors’ signatures belong to this group. A headless statue has been dated to Nero’s reign and tentatively identified with Claudia Octavia, Claudius’ daughter who married Nero in A.D. 53.41 A cuirassed torso is attributed to Nero himself.42 Nero’s head was later replaced with that of Titus. Another headless female statue was identified by Hitzl with Flavia Domitilla the Elder and by Renate Bol with Livia, Claudius’ grandmother and wife of Augustus.43 Finallly, the emperor Vespasian, clearly a later addition, is recognized in another cuirassed torso.44 From the late reign of Claudius and the reigns of Nero and the Flavians, we have three honorary portraits of local women in Pentelic marble. All are signed by Athenian sculptors. The statues and their 32

Hoepfner 1971, 11-54; Herrmann 2000, 385, fig. 14; Palagia 2013, 147, fig. 9.2. Mallwitz – Herrmann 1980, 198-199, no 139. 34 Olympia Museum 45. Moreno 1995, 91-93. 35 Olympia Museum 110. Hitzl 1991, 34-38, pls. 2-7. 36 Hitzl 1991, 19-24, figs. 2-4, pl. 1. 37 Olympia Museum 125. Hitzl 1991, 38-43, pls. 8-13, 14a-b. 38 Olympia Museum 143. Hitzl 1991, 43-46, pls. 14c, 15-19. 39 Hitzl 1991, pl. 14a. 40 Hitzl 1991, pl. 14c. 41 Berlin, Staatliche Museen, Antikensammlung Sk 1400. Hitzl 1991, 49-52, pls. 26-29. For the identification, see Bol 2011, 116- 122, fig. 7. 42 Olympia Museum 126. Hitzl 1991, 46-49, pls. 20-25. 43 Olympia Museum 142. Hitzl 1991, 55-56, pls. 35-37; Bol 2011, fig. 8. 44 Olympia Museum 127. Hitzl 1991, 52-55, pls. 30-33. 33

80

Chapter Four

inscribed pedestals came to light in the pronaos or near the east side of the Heraion. The portrait statue signed on its plinth by the Athenian Eleusinios has been associated with the base of Claudia Alcinoa.45 Her husband, Lucius Vetlenus Florus, is known to have dedicated a victory monument in honour of Tiberius Claudius Aphrodisius in A.D. 53, her portrait must therefore date from the same period.46 The second portrait, signed on the drapery by the Athenian sculptor Eros, is associated with the statue base of Numisia Teisis, granddaughter of Claudia Alcinoa, which is dated to the second half of the first century A.D.47 Finally, a third portrait statue in the type of the Great Herculaneum woman, also signed on the drapery by the Athenian Aulos Sextus Eraton, was associated by Renate Bol with the statue base of Antonia Cleodice from the pronaos of the Heraion and with a head sporting a Trajanic hairstyle.48 So far, we have noted in the sanctuary of Zeus intense activity by Athenian sculptors creating portraits in Pentelic marble during the second half of the first century A.D. The exception to the rule is the over-life-size portrait of a woman in Parian marble (Fig. 9)49 found in a niche in the cella of the Heraion, right next to the findspot of the Hermes of Praxiteles. This statue lacks a base. To my knowledge, it is the only Roman statue in Parian marble from the sanctuary of Zeus. Her statuary type is the same as that used for Numisia Teisis and Claudia Alcinoa. She was originally identified with Poppaea Sabina, who was married to Nero from A.D. 62 until her death in 66.50 There are problems with this identification, however. To begin with, her hairstyle is a mixture of styles of the Neronian and early Flavian periods: the corkscrew locks over the shoulders recall Agrippina the Younger, while the rows of heavily drilled curls over the forehead follow early Flavian fashions. The uncertain iconography of Poppaea and her lack of coin portraits have cast doubts on her identity and she has come to be known as a local woman shown as a priestess on account of her twisted fillet, and by analogy with the other 45

Olympia Museum 141. Zoumbaki 2001, 295, K 33; Krumeich 2008, 85, pl. 15,4; Trimble 2011, 125-126, 363-364, fig. 3.4. 46 Zoumbaki 2001, 246, B 10. 47 Olympia Museum 140. Zoumbaki 2001, 336, N 26; Krumeich 2008, 85, pl. 15,5; Trimble 2011, 188-191, 363-364, fig. 3.5. 48 Olympia Museum 145 (head), 139 (torso), 193 (pedestal of Antonia Cleodice). Zoumbaki 2001, 211, A 78; Bol 2008; Trimble 2011, 121, 363-364, figs. 3.2 and 3.3. 49 Olympia Museum 144. Krumeich 2008, 83-84, pls. 14.3 and 15, 1; Palagia 2010b,431 n. 3, fig. 1. 50 Treu 1897, 259.

Dating the Corner Figures of the West Pediment

81

private female portraits that we have already discussed.51 Her findspot within the cella, however, and the exceptional use of Parian marble, may indicate an imperial commission that was not executed in Olympia by Athenian sculptors like the other portraits but imported from an unknown workshop.52 We know that the Heraion was used as a secondary location for statues removed from elsewhere, for Pausanias (5.17.4) tells us that the statues of Eurydice and Olympias, originally dedicated by Philip II of Macedon in the Philippeion, were eventually moved to the Heraion.53 If this statue is indeed a portrait of a member of the imperial family, she may have been initially set up in the Metroon and later removed to the Heraion. Her statuary type was employed in Asia Minor for another imperial image of the Julio-Claudian dynasty, that of Agrippina the Younger, in the reliefs of the Sebasteion of Aphrodisias, where she is shown as a new Demeter holding ears of grain.54 The sanctuary witnessed a second period of intense activity by sculptors working in Pentelic marble in the mid-second century A.D., when Herodes Atticus dedicated a Nymphaion decorated with 24 statues of the imperial family, and of his own and his wife Regilla’s family.55 The non-imperial portraits were in fact dedicated by the Eleians. Marble testing of the statues of the Nymphaion has shown that all of them were carved in marble from the same quarry on Mt. Pentelikon.56 Now that we have concluded our brief survey of the use of Parian and Pentelic marbles in the sanctuary, we come back to the question of the date of the replacement corner figures of the west pediment of the temple of Zeus. It should be pointed out that the fact that replacement figure A (Fig. 1) is an almost exact copy (albeit in mirror image) of V (Fig. 2) indicates a date not before the first century B.C. when exact copies of Greek originals begin to appear.57 Replacement figures B (Fig. 3) and U (Fig. 4), on the other hand, are new creations in the manner of the original pedimental sculptures. Their draperies differ enough to show that they are not mirror 51

For the arguments against an identification with Poppaea, see Krumeich 2008, 83-84. 52 Her findspot also prompted Kruse (1975, 100-101) to identify her with an empress. For another imperial portrait found within a cella, see the empress Thebes Museum BE 63, probably from the reign of Claudius, excavated in the temple of Artemis at Aulis, Palagia 2003. 53 Palagia 2010c, 35, 41. 54 Smith 2013, 134-138, C 5, pl. 52. 55 Bol 1984. 56 Kane et al. 1999. 57 Cf. Ridgway 1984, 9.

82

Chapter Four

images. We do not know if they were inspired by fifth-century originals that had to be discarded on account of heavy damage or whether they were not replacements but later additions to the composition. The intense concentration of Pentelic marble statuary in the second half of the first century A.D., at any rate, may suggest repairs to the architectural sculptures of the temple carried out at the same period. The peculiar hairstyle of figure B (Fig. 7) with short, flat locks parted in the middle and curling inwards unlike any other sculptured head in the temple of Zeus, recalls the hairstyle of Agrippina (Fig. 10)58 and may point us in the direction of a Claudian or Neronian repair to the corner figures. Until further evidence comes to light, this proposal may be added to the several others that have been put forward regarding the enigmatic corner figures.

58 Fig. 10: Athens National Museum 3554, from the west slope of the Athenian Acropolis, Rhomiopoulou 1997, 43, no. 31.

Dating the Corner Figures of the West Pediment

83

Fig. 1. West pediment A. Pentelic marble. Olympia Museum. Photo: Olga Palagia.

Fig. 2. West pediment V. Parian marble. Olympia Museum. Photo: Olga Palagia.

84

Chapter Four

mpia Museum m. Photo: Fig. 3. Weest pediment B. Pentelic marble. Olym American School of Classical Stu udies at Atthens: Alison n Frantz P 124. Photographiic Collection PE

Fig. 4. Weest pediment U. Pentelic marble. Olym mpia Museum m. Photo: American School of Classical Stu udies at Atthens: Alison n Frantz P 126. Photographiic Collection PE

Dating the Corner Figuress of the West Peediment

85

B Olympia Museum. M Photoo: American School S of Fig. 5. Wesst pediment B. Classical Sttudies at Athhens: Alison Frantz F Photoggraphic Colleection PE 128.

86

Chapter Four

m Olym mpia Museum m. Photo: Fig. 6. Easst pediment O. Parian marble. American School of Classical Stu udies at Atthens: Alison n Frantz P 85. Photographiic Collection PE

Dating the Corner Figures of the West Pediment

87

Fig. 7. West pediment B. Olympia Museum. Photo: American School of Classical Studies at Athens: Alison Frantz Photographic Collection PE 18.

88

Chapter Four

m Olym mpia Museum m. Photo: Fig. 8. Weest pediment Q. Parian marble. American School of Classical Stu udies at Atthens: Alison n Frantz P 147. Photographiic Collection PE

Dating the Corner Figures of the West Pediment

89

Fig. 9. Portrait statue from the Heraion. Parian marble. Olympia Museum 144. Photo: Olga Palagia. Fig. 10. Head of Agrippina the Elder. From the west slope of the Acropolis of Athens. Athens National Museum 3554. Photo: Olga Palagia.

CHAPTER FIVE NOCH EINMAL RECHTS UND LINKS AM ZEUS-TEMPEL VON OLYMPIA TONIO HÖLSCHER

Abstract The paper discusses the textual evidence for the correct reconstruction of the East pediment. It is argued here that the expressions “left” and “right” should be understood from the perspective of the ancient viewer.

Pausanias beschreibt bekanntlich in seiner Periegese (5, 10, 6-7) die Komposition im Ostgiebel des Zeus-Tempels von Olympia in Bezug auf die zentrale Gestalt des Gottes: Auf der linken Seite des Zeus ( ) befinden sich die Figuren von Oinomaos und Sterope bis zu dem Flussgott Kladeos, auf seiner rechten Seite ( ) die von Pelops und Hippodameia bis zu dem Flussgott Alpheios Die entscheidende Frage aber, ob „rechts“ und „links“ von der Figur des Zeus aus gedacht oder aus der Sicht des Betrachters gemeint ist, ist von der Forschung von Anbeginn heftig diskutiert worden und hat bis heute keine aus dem Wortlaut heraus begründete Lösung gefunden1. Viele Forscher/innen halten die Frage für kaum zu entscheiden, gehen für die Rekonstruktion von anderen Indizien aus und verstehen dann die Formulierungen des Pausanias in diesem Sinn.

* Ich danke András Patay-Horváth herzlich, dass er mich, obwohl ich an dem Kolloquium nicht teilnehmen konnte, aufgefordert hat, meinen Widerspruch gegen seine Ergebnisse – den ich sonst hätte auf sich beruhen lassen – in diesem Band zu veröffentlichen. Ich habe seine Entgegnung gegen die hier begründete Ansicht wahrgenommen, bleibe aber bei meinem Verständnis der Stelle bei Pausanias und verzichte darauf, meinerseits noch einmal zu antworten. 1 Untersuchungen zum entsprechenden Sprachgebrauch des Pausanias: Michaelis 1876, 162-169 (Geographie, Topographie und Gebäude); Müller 1922, 352-354 (Bildwerke).

Noch einmal rechts und links am Zeus-Tempel von Olympia

91

András Patay-Horváth hat kürzlich noch einmal einen Versuch unternommen, Pausanias aus sich selbst heraus zu verstehen und dadurch ein unabhängiges Zeugnis für die Rekonstruktion zu gewinnen, dessen Verständnis nicht schon von vorgängigen Ergebnissen beeinflusst ist2. Er weist darauf hin, dass der Autor in einer späteren Partie (5, 26, 2) „neben der linken Seite des großen Tempels“ ( ) Weihgeschenke des Mikythos beschreibt, die mit Sicherheit im Norden des Tempels aufgestellt waren. Aus dem Kontext gehe hervor, „dass Pausanias vor der Ostfront des Tempels steht (. . .), als er die nördliche Seite des Tempels als die linke Seite bezeichnet“; das sei „merkwürdig, weil die Nordseite des Tempels ihm genau zur rechten Hand liegt“. Offenbar habe „der Tempel, genauso wie ein Fluss, eine festgesetzte linke und rechte Seite“, die bei dem Kultbau von der Ausrichtung der Kultstatue bestimmt sei. Ebenso müssten demnach in der Beschreibung des Giebels „rechts“ und „links“ von der Gestalt des Zeus aus gedacht sein. Nach Pausanias hätten darum die als „Oinomaos“ und „Sterope“ gedeuteten Figuren in der südlichen, „Pelops“ und „Hippodameia“ in der nördlichen Hälfte des Giebels gestanden3. Diese Interpretation scheint jedoch die Bewegungsrichtungen in der Beschreibung des Pausanias nicht genau zu beachten4. Die Aufzählung der Weihgeschenke des Mikythos steht bekanntlich im Rahmen einer zusammenfassenden Beschreibung der figürlichen Weihgeschenke: Zunächst bespricht er die Bilder des Zeus (5, 21-24), dann andere Anatheme, die nicht Zeus darstellen (5, 25-27), schließlich die Standbilder von Athleten (6, 1-18). Die Denkmäler der zweiten Gruppe, zu denen auch die des Mikythos gehören, führt er in einer topographisch angelegten Reihenfolge auf. Unter den ersten Werken stehen eine Gruppe Betender Knaben von Agrigent und zwei Statuen des jugendlichen Herakles auf oder an der Mauer der Altis, wohl im Süden des Zeus-Tempels (5, 25, 57). Von den folgenden Denkmälern sind das Weihgeschenk der Achäer

2

Patay-Horváth 2007, besonders 170-174. Dass Patay-Horváth die Deutungen des Pausanias nicht übernimmt (m. E. ohne ausreichende Gründe), spielt hier zunächst keine Rolle 4 Die Frage, ob ‚links‘ an dieser Stelle von den nach Osten ausgerichteten Tempel oder von dem nach Westen gerichteten Periegeten aus gemeint ist, wurde schon bald nach den ersten Ausgrabungen diskutiert: Hitzig – Blümner 1901, 444. Die m. E. einzig mögliche Lesung vertrat bereits Dörpfeld 1897, 87. 3

92

Chapter Five

mit den Troianischen Helden und die Nike der Messenier im Bereich vor der Front des Tempels gesichert (5, 25, 8-10; 26, 1)5. Die anschließenden Anatheme des Mikythos waren nach Pausanias „zahlreich und nicht in einer Reihe aufgestellt“ (5, 26, 2-5)6. Zuerst nennt er Statuen von Amphitrite, Poseidon und Hestia, nahe bei einer Gruppe von Iphitos und Ekecheiria, die er zuvor in der Peristase des Zeus-Tempels vor der nördlichen Pronaos-Säule beschrieben hatte (5, 10, 10); die Statuen des Mikythos standen also, entweder innerhalb oder außerhalb, nahe bei der Nordost-Ecke des Tempels. Die danach folgenden Anatheme des Mikythos befanden sich mit Sicherheit nördlich des Zeus-Tempels, wo auch die Basis-Blöcke gefunden wurden, die ihnen überzeugend zugewiesen wurden. Aus Pausanias‘ Beschreibung zahlreicher Details kann man zweifellos entnehmen, dass er diese Werke nicht aus der Distanz vor der Front des Tempels, sondern aus der Nähe betrachtet hat: Das heißt, von dem Weg aus, der zwischen dem Tempel und dem Pelopion nach Westen führt. Hier nennt er nun eine große Zahl weiterer Bildwerke, die Mikythos aufgestellt hatte. Der Wortlaut ist klar: Er bedeutet nicht „neben der linken Seite des großen Tempels“, sondern „an der Seite des großen Tempels zur linken Hand“, also links vom Weg des Pausanias. Die Angabe bezieht sich als attributive Erklärung auf , sie lokalisiert die Seite des Tempels in Bezug auf die Blickrichtung des Autors7. Diese Lesung wird bestätigt durch die Fortsetzung der Beschreibung. Nachdem Pausanias noch einige Werke genannt hat, die neben den Weihgeschenken des Mikythos standen, eine Athena, eine Nike und Taten des Herakles (5, 26, 6-7), fährt er fort (5, 27, 1): „Gegenüber diesen genannten Anathemen stehen andere, in einer Reihe angeordnet, nach Süden blickend, ganz nah bei dem Bezirk des Pelops“. Von diesen beschreibt er besonders ausführlich das Weihgeschenk des Phormis (5, 27, 1-8). Alle drei Angaben: „gegenüber“, „nach Süden blickend“ und „nahe beim Pelopion“ weisen eindeutig auf eine Position nördlich, also rechts vom Weg in der Blickrichtung des Pausanias.

5

Zu den Gruppendenkmälern der Akragantiner und Achäer und ihren Standorten s. zuletzt Ioakimidou 1997, 98-99, 255-262 (Akragas); 82-87, 213-2225 (Achäer). Zur Nike der Messenier s. Herrmann, K. 1972; Hölscher 1974. 6 Zu den Anathemen des Mikythos s. Eckstein 1969, 33-42; Maddoli -Saladino 1995, 346-350; Ioakimidou 2000. 7 Zu den sprachlichen Möglichkeiten der Angabe von rechts und links s. Schwyzer 1928, 249-253. Für Beratung in dieser Frage und den Hinweis auf den genannten Aufsatz von Schwyzer danke ich Jonas Grethlein herzlich.

Noch einmal rechts und links am Zeus-Tempel von Olympia

93

Wenn somit die Bildwerke des Mikythos nicht links vom Tempel, sondern links von Pausanias und seinem Weg nördlich des Tempels standen, so fällt dies Zeugnis zunächst für das Verständnis der Beschreibung von „rechts“ und „links“ im Ostgiebel fort. Natürlich wird man daraus nicht gleich schließen, dass Pausanias auch bei dem Giebel die Seiten aus seiner eigenen Perspektive beschrieben haben muss. Die Frage ist aber weiterhin so offen, dass die anderen Indizien für die Rekonstruktion wieder voll zur Geltung kommen können. Unter diesen scheinen mir die Argumente für die Anordnung von Georg Treu, die in neuerer Zeit vor allem von Erika Simon, Hans-Volkmar Herrmann und zuletzt von Helmut Kyrieleis dargelegt wurden, von unbestreitbarer Überzeugungskraft zu sein: Pelops und Hippodameia auf der südlichen, Oinomaos und Sterope auf der nördlichen Seite von Zeus8. András Patay-Horváth hat selbst einen Hinweis gegeben, wie diese von ihm nicht favorisierte - Lösung mit der Formulierung des Pausanias vereinbar sein könnte: Wenngleich der Perieget „rechts“ und „links“ bei Bildwerken in der Regel von der Figur aus meint, so geht er bei Werken der Architektur vom Standpunkt des Betrachters aus9. Der Giebel des Zeus-Tempels war offensichtlich so stark in die Architektur eingegliedert, dass Pausanias auch die Komposition der Figuren in dieser Weise beschrieb.

8

Treu 1897, 44-68, 114-129; Simon 1968; Herrmann 1987, 125-148; Kyrieleis 2013. 9 Patay-Horváth 2007, 172.

CHAPTER SIX AN DER LINKEN SEITE DES GROßEN TEMPELS (PAUS. 5.26.2) ANDRÁS PATAY-HORVÁTH

Abstract This is a short reply arguing against the view expressed in the previous chapter. A new collection of comparable expressions found in the work of Pausanias is used to underpin the notion that the temple had a left and right side by default which were independent from the actual viewpoint of the ancient observer. In this way the passage under discussion can be used as a relevant piece of evidence for the placement of the central figures in the East pediment.

An dieser Stelle möchte ich nur einen Punkt aus dem Diskussionsbeitrag von Prof. Hölscher herausgreifen, da dieser Punkt m.E. entscheidend für das Verständnis des Pausaniastextes und auch für die Anordnung der Mittelfiguren im Ostgiebels ist. Es geht dabei zunächst um die Frage, wie die Aussage des Pausanias (5,26,2) „ ” übersetzt werden soll. Entweder sollen diese Worte „neben der linken Seite des großen Tempels“ bedeuten, oder „an der Seite des großen Tempels zur linken Hand, also links vom Weg des Pausanias.” Ich möchte nicht bezweifeln, dass der Wortlaut vielleicht auch die zweite, von Prof. Hölscher vorgeschlagene Übersetzung bzw. Interpretation zulassen würde, glaube aber dass die erstgenannte entschieden besser ist, da sie der allgemeinen Bedeutung der attributiven Wortstellung von zwischen Artikel und dazugehörigem Substantiv gerecht wird, die andere aber nicht. Wie auch Herr Prof. Hölscher ausdrücklich betont, bezieht sich „als attributive Erklärung auf ”.1 Dass 1

Um die Statuen im Sinne der von Hölscher vertretenen Übersetzung zu lokalisieren hätte Pausanias wahrscheinlich andere Ausdrucksmöglichkeiten (etwa

An der linken Seite des großen Tempels (Paus. 5.26.2)

95

der Periheget die Statuen nicht aus einer größeren Entfernung betrachtet, schließt nicht aus, dass er die linke Seite des Tempels als Bezugspunkt in seine Beschreibung einführt. Eine solche Lokalisierung kann dem Leser bzw. dem Besucher mindestens genauso hilfreich und aussagekräftig gewesen sein, wie die andere, zumal dadurch die reale Position der Weihgeschenke nördlich des Tempels wesentlich klarer zum Ausdruck kommt: sollte Pausanias „an der Seite des großen Tempels zur linken Hand” gemeint haben, dann würde in seiner Beschreibung nichts darauf hinweisen, dass er nun die Frontseite des Tempels verlassen hat und sich an der nördlichen Seite des Tempels bewegt. Ich sehe auch keinen Grund, warum die Fortsetzung der Beschreibung unser Verständnis der früheren Lokalisierung beeinflussen könnte, da die Bezeichnungen „gegenüber” „nach Süden blickend” und „nahe beim Pelopion” auf den großen Tempel überhaupt nicht mehr Bezug nehmen und unabhängig von der hier diskutierten Bedeutung von 5,26,2 sinnvoll und verständlich sind. Die genannten Weihgeschenke standen offensichtlich nördlich vom Zeustempel, aber ob sie sich zur rechten oder linken Hand des Pausanias befanden, wissen wir einfach nicht, da der Periheget darüber nichts mitteilt. Obwohl also der Wortlaut m. E. klar ist und nur „an der linken Seite des großen Tempels” bedeuten kann, ist es fraglich aus welcher Sicht die Lokalisierung erfolgt, da dies im Text nicht expressis verbis genannt ist. Zunächst soll in diesem Zusammenhang der allgemeine Sprachgebrauch des Perihegeten darauf geprüft werden, wie er die Angaben links/rechts benutzt. Dank der Digitalisierung des gesamten Textes ist es relativ einfach, sämtliche zehn Bücher nach ähnlichen Fällen durchzusuchen und dabei ergibt sich, dass Pausanias ziemlich oft etwas „links / rechts” von seinem Weg beschreibt und diese Lokalisierung folgendermassen ausdrückt: / . Oft gibt er auch an, in welche Richtung man gehen bzw. schauen muss, um die Angabe richtig zu verstehen.2 Allgemein wird die Bezeichnung / bei Pausanias nicht in attributiver sondern in prädikativer Wortstellung gebraucht, mit der Angabe eines Bezugspunktes in Genitiv und/oder der Bewegungsrichtung (part. impf. act.) in Dativ. Ohne diese beiden kommen diese Bezeichnungen selten vor, (z.B. 2,2,1; 2,17,3; 3,19,3; 8,32,4) und es ist anzunehmen, dass in diesen Fällen die Blick- bzw. Bewegungsrichtung / 8,37,8. 2 z.B. 2,24,2 2,11,3

…

) gewählt. Vgl. etwa 5,15,3; 5,20,10 oder oder 2,12,2 '

, .

96

Chapter Six

aus dem Kontext so klar war, dass es ihm unnötig erschien, nochmals ausdrücklich darauf hinzuweisen. Wahrscheinlich war also die Bezeichnung auch in 5, 26,2 ohne weitere Angaben verständlich bzw. unmissverständlich, auch wenn die Formulierung ungewöhnlich ist. Im allgemeinen geht Pausanias tatsächlich vom eigenen Standpunkt (bzw. von der Orientierung des mutmaßlichen Betrachters) aus, aber in manchen Fällen (bei Statuen, die Götter oder sterbliche Personen darstellen und bei Flüssen) ist mit der Angabe / + gen. offensichtlich die festgelegte und von der Perspektive des jeweiligen Betrachters unabhängige rechte/linke Seite gemeint.3 Die sprachliche Formulierung ist also an sich nicht entscheidend und der Sinn der Angabe hängt vom jeweiligen Objekt ab, dessen rechte/linke Seite erwähnt wird. Ein Bezug auf die Blickrichtung bzw. auf den Weg des Autors scheidet bei einer attributiven Wortstellung auch nicht aus, da an einer vergleichbaren Stelle (3,13,9) wohl in diesem Sinn verstanden werden muss, ist aber inhaltlich in 5,26,2 sehr unwahrscheinlich. Pausanias könnte nämlich die Weihgeschenke des Mikythos nur dann an der linken Seite des Tempels sehen, wenn er von Westen nach Osten oder (etwa beim Pelopion stehend) von Nord nach Süden blicken würde. Die erstgenannte Möglichkeit ist mit an Sicherheit grenzender Wahrscheinlichkeit auszuschliessen, und die zweite erscheint auch nicht besonders plausibel, da beide einen Wechsel seiner Bewegungsrichtung von Ost nach West voraussetzen würden und daher wohl in der Beschreibung Erwähnung finden müssten. Deswegen scheint 5,26,2 die linke Seite des Tempels nicht vom Standpunkt des Betrachters verstanden zu haben und somit bleibt eigentlich nur die Möglichkeit, dass die Angabe sich auf eine vom Standpunkt und Blickrichtung des jeweiligen Betrachters unabhängige linke Seite des Tempels bezieht. Dieser Gedanke ist, wie ich früher darzulegen versuchte, keineswegs unmöglich, er ist eher naheliegend, da dabei die Perspektive der im Tempel verehrten Gottheit zugrunde gelegt wird. Das lässt sich bei Pausanias auch an weiteren Stellen beobachten bzw. vermuten. So wird die Zeusstatue der Lakedaimonier, dessen inschriftlich gesicherte Basis südöstlich vom Zeustempel gefunden wurde4, in 5,24,3 3

Müller 1922, 352-353. IvO 367. Dass die Basis vermauert und nicht in situ gefunden wurde, bedeutet noch nicht, dass sie ursprünglich sehr weit von diesem Punkt, geschweige nördlich vom Eingang gestanden hätte (so die Annahme bei Hitzig – Blümner 1901, 434 mit der früheren Literatur). 4

An der linken Seite des großen Tempels (Paus. 5.26.2)

97

folgendermassen lokalisiert: . Da östlich unmissverständlich angegeben ist, muss die Bezeichnung „rechts vom Tempel” die südliche Seite bedeuten und zwar wahrscheinlich ebenfalls unabhängig von der Perspektive des Betrachters, der sich laut 5,24,1 vom Buleuterion her, also aus südöstlicher Richtung dem Tempel nähert. Dass die Perspektive der Gottheit für die Angaben links/rechts vom Tempel ausschlaggebend ist, kann am deutlichsten beim Tempel der Athena Chalkioikos in Sparta beobachtet werden: 3, 17, 5-6 , .... … , . Offensichtlich spricht Pausanias in diesen Fällen vom Tempelgebäude der Athena Chalkioikos, aber genannt wird in jedem Fall nur die Göttin, und nicht der Tempel. So ist es wahrscheinlich, dass die Angaben, wie bei Statuen üblich, nicht vom Betrachter aus gemeint sind. Beim Apollotempel in Delphoi kann die Anordnung der Schilde nicht eindeutig rekonstruiert werden und es wird angenommen, dass links/rechts vom Betrachter aus zu verstehen sind, aber es ist ebenso möglich, die Bezeichnungen im hier dargelegten Sinn zu interpretieren.5 Zusammenfassend ist also festzuhalten, dass Pausanias in 5,26,2 eine bei ihm sehr ungewöhnliche und inhaltlich zunächst nicht ganz klare Formulierung bietet, die die Statuen des Mikythos zwar eindeutig an der linken Seite des Tempels lokalisiert, aber nicht angibt aus welcher Sicht die linke Seite des Tempels gemeint ist. Wahrscheinlich erscheint jedoch, dass dabei nicht der Standpunkt und die Blickrichtung des Betrachters entscheidend war, sondern dass der Tempel selbst eine festgelegte linke Seite hatte wie ein Fluss, ein Schiff, ein Mensch oder eine Gottheit. Die ähnliche Formulierung bei der Beschreibung des Ostgiebels (5,10,7: ) bleibt also m.E. weiterhin als ein wichtiges Indiz für die geschlossene Anordnung der Mittelfiguren.

5

10,19,4

, , .

CHAPTER SEVEN THE INVISIBLE GOD: THE REPRESENTATION OF DIVINE INTERVENTION IN THE EARLY CLASSICAL PERIOD TOSHIHIRO OSADA

Abstract In this chapter I discuss an aspect of the sculptural decoration on the temple of Zeus at Olympia which remains open to interpretation, but which, despite the fact that it appears to raise critical issues concerning the contemporary religious mentality, has seldom been addressed. I am referring to the representation of the “invisible god.” In each of the three relief scenes found in the temple decoration (the metope of Atlas, and the eastern and the western pediments), Athena, Zeus and Apollon were respectively represented as invisible to the heroes depicted there. The designers of these scenes appear to have explored the possibility of representing the existence of the gods through the depiction, in a variety of forms, of the roles they played in the activities of each hero. These depictions of divine interventions into earthly events seem to have mirrored the religious experiences of the Greeks at that time.

Introduction I start by referring to one of the metopes of the Temple of Zeus at Olympia (fig. 1). While Atlas carries the apples of the Hesperides, Herakles takes his place carrying the globe. The goddess Athena, the protectress of heroes, stands behind Herakles. Unseen by the hero, she is gently supporting the globe, to assist his labour.1 1

Ashmole – Yalouris 1967, 28, figs. 188–93; Fuchs 1983, 411 fig. 460; LIMC V (1990) no. 1705 pl. 11 s. v. Herakles (J. Boardman); Knell 1990, 83 fig. 119;

The Invisible God

99

As demonstrated by this scene, in Greek art the deity was sometimes depicted as invisible to the heroes portrayed. This chapter focuses on this iconography of the invisible god. It would seem to represent a kind of divine intervention – suggesting that the unperceived support by the deities demonstrates their intervention in the mortal world. In my opinion, this type of representation appears to have developed particularly during the first half of the fifth century B.C., although to date there has been very little scientific investigation of this. The representation of divine intervention seems to have been deeply rooted in the religious mentality of the ancient Greeks. It primarily expressed, albeit in mythical themes, a direct relationship between mortal and immortal, and must therefore reflect the religious view of the contemporary Greeks. In this chapter the depiction of divine intervention in Greek art will be analysed. I will first discuss the literature, then Greek art, and then finally the historical tradition.

Literature Firstly, mythical scenes in literature must be examined, particularly those in the Iliad and the Odyssey. The question is whether Greek deities in literature were described as visible or invisible to the mortal characters present. The answer is that both cases can be found. On the one hand the Greek gods were often depicted as visible to mortals; they appeared to the heroes, and often had conversations with them. In the Iliad (5, 335–49) there is even a scene where Diomedes attacks the goddess Aphrodite: … then the son of great-souled Tydeus thrust with his sharp spear and leapt upon her, and wounded the surface of her delicate hand, and forthwith through the ambrosial raiment that the Graces themselves had wrought for her the spear pierced the flesh upon the wrist above the palm and forth flowed the immortal blood of the goddess … But over her shouted aloud Diomedes good at the war-cry: “Keep thee away, daughter of Zeus, from war and fighting. Sufficeth it not that thou beguilest weakling women?”2

On the other hand, however, the gods were often depicted as invisible to the mortal characters. In the Odyssey (16, 157–64) there is a scene where the figure of the deity is only perceivable by specific, important heroes. Only Odysseus can see the appearance of the goddess Athena,

Stansbury – O’Donnell 1999, 74–75 pls. 31–32; Barringer 2005, fig. 5. 2 Translation by A. T. Murray.

100

Chapter Seven

while she remains completely unnoticed by the other less important characters in the story.3 but she drew near in the likeness of a woman, comely and tall, and skilled in glorious handiwork. And she stood over against the door of the hut, shewing herself to Odysseus, but Telemachus did not see her before him, or notice her; for in no wise do the gods appear in manifest presence to all. But Odysseus saw her, and the hounds, and they barked not, but with whining slunk in fear to the further side of the farmstead. Then she made a sign with her brows, and goodly Odysseus perceived it … 4

The religious historian H. S. Versnel refers to various types of appearance of Greek deities in classical literature, and cites the example of the visible and invisible god. The gods could also disguise themselves as mortal, for example as a hero or as a nurse,5 or they could appear in animal form, for example as a dolphin.6

Representations When the deities are represented with mythical heroes in Greek art, throughout the periods, the most popular theme was single combat between two warriors. A typical composition shows two fighting heroes and two protecting deities depicted as onlookers. The gods appear to be represented as supporting from behind. However, in this chapter we deal with examples where there is clearer intervention by invisible gods, and where they are more actively taking part in the earthly events of the mortals.

The Vase-Paintings in the Late Archaic Period The iconography which can be interpreted as representing invisible gods seems to have begun during the late archaic period. As far as I have observed, the earliest examples are found in attic vase-paintings, notably in those that depict any of the following three themes of the Trojan War.

3

Versnel 1987, 44f. Translation by A. T. Murray. 5 Athena disguised as Mentor, a friend of Odysseus: Hom. Od. 2, 268; Demeter as a nurse: Hom. h. Demeter; Helen as a mortal woman: Hdt. 6, 61. See Versnel 1987, 45. About representations of encounters between mortal and immortal in Greek art, see Klöckner 2010; Platt 2011, 31–50. 6 Hom. h. Apollon. See Versnel 1987, 47. 4

The Invisible God

101

(a) Achilles and Ajax playing a game. A standard composition of this episode shows the two heroes crouching and playing dice on either side of a table, with Athena standing behind it. The goddess seems to represent either a protectress of Achilles, or Athena Skiras, whose oracle involved lottery.7 The representation could also, however, be interpreted as a scene of an epiphany; the goddess might have emerged, as a visible goddess, to warn the heroes about the attack of the Trojans.8 (b) Hermes weighing the destinies of Memnon and Achilles. The theme of Psychostasia appeared most frequently in attic vase-paintings in the late sixth and the early fifth century B. C.9 The same episode was also depicted on the eastern frieze of the Siphnian Treasury, from ca. 525 B. C.10 Many variations in composition can be found; the fullest version, however, depicts the battle between Achilles and Memnon, the supplication to Zeus and Hera by Eos and Thetis for their children’s lives, and Hermes’ weighing of the destinies of the two heroes. The god between the two warriors could be represented as invisible.11 (c) The Achaeans voting on whether Achilles’ arms should go to Ajax or Odysseus (fig. 2). When Odysseus and Ajax quarrelled over possession of the arms of the dead Achilles, the Achaeans decided it by voting. In the representations the victor, Odysseus, usually rejoices, while the loser, Ajax, is in despair. The protectress of Odysseus, the goddess Athena, stands in the middle, her hand raised, to show the victory of her protégé.12 In this image, the figure of Athena seems to be invisible to the two depicted heroes. 7

See for example the vase-painting in Toledo Museum of Art 63.26: Para 149, 23bis; Boardman 1974, fig. 227; Spivey 1994, 50. See LIMC I (1981) 37–200 s. v. Achilles (A. Kossatz-Deissmann). The goddess Athena is represented in nos. 402– 409 (black-figure), 418, 420, 423 (red-figure). Concerning the representation of the invisible Athena, see also Patay-Horváth 2007, 168 n. 12. 8 LIMC I (1981) 103 s. v. Achilles (A. Kossatz-Deissmann). 9 See the list of Knittlmayer 1997, 100–108, 139–40. 10 Schefold 1978, 60; Brinkmann 1994, 75ff. 11 See for example the vase-painting in Roma: Villa Giulia 57912. ARV 72, 24; 1628; LIMC I (1981) no. 804 pl. 136 s. v. Achilles (A. Kossatz-Deissmann); Knittlmayer 1997, 100, 140, E2, pl. 24, 1. 12 Spivey 1994. See also Robertson 1992, 87–89; LIMC I (1981) 325ff. s. v. Aias I (O. Touchefeu). For a typical example see the Kylix painted by Douris: Vienna, Kunsthistorisches Museum 3695 (= Masner 325). ARV 429; CVA I pls. 11ff.; LIMC I (1981) 326 no. 81 pl. 243 s.v. Aias I (O. Touchefeu); Schefold – Jung 1989, fig. 236. See also Patay-Horváth 2007, 168 n. 12.

102

Chapter Seven

The Early Classical Period Both pediments of the Temple of Zeus at Olympia show almost the same composition as the voting scene.13 The standing deities are also represented in the middle of the heroes. The statue of Zeus in the eastern pediment is probably one of the most impressive examples of a representation of an invisible god. According to traditional interpretation the scene depicts the start of the chariot race by Oinomaos and Pelops. However A. Patay-Horváth challenges this interpretation; in his opinion the scene represents the reconciliation between Agamemnon and Achilles.14 I cannot enter into a detailed consideration of the interpretation here, but at least one thing is certain: Zeus Horkios, the god of oaths, is standing in the middle between the two heroes, while they are making a pledge.15 As the guarantor of vows, this god would have represented the embodiment of justice.16 The designers of the Olympian sculptures were evidently very interested in investigating the possibilities of representing the gods as invisible, since in the western pediment they also created a figure of Apollo, standing still, showing little involvement in the Centauromachy depicted there.17 The young god raises his right hand like a judge, to show the victory of the Lapiths. The Atlas metope, mentioned earlier, is the third example of the representation of an invisible god at Olympia (fig. 1).18 In the early classical period in particular, the representation of divine intervention seems to have extended its spectrum. My next example is a vase-painting from ca. 460, by the Niobid Painter, which depicts a scene from the story of Troy.19 Paris is trying to shoot Achilles, and at first the arrows miss him, three of them have fallen on the ground; but Apollo 13

Both pediments of the temple of Aphaia at Aigina were also referred to as examples of the representation of an invisible god: Patay-Horváth 2007, 168 n. 12. See also Osada 2008, 218. 14 Patay-Horváth 2007. 15 Schuchhardt 1971, 299; Robertson 1975, 277–28. See also Der Kleine Pauly V (1979) 1518–19, s. v. Zeus (D. Wachsmuth). 16 Previous interpretations of the meaning of the intervention of Zeus into the mortal world depicted on the eastern pediment of the Olympian temple are analysed in Osada 2008. 17 Buschor – Hamann 1924, 10ff.; Ashmole – Yalouris 1967, 17ff.; Stewart 1983, 143–44; Tersini 1987, 145–52; Sinn 1994, 585–602; Heiden 2003, 186; Osada 2008, 218; Barringer 2005, 236; Westervelt 2009, 147. 18 Schefold – Jung 1988, 163–64. 19 Pelike of the Niobid Painter, Bochum S1060: Schefold – Jung 1989, fig. 230; LIMC I (1981) no. 92 pl. 394 s.v. Alexandros (R. Hampe).

The Invisible God

103

intervenes – he allows an arrow to pierce the heel of Achilles, the strongest hero, which results in his death. The next example concerns a statue of a god: at the fall of Troy Ajax attacks Kassandra, and the princess clings to the statue of Athena, begging to be saved by her. In a vase-painting of ca. 450 Athena herself stands behind her statue, from where she gazes at the behaviour of Ajax (fig. 3a, b).20 The message of this image appears to have been the revenge of the deity. The goddess observes the earthly occurrence, and will take revenge at any cost, without fail.21 The theme of my next example also belongs to the Iliupersis. Menelaos threatens Helen, while she begs him for mercy. He tries to punish her, but her beauty forces him to drop his sword. Salvation through the power of Aphrodite is embodied in the figure of flying Eros.22 This same composition was also represented in two metopes of the Parthenon temple (fig. 4).23 The representation of divine intervention is again found in several mythical scenes. In an example by the Kleophrades Painter, Aphrodite intervenes in the battle between Aeneas and Diomedes (fig. 5), while in an example by the Penthesileia Painter, Eos intervenes in the battle between Memnon and Achilles.24 The extension of the spectrum of depictions of invisible gods also seems to have spread outside the boundaries of myths: a completely new type of representation of Greek gods, which allows them to appear before 20 Amphora of the Polygnotos-group, Cambridge, Corpus Christi College (ARV2 1058.114; Add2 323): Mangold 2000, 173 cat. II 61 fig. 33; LIMC I (1981) no. 54 pl. 261 s. v. Aias II (O. Toushefeu). 21 See also Auffarth 2010, 470. 479–80, who mentions the later representations, painted on Lucanian kraters, of invisible deities standing behind their statues. 22 Oinochoe of the Heimarmene Painter, Vatican, Museo Gregoriano Etrusco, 16535 (H 525) (ARV2 1173.3; Add2 339): LIMC IV (1988) no. 272bis pl. 340 s. v. Helene (L. Kahil); Gold 2000, cat. IV 35 fig. 45. 23 LIMC IV (1988) no. 283 pl. 343 s. v. Helene (L. Kahil). 24 Kylix by the Kleophrades Painter, London E73: ARV2 192, 106; Add. 94; LIMC I (1981) no. 37 pl. 298 s. v. Aineias (F. Canciani); Schefold – Jung 1989, 188–89 fig. 167. Kylix by the Pentesilea Painter, Ferrara, T 18 C. VP: ARV2 882, 35. 1673; Paral 428; Add. 148; LIMC III (1986) no. 315 s. v. Eos (C. Weiss); Schefold – Jung 1989, 255–56 fig. 229. See also the Etruscan amphora from ca. 470/60, Würzburg, 799: LIMC I (1981) no. 41 pl. 299 s. v. Aineias (F. Canciani). It also depicts an intervention by Aphrodite. See also the vase-painting of the stamnos by the Berlin Painter from 490 BC, which shows the battle scene of Achilles and Hektor, with Athena standing between the two: Munich 2406. Beazley ARV2 207, 137; LIMC I (1981) no. 566 pl. 115 s. v. Achilles (A. Kossatz-Deissmann).

104

Chapter Seven

real earthly citizens. A few examples from attic vase-paintings show the scene in which Athena Ergane, the goddess of craft-workers, visits a workshop.25 As W. Burkert points out, out of all twelve Olympian gods, the goddess Athena was considered the most accessible to mortals.26

Historical Tradition So far we have observed the images of deities. The question remains why the representation of invisible gods developed so particularly in Greek art during the earlier phases of the classical period. Why did the depiction of divine intervention become popular in sculpture and in vasepainting at this time? Let us now consider the historical tradition, and examine how the appearance of Greek deities before mortals is described in literature. The representation of an invisible god seems to have reflected the contemporary religious views of the Greeks, since the monuments told stories of divine assistance. But why was this type of representation, a kind of confession of faith, peculiar to this period? In Greek folkloric tradition, until the classical period, incidences of divine appearance before mortals are relatively rare. As J. D. Mikalson mentions, in folk literature the Greek gods seldom appear in their mortal form, at least not until the fourth century in Athens.27 The deities certainly exhibited their will to mortals, but they manifested themselves indirectly, utilizing a variety of means, such as storms, lightning, voices, dreams and other omens.28 However there was one exception, a historical event, in which many stories were told of the appearance together of gods and heroes. This was the Persian war. In and after this war folk stories referring to interventions and epiphanies by the gods and heroes remarkably increased in number. In the case of the battle of Marathon in 490, Pan appeared in the Arcadian mountain.29 The figures of Theseus and the local hero Echetlaeos were also witnessed in Marathon.30 In 480, the heroes Phylakos, Autonoos, and 25

The invisible Athena Ergane is represented on the Hydria Milano: Private Collection, ARV2 571, 73; LIMC IV (1984) no. 47 pl. 708 s. v. Athena (R. Fleischer). 26 Burkert 1985, 141. 27 Mikalson 1983, 53. See also Platt 2011, 1–27. About the battle epiphanies see also Platt 2011, 218 n. 17. 28 Versnel 1987, 50–51; Platt 2011, 10. 29 Hdt. 6, 105; Paus. 1, 28, 4; 8, 54, 6; Versnel 1987, 49; Platt 2011, 55–56, 147. 30 Plut. Thes. 35, 7; Paus. 1, 32, 5; Versnel 1987, 44. 47; Platt 2011, 16 n. 57, 218

The Invisible God

105

the Aiakids emerged, to relieve the Greeks.31 A reason why so many historical traditions were born after these battles could be that this was a folk war; battles against the barbarians seem to have engendered stories of salvation by the supportive, native gods and heroes. If we now turn to the art history, we find interesting testimonies by Pausanias and others. They tell of a painting of the Persian war in the Stoa Poikile at the Athenian Agora, which must have been the inspiration for many later monuments.32 According to Pausanias it included a scene depicting the rout of the Persians, and Athenian warriors such as Miltiades.33 He also mentions the representation of a deity along with four heroes: Athena with Theseus, Herakles, Marathon and Echetlaeos. The literary descriptions alone do not give us a clear idea of how the figures of the gods and the heroes were depicted in the painting. Pausanias describes Theseus as raising himself from the earth, but neither he nor any other source gives us details of the other figures. However, as E. Simon, E. B. Harrison, and T. Hölscher also point out, these mythical figures do seem to have been represented as invisible to the mortals.34 This painting of the Marathonomachy represents the deities, the heroes, and ordinary Athenian citizens in one composition, and so expressed direct participation by the immortals in earthly, historic events. This rendered divine control in visual form, and thus publicly stated to Athenian citizens that divine assistance really existed.

Conclusion According to historical tradition, contemporary Athenians would have considered the Persian war not merely as a political or diplomatic incident, but rather as a religious phenomenon. They would have interpreted the result of the war as an issue of faith. According to Aeschylus the cause of the Greeks’ victory lay not in the fighting abilities of the two sides, but in the fact that the barbarians destroyed Greek sanctuaries, thus disrespecting the worship of the deities.35 These various art forms seem to have n. 17. 31 Hdt. 8, 38–39 (Pylakos, Autonoos), 64, 83–84 (the Aiakids); Versnel 1987, 44; Platt 2011, 16 n. 57; 155–56. 32 All the literary traditions mentioning the Marathon painting in the Stoa Poikile are listed in Harrison 1972. See also Hölscher 1973, 50–68; Stansbury-O’Donnell 1999, 143–44 fig. 62. 33 Paus. 1, 15. 34 Simon 1963, 36; Harrison 1972, 377; Hölscher 1973, 63–64. 35 Aischyl. Pers. 800–842; Castriota 2000, 443–44; McNiven 2000, 87. See also

106

Chapter Seven

influenced each other, and a characteristic religious culture was developed, but under the surface of the various themes, a common message seems to have been buried: that the deities were constantly watching over mortals and intervening in their lives.

Raeck 1981, 228–29.

The Invisible God

107

Fig. 1 Olympia, Zeus Temple, Atlas metope. Photo: © Hirmer Photoarchiv.

108

Chapter Seven

Fig. 2 Vienna, Kunsthistorisches Museum 3695, Attic red-figure Kylix by Douris. Photo: © Kunsthistorisches Museum, Vienna.

Fig. 3a, b Cambridge, Corpus Christi College, Attic red-figure Amphora attributed to the Polygnotos-group. Lewis Collection. © Fitzwilliam Museum, University of Cambridge. By permission of the Master and Fellows of Coprus Christi College, Cambridge.

The Invisible God

109

Fig. 4 Basel, Skulpturhalle, model of north metopes 24 and 25, Parthenon. Photo: © Dieter Widmer (E. Berger, Der Parthenon in Basel. Dokumentation zu den Metopen, Basel 1986, pl. 3,3).

Fig. 5 London, the British Museum E73, Attic red-figure Kylix by the Kleophrades Painter. Photo: © Trustees of the British Museum.

PART III CULT AND HISTORY

CHAPTER EIGHT THE GREAT HECATOMB TO ZEUS OLYMPIOS: SOME OBSERVATIONS ON IVO NO. 14 JULIA TAITA

Abstract Some literary sources document the offering of a magnificent sacrifice to Zeus Olympios, called the “hecatomb,” which was performed on the big ashes altar of the god and represented the peak of the rites in the quadrennial panegyris. No author dwells upon the details of the ceremony, stressing only its colossal character, with the immolation of a very high number of bovine victims. In this context an important piece of evidence is represented by IvO 14 from the second quarter of the fifth century BC. It is the only lex sacra where the term “hecatomb” appears, and it probably aimed at regulating the criteria for choosing the animals to be immolated and at defining the financial means to cover the high costs of such a huge sacrifice. This inscription raises different questions, which are discussed within a broader examination of the archaeological, osteological, epigraphic and literary evidence, in order to define the features of the Olympic hecatomb.

1. In the first century AD the philosopher Epictetus remarked critically on a very widespread tendency of his times, that of going to the sanctuary of Olympia to admire the colossal gold-ivory statue of Zeus inside the temple: […] , '[ ] (Diss. 1.6.23). The wish to see the famous masterpiece of Pheidias was so intense in antiquity that to die without having seen it was considered a

I am very grateful to Cpt. Fred Stålfjell for his friendly help to improve the English of my presentation and to Dr. Kingsley Elliot Kaye for his collegial stylistic revision of this paper.

The Great Hecatomb to Zeus Olympios

113

huge (ibid.).1 However, the main focus of the rites in honour of Zeus was neither the temple nor the statue of the god: it was a big open-air altar extant since the Early Iron Age, long before the fifth century, when his temple was built and Zeus’ statue was enshrined. Because of the high number of pilgrims visiting the sanctuary, Pindar indeed defines the altar of Zeus as o (Ol. 1.92). Although according to Pausanias (5.14.1–15.12) there were sixty-nine altars in Olympia,2 the main altar in the Altis is called sic et simpliciter , without any further specification, even in the oldest inscriptions that mention it, from the last quarter of the sixth to the middle of the fifth (ibid. 14.8; 6.17.1) century.3 Pausanias, who defines it as and as (5.14.4), is the only literary source which describes its structure in some detail, dating to the time of his visit to the sanctuary, in the second half of the second century AD (ibid., 13.8–11). The altar of Zeus had a stone basement, called prothysis, where the first phases of the sacrifices were performed: , , (ibid., 13.9). This basis was surmounted by a big heap, presumably of conical shape. It was formed by ashes and the remains of the sacrificial animals’ bones; at Pausanias’ time it was about 22 feet (6.50 m) high.4 Only the thighs – or, more precisely, the thigh bones5 – of the victims were carried up to the top and burned: < > (ibid.). The reason for this procedure was the special mantic technique used in Olympia, called empyromancy. Pindar mentions the (Ol. 8.2–4): the Olympic seers who were able to foresee the will of Zeus by watching and deciphering the shape and the intensity of the flames whilst burning the ' (Ol. thigh bones.6 The presence of the oracle 1

Statue of Zeus: Kyrieleis 2011, 47–52; Heilmeyer 2012, 118–19. Maddoli 1995, 259–82; Hölscher 2002, 334–36; Mylonopoulos 2011b, 57–59. 3 IvO 5 (= IED 3: 525–500), l. 2; IvO 10 (= IED 14: 475), l. 4; IvO 13 (= IED 19: 475–450), l. 6. The altar is not mentioned by an older inscription, NIO 1 (= SEG 42, 373 and 57, 395: about 570–530). 4 Weniger 1913, 250, followed by Yavis 1949, 211, remarks that the dimensions of the altar mentioned by Pausanias (perimeter of the prothysis: 125 feet = 37 metres; perimeter of the ash altar: 32 feet = 9.50 metres) should probably have been indicated by the Olympic exegetai and reflect the layout of the bomos after its last spring setting up (cf. below n. 8). 5 On the distinction between thighs and thigh bones and the material and symbolic meaning of offering this part of the animal body cf.: Maddoli 1995, 258; Ekroth 2009, 127–28, 144–46, 149. 6 Empyromancy in Olympia: Taita 2007, 92–95, with previous bibliography. 2

114

Chapter Eight

6.70) explains Pindar’s definition of the altar as o (Ol. 6.5) as well as the expression in the official language of the Olympic bureaucracy. The inscription IvO 4 (= IED 9: 500–475), l. 7, refers to the altar of Zeus, pointing to the activity of the manteia which was carried out on the bomos.7 Because of the singularity of the ritual procedure, the main part of the altar was made of the ashes of the sacrifices carried out here: (Paus. 5.13.8). Moreover, once a year, the ashes from the rites in the Prytaneion were brought out of this building, mixed exclusively with the hard calcareous water of the river Alpheios, and plastered onto the ash mound (ibid.,11).8 In this way the mixture became so hard and resistant that it was even possible to carve steps into it, leading to the top of the altar. The rather ambiguous formulation of some passages in Pausanias’ description of the altar, and the practically total absence of material traces of its structure,9 have led to many different hypotheses as regards its real Ekroth 2009, 146, stresses the combustible qualities of the thigh bones as an additional reason for choosing them as the part to be burnt: “Recent experiments with burning fresh bones stripped of meat have demonstrated that they burn very well […]. The bone fire has high, large and durable flames with an intense heat and there is rapid extinction of the embers. Such a short, spectacular fire would suit the practical requirements of a Greek thysia very well […]” and, of course, the needs of an empyromancy oracular rite too. 7 The prescribed penalty for the theokolos, the highest Olympic cultic magistrate, in case of abuse of his authority, was to be removed from the altar of Zeus, where he performed his sacrificial duties together with the manteis: ’ (IvO 4 = IED 9, l. 7). On this inscription see Taita 2007, 120–22. 8 The annual setting up of an altar was a quite common procedure in many sanctuaries, e.g. in Athens, Eleusis and Delos. For Olympia we don’t know the details of the ritual ashes’ moving from the altar of Hestia in the Prytaneion to the altar of Zeus in the Altis. Weniger 1913, 256, observes: “Wie man mit der Opferasche verfuhr, ob man sie, wie anderwärts, mit Schaufeln abhob, in eigenen Behältern zugleich mit der vom Herde verwahrte oder auf dem Altar unter freiem Himmel liegen liess und sich durch Aufstellung eines metallenen Rostes half, ist nicht gesagt. Der Schriftsteller fand keine Veranlassung, über diese Einzelheiten der Dienerarbeit zu berichten […].” These build-up operations were probably carried out by the Olympic architect (attested by IvO 62,14) under the superintendence of the manteis: Weniger 1913, 250. For the exclusive use of the Alpheios’ water in the annual re-shaping of the altar cf.: Schleif 1934, 148–49; Taita 2007, 87–88. 9 Mallwitz 1972, 84; Heilmeyer 2012, 113: “[…] von dem […] Zeus-Altar ist kein Bauelement erhalten […]” and 119; Senff 2012a, 11: “[…] wurden bei den Ausgrabungen keinerlei Reste gefunden”; Fuchs 2012, 21; Herrmann 2013b, 113; Moustaka 2013, 119. See also below, n. 91.

The Great Hecatomb to Zeus Olympios

115

shape.10 The most convincing suggestion, which satisfies all elements mentioned by the Perieget, was offered by H. Schleif in 1934 and today is widely accepted.11 Pausanias does not dwell upon the description of the sacrifices in honour of Zeus. According to the author, (5.13.10): the sacrifices were offered not only during the big panegyris, every four years, but also in the daily administration of the cults by the Eleans12 and sometimes by private citizens. He speaks generically of , “sacred victims,” without details about quantity, type, or provenance of the animals. Some other literary sources offer more information on the performance of the great ceremony within the framework of the panegyris. Timaeus of Tauromenion reports that , , , (FrGrHist 566 F 41A). According to the famous mirabilium – i.e. the submarine connection believed to exist between the mouth of the river Alpheios and the spring Arethusa in Sicily13 – the amount of dung from the washing of the animals in the Alpheios was so enormous that even in the far Sicilian town of Syracuse, beyond the Ionic sea, the water of the Arethousa poured out impure. Setting aside this legend, the significant fact is the contamination of the river’s waters caused by the huge amount of sacrificed animals. The same detail is mentioned by Polybius and Strabo. They give a further indication about the type of the victims: (Strab. 6.4.2 = FrGrHist 566 F 41C) and (Polyb. 12.4d = FrGrHist 566 F 41B). This information about the bovine victims is confirmed by Euripides: ' […] (Hipp. 532–37) and Dio Chrysostomus, who mentions the sacrifices of 10

Adler 1892, with Taf. CXXXII; Trendelenburg 1902, 27–39, with Taf. II–III; Weniger 1913, 250–55; Schleif 1934, 153–56, with objections to the previous restorations; Yavis 1949, 211–13, with fig. 54. 11 Schleif 1934, 142–53, with Abb. 8–9. His reconstruction hypothesis has been accepted, among others, by: Mallwitz 1972, 19–20, with fig. 8; Herrmann 1972, 67–68, with Abb. 34; Heilmeyer 2012, 113–14. 12 The cultic official who was entrusted with the daily sacrifice to Zeus was the , as documented by IvO 61, 12; 62, 13; 64, 22; 68, 6; 78, 5; 83, 6; 84, 21; 86, 17; 92, 24; 92a, 1. 13 Already in Ibycus (PMG 323) and then Pindar (Nem. 1.1–2 and schol. ad loc., 3 Drach.), Polybios (12.4d = FrGrHist 566 F 41B), Strabo (6.4.2 = C 270–71), Pausanias (5.7.2; 7.24.3; 8.54.3), Plinius (Nat. Hist. 2.225) and Seneca (Nat. Quaest. 3.26.5). Cf. Maddoli 1995, 212–13.

116

Chapter Eight

(12.51). The first author offering a specific definition of the great Olympic sacrifice is Lucian, who describes as a topos of the activities of Zeus (Bis. Acc. 2), although without indication of the features of the hecatomb itself. Using the statements in the quoted sources, modern scholars have accepted that the sacrifice on the big altar of Zeus was the religious highlight of the Olympic panegyris. A huge quantity of animals would have been offered to the god: considering the etymology of the word , one hundred, or rather more precisely, one hundred oxen.14 L. Weniger observed: “Es liegt kein Grund vor, darunter [i.e. unter der Hekatombe] nur eine das gewöhnliche Mass überschreitende Darbringung zu verstehen. Opfer von hundert Rindern und noch grössere wurden anderwärts ausgeführt: warum hätte man in Olympia sparen sollen?”15 However, once we set aside the testimony in the literary tradition, what do we really know about this sacrificial ceremony? Thucydides says that it was carried out (1.126.5). In the collective imagination of the Greeks this rite was legendary, and yet concrete details about how it was performed are scarce. As T. Hölscher pointed out, “es ist bedauerlich, dass die Rituale, die hier vollzogen wurden, sich nicht in wirklicher Anschaulichkeit rekonstruieren lassen.”16 There are indeed 14

Weniger 1913, 248: “Und dass Grossopfer von hundert Rindern im olympischen Zeusdienste dargebracht wurden, hätte man nie bestreiten sollen”; 259: “die unermessliche Opferung”; Mallwitz 1988, 79: “[…] the focus was on the sacrifice at the great altar, where, in good times, perhaps more than one hundred cattle […] were slaughtered […]”; Renfrew 1988, 177; Lee 2001, 51–52: “The feasting could have occurred at the midday meal, but, given the logistics of cooking so many animals for so a large crowd, more likely in the evening after the afternoon competition […]”; Miller 2003, 18: “[…] the central element in the procession was the hundred oxen for the hekatombe […]” and 35, n. 77; Sinn 2004, 119: “[…] eine Hekatombe, also ein Aufgebot von einhundert Stieren geopfert […]”; Fuchs 2012, 21: “[…] 100 Rinder am Hauptaltar des Zeus geschlachtet wurden”; Gehrke 2012, 32: “[…] dem grossen Opfer von 100 Stieren, der Hekatombe, dem religiösen und damit eigentlich Höhepunkt des Festes […]”; Heilmeyer 2012, 113; Herrmann 2013b, 110, “Schlachten von 100 Stieren”; Moustaka 2013, 120: “die grossen Stieropfer.” The statements of Bocher 2012, 142, “[…] den Hekatomben-Opfern, bei welchen Zeus jedes Jahr einhundert Rinder geopfert wurden,” and of Wacker 2012, 271, “[…] das Hauptopfer […], bei dem laut Pausanias 100 Stiere geschlachtet wurden […] ,” are clearly incorrect. 15 Weniger 1913, 248; see also Weniger 1905, 200: “[…] die Zutritt bedeutender Massen von Opfertieren […].” 16 Hölscher 2002, 338; but see already Ziehen 1939, 18: “Leider sind wir nur

The Great Hecatomb to Zeus Olympios

117

many aspects that need to be examined: the exact place where the rite was carried out; the true proportion of the sacrifice, to ascertain if it was really as colossal as depicted by the ancient authors; the type of animals which were immolated to Zeus, to determine if they were actually only bovine victims or others too; and – a relevant question with regard to the study of the sanctuary’s sacred finances – the provenance of the sacrificial victims.

2. A very important piece of evidence concerning this sacrifice comes directly from the sanctuary, but it has not received due attention from scholars up to now: the inscription IvO 14 (= IED 18). This is roughly contemporary to the construction of the temple of Zeus, paleographically dated by S. Minon to the second quarter of the fifth century.17 The text, inscribed on a small bronze tablet (20.4 x 9.7 cm), is unfortunately fragmentary. However, from the content in the first line, it may be assumed that we are reading the very beginning of the inscription or, at least, the beginning of a section of it.18 In line 5, which is almost half-cut, the very important expression [ - - ] appears: this means a “perfect,” a “fully performed” hecatomb. Amongst the many published and unpublished inscriptions on sacred subjects from Olympia, this is the only lex sacra where the term “hecatomb” is mentioned. It thereby manifests the official use of this word in the language of the Olympic bureaucracy to designate the sacrifice to Zeus. So the word in the quoted passage from Lucian has to be interpreted as a precise reflection of the Olympic sacrifice’s official name. The reference to the very rite which took place during the panegyris and not to any other ceremony for Zeus is demonstrated by the presence of the term [ - - ] (l. 4), the well-known figure of the competition judge,

ungenügend darüber unterrichtet und können uns nur ungefähr ein Bild davon machen […].” Miller 2003, 35 n. 77, observes: “[…] his [i.e. of Weniger 1904, 125–31] interpretation has never been challenged […]. I, too, accept it, but with a hope that stronger evidence will some day emerge.” 17 Minon 2007, 127. This inscription dates back to the time which represents the highlight of the Panhellenic success of the Olympic games. As the archaeological evidence shows, the biggest concentration of excavated water wells in the sanctuary is to be placed between the last quarter of the sixth and the first half of the fifth century: Mallwitz – Herrmann 1999, 194–95, with Abb. 112; Gauer 2012b, 102–3. 18 See below n. 27.

118

Chapter Eight

operating almost exclusively during the Olympic festival.19 Furthermore, the expression in line 2, F [ - -], recalls the very similar formulations in Pausanias’ description of the state sacrifices to Zeus – […] (5.13.10) and […] , (ibid., 9) – attesting that the inscription is specifically related to the official sacrifice offered by the Eleans. The use of the simple verb – not , as in the leges sacrae older than IvO 1420 – means “to carry out a collective ritual action.”21 Therefore, the Eleans, in charge of the organization of the whole Olympic festival, were also responsible for conducting and financing the hecatomb. This inscription was meant to regulate the performance of the sacrifice, wherein the competition judge seems somehow to be involved. The action of the hellanodikas – or of the hellanodikai22 – is described by the perfectly readable verb . This is the aorist infinitive from and indicates the operation of observing, examining, and checking, in a general sense and in different fields, not necessarily with a specific religious meaning.23 The hellanodikas had to carry out a control on something or someone that was presumably mentioned in the inscription. The verb seems thereby to allude to the so-called procedure of the dokimasia. Pausanias testifies that such a function was not outside the range of the judges’ competences. They had to verify ( ) the requirements which men and boys, horses and foals needed to fit to be admitted to the games, and had to issue a final decision, positive or negative, as to their participation.24 In the ritual context suggested by IvO 1425 it has to be assumed that the hellanodikas had to control and evaluate 19 For an overview of the figure and functions of the Olympic hellanodikai cf. Oehler 1912 and Ziehen 1939, 29–31. 20 NIO 1, l. 3; IvO 5 (= IED 3), l. 2. 21 A new examination of the two main meanings of the verb is offered by Parker 2009, part. 171: “The linguistic intertwining of sacrifiant, sacrificateur and the participants at large (the ) through the ambiguities of the single term is an aspect of the irreducible complexity of sacrifice itself.” 22 Because of the fragmented character of the text, it is not possible to ascertain whether the substantive is declined in the singular or in the plural form: Minon 2007, 129, 533. 23 Meister 1889, 27; Minon 2007, 129. 24 Paus. 5.24.10: ; 6.2.2; 23.2; Philostr. Gymn. 25–26. About this dokimasia function of the Olympic judges see also: Ziehen 1939, 8–10; Frisch 1988; Lee 2001, 31–32; Sinn 2004, 126–27. 25 Anyway, there is no element in the inscription that supports the hypothesis of Blass 1884, 328, no. 1160: “Es scheint ein Opferorakel des olympischen Zeus auf

The Great Hecatomb to Zeus Olympios

119

not the characteristics of athletes or competition animals, but the quality of the sacrifices. The designated victims had indeed to be immolated to Zeus in order to perform an hecatomb that must be : fully and perfectly accomplished. This means perfect in its content because of the quality and the fitness of the chosen animals,26 and in its form because of the correctness of the ritual procedure. The inscription IvO 14 is therefore relevant in two respects. First, it attests to a function of the Elean hellanodikai, the dokimasia in a sacred sphere, which is not testified to in any other literary source or inscription. These magistrates were bestowed with such judgmental competency not only during the competitions, but also in the preceding phases, in connection with athletes, horses, and sacrificial animals. Secondly, this lex sacra was intended to define the criteria of an indispensable preliminary act to the hecatomb, i.e. the choice of the animals, whose provenance and quality had to be ascertained and accepted. Hence we have to examine how the victims for the sacrifice of the Eleans were provided.

3. There is no explicit information in the text about the provenance of the animals. However, an interesting hint is offered by the expression at the beginning of the inscription, the genitive , from , “the tenth parts” or, more precisely, “the supplementary tenth parts.”27 This is followed by another element, which remains unidentified because of the lacuna in the tablet. The , “tenths,” were usually the 10% of the harvest, i.e. a percentage of agricultural products, mostly cereals, vegetable, fruits, and oil.28 In the Olympic area there is a very significant parallel regarding tenths devoted to religious purposes.

eine Anfrage der Eleer mitgeteilt zu werden.” 26 Minon 2007, 129, 533. On the physical criteria defining a victim as teleia and for other examples of dokimasia on sacrificial animals cf.: Gauthier 1984, part. 846–48; Lombardo 1992, 423, 432; Chandezon 2003, 292; Hermary 2004, 95–101; Feyel 2006; Brulé – Touzé 2008, 111–26; Chaniotis 2011, 36 and n. 347. 27 Minon 2007, 128. We wonder whether the genitive at the beginning of the inscription might have the same function as in another epigraphical text from Olympia, i.e. a genitivus pertinentiae, where the subject, the substantive F , may be understood. In this case the genitive would represent the titulus of the inscription, as in NIO 1, l. 2, F , “the provision of / about the Kronikà,” i.e. the rites for Kronos. The publication of the full text of NIO 1 is being prepared together with Peter Siewert. 28 Jim 2014a, 47–48.

120

Chapter Eight

Describing the sanctuary for Artemis Ephesia near Skillous, south of Olympia, Xenophon states that he (Anab. 5.3.9). This use of an annual offering of the tithe to the goddess from the land’s products is also reported among the provisions of the lex sacra on the stele near the temple: (ibid., 13).29 In analogy to this local sanctuary, the reference to the epidekata in the inscription IvO 14 might imply that the sanctuary of Zeus also owned some cultivated properties from which the tenths came. Although for Olympia we do not have any direct testimony in this regard, we should consider an interesting parallel with the sanctuary of Apollo in Delos. Some inventories of the gods’ rented properties in the Hellenistic age employ exactly the same term as in our Olympic inscription, , “the supplementary tenths,” to define a special leasing condition: the payment of an annual rent on the land lot consisting in the tenth plus the tenth on the tenth from the products.30 Of course we cannot be certain that the expression has the same technical meaning in classical Olympia as in Hellenistic Delos. But this singular term at least suggests that a procedure similar to the one attested in Delos should be assumed for the Olympia region too. We can thereby suppose that the properties owned by the sanctuary, presumably located in the Alpheios valley, were leased. The lessees had to hand over their agricultural contribution, i.e. at least a tenth of the land’s products, to the administration of the sanctuary, probably once a year. Such contributions were then exploited in connection with the hecatomb, as in other Greek poleis and sacred centres. The sacrificial system documented for the Athenian patrioi thysiai and the local one of the Attic demes implied 29

Jim 2014a, 193: “The land was leased out, and tenants were required to pay a dekate of the land’s produce […]”; 40 n. 45; 47–48; 257–58. 30 Chandezon 2003, 114–33, part. 116–24, nn. 25–29. The Delos inventories – only the one going back to 250 is fully readable; the others are fragmentary (ibid., 117) – concern lots of land which had been assigned for ten years. These properties are of various kinds, according to the different geomorphological features of Apollo’s possessions: some were mainly destined to husbandry (in Delos), others seem to have had a chiefly agricultural character (as Rheneia and Mykonos): ibid., 115, 128–31. Another meaning of , mainly attested for Athens and Attica, is “tenth from confiscated property,” which was usually paid into Athena’s treasury: Jim 2014a, 264–66. However, it seems likely that, in the case of IvO 14, this peculiar meaning has to be excluded. The cultic context of the inscription and the strong analogy with the sanctuary of Delos suggest indeed that the mentioned come exclusively from the sacred properties of Zeus, not from lands which had been confiscated from private people.

The Great Hecatomb to Zeus Olympios

121

indeed that the funds for sacrifices were raised through the leasing of sacred properties.31 The same procedure is testified e.g. for the sanctuary of Athena Ilias in Ilion.32 Therefore, we may assume that the Elean administration collected the tenth parts of the incomes from the sacred properties of Zeus. These, or at least a part of them,33 were then sold to finance the purchase of the animal victims for the hecatomb.34 For the small sanctuary of Artemis Ephesia, Xenophon informs us also that (Anab. 5.3.9), a part of the immolated animals came from the sacred pasturage of the goddess. However, in this case it was just a part and, for Olympia, we don’t have any hint about animals sacred to Zeus which were devoted to sacrificial purposes. From the inscription IvO 7 (= IED 4: 525–500), line 2, we can indeed infer that the sanctuary could rely on bovine herds dedicated to the divinity.35 But, as we know from the better-attested cases in Delphoi and Delos, these herds and flocks, mainly composed of female animals, were at first meant to give an income to the sanctuary, because their skins, wool, milk and derived products were usually sold.36 So we may consider that the Elean administration had to buy the victims both for the big sacrifice every four years and for the small ones every day. The purchase was

31

Rosivach 1994, 120–46. IK Ilion 10, ll. 23–24: [ ] [ ] . See also Chandezon 2003, 188. 33 A part of the agricultural tenths might have been kept by the administration and destined for the bloodless monthly offerings on all the altars in the Altis. As Pausanias remarks, the Eleans (5.15.10): these wheat corns might have come from the landed properties of the sanctuary. Of course some funds from the tenths might also have been allocated to pay for other miscellaneous expenses involved in the sacrifices: Rosivach 1994, 143. 34 On the purchase of sacrificial victims see in general: Rosivach 1994, 108–14; Chandezon 2003, 291–93. 35 The first provision of this inscription orders that, in the case of sexual intercourse in the sanctuary, the guilty , i.e. had to accomplish a full purification and to be subject to another sanction which involved an ox. As this animal doesn’t seem to have been destined for sacrifice – for this purpose the victims were generally swine – we may assume that the ox had to be handed over to the Elean administration for the herds sacred to Zeus. On this provision of IvO 7 cf. also: Taita 2007, 26; Minon 2007, 31–33. 36 Jameson 1988, 104; Chandezon 2003, 39 (Tegea); 54–69 (Delphoi); 114–39, part. 126–28, 132–33 (Delos); 186–88 (Ilion); for other examples, 283–93; Hermary 2004, 95–96. Otherwise the oxen could be exploited for agricultural activities. 32

122

Chapter Eight

probably financed by the proceeds from the sale of the epidekata, but, evidently, from some other income sources too.37 The rich dossier of Hellenistic inscriptions from the sanctuary of Artemis Leukophryene in Magnesia on the Meander attests that the theoroi, the sacred envoys announcing the upcoming festival, received some gifts from the visited poleis. Beside a few accommodation benefits, quite often also a (or ) is granted: this is a sum of money specially destined for the public sacrifice in honour of Artemis during the feast.38 Unfortunately, for Olympia we don’t have any direct evidence about this kind of gift being employed. However, a decree from Delphoi offers an interesting hint in this regard. The koinon of the Aetolians accepts the newly instituted festival Nikephoria in Pergamon of 182 and decides to give the theoroi [ ] (Syll.3 629, ll. 26–27). Another important indication comes from a Geloan decree which recognizes the new Asclepieia on the isle of Cos of 242: the polis establishes that ] [ ] [ ] , [ ] [] , [ ] [ ] (Asylia 49 = SEG 12, 380, ll. 30–33). For the new festivals of the Hellenistic age, the old traditional panegyreis of Olympia, Delphoi, Nemea, and Isthmia represented a model to imitate in the organizational procedures as well. So, although the Aetolian decree mentions only the granting of two other privileges to the Olympic theoroi

37

Since it may be assumed that in Olympia there were probably two distinct treasuries with different destinations of the respective funds – or, at least, two separate accountings of one treasury – (Taita 2014, 136), we can suppose that the sacrifices were financed by means which ought to have been used strictly for cultic expenses. Such a procedure would be analogous to the better-attested systems in Athens, for the so-called epithetoi heortai (Rosivach 1994, 59–60), and in other Greek poleis, where the state administration allocated special funds – in Athens some of the ordinary revenues of the polis, mainly from indirect taxes – to buy sacrificial victims: Rosivach 1994, 108, 115–20, 146–47. 38 IvM 30, 1; 31, 34–36 ( ); 36, 19–20; 44, 32–33; 52, 33; 59a, 6–7; 60, 24–25; 82, 10– 12; 83, 12–13; 85, 22; 87, 20. See further FD III, 1, 308, l. 11. On this peculiar meaning of aparché cf.: Boesch 1908, 77–82, 90, 104; Perlman 2000, 46–60; Jim 2014a, 203, 236–49, part. 244–47 on Magnesia; Jim 2014b, 20. From the Magnesia epigraphical dossier we know that this sum could be a variable amount, depending on the economic power of the polis, on average about 100 drachmas (IvM 60, 25– 26; 85, 22; 87, 20). Other amounts: 15 drachmas (IvM 36, 20); 20 drachmas (IvM 82, 11–12); 150 drachmas (IvM 31, 35–36; 44, 32–33); 2 staters (IvM 30, 1).

The Great Hecatomb to Zeus Olympios

123

– the ekecheiron and the xenia,39 not the aparché – and the Geloan document alludes generically to a certain sum of money for the sacrifice and the procession to Asclepius, we may assume that the states they visited also provided the Olympic theoroi with the aparché to support the great Elean hecatomb.40 Given this background, we may wonder whether the missing expression following in line 1 of IvO 14 might perhaps contain an allusion to the aparché. Along with the funds from the selling of the tenths, the grants of the invited poleis also would have been used by the Elean administration to buy the animal victims.

4. The financial contribution of the Greek poleis to the Olympic hecatomb may also enlighten us regarding a sacrificial usage at which the literary sources hint. In the homonymous comedy of Aristophanes, Lysistrata reminds the Athenians and Spartans of their practice of performing together the sacrificial rites in the Panhellenic sanctuaries, sprinkling from one cup on the altars: / / , , (vv. 1129–32). Isocrates, praising the Greek panegyreis, stresses their special feature of making prayers and sacrifices in common, (Paneg. 43).41 For a polis, the act of taking part in a big festival meant having the opportunity and the honour to sacrifice together with the community that had organized the feast as well as the other cities – thus, in effect, with the whole Hellenic community. This joint performance is technically designated as by many decrees of the poleis invited to the Leukophryeneia in Magnesia.42 When the Spartans were excluded 39 On these two gifts see: Boesch 1908, 74–77, 79–88, 104; Perlman 2000, 46–60; Jim 2014a, 245–46 and n. 154. 40 Also Jim 2014a, 247, remarks: “As the new Panhellenic games modelled themselves on the characteristics of the traditional Panhellenic festivals […], it is reasonable to think that monetary aparchai in the Hellenistic documents might be an extension of the traditional honour for the gods at the four traditional festivals.” See further Jim 2014b, 20. 41 See also the pregnant expression of Plato, Leg. 12.950E:

. IvM 31, 24 ( ( ) ), 30–31 ( [] ( ) ); IvM 32, 29–30; IvM 44, 17. Among other festivals see e.g. Choix Delphes 73, ll. 10–11 ( ] [ ] [ ) and ll. 20–21 about the Ptolemaieia in Alexandria. 42

124

Chapter Eight

from the Olympic games of 420, they , (Thuc. 5.50.2).43 Unlike all other Greek poleis that sent their theoroi to Olympia – with the exclusion of the Triphylian Lepreon – the Spartans were forced to sacrifice in their own town, as they could not perform their religious duties in the Panhellenic sanctuary, as usual, together with the Eleans and the other Greeks.44 Therefore, we may assume that each polis invited to Olympia carried out its own sacrifice, but still within the framework of the great hecatomb, which in this way could reach really huge proportions.45 However, because of the lack of direct evidence, it’s difficult to define the exact chronological relation between the Elean hecatomb and the rites performed by the single theoriai, nor to establish whether the sacrifices of the other poleis followed the Elean one on the same day or rather at a later moment.46 From the literary sources we learn that the theoriai performed their own , bringing their , the ritual objects, mainly vessels, needed for the sacrifices.47 A passage of Ps.-Andocides refers to the pompé and the pompeia of the Athenians in Olympia, mentioning (4.29). A fragment of Polemon of Ilion, reported by Athenaeus (11.59), lists all the precious gold and silver items kept in 43

On this episode see Roy 1998. Gribble 2012, 52 n. 35, properly remarks: “[…] it was participation in the sacrifice, the religious festival itself, which established and renewed the participation of the city and its religion in the Panhellenic cult”; see further ibid., 63. 45 See the observation of Plutarch, Demetr. 11.1: . Cf. Ziehen 1939, 18: “[…] eine große Prozession stattfand, an der auch die Festgesandtschaften der auswärtigen Staaten mit ihren Opfertieren teilnahmen. Sie endete am Altar des Zeus, wo sowohl die Elier selbst wie die durch ihre Theoren vertretenen fremden Gemeinden ihre Opfer darbrachten”; Kavoulaki 2011, 140– 41: “The broader frame of the major event allowed the formation of and accommodated units arranged and conducted towards the same goal. Directing and directed towards the same destination, these groupings […] were set on the move, forming the prosodos to the altar.” 46 A problem which is still open is the definition of the moment when the Elean hecatomb exactly took place. About this question and the broader one of the sequence of the sacred and agonistic events during the panegyris see e.g.: Weniger 1904; Ziehen 1939, 10–16; Mallwitz 1972, 69–70; Lee 2001; Wacker 2012, 270– 71. From the description of Philostratus (Gymn. 6) it seems that the Greek delegations sacrificed soon after the Elean rite: , . 47 On the meaning and use of the word cf.: True 2004; Kavoulaki 2011, part. 138–39. 44

The Great Hecatomb to Zeus Olympios

125

the Olympic treasuries of two colonies, Metapontum and Byzantium.48 They cannot all be considered as pure votive objects for the god, without any functional purpose, first because of the high number of some of them. It is indeed likely that the 132 silver phialai of the Metapontines were used by the members of the sacred delegation and perhaps also by the citizens of the colony taking part in the rites and in the following sacred meal. Even some of the small communities in the Alpheios valley, next to Olympia – the Amphidoloi, the Akroreioi together with Alasyeis, the Letrinoi, Skillous – and the polis of Elis dedicated a few bronze objects to Zeus, which were used for their sacrifices.49 Together with their pompeia, the Greek theoriai conducted their sacrificial victims to the altar of Zeus. For some sanctuaries we are informed that sometimes the theoriai led the animals on the hoof from their original pasture, but we are not always told about their number or type. According to Plutarch, when Nicias was architheoros to the Delia of 417, he reached the isle of Rheneia (Nic. 3.5).50 For Olympia we have no traces of this use. Of course we may assume that the transport of animals to the sanctuary might have been the case – perhaps partially – for some continental poleis. But it would have been far more practical for the Greek theoriai to buy the animals directly on the site, as the Eleans did. We do indeed have some epigraphical evidence about this procedure: before leaving, the theoriai received from their polis a sum of money specifically issued to buy the animals for their own pompé and sacrifice.51 48

Athen. 11.59: […] , , , , . […] , , , , , . The careful and precise list of the furnishings in the treasuries suggests that Polemon could rely on written inventories of the single items, which should have been kept in the sanctuary’s archive: on this point also Angelucci 2011, 337. On Greek inventories see Scott 2011. 49 Siewert 1991b, 83: “Ob diese Geräte bei den monatlichen Weinspenden an fast allen Götteraltären Olympias eine Rolle spielten, oder bei dem großen penteterischen Zeusopfer oder bei der abschließenden Siegesfeier im Prytaneion, bleibt zu untersuchen.” On these local furnishings cf. further Taita 2007, 46–48. 50 The carriage of mainly small-size animals by boat for sacrificial purpose is attested already by Homer (Il. 1.142–43: ' ; 309–10; 431; 438) and then in the classical age (e.g. Dem. In Meid. 167). For further examples see Rosivach 1994, 158–60, about the proportions of imported victims in Athens in the fifth and fourth century. It was thereby a well-known usage, which had been practiced since the early archaic age without major logistic obstacles. 51 The polis of Cos gave to the theoria to the Delphic Pythia of 278 the sum of 400 drachmas to buy an ox with golden horns (Syll.3 398, ll. 23–24:

126

Chapter Eight

A Cretan decree from the polis of Praisos of the beginning of the third century regulates the juridical relationships with the subordinate community of Stalai. Whenever they travelled outside Crete, and especially for the theoriai to Delphoi and Olympia, Stalai had to cover all the related costs together with Setaia, another subordinate community of Praisos: [ ] , [ ] , [ ] .52 These costs are defined by the generic expression (attic ), “the needed things.”53 We may assume that Praisos’ two dependent communities had also their share in the financial contribution for the sacrifices in Olympia and Delphoi. Therefore, from the considerations outlined, we have solid arguments on the basis of which to accept the view that both the polis of Elis, the organizer of the hecatomb, and the Greek theoriai used to purchase the sacrificial victims – or at least the greater part of them – directly in Olympia or in the Olympia area.54 The archaeological research of B. Eder and H. Baitinger has demonstrated the existence of an agora south of the Olympic Altis, in the area of the Bouleuterion.55 Here, at least in connection with the panegyris every four years, a big market took place,

; ll. 43–45: ). Accepting the newly founded Aetolian Soteria after 240, the polis of Chios established to grant to the theoroi a certain sum of money in Delphoi (Choix Delphes 78, l. 30). 52 Chaniotis 1996, 383–93, no. 64, B1, ll. 12–15. 53 Chaniotis 1996, 391–92. 54 According to the literary sources, very rich and powerful people who had achieved a victory in the games used to offer huge sacrifices to Zeus, inviting all those present to the banquet. However, these rites were performed in a purely euergetic form beside the official Elean hecatomb. Pindar records that Psaumis from Camarina, after his Olympic victory with the mules’ chariot (presumably in 456), / (Ol. 5.5–6: on this passage see Currie 2011, 304–5 and n. 143). When Alcibiades took part in the games of 416, the inhabitants of Chios provided him with (Ps-Andoc. 4.30). The text doesn’t offer any information about the provenance of the sacrificial victims, i.e. whether the Chians had sent him the animals or a money gift to buy them in Olympia. On this episode see Gribble 2012, 58–59, 63–67. For other examples of this kind of sacrifices on private initiative cf.: Taita 2007, 26 n. 30; Currie 2011, 301–6. 55 Baitinger – Eder 2001, 190–5; Baitinger – Eder 2003, 102–4; Kyrieleis 2011, 111–12.

The Great Hecatomb to Zeus Olympios

127

where every kind of ware was sold.56 But this market was not the only one in the sanctuary’s region. Strabo reports that at Alesion, a polis on the mountain road between Elis and Olympia, (8.3.10 = C 341): the inhabitants of the neighbourhood held a market every month. We may assume that there had to be a lot of animal traders too among the many sellers in the markets of Alesion and Olympia, just as in the panegyris described by Pausanias for the festival of Isis at Tithorea, in Phocis: , (10.32.15). Considering the region of Olympia and the Alpheios valley, so rich in herds and flocks of every kind, the number of such traders during the Olympic panegyris must have been very high.57 Unfortunately we are not informed about the details of the purchasing system of the sacrificial victims from these merchants. We don’t know if every trader was authorized to sell the animals for sacrificial purposes, or only the one/ones who had gained the selling right by offering the best price to the Elean administration. Such a procedure is attested by the lex sacra of Andania, in Messenia, regulating the mystery cult of the Great Gods. The initiated awarded the victims’ supply to the best offerer: , , , (IG V, 1, 1390, ll. 66–67: 92/1).58 Thanks to these local markets, the Eleans and the Greek delegations had a ready opportunity to buy the animals for the sacrifices to Zeus. The Olympic hellanodikas or hellanodikai mentioned in IvO 14 had thereby to control the quality of the animals that had been locally purchased. They had to check if the victims corresponded to the defined criteria, in order to perform a truly teleia hecatomb. We are not enlightened about the special regulations used at Olympia. The hellanodikai had probably been taught in this respect by the Elean nomophylakes during their ten-month training in the Hellanodikeon of Elis: 56 Ziehen 1939, 35–36; Sinn 2004, 172; Kyrieleis 2006, 102; Kyrieleis 2011, 111– 14. 57 About animal breeding in Elis see below n. 101. For the common practice of selling and purchasing animals at fairs Epictetus (Diss. 2.14.23) observes: […] . , . On such livestock markets Chandezon 2003, 281–82, remarks on their importance especially for some regions: “[…] dans des régions plus isolées, les foires de panégyrie pouveint probablement servir de cadre à ces échanges.” 58 Cf. also Taita 2014, 118–21.

128

Chapter Eight

< > (Paus. 5.24.3). Among the many rules on the competitions, , which the hellanodikai had learnt, there were certainly ones about the Olympic hecatomb and the choice of animals which were suitable, i.e. teleioi, for the sacrifice. As V. J. Rosivach remarks, according to different literary sources, “an animal whose deciduous teeth have all been replaced with permanent ones is said to be ” – but this adjective usually indicates only the age of the animals and not their ritual purity.59 As cattle and sheep were normally considered teleioi at the age of 3.5 to 4 years, but their productive life, in terms of offspring and milk, had begun about two years earlier, we might surmise that they had been in service for two or more years before being sacrificed.60 We might thereby assume that the hellanodikai had learnt how to determine an animal’s age by checking its teeth. During the official procedure of the dokimasia they had to reject exemplars that were too young, and perhaps likewise ones which were too old. In this perspective we might wonder whether the first word-fragment in line 3 of the inscription, [--] , might be integrated as the comparative form [ ] (= attic [ ] ).61 It could perhaps be an allusion not to “older people,” but to “older,” i.e. “rather old,” animals, in an absolute meaning: maybe, that because of their advanced age, they could not be accepted for sacrifice to Zeus.62 Beside the age, the hellanodikai had to ascertain the physical integrity of the animals, i.e. whether they were not only fully grown – and probably not too old – but also physically unblemished. Finally, as we know from the dokimasia procedure of other

59 Rosivach 1994, 91, 148–53, part. 150–51. On the checking of the teeth see further: Hermary 2004, 99; Feyel 2006, 44–45. 60 One of the main reasons for choosing adult victims may be represented by the need to maintain a relatively constant size of the herds by culling excess animals, because of the limited land available for grazing, as Rosivach 1994, 144, points out: “[…] the animals, both young and old, which were sold for sacrifice were […] animals which would have to be culled in any case.” See also ibid., 89–93, 151, and further Jameson 1988, 97–103, part. 103: “[…] the sacrificial calendar was close to that of seasonal availability from the annual increase of younger animals and the culling of older ones […],” and 100: “Unwanted adult animals are usually culled from the flock in the summer,” when indeed the Olympic hecatomb took place. On the choice of adult animals see also Hermary 2004, 100–101. 61 This restoration, suggested by Blass 1884, 328, was accepted by Dittenberger 1896, 37, and Minon 2007, 128. 62 Rosivach 1994, 153, remarks on the lower value – and consequently lower price – of older animals as sacrificial victims because of their lower potential as milkand wool-producers. Cf. also Hermary 2004, 100.

The Great Hecatomb to Zeus Olympios

129

sanctuaries, they needed to check that the victims fulfilled other possible requirements concerning type, gender, and colour.63

5. The preserved part of the inscription IvO 14 does not give any concrete details about the performance of the hecatomb, neither the number or the species of the animals involved, nor the course of the ceremony itself. As already remarked, the ritual of the hecatomb is not even mentioned in other Olympic inscriptions. However, we can consider another particular kind of evidence, the osteological remains, to assess if they confirm or even amplify the picture of the big bovine sacrifices we find in the literary sources. Despite the legendary proportions of the rites, we cannot rely on large quantity of animal bones in Olympia,64 unlike the finds in some other Greek sanctuaries. The zooarchaeological remains were noted already during the first phase of the Olympia excavations at the end of the nineteenth century.65 But only it is in recent years, in connection with the archaeological research carried out by H. Kyrieleis between 1987 and 1996 in the Pelopion area, that these few traces have been scientifically examined. They were found in the so-called “schwarze Schicht,” a layer of black ashes spread between the Pelopion and the Heraion.66 The layer is made of dark earth, mixed with black combustion ashes, charcoal and very small pieces of burnt animal bones.67 Furthermore, it contains hundreds of terracotta and bronze Protogeometric to Late Geometric vessels and votive figurines. This layer thereby represents the result of early cult activities, i.e. the leftovers of a deposit both sacrificial and votive. It was levelled a 63

About the criteria defining a teleia victim see above, n. 26. It is very likely that in Olympia too, as in other sanctuaries, the examined and approved animals received a mark to allow an easy identification, as the lex sacra of Andania expressly provides: , (IG V 1, 1390, ll. 70–71). 64 Beside the altar of Zeus, other osteological leftovers have been found in the altar’s sacrificial deposit layers of the Late Archaic Artemision, in the south-east area of the sanctuary: Benecke 2006b; Heiden 2012, 146; Senff 2012b, 200. 65 Furtwängler 1897, 4: “Der Boden war mit Knochen und Kohlenresten durchsetzt,” but without any remark about the bones; Kyrieleis 2006, 27–33, with excerpta from the excavation diaries of the “Alte Grabung.” 66 Kyrieleis 2006, 33, with Beilage 12. 67 Mallwitz 1972, 85; Mallwitz 1988, 81–85, 102; Mallwitz – Herrmann 1999, 221–22; Kyrieleis 2002, 216–19; Kyrieleis 2006, 16–17, 22–23, 27, 35–36, 42–47; Heilmeyer 2012, 115–17.

130

Chapter Eight

few times and has to be connected with an altar, very probably the altar of Zeus, the main divinity in Olympia.68 Considering the stylistic features of the clay and bronze dedications found in it, the end phase of the black layer seems to date back to the late seventh century.69 This altar of Zeus was a “Brandopferaltar,” which must have been in use for more than four centuries, from the late eleventh to the end of the seventh century, and must be placed south of the later Heraion and in proximity to the north side of the later Pelopion.70 So it was located in a more westward position than the one described by Pausanias in the second century AD, which was […] , (5.13.8), “about equally distant from the Pelopion and the sanctuary of Hera, but in front of both.”71 Therefore, it has to be assumed that there have been different cult places for Zeus and his hecatomb, in different ages, inside the Olympic Altis. From this old altar only the few material traces mentioned above are left.72 The bones of two excavated sections in the Pelopion area have been thoroughly examined by N. Benecke, but only for one of them can the animal species still be recognized, and there only partially.73 About 800 pieces have been found here, mostly fragments of mammals’ bones from every part of the skeleton, as well as teeth. Because of the tiny dimensions 68

Mallwitz 1988, 85; Kyrieleis 2002, 218–19; Kyrieleis 2006, 35–36, 41. The archaeological evidence shows that two distinct and overlapping black layers are actually recognizable; the older one might date back to the first half of the seventh century: Mallwitz 1988, 85–86; Kyrieleis 2006, 33–36, 47. For this reason it seems appropriate to assume that the black layer doesn’t represent the remains of one and only removal operation of the sacrificial deposit, but the leftovers of subsequent levelling and spreading actions which might have occurred during the seventh century. On the dating of the black layer see also: Kyrieleis 2002, 216–19; Kyrieleis 2006, 47–55; Moustaka 2013, 120. 70 Kyrieleis 2006, 39–42, 46–47, 50. 71 For discussion of the exact position of the altar mentioned by Pausanias in the Altis cf.: Trendelenburg 1902, 7–24; Schleif 1934, 140–41. 72 Kyrieleis 2006, 46 n. 158, observes that also “viele kleine Splitter aus Kalkstein” have been found among the sieved materials from the black layer. These small fragments have been interpreted as remains of the altar’s stones, which had broken off because of the high temperatures during the victims’ burning and had heaped up, together with the ashes, at the end of the sacrifice. See also ibid., 41: “Es spricht […] manches dafür, in den Steinsetzungen am Nordrand des Pelopion Reste des Altares zu erkennen […]” and 42. 73 The study has been conducted on remains found in the so-called “Fläche P 16” (under the Pelopion’s enclosure wall) and “Fläche P 28” (“unter Stein”): Kyrieleis 2006, 16–17, 22–25, 43–44; Benecke 2006a. The species can be determined only for the fragments of bones in the “Fläche P 16.” 69

The Great Hecatomb to Zeus Olympios

131

of the finds, the species has been ascertained only for 172 pieces: 55% of the findings belong to sheep and goats, 34% to oxen, and 9% to pigs.74 The age of the animals can be determined only in a few cases: the bones never belong to young specimens, but always to fully grown animals, whose physical development had been completed, and this is also the case in regard to the teeth.75 As we have seen, only as adults or at least subadults who had reached bodily maturity, could they be considered as teleioi, “complete,” and so perfectly suitable for a teleia hecatomb. In this respect the osteological remains from the Protogeometric and Geometric age seem to confirm the content of the later lex sacra IvO 14 concerning the probable criteria for choosing adult victims. According to Benecke, a lot of the bones show traces of combustion, to various degrees, but none of them any mark of butchery operations.76 From the analysis of the finds in both portions of the Pelopion area it might be thereby assumed that, at the time of the older altar of Zeus, the whole body of the animal – or, at least, of a certain number of animals – was burnt.77 Moreover, as most of the bones and teeth show a very high level of fragmentation, it can be supposed that the original Olympic ritual consisted of a sort of combustion sacrifice, followed by the sacrificial meal and also by the intentional crushing of the skeletal remains, which were then deposited together with the leftovers of the cultic activity, including the votive objects.78 Thanks to the zooarchaeological evidence, we can thereby reconstruct a picture of the hecatomb’s older form where bovine animals certainly played an important role, as in the rites from the classical age onwards, according to the literary sources. But in the earliest centuries of the Olympic cult, sheep and goats presumably represented the largest part of the victims in the 74

In the “Fläche P 16” a deer’s bone and two fragments of a cockle (Cerastoderma edule) have also been discovered: Benecke 2006a, 247. On the zooarchaeological finds in Olympia cf. further Kyrieleis 2002, 219. 75 Benecke 2006a, 247: “Knochen von Tieren im ersten Lebensjahr fehlen sowohl unter den Funden von Schaf/Ziege also auch unter den Funden von Rind und Schwein.” 76 Benecke 2006a, 247: “[…] keine Zerlegungsspuren, wie Schnitt- und Hackspuren […].” 77 Benecke 2006a, 247: “Das Tieropfer selbst bestand im Verbrennen wohl vollständiger Tierkörper […],” and 248. Cf. also Kyrieleis 2006, 16, 42–43. On the meaning of holocaust rituals, which mostly concerned small victims, see Ekroth 2008a, 89–93. She also remarks (90) that a holocaust could often open a traditional thysia sacrifice of a larger animal. This assumption raises the question whether holocausts, in the form of purely opening rites, might have occurred in Olympia too in the Early Iron Age. 78 Benecke 2006a, 248; Kyrieleis 2006, 16–17, 43–44, 47.

132

Chapter Eight

hecatomb, contrary to the testimony of the authors from later times, who stress the almost exclusive bovine component of the big sacrifice. The mixed composition of the Olympic hecatombs in the Early Iron Age recalls other sacrifices, also called hecatombs,79 described by Homer: to Apollo ' (Il. 1.315–16), where the rites are designated as , i.e. ;80 or (Od. 1.25). The most remarkable difference between the older ritual and the later one seems to be the treatment of the skeleton bones. In his description of the altar of Zeus, Pausanias reports: , , · < > (5.13.9). The verbal form , “it is in the custom,” reflects a ritual habit which, at Pausanias’ time, would have been in use for a long period and consisted in the burning of the thigh bones alone, after the meat had been removed. The terminus postquem for this ritual procedure would be later than the end of the seventh century or the very beginning of the sixth, when the original altar, where the total combustion took place, was levelled for the last time. The discrepancies between the features of the Olympic hecatomb rituals in different ages are very instructive. As it has been shown for other archaeological sites in Greece, the zooarchaeological evidence does not contradict the information offered by inscriptions and literary sources; the pictures may be different where they refer to different times, simply because the “ritual practises were neither static nor linear.”81 The very first sacred centre of the Olympic sanctuary was moved eastwards inside the Altis around 600, but probably not immediately to the spot we know from Pausanias. The results of the recent archaeological excavations by A. Moustaka and K. Herrmann in the area of the Heraion’s altar, east of the temple of Hera, show an interesting picture. Both the temple and some special joint slabs in the altar’s foundation – the so-called

79

For the detailed description of a homeric hecatomb see Il. 2.430–68. Cf. also Hermary 2004, 110–11. 80 “Teleiai hecatombs” are mentioned also in Hom. Il. 2.306 (to the immortal gods); Od. 4.582; 13.350 (to the Nymphs); 17.50 and 59 (to all gods); stressing the need to respect the precise ritual of a “teleia hecatomb”: · ' (Od. 4.352– 53). 81 Trantalidou 2013, 61 and 80. Ekroth 2009, 126, further remarks that “[…] the written and iconographical sources present us with the ritual, as the Greeks desired it to be, while the bones can show us how it actually was.”

The Great Hecatomb to Zeus Olympios

133

“Falzplatten”82 – date back to the years between the end of the seventh and the beginning of the sixth century.83 This is the same period when the older altar was levelled for the last time, with the production of the most recent phase of the black layer. Therefore, Moustaka, Herrmann, and also H. Kyrieleis assume that the first altar of Zeus would have been moved eastwards from its original position on the north side of the Pelopion in order to clear space, which was destined for the construction of the first stone temple in Olympia, the Heraion.84 Some archaeological elements indeed suggest that a very well-organized place for sacrifices was set up east of this temple. The Falzplatten formed part of a flooring that achieved an almost waterproof effect: the floor had to resist heavy and frequent scrubbing or intensive washing, probably to remove the bloody remains of the slaughter operations which were performed on that spot.85 Some of the slabs display characteristic holes: they have been connected with the need to tie down the animals to posts before their slaughtering.86 Equipment with a similar aim is attested for the later age too on the southern and eastern steps of the Heraion: these fifteen holes are called “Seilöhre” and apparently do not have any logical building purpose.87 Some stone bases with a singular iron ring on their upper surface have been found south of the south-eastern corner of the Heraion; they were probably put in a row along the south side of the temple, about 8 metres away from it, with the same function as the Seilöhre, i.e. to tie the future victims.88 Considering 82

There are more than 60 pieces: Herrmann 2013b, 106, with Abb. 4, 107–14; Moustaka 2013, 117–19. These are already mentioned by Mallwitz 1972, 138, and Mallwitz 1988, 86. 83 Mallwitz – Herrmann 1999, 220, 222; Kyrieleis 2006, 50–51; Herrmann 2013b, 114; Moustaka 2013, 115–20. 84 Kyrieleis 2006, 50–55; Herrmann 2013b, 114; Moustaka 2013, 120–21. 85 Herrmann 2013b, 110; Moustaka 2013, 120. See already Mallwitz 1972, 88: “Vielleicht haben wir die Sohle eines Beckens vor uns, die auf diese Weise abgedichtet wurde.” 86 Herrmann 2013b, 110–11. 87 Schleif 1934, 142; Herrmann 2013b, 110, 111, with Abb. 6, 114 (earlier than 600); Moustaka 2013, 120. Cf. also Bocher 2012, 142. 88 Herrmann 2013b, 111–13, with Abb. 7–8. Considering some archaeological factors, Herrmann believes that the practice of tying the animals to a ring driven into a stone came later than the habit of tying them to a post in the steps of the Heraion and in the slabs of the altar’s foundation. However, these two methods, according to this scholar, might even have been followed at the same time, at least for a certain period. Cf. further Moustaka 2013, 120, and already Schleif 1934, 142. Similar stones with rings attached to blocks before the altar have also been found in the sanctuary of Zeus Olympios in Dion (Hermary 2004, 88; Herrmann

134

Chapter Eight

the outlined archaeological evidence, it may be assumed that, from the beginning of the sixth century, there would have been a specific place east of the Heraion with a sacrificial function. However, no bone remains have been found up to now in this part of the Altis. According to Kyrieleis, Moustaka, and Herrmann, this new sacrificial spot could be considered the successor of the older altar in the area of Pelopion. It might have been in use for about a century, until the beginning of the big building project of the Eleans in the Altis, which started from the second quarter of the fifth century.89 The focus of this project was of course the new temple. In connection with it – as suggested by the mentioned scholars – the new altar for Zeus was also constructed, with its architectonically well-defined structure as described by Pausanias.90 Unfortunately we do not have any archaeological or osteological leftovers that can confidently be associated with it.91 Therefore, it might even be supposed that the prothysis of this altar had been set up earlier than in the second quarter of the fifth century, i.e. around the beginning of the sixth, when the older altar was definitively levelled. Since the space on the stone basement was too narrow to perform a big hecatomb, it should be indeed assumed that a place next to the altar was needed to carry out the proper operations of animal slaughter.92 The described area of the altar’s 2013b, 113) and in the shrine of Apollo in Claros, where the hundred blocks discovered date back to the end of the Hellenistic age (Hermary 2004, 111: “[…] illustrent la pratique de grands sacrifices simultanés de bovins, qui pouvaient effectivement concerner une centaine d’animaux,” 134; Herrmann 2013b, 113). 89 Mallwitz – Herrmann 1999, 240–45, part. 242; Moustaka 2013, 120–21. 90 Mallwitz 1972, 84. According to Moustaka 2013, 121, this altar should be placed to the west of the spot generally accepted on the basis of Pausanias’ description. 91 See above n. 9. Although we don’t have any bone material deriving from the activity at the altar (Heilmeyer 2012, 117: “[…] die gewaltige Menge an Asche des von ihm [Pausanias] detailliert beschriebenen großen Zeus-Altars ist im Gegensatz zu den frühen Aufschüttungen ganz verschwunden”; for the possible reasons for the complete disappearance of any trace of the ashen material see: Yavis 1949, 213; Heilmeyer 2012, 115; Senff 2012a, 11; Moustaka 2013, 119), the slaughtering of bovines is nonetheless documented by some butchering refuse which has been discovered in the water wells. Baitinger 2001, 89, records the finding of three bull skulls that still presented traces of the spear which transfixed them. Gauer 2012b, 99, gives an interesting and unfortunately only autoptical proof about an ox’s tongue: “In einem Brunnen, der durch eine Tonschicht luftdicht abgeschlossen war, fand sich, wie frisch abgeschnitten, eine Rinderzunge. Sie ist leider undokumentiert in den Händen und vor den Augen der Ausgräber zerfallen.” 92 This lack of space was stressed already by Weniger 1913, 248: “Man war daher genötigt, die vorliegende geräumige Fläche der Altis mit zu verwerten, welche offenbar schon von Anfang an, […], dazu benutz wurde. Man hatte sie recht

The Great Hecatomb to Zeus Olympios

135

foundation east of the Heraion might thereby be interpreted as a kind of service area for the big altar of Zeus: here the victims were kept before being conducted to the altar and here their slaughtering had to be completed, once it had probably begun on the prothysis in connexion with the first phases of the rite.

6. The performance of a teleia hecatomb became a superb spectacle, with people coming to Olympia not only to watch the games but also to attend such a great sacrificial ceremony with its huge number of victims. The act of seeing ( ) the actions conducted on the big altar and in its proximity from the whole surrounding area appears evident in the technical meaning of the word in Xenophon (Hell. 7.4.31). As has been archaeologically proved, the Olympic “theatre” surrounded the “Festplatz,” i.e. that part of the Olympic Altis east and north of the big bomos. This was delimited, on the east, by the western embankment of the stadium and, from the fourth century, by the Echohalle; on the north, by the slopes of the treasures’ terrace.93 Such a large space was required by the high number of people and animals entering the Olympic Altis from its south-east corner, along the older “Feststrasse.” This processional way was 63 metres wide between the Bouleuterion and the Hestia-Halle, so that it could be accessible for carts, and, in the entrance area to the Altis, it

eigentlich zu diesem Zwecke freigelassen.” According to his calculations, 253–54, the length of the prothysis was enough to slaughter only three oxen at once. See further: Schleif 1934, 142: “Nicht die Schlachtung, sondern nur die eigentliche Opferzeremonie – Segnung, Zerteilung, Festgesänge – fand auf der Prothysis statt”; ahin 1972, 35: “Die kleinen Tiere, wie Schafe, dürfte man aber auf der Schulter auf die Prothysis getragen und dort geschlachtet haben. Die Rinder müssen jedoch irgendwo unten, auf dem Erdboden, geschlachtet worden sein.” 93 Schleif 1934, 156; Mallwitz 1981, 103–6: “[…] die Altis im frühen 4. Jahrhundert sowohl im Norden als auch im Osten von einer abgetreppten Stützmauer begrenzt war, die zu Zeiten des Zeusfestes als Zuschauerraum – Theatron – diente […]”; Mallwitz 1988, 103; Sinn 1996, 54–57; Mallwitz – Herrmann 1999, 256 with n. 288; Hölscher 2002, 336, 338. Already Adler 1892, 213–14, observed: “Keine andere Platzlage wie diese gestattet eine so günstige Unterbringung einer sehr grossen Menschenmenge auf drei Seiten, im Norden, Osten und Süden, wobei die Schatzhäuserterrasse, die Echohalle und die Nord- wie Osthalle des Zeustempels sowohl wegen ihrer erhöhten Fussbodenlage als auch ihrer Schattenspendung halber ganz besonders zweckmässig ausgenutzt werden konnten.”

136

Chapter Eight

always maintained a width of 40 metres, to ensure easy access.94 In view of the fact that a lot of animals entered the Altis, where monuments stood and precious votive objects were exposed, it has to be assumed that the Elean administration had taken measures to prevent possible damage from the animals. In this perspective we might thereby interpret the mention of the singular figure of the , the shepherd, in a lex sacra which goes back to 500 and grants a few privileges to the Greeks from Epidamnos, Libya and Crete visiting Olympia: [---- ] , < > [ ’ ----] (NIO 4 = IED 8, l. 7). Because of the fragmentary context, we cannot define the details of this provision, ascertaining if the nomeus had to look after the animals for the sacrifice or the horses for the competitions or, again, the draught animals of the pilgrims or, perhaps, all three categories together. However, the reference to the agalmata, which were mostly concentrated inside the Altis, makes it likely that he had to watch over the victims for the rite, especially – but perhaps not only – on the occasion of the big hecatomb. In case of misconduct, the nomeus had to pay a fine of five drachmas to Zeus.95 In conclusion, in regard to the questions asked above, we have discussed different aspects of the grand Olympic hecatomb: the place where the ritual was carried out, the big ashes altar which was displaced inside the Altis at least twice during the long history of the sanctuary; the species of animals preferably offered to Zeus, mostly bovines but also sheep, goats, and pigs; the criteria for the choice of the suitable animals, teleioi because fully grown and unblemished; the way the victims were provided and paid by the Eleans, mainly through purchasing them on the local market by funds from the agricultural tenths or by donations from the Greek theoriai. But one point still remains to be clarified, i.e. the real proportions of the rite. If we assume that the quantity of sacrificial meat to be distributed 94

Mallwitz – Herrmann 1999, 256, 258, 263. A water well has been found before the Bouleuterion: ibid., 262–63, with Abb. 165. It was covered by a kind of roof and had a long life in comparison with many other wells in Olympia. Because of its key position along the processional way, this well always maintained a very important function, i.e. to let the people clean and purify themselves before entering the Altis. 95 Inscriptions from the second half of the fourth century attest a similar practice in Delphoi. Some accounts of the sanctuary indeed mention the salary paid to , who had to watch over the animals in the days between their purchase and their sacrifice: CID II, 34, I, 30–31, II, 33–34, 49–51; CID 46, 4–5; CID 50, II, 3–4; CID 55, 13–15. Since the fee for such tasks is generally very low (on average 1.5 obolos), it is likely that the relatively high sum of five drachmas in NIO 4 has to be interpreted not as a salary but as a fine.

The Great Hecatomb to Zeus Olympios

137

should have been able to fulfil the nourishment needs of all the visitors, we may try to estimate how many animals were required for the average number of pilgrims in a panegyris. In the heyday of the games, i.e. from the last quarter of the sixth to the first half of the fifth century,96 the number of visitors in Olympia might have been around 24,000 units, increasing during the second half of the fifth and the fourth century to the supposed figure of 40,000 to 50,000 people.97 For the fifth century we might thereby consider an average of about 35,000 people. F. S. Naiden has recently offered interesting reflections on the number of persons partaking in sacrificial meals and the quantity of meat needed to feed the whole sacrificial community. Since a bovine can provide an average quantity of about 80 kg of meat and each person can eat an average portion of about a kilogram,98 we might calculate that about 440 bovines were needed for about 35,000 people. Of course, we could also consider a lower estimate of meat per person, about half a kilogram; in this case about 220 bovines were required. Although such calculations are purely theoretical, they are nonetheless instructive, as they can give an approximate idea of the quantity of animals – supposing that in the hecatomb only bovines were slaughtered – which were demanded once every four years to satisfy the supplying needs of the large Olympic crowd. Therefore, it may be accepted that the hecatomb of every panegyris really did attain enormous proportions, with at least 200 bovine victims.99 96

See above n. 17. Such a flow of people has been calculated considering the capacity of the Late Archaic Stadium (phase II of the late sixth century) and of the Late Classical one (phase III of the middle of the fourth century, whose reconstruction we see nowadays: Schilbach 2012, 282): Sinn 2004, 101; Nielsen 2007, 55–57; Kyrieleis 2011, 111. See also Adler 1892, 214. However, Schilbach (ibid.) has recently remarked that the chronology of the Stadion II and III should be examined once again: “[…] die offenbar falsche Datierung von Stadion III, das um rund hundert Jahre früher, um 460/50 v. Chr. anzusetzen ist. Damit musste auch die Bauzeit von Stadion II überprüpft warden.” 98 Naiden 2012, 57–66. Cf. Ekroth 2008b, 270–72, for other estimates of a meat portion in various Greek sacrifices, part. 271, where she remarks that a kilogram portion should have been destined to a whole family. For the average quantity of meat derived from an animal in Greek times see further Rosivach 1994, 157–58. 99 In view of all the above considerations, there is nothing to support the sceptical interpretation of Trendelenburg 1902, 33–34: “[…] so wäre doch die Meinung, als müsste der Zeusaltar Raum für das gleichzeitige Schlachten oder gar Braten von hundert Stieren geboten haben, sicherlich irrig.” To compare this Olympic number with the figures of animals sacrificed in various festivals at Athens and in the Attic demes see Rosivach 1994, 68–78. Such huge sacrifices, with a number of victims 97

138

Chapter Eight

For this reason it is easy to understand how massive the problem of providing food, i.e. mainly meat, would have been for such huge and periodic gatherings. As J. M. Renfrew remarks, “the festival put a severe pressure on the local resources, and it was perhaps just as well that it only lasted for five days and that it only occurred once every four years.”100 That’s why the local market which took place during the Olympic panegyris had the purpose of providing a large offering and choice in terms of sacrificial animals for any purchaser, whether the polis of Elis, or the theoriai of the Greek poleis, or private individuals. The whole region of the Olympia sanctuary, north and south of the Alpheios and in the valley of the koile Elis, was very rich in flocks and herds of every kind, as attested by many literary sources as well as the archaeological evidence, and especially by the big quantity of small bronze and clay votive reproductions of bulls, oxen, sheep, goats, and also horses.101 The livestock represented the basis of the economic resources and of wealth in Elis. So, although the pressure would have been severe, as Renfrew stresses, it might not have been so extremely difficult to supply an adequate number of victims for the great hecatomb, at least once every four years. The economic system of a Panhellenic sanctuary like Olympia, as J. McInerney points out, was indeed subject to “enormous fluctuations,” according to the religious calendars and to the festivals, with the related prescribed animal sacrifices.102 Furthermore, as the Greeks consumed meat primarily – if not exclusively – in a sanctuary and in the framework of an official ceremony, the performance of such a grand hecatomb was a key event during the Olympic panegyris for two reasons. First, because of the religious meaning of the moment, which represented an act of devotion and celebration for the main divinity of the sanctuary, Zeus. Secondly, because of the manifold economic, alimentary, and social aspects that were closely intertwined with the cultic action. In the above reflections we have tried to illuminate some of these points, such as the type, quality, and presumable number of the victims, and the funding of their purchase at the far above a hundred heads, are anyway testified in some other cases, as for the annual festival in honour of Zeus Eleutherios at Syracuse (Diod. Sic. 11.72.2: , ) or for the great sacrifice at the Pythian games of 370 organised by Jason of Pherae (Xen. Hell. 6.4.29: , ). Renfrew 1988, 180. 101 For a review of all sources see Taita 2007, 24–30, 83–89. 102 McInerney 2010, 149. 100

The Great Hecatomb to Zeus Olympios

139

local market. But many questions connected with the performance of the hecatomb and the following sacred meal for all the pilgrims have yet to be thoroughly analysed: the Olympic cultic officials and servants who carried out the entire rite; the individual phases of the complex sacrifice ceremony, from the slaughtering of the animals to the removing of their skins and entrails and the division of their meat; and finally the distribution and consumption of the sacrificial meat, which could be cooked, boiled, or grilled, as the many bronze caldrons and iron spits found in the sanctuary testify.103

103

For the spits see e.g. Sinn 2004, 120, Abb. 39.

CHAPTER NINE THE JUSTICE OF ZEUS AT OLYMPIA JIM ROY1

Abstract This paper examines early inscriptions from Eleia posted at Olympia that refer to breaches of treaties, laws, or regulations punishable by penalties in some way concerning Zeus (fines payable to Zeus, banishment from the altar, etc.). The paper considers how these penalties were administered by the Eleians who controlled the sanctuary at Olympia and were therefore the agents of Zeus, and also how local communities viewed the justice of Zeus.

Excavation at Olympia has produced a remarkable number of Eleian inscriptions of the sixth and earlier fifth centuries B.C., a body of epigraphic material unparalleled at this date in the Greek world. Many of these inscriptions – treaties, laws, regulations, decrees, and legal decisions – deal with how men must behave and what they must or must not do, and also with the penalties for not behaving properly. In many cases these penalties involve Zeus Olympios, the principal deity of Olympia, and this paper will consider the justice of Zeus as it is seen in such inscriptions, and also how the justice of Zeus was administered in practice. The term Elis will be used for the state, or polis, Elis. Eleia will be used for area of the western Peloponnese that Elis sought to dominate, although it did not achieve lasting control until the Hellenistic period. The inscriptions considered are mainly from Eleia, and in Eleian dialect. Greek states from outside Eleia could of course be allowed to post inscriptions in the sanctuary at Olympia, and the earliest surviving examples from such states date from the late archaic and classical periods; they too will be considered, but only briefly since they are significantly different from the Eleian inscriptions and are much less directly concerned with Zeus 1

I am very grateful to Esther Eidinow for valuable information and advice on Greek curse tablets.

The Justice of Zeus at Olympia

141

Olympios. The inscriptions found at Olympia are now available in two corpora: Inschriften von Olympia by Dittenberger and Purgold, published in 1896, and Neue Inschriften von Olympia, by Siewert and Taeuber, published in 2013. There is also a valuable study of inscriptions in Eleian dialect by Minon, Les inscriptions éléennes dialectales (VIe-IIe siècle avant J.-C.), in two volumes, published in 2007. Few Eleian inscriptions from anywhere other than Olympia have so far been published (Siewert 1991a), although the Berlin Academy has now decided to produce the volume on Eleian inscriptions that has so far been missing from Inscriptiones Graecae. The inscriptions will be cited in this paper from IvO, NIO, and Minon, although many have also been edited and discussed in other publications. Elis was evidently in control of the sanctuary at Olympia by the middle of the 6th century B.C., the period when Eleian inscriptions begin to appear at Olympia. From then on ancient writers consistently represent the Eleians as running the sanctuary and its cult, including the Olympic Games. Herodotus (2.160) reports an Eleian embassy to the Egyptian Pharaoh Psammis and a discussion on how the Eleians administered the Olympic Games : the embassy may well be unhistorical, but Herodotus’ presentation of an Olympia controlled by Elis is clear. When Thucydides reports the quarrel in 421-0 between Elis and Sparta over Lepreon (5.31.15, 34.1, 49.1-50.4), he shows the Eleians as responsible for fixing the date of the Olympic truce and making the date known to the Greek world, and he also shows the Eleians administering the penalty that was imposed on Sparta under Olympic law. (The dispute over Lepreon will be mentioned again later.) Xenophon, when reporting the outcome of the war between Elis and Sparta at the end of the fifth century, says (Hell. 3.2.31) that the Spartans did not deprive Elis of control of the sanctuary of Zeus Olympios. It therefore appears that Elis ran the sanctuary at least from the mid-sixth century onwards (save for a brief Pisatan interlude in 365363/2), and it follows that the Eleians will have introduced any new regulations for the sanctuary, its cult, and the Games, and also that the Eleians will have dealt with any breach of those regulations. Eight texts survive that specify penalties for improper behaviour.2 The penalties generally take the form of the payment of a sum of money, expressed in drachmai, but in the sacred law on profanation (Minon 4 = IvO 7) the requirement is payment of an ox and also ritual purification. In 2

The eight inscriptions are: Minon 3 = IvO 5 (c.525/500 ?); Minon 4 = IvO 7 (c.525/500); Minon 5 = NIO 2 (c.525/500); Minon 6 = IvO 1 (c.525/500); Minon 7 = IvO 8 (c.500/475); Minon 8 = NIO 4 (c.500); Minon 9 = IvO 4 (c.500/475); Minon 13 = IvO 3 (c.375).

142

Chapter Nine

several cases, but not in all, the text says that payment is to be made to Zeus Olympios.3 It is not clear why some texts specify payment to the god, but not others. Probably all penalties for improper behaviour relating to the sanctuary would be payable to the god, and that fact would be understood even if it was not explicitly stated. The respective roles of the god and of the Eleians as seen in these texts are not surprising; they are in fact the typical roles of a god and his agents in a Greek sanctuary. That the Eleians were the agents of Zeus Olympios is very clear. This can be seen e.g. in Minon 6 = IvO 1, where there is reference to Zeus, and references to the iaromaos and the theokolos, both Eleian religious officials. Minon 4 = IvO 7 has no reference to Zeus in relation to punishments, but it is specified that, if there is profanation and the judgment given is not according to the graphos, then the judgment is void and the damosia wratra is final: also, the law may be changed to something that is better ‘before the god’ (po’ ton thon) with the agreement of the council of 500 complete and the damos plethuon (see Minon (2007) 2.514-516 on this phrase). In this case the council and the damos of Elis have a major role. The Eleians therefore were able to make laws and regulations for the sanctuary, to judge those who breached the rules, and to impose and execute penalties. The authority of Zeus could also be invoked in other ways. For instance, inscriptions published in the sanctuary could be placed under the protection of the god. This was not done for the inscriptions mentioned above, but examples are Minon 16 = NIO 5A, of c.475-450 (Minon) or mid-C5 (NIO), a grant of citizenship and thearodokia by the Eleians, where the pinax is an agalma of Zeus; and Minon 30 = NIO 8, dated c.370/365-c.324 (Minon) or 360-324(?) (NIO), an amnesty for Eleian exiles, which specifies that, if anyone causes the stele to disappear, he should suffer the same penalty as a stealer of offerings. It might of course be supposed that any inscription published in the god’s sanctuary would enjoy the protection of the god. Minon 10 = IvO 9, of c. 500/475, is however a special case; it is an alliance between Elis and the Ewaioi, and it prescribes that any private person, official, or damos damaging the inscription shall be liable for the sacred fine (epiaron), i.e. that they should suffer the same penalty as for a breach of the alliance, namely the payment of one talent of silver to Zeus Olympios. This very heavy penalty for damage to the inscription is presumably because for the Greeks destruction of the text recording a treaty was a symbolic gesture amounting to the renunciation of the treaty (Alonso Troncoso 2012: 4).

3

On payments to Zeus Olympios, including fines, see Taita 2014.

The Justice of Zeus at Olympia

143

The discussion so far has been about the management of the sanctuary and the games held there, and the punishment of those who misbehaved in the sanctuary. It was natural that the god be involved in the management of his sanctuary, through his human agents, the Eleians. But in other cases the god was involved in matters that went beyond the sanctuary. Minon 19 =IvO 3, of c.475/450, records the condemnation of Timokrates son of Malex for sequestration of a thearos (or possibly more than one). He is to pay the penalty (ta dikaia), namely two draught animals valued at 2 talents and the animals’ harness valued at 30 minai. (The animals were presumably valuable horses: see Minon’s commentary.) In addition he is to be excluded from the altar by the proxenoi and the priest. It was also decided by the damos that two men, thearoi, should sell ... (the remainder is too fragmentary to give clear sense). The priest is presumably the priest of Zeus Olympios, an Eleian, and the damos is clearly the Eleian damos. The further decision (“also”) shows that the case came before the citizens of Elis, presumably in the assembly: it thus shows that the case was dealt with by Elis. And Zeus Olympios was directly involved in the punishment, since Timokrates was excluded from the altar (and consequently from all activities in the sanctuary that required a preliminary sacrifice on the altar). Minon 20 = IvO 2 (c.475/450) is a decision (wratra) by the Eleians concerning Patrias the scribe (gropheus). If anyone should curse Patrias, that person is to be banished as if for cursing an Eleian. If “he who holds the highest office” and the basilaes do not impose the penalties, each of them is to be fined 10 minai sacred to Zeus Olympios. A hellanodikas shall pronounce (on those who wrong Patrias: see Minon’s commentary), and the zamiorgia (i.e. the body of damiorgoi) shall pronounce the other penalties; if it does not pronounce, it shall pay double in the mastraa (meeting of the mastroi). If anyone whips the accused, he shall be liable for a fine of 10 minai, if he acted deliberately; and Patrias shall suffer the same penalties if he wrongs anyone. The pinax is sacred at Olympia. Here clearly Eleian officials are acting (or being punished if they do not act). The text concerns purely Eleian matters, save that the scribe Patrias is clearly not Eleian. Zeus is involved in that fines are to be sacred to him. Minon 30 = NIO 8 (370/365-c.324 (Minon) or 360-324 (?) NIO) is an amnesty for Eleian exiles, with other regulations concerning exiles. It forbids banishment of descendants, male or female, and confiscation of their property: if anyone banishes (descendants) or appropriates their property, he shall be banished before Zeus Olympios as for a crime of blood, and whoever wishes may curse him with impunity. This is clearly

144

Chapter Nine

an Eleian decision on a purely Eleian matter. Zeus is involved in the banishment: see later on the phrase ‘before Zeus’. Minon 22 = IvO 16 (c.450/425) is a decision concerning the Skillountians. (Skillous was a small community just south of the Alpheios, and at the time of the inscription a subordinate ally of Elis.) The inscription is long and very fragmentary, and its meaning is often unclear. However it is in Eleian dialect, and it seems to be a decision taken by Elis concerning Skillous. (See Minon ad loc. for the relevant arguments, and the conclusion that the text emanates from Elis.) There are several references to fines that are to be paid as sacred to Zeus Olympios. Notably this text concerns not only Elis but the separate polis Skillous, admittedly subordinate to Elis. Minon 10 = IvO 9 (c.500/475) is an agreement (wratra) between Elis and Ewa (or Eua), in fact a primarily military alliance for 100 years. If either party breaches the alliance, they are to pay a talent of silver sacred to Zeus Olympios. Ewa is otherwise unknown: it cannot be the community Eua of the Thyreatis in the eastern Peloponnese, since it would make no sense at this period for Elis to form a military alliance with a very small community on the other side of the Peloponnese; Ewa must in fact be an otherwise unknown community of Eleia. What we know of the political situation in Eleia at the time of this text is that Elis was extending its influence in the area by creating subordinate allies (the ‘alliance’ mentioned in Minon 5 = NIO 2), and this text evidently shows the creation of such an alliance. The treaty apparently treats Elis and Ewa as being on an equal footing, but in fact Elis was dominant. Both parties are subject to the authority of Zeus, but it would have fallen to the Eleians to decide whether a breach of the treaty had occurred and, if so, whether a fine was liable to be paid, and then to enforce payment. The power of the Eleians at Olympia thus made the alliance unequal. These last two texts concern Elis’ relations with a community outside the Eleian polis, albeit in both cases a community subordinate to Elis. Here it is worth recalling an incident recorded not on an inscription but by Thucydides (5.31.2). At some time before the beginning of the Peloponnesian War, Lepreon appealed to Elis for an alliance because of a war against unspecified Arkadians. The Eleians granted the alliance on condition that Lepreon cede half its territory to Elis, and the Eleians then rented the land back to the Lepreates for one talent (presumably 1T per annum: Gschnitzer 1958: 15) payable to Zeus Olympios. Why was this money payable to Zeus? The alliance between Elis and Lepreon was a military and political matter rather than a religious one, but Elis apparently decided that Zeus Olympios should be involved, presumably to make

The Justice of Zeus at Olympia

145

alliance more binding. (The Lepreates in fact ceased to pay during the first phase of the Peloponnesian War, and trouble developed: this will be discussed later.) Two texts were published not by the Eleians but by other local communities, whose use of Eleian dialect shows that they belonged to Eleia. The first is Minon 12 = IvO 11 (c.475), a wratra of the Chaladrioi for Deukalion, who receives rights and honours. The Chaladrioi are otherwise unknown. They may have been subordinate allies of Elis, or a community of Eleia not yet linked to Elis by a subordinate alliance. Some scholars have supposed that they were part of the Eleian state itself; if true, that would suggest that Elis had a federal structure, but there is no direct evidence of federation and very little indirect evidence that suggests it. If anyone exercises syla against Deukalion, that person shall be banished before Zeus, unless the damos decides otherwise (see later on the phrase ‘before Zeus’). There is no clear evidence that Elis was directly involved here. The decision is formulated by the Chaladrioi, and the ‘damos’ mentioned is presumably the damos of the Chaladrians. The second text is Minon 14 = IvO 10 (c. 475). It is a treaty (wratra) between the Anaitoi and the Metapioi, forming a friendship for 50 years. If either party does not abide by the treaty, the proxenoi and the mantier (seers) shall exclude them from the altar. If they breach the oath, the iaromaoi at Olympia shall give judgment. Both the Anaitoi and the Metapioi are otherwise unknown. Like the Chaladrioi of Minon 12, they may be subordinate allies of Elis or communities of Eleia not yet directly linked to Elis, or, as some have supposed, part of the Eleian state itself. It has also been suggested that Elis may have imposed this treaty on the Anaitoi and the Metapioi as subordinate allies, but that can only be conjecture (Alonso Troncoso 2012: 7-8). In any case the two communities clearly subjected themselves to the justice administered by Eleian officials at Olympia. Zeus Olympios was involved because the penalty for a breach of the treaty was exclusion from the altar of the god, put into effect by the god’s human agents at Olympia. Alonso Troncoso points out (2012: 7) that this penalty is different from that set out in the wratra between Elis and Ewa, where the penalty for a breach is a fine of one talent of silver, payable to Zeus. As Alonso Troncoso says, Elis would not have accepted the possibility – however unlikely to be put into practice – that it might be excluded from the altar of Zeus at Olympia. However these two texts, the wratra of Elis and Ewa and the wratra of the Anaitoi and the Metapioi, stand out because in these two cases other communities of Eleia, the Ewaioi, the Anaitoi, and the Metapioi, submit to the justice of Zeus as administered by Eleian officials at Olympia.

146

Chapter Nine

However, before the point is examined further, some other points deserve attention. One is banishment ‘before Zeus’. The phrase ‘before Zeus’ or ‘before the god’, expressed by the preposition po(ti) (the equivalent of Attic pros) followed by the accusative, occurs in three Eleian texts, in relation to legislation (Minon 4 = IvO 7, c.525-425), to an oath (Minon 22.11-12 = IvO 16, c.450-425), and to banishment (Minon 12 = IvO 11, c.500-475, decree of the Chaladrians). Latte (1920: 62-68) discussed the Chaladrian text at length, taking po(ti) to mean ‘to’ in a literal sense and arguing that the person so banished would ‘go to Zeus’ (63: “er soll dahinfahren zu Zeus”) and so be excluded from human society and become vogelfrei. However there are arguments against this view. One major argument is the parallel usage found on Greek curse tablets (defixiones), where pros followed by the accusative is used to invoke one or more deities: e.g. IG.III.Pars 3.85, invoking Hermes, and Jordan (1985) no. 42, invoking Hermes, Ge, and Persephone. In these texts, particularly those where more than one deity is invoked, pros can hardly mean ‘to’ in the sense of movement to. The common current opinion was expressed by Faraone: “The most common elaboration of the direct binding formula is the addition of the name(s) of a deity or deities who appear as witnesses or overseers of the act”.4 The use of po(ti) plus Zeus or ‘the god’ in the accusative at Olympia seems parallel to pros and the accusative in curses. However, Eleian usage changed in the 4th century, and po(ti) is found followed by the genitive in Minon 30 (370/365- c.324) = NIO 8 (360324?) to express banishment ‘before Zeus’. (This text was not known when Latte wrote.) Minon translates (p. 199) “qu’il soit (lui-m me) banni pour meurtre devant Zeus l’Olympien”, and in NIO Siewert and Taeuber translate (p.38) “so soll er im Namen des Zeus Olympios verbannt sein wegen Blutschuld”. The same usage with the genitive is found in the Triphylian decree Minon 28 (399-369) in the phrase “he shall be impious before Athena”, referring to the goddess of the temple found at Mazi in Triphylia. Minon argues plausibly (2.435) that the change in usage from accusative to genitive after the preposition po(ti) does not involve any significant change in meaning. (It is however worth noting that po(ti) followed by the genitive cannot mean ‘to’, as Latte’s argument would require.) In view of the evidence it seems safe to accept Minon’s view (2.434) that “le nom du dieu est invoqué comme garantie de la qualité des lois, des serments ou de la sanction que constitue le bannissement”. The 4

Faraone 1991: 5. See also Eidinow 2007: 143-159 on curses, and particularly 151-2 on the invocation of the god, and Kropp 2010: 363-4. The surviving texts of binding curses are presented in Eidinow 2007: 356-459.

The Justice of Zeus at Olympia

147

Eleians, and other communities of Eleia such as the Chaladrians and the Triphylians, appear to have used originally a phrase invoking Zeus Olympios comparable to the invocation of a deity or deities in defixiones, and then to have continued the usage in the fourth century with – for some unexplained reason – a change in the case that followed the preposition. A different point has been raised by Mello (2008). She argues that in these Eleian texts the word wratra refers to an oracular pronouncement, and, further, that the personnel of the oracle at Olympia was separate from the Eleian magistrates so that an Olympic wratra had independent authority. This, if true, would explain e.g. the wratra between Elis and Ewa as a treaty on equal terms between the two communities. However, if one considers the terms of that treaty, it is very difficult to see how the penalties for breach of the treaty would in practice be applied equally to both, and so it is difficult to accept Mello’s arguments. A different point has been made recently by Alonso Troncoso (2012). After examining the surviving treaties from Olympia from the 6th and 5th centuries, he suggests that there are differences between the Eleian treaties and those from elsewhere. One important difference is that the non-Eleian treaties do not refer to Zeus Olympios. In the treaty between Sybaris and the Serdaioi (NIO 6, 530-510 B.C.), for instance, there are named as ‘proxenoi’ Zeus, Apollo, the other gods, and the polis Poseidonia, but Zeus is not given the epiclesis Olympios and there is no mention of any punishment for breaching the treaty. Also the non-Eleian treaties are not military, unlike the treaty between Elis and Ewa. The known non-Eleian treaties are set up by western Greeks, and, strikingly, there is nothing from the Peloponnese outside Eleia: Alonso Troncoso points out, however, that such an observation depends on which texts have happened to survive, and that the treaty between Sparta and Tegea that was set up, as Aristotle says, “on the Alpheios” (Bengtson 1962: no. 112) may well have been at Olympia. Nonetheless it seems clear that invoking the authority of Zeus Olympios to punish breaches of treaties or other texts setting standards of behaviour was a specifically Eleian practice. This brings us back to the two texts that made minor communities of Eleia liable, if they breached a treaty, to a punishment involving the authority of Zeus Olympios. One was the treaty between Elis and Ewa, and the other was the treaty between the Anaitoi and the Metapioi. In both cases, if punishment had been required, it would have been carried out by Eleian officials at Olympia. This suggests that Elis profited from its control of Olympia by using the authority of Zeus to increase Eleian influence over its smaller neighbours in Eleia. It is worth considering alongside these texts Thucydides’ account of the quarrel between Elis and

148

Chapter Nine

Sparta in 420 (Thuc. 5.31.1-5, 34.1, and 49.1-50.2: see Roy 1998). Lepreon had sought alliance with Elis, and, to get it, agreed to cede half its territory to Elis, and Elis then rented the land back to the Lepreates for an annual payment of one talent to Zeus Olympios: during the first phase of the Peloponnesian War Lepreon ceased to pay, and asked Sparta to decide whether it should pay: Sparta decided that Lepreon should be autonomous, and, to protect it, sent a garrison and later increased its strength. Elis claimed that sending the later body of 1,000 hoplites to Lepreon breached the Olympic truce for the Games of 420. Elis then obtained a judgment against Sparta under Olympic law: the Spartans were banned from sacrificing at Olympia or competing in the Games, and were fined 2,000 minai, i.e. 2 minai for each man involved in the breach of the truce according to Olympic law. It is not clear what body gave this judgment against Sparta, but it clearly reached a verdict favourable to Elis, and it appears that the Spartans were not represented at the hearing. Sparta sent ambassadors to Elis to appeal against the verdict, saying that the Olympic truce had not been announced at Sparta when the troops were moved into Lepreon. Elis then proposed two solutions, but both of the Eleian proposals would have involved recognition by the Spartans that they had broken the Olympic truce, and the Spartans rejected both. The matter was then left unresolved until some years later. It is clear that the Eleians were trying to regain control of Lepreon, and it is also clear that both the Spartans and the Eleians believed that the Eleians had the power to resolve the situation that had arisen from the verdict against the Spartans, though they could not agree on what should be done. It therefore appears that the Eleians were using Olympic law, and the authority of Zeus Olympios, as an instrument in order to achieve their political goal of recovering Lepreon. It is therefore not surprising that Eleian control of Olympia was also used on other occasions to strengthen Eleian influence in Eleia, as for instance in the alliance between Elis and Ewa. Likewise Eleian power at Olympia lies behind the relations between the Anaitoi and the Metapioi seen in the treaty between them. To conclude. The justice of Zeus at Olympia was put into effect by the god’s human agents, the Eleians. To a considerable extent such activity concerned the suppression of improper behaviour in the sanctuary, and was entirely normal in the Greek world. Sometimes, however, the authority of Zeus Olympios was extended to other matters, and in such a way as to increase Eleian influence over Elis’ smaller neighbours in Eleia. It is clear that by the outbreak of the Peloponnesian War Elis had built up its power in the western Peloponnese by bringing all of Eleia that did not belong directly to the polis Elis into a hegemonial alliance. There is no

The Justice of Zeus at Olympia

149

evidence that other Greek states tried to interfere with this expansion until Sparta took an interest in Lepreon in the late 420s, and it is unlikely that other Greeks would have objected if Elis exploited the justice of Zeus as an instrument to strengthen its influence over the other communities of Eleia. Cases in which Elis tried to use its authority at Olympia against states outside Eleia were probably very rare,5 and the attempt to do so against Sparta in 420 was a failure.

5

A notable dispute between Elis and Athens arose in 332. Athens objected strongly to an Eleian decision to fine the Athenian Kallippos, who had won the pentathlon but was punished for bribing his opponents, and Hypereides was sent as ambassador to Elis to have the fine revoked, but failed. Athens then maintained its objections, possibly for 20 years, until the Delphic oracle resolved the dispute. See Nielsen 2007: 85 n. 291, citing the ancient evidence and modern literature. Nothing in the surviving ancient evidence, however, suggests that in this case Elis had tried to exploit its authority at Olympia to its own advantage.

CHAPTER TEN CHARLES SELTMAN, LE TEMPLE DE ZEUS ET LE MONNAYAGE D’ARGENT DES ÉLÉENS FRANCK WOJAN

Abstract In Charles Seltman’s opinion (in his famous book The Temple Coins of Olympia), the Elean silver coinage was a coinage of Olympia. He also believed that the two Elean mints were located in the sanctuary of Olympia, one of them (the Zeus-mint) in or near (Seltman says “in the precincts of”) the temple of Zeus. The aim of this article is to reconsider these questions, one century after the publication of Seltman’s book, and to show that the Elean coinage was in fact a civic coinage (and not a temple one) and that the Elean mint was never located in the temple of Zeus. .

En 1921 paraissait l’ouvrage de Charles Seltman intitulé The Temple Coins of Olympia1, qui réunissait en fait trois articles de l’auteur parus précédemment dans la revue Nomisma : VIII (1913), IX (1914) et XI (1921). L’intérêt principal de cette monographie était d’établir pour la première fois le corpus des statères éléens tout en proposant diverses remarques sur les fractions du statère. Dès la préface, Seltman jetait également les bases de son étude du monnayage éléen, qui reposaient principalement sur les deux hypothèses suivantes : 1) le monnayage des Éléens était en réalité un monnayage d’Olympie ; 2) l’atelier monétaire des Éléens se trouvait à Olympie. Pour étayer sa démonstration, Seltman s’appuyait sur cinq arguments2, que je résume brièvement : 1

L’auteur est revenu par la suite sur le monnayage des Éléens, mais sans remettre en question les conclusions auxquelles il avait préalablement abouti : Seltman 1921, p. VII-IX, Seltman 1948 et Seltman 1951. 2 Seltman 1921, p. 1-3.

Charles Seltman, le temple de Zeus et le monnayage d’argent des Éléens 151

1) Les débuts du monnayage éléen étaient fixés à la fin du VIe siècle avant J.-C. et donc déconnectés du synécisme d’Élis daté de la fin des années 470. 2) Le monnayage des Éléens était rattaché au sanctuaire d’Olympie, car ce sanctuaire commun, qualifié de « plaque tournante du monde grec3 », occupait une position essentielle dans le monde grec, tant aux niveaux religieux, politique qu’économique. 3) Sur certains statères figurent des légendes monétaires en rapport direct avec le sanctuaire d’Olympie : il s’agit d’une part de la légende « / », qui entoure un Zeus lançant le foudre dans un carré incus, et, d’autre part, de la légende « », qui accompagne une tête féminine. 4) Les types monétaires éléens ne représentent en fait que des divinités propres au sanctuaire d’Olympie, Zeus et Héra ; aucune divinité poliade d’Élis ne figure sur ces monnaies soi-disant éléennes. 5) Les monnaies éléennes étaient en réalité frappées dans deux ateliers distincts du sanctuaire d’Olympie, en relation avec les temples de Zeus et d’Héra. Un siècle plus tard, presque jour pour jour après la parution du premier article de Seltman, d’indéniables progrès ont été réalisés tant dans la numismatique que dans l’histoire des Éléens. Comme le livre de Seltman n’a été ni complété ni remplacé, ce symposium international sur le temple de Zeus à Olympie est l’occasion de revenir notamment sur la cinquième hypothèse de travail de Seltman, qui, pour simplifier, voyait dans le temple de Zeus à Olympie l’atelier monétaire des Éléens. Nous tenterons ainsi de répondre à trois questions : 1) Quelle est la nature du monnayage éléen ? 2) Où se trouvait l’atelier monétaire des Éléens ? 3) Quelles nouvelles approches pouvons-nous aujourd’hui proposer à propos du monnayage des Éléens et de ses liens éventuels avec le temple de Zeus à Olympie ?

La nature du monnayage des Éléens Il existe une certaine ambiguïté à propos de la nature du monnayage des Éléens, et cette ambiguïté trouve en partie son origine dans l’ouvrage de Seltman. La question est simple : le monnayage des Éléens doit-il être défini comme un monnayage civique, c’est-à-dire celui d’une cité, ou comme un monnayage de sanctuaire ? Dit autrement, faut-il parler du 3

Traduction de l’expression « hub of Hellas » utilisée par l’auteur.

152

Chapter Ten

monnayage « des Éléens » ou du monnayage « d’Olympie » ? Cette question, en apparence banale, a pourtant divisé et divise encore le monde de la numismatique en deux camps : d’un côté, ceux qui attribuent ce monnayage aux seuls Éléens et le considèrent par conséquent comme un monnayage civique ; de l’autre, les tenants d’un monnayage d’Olympie, qui estiment que le sanctuaire de la vallée de l’Alphée avait la primauté sur les Éléens. George F. Hill fut le premier, à ma connaissance, à analyser ainsi le monnayage des Éléens4 : « It is not strictly accurate to speak of the coinage of Elis, as if it were a single city state. It should not be forgotten that the coinage of the Eleans was especially connected with Olympia, and probably had more of the character of a temple-coinage than any other considerable series known to us. ». À la même époque, Ernest Babelon décrivait de la même manière le monnayage éléen : « …les monnaies qui portent le nom des Éléens ont été frappées tous les quatre ans, à l’occasion de la célébration des jeux olympiques, à Olympie, cité religieuse qui dépendait d’Élis. » 5 ; « …ce sont eux aussi [les Hellanodices], sans nul doute, qui font frapper les admirables médailles-monnaies… »6. Seltman s’inspira certainement de ces hypothèses, mais il leur donna une ampleur supplémentaire comme l’indique d’ores et déjà le titre choisi pour ses trois articles puis son ouvrage de synthèse : « The Temple Coins of Olympia ». Dès la première phrase de son introduction, il indiquait : « It is only after a very careful study of the matter that I have assigned the coins usually spoken of as Elean to Olympia. »7 Le numismate britannique attribuait donc ce monnayage au sanctuaire d’Olympie et reléguait, pour ainsi dire, les Éléens à un rang secondaire. Une décennie après la parution de l’ouvrage de Seltman, Josef G. Milne8 alla plus loin que Seltman en affirmant que le monnayage éléen n’était composé que de simples « médailles commémoratives des Concours »9. Colin M. Kraay10, dans les années 1970, avait un point de 4

Hill 1906, p. 77. Babelon 1914, col. 699. Cette hypothèse avait été formulée sept ans auparavant (Babelon 1907), col. 891 : « C’est ainsi que [les monnaies frappées avec la légende] / désignerait les monnaies frappées par les administrateurs du temple de Zeus à Olympie… » et col. 894 : « Si les monnaies à la légende / ont été ainsi frappées par les soins des Hellanodices… ». 6 Babelon 1914, col. 708. 7 Seltman 1921 p. 1. 8 Milne 1931, p. 171-180. 9 Milne 1931, p. 180. Traduction de l’expression « souvenirs » utilisée par l’auteur. 5

Charles Seltman, le temple de Zeus et le monnayage d’argent des Éléens 153

vue similaire : « There can be little doubt that Seltman was right to attribute this coinage to the great festival centre of Olympia, which was normally controlled by the Eleans, rather than to the city of Elis itself. (…) …it was not a normal coinage of state (…), but rather a souvenir of the festival which passed only at its bullion value… ». Récemment encore, Sélènè Psoma11, en s’appuyant notamment sur les arguments développés par Alan Walker12, suggérait d’interpréter ainsi le monnayage des Éléens : « There are a number of good reasons to suggest (…) that both series should be interpreted as festival coinages ». Revenons d’abord sur l’argument de Milne repris par Kraay. Suivre cette thèse au pied de la lettre revient à remettre en cause non seulement la nature même du monnayage des Éléens, mais, surtout, à lui nier toute valeur économique et financière13. Les arguments de ces auteurs me paraissent contestables, car ils vont à l’encontre de tout ce qui est observé dans le monde grec en matière numismatique, financière et économique. Il revient, me semble-t-il, à Hélène Nicolet-Pierre d’avoir émis un doute sur l’interprétation de Seltman : « ces monnaies [éléennes] en effet portent dès l’origine l’ethnique des Éléens et des types en rapport avec le culte le plus important célébré sur leur territoire ; rien n’indique qu’elles aient été émises ou contrôlées par une autorité sacerdotale »14. Les deux arguments avancés par Nicolet-Pierre pour contredire Seltman sont particulièrement intéressants. La définition que les numismates et les épigraphistes donnent de l’ethnique monétaire ne laisse plus guère de place au doute. D’après Philippe Gauthier, même s’il emploie le conditionnel, « on pourrait dire que l’ethnique figurant sur les monnaies renvoie tout à la fois à l’autorité qui émet la monnaie et à la communauté qui en est la source et qui, la première, utilise cette monnaie. »15. Or, sur les monnaies éléennes figurent systématiquement, à de rares exceptions près, l’ethnique des Éléens, qui est écrit parfois en entier, sous la forme ancienne « » (pour ), puis « », mais le plus souvent sous forme abrégée, avec les 10

Kraay 1976, p. 103. Psoma 2008, p. 240-241. 12 Walker 2004, sans pagination. 13 D’après ces auteurs, l’Élide n’était pas, en dehors des Olympia, un territoire attractif et, par conséquent, y frapper monnaie n’avait que peu d’intérêt. Ces « médailles » n’auraient pas été utilisées comme instrument de valeur et d’échange, mais elles auraient été vendues aux athlètes, aux délégations des cités et aux visiteurs du sanctuaire. 14 Nicolet-Pierre 1975, p. 7. 15 Gauthier 1975, p. 166. 11

154

Chapter Ten

deux premières lettres (« »), digamma – alpha. On remarquera en tête la présence d’une lettre issue de l’alphabet grec archaïque, le digamma ( ), qui a été utilisée en Élide jusqu’à l’époque hellénistique et que l’on retrouve dans les inscriptions épigraphiques comme dans les légendes monétaires16. L’ethnique des Éléens figure la plupart du temps au revers, parfois au droit, des monnaies. L’identité de l’autorité émettrice est donc bien celle des Éléens et ces monnaies ne peuvent pas être considérées autrement que comme éléennes17. Passons rapidement à l’argument n° 3 de Seltman : il est plus technique et se rapporte à deux légendes monétaires. Ces deux légendes – « / » et « » – ne peuvent pas être considérées, je pense, comme des preuves déterminantes, car d’un point de vue numismatique et historique, elles s’inscrivent dans un contexte précis et valable pour une courte période. La légende archaïque « / » doit sans doute être rattachée aux débuts du monnayage éléen, juste avant que les Éléens ne forment une cité par synécisme, quand ils assurèrent définitivement leur contrôle sur la basse vallée de l’Alphée et sur le sanctuaire d’Olympie18 dans la première moitié du Ve siècle avant J.-C. Quant à la légende « », elle ne sert qu’à identifier la nymphe éponyme représentée au droit ou au revers de certains statères datables des lendemains de la guerre qui a opposé les Éléens et les Arcadiens dans la seconde moitié des années 360 avant J.-C. Cette maigre production monétaire à l’effigie de la nymphe Olympia aurait permis aux Éléens de « commémorer », si l’on peut dire, leur reprise en main du sanctuaire d’Olympie19. Cependant, dans l’argumentaire développé par Seltman, les arguments nos 2 et 4 semblent apparemment en contradiction avec ce que nous venons d’affirmer. D’une part (argument n° 2), le sanctuaire commun d’Olympie était, avec celui de Delphes, l’un des plus célèbres du monde grec et sa popularité ne s’est jamais démentie tout au long de l’Antiquité. Or, depuis l’époque archaïque, le sanctuaire était convoité par les Éléens. Si, dans le détail, les événements restent mal connus, il semble que, au début du VIe 16

IED p. 255-273. Ce que reconnaissent Walker (Walker 2004, sans pagination) : « While the legend on the coin tells us the Eleans minted the coins… » et Psoma (Psoma 2008, p. 240) : « …was issued by the Eleans (…) : we find the initials of their ethnikon (…) on the reverse of the very first staters ». 18 Cette hypothèse avait été suggérée naguère par Babelon (Babelon 1907, groupe A, col. 885-896). Elle mérite, je pense, d’être réexaminée et approfondie. 19 Ritter 2002, p. 47-52. 17

Charles Seltman, le temple de Zeus et le monnayage d’argent des Éléens 155

siècle avant J.-C., les Éléens exercent leur autorité sur la basse vallée de l’Alphée et administrent le sanctuaire d’Olympie20, et ce, bien que cette autorité éléenne sur le territoire que nous appelons l’Élide n’ait pas été complète avant l’époque hellénistique21. D’autre part (argument n° 4), la figure de Zeus est omniprésente dans le monnayage éléen : soit le dieu lui-même est représenté soit il apparaît sous la forme de l’un de ses attributs (le foudre, l’aigle). Or, si l’on en croit le témoignage de Pausanias, Zeus n’est pas la divinité poliade des Éléens, ni Héra d’ailleurs22. De plus, la représentation éphémère de Nikè rappelle davantage les Concours sportifs d’Olympie que la ville d’Élis. A priori inattaquable, cet argument, qui repose sur les types monétaires, peut être retourné pour deux raisons : d’une part, ce n’est pas obligatoirement la divinité poliade qui figure au droit d’une monnaie grecque. Cette règle, qui s’inspire du modèle athénien, connaît des exceptions23. D’autre part, si l’on considère le monnayage éléen dans son ensemble, les représentations d’Apollon, du cheval et l’ethnique dans une couronne ne constituent pas des références directes au sanctuaire d’Olympie. Comme nous l’avons dit, les Éléens ont cherché, depuis l’époque archaïque, à prendre le contrôle du sanctuaire d’Olympie24. La lecture de la Périégèse montre même qu’Olympie peut être défini non seulement comme un sanctuaire « commun », mais aussi comme un sanctuaire « éléen » : Pausanias n’affirme-t-il pas que « Sacrifient à Zeus, même sans qu’il y ait panégyrie, les simples particuliers et les Éléens chaque jour. » 25 ? Par conséquent, si Olympie peut être défini autant comme un sanctuaire « éléen » que « commun », il n’est sans doute pas exagéré de 20

Voir notamment Möller 2004, p. 249-270, et Taita 2007. Roy 1999. 22 Pour mémoire, rappelons que Pausanias, lorsqu’il décrit la ville d’Élis (Description de la Grèce, VI, 23, 1- 26, 3), mentionne les sanctuaires d’Artémis Philomeirax (23, 8), d’Apollon Akésios (24, 6), des Grâces (24, 6), de Tychè et de Sôsipolis (25, 4), de Dionysos (26, 1) et d’Athéna (26, 3), les temples de Silène (24, 8), d’Aphrodite (25, 1) et d’Hadès (25, 2), ainsi que les autels dédiés à Héraclès de l’Ida, Éros, Antéros, Déméter et Coré (23, 3), de Zeus (24, 3), le tombeau d’Achille (24, 1) et les statues d’Apollon Akésios, d’Hélios et de Séléné (24, 6), et d’Éros (24, 7). 23 On peut citer pêle-mêle l’épi de blé à Métaponte et à Sybaris, le bouclier béotien, les quadriges vainqueurs des monnaies syracusaines, etc. 24 Möller 2004 ; Taita 2007, p.69-81, 111-116 et 140-142. De plus, selon Anne Jacquemin, (Jacquemin 2001, p. 184) : « La mainmise éléenne sur Olympie permit à cette communauté du nord de dominer tout le pays et les rituels traduisent cette volonté d’affirmer l’unité des populations de l’Élide Creuse et de la Pisatide. » 25 Pausanias, V, 13, 9 (traduction de J. Pouilloux, CUF). 21

156

Chapter Ten

qualifier le Zeus d’Olympie d’ « éléen » également. L’importance et la célébrité du Zeus de l’Altis, auprès des Grecs en général et des Éléens en particulier, étaient telles que choisir l’effigie de Zeus ou l’un de ses attributs comme type monétaire principal pour orner le droit et/ou le revers des monnaies éléennes ne doit pas sembler surprenant.

L’atelier monétaire des Éléens À ma connaissance, l’atelier monétaire (argurokopeion) des Éléens n’a jamais été retrouvé lors des fouilles archéologiques menées jusqu’à présent à Olympie ou à Élis. Babelon26, quelques années avant Seltman, avait suggéré de localiser l’atelier monétaire des Éléens dans le sanctuaire d’Olympie. Mais, si le premier n’évoquait qu’ « un » atelier monétaire, on doit au second une hypothèse plus audacieuse (argument n° 5) : d’après Seltman, il devait exister, dans le sanctuaire d’Olympie, non pas un, mais deux ateliers monétaires27. Le premier (« Olympia Zeus-mint »), en relation avec le temple de Zeus, aurait frappé les monnaies aux types de la divinité principale de l’Altis, c’est-à-dire les monnaies avec les attributs principaux du dieu (l’aigle et le foudre), puis avec son effigie. Le second (Olympia « Hera-mint »), en relation avec le temple d’Héra, aurait émis les exemplaires à l’effigie de la sœur-épouse de Zeus. Pourquoi deux ateliers ? En élaborant une telle hypothèse, Seltman résolvait ainsi une difficulté majeure du monnayage éléen : l’absence constatée de liaisons de coins de revers entre les monnaies aux types de Zeus d’un côté et celles aux type d’Héra de l’autre28. Cette constatation, qui a effectivement laissé dubitatif plus d’un numismate, a été validée récemment par Walker29 et Psoma30, mais elle n’a toutefois été que partiellement confirmée par Françoise Capra, qui a travaillé sur le monnayage d’argent des Éléens31 et qui a émis à ce propos quelques timides réserves32.

26

Babelon 1914, col. 699. Seltman 1921, p. VII-VIII et 2-3. 28 Seltman 1921, p. VIII et 2-3. 29 Walker 2004, sans pagination. 30 Psoma 2008, p. 240. 31 Capra 2008, 2008, p. 113 : « Cet argument reste certes formellement valable pour l’échantillon dont nous disposons aujourd’hui… ». 32 Notamment p. 317, quand l’auteur affirme que « …les ressemblances frappantes entre certains coins réputés appartenir à l’un ou l’autre des deux ateliers, de même que le choix de thèmes indiscutablement liés à Zeus pour les types de revers 27

Charles Seltman, le temple de Zeus et le monnayage d’argent des Éléens 157

La localisation de l’atelier monétaire des Éléens à Olympie est donc une hypothèse séduisante, qui a été aussitôt acceptée et reprise par de nombreux numismates, dont Jan H. Jongkees33, Kraay34 et, plus récemment, nous l’avons dit, par Walker35 et Psoma36. Walker s’appuie sur cinq arguments, qui reprennent et/ou complètent ceux de Seltman : 1) les besoins monétaires du sanctuaire étaient accrus à l’occasion des Concours sportifs, notamment en raison des nécessaires opérations de change ; 2) les types monétaires font référence explicitement aux Concours sportifs et à Olympie ; 3) l’existence d’un seul atelier monétaire doit être exclue en raison de l’absence de liaisons de coins de revers entre les « Zeus » et les « Héra » ; 4) l’immense marché que constituait la panégyrie, à l’image des Panathénées à Athènes, est la preuve d’une intense circulation monétaire ; 5) le poids des statères, qui suivent normalement l’étalon éginétique, est inférieur à la moyenne attendue. Admettons que l’atelier monétaire (ou les deux ateliers) se trouvait dans le sanctuaire d’Olympie. Où faut-il le localiser ? Seltman proposait la solution suivante, « situated in the precincts of the temple of Zeus… in the precincts of the temple of Hera »37, mais le terme qu’il emploie ne me paraît ni clair ni précis. « Precinct » peut se traduire en français par « l’enceinte », le « pourtour », les « alentours », mais à quoi pensait précisément Seltman ? On a parfois traduit en français son expression simplement par « le temple de Zeus », ce qui sous-entendait pour certains l’intérieur du temple (?). Naturellement, cette interprétation est a priori exclue, ne serait-ce que pour des raisons évidentes de sécurité, voire de place. Seltman pensait-il à un local proche de chaque temple ? Si oui, dans l’espace contigu ou proche des temples de Zeus et d’Héra ? Aucune découverte archéologique n’a, à ma connaissance, corroborée l’affirmation du numismate britannique. Quant à Walker, s’il se démarque de Seltman sur ce point, il rappelle à juste titre que le site d’Olympie n’a livré aucune trace d’activité artisanale en liaison avec un atelier monétaire, avant de proposer la solution suivante : « it is very likely that the Elean mint at Olympia was located associés aux droits portant le profil de la déesse [Héra], nous paraissent être des éléments solides d’unité entre les deux productions. » 33 Jongkees 1939, p. 219-229, pour qui les monnaies ont été frappées à Olympie pour le compte des Éléens (p. 219). 34 Kraay 1976, p. 105. 35 Walker 2004, sans pagination. 36 Psoma 2008, p. 240. 37 Seltman 1921, p. 2.

158

Chapter Ten

somewhere that has either not been excavated, or was washed away long ago by flooding from the rivers Kladeos and Alpheios. »38. Si l’existence de deux ateliers monétaires situés à quelques mètres de distance l’un de l’autre dans un prestigieux sanctuaire commun était avérée, il resterait alors à expliquer leur fonctionnement, soit concomitamment soit à des périodes différentes. Une fois encore, Seltman s’appuie sur son observation de l’évolution du monnayage éléen : entre la fin du VIe siècle et le début du IIe siècle avant J.-C., il avait distingué 14 groupes (dont le critère principal est le type de droit) : 11 étaient attribués à l’atelier de Zeus (Groupes A à L39) et 3 à l’atelier d’Héra (groupes E1, E2 et GH). Mais, comme les monnaies sur lesquelles figure la tête d’Héra sont chronologiquement postérieures à celles aux types de Zeus, Seltman en concluait, pour le Ve siècle, qu’un seul atelier – celui de Zeus – avait fonctionné des débuts du monnayage jusqu’à l’ouverture de l’atelier d’Héra à la fin des années 420 avant J.-C. Cette date ne doit rien au hasard : l’apparition de l’effigie d’Héra dans le monnayage éléen a été mise en relation avec l’alliance militaire conclue entre les Éléens, les Argiens, les Mantinéens et les Athéniens au cours de la « Guerre du Péloponnèse » ; ce serait donc une effigie qui associerait la Héra d’Olympie avec la Héra argienne40. Mais cette hypothèse a été contestée dès l’origine par Seltman, qui y voyait davantage une simple coïncidence41. Minon, dans son étude des inscriptions éléennes, ne remet pas en cause cette datation, mais la discussion est plus serrée qu’on ne pourrait le croire42.

38

Cf. n. 27. En réalité, les groupes sont numérotés de A à H et de J à L. Il n’existe pas de groupe I rattaché à l’atelier de Zeus. 40 Thucydide, V, 47. Gardner 1887, p. XXXVI-XXXVII ; Head, 19112, p. 421422 ; Babelon 1914, col. 731, Seltman 1921, p. 3 et « group E ». 41 Seltman 1921, p. 86 : « My reason for tentatively assigning the first of these coins issued from the mint of Hera to the year B.C. 421 has no connection whatever with the fact that in that year incidentally Elis concluded an alliance with Argos and Mantinea. That was a political event which concerned the city of Elis. But what concerned Olympia was the fact that the next year B.C. 420 was an Olympiad. » 42 IED p. 262-265. Il peut sembler étonnant, en effet, à la lecture de la démonstration de l’auteur, que l’apparition d’un à l’initiale du nom d’Héra, cinquante ans avant la généralisation de cette lettre dans les textes éléens, témoignerait par conséquent de l’influence des Argiens en Élide dans les années 420 avant J.-C., influence qui n’aurait concerné que la monnaie et pas d’autres domaines. 39

Charles Seltman, le temple de Zeus et le monnayage d’argent des Éléens 159

Seltman concevait ainsi les rythmes de la production monétaire éléenne43 (fig. 1) : « ATELIER DE ZEUS »

« ATELIER D’HÉRA »

Groupe

Datation

Nombre de statères

Groupe

Datation

Nombre de statères

A B C D E

510-471 471-452 452-432 432-421 421-365

33 37 55 15 32

G H J K L TOTAL

364-343 343-323 323-271 271-191 191- ?

27 12 15 11 2 239 statères

— — — — E1 E2 GH1

— — — — 421-385 385-365 364-323

— — — — 48 17 60

— — —

— — —

— — — 125 statères

Fig. 1 : Les rythmes de la production monétaire éléenne, d’après Ch. Seltman On constate que les deux ateliers n’auraient fonctionné parallèlement que sur une période d’un siècle, estimée entre les années 420 et 320 avant J.-C. Le reste du temps, l’atelier de Zeus peut donc être considéré comme le principal atelier monétaire des Éléens. Mais lorsque l’on regarde la ligne « total », on s’aperçoit que l’atelier d’Héra a émis en quantité la moitié des statères attribués à l’atelier de Zeus, alors qu’il n’a fonctionné qu’un siècle environ contre trois siècles pour celui de Zeus. La période 420-323, si cette datation est avérée, est celle de la production monétaire la plus intense44.

43

Il n’existe pas de groupe I. Quant au groupe F, nous l’avons supprimé du tableau car il s’agit des deux monnaies frappées au nom des Pisates. 44 Ces chiffres sont désormais à corriger avec les données fournies par Françoise Capra (Capra 2008).

160

Chapter Ten

Nouvelles approches du monnayage des Éléens et de ses rapports éventuels avec le temple de Zeus à Olympie Un siècle après la parution des premiers articles de Seltman, notre regard sur le monnayage des Éléens s’est inévitablement modifié : quelques articles et deux thèses de doctorat récentes – celle de Françoise Capra en 2008 et la mienne en 201145 – montrent que le monnayage des Éléens n’a pas encore révélé tous ses secrets. Reprenons et discutons les deux arguments principaux de Seltman : 1) Le monnayage des Éléens était en réalité un monnayage d’Olympie. 2) L’atelier monétaire des Éléens se trouvait à Olympie. En ce qui concerne le premier argument, je vois davantage dans le monnayage des Éléens un monnayage civique qu’un monnayage de sanctuaire, en dépit des arguments avancés encore récemment par Walker et par Psoma46. La « capitale » politique de l’Élide est Élis. L’ethnique monétaire, dont on a déterminé l’importance, met en avant systématiquement les Éléens et non le sanctuaire d’Olympie47. Quant aux deux légendes monétaires présentées par Seltman, elles constituent, nous l’avons vu, davantage une exception qu’une règle. Les liens qui unissaient la ville d’Élis et le sanctuaire d’Olympie montrent qu’une certaine prééminence revenait à la ville plutôt qu’au sanctuaire. En effet, par exemple, quand on lit la description de la ville d’Élis faite par Pausanias48, on constate que de nombreux bâtiments situés sur et autour de l’agora ont leur double dans le sanctuaire d’Olympie49 (et non l’inverse). De plus, les athlètes et leurs entraîneurs, qui venaient

45

Wojan 2011. Walker 2004 ; Psoma 2008. 47 La question est de savoir si l’ethnique monétaire l’emporte sur le type monétaire qui figure au droit de la monnaie. L’ethnique monétaire garde, je pense, une prééminence sur le type monétaire qui, lui, peut changer sans que l’utilisateur soit gêné, comme en témoigne le monnayage éléen où, sur les statères, les types de Zeus (Zeus lui-même, l’aigle, le foudre) et d’Héra cohabitent en quelque sorte. 48 Pausanias, VI, 23, 1- 26, 3. 49 La liste suivante n’est pas exhaustive : les trois gymnases (23,1 ; 23, 4 ; 23, 5), le Bouleutèrion (23, 7) l’Hellanodikéôn (24, 1) et le prytanée (non indiqué par le Périégète). 46

Charles Seltman, le temple de Zeus et le monnayage d’argent des Éléens 161

concourir aux Olympia, devaient au préalable séjourner à Élis50 avant que la procession ne les emmène au sanctuaire51. L’argument selon lequel les poids des statères sont éloignés de ceux observés traditionnellement dans les cités ayant choisi l’étalon éginétique peut être également contourné. D’après Capra, qui a analysé le poids de quelques 1200 statères52, on constate effectivement une irrégularité flagrante des poids mesurés : la moyenne est de 11,84 g, avec un écarttype de 0,42 (plus de 3,5 %) ; la médiane s’élève à 11,94 g. Mais, si l’on ne tient compte que des 10 % des monnaies les plus lourdes53, on trouve alors une moyenne de 12,34 g, c’est-à-dire un chiffre proche de la « normale », environ 12,2 g. En ce qui concerne la seconde affirmation de Seltman, nous pouvons lui opposer les arguments suivants : Le principal reproche est qu’un atelier monétaire coûte cher à la cité. Par conséquent, il paraît difficilement envisageable que les Éléens aient pu entretenir deux ateliers, qui plus est fonctionnant à quelques mètres de distance, que ce soit séparément et/ou concomitamment. De manière plus anecdotique, l’atelier de Zeus était censé frapper les monnaies aux types de Zeus et l’atelier d’Héra les monnaies à l’effigie de la déesse. Mais, dans ce cas, de quel atelier sont sortis les « Olympia » et, plus tard, les « Apollon » ainsi que les « bronzes à l’ethnique » ? Il n’est guère raisonnable d’envisager un nombre supérieur d’atelier monétaire. Enfin, les archéologues n’ont retrouvé aucune trace d’activité monétaire à Olympie, dans ou à proximité des temples de Zeus et d’Héra, ni même dans ou à proximité du sanctuaire. On sait pourtant que le travail des métaux n’était pas exclu du sanctuaire d’Olympie si l’on pense, par exemple, à l’atelier de Phidias, dans lequel le sculpteur athénien avait réalisé la statue chryséléphantine de Zeus54. Il est vrai aussi que, à Élis, aucune trace d’activité monétaire n’a été décelée. L’atelier monétaire des Éléens pouvait-il se trouver à Olympie, ne serait-ce qu’en raison des Olympia ? Rappelons une évidence : les Olympia se déroulaient sur quelques jours, chaque cinquième année, lors

50

Pausanias, VI, 23, 1-2. Deux routes permettaient de relier Élis à Olympie, celle du "littoral" et celle de "montagne". D’après Pausanias (V, 25, 7), la route littorale était appelée « voie sacrée » (hiéra). On en déduit généralement que la procession devait l’emprunter. 52 Capra 2008, p. 97-99. 53 Cette méthode a été suggérée par Seltman lui-même (Seltman 1921, p. 109). 54 Pausanias, V, 15, 1. Les résultats des fouilles archéologiques de l’ « atelier de Phidias » ont été publiés. Cf. Mallwitz & Schiering 1964 ; Schiering 1991. 51

162

Chapter Ten

du mois olympique55. La panégyrie et les Concours sportifs attiraient certes une foule immense, mais sur une courte période (grosso modo, un mois tous les quatre ans). En dehors de cette période festive, Olympie n’accueillait que le personnel du sanctuaire et les visiteurs de passage. Comme une cité frappe monnaie en fonction de ses propres besoins, on peut se demander, d’une part pourquoi l’atelier monétaire des Éléens aurait été aménagé à une cinquantaine de kilomètres de la ville, fût-ce dans un prestigieux sanctuaire et, d’autre part, pourquoi la production monétaire doit être rattachée uniquement et exclusivement aux Olympia. D’un point de vue strictement numismatique, les dénominations émises par l’atelier monétaire éléen sont essentiellement des statères, peu compatibles avec les achats du quotidien, mais plus appropriés en revanche pour financer de grosses dépenses. Par exemple, au cours du Ve siècle avant J.-C., les Éléens ont connu quatre événements majeurs, qui ont eu très certainement des conséquences en matière de politique monétaire et financière : le synécisme d’Élis à la fin des années 470 ; les relations entre les Éléens et les autres communautés non-éléennes du territoire que nous appelons l’Élide ; la construction du temple de Zeus, puis la réalisation de la statue de Zeus ; la « Guerre du Péloponnèse » et ses conséquences (les opérations militaires, les changements d’alliance, et l’invasion lacédémonienne du territoire éléen à l’issue du conflit). Ces quatre événements ont-ils davantage influencé la politique monétaire des Éléens que les Olympia ? La question reste ouverte et débattue. Sans minimiser l’importance des Concours sportifs, qui attiraient chaque cinquième année des athlètes et des visiteurs venus des quatre coins du monde grec, ce serait une erreur de négliger les besoins financiers des Éléens en dehors du mois olympique56.

Conclusion L’ouvrage de Seltman sur le monnayage éléen, et plus précisément les statères, reposait sur les deux postulats suivants : 1) Le monnayage des Éléens était en réalité un monnayage d’Olympie ; 2) L’atelier monétaire des Éléens se trouvait à Olympie, en relation avec le temple de Zeus et celui d’Héra. Ces deux affirmations me paraissent contestables et, même si le classement établi naguère par Seltman garde en partie toute son autorité, il 55

IED p. 52-54. Dans le même sens, Patay-Horváth 2013a, p. 1-28, et Patay-Horváth 2013b, p. 61-83. 56

Charles Seltman, le temple de Zeus et le monnayage d’argent des Éléens 163

nécessite désormais des corrections et des ajustements. De plus, certaines analyses du numismate britannique – notamment ses interprétations et ses datations – doivent être réexaminées en fonction d’autres critères. Pour ma part, le monnayage des Éléens doit être davantage regardé comme un monnayage civique que comme un monnayage de sanctuaire57. Il doit être étudié comme un monnayage civique qui a bénéficié à un sanctuaire autant qu’à une cité, car c’est sur le territoire d’une cité que se trouvait ce prestigieux sanctuaire. Le monnayage des Éléens a été émis pour satisfaire les dépenses traditionnelles de la cité, dépenses dans lesquelles étaient inclus le fonctionnement et l’embellissement d’un sanctuaire, sanctuaire qui était à la fois « éléen » et « commun ». Reste l’énigme de la localisation de l’atelier monétaire des Éléens : à Élis et/ou à Olympie ? Soit la primauté doit être donnée à la ville principale de l’Élide – on dirait de nos jours la « capitale » – et, dans ce cas, l’atelier monétaire avait toutes les chances de se trouver sur les bords du Pénée ; soit le sanctuaire d’Olympie avait réellement la prééminence sur Élis, auquel cas l’atelier monétaire éléen avait toute sa place sur les bords de l’Alphée et du Kladéos. Mais, il faut reconnaître qu’il s’agit là de considérations modernes. Peuvent-elles s’appliquer à l’époque antique ? Rien n’est moins sûr. Néanmoins, une solution peut être avancée si l’on considère un atelier monétaire antique comme une structure souple et itinérante. Dans ce cas, l’atelier monétaire des Éléens pouvait se trouver à Élis, car, rappelons-le, une cité frappe monnaie en fonction de ses propres besoins et les besoins monétaires des Éléens ne peuvent pas être résumés aux seuls Concours sportifs d’Olympie. Cependant, le temps du mois olympique, l’atelier pouvait fonctionner sur le site d’Olympie pour les nécessaires opérations de change. Quoi qu’il en soit, il ne semble guère pertinent, à propos des Éléens, de vouloir séparer la ville de son sanctuaire extra-urbain : le territoire comme le monnayage des Éléens avaient tous deux une dimension éléenne et olympique. Quant au temple de Zeus, il faut le rendre – hélas ! – à la divinité et laisser les Éléens battre monnaie anywhere in Eleia, except in the temple of Zeus…

57 Capra 2008, p. 316. Contra, Walker 2004, sans pagination, mais qui rend toutefois aux Éléens la part qui leur revient.

CHAPTER ELEVEN LUCIAN ON HERODOTUS: A POSSIBLE SECOND CENTURY AD VIEW ON THE WEST PEDIMENT JESSICA SUSANNE KRAUSE

Abstract It is often taken as a fact that Herodotus visited the Olympian Games in the second half of the fifth century BC, using this occasion to perform his Histories in the sanctuary of Zeus. We are told that the exact place of his performance was the opisthodomos of the temple itself. This even led W. Dörpfeld to reconstruct a stone bench in this part of the temple, although such a bench is archaeologically not verifiable. What is often not considered, however, is that the first mention of this performance comes from Lucianus, in his work Herodotus or Aëtion (Lukian., Herod. 1) written in the second century AD. This late date might suggest that the visit of Herodotus to Olympia was made up by the author in order to construct a story that served his needs by framing his own writing in a particular manner. But why should he have chosen to let Herodotus hold his speech in the back of the temple, right below the west pediment, where the space for spectators would have been extremely limited due to the sacred oil tree and the enclosure? Herodotus would perhaps even not have been visible to visitors in the wider sanctuary grounds. And was the temple itself an appropriate place for such performances? Here, I highlight the possibility that Lucianus placed “his” Herodotus right below the Thessalian Centauromachy during the wedding festivities in order to use it as a metaphorical illustration of the Histories, using the term “hybris” as a key element. These considerations include close observations of the historical context of the erection of the temple and its sculptural decoration, with special emphasis on Elean–Athenian relations at that time.

Lucian on Herodotus

165

Introduction Images have always carried different meanings, both in the time of their composition as well as in subsequent periods. In this chapter I focus on a possible interpretation in one of these later cases, and attempt to explicate the different levels of interpretation embedded within one written source.1 To do so, it is important not only to take ancient texts as collections of facts, but also to work with them in a more elaborate way, as I demonstrate later on. To analyse the written sources of antiquity2 in a deeper manner we have to interpret them just as we interpret more recent or even contemporary texts. This is certainly true for the texts of Lucian of Samosata. This second century AD author wrote texts of different genres3 and has for good reasons been labelled as a satirist in modern times.4 As we will see, Lucian used his texts as a whole as well as every single word within them to play with his audience and his readers. His works seem to be specially crafted for this purpose. For archaeologists, it is still not entirely common to work with ancient texts in the manner described – i.e. to search for the imbedded intentions and interests of both the author and the reader and to spot the passages that contain more than is actually spelled out. It has happened 1

I would like to thank the audience of the conference at Budapest for their support. Previously, I also had the chance to present preliminary ideas to the KHAN conference at Kiel (March 21st–22nd, 2014). My thanks go also to the organizers of this conference. For support at Kiel University I would like to thank PD Dr. Konrad Hitzl who – as always – offered his help for both the talk and the paper, as well as Prof. Dr. Josef Wiesehöfer and Prof. Dr. Lutz Käppel, who gave me their opinion on my ideas at various stages. Very special thanks go to my colleague Kleoniki Rizou, to whom I owe more than simply the hint at Euripides and Inga Quandel. I would also like to express my deep gratitude to Prof. Dr. Hans Rupprecht Goette, who helped me a great deal by creating the sanctuary plan used in this paper. 2 All abbreviations of ancient authors and their writings follow the DNP online abbreviations (http://www.brillonline.nl/entries/brill-s-new-pauly/ancient-authorsand-titles-of-works-Ancient_Authors_and_Titles_of_Works [accessed September 26th, 2014] with additions for missing abbreviations). The personal names in this paper are given in their commonly used form that might therefore be either their Latin or Greek version. 3 In his writings we find not only prolaliai (pre-speeches) but as well ekphraseis, enkomia, instructive writings, folk tales and apologetic texts; for a discussion of Lucian’s prolaliai see Nesselrath 1990. 4 See for example Anderson 2009, 6; for a lengthy discussion Whitmarsh 2001, 247–293. For an overview on the reception of Lucian from the 19th century onwards see Jones 1986, 1–5.

166

Chapter Eleven

more than once that – for various reasons – ancient sources have been taken very literally. A good example is Lucian’s prolalia “Herodotus sive Aëtion.” This short introductory speech5 is the first account we have concerning Herodotus’ alleged performance of his Histories at Olympia.6 Even though it was written more than 500 years after the death of Herodotus, archaeologists and others still take this information as fact.7 There might be gaps in our source material, and there might have been older written or oral accounts about this event,8 but the construction of Lucian’s text is too conspicuous by far to take it as a collection of mere facts. This chapter therefore aims on the one hand to present a new idea concerning how to read and work with this prolalia, and on the other hand it argues for a more critical use of written sources by archaeologists in general.

Written Sources and Classical Archaeology To deal with the second point first, I would like to give an example of how insufficiently interpreted written source material might influence archaeology. In 1892, Wilhelm Dörpfeld issued the first plan of the 5

For the term “prolalia” see Nesslrath 1990, 11–117. The only other account mentioning the performance to have taken place in Olympia is the Suda article on Thucydides (414). Three other accounts of a performance tell the same story of the young Thucydides listening to Herodotus’ performance, but without giving a place: Marcellinus § 54 (see in general Maitland 1996 as well as Bux 1930 (with older literature) for further information on this source) as well as Photios (cod. 60; he even gives as a place for the performance the house of Thucydides’ father, see Schöll 1855, 412) and Tzetzes (Poppo 1, 1, 321 [cited after Bux 1930, 1478]). Interestingly, the fact that Marcellinus (who will be discussed further as a model for the two similar later sources Photios and Tzetzes, who are very likely to be based on Marcellinus; see Bux 1930, 1478) only mentions a performance of Herodotus in front of Thucydides and his father. This has the consequence that the Suda account appears as though it were a symbiosis of the account by Lucian as well as of that of Marcellinus (see below). 7 Of course I do not intend to attack archaeologists in general, as the majority of them take the use of written sources very seriously. It could also be due to the character of some of those cases in which not enough caution is taken with the texts, as they are written for a broader audience and are therefore not to be considered as “proper” scientific works in the strict sense. Nevertheless I think that even in these cases more attention should be paid not to mix up fact and fiction. 8 One example for the use of older written sources by ancient writers is Athenaios (Ath. 14, 12) where he cites Dikaiarchos’ “Book on Olympia.” 6

Lucian on Herodotus

167

Temple of Zeus at Olympia in his volume Die Baudenkmäler (Dörpfeld 1892); there he gives the position of every stone he as well as others came across during the German excavations that had been ongoing since 1875.9 Additionally, he added to the plan a sitting bench on the eastern wall of the Opisthodomos of the temple, even though a careful onlooker might have noticed that the plaster of the floor stretches right beneath this structure. A look in the accompanying text volume10 gives the following explanation for this addition: “Letztere [i.e. die Hinterhalle] diente als Versammlungsraum und enthielt daher eine lange Sitzbank aus Stein.”11 Dörpfeld does not mention the bench again. This tells us that even a renowned archaeologist such as Dörpfeld – a pioneer of Classical Archaeology – was not immune to the influence of written sources. The only thing that now can be “read” out of this bench is that for Dörpfeld it was absolutely clear that there was a sitting bench – possibly because to him this was the place where Herodotus performed.12 The “German Quellenkritik” for which German Ancient Historians are famous, as brought forward by for example Johann Gustav Droysen,13 should be more carefully applied in archaeological works.14 After these more general remarks on how not to use written sources, I would like to concentrate on a proposal for a new interpretation of a certain aspect of Lucian’s “Herodotus sive Aëtion.” As has been already hinted at, apart from other clues on the meaning of certain pictures, as for example context and combination patterns or their historical background, a very useful source for archaeologists to support the understanding of images in antiquity is ancient texts. This goes especially for sources that come later than the images concerned. This is 9

Dörpfeld 1892, Pl. 8. 9. Dörpfeld 1892, 4–27. 11 Dörpfeld 1892, 9 f. 12 It might also be that there were other reasons why Dörpfeld added the bench to his plan, as for example marks in the eastern wall or the like. But as far as I can see there is nothing that hints at such a bench. Also, even if there really existed a bench it would by no means be clear that it was erected for sitting on. A use of such a bench to display offerings or instruments connected to the cult could also be possible. 13 It is also important here to mention the names of Barthold Georg Niebuhr and Leopold Ranke, who have also played their part in the development of this important “technique.” 14 For example C. Wacker in his book on the Gymnasium of Olympia (Wacker 1996, 131–37) takes the account of Lucian for more or less granted, even though he himself states that this might not be a valid source. He even manages to give us a date for the performance without giving any references for that fact or its source. 10

168

Chapter Eleven

the case for example when Lucian tells us in “Herodotus sive Aëtion” 3, that the Erotes carrying weapons in the depiction of the wedding night of Alexander the Great and the Sogdian Princess Roxane, composed by the famous painter Aëtion, are symbols for Alexander’s ferocity and his desire for further warfare.15 This example has to be labelled as an ekphrasis, a description of – in this case – a certain piece of art, the genre that is most promising for furnishing relevant explanations, alongside such descriptive texts as Pausanias’ Description of Greece. In the latter kind of written sources we are often even told what the meaning of a certain piece of art was. But ancient literature is just as complicated as ancient images. This is why I would like to argue for the use of ancient literary sources in a less literal way, and for analysing and interpreting the texts and their historical and/or literary background. By doing so one might be able to understand possible implicit hints at pictures and monuments that could be hidden within them (fig. 1). All my ideas are based on a careful reading of Lucian’s text “Herodotus sive Aëtion” itself, as well as on an analysis of other texts written both by the author himself and others. I have also tried to integrate the archaeological record – in this case the architectural sculpture of the Western Pediment – into my argumentation. With this “way of interpretation,” I am following general assumptions that might be taken as probable – or not. The following is a rough model with which I would like to demonstrate what interpretations are possible. I would also like to underline the depth of the many “layers” of this kind of written source.

Multilayered Written Sources It seems that there are more than one of these “layers” within the writings of Lucian. The writer himself makes this apparent in one of his prolaliai, namely “Bacchus” (Lucian. Bacch. 5 f.), as well as in his speech “De Domo” (Lucian. Dom. 21). In the first case the Syrian Lucian alludes in an allegorical manner to the different kinds of listeners that come to his performances to hear different aspects of his texts.16 He also adds that sometimes the members of the audiences are surprised to find other 15

Of course Lucian would have been aware that this interpretation mirrors the historical facts: when Alexander married Roxane the Indian part of his campaign was yet to come. 16 He is telling the tale of the invasion of India by the army of Dionysus from the Indian perspective and uses the reported reactions by the Indians as metaphors for the reactions of different parts of his audiences, see Nesselrath 1990, 137 f.

Lucian on Herodotus

169

motives in his writings. This information is itself disguised as an allegory. When Lucian says that “if those that come to whatever performance, find iron instead of ivy [on the staffs carried by the entourage of Bacchus], they don’t dare to praise [it], confused by the incredibility of things” (Lucian. Bacch. 5), this has to be interpreted as saying that the listeners are surprised or even taken aback when they discover that his writings contain hidden serious topics and not just funny ones. The statement that texts are capable of possessing more than just one “layer” of content becomes even more clear when combined with the second example. It reads: The exactness of execution and the useful (elements) of ancient history [in the pictures] are truly of great interest and ask for well educated onlookers. And so that you are not only looking at them forgetting about us, I will as well as I can start to paint those topics with words for you. I believe this will cause you delight, listening to things you also admire when looking at them. (Lucian. Dom. 21)

The evidence that Lucian takes it for granted that different parts of the audience have different interests in his images as well as in performed talks, depending on their level of education, becomes very clear in this quote. Only those who know about the stories painted or told will find delight in them. This is a very complex way17 of extracting the “hidden” information from texts. But in my point of view it is a valuable task, as it delivers unique insights into ancient mentality. First of all the question arises how clues on such “hidden” information can be detected. At the very start there has to be a part that seems as though there is more behind it than just a mere mention of a certain fact. This always depends on the special knowledge of the reader. As I have mentioned before, I will focus on the accounts of Lucian in which he presents us with historical figures performing in the Opisthodomos of the Temple of Zeus at Olympia, namely Herodotus (Lucian. Her. 1) and Peregrinus Proteus (Lucian. Peregr. 32 f.; Fug. 1–2. 7). The reason why these two seem to be suspicious – at least to archaeologists – is the place of performance, specifically given as the back chamber of the most important temple of the sanctuary. But first of all let us take a closer look at Lucian’s “Herodotus sive Aëtion”:18 The structure of this prolalia generally consists of three distinctive parts. In the first, Lucian gives an elaborate description of the visit to the Olympic Games by the fifth-century BC historian Herodotus 17 18

As might become obvious from fig. 1. For a broader discussion see Nesselrath 1990, 117–22.

170

Chapter Eleven

and the performance of his Histories there. He praises Herodotus in almost every way possible (Lucian. Her. 1–3). The second part of the text can be identified as an ekphrasis of its own right, since it contains the description of a famous painting executed by the well-renowned painter Aëtion, already described above (Lucian. Her. 4–6). The third and last part gives the reader an account of the very situation Lucian himself is in at the moment of the speech’s delivery: he is performing right in front of the Macedonian assembly. He links his own situation to the description of Herodotus’ performance and draws parallels between them, therefore making clear that the writer of the Histories for him serves as some kind of role model. He hopes to be as successful as the “Pater Historiae” (Lucian. Her. 7–9).19 But the claimed fact of Herodotus’ performance in Olympia seems to be rather untrustworthy as concerns its historical value.20 As was said in the beginning, it is very interesting that the statement by Lucian is the earliest account of a performance of Herodotus in Olympia, although Lucian lived and wrote in the second century AD. It is easily imaginable that Herodotus had been to Olympia,21 but in my view Lucian’s account is insufficient as a proof. This is even more true for the second account of this performance. In the entry on Thucydides in the Byzantine Suda we can find the story of how Thucydides shed tears while listening to Herodotus in Olympia, which caused Herodotus to point out to Thucydides’ father Olorus that his son was bound to become a big name since he had so much passion for history (Suda 414 [Thucydides]). This very late account (tenth century AD) from the lexicon – which in this case gives no sources for its “facts” – seems to be a mixture22 of the description of Herodotus’ performance in Olympia given by Lucian and another very problematic written source, the biography of Thucydides by Marcellinus.23 This text is a late antique collection24 of three different texts written by three different authors25 that 19

This “design” of Herodotus as Lucian’s perfect role model has a big impact on the speculations which follow: I assume that the details of the performance were entirely made up by Lucian and that therefore the Olympia which has to be the “stage” for the following thoughts must be the Olympia of the second century AD – Lucian’s Olympia (see fig. 2). This is due to the fact that Herodotus’ performance had to follow Lucian’s needs in order to be an example for himself. 20 Cf. Schöll 1855, 411 (with older literature). 21 Jacoby 1913, 242, who in general doubts the historical value of Lucian’s account; Schöll 1855, 417; Dahlmann 1823, 18–20. 22 Schöll 1855, 410. 23 For the later sources most likely to be based on Marcellinus (and thus represented by this source here) see no. 6. 24 Bux 1930, 1451.

Lucian on Herodotus

171

has survived up until today as a Byzantine manuscript from the tenth or eleventh century AD (Cod. E, Palatina Heidelberg, 252)26 to name but one example. Even the time of creation of the texts is unclear, though it is speculated that they date from the fifth to the sixth century AD.27 Marcellinus28 gives us the slightly different account of the boy or young adult Thucydides listening to Herodotus, along with Olorus. In this case Marcellinus gives no information about his sources (he just states “there is a story”), even though he does give information on different occasions in the first part of the biography of Thucydides.29 He also does not give a location for this meeting of the two authors.30 The story itself makes the reader immediately think of Athens as a setting where both Thukydides and his father lived and where Herodotus is also believed to have spent time.31 It might well therefore be that the compiler of the Suda, who was definitely a very well-read man, was aware of the accounts by Lucian32 and Marcellinus, and blended them into one single event.33 After all, the Suda entry has to be considered as too late as well as too uncertain to have factual value for the interpretation of Herodotus’ performance.34 Coming back to the Suda’s account on the performance at Olympia: it mentions no location for the performance in the sanctuary. It is very unusual that Lucian should give an exact place where the performance was being held, when all other sources35 – at least those I am aware of – give no information about a specific place for orations at all, just like the Suda.36 In general it is rarely the case that sources tell us about writers or 25

Compare Bux 1930, 1453 f.; see also Maitland 1996, 538, 541. Bux 1930, 1450; for more codices see Maitland 1996, 538 f. 27 See for example Bux 1930, 1468. 28 As well as the later sources depending on him, see no. 6. 29 See for example the passage (§ 3) in which Marcellinus describes the linage of Thucydides, where he cites Didymos, himself using Pherekydes, as a proof for his information; for further information see Bux 1930, 1458. 30 Compare Bux 1930, 1478 f. 31 Even though this is most often taken for granted we have no direct “proof” for his stay at Athens: Jacoby 1913, 227. 237–246; Dahlmann 1823, 33. 32 Schöll 1855, 412. 33 On the imaginary character of these written sources see Bux 1930, 1478; for older literature on this aspect see Maitland 1996, 547 no. 39. For Marcellinus being a source for the Suda see Maitland 1996, 558. 34 This criticism of the Suda as a source for antiquity works for this very special case, of course. For other parts however the lexicon is a very important help in reconstructing the corpora of otherwise lost authors. 35 With the exception of Lucian and Philostratus, see below. 36 For the question of performances and other intellectual activities at Olympia in 26

172

Chapter Eleven

rhetores that performed in Olympia.37 This might be due to the fact that Olympia was never used for hosting events other than sportive competitions,38 and that therefore the Altis was not considered as an appropriate stage for the performative arts apart from notable exclusions in the later Roman Empire. We have to separate older sources (dating before BC) from younger ones when looking at the reports of performances that took place in Olympia. It becomes apparent that we only have two older written sources informing us about rhetores or people otherwise linked to public speaking being present at Olympia at all.39

Spoken Word Performances at Olympia The oldest account of the presence of a famous rhetor in Olympia is by Plato, who tells us about the visit of Hippias of Elis to the sanctuary but without giving further details (Plat. Hipp. min. 364c–364a).40 A more precise description of the events taking place in Olympia is given by Diodorus Siculus, when he writes about the show the tyrant Dionysius of Syracusae has set in motion (Diod. 14, 109).41 He mentions a number of rhapsodes citing aloud from the poems written by the tyrant himself. The description of the tents that have been erected for this spectacle in an open space within or close to the Altis is closely linked to the mentioned performances and could have been a suggestion for the place of performance by the author. But, again, a particular place in this case is not mentioned. Additionally one might also count Isocrates’ Panegyricus (Isoc. Ora. 4), which is fashioned as to be held before a great audience maybe in Olympia (Isoc. Ora. 4, 1), but it is not clear whether it was really performed there since we have no sources for this event of which I am aware.

general see Kokolakis 1988; Weiler 1997. 37 Herrmann 1972, 178. 38 See also Herrmann 1972, 178; compare with Suet. Nero 23. 39 For example we don’t know where Pindar performed his victory odes for the different panhellenic games. 40 A further setting for the dialogue itself is not given. What is stated in the dialogue is that before the men’s meeting Hippias spoke in a closed room in front of a huge audience (Plat. Hipp. min. 364b). In my view the festival at Olympia and Hippias coming from Elis to take part in it is mentioned here just to emphasize Hippias’ fame and offspring. 41 It might be that Diodorus Siculus is using older writings as sources for his information as he frequently did so with the writings of Ephoros. However this has to remain unclear.

Lucian on Herodotus

173

Looking at the more numerous but still relatively few more recent accounts of performances in Olympia, more or less the same pattern can be found. Lucian himself, even in fact in his “Herodotus sive Aëtion,” is to be mentioned first (Lucian. Her. 3). He is the only writer who gives us names of performers in Olympia, namely Hippias of Elis, mentioned before, Prodikos of Keos, Polos of Akragas, and Anaximenes of Lampsakos. All of these performances can be dated to the fifth century BC, except that of Anaximenes which dates to the fourth century BC. Lucian gives no indication for the places of performances42 even though he locates Herodotus very precisely. The same is true for the mention of a fictitious rhetor’s performance in Lucian’s “Pseudologista” – the man is performing at Olympia – but nothing more specific about the staging is mentioned (Lucian. Pseudol. 5). Dio Chrysostom’s “Olympian Discourse” (Dion. Chrys. 12) from the second half of the first or the beginning of the second century AD is staged in Olympia, and the text suggests that the performance was held before the big assembly and close to the Temple of Zeus (Dion Chrys. 12, 21). Still, it has to be mentioned that the setting could be a literary topos.43 A second writer from the third century AD mentions visits from prominent intellectuals to Olympia. Diogenes Laërtios states that the sage Thales has been to Olympia (Diog. Laert. 1, 39), as well as reporting that the death of the sage Chilon took place in Olympia (Diog. Laert. 1, 72). In both cases the visits are not explicitly linked to performances. This mode of report can also be found when he mentions Plato’s visit to the Olympian sanctuary (Diog. Laert. 3, 25). In addition, in his “Deipnosophistae” Athenaios also tells us about a performance of the “Katharmoi” of Empedokles by the rhapsodist Kleomenes (Ath. 14, 12).44 No place of performance is mentioned. The fourth-century AD writer Themistius also generally mentions the presence of speakers in Olympia without further stating which sites they visited (Them. Or. 18, 342). Reasons for these inexactnesses both in Lucian and the other sources mentioned may either be that those facts were just of no importance or – more likely – that everyone was aware of where to locate those activities in the sanctuary of Zeus. This maybe compared to modern examples: when opera singers are said to have performed in Milan it is clear that the exact

42 Except for the performance of Hippias, to my knowledge Lucian is our only source for these performances. As in the case of Herodotus it has to be doubted that Lucian is giving a historically correct account. 43 Similar to the case of Isocrates, see above. 44 Dahlmann 1823, 22 n. 27 b (with wrong quotation).

174

Chapter Eleven

place of the performance must have been the Scala.45 Maybe a further look at the “Pseudologista” of Lucian might give us a better hint (Lucian. Pseudol. 5). Lucian has the (fictitious) man performing in front of all the participants at the feast in Olympia without a further location being mentioned. In my view this might hint at the use of a special part of the Altis, the Theatron which is mentioned by Xenophon (Xen. Hell. 7, 4).46 This special location is mentioned by Xenophon in relation to the Stadion IIIa and the older “Schatzhausterrasse” (indicating a period in the first half of the fifth century BC47). It might nevertheless be that this particular area in whose midst48 performances took place retained its purpose in times when the mentioned stadion was long gone and the “Schatzhausterrasse” had already been altered, which of course also goes for the times of Lucian.49 I will return to this idea. Apart from the sources mentioned above that give no specific places for performances or do not mention performances at all, there is a slightly more recent writer than Lucian who gives a place of performance in Olympia: Philostratus, a very important source, as he is the author of the “Lives of the Sophists” and is well informed about their behaviour. He gives the unspecified “steps of the temple” as the place of performance by the saint-like rhetor Apollonius of Tyana (Philostr. VA 4, 31). He also gives us the very same position for the performances of Gorgias of Leontinoi, that must have taken place in the fifth century BC (Philostr. Epistula 73).50 How much Philostratus considers this placement on the steps of “a temple” as an honorary position is shown by another account. In his “Vitae 45 This is a very good example of the use of multiple layers of information today. Of course the hint at the Scala will only be understood by people familiar with the opera and its traditions. Therefore the information given tacitly by the speaker is also used to give credit to the speaker’s knowledge in a very sublime and rather sophisticated way. 46 For the existence of the Theatron and its positioning see Mallwitz 1981. The place might well have kept this function even though in Lucian’s time the spectators were positioned on the new “Schatzhausterrasse” or within the “Echohalle”; Sinn 1994, 589-592 (with older literature). 47 Mallwitz 1981, 104. 48 Or the free space facing it, including the Great Altar of Zeus. 49 For a more precise dating of the alterations in this part of the Altis see Mallwitz 1981, 103–5. 50 Of course this information has to be considered as not very trustworthy since again Philostratos was writing his letter more than 500 years after the death of Gorgias. And apart from that, the portion of the letter concerning Gogias has to be understood as praise for the author and might therefore exaggerate the facts.

Lucian on Herodotus

175

Sophistarum” Philostratus tells us about the dedicatory speech that Polemon delivered at the Temple of the Olympian Zeus at Athens that was finally completed under the reign of Hadrian (Philostr. VS 533). He also delivered this speech from the steps of the temple, which in this case can only be the steps of the entrance side. A similar position for Apollonius in Olympia51 is very likely to have been kept in mind by Philostratus. Apart from these elaborate positionings for famous orators, the written sources tell us of the appearance of lesser orators in the crowd in the sanctuary grounds, once for Olympia and once for the sanctuary of Poseidon at the Corinthian Isthmus. For Olympia it is again Diogenes Laërtios who paints a vivid picture of the buzz in the sanctuary during the games to which the loud shouting of the rhapsodes is added (Diog. Laert. [attributed to] Epistula 38).52 For the Isthmia it is Dio Chrysostom who gives a very similar testimony of what is happening around the Temple of Poseidon (Dion. Chrys. 8, 9).53 To sum up: the Opisthodomos of the Temple of Zeus is never mentioned as a location for the performative arts and the evidence for performances in Olympia at all is rather scarce. In addition we do not have even a single account that tells us that opisthodomoi54 were used for performances at any other place.55 The term “opisthodomos,” which is used rarely in written sources and inscriptions,56 is almost exclusively57 used in connection to its function as a treasury.58 The best-known example 51

The story may have taken place in that way, or not. Again, the problem is whether Philostratus gives us historical facts or is making up the stories about Apollonius. 52 Weiler 1997, 210 n. 239. 53 Weiler 1997, 197. 54 In the canonical sense of the word, denoting the back chambers of temples; Hollinshead 1999, 212; Coulton 1977, 191; Dinsmoor 1973, 393. 55 Hollinshead 1999, 210–14. 56 For example the decrees of Kallias, IG I³ 52A, B; Hollinshead 1999, 210 f. For discussion of the identification of this structure see also Hollinshead 1999, 211, 213 (with older literature). 57 For a very interesting case of this rule being broken see Plut. Demetrios 23, 3, describing how Demetrios Poliorketes inhabited the Opisthodomos of the Parthenon, as well as Diod. 14, 41, 6, who mentions that Dionysios of Syrakus had weapons constructed even in the opisthodomoi of the temples; Hollinshead 1999, 212 (with older literature). It must be kept in mind that it might be that these sources are following a tradition of criticizing both rulers and are therefore not necessarily to be taken as facts. 58 For the use of opisthodomoi see Ach. Tat. 3, 6, 2 f. who states that a painting was to be seen in an opisthodomos from the outside, and Dion. Chrys. 11, 45 who tells

176

Chapter Eleven

for this is of course the Opisthodomos of the Parthenon on the Athenian Acropolis.59 If a use like this was the rule, as it seems to have been, we could also speculate that the Opisthodomos of the Temple of Zeus at Olympia could be closed60 in a similar way as its Pronaos. It would therefore not have been suitable to host audiences. So how come Lucian gives this detailed but suspicious place for the performance by Herodotus?

Lucian’s Characters: Herodotus and Peregrinus A closer look at the text by Lucian tells the reader that the “Herodotus” described by the author61 is made up entirely to serve as the perfect role model he is comparing himself to.62 He draws parallels between the situation of Herodotus, who is to perform for all the Greeks assembled in Olympia, to his own performance in Macedonia in front of the Macedonian assembly,63 as has already been pointed out.64 In this attempt to make “his” Herodotus suitable for comparison he even changes parts of Herodotus’ biography.65 He states that Herodotus came straight from Caria to Olympia after having finished his works. As far as we know or assume, Herodotus was working on his Histories up to an old age and might indeed never have finished them.66 Even more, it is quite likely that he spent some time in Athens after he left Caria.67 So it might well be that Lucian is not giving facts but a literary topos both in case of Herodotus’ biography as well as in the case of his performance. The two other writings in which Lucian has a speaker positioned in the Opisthodomos concern the cynic Peregrinus Proteus. From the writings us about the Chest of Kypselos being on display in the Opisthodomos of the Temple of Hera at Olympia; Hollinshead 1999, 213. I do not however agree with her view that opisthodomoi were used for recitation, as the only examples she gives are the three accounts by Lucian for Herodotus and Peregrinus Proteus (and the heralds) which, as I try to show here, are not fact but fiction. Therefore they are not to be used to identify the function of opisthodomoi. 59 Hollinshead 1999, 2011 f. 60 For the responsibility of closing and sealing the Opisthodomos on the Athenian Acropolis see again decree A of Kallias mentioned above, n. 56; Hollinshead 1999, 210 f. 61 Cf. Dahlmann 1823, 18. 62 See also above and n. 19. 63 Cf. Dahlmann 1823, 18. 64 See also Nesselrath 1990, 118. 65 Compare for example Jacoby 1913, 225; Schöll 1855, 413. 66 Jacoby 1913, 222. 232. 372–79. 67 See n. 31; for the direction of the “Histories” at Athens see below.

Lucian on Herodotus

177

themselves it is more than clear that Lucian despised the man who was his contemporary.68 The trait of Peregrinus’ character, which is to be considered as the most important one here, is that of the imposter. Whenever Peregrinus has the opportunity, he copies the deeds of others, which is the reason for his nickname Proteus, deriving from the mythical sea god Proteus as he is already described by Homer (Od. 4, 349–570). This Proteus is capable of changing his physique into whatever form he desires – just as Peregrinus himself copies his more glorious predecessors in word and deed.69 This eagerness to live and die after the model of others is most clear in his wish to die on the pyre like Heracles. But he fails in this death wish, as well as in his possibly hidden wish to be like Herodotus: his death is no more than a mere joke, and he cannot deliver his speech to its end because his voice breaks down (Lucian. Peregr. 33). It would therefore fit Lucian’s line of thinking very well to have an intertextual reference inserted here. Accepting this theory, Peregrinus would then pose as Herodotus by standing in the exact same spot as he did while performing. This would have of course only been understood by those people familiar with Lucian’s other works and would have therefore functioned as an additional “layer” of information to the text. Lucian might well have given us two hints that the place of Peregrinus’ performance was made up. The first hint would be the utterance of the “Lucian” in his “De morte Peregrini,” an eyewitness to the performance. He tells his readers or listeners70 that after Peregrinus had died he met several people on his way back to Olympia. They were asking him what had occurred and so he answered truthfully to those who seemed to be smart and added made up facts for those he considered stupid. In this manner he sets into effect another “layer” of the text: the audience now has to decide for itself whether it belongs to this sort of people or the other and therefore decide whether to accept Lucian’s story as fact – or not. The second hint is of a more factual nature. In the same text Lucian tells us71 that Peregrinus’ talk was held after the contest of heralds that also took place in the Opisthodomos. Pausanias gives us a very specific place for this contest, a Bomos close to the entrance of the stadium (Paus. 5, 22, 1).72 Coming back to the assumption that this area was once part of 68

For example Lucian.Peregr. 1 f. See for example his Christ-like behaviour or his attempt to follow the saint-like Egyptian (Lucian. Peregr. 11. 17). 70 Lucian. Peregr. 39. 71 Lucian. Peregr. 32. 72 See also Pilhofer u. a. 2005, no. 108 (P. Pilhofer). 69

178

Chapter Eleven

the Theatron mentioned by Xenophon (see fig. 2),73 this position seems to make sense for heralds who had to give their talks in front of huge audiences.74 Due to this fact Pausanias’ account has to be given priority over that of Lucian. And if he was making up something in the heralds’ case, why not also in the case of Peregrinus? Following this question and the foregoing argumentation, the first reason for the invention has to be: the place of performance is an invention of Lucian in all three cases, even though all three texts each represent different genres of writing. And if they were inventions, for what reason did Lucian make them up? What can be stated is that if this information about the two speakers (Peregrinus mentioned twice) standing inside or in front of the Opisthodomos is fictitious, every person from antiquity interested in culture would have instantly known – for Olympia and its festivals were still very famous, especially in the Greek world.75

The Possible Illustration As mentioned before, images possess different meanings and so do literary texts. The writings may be full of hints at political circumstances, other works of literature – which is very common especially for quotations or quasi-quotations of the Homeric Hymns – intercorporal references,76 other authors’ styles of writing or public speaking, and so on. Of course these “layers,” as one might call them, are different from author to author. To see whether a writer is really using different “layers” to communicate different aspects of his topic to different parts of his audience, the whole corpus of his works has to be analysed. This must be done in order to see whether it could be possible that there is more behind such an invention. As we have already seen, Lucian’s writings are full of hints directed at a well-educated and well-read segment of the audience. Another very good example is the following comparison from his writing “De Saltatione” (Lucian. Salt. 50): “And also when you come to 73 The plan given has been created after Herrmann 1972, 129 (alterations partly after ideas by the author, created by H. R. Goette, DAI Berlin). To mark a later place for performances, the 3rd century Odeion (as shown on the sanctuary plan by K. Hermann in Olympia 1875 – 2000 [attachment]) has been added to it. 74 For the acoustical problems that would come with a recitation of the Herodotean “Histories” if held in an acoustically non-suitable place see Dahlmann 1823, 21– 23. Dahlmann understands these difficulties as the reason why Lucian positioned “his” Herodotus on the steps on the back of the temple. 75 Cf. Dahlmann 1823, 18 f. 76 As has been shown for the case of Peregrinus and the possible hint at Herodotus.

Lucian on Herodotus

179

Aitolia, the mimic dance also covers many topics there: […] (and) next Nessus and the jealousy of Deinaneira that caused the pyre on the Oëta.” He names this pair of mythological figures (as well as the other figures) to prove that a mimic dancer had to know all the characters from the different Greek mythological landscapes to have them in his repertoire. Not only does he mention them, but he also hints at the deliberate death of Heracles on the pyre, where he put his life to an end after the cloth with the dead Centaur’s blood, given to him by the jealous Deianeira, had intoxicated the hero. This is clearly a different “layer” of the text, as this information again only makes sense for those who are familiar with the myth. All the other people in the audience would simply not understand this information. Also his hints at the Homeric Hymns and the works of Thucydides and Herodotus77 are enormous, as can be seen from various instances throughout all of his works.78 So it is plausible that he made these scenes up for special reasons, which he wanted to be understood by whoever was able to get the hint. But what could he have possibly been hinting at? The most significant feature of the western portion of the Temple of Zeus is undoubtedly the pedimental sculpture group that depicts the fight of the Lapiths and Theseus with the Centaurs at the Wedding festivity of Peirithoos, King of Lapiths.79 Could there be a relation between the possible invention of Lucian and this picture? To answer this question one has to turn once again to the opus of Lucian as a whole to see whether this reading would fit the author’s line of thinking. We have already seen that Lucian sees pictures as being able to symbolize certain topics that are not literally part of the picture but included in them as metaphors, allusions, or allegories. If one considers the pedimental group, which allusions could possibly be seen in them? First of all, the fight of the Centaurs against the human Lapiths, who had invited them for the sacred wedding festivity but were attacked by their relatives, is a very obvious example of hubris.80 Hubris denotes – in this 77

See below for the example of the “Dipsades.” For the most significant example of the style of writing and the overall design of his writing of “Dipsades,” which is clearly a sort of “homage” on the different Logoi of Herodotus, see Nesselrath 1990, 122–25; compare also n. 91. 79 Following the description of the group by Pausanias (Paus. 5, 10, 8), now widely been accepted to be true. The most recent article (Kyrieleis 2013) on the sculptures of the Temple of Zeus and therefore also the western pediment gives a very good overview of the research on the sculptures, as well as expressing opinions on especially the western pediment that are shared by the author. 80 Cf. Barringer 2008, 18. 78

180

Chapter Eleven

sense of the word – the crossing of divine or natural borders by humans. This is considered to be sacrilegious. In my view this is also the reason why the god Apollo is depicted in this scene, as identified by various archaeologists.81 One of the primary topics of Herodotus’ Histories is also hubris: for example, Kroisos crossing the Halys and as a result losing his kingdom has to be understood as an example of hubris (Hdt. 1, 46–56. 72– 91). This goes also for Xerxes crossing the Hellespont (Hdt. 7, 33–37) and finally being defeated by the Hellenic League. But in Herodotus there is also a detectable warning: a warning towards the Athenians,82 who have to be seen as the primary recipients of the Histories:83 if they should further follow their hegemonial politics they would fail and fall, as did Kroisos and ultimately Xerxes. Both aspects might have been visible in the pedimental group. Apart from the “obvious” hubris there is also a very clear hint towards Athens: Theseus is a very prominent figure of the group and is at the time of the artwork’s origin in this specific mythological context very likely a symbol for Athens.84 This goes also for the times of Lucian – as is shown by Plutarch’s description of the life of Theseus. It has also been argued for a long time that the pediment, as well as other depictions of this type of Centauromachy, were understood as allegories of the Persian Wars.85 Erika Simon86 argued that this reading would be impossible, as Centaurs are related to the Lapiths and are therefore not suitable for an allegory of the Persians. Still, she omits that there were many Greek poleis fighting with the Persians against the Hellenic League, a fact that caused a lot of dispute in the Greek world in the aftermath of the Wars – in fact, the court which decided on the punishments for these cases of medismos had been installed in Olympia.87 If it were true that the depiction in the pediment could be seen as an allegory for a specific aspect of the Persian Wars and the prominent role 81

For the discussion of the identification of this pedimental figure see Sinn 1994, 593–96 (with older literature); see also Kyrieleis 2013, 95 f. 82 For a discussion of the link between Herodotus’ “Histories” and Thucydides’ “Peloponnesian War” and the implied warning for Athens see Forsdyke 2006; Raaflaub 1987. 83 Jacoby 1913, 227. 237–42. 84 See for example Kyrieleis 2012/13, esp. 105. For an extensive argument concerning this interpretation that cannot be part of this chapter see my forthcoming dissertation, which will deal with the phenomenon of public and official depictions of the hero outside Athens. 85 For a short overview see Kyrieleis 2012/13, 94 f. 86 Simon 1975, 118. 87 Kyrieleis 2012/13, 102; cf. Sinn 1994, 599 f. (with older literature).

Lucian on Herodotus

181

Athens played in them, it would have been the perfect backdrop or even illustration for the performance by Lucian’s “Herodotus.” This would of course only have been understood by those familiar with the different layers of the Herodotean Histories. That Lucian was aware of these different “layers” of the Histories is obvious: such accounts as the “Dipsades” show that Lucian was very familiar with Herodotus and his work, as he even imitates his style and often refers to his work, as pointed out earlier.88 Also, it is just as obvious that Lucian was aware of the allegorical functions of pictures, as is shown by his explanation of the meaning of the aforementioned painting by Aëtion. In this new context it is quite interesting that Lucian alludes to the picture at all, since it does not necessarily add to the three-part prolalia with (1) the encomion on Herodotus, (2) the seemingly rather unrelated description of the picture,89 and (3) the comparison of Lucian’s own situation to the role model of Herodotus. The only function that the encomion might have is that of a link between the parts, but not even a necessary one. Lucian presents a painting of Macedonian content that was somehow linked to Olympia in the past, to bring together Herodotus’ place of performance and his own audience. It could also well be that this explanation of the picture serves as an explanation of the “hidden” illustration of the Herodotean Histories. This would also fit the title of the speech: “Herodotus sive Aëtion” suggests, as do the other titles of speeches by Lucian in that manner,90 that both parts of the title connected with the particle “sive” (Greek “ ”) have something in common, even that in fact in a way they denote things that can be seen as alternatives for one another. So far this is only a theory that is not verified. But again a closer look at the writings of Lucian could be helpful to see whether this reading could have possibly been intended by Lucian. Lucian twice mentions Herodotus in the same context: he quotes him once (Lucian. Salt. 78), and also has him addressing the audience in person by summoning his ghost (Lucian. Dom. 20),91 to the effect that in the question of whether things that can be 88

Nesselrath 1990, 122–25. Nesselrath 1990, 119. 90 For example (following the Greek titles of the same structure closely) “Charon sive Contemplantes,” “Revivescentes sive Piscator,” “Philopseudes sive Incredulus,” “Somnium sive Vita Lucani” etc. 91 Actually those are allusions to Hdt. 1, 8, a fact that only someone truly familiar with the writings of Herodotus would have recognized; see the translator’s insertion at Lucian. Salt. 78 in the German translation by August Friedrich Pauly in 1827 (Pauly 1827), as well as the note to Lucian. Domo. 20 by the same author in 89

182

Chapter Eleven

seen are to be preferred to those that can be heard, images would have to be given the advantage. Thus it becomes clear that he perceives pictures to be more powerful and more persuasive than words. This clearly shows that Lucian’s perception of Herodotus is that of a person who would use visual features to underline his oral performances, as these illustrations would help to prove his point with their power and persuasiveness. In “De Domo” Lucian argues for the advantages of talking in a visually stimulating setting – as would have been the space right under the West Pediment. To bring this chapter to an end I swiftly come back to Peregrinus Proteus. There could possibly also have been a link to the pediment hidden in the two accounts mentioning him: Lucian describes the behaviour of Peregrinus as pure hubris, as it is also depicted in the pediment. Additionally, as I have mentioned before, the agony that leads Heracles to burn himself alive on the pyre is caused by Centaur blood – that flows in streams in the pediment. A possibly even more complicated system of references might have been the intention of Lucian in his satire “Fugitivi” (Lucian. Fug. 7). Here it is the description of the “false” philosophers that as Cynics “bark” in the Opisthodomos that is to be paralleled with the author’s own general comparison of these wannabe philosophers with the Centaurs at the Wedding festivity in his “Symposium.”92 Again this would have to be labelled as an intercorporal reference, whether the “Fugitivi” hints at the “Symposium” or the other way round. Also, the figure of Heracles in this writing and the account of his deeds is surprising, especially the cleaning of the stables of Augias. This deed has always been located in the Pisatis and is thus strongly linked to the polis of Elis that was in charge of Olympia and the surrounding landscape of Elis. The places where the deeds of Heracles have been depicted in Olympia have been the Pronaos of the Temple of Zeus – and its Opisthodomos. I am not implying that this interpretation has necessarily to describe Lucian’s intention – but it shows how much potential for interpretation is to be found in ancient written sources. And just a small addition: Lucian would certainly not have been the first writer who might have used an obvious “mistake” in his text to hint at something else. In his “Heracles” – a tragedy we have to assume to have been known by Lucian – Euripides once locates the Centauromachy of Heracles correctly in Arcadia93 but gives its location at the second instance

1831 (Pauly 1831). 92 In the Greek original: 93 Eur. Herc. 182.

.

Lucian on Herodotus

183

as Thessaly94 – the location of the Centauromachy fought by Theseus.95 As Theseus is at least the second hero of Euripides’ story and at the time highly prominent in Athens, it might well be that this “mistake” was intentional to point at the Attic hero.96 This could have been a move not too different from what might have been intended by Lucian with the hint at the depiction of exactly the same Centauromachy at Olympia.97

Conclusion In the first part of this chapter it has been shown that the performance of Herodotus in the Opisthodomos of the Temple of Zeus at Olympia as told by Lucian (Lucian. Her. 1) is very likely an invention. To underline this point, the very limited number of ancient written accounts of spoken word performance at Olympia have been discussed as regards the places of performances mentioned in them. The result was that in general they give no locations at all. The only exceptions to this rule are Lucian himself with the positioning of Herodotus and Peregrinus Proteus in the opisthodomos of the temple, and Philostratus’ accounts of the speeches by Apollonius of Tyana and Gorgias of Leontinoi held on the steps of the temple. While for the latter accounts, the location is very likely a position of honour, the locations given by Lucian were analysed against the broader background of Lucian’s works in general. This analysis showed that the setting of the speakers might have served special purposes for Lucian. He could have embedded a further “layer” of content to his “Herodotus sive Aëtion” – a hint at the desire by Herodotus to illustrate his performances with images. He would also have been able to make an intercorporal reference to exactly this passage by having the impostor Peregrinus standing in the same spot while trying to perform just like Herodotus without having to mention the historian. The illustrative qualities of the Western pediments’ sculptures have also been found to be multilayered, as they fit different aspects of Herodotus’ Histories.

94

Eur. Herc. 373. Papadopoulou 2005, 161 f. 96 Papadopoulou 2005, 162. 97 Compare the relation between literature and monuments in Papadopoulou 2005, 161 f. (with older literature). 95

184

Chapter Eleven E

Fig. 1: Scheeme of Proceeding for the use of writteen sources in Classical Archaeologyy (by the authhor)

Lucian on Herodotus

185

Fig. 2: Portion of the Plan (to the north) of the sanctuary of Zeus. The space marked by the letter “H” gives an idea of the possible place of the Bomos of the Heralds mentioned by Pausanias.

PART IV DIGITAL TECHNOLOGIES

CHAPTER TWELVE THE EAST PEDIMENT AND THE TEMPLE OF ZEUS RECONSTRUCTED IN VIRTUAL REALITY ANDRÁS PATAY-HORVÁTH

Abstract The digitization of the extant fragments and a three-year project enabled the production of a virtual 3D reconstruction of the east pediment and its sculptural decoration. In addition, the complete temple as well as the famous statue made by Pheidias were also reconstructed virtually, making thus the visualization of the long and complicated history of the entire monument possible. The model is highly flexible and can thus be adapted to illustrate and to test different scholarly hypotheses concerning some details, e.g. the arrangement of the central group of the east pediment or the effects of different lighting conditions. It also allows the non-specialist user to manipulate the individual pieces of sculpture, to familiarize him- or herself with their original appearance and position on the building and finally to observe minor details and to learn more about the problems involved in reconstructing ancient works of art.

This chapter is a summary report on the results of a three-year project concerning the virtual 3D reconstruction of the east pediment and the entire temple.1 Figure 1 shows a traditional 3D model of the temple and below it the virtual one, and although it is not apparent at first, it is absolutely clear that the virtual one is far more useful. First of all, it is flexible and can be adapted to represent the most recent results of research and offers

1

Most important recent publications on this subject by the author include: PatayHorváth 2011b, 2011c, 2013.

The East Pediment and the Temple of Zeus Reconstructed

189

additional features, which would hardly be feasible with any traditional model. The best illustration of the possibilities is the simulation of the lighting conditions in the interior of the temple. Figure 2a shows a recent attempt without the help of a digital model: sunlight comes only through the doors, falls on the shallow pool filled with olive oil in front of the statue and illuminates just the footstool and the feet of the god.2 Figure 2b shows the same in the virtual 3D model. As anticipated, this arrangement would not be particularly satisfactory, because the upper part of the statue remains dark. So another hypothesis was put forward assuming a hole in the roof (Fig. 3a). The actual remains of the temple do not endorse this idea, and such a solution would be not only singular but not even satisfactory, since the head of the statue would still remain in the dark. The placement and the measurements of the pool in front of the statue are absolutely certain, but as shown by the digital model (Fig. 3b), this cannot effectively be used to illuminate the head of the statue. The third and best illumination would be by direct light from above, a possibility that is perfectly feasible, if we suppose an open ceiling, because the rooftiles made of white marble are translucent3 and were most probably employed exactly in order to achieve this lighting effect. This possibility (Fig. 4) was not considered earlier, and was impossible to test without a digital model, but is actually so attractive that even A. Hennemeyer, who previously opted for the hole in the roof, has recently changed his mind and accepted it as the most probable reconstruction.4 The virtual model can thus be used not just for visualizing research results but also to test earlier hypotheses and to improve our understanding of the monument. This is actually even more true in the case of the pedimental sculptures. The initial problem, the arrangement of the five central figures of the east pediment, is a complicated question, and as shown in Fig. 5, there are basically four different arrangements, all of which have already been selected by certain archaeologists for various aesthetic, technical or other considerations. The position of the male figures is indicated by the terms closed or open, the arrangement of the females by Type „A” / „B”. Despite the great number of studies, and two monographs exclusively dealing with the east pediment,5 no-one was able to present a fully 2

Hennemeyer 2011, 101–104. Ohnesorg 1993, 118–119 with Plin. Nat.Hist 36,46. 4 Hennemeyer 2012, 123; 2013b, 4-5. 5 Säflund 1970. 3

190

Chapter Twelve

satisfactory and convincing reconstruction. It is characteristic of the situation that a pair of renowned English and Greek authors presented two completely different reconstructions side by side in the same volume on the sculpture of the temple.6 During the 1970s and 1980s archaeologists apparently grew tired of a seemingly unproductive debate and gradually agreed on a reconstruction (open Type “A”: Fig. 6), which was proposed by a few authoritative scholars supporting their view by different considerations.7 But a major methodological problem persisted. In general, scholars were accustomed to discuss the reconstruction and the interpretation together and the reconstruction was normally adapted to the interpretation, which is logically the wrong way, of course; evidence, which could be used to establish the correct reconstruction independently from the interpretation, was usually neglected. Actually, of all the arguments in favour of this reconstruction, there is only a single one which is independent from the interpretation given by Pausanias (5.10. 6-7). It is a principle of supposedly universal relevance, but is formulated in slightly different ways by its proponents. "Following the iconographic principles of other pedimental compositions, Zeus does not actually partake in the mythical event, but is to be imagined as being invisible to the participants. For this reason it is unthinkable that Oinomaos and Pelops are turned toward him as if in a "sacra conversazione". Instead, each must be turned slightly away from Zeus..."8

This principle concerns the invisibility of the central god and is corroborated by referring to the west pediment of the same temple and both pediments of the Aphaia temple on Aegina. However, basically all of them are depictions of battle or combat scenes and in these cases the victorious warriors naturally turn against their defeated enemies and not towards the central god(dess). But the east pediment simply does not depict a battle or combat, so these parallels are not especially relevant. I also refer to some vase paintings, where the central god is clearly meant to be invisible, but the human figures are turned towards her.9 The 6

Ashmole - Yalouris 1967. Simon 1968, Herrmann 1972, Stewart 1983, Kyrieleis 1997. 8 Kyrieleis 1997, 14. 9 All the depictions of the dice game between Aias and Achilles, which contain a central image of Athena, show the protagonists turned towards each other and the goddess. The name vase of the Nausikaa Painter is also an obvious example for mortal figures turned towards an invisible goddess. On the subject of the invisible gods see also the discussion by T. Osada (chapter seven of the present volume). 7

The East Pediment and the Temple of Zeus Reconstructed

191

position of the protagonists obviously does not depend on the visibility of the god, but is determined by the action. The same is to be expected in a pedimental composition as well. Another, more general formulation was also adduced for the open arrangement of the central group. "The principle of late archaic - early classical pedimental composition, as applied on the west pediment, on Aegina, the Parthenon and elsewhere, that a converging movement from the wings to the center should be balanced by a centrifugal movement operating from the center outward." 10

In this case, as an additional piece of evidence the west pediment of the Parthenon is cited, but even in this form, the principle does not seem to be relevant for our problem, since in the centre of the east pediment there is no movement at all, neither converging nor centrifugal. There are only static figures standing in a rigidly vertical position and in such a case, there is actually no point in comparing the compositions mentioned so far. The pediment of the Alcmeonid temple in Delphoi would be a much better comparandum and clearly shows that a figure turning towards the centre is not inconceivable. All the other considerations which are independent from the interpretation actually point to the closed arrangement and especially to Type A, which is known as the Studniczka- Buschor version. Three major points (optical corrections, findspots and the indications „to the left/right of Zeus” given by Pausanias) were already discussed in an earlier paper11 and I will therefore concentrate here only on the 3D reconstruction of the statues and their original architectural framework. Plaster casts and models belong to this category of 3D reconstruction as well and their history should be discussed here too. The fine coloured models in Fig. 1 were produced on a miniature scale in Berlin by the sculptor H. Grüttner immediately after the discovery of the fragments. They do not represent a final, elaborate version of the reconstruction, but just a first attempt. That they are in fact inaccurate was already realized by G. Treu, who replaced them soon, already at the end of the 19th century, by real-size models, which were made by using the plaster casts of the fragments and by restoring the missing parts in plaster as well. The most important result reached by G. Treu, who spent more than a decade on experimenting with them, was that one of the four possibilities (open Type „A”) was definitively ruled out purely because of the lack of 10 11

Stewart 1983, 135. Patay-Horváth 2007. Cf. also chapter six in this volume.

192

Chapter Twelve

space. Treu stated explicitly that figures G and K cannot be placed next to each other, because their arms would come into collision. Obviously enough, he was absolutely convinced that this arrangement is physically impossible and invited everybody to verify this statement with the life-size plaster models. 12 This has been done by various scholars following him, and no one questioned this observation, even if both Studniczka and Bulle arrived at an arrangement which is different from the one suggested by Treu.13 Thus an absurd situation emerged: today the most widely accepted reconstruction (open Type “A”) is precisely the one which was deemed technically impossible by Treu. Obviously, this would not present a problem if Treu’s results had been thoroughly tested and clearly refuted, i.e. if anyone had showed that Treu had experimented with ill-restored models or had come to wrong conclusions for some other reason. Instead, everyone (except for the architect P. Grunauer14) has ignored his arguments and his results. It is interesting to note that G. Treu advocated the arrangement widely accepted today, while he only had the miniature models at his disposal, but later his experiences with the life-size models made him change his mind. This was the reason to investigate the problem in more detail. A project making use of the latest technological innovations started by creating highly accurate virtual 3D models of the statues: the extant fragments were scanned in 3D and the missing parts were modelled virtually. The genesis of a single model is shown here in Figure 7, the model of the entire pediment in Figure 8. (Original and reconstructed parts are clearly differentiated by their colour.) The difficulties resulting from the large size and the unmovable placement of the fragments near the wall were overcome in most cases very successfully and all the figures of the pediment (including minor fragments in the storerooms) were scanned in two weeks.15 Some rear parts, however, proved to be entirely inaccessible for the scanner. As these parts were in most cases only roughly hewn from the 12

Treu 1897, 120: „Sodann ergeben sich bei einer Aufstellung von K* neben Pelops unüberwindliche räumliche Schwierigkeiten. Es wird, wenn man Pelops die richtige, durch die Rückendübel angezeigte Dreiviertelsdrehung zur Ecke hin giebt, unmöglich, seinen Speer an dem schleierfassenden linken Arm von K* vorbei zu bringen. Davon überzeugt ein Versuch mit den Abgüssen in dem richtig gebauten Rahmen ohne weiteres.” 13 Studniczka 1923, Bulle 1939. 14 Grunauer 1981, 287-288. 15 Patay-Horváth 2011a.

The East Pediment and the Temple of Zeus Reconstructed

193

block, their exact rendering is actually irrelevant for the reconstruction. Moreover, they are sufficiently documented in drawings and photographs, and could therefore be approximately completed during the processing of the scans. Modelling was done by a sculptor and by an architect-designer independently from each other. Both were instructed and supervised by the author, but both of them used different software packages and adopted different methods for the reconstruction of the missing parts. The resulting models were, however, not substantially different regarding their poses and their proportions.16 The pedimental frame was created in 3D max Studio and the completed models of the statues were inserted into this frame in order to test the feasibility and the aesthetic effects of each possible reconstruction. The advantages of the digital model are, I think, obvious. One can e.g. adopt the position of an ancient viewer standing in front of the temple, or remove the architectural frame and look at the statues from above. Switching between these views is a matter of seconds and experimentation with the exact position, rotation of the figures, takes not more than a few minutes. To change the pose or rendering of the reconstructed parts takes a bit more time, but is far easier than modelling them from plaster or other real materials. So one is able to test hundreds of possibilities in a reasonable time. According to the experiments carried out with the digital 3D models, one can finally conclude that the reconstruction which is most widely accepted today (open Type „A”), is technically the most difficult to realize and therefore not really probable. The most problematic part, described by Treu, where the spear of the warrior and the shoulder of the lady come very close to each other, is shown in Figure 9 from different angles and is highlighted in red. The problem is, however, not as compelling as Treu stated, since the collision might be avoided by a slightly different positioning of the right arm of the warrior. The real problem is related to the spears, and is an iconographic one: both open arrangements would be feasible only if we ignored a general convention of ancient Greek art regarding the way of holding a spear: spears are usually much taller than the human or divine persons holding them and the figures normally grasp neither the tip of the spear nor the shaft directly below it. The length of the spear and the handling dictated by the open arrangement is therefore not only awkward, but hardly attested.

16

Patay-Horváth 2011b.

194

Chapter Twelve

The only example, which can be adduced as a parallel, and was already mentioned by Treu, is a grave stele from Thessaly,17 a provincial work of art, which can scarcely prove that the spear was rendered in this way on the pediment as well. The fragment, which Treu interpreted as a hand fragment holding the tip of the spear, is known only in a simple sketch published by him.18 No photograph has been published afterwards and the fragment seems to be lost. At least nobody was able to locate it in the museum of Olympia and therefore it was impossible to scan it. Even if it belonged to figure G of the pediment, it is by no means apparent that it is intended to depict the spearhead. So, contrary to the opinion of Treu, it does not suggest the open arrangement and the closed arrangements should be clearly preferred. By considering every piece of available evidence (e.g. find circumstances), one can finally conclude that the most probable reconstruction is the closed arrangement Type „A”. This is, at least my conclusion. Paraphrasing what Treu wrote more than a century ago, I hereby invite everybody interested in the subject to test this conclusion with her/his own eyes by looking at the 3D models on the screen.19 One problem should be discussed in addition.20 The scanning of the miniature plaster models actually confirmed Treu’s statement regarding their inaccuracy. If scaled to the original dimensions and compared with the digital models of the fragments, it is clearly visible that the differences are sometimes quite remarkable, especially in the case of the male figures. Treu's conclusions based on the long experimentation with the large plaster models were on the contrary still puzzling, since they were markedly different from the results achieved with the digital models. It could be easily demonstrated that the difference was not due to the fact that the pediment was in Treu’s day reconstructed with slightly different (smaller) dimensions, because the same digital models placed in the virtual reconstruction of the pediment using the former, smaller, dimensions yielded the same result as with the recent ones. The discrepancy was thought to be caused by the slightly different rendering of some 17

Treu 1897, 47; Biesantz 1965, 16 Kat. No. 26. The relief dates from the second half of the 5th century BC and is now in the National Archaeological Museum of Athens (No. 734). 18 Treu 1897, 47 Fig. 59. 19 They are all included in 3D pdf format on a CD-ROM (Patay-Horváth 2011c) and are also available online (Patay-Horváth 2012c). Users may observe the models from any point of view and enlarge any part of them. 20 For a more detailed discussion see Patay-Horváth 2012d.

The East Pediment and the Temple of Zeus Reconstructed

195

reconstructed parts (or because of different poses adopted for some limbs) and to test this hypothesis, the preserved parts of the life-size models were also scanned in Dresden. Although regrettably little is preserved of them and even the preserved parts were subsequently altered by later scholars using them, the digitization could clearly show that the reconstructions realized in plaster on the original scale by Treu were not markedly different from those in the virtual reconstruction. So the discrepancy between the results cannot have been caused by the differences between the details of the two reconstructions realized in plaster and in virtual reality respectively. The explanation was to be sought outside the central group and beyond the reconstruction of the pedimental frame. During a final check of the whole reconstruction t eventually turned out that the different results were simply caused by the figures flanking the central group. In fact, the ones adopted for these positions by Treu (E and L) were markedly different from those generally accepted today (B and O) and affected the positions of the central figures by reducing the space available for them. Having realized this and adjusting the digital reconstruction in this respect to the one suggested by Treu, the same result emerged, as described by him: the collision of figures G and K became unavoidable. Treu’s statement was thus confirmed and explained without compromising either method or any of the resulting models. At the same time, the inaccuracy of the miniature models emerged again with a remarkable clarity: if arranged in the same way, they absolutely do not reveal the problem of the collision between figures K and G. In this way, recent technological innovations can contribute to solving an old puzzle of classical archaeology. In addition, the models of the individual figures can be used for further research and for visualization.

196

Chapter Twelve T

Fig. 1. Tradditional and virtual v 3D reeconstruction model of thee temple. (above: 1:100 plaster moddel from the end e of the 19tth century; peedimental figures madee by H. Grüttnner. Dresden, Albertinum)

The East Pediment and the Temple of Zeus Reconstructed

197

Fig. 2. Illumination of the chryselephantine statue through the door: (a) reconstruction drawing after Hennemeyer 2011, Fig. 1; (b) virtual 3D model

Fig. 3. Illumination of the chryselephantine statue through a hypothetical window in the roof: (a) reconstruction drawing after Hennemeyer 2011, Fig. 2; (b) virtual 3D model

198

Chapter Twelve T

Fig. 4. Illum mination of thee chryselephan ntine statue thhrough the tran nslucent marble tiles

Fig. 5. Concceivable reconnstructions of the central grooup of the easst pediment

T The East Pedim ment and the Tem mple of Zeus Re Reconstructed

199

Fig. 6. Openn arrangementt Type „A”: reeconstruction drawing afterr Herrmann 1972.

Fig. 7. Figurre G: original fragments (leeft), virtual 3D D model of thee fragments (ccentre), compllete virtual 3D D reconstructioon of the figurre (right)

Fig. 8. Virtuual 3D modelss according to the closed arrrangement Ty ype „A”

200

Chapter Twelve

Fig. 9. Figures G and K according to the open arrangement Type „A”.

CHAPTER THIRTEEN POLYKLEITOS’ WORKS “FROM ONE MODEL”: NEW EVIDENCE OBTAINED FROM 3D DIGITAL SHAPE COMPARISONS KYOKO SENGOKU-HAGA, YUJIN ZHANG, MIN LU, SHINTARO ONO, TAKESHI OISHI, TAKESHI MASUDA AND KATSUSHI IKEUCHI

Abstract Polykleitos was famed for his statues of athletes, and Pausanias names six such statues dedicated in Olympia. But his works were also criticized because they were “from one model” (Pliny nat. hist. 34.56). These words have usually been interpreted as referring to his monotonous style, but with the help of the 3D Digital Form Comparison Method, which enables us to distinguish millimetre differences, the authors here confirm that he literally used one model for the different statues. Our comparisons reveal that the sculptor used the model of the Doryphoros for the right foot and the face of the Diadoumenos, adding some changes in the latter case. Regarding the century-long discussion about which of the three Amazon statues was created by Polykleitos, Pheidias, and Kresilas but never securely attributed, we found that the face features of the Sosikles-type Amazon closely match those of the Doryphoros and the Diadoumenos, which means that it is the Sosikles Amazon that was created by Polykleitos.

Introduction Polykleitos was undoubtedly one of the greatest sculptors in Classical Greece. He is named as the creator of the Doryphoros, also known as the “canon” of the male figure (fig. 1). He was renowned as a sculptor of human figures and cast many bronze statues of victorious youths. But none of his original works have survived. Our knowledge of his oeuvre comes

202

Chapter Thirteen

from some descriptions in Roman literature, some original statue bases, and through a quantity of “Roman copies.” In Olympia, for example, Pausanias referred to five athlete statues made by someone named “Polykleitos.” But he also states that the statue of Agenor was made by “Polykleitos, not the one who made the statue of Hera, but the pupil of Naukydes,” namely a pupil of the great master’s pupil.1 Four bases bearing Polykleitos’ signature have been found in the excavation of the sanctuary at Olympia.2 In total, the evidence indicates six athlete statues made by someone named “Polykleitos,” of which at least one – and maybe more – have been attributed to a Polykleitos the Younger.3 Discussion continues, but in general three works – a young boxer Kyniskos, a pentathlete Pythokles, and a young wrestler Xenokles – are considered to be attributed to the great master himself given that their inscription dates correspond to his active period.4 If at least three of his statues stood at Olympia in the age of Pausanias, namely 600 years after Polykleitos’ time, many more must have existed all over Greece during his lifetime. According to Pliny (nat. hist. 34. 16–17), statues of all the winners in the Olympic Games were dedicated in the sanctuary, with life-size statues for those who had won three times. Pausanias (6.18.7) suggests that this custom originated in the second half of the sixth century BCE and that the production and dedication of bronze athlete statues flourished even more in the course of the following century, along with the development of bronze casting techniques.5 Given that Polykleitos was absolutely the most renowned andriantopoios among all of the sculptors of the High Classical period,6 we might well imagine that many athletic winners wanted him to cast their “own” statues. How did he create all of those athlete statues? The indirect lost-wax casting technique, in which original clay models made by artists themselves are preserved, is thought to have been invented in the late sixth

1

Paus. 6.2.6 (Antipater, a young boxer); 6.4.11 (Kyniskos); 6.6.2 (Agenor); 6.7.10 (Pythokles); 6.9.2 (Xenokles). 2 IvO 149 (statue of Kyniskos, but without the sculptor’s name); IvO 162, 163 (Pythokles); IvO 164 (Xenokles); 165 (Aristion, but the lettering of the inscription is dated the mid-fourth century BCE). 3 Ridgway 1995. As for the Polykleitos the Younger, see: Neudecker 2007b, Polyclitus 2 (with bibl.). 4 Ridgway 1995. 5 Mattusch 1988, 99–118. As for the dedications of athlete statues in Olympia, see also Barringer 2008, 48–52. 6 Aristot. Eth. Nic. 6.7 (1141a, 9–13).

Polykleitos’ Works “From One Model”

203

century and become popular by the middle of the fifth century BCE.7 Polykleitos was active specifically in this period and his works were, according to Varro, “quadrata” (square) and “paene ad unum exemplum” (Plin. nat. hist. 34.56). Usually these words are interpreted as referring to his monotonous style, as in “after the same pattern” (J. J. Pollitt), or “stereotyped” (A. Stewart).8 But literally it means that they “almost derive from one model” (B. S. Ridgway).9 Is it possible that Polykleitos created his statues “from one model”? Or, indeed, did he ever reuse one model for various statues?

Use of Models in Antiquity In the late Hellenistic and Roman Imperial periods – more specifically, from the first centuries BCE onward – sculptors used models to create both bronze and marble statues. Pasiteles, the famed and influential sculptor (Plin. nat. hist. 36.39), native of a Greek city in Southern Italy and active in Rome in the first half of the first century BCE, “said that the model (plastice) was the mother of relief carving, of bronze statuary and marble sculpture, and although he was excellent in all these arts, he never made anything before he had made a model” (35.156).10 It must not be a mere coincidence that one of the earliest “mechanical copies” in marble based on a Classical original, the Diadoumenos from Delos (fig. 2), dates to around the same period.11 Most likely Pasiteles was very positive about models because this new exact copying method had been just invented and introduced to Italy. He not only cast bronze statues by taking molds from models, he was also able to carve marble statues by mechanically copying the form of models. In fact, a quantity of ancient plaster casts found at Baiae reveal that some of them derive from famous Classical masterpieces.12 These casts, which are all section models and presumably taken with molds, part by part, from their originals, were kept in a sculptor’s workshop, evidently for 7

Mattusch 1988, 19–22, 105–6; Stewart 1990, 38–39; Hemingway 2000, 40–43. Pollitt 1990, 76; Stewart 1990, 264, T62. 9 Ridgway 1995. Borbein 1996, 70, translates it as “designed largely on the basis of a single model,” but he did not delve into this problem. 10 Borda 1953; Donderer 1988; Moreno 1994, 735–40; Fuchs 1999, 89–83; Neudecker 2007a. 11 Ridgway 1984a, 9; Niemeier 1985, 106–7. Pfanner 1989, 187–92, mentions an unfinished male statue from Rheneia near Delos (now in Athens, National Museum), which also dates back to the first century BCE. 12 Landwehr 1982; Landwehr 1985; Landwehr 2010. 8

204

Chapter Thirteen

repeated use. Thus, Roman sculptors could efficiently reproduce Classical masterpieces in their workshops by using sets of these section models. But they may have been used for more than just simple reproductions. Theoretically they could have used each section model independently to create variations or even a new eclectic work. Let’s consider the case of a bronze group, the so-called “Dancers” found in the Villa dei Papiri at Herculaneum, generally dated to the Augustan period (fig. 3).13 In 2007 we 3D-scanned these five female statues, and we confirmed that some of their feet have exactly the same form. (The rest of the feet, whose forms did not match at all, seem to have been deformed or restored in later periods.) At least three – or maybe all of them were cast from a single pair of foot models, which means the sculptor used single foot models to cast five varied statues and thus make his work easier.14 Of course this is an early Imperial period example and we are not sure if this repeated use of models dates back to earlier periods. But given that the indirect lost-wax casting method, which was already popular in the middle of the fifth century BCE, enabled sculptors to keep their original clay models in their workshops, the repetitive use of models must also date back to this period. In fact, Sabbione measured lengths of some parts (with a normal measure) of the two bronze Riace Warriors, both most probably created in Polykleitos’ time,15 and given that the differences between the two are in most points within 1 cm, supposed that they may have been cast from the same models.16 Then there is also the possibility that Polykleitos used this same method in his works. Varro’s critiques of Polykleitos’ works, cited above, as “almost from one model” may contain some truth, even if Varro may not have fully understood the real meaning of the words. In spite of his fame, there are only two works certainly (or almost certainly) attributed to him: the Doryphoros (the “Spear-Bearer,” fig. 1) and the Diadoumenos (the “Youth tying his head-band,” fig. 2). There are also three Amazons, of which only one was to be attributed to Polykleitos, but we don’t know 13

Naples, Museo Archeologico Nazionale, inv. nos. 5604–5, 5619–21: Ruesch 1908, 209–10, nos. 843–47; Tölle-Kastenbein 1986, 75–79, no. 58b; Pozzi 1989, 130, nos. 166–70 (L. A. Scatozza Höricht); Pettenò 2002, 309–17; Mattusch 2005, 195–208; Ciardiello 2007; Moesch 2009, 136–41, no. 73 (Augustan). 14 Sengoku-Haga et al. 2010. 15 In summary, see La Rocca 1996; Micheli – Vidale 1996. Contra, Ridgway 1984b (100 BCE – second century CE). See also Mattusch 2002, 111–15. 16 Sabbione 1984; Mattusch 2002, p. 114.

Polykleitos’ Works “From One Model”

205

which one. Thus we decided to check these statues to see if we could find the repeated use of a single model in the case of Polykleitos.

Research Method: 3D Shape Comparison To check for the reuse of a single model, we need to examine the subtle form differences between two statues, but neither comparisons with trained eyes nor the measurements of lengths are sufficient, because these methods do not allow an accurate 3D comparison. So we used a 3D laser scanner, specifically a Konica-Minolta Vivid 910/9i, with an accuracy of ±50 m. We demonstrated the essential 3D shape comparison method for the feet of the five bronze statues in Naples. Specifically, first we scanned all the statues (fig. 4) and created their digital models (fig. 5),17 from which we cut out and superimposed two of the foot areas (“rigid alignment,”18 fig. 6). But still the minute differences are not very clear. So we plotted the distances between matching points on each of the forms, and visualized them with colour-mapping (fig. 7).19 Using rainbow colours as a scale with green denoting a perfect match, namely 0 mm difference, we adjust the threshold as appropriate, but in this case, +2 mm distances are red, and –2mm blue, over ±2mm black. Here you see that almost the entire foot surface is coloured green. That means the two feet have exactly the same form. This procedure allows us to distinguish millimetre differences and, as a result, reveal new information about ancient sculpture. In this case, as mentioned above, we found that this sculptor cast some of the feet from a single model to expedite the creation of five different foot images.

The Precision of “Mechanical” Roman Copies We decided to apply this new method to the statues of Polykleitos. But, unlike the five bronze statues discussed above, there is another big problem in the case of Polykleitos’ works. We do not have his original bronzes, but rather only the Roman copies, most of which are marble. They are very tantalizing echoes of their originals but the copy quality varies.

17

Ikeuchi – Miyazaki 2007; Masuda 2002. Besl – Mckay 1992; Ikeuchi – Miyazaki 2007. 19 Masuda 2005. 18

206

Chapter Thirteen

However, they are not freehand copies; they were actually all fabricated via a mechanical process. The bronze copies were cast with molds taken from plaster master models (like those of the Baiae casts), which in turn, were cast from the bronze originals. Thus the bronze copies maintain their original forms, even though some shrinkage may have occurred in the process of copying. Besides, the production of accurate marble copies using the “mechanical copy” system began in the first century BCE.20 Details of this system are still not clear, but it is generally accepted that they used some form of instrument and copied many points of a model to a marble block, and, by connecting these points, they arrived at the same form as the model (fig. 7).21 In this manner, Roman copy makers could reproduce numerous precise Roman copies based on the plaster master models kept in their workshops. No wonder an especially large number of copies were created from the Doryphoros, given that it was the “canon” for the other sculptors.22 We chose four examples from the extant copies to check the precision of the Roman copies: the bronze herm from Herculaneum (hereafter called “Bronze Head,” fig. 8),23 the marble herm in Naples (“Marble Head”),24 the marble statue from Pompeii (“Pompeii Statue,” fig. 1),25 and the marble statue in Minneapolis (“Minneapolis Statue”).26 In this last case we scanned a modern plaster cast conserved in Munich instead of the ancient Roman copy. This section presents an outline of the process we used in our comparisons of the four Doryphoros copies and the results obtained, with 20

Ridgway 1984a, 31–36. See above, n. 11. Pfanner 1989. 22 Regarding the replicas of the Doryphoros, see Kreikenbom 1990, 163–80, III 1– 29, III 42–61. 23 Naples, Museo Archeologico Nazionale di Napoli, inv. 4885, found in the Villa dei Papiri, Herculaneum: Ruesch 1908, no. 854; v. Steuben 1973, 11–27; Pozzi 1989, 136, no. 195; Kreikenbom 1990, 174, III 42 (Augustan); Mattusch 2005, 276–77. 24 Naples, Museo Archeologico Nazionale di Napoli, inv. 6412, found in Herculaneum: Ruesch 1908, no. 147; v. Steuben 1973, 28–29; Zanker 1974, 8–9, 38, pl. 39 (2,4); Pozzi 1989, 142, no. 226; Kreikenbom 1990, 174, III 43 (Augustan). 25 Naples, Museo Archeologico Nazionale di Napoli, inv. 6011, found in the Palaestra of Pompeii: Ruesch 1908, no. 146; v. Steuben 1973, 31–53; Zanker 1974, 8, pl. 5,1; 7,3; Pozzi 1989, 100, no. 31; Kreikenbom 1990, 163, III 2 (late Tiberianearly Claudian). 26 Minneapolis, Institute of Fine Arts: Meyer 1995; Hallett 1995; Kreikenbom 1990, 163, III 1 (high Augustan). 21

Polykleitos’ Works “From One Model”

207

the detailed procedure notes and analysis to be published elsewhere.27 First, the comparison of the Pompeii Statue and the Minneapolis Statue as a whole demonstrates that they differ considerably; with particularly notable differences found in their arms and legs. A part-by-part comparison reveals a much closer match. In addition we noticed that some sections matched quite closely, namely the heads, hands, and feet (or toes). They were copied more accurately. In fact, in this instance and in other statues, we found more precise copying in the heads and feet.28 Next we compared the four heads. During these comparisons we noticed that the Bronze Head is about 2% smaller in size than the Marble Head and the Minneapolis Statue. The size of the latter two is almost the same, but the head of the Pompeii Statue is about 3% bigger than the other two marble heads, thus about 5% bigger than the Bronze Head. (The scale ratios were calculated automatically to minimize the gaps of the two compared objects29). We compared them after scale normalization and, as a result, we found that the distances of the nearest points between the two heads are within 2 mm in any points – namely, the form of the Bronze Head is quite similar to the other heads. This indicates that, though excavated in the eighteenth century and possibly restored by a sculptor of that period, the Bronze Head effectively maintains its ancient form and was neither crushed nor distorted. In short, (1) we can use Roman copies in place of their originals for head and foot comparisons, but we must remember possible slight variations in scale. (2) The Bronze Head is the best copy among the four heads of the Doryphoros, although the scale of its head is about 5% smaller than that of the Pompeii Statue.

The Doryphoros and the Diadoumenos Polykleitos’ two works, the Doryphoros and the Diadoumenos, were created in two separate periods of his life: the latter is dated around 430– 420 BCE, namely about twenty years after the creation of the former, when he was already at the end of his career.30 They differ in height: the Doryphoros from Pompeii is 200 cm tall and the Diadoumenos in Athens, 188 cm (measured in both cases from the upper surface of the base to the highest point of the head, measurements taken from their 3D models). 27

See Sengoku-Haga et al. forthcoming. Zhang et al. 2013. 29 The classical ICP plus scaling was applied: Horn 1987. 30 Stewart 1990 (Doryphoros ca. 440 BCE, Diadoumenos ca. 430 BCE); Borbein 1996 (Doryphoros 450–440 BCE and Diadoumenos around 420 BCE). 28

208

Chapter Thirteen

They also differ in style – the former representing “manly youth” (viriliter puerum) and the latter “soft youth” (molliter juvenem) (Plin. nat. hist. 34.55) – and their heads differ considerably in impression. The face of the Doryphoros is very rigid and vigorous, while that of the Diadoumenos, soft, natural, and expressive. The hair of the Diadoumenos is fluffier, the face contour more rounded and slightly younger in appearance. The head is inclined and looking downward, with the eyelids slightly closed. But, since part of the received impression may come from the inclination of the head and carving style of the copy-maker, it is worth trying our new 3D form comparison method.

3D Form Comparisons In terms of the Diadoumenos of Polykleitos, the late Hellenistic marble statue found in Delos, now in Athens, is the best copy that still maintains its head (fig. 2).31 Datable to around 100 BCE, it is one of the earliest mechanical copies in the Greek and Roman world.32 Given practical considerations, we scanned the modern plaster cast in Munich rather than the ancient marble copy in Athens. We used the Bronze Head in Naples (fig. 8) as a sample of a Doryphoros’ head and the Pompeii Statue (fig. 1) for the feet, given that the right foot of the Minneapolis Statue is broken. First we normalized the scale of the heads – more precisely, we enlarged the Diadoumenos head by 1.01 – and compared them. In contrast to our first impression, the facial features of the two heads do not differ as much as we supposed (fig. 9). Mentioning some differences, the area from beneath the left eyebrow to the root of the nose appears in red, beneath the eyelids also in red, the both sides and the end of nose blue, and also the lower contours of the lips, blue. These factors indicate that the left side of the root of nose is less carved on the Diadoumenos’ head, its eyelids are thicker, the nose line less carved, and the nose slightly shorter, which may cause a slightly higher positioning of its mouth. Regarding the foot forms, we again normalized their scales and compared them. The results show that the right feet have similar forms (fig. 10), but the left feet reveal more difference (fig. 11). It is a natural result, seeing that the Diadoumenos’ left heel is not raised as high as the Doryphoros and thus their toes and insteps have different forms.33 31 Athens, National Museum, inv. 1826, found in the House of Diadoumenos, Delos: Kreikenbom 1990, 188, V 1; Kaltsas 2002, 111, no. 201. As for the copies of the Diadoumenos, see Kreikenbom 1990, 188–203, V 1–60. 32 See above, n. 11. 33 The height of the raised left heel of the Diadoumenos is 9 cm, while that of the

Polykleitos’ Works “From One Model”

209

Regarding the right foot, the ratio of the Doryphoros (Pompeii Statue) and the Diadoumenos is 1 to 0.937. (The left foot ratio is 1 to 0.978 but this numerical value may not reflect reality because the two feet of different forms are not well aligned.) Because the ratio of the Diadoumenos and the Bronze Head is 1 to 1.006 and that of the Pompeii statue and the Bronze head 1 to 0.946, the head ratio of the Pompeii Statue and the Diadoumenos is 1 to about 0.940, which approximately corresponds to the scale of the right foot.

Discussion The comparison results demonstrate that (1) the face and the right foot of the Diadoumenos repeat quite similar forms as the Doryphoros and (2) in terms of these parts, the Diadoumenos is about 6% smaller than the latter (Pompeii Statue). Regarding (1), the high degree of matching does not seem to be a mere coincidence; the positions of the eyes, brows, noses, and mouths are almost the same. This resemblance would not have occurred if Polykleitos had modelled the two heads independently. Most probably Polykleitos, instead of modelling all of the section models from the clay work, utilized models of the Doryphoros head and right foot in his later works. Of course we must admit that they are different in scale and, even after the scale normalization, the two faces do not match perfectly. This difference may be partly the result of the late Hellenistic copy-maker’s soft carving, but we may as well suppose that the original sculptor, Polykleitos, reworked the model to vary the facial expression, given that he remodelled the hair on the same head to reflect the latest fashion. He most likely thickened the eyelids and made them slightly closed to create the downward gaze appropriate for the head movement; he may also have remodelled the lips so that they were softer and rounder, thus giving the face a more youthful, fresh appearance. But if Polykleitos used the same model for both heads and the right feet, why do they differ 6% in scale?34 Considering the model reuse process, the master model of the Doryphoros (or maybe a bronze replica?) Doryphoros (Pompeii Statue), 12 cm. (Measured on the plaster casts in Munich.) 34 Also the heights of the statues are 200 cm (Pompeii Statue) and 188 cm (Diadoumenos), thus their ratio is 1 to 0.94, too. But we should be cautious about the total heights; as having seen above, Roman copy-makers worked less accurately regarding legs and arms. Actually, the height of Minneapolis Statue is 200 cm, which does not correspond in terms of the head scale to the Pompeii statue.

210

Chapter Thirteen

must have been kept in the workshop, from which Polykleitos would have made a mold when he wanted to create the Diadoumenos. From this mold he made a new clay model, that of the Diadoumenos, adding some modifications. We can presume that the master model was dried and fired to give it a hard surface that would be durable through repeated use,35 and this must have caused the considerable shrinkage in its scale.36 He used the old Doryphoros model for the Diadoumenos’ right foot, but he modelled a completely new left foot, since he wanted to have its heel lower than that of his earlier work. He represented the Doryphoros as walking slowly forward, while the Diadoumenos is shown standing thoughtfully in a relaxed position with a less raised heel. He never hesitated to try new expressions, even while he reused his old models to lighten the work involved in each new production. According to Plutarch, Polykleitos said, “when the clay modelling comes to the nails (namely in details), the work is the hardest.”37 The procedure of applying some section models of the “canon” to his new bronze statues must have been easier for him than creating completely new detailed section models for each work.

The Three Amazons Pliny (nat. hist. 34.53) records four sculptors’ names who each made his own Amazon dedicated to the sanctuary of Artemis at Ephesos: Polykleitos, Pheidias, Kresilas, and Phradmon.38 The first three are renowned as the great artists of the fifth century BCE. Three Amazon types have been distinguished among the numerous Roman copies of the fifth century style and they are called, respectively, the Sosikles (or Capitolino) type, the Sciarra (or Copenhagen-Tivoli) type, and the Mattei type (Fig. 12). The identification of Polykleitos’ Amazon swings mainly between the Sosikles Type and the Sciarra Type.39 Without any conclusive 35

New York, Metropolitan Museum of Art, Rogers Fund, 1906 (06.1151): Hemingway 2000, 40–41, fig. 19. 36 The rate of shrinkage for terracotta statuettes is 12–20% (Knoop 1988, 82–83), but we have to take into account the fact that in the case of a terracotta statuette both the mold and the statuette were fired (Bell 1981, 120), and thus the shrinkage must have been twice as much as the case of Polykleitos’ works. 37 Plut. mor. 86a; Plut. mor. 636c: Pollitt 1990, p. 77. 38 As for the reading of the fourth name Cydonis as Phradmon’s or Kresilas’ ethnikon, see Stewart 1990, 262, T 60; DNO 1229. 39 The Sciarra type was suggested by Furtwängler, whose opinion has been followed by many scholars, and was recently supported by R. Bol: Furtwängler

Polykleitos’ Works “From One Model”

211

evidence, scholars have continued to attribute each type to one of the three sculptors for more than a century. But now, with the help of the 3D shape comparison method, we may obtain new evidence that will allow us to conclude this long-continuing discussion.

3D Shape Comparisons In order to compare the forms of heads and feet of the three Amazons with those of the Doryphoros and the Diadoumenos, prior to 2014 we scanned the plaster casts of the Sosikles type statue (Rome, Museo Capitolino, inv. 651)40 and the Sciarra type statue (Copenhagen, Ny Carlsberg Glyptothek, inv. 1658)41 in the Cast Museum in Munich; the headless Mattei type Amazon statues (Villa Adriana, Antiquarium, inv. 2266) and the Sciarra type (Villa Adriana, Antiquarium, inv. 2255) at Tivoli;42 the fragment of an Amazon head in Rome (Rome, Museo Nazionale Romano, Palazzo Massimo, inv. 124666).43 In addition, in September 201444 we scanned a bronze herm of the same Amazon type found in Herculaneum at Naples (Naples, Museo Archeologico Nazionale, inv. 4889, fig. 13)45 and the Sosikles head in the Cast Museum of Munich (Rome, Palazzo dei Conservatori, inv. 1091, fig. 14).46

1893, 286–91, 293–96, 449–50; Becatti 1951; Beschi 1965, 269–70; Weber 1976; Borbein 1996, 82–84; Hafner 1997; Bol 1990; Bol 1998; Delivorrias 1995. The Sosikles type was strongly suggested by von Steuben who measured numerous parts of the Doryphoros and the Sosikles Amazon searching for the meaning of the “canon”: v. Steuben 1973, 56–71; v. Steuben 1993; v. Steuben 1994; Stewart 1990, 162–63; Rolley 1999, 39–42. 40 Helbig, II, no. 1393 (v. Steuben); v. Steuben 1973, 56–65; Bol 1998, 188, no. II.2. 41 Poulsen 1951, no. 54; Ridgway 1974, 7–15 Augustan ; Bol 1998, 175–76, no. I,4; Moltesen 2002, 207–11, cat. 60. 42 Mattei type in Tivoli: Helbig IV, no. 3200 (v. Steuben); Raeder 1983, no. I 90; Bol 1998, 210–11, no. III.7. Sciarra type in Tivoli: Helbig IV, no. 3201; Bol 1998, 210–11, no. III.7. 43 Helbig III, no. 2261 (v. Steuben); Giuliano 1979, 223–25, no. 140 (D. Candilio); Raeder 1983, 77–78, no. I,64, p. 307; Bol 1990, pp. 213–39; Bol 1998, 187, I.26,2 44 After the Budapest symposium we scanned three less restored Amazons’ heads to get better data and more objective results. Thus we are obliged to change our conclusion. 45 Bol 1998, 187, no. 1.26,1; Matusch 2005, 278–82. 46 Helbig4 II, no. 1592 (v. Steuben); v. Steuben 1973, 65–67; Bol 1990, 568–69, no. 87m; Bol 1998, 195–96, no. II.16.

212

Chapter Thirteen

Regarding foot comparisons between the three Amazon types and the two male statues of Polykleitos, we used the Sciarra type of Copenhagen, the Sosikles statue of Capitolino, the Mattei type in Tivoli, the Doryphoros in Naples (“Pompeii Statue”) and the Diadoumenos in Athens, which all have extant ancient feet. The results were very ambiguous (tables 1, 2): we could not identify the best matching pair. The feet of the Sosikles Amazon show a major difference from the Doryphoros and the Diadoumenos, naturally because her standing leg is the left leg, as opposed to the other statues, which all stand on their right legs. Neither the Mattei type’s left foot nor that of the Sciarra type match that of the Doryphoros or the Diadoumenos. However those results were also expected, given that, as we have already observed, the forms of their raised feet vary due to the specific poses shown. As for the right feet, maybe there is some similarity between the Sciarra type and the Doryphoros (at least the lengths of the toes are very close), but they are far from a perfect match. Next we tried a face comparison. For the Amazons’ heads, we used the 3D head models of the Sosikles head in Conservatori (cast in Munich), the Sciarra statue in Capitolino (cast in Munich), and the bronze head in Naples. Although none of the Mattei type statues was found with its head, we will consider the Rome-Naples type as the Mattei type, because the fragmentary head in Rome was found in Tivoli where the Mattei body was excavated and because there is no other – more probable – Amazon head type attested by several copies.47 All have perfectly preserved faces, except for a small part of the nose tip of the Sosikles head, which is a modern restoration. The scale ratio of the Doryphoros (Bronze Head), the Diadoumenos, the Sosikles head, the Sciarra statue, and the bronze Mattei head is 1.00 : 0.99 : 0.99 : 0.95 : 1.01. We normalized the scale before comparison. The results are quite clear (table 3). The comparison of the Doryphoros with the Sciarra type shows black areas along the nose lines and on the chin, while that with the Mattei Type has the black areas around the left eyes, and the comparison with the Sosikles Type has almost no black area. The restored nose tip and the left wing of the nose are a little bit narrower, but otherwise the gap is less than 2 mm. The closeness is comparable with that of the Doryphoros and the Diadoumenos pair (fig. 9). 47 Furtwängler 1893, 296–303; Becatti 1951; v. Steuben in: Helbig III, no. 2261. In contrast, Bol 1998, 187, no. I.26 put this head type into the variant of the Sciarra type, and Mattusch 2005, 278, calls it a “Polykleitan” young woman. Weber 1978 and Boardman 1985, 217, suggest that the Petworth head should belong to the Mattei body.

Polykleitos’ Works “From One Model”

213

Discussion The gap between the faces of the Doryphoros and the Sosikles Amazon is as close as that between the Doryphoros and the Diadoumonos, thus in the range of works from the same model. What is more, the faces of the Diadoumenos and the Sosikles Amazon are the same size: the ratio of both heads to that of the Doryphoros (Bronze Head) is the same, 0.99. This suggests that the Diadoumenos head and the Sosikles head were created by the same process, namely, their master models were each made with a mold from the face model of the Doryphoros. Naturally, multiple modelling processes using the same models could most easily occur in the workshop of the same sculptor: that of Polykleitos. Thus the evidence strongly suggests that it is the Sosikles type that was created by Polykleitos. We could not find any significant results in the feet comparison, because, contrary to the case of the Doryphoros and the Diadoumenos, he chose to use the left leg as the standing leg in his Amazon statue, which prevented him reusing the foot-models of his earlier work.

Conclusion Through the use of 3D shape comparison we ascertained that Polykleitos made the master head-models of his later works from that of the Doryphoros, changing their hair to a more natural style and possibly revising some details suitable for the pose, age, and gender of the new statues. Except for the head, the right foot of the Diadoumenos is the only part whose reuse of the older section-model we confirmed, but this does not exclude the possibility that he might have applied other section-models in his later works as well. We do not know if this reuse of the master model was purely a laboursaving device or if it is related to the concept of the “canon.” If the latter holds true, then Polykleitos considered that the Doryphoros’ facial features were one of the key points of his canon, given that he repeated exactly the same shapes in his later works. Modern scholars often regard the hair styles of ancient statues as important but, at least in terms of his later works, Polykleitos seems to have considered the hair as changeable. We identified the Sosikles type (fig. 12), known through about thirty Roman copies, as the Amazon of Polykleitos. The number of extant Sosikles-type copies is roughly twice that of the other two types. The impression of the head (fig. 14) is totally different from that of the Doryphoros: the oval face has delicate features, the hair locks are

Chapter Thirteen

214

modelled like entangled cords and wave backwards naturally and irregularly, and the eyes are weary and full of emotion. This head reminds us the words of Strabo: “the Argive Hera of Polykleitos was the most beautiful in its skill (techne) in the world but in costliness and size inferior to those of Pheidias.”48 Without a doubt he surpassed other sculptors in the carving of a beautiful female face. Which, then, we might ask, of the remaining two Amazons was made by Pheidias, the great creator of the Athena Parthenos and the Zeus Olympios? Lucian gives us a hint. According to Lucian, “Athena Lemnia” and the “Amazon leaning on her spear” should be the most highly praised works by Pheidias.49 Provided that the Sciarra type has no lance, Pheidias’ Amazon should be the Mattei type, while Kresilas’ is the Sciarra type. But for the moment we would rather suspend final attribution of the two Amazons until we finish our 3D shape comparison of the Pericles’ head and Diomedes statue, usually attributed to Kresilas.

Acknowledgements For generous permission and kind assistance during the scanning, I am particularly grateful to Dr. Ingeborg Kader, Museum management and chief keeper of the Museum für Abgüsse Klassischer Bildwerke München, Dr. Teresa Elena Cinquantaquattro, the superintendent of the Soprintendenza Speciale per i Beni Archeologici di Napoli e Pompei, Dr. Valeria Sampaolo, the director of the Museo Archeologico Nazionale di Napoli, Dr. Elena Calandra, the superintendent of the Soprintendenza per i Beni Archeologici del Lazio and Dr. Benedetta Adembri, the director of the Archaeological Area of Villa Adriana. This work is supported by JSPS KAKENHI Grant Numbers 19520088, 23320040, and 23652018.

48 49

Strabo, 8.372: DNO 1214. Lucianus, Imagines 4.

Polykleitos’ Works “From One Model”

215

Fig. 1. (left) Polykleitos’ Doryphoros found in the Palaestra of Pompei (here called “Pompei Statue”) Naples, National Archaeological Museum, inv. 6011 Fig. 2. (right) Polykleitos’ Diadoumenos found in the House of Diadoumenos on Delos. Athens, National Museum, inv. 1826

216

Chapter Thirteen

Fig. 3. Bronze group of the so-called “Dancers” found in the Villa dei Papiri, Herculaneum. Naples, National Archaeological Museum, inv. 5604-05, 5619-21

Fig. 4. 3D scan in the National Archaeological Museum in Naples in 2007

Polykleitos’ Works “From One Model”

217

Fig. 5. Alignment of two 3D foot models

Fig. 6. Visualization of the distances with color-mapping. Here the threshold is 2mm.

218

Chapter Thirteen

Fig. 7. “Mechanical copy” in marble (Pfanner 1989)

Fig. 8. Polykleitos’ Doryphoros copied by Apollonios as a herm, found in the Villa dei Papiri in Herculaneum (here called “Bronze Head”) Naples, National Archaeological Museum, inv. 4885

Polykleitos’ Works “From One Model”

219

Fig. 9. 3D shape comparison of the Doryphoros (Bronze Head, left) and the Diadoumenos (right). The distances to the latter are plotted on the former. The scale of the Doryphoros and the Diadoumenos is 1 : 0.978. Threshold 4mm.

Fig. 10. A comparison of the right feet of the Doryphoros (left) and the Diadoumenos (right). The scale of the Doryphoros and the Diadoumenos is 1 : 0.978. Threshold 4mm.

Fig. 11. A comparison of the left feet of the Doryphoros (left) and the Diadoumenos (right). The scale of the Doryphoros and the Diadoumenos is 1 : 0.978. Threshold 4mm.

220

Chapter Thirteen

Fig. 12. Amazon types. From left to right: Sciarra, Sosikles, Mattei (after Boardman 1985 fig. 195)

Fig. 13. (left) Bronze herm of the Mattei type Amazon found in the Villa dei Papiri in Herculaneum. Naples, Nat. Arch. Mus., inv. 4889 Fig. 14. (right) Amazon head of the Sosikles type Rome, Mus. Cap., inv. 1091

Polyklleitos’ Works “F From One Moddel”

221

Table 1 A ccomparison off the left feet of o the three Am mazon types with w the Doryphoros and the Diadoumenos. Thrreshold 4mm.

Table 2 A ccomparison off the right feett of the three A Amazon typess with the Doryphoros and the Diadoumenos. Thrreshold 4mm.

222

Chapter Thirteen

Table 3 Face comparison of the three Amazon types with the Doryphoros and the Diadoumenos. Threshold 4mm.

CHAPTER FOURTEEN VIRTUAL ENVIRONMENTS AND TECHNOLOGICAL SOLUTIONS FOR AN ENRICHED VIEWING OF HISTORICAL AND ARCHAEOLOGICAL CONTEXTS FRANCESCO GABELLONE

Abstract Specialist studies related to historical and archaeological contexts are usually not properly disseminated. The vast academic research literature is rarely reflected by a public awareness of the importance of those results. In many cases, peculiar features of cultural heritage are totally misunderstood, causing a relevant loss in terms of economic exploitation, or even the near oblivion of the heritage site itself. This phenomenon is common in archaeological areas all over the world, where the importance of ancient architectonic features is underestimated due to the scantiness of the remains. Several research groups interested in these issues have developed methods and technologies aimed at allowing tourists to be more engaged during their visits. This chapter elucidates the experiences gained during a decade of research undertaken by the ITLab (Information Technologies Lab, CNR IBAM), aimed at studying, valorizing, and promoting the use of historic and archaeological contexts. The emerging and promising technologies that have marked a significant advance in the field of virtual archaeology will be discussed, with emphasis on their potential in enhancing communication processes and “edutainment.”

Virtual Environments and Enriched Vision The use of the term Virtual Heritage has become fairly common. Anyone concerned with cultural heritage, museums, historical, and archaeological research employs it to define innovative methodologies

224

Chapter Fourteen

and technologies based on 3D digital technology used to represent, communicate, and transfer diverse information about Cultural Heritage. The term Cultural Heritage defines all those material and immaterial components that make up the cultural patrimony of a particular society. This category includes tangible objects such as buildings, archaeological sites, works of human ingenuity, works of art, manuscripts, sites with special natural characteristics, and sites with great scientific or anthropological value, but also immaterial heritage such as oral traditions, popular songs, and artistic forms of expression that are closely tied to a cultural identity. In order to preserve sites or ecosystems that have particular cultural or natural importance, UNESCO draws up a list every year called the World Heritage List (fig. 1), where places and monuments that deserve special attention from institutions and the public are included. During the early 1990s, the phrase Virtual Reality (coined by the IT guru Jaron Lanier in 1989), spread rapidly amongst people working in the fields of communication and scientific visualization. “Seize reality through illusion” is the most convincing synthesis of what this new discipline represents in the simulation of physical processes, the interactive manipulation of machinery, and the representation of material reality. The birth of Virtual Reality marked a new path forward and somehow afforded us a new view of reality, allowing alternative “readings.”1 Soon, this new way of representation and simulation became associated with the term virtual environment, used to indicate those interactive threedimensional models that could be navigated and which simulated a place, a building or a synthetic scheme in real time. A virtual environment doesn’t necessarily have to be a digital copy of a material object, because, as already noted, its utility is undoubtedly linked to the possibilities of introducing new ways of acquiring knowledge (fig. 2).2 A virtual environment can be created either by starting with a realistic representation of a material element that really exists,3 or with an abstract and schematic reconstruction. In fact, a virtual environment is like a “microscope for the mind”4 that allows you to elaborate amplified projections of the material world, to “look beyond” simple appearances and to make logical connections between elements grouped together. This makes it clear to what extent the world of simulation and representation have benefited from the use of virtual environments, especially evident in 1

Diodato 2005. Gabellone – Giannotta 2013b. 3 Gabellone et al. 2013a. 4 Lévy 1997. 2

Enriched Viewing of Historical and Archaeological Contexts

225

sectors such as industrial design, rapid prototyping, augmented reality for medicine, the entertainment industry, the training of pilots and soldiers, the simulation of emergency plans, and also in all of those areas where digital simulation permits the saving of precious human resources or the safe execution of risky activities. For those involved with the promotion of Cultural Heritage, possibly the most interesting aspect is the association of virtual environments with a new way of learning and communication, not just dealing with the characteristics of the actual physical heritage, but above all considering the stimulus that comes from interpretation and problems of reconstruction tied to long-lost civilizations – in short, to archaeology. The use of simulation techniques coincides with a clear attempt to reply to one of the fundamental issues of modern archaeology: how to reconstruct an ever-wider picture of ancient civilizations with the highest possible level of trustworthiness and reliability. One of the aims of every generation of archaeologists is to study a monument in order to reconstruct it. But the input provided by the arrival of these new technologies has demonstrated how important it is to combine historical and humanistic knowledge with the enormous possibilities offered by IT, which is not just useful for interpreting objects and works of art, but also and above all for transmitting the knowledge acquired to a vast, heterogeneous audience, at many levels of interest and comprehension. I believe that archaeology is enjoying a moment of great vitality and awareness on the part of the wider public, thanks to the communicative force of those new communication tools, with which it is possible not only to recreate shapes and materials of the past, but also to evoke details of everyday life (fig. 3). The explosion on the web of virtual, thematic museums, virtual collections and galleries, that all have the same desire to promote and publicize archaeological sites and monuments using Computer Vision and virtual environments, testifies to this interest. In this context, the 3D image is no longer just a pure iconic representation, an artistic fantasy, but becomes a tool for synthesis that transmits and communicates information deriving from scientific research in a graphic form, and represents the elements necessary for a correct interpretation and understanding of information in a very direct fashion.5 The integration of diverse forms of knowledge allows for a demonstrated ease of learning, understanding, and transmission, not just of the historical-archaeological or morphological aspects of an object, but also of technical aspects used in the hard sciences 5

Gabellone 2009.

226

Chapter Fourteen

and usually represented by graphs and tables. Archeometric data that is essential to the reconstruction processes can be integrated within virtual scenes, giving quality to an often-misunderstood field. “The Virtual” becomes a starting point for the creation of an artificial world that enriches and decodifies the “real world,”6 recreating it in a 3D form suited to a simplified interpretation of the information it is designed to transmit. There is no doubt that a three-dimensional usage permits a direct reading of complex information and that many 2D representations of architectural detail or structural anomalies with notable three-dimensional aspects can be difficult to read, especially for non-experts. In virtual scenarios, the 3D real-time visit can be associated with almost any media, in a single environment of use; the users can explore the 3D model from any angle, but can also activate links at any given moment that will take them to a theoretically limitless source of information. This is the start of a voyage of discovery that goes beyond simple aesthetics, where the object is no longer exposed in its purely morphological or aesthetic state, but where the visit can lead to resolution and avoid exclusive contemplation of the object in itself. This voyage leads principally to the deciphering of the specific contents of the object being studied, starting with the visible contents and concluding with the invisible ones; that corpus of information, anomalies, and hidden structures that lies beneath the skin of the monument and that only artificial vision can reveal. The object, as a container of values, can be analysed in its mineral, petrographic, chemical, and physical components, but should also be explored in its relationship with its original context. In this way, virtual space can accelerate and strengthen cognitive capacities, capable of generating ‘virtuous’ processes of extremely effective learning, based on metaphors of the real world, but easy to use and understand (fig. 4–5).7

Virtual Environments for the Fruition of Inaccessible Contexts Projects of promotion and digital fruition of inaccessible monuments or artefacts separated from their original context of provenance have become rather popular. The international scene offers numerous cases of sites for which a physical approach is limited. On one side, there are monuments located in private properties or in problematic environments

6 7

Rheingold 1993. Gabellone 2009.

Enriched Viewing of Historical and Archaeological Contexts

227

unapt to be open to the public, or sites whose state of preservation and climatic conditions are not compatible with the presence of visitors. On the other side, there are categories of monuments, without particular musealization issues, which due to their articulation or location have a limited level of accessibility. In addition, there are some categories of monuments that do not have particular problems of musealization, but which are not easily accessible due to their location, or because their morphology makes visiting difficult. To these main groups, some monuments can be added for which public access is restricted because of conservation strategies. That is the case for those contexts where the high number of visitors represents a serious threat, particularly over a longer period. An emblematic example is Herculaneum and Pompeii, where the unconscious damage daily inflicted by visitors is worse than that caused by the eruption of Vesuvius. “Save Pompeii from the damages of time and tourists” is the title of an article published in 2009 in the New York Times art history section, wherein the author invited Italians to do their best to find a solution to the damage caused by tourists to one of the most famous archaeological world heritage sites. What can be done to limit that damage? What can be done for those monuments with limited accessibility? How can the decontextualized heritage be properly musealized? A viable solution is offered by certain technologies of visual computing, which are constantly developing communication tools to implement the experience of a visit (fig. 6–7). As mentioned above, the use of virtual environments enhances and enriches the comprehension, and sometimes this approach represents the transmission of cultural messages embodying the awareness of their importance and value, and not just a method to allow the visit to the sites. In the “Marta Racconta, storie virtuali di tesori nascosti”8 project there are three distinct levels of transmission: the interactive virtual visit of the “Hypogeum of festoons,” the non-interactive video where the hypogeum is described and contextualized, and finally a series of documents in which the artefacts included in the reconstruction (fig. 8) are elucidated. A traditional approach is a multimodal visit in which heterogeneous information is combined and presented in different ways to produce the best educational and communicational impact. In a virtual visit the user can visualize the site as it is nowadays, with the option of getting information about grave goods and pigments used for the frescoes. In the non-interactive video, based on the production of a stereoscopic clip, the monuments under study are presented in a narrative way, where their relation with the original historical and cultural context is 8

Gabellone – Giannotta 2013a.

228

Chapter Fourteen

recreated. After a general introduction about cult practices and funerary rituals in Greek culture, the user is informed about the original aspect of the monuments – that is, the reconstruction. The main aim of this approach is to take the user/visitor to another level of comprehension of the presented object. To comprehend means to know before understanding. The principal components of every comprehension process are essentially twofold: comprehension and motivation. The first is a cognitive component and corresponds to the comprehension of the object of study. The second is more dynamic and coincides with one’s motivation – that is, the interest which drives us to know and understand. If motivation determines comprehension, it can be inferred that the digital technology characterized can have a positive effect on both factors of comprehension, producing a successful instance of communication from an instance of strong motivation.9 In this context the virtual reconstruction, as the final step in multidisciplinary research, represents the most direct and simple way to communicate the splendour of the past, enriching it with emotional values deriving both from the narrative and from the related visual effects.

9

Gabellone et al. 2013b.

Enriched Viewing of Historical and Archaeological Contexts

229

Fig. 1. Sites of the World Heritage List

Fig. 2. A mobile device showing the actual state and original state of an inaccessible house in Pompeii (Casa dei Ceii).

230

Chapter Fourteen

Fig. 3. Reconstruction of a residential area of ancient Syracuse, in Greek period

Fig. 4. The reconstruction of ancient Metapontum sanctuary

Enriched Viewing of Historical and Archaeological Contexts

231

Fig. 5. The reconstruction of Temple of Apollo in Syracuse

Fig. 6. An inaccessible monument: The Hypogeum of Festoons in Taranto.

232

Chapter Fourteen

Fig. 7. Virtual 3D reconstruction of the same monument.

Fig. 8. The objects found in the Hypogeum and used in the virtual reconstruction in the current exhi-bition of the Nat. Arch. Mus. of Taranto.

CHAPTER FIFTEEN CLOUD-BASED COLLABORATIVE FRAMEWORK FOR REMOTE REAL-TIME INTERACTION WITH LARGE-SCALE 3D DATA YASUHIDE OKAMOTO, GREGORIJ KURILLO, TAKESHI OISHI, KATSUSHI IKEUCHI AND RUZENA BAJCSY

Abstract Recent advances in 3D technologies have enabled us to easily obtain largescale 3D data from real cultural properties for the purpose of archaeological research and preservation. Cloud computing technologies also allow us to store and share such 3D data among archaeologists around the world. 3D data sharing can facilitate collaborative activities and discussions between archaeologists from the view point of visibility and interactivity. Current cloud-based platforms for data sharing, however, do not provide real-time capabilities for interaction with large 3D models. We propose a novel collaborative framework for navigation, manipulation, and annotation of scenes with very large archaeological 3D models (with several 100-million vertices). Our system can visualize large 3D models in real time, and allows remote multiple users to share collaborative works with those models in a common virtual world: such as reviewing cultural models, comparing multiple objects, and rebuilding broken pieces. Our system also provides several logging mechanisms that geographically distributed users can take advantage of during real-time discussion of 3D models.

234

Chapter Fifteen

Introduction Visualization and interaction with three-dimensional (3D) models has become ubiquitous in the last decade due to advances in CPU and GPUbased computing, rendering capabilities, and availability of low-cost largesize 2D and 3D display systems. 3D models are now also easy to generate thanks to advances in 3D scanning technologies used in rapid prototyping, cultural heritage preservation, 3D medical imaging, and others. In addition, new interactive technologies such as touch screens, depthsensing cameras (e.g. Microsoft Kinect http://www.microsoft.com/enus/kinectforwindows/), and other gesture control devices promise new and more intuitive ways of interacting with these data. The re-birth of distributed computing in the form of the cloud has facilitated new communication and collaboration tools, which enable multiple users to share and edit data interactively. The majority of the collaborative tools are, however, still limited to interacting with documents, images, and slides. In light of these recent technological advances, we propose a cloudbased collaborative architecture for sharing, rendering, and real-time interaction with large-scale 3D models. In our work we consider 3D models represented either as a dense cloud of 3D points, such as data obtained with laser scanning or photogrammetric 3D reconstruction, or as a 3D polygon mesh generated by geometric 3D modelling. We are considering 3D models that consist of several million to hundreds of millions of vertices (or more), which are impractical to store locally or to share between multiple users. Such large-scale 3D data are found in many scientific and engineering disciplines, including computer aided design (CAD), geoscience, aeronautical engineering, architectural design, cultural heritage preservation, modelling from medical imaging, and others. Real-time access to such large-scale 3D models by multiple users requires the introduction of novel techniques for rendering, network transmission, and user interaction. To overcome the network bandwidth limitations of the cloud connectivity, we approach the 3D data streaming in a hierarchical fashion, by transmitting only the required level of detail (LOD) of the model based on the user’s virtual viewpoint. The level of detail is dynamically increased as more geometrical information is streamed from the server to each client. To facilitate real-time collaboration for multiple users, our framework implements a distributed scene management system which includes several navigation and interaction tools available to the users to manipulate 3D objects and annotate the data points. In addition, the

Cloud-based Collaborative Framework

235

framework allows tracking and logging of the interaction history in 3D space, which can be used to study interaction in collaborative 3D environments. Finally, we provide users with the ability to transmit their motion and their images into the virtual space via 3D real-time generated or gesture-driven avatars facilitated by the Microsoft Kinect camera. We argue that user embodiment will further enhance geographically distributed collaboration. The importance of user embodiment was identified as a key component for successful distributed collaboration in the early stages of shared virtual environments.1 The main contribution of this work is its novel type of 3D collaborative system, which can handle very large models (over 100 million vertices) by multiple users, implemented in the cloud computing environment. Other 3D collaborative services for sharing and viewing models can only handle models of much lower dimensions. Furthermore we introduce immersive and natural interaction with such data via embodiment through the Microsoft Kinect camera. Finally, our 3D collaborative framework provides several annotation tools that geographically distributed users from various disciplines of science and engineering can take advantage of during the discussion of 3D data.

Related Work Geographically distributed real-time interaction and collaboration in 3D environments has been an important research area from the early days of virtual reality. More recently, several commercial and open-source collaboration tools aimed at visualization, sharing, interaction, and annotation of 3D models have been developed. The most popular distributed 3D environment for sharing and interaction with 3D models is Second Life (Linden Research, Inc., 2003: http://secondlife.com) which features a large virtual world populated by avatars representing remotely connected users. Second Life, whose focus is on social interaction rather than modelling, does not offer a scalable system which can handle large datasets nor a wide range of tools for 3D model interaction. Another example of a collaborative 3D framework is Eon Coliseum,2 which includes presentation-style interaction with simplified 3D models (and other data). Several more recently released commercial products are focused on distribution and sharing of computeraided-design (CAD) models in real-time. Hypercosm (http://hypercosm. 1 2

Benford et al. 1995. Eon Reality, 2009: http://www.eonreality.com/product_coliseum.html.

236

Chapter Fifteen

com) allows users to create 3D simulations and explore the 3D scenes jointly through a web-based interface. CadFaster (http://www.cadfaster. com/index.php/Products), on the other hand, is a more general-purpose collaborative mark-up and co-viewing application for sharing and viewing 3D CAD models. A similar concept is featured in Bloom Unit plug-in (http://www.bloomunit.com) for Trimble SketchUp (http://www.sketchup. com) which allows users to share and also edit 3D models via a cloudbased service. A collaborative 3D tool for digital artwork, in particular for game design, is featured by Verold Studio (http://www.verold.com) which allows users to share models in real time, add/remove model instances, and change their material properties or lighting. Another important component that we address in this work is the adaptive streaming and rendering of large 3D mesh models. One of the first approaches to streaming of mesh data was the progressive mesh introduced by Hoppe,3 which enables 3D models to change resolution by precomputing the sequence of vertex splits. Adaptive Tetrapuzzles4 has hierarchical mesh structures which contain small partial meshes in multiresolution, that can reconstruct 3D surfaces in arbitrary resolution by traversing the tree structure and selecting only meshes to satisfy the display quality. QSplat5 has point-based representations and also uses a hierarchical data structure, whose hierarchical structure is constructed with encoded simplified points. Far voxels6 applies not only the points and triangle meshes, but also view-dependent colour voxels that are used as the rendering primitives in the LOD hierarchy. Other systems have also focused on network streaming for 3D meshes by using the hierarchical structure represented in simplified and partitioned meshes.7 In contrast to these research and commercial software frameworks for collaborative viewing and sharing of 3D models, our focus is on supporting interaction with large 3D models of several hundred million vertices, which are typically produced as a result of 3D object scanning rather than computer-aided design (although our framework also support the latter). To access and display such huge data on commodity hardware, we propose novel techniques for adaptive streaming and rendering which we integrate in a unified cloud-based framework that supports the geographically distributed collaboration of multiple users, as described in the next few sections. 3

Hoppe 1996. Proposed by Cignoni et al. 2004. 5 Rusinkiewicz - Levoy 2000. 6 Proposed by Gobbetti - Marton 2005. 7 Kim - Lee 2012. 4

Cloud-based Collaborative Framework

237

3D Collaboration Framework In this chapter we propose a novel collaborative framework for 3D and cloud computing technologies. Our proposed system has three main functions: 1. collaborative interaction between remote users 2. display and sharing of large 3D models 3. 3D telepresence-based interaction Primarily, our collaborative system was developed for visualization and real-time interaction with 3D triangle mesh models obtained by laser scanning. The architecture thereby developed facilitates collaboration of multiple users through the Internet. Using the centralized storage provided by the cloud computing architecture, the users can share and discuss the 3D models with other remote users. The geometry of the models is transmitted from the cloud server on demand with the optimal level of detail, depending on the user’s available network bandwidth and their location in the virtual space. If the user is far away from the model, only a minimal set of vertices to describe the geometry will be transmitted. The presented framework allows multiple users to work independently or jointly by first logging into the shared virtual space. In this 3D environment, the users can add or remove objects, change their position, orientation, and other properties (e.g. colour, material). Furthermore, we provide several ways of navigating through the virtual space (e.g. first person, third person, and look-around) and several tools to support annotation and discussion. To emphasize the presence of users in the shared virtual space, our system renders 3D avatars at their viewpoint location. As the users navigate through the space, their avatar follows their position. Moreover to realize a more intuitive immersive environment, our system can also generate the users’ avatars in real time through the Microsoft Kinect camera.

Server Structure and Protocol For this framework, we developed a cloud-computing-based architecture (see fig. 1) where the server is located on the Microsoft Windows Azure cloud computing platform (http://www.windowsazure .com) while a stand-alone client program runs on a personal computer or a laptop with a Windows operating system (see fig. 2 for more details). The communication between the two is realized via the TCP-IP network

238

Chapter Fifteen

protocol. On the server side there are three server programs which are in charge of the three major functions outlined above. The collaboration management server controls active sessions from each user. This server receives information about users’ viewpoints, active objects and their location, the usage of input tools, and messages between the users. The server distributes the information to all clients in order to maintain the consistency of the shared virtual environment. The 3D display server provides 3D data represented with a hierarchical data structure encoding different levels of detail. When a user is browsing a model, the client program sends request messages containing the model ID and the user’s active viewpoint. The 3D server retrieves partial 3D geometric data and optionally prerenders texture data based on the user’s current viewpoint to increase the quality of the rendering while requiring minimal geometry to be downloaded from the server. The details of the display method are described below. Finally, the user image server manages the transmission of 3D video data for the rendering of real-time 3D avatars. Colour and 3D data captured by the Kinect camera are first compressed and then transmitted via the user image server to the other connected clients for display.

Collaboration System Our system is focused in particular on collaborative interaction with complex 3D objects and the discussion process involved. Therefore we provide several annotation tools and logging mechanisms to capture this process and provide the history of the events. There are several areas where there is a strong need for real-time discussion alongside the visualization of 3D models. For example, in the field of archaeology it has become more common to use 3D archiving technologies to preserve the cultural heritage and to explore various hypotheses about archaeological artifacts. Current practices involve the use of file and desktop sharing technologies, Skype, annotations via Adobe Acrobat 3D files, and others. In this work there is a need to visualize, compare, and analyse different scanned artifacts, while being able to discuss the findings with other collaborators who may be geographically distributed. Our framework was designed with such a scenario in mind. In what follows, we describe the modes of navigation and various tools for manipulation and annotation of objects, and the network protocols behind them.

Cloud-based Collaborative Framework

239

Navigation and Manipulation To support intuitive interaction with 3D models, we implemented several different navigation modes. The interface of the client applications provides users with different modes of navigation, such as panning, rotating, and travelling forward and backward, in a similar style as used in general 3D modelling software. The users can explore the environment in the walking mode where the rotation of the view is centred around the current viewpoint position. When interacting with the objects, the user can select an object and rotate the view around its centre point. The objects can also be viewed automatically from their corresponding principal planes. In additions, users can perform various manipulations with the 3D objects as shown in fig. 3. They can remove existing objects or add new instances by retrieving the data from the server storage. Furthermore, they can change the location, orientation, scaling, and colour/material information of the models. Any changes to the scene are immediately transmitted via the cloud-based server to other clients. If multiple users access the same object at the same time, the server process gives a permission only to the first user who initiated the interaction and locks the object from editing by the other users. The lock is lifted after the first user releases the object.

Collaborative tools To facilitate more efficient interaction with 3D models and to support the discussion process, we propose several tools for annotation, feature marking, and pointing inside the client application. Users can create annotations directly on the 3D models by placing a 3D marker on the surface of the object and attaching meta information, such as description, images, and HTML links (fig. 4). When the user creates the annotation, the server also stores the information about the creator’s viewpoint, which can be retrieved by other users at later points when reviewing the annotations. The annotation tools allow multiple users to edit the meta-data in a similar fashion as a Wiki. The annotation data can be provided in HTML format and also accessed via a web browser. When users discuss a particular section of a large 3D model, they may need to indicate a particular target using a virtual laser pointer as shown in fig. 5. The pointer tool implemented follows the mouse cursor as the host user hovers the mouse over the scene. The remote users can see the pointer as an arrow originating at the arm of the avatar and ending at the surface of an object.

240

Chapter Fifteen

During a discussion between two or more users, one user may be the leader explaining different parts of the 3D model to other users. In order to allow users to easily follow such discussion we include an option of tethering your viewpoint to the primary user. In this way each user’s view automatically follows the viewpoint changes made by the leader. During collaborative interaction, the communication channel between users plays an important role in conveying information through discussion. In the current implementation, our system provides a simple text-based chat interface that allows users to converse while sharing the 3D space. The conversation can also be stored on the cloud server. In the future, we plan to add an audio communication channel to current video communication. At this time, we are using Skype to provide users with voice chat support.

History Logging All activities, such as navigation, manipulation, use of tools, and communication through text chat are recorded in the history log on the cloud server storage. Such information is very useful to review previous activities and results of discussions, and to replicate them for future user studies and evaluations. Furthermore, the data could be also used to optimize the caching mechanism on the server side while generating more detailed pre-rendered images for more often-accessed parts of a large 3D model.

Network Protocols In our collaborative framework, all interactions and activities are routed through the cloud and managed by the collaboration management server. This server has the following elements: a list of connected users, a list of object instances, status of tool usage, and a message queue. These data are recorded in the tables on the cloud storage, while the management server updates them whenever any of the parameters are changed. The changes are also shared among all the connected users in real time to keep the distributed 3D environment consistent among users (see fig. 6). Newly connected clients are updated with the initial state of the 3D world at the time of connection to the server. To enable fast and efficient communication, we implemented asynchronous protocols on the server as follows.

Cloud-based Collaborative Framework

241

Each client sends their current viewpoint position and direction to the management server in every rendering frame, while the server distributes this information to other clients. The management server also keeps track of the state of all active 3D objects in the virtual world using a scene graph representation. Each 3D object is managed as an instance in the scene graph structure, which describes the parent–children relationship between the instances. Clients can add new instances, copy existing instances, change connections to parent nodes, and remove them entirely from the scene graph. The status of the collaborative tools, such as annotations and pointing tools, are also managed by this server. The information sent includes tool type, targeted object ID, the position, and the host user. A message queue is used for the chat function and other notifications from the server.

3D Display via Network To accomplish the distribution and rendering of large-scale 3D models with an interactive frame rate, our display server employs the novel levelof-detail (LOD) rendering method, which was based on Adaptive Tetrapuzzles.8 In this section, we first describe the improved LOD display method and provide details on the mesh representation and distribution protocols implemented for real-time streaming and rendering. We also introduce a hybrid method which combines pre-rendered image data and low-level geometry to improve the rendering quality and speed for clients with a low-bandwidth connection.

Hierarchical 3D Mesh Construction (Offline Process) The mesh hierarchy construction method of Adaptive Tetrapuzzles can be easily implemented and quickly run on small and medium models. In the case of models with very large number of triangles, the splitting process can be complex and time-consuming. In our framework, we therefore perform the splitting process not by triangles but rather by small meshes as described in fig. 7. In the first step, we define a voxel space of pre-determined resolution and generate a group of small meshes. Then, we decompose an input mesh model into smaller meshes using the voxel space. We sort each triangle to a single voxel which contains at least one of its vertices, and assign a

8

Cignoni et al., 2004.

Chapter Fifteen

242

group of meshes to the corresponding voxel so that it contains fewer triangles than the predefined value Nv . In the second step, we convert those divided meshes into a graph representation, G(V, E); whose vertices v V correspond to one mesh assigned to one voxel, and whose edges e E represent the adjacency between two voxels. Then we recursively split the graph into two subgraphs by applying a graph-cut algorithm until the number of triangles in each sub-graph becomes less than a pre-defined threshold Nl, defining the maximal number of triangles in a section. Finally we simplify meshes at the non-leaf nodes to reduce the size to a pre-defined number Nn, where we use a quadric error metric simplification.9 We store the quadric error value epsilon, meaning an approximated distance between original and simplified meshes, as the level of simplification used for the calculation of mesh resolution.

Display Method (Online Process) The hierarchical data is constructed beforehand and uploaded to the server. When a user connects to the server and sends the request for a selected 3D model, it also includes its current 3D viewpoint in order to obtain a mesh with higher detail for a particular viewing direction. The server process traverses the constructed hierarchy from the root in depth-first order based on the user’s viewpoint. If the process finds a leaf node, or a node whose simplified mesh has sufficient resolution, the process adds that mesh to a rendering list, and tracks back. The necessary resolution is calculated by the projected quadric error, where is the quadric error of that node, and r is the distance between a viewpoint and the centre of the mesh. When the processed node is determined to be outside the visibility cone or occluded, the algorithm will immediately track back without adding any nodes to the rendering list. After traversing the entire tree structure, the only the geometric data of nodes included in the final rendering list is sent to the client side. The client process puts received data into the rendering pipeline, and renders the model.

Hybrid Rendering Method Our display server also has an option to use pre-rendered images to enhance the quality of the rendering when the client is not able to 9

Garland - Heckbert, 1997

Cloud-based Collaborative Framework

243

download high-resolution geometric data with sufficient speed. The method is based on Grid-Lumigraph.10 The image data are rendered on the server with the geometry of the highest level of detail. The generated images are then used on the client side to texture map a model that is of lower geometric complexity while attaining a relatively high level of detail for the rendering. With this approach, we can adapt to the bandwidth or other limitations of the clients. When using the hybrid rendering mode, we preliminarily construct a pre-rendered image repository for each model offline. We evenly sample the voxel space around the object, and render images of that object from the defined points in six directions along the main axes, i.e. positive and negative directions of x, y, and z. Each image is stored on the server side with unique index represented by (xi , yi , zi , d), which denote the location of the grid point and the viewing direction. Once the image data is uploaded to the server, a client requests a particular model and sends its current viewpoint. The server processes the request and retrieves the pre-rendered images from the repository for the closest viewpoint. The images are sent to the client along with the sample point parameters. During the display process on the client side, the retrieved images are rendered on the geometric data as texture projections (see fig. 8) and multi-pass rendering powered by the GPU.

Virtual Presence via Kinect For collaborative activities between geographically distributed users in a shared virtual environment, the form of presence in this environment is an important component affecting the experience of interaction. Avatars are commonly used in online gaming and social metaverse applications. In our framework, we include by default a simple skeletal avatar to represent each user in the 3D environment. To further enhance the collaborative experience, users can generate realistic 3D avatars using Microsoft Kinect. The Kinect camera captures colour and depth image sequences which are transformed into a 3D mesh and rendered in the shared virtual environment. Such realistic avatars allow users not only to express their presence, but also to communicate with other users via body and hand gestures in a more intuitive way. Our system also supports simple gesture-based interaction for some of the manipulations with the 3D models. For example, users can use their

10

Okamoto et al. 2011.

244

Chapter Fifteen

hand gestures to grasp, move, and rotate objects or navigate through the 3D space. We use Microsoft Kinect SDK to interface with the camera and obtain the colour and depth frame. As described in fig. 9, the client process extracts the bounding box of the image area covered by the user. The image data is then compressed for transmission to the server. The image data is compressed using standard JPEG compression while the original depth data is compressed using loss-less compression, such as ZIP. The original depth data is reduced from 13-bits per pixel to 8-bits by quantizing the range of detected user’s depth to 256 depth levels. The client process then sends the combined colour and depth package to the server. The server process sends those images to remote users, where the data are decompressed and rendered as a 3D point cloud/mesh in the same 3D space as the static 3D models. As described above, the server framework has a dedicated server instance that runs independently of the collaboration, management and 3D model-sharing services to avoid processing slowdown from the workload of Kinect image transfer. The transmission rate of the Kinect data is determined from the available throughput and can be adjusted appropriately for low bandwidth connections.

Case Study We applied our system to the collaborative discussion in an archaeological project aimed at researching copying of sculptures in the ancient Roman stone workshop as shown in the left of fig. 10. The main goal of this project is to compare and analyse many sculptures and fragments of different Roman periods and determine which of the sculptures are copies of others and what was the copying propagation path. The researchers and collaborators in this project are located in several different countries; however, they need to collaborate closely and share information about targeted workshop and excavation sites and corresponding artifacts. In our preliminary tests, we demonstrated the utility of our remote collaborative system for discussion between archaeologists from several different countries, as shown in the centre and right of fig. 10. In this demonstration, the experts were comparing different 3D models of the sculptures obtained via scanning, discussing particular features, and recording their observation results in the form of annotations and notes. Based on these preliminary tests and feedback from the archaeologists, we

Cloud-based Collaborative Framework

245

updated the design of several collaborative tools, the methods for object interaction, use of annotations, and pointer tools.

Conclusion and Future Work In this chapter, we set out our design for a 3D collaborative framework based in the cloud computing environment that can provide a collaborative interface and tools, distribute 3D models, and facilitate immersive telepresence via Kinect avatars. We introduced a novel hierarchical data structure and hybrid rendering system that allows for optimal transmission of only the minimum required geometric and image information for sharing and rendering of very large 3D models. In our system, the users are able to share the 3D space while being able to freely navigate the virtual space, interact with various objects, and communicate with other users. In the future, we will implement additional tools for collaboration and communication between remote users in various disciplines of science and engineering. Based on the feedback from the preliminary tests with the archaeologists, we believe that several features of our client software can be re-used for other types of users, while several specialized tools are needed for a particular discipline. We plan to implement various tools for geometry measurements and tools for real-time editing of large 3D meshes. Furthermore, we plan to evaluate the efficiency and effectiveness of our system for remote collaboration as compared to more traditional forms of remote collaboration such as the use of desktop sharing. We will also explore how the proposed framework can be extended to more immersive interaction using large 3D displays, voice chatting functions, and more intuitive gesture-based interaction.

246

Chapter Fifteen

Fig. 1. The overview of the collaborative framework for real-time large 3D model interaction. Each client is connected to all three servers.

Fig. 2. The client application window. On this system users can share, manipulate 3D objects, and communicate with other users.

Cloud-based Collaborative Framework

247

Fig. 3. Collaborative manipulation. Each user can manipulate 3D objects in the shared 3D space with the following operations: translation, rotation, change of scale, and change of color/material properties.

Fig. 4. The annotation tool. Users can add annotation markers on the surface of the 3D object and create information pages formatted in HTML that can be shared among remote users.

248

Chapter Fifteen

Fig. 5. Pointer tool. User A is using the pointer tool (in the first person perspective) as a host user while remote users, such as User B, can see the avatar of User A pointing at the specific feature with an arrow.

Fig. 6. Network protocol for collaboration. Each client sends their viewpoint, the states of active objects, the usage of collaborative tools, and chat messages in each display frame. The server sends all connected users’ states, the changed states of manipulated objects, the usage of tools by other users, and messages.

Cloud-based Collaborative Framework

249

Fig. 7. Hierarchical 3D mesh. The left image represents the process of recursive decomposition of meshes based on the voxel space. The input model is decomposed as a tree structure, simplified by each tree node, and converted to small mesh-patch based LOD representation such as shown on the right.

Fig. 8. Hybrid rendering using geometry and image. We pre-render the appearance of the 3D object from grid points around that object. In the online display process, the images rendered at the closest points to the user’s active viewpoint can be used to improve the rendering quality by projecting it onto the sparse geometric data as a texture.

250

Chapter Fifteen

Fig. 9. Sharing process of user’s 3D images. We clip, quantize, and compress the depth and color images from Kinect capture data. That data is transferred via the cloud and shared with other users.

Fig. 10. Case study of the collaboration work in the Roman sculptures’ research project. The comparison of ancient Roman sculptures (top left) and the scenes of demonstration and presentation of collaborative discussion between remote locations.

BIBLIOGRAPHY

Adler, F. 1892. Der grosse Altar des Zeus in der Altis, in: Curtius – Adler 1892, 210–214. Alonso Troncoso, V. 2012. Olympie et la publication des traités internationaux, Rivista di Diritto Ellenico 2, 3–24. Anderson, G. 2009. It’s how you tell them. Some Aspects of Lucian’s Anecdotes, in: A. Bartley (ed.), A Lucian for our Times, Cambridge, 3– 10. Angelucci, M. 2011. Polemon’s contribution to the periegetic literature of the II century B.C., Hormos. Ricerche di Storia Antica 3, 326–341. Ashmole, B. – Yalouris, N. 1967. Olympia. The Sculptures of the Temple of Zeus, London. Auffarth, Ch. 2010. The Materiality of God’s Image. The Olympian Zeus and Ancient Christology, in: J. N. Bremmer – A. Erskine (eds.), The Gods of Ancient Greece: Identities and Transformations. Edinburgh Leventis Studies 5, Edinburgh, 465–480. Babelon, E. 1907. Traité des monnaies grecques et romaines, I.2, Paris. —. 1914. Traité des monnaies grecques et romaines, III.2, Paris. Baitinger, H. – Eder, B. 2001. Hellenistische Stimmarken aus Elis und Olympia: neue Forschungen zu den Beziehungen zwischen Hauptstadt und Heiligtum, JdI 116, 163–243. Baitinger, H. – Eder, B. 2003. Tesserae di voto ellenistiche da Elide e da Olimpia. Nuove ricerche sulle relazioni fra la città e il santuario, Geographia Antiqua 12, 95–109. Baitinger, H. 2001. Die Angriffswaffen aus Olympia. OlForsch 29, Berlin. Bankel, H. 1991. Akropolis-Fußmaße, AA 1991, 151–163. —. 1993. Der spätarchaische Tempel der Aphaia auf Aegina, DAA 19. Berlin. Barringer, J. M. 2005. The Temple of Zeus at Olympia, Heroes, and Athletes, Hesperia 74, 211–241. —. 2008. Art, Myth, and Ritual in Classical Greece. Cambridge. —. 2009. The Olympic altis before the temple of Zeus, JdI 124, 223–249. Becatti, G. 1951. Problemi Fidiaci. Milano. Bell. M. 1981. Morgantina Studies, I. The Terracottas, Princeton. Benecke, H. 2006a. Die Tierreste, in: Kyrieleis 2006, 247–248.

252

Bibliography

—. 2006b. Animal sacrifice at the Late Archaic Artemision of Olympia: the archeozoological evidence, in: U. Tecchiati – B. Sala (Hg.), Studi di archeozoologia in onore di Alfredo Riedel, Bolzano, 153–160. Benford, S. et al. 1995. User embodiment in collaborative virtual environments, in: Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (CHI ’95), 242–249, 1995. Bengtson, H. 1962. Die Staatsverträge des Altertums, Vol. 2: Die Verträge der griechisch-römischen Welt von 700 bis 338 v.Chr., München. Berger, E. 1984. Zum Maß- und Proportionssystem des Parthenon: Ein Nachwort zur Diskussion des Bauentwurfes, in: E. Berger (Hrsg.), Parthenon-Kongreß Basel 1982, Mainz, 119–174. Beschi, L. 1965. Policleto, in: EAA VI, 266–274. Besl, P. J. – Mckay, N.D. 1992. A method for registration of 3-D shapes, IEEE Transactions on Pattern Analysis and Machine Intelligence, 14(2), 239–256. Biesantz, H. 1965. Die thessalischen Grabreliefs, Mainz. Blass, F. 1884. Die eleischen Inschriften, in: H. Collitz (Hg.), Sammlung der griechischen Dialekt-Inschriften. I., Göttingen. Boardman, J. 1974. Athenian Black Figure Vases. A Handbook, London. —. 1985. Greek Sculpture. The Classical Period. A handbook. London. Bocher, S. 2012. Pelopion und Heraion und ihre Bedeutung im Heiligtum von Olympia, in: Mythos Olympia 2012, 137–143. Boedeker, D. 2001. Paths to Heroization at Plataea, in: D. Boedeker - D. Sider (eds.), The new Simonides. Contexts of Praise and Desire, Oxford, 148–163. Boesch, P. 1908. . Untersuchung zur Epangelie griechischer Feste, Göttingen. Bol, P. C. 1988. Die ‚Artemis von Lusoi‘. Eine klassische Wiedergabe eines frühgriechischen Kultbildes in: M. Schmidt (ed.), Kanon. Festschrift E. Berger (AntK Beih. 15), Basel, 76–80. Bol, R. and P. 2011. Die Statuengruppe der iulisch-claudischen Dynastie im Olympischen Metroon, in: S. Pingiatoglou – T. Stefanidou-Tiveriou (eds.), , , Thessaloniki, 107–123 Bol, R. 1984. Das Statuenprogramm des Herodes-Atticus-Nymphäums (OlForsch 15), Berlin. —. 1990. Die Amazone des Polyklet, in: H. Beck – P. C. Bol – M. Bückling (eds.), Polyklet. Der Bildhauer der griechischen Klassik. Ausstellungkatalog, Frankfurt am Main, 213–239. —. 1998. Amazones Volneratae. Untersuchungen zu den Ephesischen Amazonenstatuen, Mainz.

New Approaches to the Temple of Zeus at Olympia

253

—. 2008. Die Bildnisstatue der Antonia Cleodice im Kontext ihrer Aufstellung im Olympischen Heraion, in: Amicitiae Gratia, , Athens, 149–156. Borbein, A. H. 1996. Polykleitos. in: O. Palagia – J. J. Pollitt (eds.), Personal Styles in Greek Sculpture, (Yale Classical Studies 30) Cambridge, 66–90. Borda, M. 1953. La scuola di Pasiteles. Bari. Brinker, W. – Garbrecht, G. 2007. Die Bedeutung der Zisternen im Wasserversorgungssystem des Burgbergs von Pergamon, in: Chr. Ohlig (ed.), Antike Zisternen (= Schriften der Deutschen Wasserhistorischen Gesellschaft e.V. 9). Siegburg: DWhG, 117–144. Brinkmann, V. 1994. Beobachtungen zum formalen Aufbau und zum Sinngehalt der Friese des Siphnierschatzhauses, München. Brulé, P. – Touzé, R. 2008. Le hiereion: phusis et psuchè d’un medium, in: V. Mehl – P. Brulé (ed.), Le sacrifice antique. Vestiges, procédures et stratégies, Rennes, 111–138. Buchert, U. 2000. Denkmalpflege im antiken Griechenland. Maßnahmen zur Bewahrung historischer Bausubstanz, Frankfurt a. M. Bulle, H. 1939. Der Ostgiebel des Zeustempels zu Olympia, JdI 54, 137– 218. Burkert, W. 1984. Die orientalisierende Epoche in der griechischen Religion und Literatur, Heidelberg. —. 1985. Greek Religion (tr. by J. Raffan). Cambridge. Buschor, E. – Hamann, R. 1924. Die Skulpturen des Zeustempels zu Olympia, Marburg. Buschor, E. 1932. Glockenkrater in Neapel. Pelops und Oinomaos, in: A. Furtwängler – K. Reichhold, Griechische Vasenlamalerei Auswahl hervorragender Vasenbilder (Serie III), München, 151–159. Bux, E. 1930, s. v. Marcellinus [49], RE 14, 1450 –1487. Büsing, H. 1982. Metrologische Beiträge, JdI 97, 1–45. —. 1986. Zweierlei Maß am Artemision von Ephesus?, AA 1986, 205-208. Capra, Fr. 2008. Le monnayage d’argent d’Élis. Images monétaires et organisation de la production, du VIe siècle au IIe siècle avant J.-C., thèse de doctorat inédite, université Paris IV-Sorbonne. Castriota, D. 2000. Justice, Kingship and Imperialism. Rhetoric and Reality in Fifth-century B.C. Representations Following the Persian Wars, in: B. Cohen (ed.), Not the Classical Ideal. Athens and the Construction of the Other in Greek Art, Leiden, 443–79. Chandezon, C. 2003. L’élevage en Grèce (fin Ve – fin Ier s. a.C.). L’apport des sources épigraphiques, Paris.

254

Bibliography

Chaniotis, A. 1996. Die Verträge zwischen kretischen Poleis in der hellenistischen Zeit, Stuttgart. —. 2011. Greek festivals and contests: definition and general characteristics, ThesCRA VII, 4–43. Chantraine, H. 1958. Zur Münzprägung von Chalkis im 6.– 5. Jahrhundert, JNG 9, 7–17. Ciardiello, R. 2007. Le sculture della Villa dei Papiri a Ercolano: nuove metodologie e tecniche diagnostiche, CronErcol 37, 161–169. Cignoni, P. et al. 2004. Adaptive tetrapuzzles: efficient out-of-core construction and visualization of gigantic multiresolution polygonal models. ACM Trans. Graph., 23 (3):796–803. Cook, B. F. 2004. Greek Inscriptions (4th ed.), London. Coulton, J. J. 1977. Ancient Greek Architects at Work, Ithaca. Currie, B. 2011. Epinician choregia: funding a Pindaric chorus, in: L. Athanassaki – E. Bowie (eds.), Archaic and Classical Choral Song. Performance, Politics and Dissemination, Berlin, 269–310. Curtius, E. – Adler, F. – Hirschfeld, G. (eds.) 1879. Die Ausgrabungen zu Olympia. vol. 3: Übersicht der Arbeiten und Funde vom Winter und Frühjahr 1877 – 1878. Berlin: E. Wasmuth. Dahlmann, F.C. 1823. Herodot: Aus seinem Buche Sein Leben, Kiel. Davison, C. – Lundgreen, B. 2009. Pheidias. The Sculptures and Ancient Sources, Bulletin of the Institute of Classical Studies supp. 105. London: Institute of Classical Studies. Delivorrias, A. 1995. Polykleitos and the allure of feminine beauty, in: Moon 1995, 200–217. Dinsmoor, W.B. 1950. The Architecture of Ancient Greece, New York. —. 1938. Archaeology and Astronomy, Proceedings of the Am. Philos. Soc. 80, 95–173. —. 1941. An Archaeological Earthquake at Olympia, AJA 45, 399–427. —.. 1961. The Basis of Greek Temple Design: Asia Minor, Greece, Italy, in: P. Romanelli (ed.), Atti del settimo congresso internationale di archeologia classica 1, Roma, 355–368. —. 1973. The Architecture of Ancient Greece (3rd ed.), New York. Diodato, R. 2005. Estetica del virtuale. Milano. Donderer, M. 1988. Nicht Praxiteles, sondern Pasiteles. Eine signierte Statuenstütze in Verona. ZPE 73, 63–68. Dörpfeld, W. 1890. Metrologische Beiträge, AM 15, 167–187. —. 1892. Der Zeustempel, in: Curtius – Adler 1892 text 2, 4–22, vol. plates 1, pl. 2, 10–17. Berlin: Asher. —. 1897. Lageplan der antiken Bauwerke, in: E. Curtius – F. Adler (eds.), Olympia. Die Ergebnisse der von dem deutschen Reich veranstalteten

New Approaches to the Temple of Zeus at Olympia

255

Ausgrabung, vol. 1. Topographie und Geschichte von Olympia, Berlin, 69–88. —. 1913. Die Beleuchtung der griechischen Tempel, Zeitschrift für Geschichte der Architektur 6, 1913, 1–12. —. 1935. Alt – Olympia. Untersuchungen und Ausgrabungen zur Geschichte des ältesten Heiligtums von Olympia und der älteren griechischen Kunst. Berlin: Mittler & Sohn. Eckstein, F. 1969. Anathemata. Studien zu den Weihgeschenken Strengen Stils im Heiligtum von Olympia, Berlin. Eidinow, E. 2007. Oracles, Curses, and Risk Among the Ancient Greeks, Oxford. Ekroth, G. 2008a. Burnt, cooked or row? Divine and human culinary desires at Greek animal sacrifice, in: E. Stavrianopoulou – C. Ambos (eds.), Transformations in Sacrificial Practices. From Antiquity to Modern Times (Performanzen – Performances, 15), Berlin, 87–111. —. 2008b. Meat, man and god. On the division of the animal victim at Greek sacrifices, in: A. P. Matthaiu (ed.), Mikros ieromn m n. Meletes eis mn m n Michael H. Jameson, Athens, 259–290. —. 2009. Thighs or tails? The osteological evidence as a source for Greek ritual norms, in: P. Brulé (ed.), La norme en matière religieuse en Grèce ancienne. Actes du XIe colloque du CIERGA (Rennes, septembre 2007), Liège, 125–151. Faraone, C. A. 1991. The agonistic context of early Greek binding spells, in C. A. Faraone and D. Obbink (eds.), Magika Hiera: ancient Greek magic and religion, New York, 3–32. Feyel, C. 2006. La dokimasia des animaux sacrifiés, RPh 80, 33–55. Forbat, F. 1935. Der Fußboden im Innern des Zeus-Tempels, in: W. Dörpfeld, Alt – Olympia. Untersuchungen und Ausgrabungen zur Geschichte des ältesten Heiligtums von Olympia und der älteren griechischen Kunst. Berlin: Mittler & Sohn. pp. 226–247. Forsdyke, S. 2006. Herodotus, Political History and Political Thought, in: C. Dewald – J. Marincola (eds.), The Cambridge Companion to Herodotus, Cambridge, 224–241. Frisch, P. 1988. Die Klassifikation der bei den griechischen Agonen, ZPE 75, 179–185. Fuchs, K. 2012. Götter und Kulte in Olympia, Antike Welt 2012, 4, 20–23. Fuchs, M. 1999. In hoc etiam genere Graeciae nihil cedamus. Studien zur Romanisierung der späthellenistischen Kunst im 1. Jh. v. Chr., Mainz. Fuchs, W. 1983. Die Skulptur der Griechen, München. Furtwängler, A. 1880. Die Bronzefunde aus Olympia und deren kunstgeschichtliche Bedeutung, AbhBerlin 1879.

256

Bibliography

—. 1893. Meisterwerke der griechischen Plastik, Leipzig/Berlin. —. 1893. Zu den olympischen Skulpturen, in: A. Furtwängler – G. Körte – A. Milchhöfer (Hrsg.), Archäologische Studien ihrem Lehrer Heinrich Brunn, Berlin, 67–91. Gabellone, F. 2009. Ancient contexts and Virtual Reality: From reconstructive study to the construction of knowledge models, Journal of Cultural Heritage 10, 112–117. Gabellone, F. et al. 2013a. From museum to original site: 3d environment for the virtual visit of finds re-contextualized in their original provenance, in: Digital Heritage International Congress, 28 Oct – 1 Nov,2013, Marseille, France Vol.2, Marseille, 215–222. Gabellone, F. et al. 2013b. Development of realistic Virtual Environment for a project of Museum Communication, in: 6th International Congress “Science and Technology for the Safeguard of Cultural Heritage in the Mediterranean Basin”, 22 – 25 October, Athens, vol. III, 41–49. Gabellone, F., Giannotta, M.T. 2013a. “Marta Racconta”: a project for the virtual enjoyment of inaccessible monuments, CHNT 18, International Conference on Cultural Heritage and New Technologies, Stadtarchäologie, Wien, http://www.chnt.at/wp-content/uploads/Gabellone_etal_2_2014.pdf. Gabellone, F. Giannotta, M.T 2013b. Monumenti inaccessibili della necropoli greca di Taranto: un modello di fruizione virtuale basato su interfacce naturali, paper presented at XVIIIth International Congress of Classical Archaeology, Merida, 13-17 May 2013. Gardner, P. 1887. A Catalogue of Greek Coins in the British Museum. Peloponnesus (excluding Corinth), London. Garland, M. - Heckbert, P. S. 1997. Surface simplification using quadric error metrics, in Proceedings of the 24th annual conference on Computer graphics and interactive techniques, 209–216. Gauer, W. 2012a. Der Zorn des Zeus und die klassische Kunst der Griechen. Einladung zu einer Griechenlandreise. Wiesbaden. —. 2012b. Brunnenfunde, Festgesandschaften und Festgesellschaften, in: Mythos Olympia 2012, 99–103. Gauthier, P. 1984. La dokimasia des victimes. Note sur une inscription d’Entella, ASNSP III 14, 3, 845–848. Gauthier, Ph. 1975. Légendes monétaires grecques, in: J.-M. Dentzer - Ph. Gauthier - T. Hackens (éds.), Numismatique antique. Problèmes et méthodes, Nancy-Louvain, 165–179. Gehrke, H.-J. 2012. Olympia in der Geschichte, in: Mythos Olympia 2012, 29–35.

New Approaches to the Temple of Zeus at Olympia

257

Gerkan, A. von 1930/1959. Grundlegendes zur Darstellungsmethode. Kursus für Bauforschung. Veranstaltet von der Koldewey-Gesellschaft München, 12.–14. Juni 1930, reprint in: A. von Gerkan (ed. by E. Boehringer), Von antiker Architektur und Topographie. Gesammelte Aufsätze. Stuttgart: Kohlhammer, 99–106. Ginouvès, R. – Martin, R. 1985. Dictionnaire methodique de l’architecture grecque et romaine I. Matériaux, techniques de construction, techniques et formes du décor, Roma. Giuliano, A. (ed.) 1979. Museo Nazionale Romano. Le sculture, 1,1, Roma. Gobbetti, E. - Marton, F. 2005. Far voxels: a multiresolution framework for in- teractive rendering of huge complex 3d models on commodity graphics platforms. ACM Trans. Graph., 24(3):878–885. Gribble, D. 2012. Alcibiades at the Olympics: performance, politics and civic ideology, CQ 62, 45–71. Griffiths, A.H. 1976. What Syagrus said: Herodotus 7.159, LCM 1, 23–24. Gruben, G. 2001. Griechische Tempel und Heiligtümer. München: Hirmer (1. edition Tempel der Griechen. München: Hirmer, 1966). Grunauer, P. 1974. Der Westgiebel des Zeustempels von Olympia. Die Münchner Rekonstruktion - Aufbau und Ergebnisse, JdI 89, 1–49. —. 1980. Die Grabungen am Zeustempel in Olymnpia im Herbst und Winter 1977/1978, in: Bericht über die 30. Tagung für Ausgrabungswissenschaft und Bauforschung 1978 in Colmar, ed. by Koldewey-Gesellschaft [s.l. s.n.] 21–28. —. 1981. Zur Ostansicht des Zeustempels, in: A. Mallwitz (ed.) 10. Bericht über die Ausgrabungen in Olympia. Berlin: W. de Gruyter, 256–301. Grüner, A. 2005. Wer erfand die Perspektive? Eine Beobachtung zum attischen Grabrelief, in: T. Ganschow – Matthias Steinhart (eds.), Otium. Festschrift Michael Strocka. Remshalden: Bernhard Albert Greiner, 107–114. Gschnitzer, F. 1958. Abhängige Orte im griechischen Altertum, München. Hafner, G. 1997. Polyklet. Doryphoros. Revision eines Kunsturteils, Frankfurt. Hallett, C. H. 1995. The replica of Polykleitos’ Doryphoros in the Minneapolis Institute of Arts: An archeological description, in: Moon 1995, 116–120. Harrison, E.B. 1972. The South Frieze of the Nike Temple and the Marathon Painting in the Painted Stoa, AJA 76, 353–378.

258

Bibliography

Haysom, M. – Wallensten, J. 2011. Current Approaches to Religion in Ancient Greece. Papers Presented at a Symposium at the Swedish Institute at Athens, 17 – 19 April 2008, Stockholm. Head, B.V. 19112. Historia Numorum. A Manual of Greek Numismatics, 2nd ed., Chicago. Heiden, J. 2003. Thessalische Lapithen in Elis. Zur Deutung des Westgiebels von Olympia, AA 2003, 183–190. —. 2012. Artemis-Altäre, in: Mythos Olympia 2012, 144–147. Heilmeyer, W.-D. 2012. Zeus in Olympia – der Zeus-Altar und das ZeusBild des Pheidias, in: Mythos Olympia 2012, 113–119. Hellmann, M.-Ch. 2002. L’architecture grecque I. Les principes de la construction, Paris. Hemingway, S. 2000. Bronze sculpture, in: R. Ling (ed.), Making Classical Art. Process and Practice, Stroud 37–46. Hennemeyer, A. 2006. Neue Forschungsergebnisse zur Cella des Zeustempels von Olympia, Bericht über die 43. Tagung für Ausgrabungswissenschaft und Bauforschung der Koldewey-Gesellschaft vom 19.–23. Mai 2004 in Dresden. Bonn: [s.n.] 103–111. —. 2006a. Athenaheiligtum von Priene. München. —. 2010. Zeustempel von Olympia, Griechenland; in: W. Nerdinger (ed.) Geschichte der Rekonstruktion – Rekonstruktion der Geschichte. Exhibition Munich 2010. München: Prestel, 218–219. —. 2011. Zur Lichtwirkung am Zeustempel Olympia, in: P. Schneider – U. Wulf-Rheidt (eds.), Licht-Konzepte in der vormodernen Architektur. Diskussionen zur Archäologischen Bauforschung 10. Regensburg: Schnell+Steiner, 101–110. —. 2012. Der Zeus-Tempel von Olympia, in: Mythos Olympia 2012, 120– 125. 448. 452). —. 2013 a. Kontinuität und Wandel. Beobachtungen am Zeustempel von Olympia, in: I. Gerlach – D. Raue (eds.), Sanktuar und Ritual. Heilige Plätze im archäologischen Befund (Menschen – Kulturen – Traditionen. Forschungscluster 4, Bd. 10). Rahden/Westf: Marie Leidorf, 19–26. —. 2013 b. Zum Umbau des Phidias im Zeustempel von Olympia, Architectura. Zeitschrift zur Geschichte der Baukunst 43, 1, 1–18. —. 2013 c. Die Wiederherstellung des Zeustempels in klassischer Zeit, in: Vinzenz Brinkmann (ed.), Zurück zur Klassik. Ein neuer Blick auf das Alte Griechenland. exhibition Frankfurt/Main, München: Hirmer, 126– 129. Hermary, A. et al. 2004. Les sacrifices dans le mond grec, ThesCRA I, 60– 134.

New Approaches to the Temple of Zeus at Olympia

259

Herrmann, H.-V. 1972. Olympia. Heiligtum und Wettkampfstätte. München. —. 1987. Die Olympia-Skulpturen. Darmstadt. Herrmann, K. 1972. Der Pfeiler der Paionios-Nike in Olympia, JdI 87, 232–257. —. 1981. Zu den Antenkapitellen des Zeustempels, in: A. Mallwitz (ed.) 10. Bericht über die Ausgrabungen in Olympia. Berlin: de Gruyter, 302–317, pl. 30–32. —. 1983. Zum Dekor dorischer Kapitelle, Architectura. Zeitschrift zur Geschichte der Baukunst 13, 1–12 —. 1999. Die Stützmauer am Kronoshügel, in: A. Mallwitz (ed.), 11. Bericht über die Ausgrabungen in Olympia. Berlin: de Gruyter, 367– 390; Taf. 84–87. —. 2000. Zur Verwendung des parischen Marmors im Heiligtum von Olympia, in: D. Schilardi et al. (eds.), : , : A’ , , 2-5 1997 = Paria Lithos: Parian quarries, marble and workshops of sculpture : proceedings of the First International Conference on the archaeology of Paros and the Cyclades, Paros, 2-5 October 1997. Athens, 379–389. —. 2013a. Restaurierungsarbeiten am Zeustempel und Philippeion, I. Voraussetzungen und Wiederaufbau der Zeustempelsäule N12, in: H. Kyrieleis (ed.) 13. Bericht über die Ausgrabungen in Olympia. Berlin: Wasmuth, 51–76. —. 2013b. Untersuchungen am Heraion-Altar. I. Die baulichen Überreste des Altars. II. Zu den Falzplatten und Relikten einer Opferstätte, in: XIII. Olympia-Bericht, 100–114. Hill, G. F. 1906. Historical Greek Coins, London. Hitzig, H. – Blümner, H. 1901. Des Pausanias Beschreibung von Griechenland III, Leipzig. Hitzl, K. 1991. Die kaiserzeitliche Statuenausstattung des Metroon (OlForsch 19), Berlin. Hoepfner, W. 1971. Zwei Ptolemaierbauten. Das Ptolemaierweihgeschenk in Olympia und ein Bauvorhaben in Alexandria (AM Beih. 1) Berlin. Hollinshead, M. B. 1999. “Adyton”, “Opisthodomos”, and the Inner Room of the Greek Temple, Hesperia 68, 189–218. Holloway, R. R. 1958. The Date of the Eleusis Relief, AJA 62, 403–408. Hoppe, H. 1996. Progressive meshes. In SIGGRAPH 96, 99–108.

260

Bibliography

Horn, B. K. P. 1987. Closed-form solution of absolute orientation using unit quaternions, Journal of the Optical Society of America, 4, 629– 642. Höcker, Ch. 1993. Planung und Konzeption der klassischen Ringhallentempel von Agrigent, Frankfurt. —. 2004. Metzler Lexikon antiker Architektur: Sachen und Begriffe, Stuttgart. Hölscher, T. 1973. Griechische Historienbilder des 5. und 4. Jahrhunderts v. Chr., Würzburg. —. 1974. Die Nike der Messenier und Naupaktier in Olympia. Kunst und Geschichte im späten 5. Jahrhundert v. Chr., JdI 89, 70–111. —. 2002. Rituelle Räume und politische Denkmäler im Heiligtum von Olympia, in: Olympia 1875 – 2002, 331–345. Hurwit, J. 2005. The Parthenon and the Temple of Zeus at Olympia, in: J. Barringer – J. Hurwit (eds.), Periklean Athens and its Legacy, Festschrift Jerome Jordan Pollitt. Austin: University of Texas Press, 135–145. Ikeuchi, K. – Miyazaki, D. 2007. Digitally Archiving Cultural Objects. New York. Ioakimidou, Chr. 1997. Die Statuenreihen griechischer Poleis und Bünde aus spätarchaischer und klassischer Zeit, München. —. 2000. Auch wir sind Griechen! Statuenreihen westgriechischer Kolonisten in Delphi und Olympia, Nikephoros 13, 63–94. Jacoby, F. 1913. s. v. Herodotus, RE Suppl. 2, 205–520. Jacquemin, A. 2001. Pausanias, témoin de la religion grecque dans le sanctuaire d’Olympie, dans Olympie. Cycle de huit conférences organisées au musée du Louvre par le Service culturel, du 18 janvier au 15 mars 1999, A. Pasquier (dir.), Paris, 181–213. Jameson, M. H. 1988. Sacrifice and animal husbandry in classical Greece, in: C.R. Whittaker (ed.), Pastoral Economies in Classical Antiquity, Cambridge, 87–119. Jeffery L. H. 1990. The Local Scripts of Archaic Greece. (Revised Edition, with a Supplement by A. W. Johnston), Oxford. Jim, T. S. F. 2014a. Sharing with the Gods. Aparchai and Dekatai in Ancient Greece, Oxford. —. 2014b. Aparchai in the great list of Thasian Theôroi, CQ 64, 13–24. Jones, C. P. 1986. Culture and Society in Lucian, Cambridge (MA). —. 2001. Pausanias and his Guides, in: S. E. Alcock, J. F. Cherry, J. Elsner (eds.), Pausanias. Travel and Memory in Roman Greece, Oxford, 33– 39.

New Approaches to the Temple of Zeus at Olympia

261

Jongkees, J. H. 1939. Zur Chronologie der Münzen von Olympia, JdI 54, 219–229. Jordan, D. R. 1985. A survey of Greek defixiones not included in the special corpora, GRBS 26, 5–46. Kaiser, S. I. 2010. Die Fragmente des Aristoxenos aus Tarent, Hildesheim. Kaltsas, N. 2002. Sculpture in the National Archaeological Museum, Athens, Los Angeles. Kane, S. et al. 1999. Investigation of the Pentelic marble sculpture of the Nymphaeum of Herodes Atticus in Olympia, Greece, in: M. Schvoerer (ed.), Archéomatériaux: Marbres et autres roches, ASMOSIA IV, Bordeaux, 317–323. Kavoulaki, A. 2011. Observations on the meaning and practice of Greek pompe (procession), in: Haysom – Wallensten 2011, 135–150. Kim, D. and Lee, H. 2012. Interactive rendering of huge 3d meshes in cloud computing. In The Eighth International Conference on Autonomic and Autonomous Systems (ICAS 2012), 38–41. Klöckner, A. 2010. Getting in Contact. Concepts of Human-Divine Encounter in Classical Greek Art, in: J.N. Bremmer - A. Erskine (eds.), The Gods of Ancient Greece. Identities and Transformations, Edinburgh, 106–125. Knell, H. 1985. Vitruvs Architekturtheorie, Darmstadt. —. 1988. Architektur der Griechen, Darmstadt. —. 1990. Mythos und Polis. Bildprogramme Griechischer Bauskulptur, Darmstadt. Knittlmayer, B. 1997. Die attische Aristokratie und ihre Helden: Untersuchungen zu Darstellungen des trojanischen Sagenkreises im 6. und frühen 5. Jahrhundert v. Chr, Heidelberg. Knoop, R. R. 1988. Remoulding and rates of shrinkage of architectural terracottas: The Archaic antefixes from Satricum (Le Ferriere, prov. Latina, Italy), MededRome 48, 75–85. Koenigs, W. 2002. Der Zeustempel im 19. und 20. Jahrhundert, in: Olympia 1875–2000, 131–146. Korff, W. F. 2009. Der Klang der Pyramiden, Hildesheim. Korres, M. 1994. Der Plan des Parthenon, AM 109, 53–120. K iv, M. 2013. Early history of Elis and Pisa: invented or evolving tradition?, Klio 95, 315–368. Kraay, C. M. 1976. Archaic and Classical Greek Coins, New York. Krauss, F. 1957 a. Beobachtungen am Zeustempel von Olympia, in: Neue Ausgrabungen im Nahen Osten, Mittelmeerraum und in Deutschland.

262

Bibliography

Bericht über die Tagung der Koldewey-Gesellschaft in Regensburg vom 23. bis 27. April 1957. [s.l. s.n.] (unpaginated). —. 1957 b. Die Säulen des Zeustempels von Olympia, in: E. Boehringer und W. Hoffmann (ed.) Robert Boehringer. Eine Freundesgabe. Tübingen: Mohr, 365–387. Kreikenbom, D. 1990. Bildwerke nach Polyklet. Kopienkritishe Untersuchungen zu den männlichen statuarischen Typen nach Polykletischen Vorbildern. ›Diskophoros‹, Hermes, Doryphoros, Herakles, Diadoumenos, Berlin. Kropp, A. 2010. How does magical language work? The spells and formulae of the Latin defixionum tabellae, in R. L. Gordon and F. Marco Simón (eds.), Magical Practice in the Latin West. Papers from the International Conference held at the University of Zaragoza, 30 Sept.–1 Oct. 2005, Leiden / Boston, 357–380. Krumeich, R. 2008. Vom Haus der Gottheit zum Museum? Zu Ausstattung und Funktion des Heraion von Olympia und des Athenatempels von Lindos, AntK 51, 73–95. Kruse, H.-J. 1975. Römische weibliche Gewandstatuen des zweiten Jahrhunderts n. Chr., Göttingen. Kunze, E. – Weber, H. 1948. The Olympian Stadium, the Echo Colonnade and the “archaeological earthquake”, AJA 52, 490–496. Kyrieleis, H. 1997. Zeus and Pelops in the East Pediment of the Temple of Zeus at Olympia, in: D. Buitron-Oliver (ed.), The Interpretation of Architectural Sculpture in Greece and Rome, Washington, 12–27. —. 2002. Zu den Anfängen des Heiligtums von Olympia, in: Olympia 1875 – 2000, 213–220. —. 2006. Anfänge und Frühzeit des Heiligtums von Olympia. Die Ausgrabungen am Pelopion, 1987 – 1996. (OlForsch 31), Berlin 2006. —. 2011. Olympia. Archäologie eines Heiligtums, Mainz. —. 2013. Pelops, Herakles Theseus. Zur Interpretation der Skulpturen des Zeustempels von Olympia, JdI 127/128, 51–124. La Rocca, E. 1996. Riace, bronzi di, in: EAA, secondo supplemento IV, Roma, 730–734. Landwehr, C. 1982. Griechische Meisterwerke in römischen Abgüssen. Der Fund von Baia. Zur Technik antiker Kopisten. Frankfurt. —. 1985. Die antiken Gipsabgüsse aus Baiae. Griechische Bronzestatuen in Abgüssen römischer Zeit. (Archäologische Forschungen 14), Berlin. —. 2010. The Baiae casts and the uniqueness of Roman copies, in: R. Frederiksen – E. Marchand (eds.), Plaster Casts. Making, Collecting and Displaying from Classical Antiquity to the Present, Berlin/New York, 35–46.

New Approaches to the Temple of Zeus at Olympia

263

Lapatin, K. 2001. Chryselephantine Statuary in the Ancient Mediterranean World. Oxford: Oxford University Press. Larfeld, W. 1902. Handbuch der griechischen Epigraphik II. Die attischen Inschriften, Leipzig. Latte, K. 1920. Heiliges Recht. Untersuchungen zur Geschichte der Sakralen Rechtsformen in Griechenland, Tübingen. Lee, H. M. 2001. The Program and Schedule of the Ancient Olympic Games (Nikephoros, Beihefte 1), Hildesheim. Lévy, P. 1997. Il Virtuale, Milano. Leypold, Chr. 2005. s. v. Tempel II. Gestaltung und Nutzung der Innenräume, ThesCRA 4, 100–112. Lombardo, M. 1992. Osservazioni sul decreto di Nacone, in: Giornate Internazionali di Studi sull’area elima. Gibellina, 19-22 settembre 1991. Atti I., Pisa – Gibellina, 421–442. Lugauer, M. 1967. Untersuchungen zur Symbolik des Apfels in der Antike, Bamberg. Luraghi, N. 2008. The Ancient Messenians. Cambridge. Maddoli, G. – Saladino, V. (eds.) 1995. Pausania. Guida della Grecia. Libro V. L’Elide e Olimpia, Milano. Maitland, J. 1996. Marcellinus‘ Life of Thukydides. Criticsm and Criteria in the Biographical Tradition, CQ 46, 538–558. Mallwitz, A. – Herrmann, H.-V. (eds.) 1980. Die Funde aus Olympia, Athens. Mallwitz, A. – Herrmann, K. 1999. Ergebnisse und Folgerungen, in: Id. (bearbeitet von K. Herrmann) (Hg.), XI. Bericht über die Ausgrabungen in Olympia. Frühjahr 1977 bis Herbst 1981, Berlin, 181–284. Mallwitz, A. – Schiering, W. 1964. Die Werkstatt des Pheidias in Olympia, OlForsch 5. Berlin: Walter de Gruyter. —. 1966. Das Heraion von Olympia und seine Vorgänger, JdI 81, 310– 376. —. 1972. Olympia und seine Bauten. Darmstadt: WBG. —. 1980. Kykladen und Olympia, in: . (Hrsg.), . , Athen, 361–379. —. 1981. Neue Forschungen in Olympia (Theater und Hestiaheiligtum in der Altis), Gymnasium 88, 97–122. —. 1988. Cult and competition locations at Olympia, in: Raschke 1988, 79–109. Mangold, M. 2000. Kassandra in Athen. Die Eroberung Trojas auf attischen Vasenbildern, Berlin.

264

Bibliography

Martin, R.1965. Manuel d’architecture grecque I: matériaux et technique, Paris. Masuda, T. 2002. Registration and integration of multiple range images by matching signed distance fields for object shape modeling, Computer Vision and Image Understanding, 87(1-3), 51–65. —. 2005. Registration and Deformation of 3D Shape Data through Parameterized Formulation (PhD thesis, The University of Tokyo) Mattern, T. 2007. Griechische Kultbildschranken, AM 122, 139–159. Mattusch, C. C. 1988. Greek Bronze Statuary: From the Beginnings Through the Fifth Century B.C., Ithaca. —. 2002. In search of the Greek original, in: E. K. Gazda (ed.), Ancient Art of Emulation: Studies in Artistic Originality and Tradition from the Present to Classical Antiquity. (Supplements to the Memoirs of the American Academy in Rome) Michigan, 99–115. —. 2005. The Villa dei Papiri at Herculaneum. Life and Afterlife of a Sculpture Collection. Los Angeles. McInerney, J. 2010. The Cattle of the Sun. Cows and Culture in the World of the Ancient Greeks, Princeton. McNiven, T. J. 2000. Behaving Like an Other. Telltale Gestures in Athenian Vase Painting, in: B. Cohen (ed.), Not the Classical Ideal. Athens and the Construction of the Other in Greek Art, Leiden, 71–97. Meister, R. 1889. Die griechischen Dialekte auf Grundlage von Ahrens’ Werk: “De Graecae linguae dialectis”. II. Eleisch, Arkadisch, Kyprisch, Göttingen. Mello, A. 2008. . “Rhetra” e le wratrai olimpiche.’ ZPE 167, 50–66. Mertens, D. 1984a. Zum klassischen Tempelentwurf, in: Bauplanung und Bautheorie der Antike, (Disk AB 4) Berlin, 137–145. —. 1984b. Der Tempel von Segesta und die dorische Tempelbaukunst des griechischen Westens in klassischer Zeit, Mainz. Mertens, D. 2006. Städte und Bauten der Westgriechen. München. Meyer, H. 1995. A Roman masterpiece: The Minneapolis Doryphoros, in: Moon 1995, 65–115. Michaelis, A. 1876. Olympische Glossen, AZ 34, 162–174. Micheli, M. – Vidale, M. 1996. Riace, bronzi di. Restauro, in: EAA, secondo supplemento IV, Roma, 734. Mikalson, J.D. 1983. Athenian Popular Religion, Chapel Hill. Miller, S. G. 2003. The organization and functioning of the Olympic Games, in: D. J. Philips – D. Pritchard (eds.), Sport and Festival in the Ancient Greek World, Swansea, 1-40. Milne, J. G. 1931. The Coinage of the Eleians, NC 11, 171–180.

New Approaches to the Temple of Zeus at Olympia

265

Moesch, V. (ed.) 2009. La Villa dei Papiri. Verona. Moltesen, M. 2002. Catalogue Ny Carlsberg Glyptotek. Imperial Rome, II, Statues. Copenhagen. Moon, W. G. (ed.) 1995, Polykleitos, the Doryphoros and Tradition, Madison. Moreno, P. 1994. Scultura ellenistica, Roma. —. 1995. Lisippo: l’arte e la fortuna, Monza. Moustaka, A. 2013. Untersuchungen am Heraion-Altar. III.-V., in: XIII. Olympia-Bericht, 115–128. Möller, A. 2004. Elis, Olympia und das Jahr 580 v. Chr. Zur Frage der Eroberung der Pisatis, in: R. Rollinger & Chr. Ulf (eds.), Archaik. Interne Entwicklungen-Externe Impulse, Berlin, 249–270. Müller, E. 1922. Beiträge zu Pausanias (abstract of a PhD dissertation, Erlangen 1921), AA 1922, 352–354. Mylonopoulos, J. 2011a. Divine Images behind Bars. The Semantics of Barriers in Greek Temples, in: M. Haysom (ed.), Current Approaches to Religion in Ancient Greece. Papers presented at a symposium at the Swedish Institute at Athen, 17–19 April 2008, Stockholm: Svenska Institutet i Athen, 269–291. —. 2011b. Das griechische Heiligtum als räumlicher Kontext antiker Feste und Agone, ThesCRA VII, 43–77. Naiden, F. S. 2012. Blessed are the parasites, in: C.A. Faraone – F.S. Naiden (eds.), Greek and Roman Animal Sacrifice. Ancient Victims, Modern Observers, Cambridge, 55–83. Nesselrath, H. G. 1990. Lucians Introductions, in: D. A. Russell, Antonine Literature, Oxford, 111–140. Neubecker, A. J. 1977. Altgriechische Musik. Darmstadt. Neudecker, R. 2007a. Pasiteles, in: Brill’s New Pauly. Antiquity, 10, Leiden/Boston, 587–588. —. 2007b. Polyclitus, in: Brill’s New Pauly, Antiquity, 11, Leiden/Boston, 511–515. Nicolet-Pierre, H. 1975. Remarques sur la chronologie relative des plus anciennes séries de statères éléens, RN, 6ème Sér., tome XVII, 6–18. Nielsen, T. H. 2007. Olympia and the classical Hellenic city-state culture, Copenhagen. Niemeier, J.-P. 1985. Kopien und Nachahmungen im Hellenismus. Ein Beitrag zum Klassizismus des 2. und frühen 1. Jhs. v. Chr., Bonn. Nilsson, M. P. 1937. s.v. Oktaeteris, RE XVII. 2, 2387–2392. Oehler, J. 1912. s.v. Hellanodikai, RE VIII. 1, 155–157.

266

Bibliography

Ohnesorg, A. 1993. Inselionische Marmordächer. Denkmäler antiker Architektur Bd. 18,2 (= Architektur auf Naxos und Paros 2), Architektur auf Naxos und Paros. Berlin: Walter de Gruyter. —. 2011. Der naxische Lichtdom: Das Phänomen lichtdurchlässiger inselionischer Marmordächer, in: U. Wulf-Rheidt – P. I. Schneider, Licht-Konzepte in der vormodernen Architektur. Diskussionen zur Archäologischen Bauforschung, 10. Internationales Kolloquium at Berlin 2009. Regensburg: Schnell+Steiner, 92–100. Okamoto, Y. et. al. 2011. Image-based network render- ing of large meshes for cloud computing. International Journal of Computer Vision, 94(1):12–22. Orlandos, A. K. 1968. Les matériaux de construction et la technique architecturale des anciens grecs II, Paris. Osada, T. 2008. Ein Gott, der nicht richtet. Das Zeusbild am olympischen Ostgiebel und die Religionsanschauung der Griechen, in: G. Grabherr and B. Kainrath (eds.), Akten des 11. Österreichischen Archäologentages in Innsbruck 23.-25. März 2006. IKARUS. Innsbrucker Klassisch-Archäologische Universitätsschriften Bd. 3, Innsbruck, 215–224. Osmers, M. 2013. „Wir aber sind damals und jetzt immer die gleichen”. Vergangenheitsbezüge in der polisübergreifenden Kommunikation der klassischen Zeit (Historia Einzelschriften 226), Stuttgart. Pakkanen, J. 2013. Classical Greek Architectural Design. A Quantitative Approach, Papers and Monographs of the Finnish Institute at Athens 18, Helsinki. Palagia, O. 2003. An imperial portrait from Aulis, in: P. Nolke (ed.), Romanisation und Resistenz, Mainz, 537–547. —. 2010a. Pheidias epoesen: attribution as value judgement, in: F. Macfarlane – C. Morgan (eds.), Exploring Ancient Sculpture: Essays in honour of Geoffrey Waywell, BICS Supplement 104, 97–107. —. 2010b. Sculptures from the Peloponnese in the Roman Imperial Period, in: A. D. Rizakis – C. E. Lepenioti (eds.), Roman Peloponnese III, Athens, 431–445. —. 2010c. Philip’s Eurydice in the Philippeum at Olympia, in: E. Carney – D. Ogden (eds.), Philip II and Alexander the Great: Father and Son, Lives and Afterlives, Oxford, 33–41. —. 2013. Aspects of the diffusion of the Ptolemaic portraiture overseas, in: K. Buraselis – M. Stefanou - D. J. Thompson (eds.), The Ptolemies, the Sea and the Nile, Cambridge, 143–159. Papadopoulou, T. 2005. Heracles and Euripidean Tragedy, Cambridge.

New Approaches to the Temple of Zeus at Olympia

267

Parker, R. 2009. , in: L. Bodiou – V. Mehl – J. Oulhen – F. Prost – J. Wilgaux (ed.), Chemin faisant. Mythes, cultes et société en Grèce ancienne. Mélanges en l’honneur de Pierre Brulé, Rennes, 167– 171. Pasquier, A. – Martinez, J.-L. (eds.) 2007. Praxitèle, exh. cat. Louvre, 23 March – 18 June 2007, Paris. Patay-Horváth, A. 2004. Pausanias und der Ostgiebel des Zeustempels von Olympia. Acta Antiqua Acad Hung 44, 21–33. —. 2005. Die Frisur der weiblichen Protagonisten im Ostgiebel des Zeustempels von Olympia. In: Ganschow, Thomas, Steinhart, Matthias (Eds.), Otium. Festschrift für Volker Michael Strocka, Remshalden, 275–283. —. 2006. The armor of Pelops In: Carol C. Mattusch, A. A. Donohue, A. Brauer (eds.), Common Ground. Archaeology, Art, Science and Humanities. Proceedings of the XVIth International Congress of Classical Archaeology Boston, Oxford, 424–427. —. 2007. Zur Rekonstruktion und Interpretation des Ostgiebels des Zeustempels von Olympia, AM 122, 161–206. —. 2008. Metallanstückungen an griechischen Marmorskulpturen in archaischer und klassischer Zeit (Tübinger Archäologische Forschungen 4), Rahden/Westf. —. 2011a. Virtual 3D Reconstruction of the East Pediment of the Temple of Zeus at Olympia – A Preliminary Report, in: K. F. Ausserer, W. Börner, S. Uhlirz (eds.), Proceedings of the 14th International Congress “Cultural Heritage and New Technologies” 2009, Wien, 653–658. —. 2011b. The complete Virtual 3D Reconstruction of the East Pediment of the Temple of Zeus at Olympia, Paper presented at 4th ARC 3D conference, Trento, Italy, 3-5 March 2011. Accessed 30th May 2014. http://www.isprs.org/proceedings/XXXVIII/5-W16/pdf/patay.pdf. —. 2011c. The Virtual 3D Reconstruction of the East Pediment of the Temple of Zeus at Olympia, Budapest (CD ROM: ISBN 978-963-284196-0). —. 2012a. Die Bauherren des Zeustempels von Olympia, Hephaistos 29, 35–54. —. 2012b. Die Baukosten des Zeustempels von Olympia, Boreas 35, 1– 10. —. 2012c. The Virtual 3D Reconstruction of the East Pediment of the Temple of Zeus at Olympia - An old puzzle of classical archaeology in the light of recent technologies, Digital Applications in Archaeology and Cultural Heritage 1, 12–22,

268

Bibliography

http://www.sciencedirect.com/science/article/pii/S2212054813000039 —. 2012d. Reconstructions of the East Pediment of the Temple of Zeus at Olympia – A comparison of drawings, plaster casts and digital models, International Journal of Heritage in the Digital Era Vol. 1 No. 3, 331– 349. —. 2013a. Überlegungen zu Datierung und Funktion der frühen Elischen Münzprägung, JNG 63, 1–27. —. 2013b. Die Perserbeute von Plataia, die Anfänge der elischen Münzprägung und die finanziellen Grundlagen der „Grossbaustelle Olympia”, Klio 95, 61–83. —. 2013c. Nikedarstellungen auf frühen elischen Silbermünzen, NumAntClass 42, 11–28. Pauly, A. F. 1827. Lucian’s Werke. (Übersetzt von A. F. Pauly, Erstes Bändchen), Stuttgart. —. 1831. Lucian’s Werke. (Übersetzt von A. F. Pauly, Zwölftes Bändchen), Stuttgart. Perlman, P. 2000. City and Sanctuary in Ancient Greece. The Theorodokia in the Peloponnese, Göttingen. Pettenò, E. 2002. Alcune considerazioni a proposito delle Danaidi e delle cosiddette danzatrici di Ercolano, in: I. Colpo et al. (eds.) Iconografia 2001. Studi sull’immagine: atti del convegno. Padova 30 maggio – 1 giugno 2001, Roma, 309–317. Pfanner, M. 1989. Über das Herstellen von Portäts, JdI 104, 157–257. Pfeiffer, R. 1941. The Measurements of the Zeus at Olympia, in: JHS 61, 1941, 1–5. —. 1949. Callimachus, vol. 1, Fragmenta, ed. by Rudolf Pfeiffer. Oxford: Clarendon. Pilhofer, P. et. al. (eds.) 2005. Lukian: Der Tod des Peregrinus. Ein Scharlatan auf dem Scheiterhaufen. (Herausgegeben, übersetzt und mit Beiträgen versehen von P. Pilhofer, M. Baumbach, J. Gerlach und D. U. Hansen), Darmstadt. Platt, V. 2011. Facing the Gods. Epiphany and Representation in GraecoRoman Art, Literature and Religion, Cambridge. Pollitt, J. J. 1990. The Art of Ancient Greece: Sources and Documents, Cambridge. Poulsen, F. 1951. Catalogue of Ancient Sculpture, Ny Carlsberg Glyptothek. Copenhagen. Pozzi, E. et al. 1989. Collezione del Museo Archeologico Nazionale di Napoli, I.2, Roma. Psoma, S. 2008. Panegyris Coinages, AJN 20, 227–255.

New Approaches to the Temple of Zeus at Olympia

269

Quatremère de Quincy, A. Chr. 1815. Le Jupiter Olympien ou l’art de la sculpture antique Paris: Didot. Raaflaub, K. A. 1987. Herodotus, Political Thought, and the Meaning of History, Arethusa 20, 1, 2; 221–248. Raeck, W. 1981. Zum Barbarenbild in der Kunst Athens im 6. und 5. Jahrhundert v. Chr., Bonn. Raeder, J. 1983. Die Statuarische Ausstattung der Villa Hadriana bei Tivoli, Frankfurt/Berne. Raschke, W. (ed.) 1988. The Archaeology of the Olympics. The Olympics and Other Festivals in antiquity, Madison. Rehak, P. 1998. Unfinished Hair and the Installation of the Pedimental Sculptures of the Temple of Zeus at Olympia, in: K. J. Hartswick – M. C. Sturgeon (Hrsg.), Stephanos. Studies in Honor of Brunilde Sismondo Ridgway, Philadelphia, 193–308. Renfrew, J. M. 1988. Food for athletes and gods. A classical diet, in: Raschke 1988, 174–181. Rheingold, H. 1993. La Realtà Virtuale, Bologna. Rhomiopoulou, K. 1997. , Athens. Ridgway, B. S. 1974. A story of five Amazons, AJA 78, 1–17. —. 1984a. Roman Copies of Greek Sculpture: The Problem of the Originals, Ann Arbor. —. 1984b. The Riace Bronzes: A minority viewpoint, in: Due bronzi da Riace: Rinvenimento, restauro, analisi ed ipotesi di interpretazione, BdA, serie speciale, 3, Roma, 313–326. —. 1995. Paene ad exemplum: Polykleitos’ other works, in: Moon 1995, 177–199. Ridgway,B. S. 1970. The Severe Style in Greek Sculpture. Princeton. Riemann, H. 1935. Zum griechischen Peripteraltempel, Düren. Ritter, S. 2002. Bildkontakte. Götter und Heroen in der Bildsprache griechischer Münzen des 4. Jahrhunderts v. Chr., Berlin. Robert, C. 1919. Archäologische Hermeneutik, Berlin. Robertson, M. 1975. A History of Greek Art, Cambridge. —. 1992. The Art of Vase-Painting in Classical Athens, Cambridge. Rolley, C. 1999. Le sculpture grec, 2. La période classique, Paris. Romano, D. G. 1981. The Stadia of the Peloponnesos, Ann Arbor. Rosivach, V. J. 1994. The System of Public Sacrifice in Fourth-Century Athens, Atlanta. Roy, J. 1998. Thucydides 5.49.1 – 50.4: the quarrel between Elis and Sparta in 420 B.C., and Elis’ exploitation of Olympia, Klio 80, 360– 368.

270

Bibliography

—. 1999. Les cités d’Élide, dans J. Renard (ed.), Le Péloponnèse. Archéologie et histoire. Actes de la rencontre internationale de Lorient (12-15 mai 1998), Rennes, 151–176. —. forthcoming, Kings of Elis, kings of Olympia, in: Kingship in Ancient Greece: from Agamemnon to Cleopatra: International Conference in honor of Pierre Carlier, Athens, 6 – 8 February 2014. Ruesch, A. (ed.) 1908. Guida illustrata del Museo nazionale di Napoli, Napoli. Rusinkiewicz, S. and Levoy, M.. 2000. Qsplat: a multiresolution point rendering system for large meshes. In Proceedings of the 27th annual conference on Computer graphics and interactive techniques, SIGGRAPH ’00, 343–352. Sabbione, C. 1984. Appendice, in: Due Bronzi da Riace: Rinvenimento, restauro, analisi ed ipotesi di interpretazione, BdA, serie speciale, 3, Roma, 211–26. Säflund, M. L. 1970. The East Pediment of the Temple of Zeus at Olympia. A Reconstruction and Interpretation of its Composition, Göteborg. ahin, M.C. 1972. Die Entwicklung der griechischen Monumentalaltäre, Köln. Schefold, K. 1978. Götter und Heldensagen der Griechen in der spätarchaischen Kunst, München. Schefold, K. – Jung, F. 1988. Die Urkönige, Perseus, Bellerophon, Herakles und Theseus in der klassischen und hellenistischen Kunst, München. Schefold, K. – Jung, F. 1989. Die Sagen von den Argonauten, von Theben und Troia in der klassischen und hellenistischen Kunst, München. Schiering, W. 1991. Die Werkstaat des Phidias in Olympia. 2. Werkstaatfunde, (OlForsch. 18), Berlin. Schilbach, J. 2012. Neue Möglichkeiten und Probleme bei der Erforschung des Stadions von Olympia, in: Mythos Olympia 2012, 282–283. Schleif, H. 1934. Der Zeusaltar in Olympia, JdI 49, 139–156. —. 1943. Die neuen Ausgrabungen in Olympia und ihre bisherigen Ergebnisse für die antike Bauforschung. Berlin: [s. n.]. Schneider, L. – Höcker, C. 2001. Die Akropolis von Athen. Eine Kunstund Kulturgeschichte, Darmstadt. Schöll, A. 1855. Herodots Vorlesungen, Philologus 10, 410–431. Schrader, H. 1941. Das Zeusbild des Pheidias, in Olympia. JdI 56, 1941, 1–71. Schuchhardt, W.-H. 1971. Geschichte der griechischen Kunst, Stuttgart. Schwyzer, E. 1928. Zu griechischen Inschriften, RhM 77, 225–261.

New Approaches to the Temple of Zeus at Olympia

271

Scott, M. 2011. Displaying lists of what is (not) on display. The uses of inventories in Greek sanctuaries, in: Haysom – Wallensten 2011, 239– 252. Seltman, Ch. T. 1921. The Temple Coins of Olympia, Cambridge (originally published in Nomisma. Untersuchungen auf dem Gebiete der antiken Münzkunde 8, 1913, 23–65; 9, 1914, 1–33; 11, 1921, 1–39 and reprinted in New York 1975). —. 1948. Greek Sculpture and some Festival Coins, Hesperia 17, 71–85. Seltman, Ch. T. 1951. The Katoché Hoard of Elean Coins, NC, ser. VI, vol. XI, 40–55. Senff, R. 2012a. Olympia – Geschichte eines Heiligtums, Antike Welt 2012/4, 10–19. —. 2012b. Neue Methoden und aktuelle Forschungen, in: Mythos Olympia 2012, 198–201. Sengoku-Haga et al. 2010. Le Peplophoroi dalla Villa dei Papiri e le misurazioni tridimensionali, in: A. De Rosa (ed.), Vesuvio: Il Grand Tour dell’Accademia Ercolanese. Dal passato al futuro. Atti del Convegno Internazionale, Reggia di Portici, 21 e 22 maggio 2010, Napoli, 93–100. Sengoku-Haga et al. forthcoming , Kopienkritik of Classical sculpture via 3D shape comparison: The case of Polykleitos. Shoe, L. 1949. Dark Stone in Greek Architecture, in: American School of Classical Studies at Athens (ed.), Commemorative Studies in Honor of Theodore Leslie Shear, Hesperia Suppl. 8. Princeton N. J., 341–352, pl. 45–46. Siewert, P. 1991a. Die Inschriften der Landschaft Eleia ohne Olympia, in: A. D. Rizakis (ed.), Achaia und Elis in der Antike. Akten des 1. Internationalen Symposiums, Athen, 19.-21. Mai 1989. (Meletemata 13), Athens, 105–107. Siewert, P. 1991b. Staatliche Weihungen von Kesseln und anderen Bronzgeräten in Olympia, AM 106, 81–84. Simon, E. 1963. Polygnotan Painting and the Niobid Painter, AJA 67, 43– 62. —. 1968. Zu den Giebeln des Zeustempels von Olympia, AM 83, 147–167. —. 1975. Versuch einer Deutung der Südmetopen des Parthenon, JdI 90, 100–120. Sinn, U. 1994. Apollon und die Kentauromachie im Westgiebel des Zeustempels in Olympia. Die Wettkampfstätte als Forum der griechischen Diplomatie nach den Perserkriegen, AA 1994, 585–602. —. 1996. Olympia. Kult, Sport und Fest in der Antike, München. —. 2004. Das antike Olympia. Götter, Spiel und Kunst, München.

272

Bibliography

Smith, R.R.R. 2013. Aphrodisias VI, The Marble Reliefs from the JulioClaudian Sebasteion, Darmstadt/Mainz Sonntagbauer, W. 1998. Zum Grundriß des Parthenon, ÖJh 67, 1998, 133– 169. —. 2003a. Sonntagbauer, Singt der Tempel wirklich? Zur musikalischen Proportionstruktur griechischer Tempel, in: B. Asamer - W. Wohlmayr (Hrsg.), Akten des 9. Österreichischen Archäologentages am Institut für klassische Archäologie der Paris-Lodron-Universität Salzburg, 6.8. Dez. 2001, Wien, 189–194. —. 2003b. Einheitsjoch und Stylobatmaß. Zu den Grundrissen des Zeustempels in Olympia und des Parthenon, BABesch 78, 35–42. —. 2004. Die olympische Festperiode oder warum das olympische Stadion 192,28m misst, in: F. Huber/ R.C.A. Rottländer (Hrsg.), Ordo et Mensura VIII, St. Katharinen, 75–91. —. 2009. Zur Genese des klassischen Tempelentwurfes, in: R. Einicke u. a. (Hrsg.), Zurück zum Gegenstand. Festschrift für A. Furtwängler 1, Langenweissbach, 37–49. —. 2013. Gebauter Rhythmus. Vom rhythmischen Denken und Bauen bei den frühen Griechen, in: W. Sonntagbauer - J. Klopf (Hrsg.), WeltSeele-Gottheit. Festschrift für Wolfgang Speyer zum 80. Geburtstag, Horn, 136–171. Spawforth, A. J. S. 2012. Greece and the Augustan Cultural Revolution, Cambridge. Spivey, N. 1994. Understanding Greek Sculpture. Ancient Meanings, Modern Readings, London. Stansbury-O’Donnell, M. 1999. Pictorial Narrative in Ancient Greek Art, Cambridge. Stewart, A. F. 1983. Pindaric Dike and the Temple of Zeus at Olympia, ClassAnt 2, 133–144. —. 1990. Greek Sculpture. An Exploration, New Haven/London. Studniczka, F. 1923. Die Ostgiebelgruppe vom Zeustempel in Olympia, Abhandlungen der Sächsischen Akademie der Wissenschaft, Phil.-Hist. Klasse 37, 3–36. Taita, J. 2007. Olimpia e il suo vicinato in epoca arcaica (Il Filarete, 244), Milano. —. 2014. Quando Zeus deve far quadrare il bilancio. Osservazioni sul tesoro del santuario di Olimpia, in: K. Harter-Uibopuu – T. Kruse (Hg.), Sport und Recht in der Antike. Beiträge zum 2. Wiener Kolloquium zur Antiken Rechtsgeschichte, 27.-28.10.2011, Wien, 107– 145.

New Approaches to the Temple of Zeus at Olympia

273

Taraporewalla, R. 2011. Size Matters. The Statue of Zeus at Olympia and Competitive Emulation, in: J. McWilliam, R. Taraporewalla, T. Stevenson and S. Puttock (eds.), The Statue of Zeus at Olympia. New Approaches. Cambridge: Cambridge Scholars Publishing, 33–49. Tersini, N. D. 1987. Unifying Themes in the Sculpture of the Temple of Zeus at Olympia, ClassAnt 6, 139–159. Tölle-Kastenbein, R. 1986. Frühklassische Peplosfiguren. Typen und Repliken. (Antike Plastik XX), Berlin. Trantalidou, K. 2013. Dans l’ombre du rite: vestiges d’animaux et pratiques sacrificielles en Grèce antique. Note sur la diversité des contextes et les difficultés de recherche rencontrées, in: G. Ekroth – J. Wallensten (eds.), Bones, Behaviour and Belief. The Zooarchaeological Evidence as a Source for Ritual Practice in Ancient Greece and Beyond, Stockholm, 61–86. Trendelenburg, A. 1902. Der grosse Altar des Zeus in Olympia, Berlin. Treu, G. 1892. Löwenköpfe von der Traufrinne des Zeustempels, in: E. Curtius – F. Adler (eds.), Olympia. Die Ergebnisse der von dem deutschen Reich veranstalteten Ausgrabung, vol. 2, Die Baudenkmäler. Berlin, 22–27. Trianti, I. 2002. Neue technische Beobachtungen an den Skulpturen des Zeustempels von Olympia, in. Olympia 1875 – 2000, 281–300. Trimble, J. 2011. Women and Visual Replication in Roman Imperial Art and Culture, Cambridge. True, M. et al. 2004. Greek Processions, ThesCRA I, 1–20. v. Gerkan, A. 1940. Der Tempel von Didyma und sein antikes Baumaß, ÖJh 32, 127–150. v. Steuben, H. 1973. Der Kanon des Polyklet. Doryphoros und Amazone, Tübingen. —. 1993. Die Amazone des Polyklet, in: H. Beck – P. C. Bol (eds.), Polykletforschungen, Berlin, 73–102. —. 1994. Noch Einmal die Amazone des Polyklet. in: Griechische Klassik. Vorträge bei der interdisciplinaeren Tagung des Deutschen Archaeologenverbandes, 24.10.1991, Nürnberg, 275–282. Vallois, R. 1923. Le portique de Philippe (Delos 7, 1), Paris. Versnel, H. S. 1987. What Did Ancient Man See When He Saw a God? Some Reflections on Greco-Roman Epiphany, in: D. van der Plas (ed.), Effigies Dei. Essays on the History of Religions, Leiden, 42–55. Wacker, C. 1996. Das Gymnasium in Olympia. Geschichte und Funktion, Würzburg. —. 2012. Das Programm der antiken Olympischen Spiele, in: Mythos Olympia 2012, 269–273.

274

Bibliography

Walker, A. 2004. The Coinage of the Eleans for Olympia, in: Coins of Olympia. The BCD Collection. Auction Leu 90, 10 May 2004. Walter-Karydi, E. 1987. Die Äginitische Bildhauerschule, Mainz. Weber, M. 1976. Die Amazonen von Ephesos, JdI 91, 28–96. —. 1989. Die Amazonenstatuen in der Villa Hadriana in Tivoli, in: Festschrift für Jale Inan, Istanbul, 31–37. Weber, U. 2011. Der Plan des Didymaion - Buchstaben auf der Euthynterie enthüllen ihn und belegen die Verwendung des attischen Fußes, in: O. Pilz - M. Vonderstein (Hrsg.), Keraunia. Beiträge zu Mythos, Kult und Heiligtum in der Antike, Berlin, 33-46. —. 2013. Versatzmarken im antiken griechischen Bauwesen (Philippika 58), Wiesbaden. Weiler, I. 1997. Olympia – Jenseits der Agonistik. Kultur und Spektakel, Nikephoros 10, 191–213. Weniger, L. 1904. Das Hochfest des Zeus in Olympia. I. Die Ordnung der Agone, Klio 4, 125–151. —. 1905. Das Hochfest des Zeus in Olympia. III. Der Gottesfriede, Klio 5, 184-218. —. 1913. Der Hochaltar des Zeus in Olympia, Neue Jahrbücher für das klassische Altertum 16, 1913, 241–260. —. 1920/21. Olympische Studien, AfRW 20, 1920/21, 41–78. Wesenberg, B. 1983. Beiträge zur Rekonstruktion griechischer Architektur nach literarischen Quellen, 9. Beih. AM 1983, 32–68. —. 1984. Der Fuß des Kallikrates, AA 1984, 547–554. —. 1995. Die Metrologie der griechischen Architektur. Probleme interdisziplinärer Forschung, in: D. Ahrens – R. C. A. Rottländer (Hrsg.), Ordo et Mensura III, St. Katharinen, 199–222. Westervelt, H. 2009. Herakles at Olympia. The Sculptural Program of the Temple of Zeus, in: P. Schultz - R. v. den Hoff (eds.), Structure, Image, Ornament, Oxford, 133–152. Willemsen, F. 1959. Die Löwenkopf-Wasserspeier vom Dach des Zeustempels (OlForsch 4), Berlin. Williams, D. 2013. The East Pediment of the Parthenon from Perikles to Nero (BICS Supplement 118), London. Wilson Jones, M. 2000. Doric Measure and Architectural Design 1: The Evidence of the Relief from Salamis, AJA 104, 73–93. Wojan, F. 2011. Les Éléens (IVe s. a. C.-IIIe s. p. C.). Recherches de numismatique et d’histoire, thèse de doctorat inédite, université François Rabelais, Tours.

New Approaches to the Temple of Zeus at Olympia

275

Wölfel, C. 1990. Erwägungen zur künstlichen Beleuchtung von Skulptur, in: W.-D. Heilmeyer – W. Hoepfner, Licht und Architektur. Tübingen, 43–52. Yavis, C. G. 1949. Greek Altars. Origins and Typology, Including the Minoan-Mycenaean Offertory Apparatus. An Archaeological Study in the History of Religion, St. Louis. Younger, J. G. – Rehak, P. 2009. Technical Observations on the Sculptures from the Temple of Zeus at Olympia, Hesperia 78, 41–105. Zanker, P. 1974. Klassizistische Statuen. Studien zur Veränderung des Kunstgeschmacks in der römischen Kaiserzeit, Mainz. Zhang, Y. et al. 2013. Classical sculpture analysis via shape comparison, in: 2013 International Conference on Culture and Computing (Kyoto 2013) 57–61. Ziehen, L. 1939. s.v. Olympia, no. 1, RE XVIII. 1, 1–71. Zinserling, G. 1957. Kultbild, Innenraum, Fassade. Eine Untersuchung zum mutterländisch-dorischen Tempel des 5. Jarhunderts, Das Altertum 3, 18–34. Zoumbaki, S. 2001. Elis und Olympia in der Kaiserzeit, Athens.

CONTRIBUTORS

Francesco Gabellone studied architecture and has founded the IBAMCNR laboratory at Lecce. He is professor for digital restoration at the Academy of Fine Arts (Lecce). His research activities focus on valorization and enjoyment of cultural heritage by means of virtual 3D technologies. He conducts several projects for museum communication and virtual archaeology (Iraq Virtual Museum, Museo Egizio di Torino, National Archaeological Museum of Taranto, Museum of Metaponto and Herakleia, Syracuse, Terracina, etc.). Arnd Hennemeyer has studied architecture at the Technische Universität München (TUM) and received his PhD in 2006 for a dissertation on the Sanctuary of Athena at Priene (published in the series Archäologische Forschungen). After many years of extensive field research on the temple of Zeus at Olympia and in Priene he is now research associate at ETH Zurich (Institut für Denkmalpflege und Bauforschung, Prof. Uta Hassler) and is working on a topic concerning attitudes to ancient architectural polychromy during the 19th century. He has published many papers concerning the architecture of the temple of Zeus and is currently preparing a research monogpraph on this monument. Tonio Hölscher was professor of classical archaeology at the University of Heidelberg from 1975 until 2009, and is still a very active emeritus, e.g. as „Chair du Louvre“ (Paris 2015). His publications include e.g. Rituelle Räume und politische Denkmäler im Heiligtum von Olympia., in: H. Kyrieleis (ed.), Olympia 1875 - 2000. 125 Jahre deutsche Ausgrabungen, Mainz, 2002, 331-345 and Klassische Archäologie. Grundwissen (Darmstadt 2006) or Griechische Kunst (München 2007) already translated to various other languages. Katsushi Ikeuchi is a Professor at the University of Tokyo. He received a Ph.D. degree in Information Engineering from the University of Tokyo in 1978. After working e.g. at the Massachusetts Institute of Technology and the Carnegie Mellon University, he joined the university in 1996. His research interest spans computer vision, robotics, and computer graphics. He has received several awards, including the IEEE RAS “most active distinguished lecturer” award and the Shiju Houshou (the Medal of Honor

New Approaches to the Temple of Zeus at Olympia

277

with Purple ribbon) from the Emperor of Japan. Publications: K. Ikeuchi and D. Miyazaki, Digitally Archiving Cultural Objects, Springer 2008; K. Ikeuchi (ed.), Computer Vision. A Reference Guide, Springer 2014. Jessica Susanne Krause, M. A. is currently working on her PhD thesis (entitled: Theseus as a Symbol of Athens? The Attic *Demos Heros* in Public Depictions Outside Attica) at the Graduate School “Human Development in Landscapes” at the Christian-Albrechts-Universität, Kiel. She co-organized a session entitled “Different perceptions of Landscapes throughout Classical Antiquity and beyond" at the Fourth International Open Workshop “Socio-Environmental Dynamics over the Last 12,000 Years: The Creation of Landscapes” in 2015. Yasuhide Okamoto is a Project Research Associate of Industrial Science, the University of Tokyo. He received the Ph.D. degree in Information Science and Technology from the University of Tokyo, Japan, in 2010. He is focusing on visualization technologies for huge 3D data, and 3D applications in the context of virtual and mixed reality. Toshihiro Osada studied 1989-1993 at the University of Salzburg. He is currently Professor in the Faculty of Art and Design, at the University of Tsukuba, Japan. His most important publications include e.g. “Ein Gott, der nicht richtet. Das Zeusbild am olympischen Ostgiebel und die Religionsanschauung der Griechen,” in: G. Grabherr et al. (eds.), Akten des 11. Österreichischen Archäologentages in Innsbruck 23.-25. März 2006. IKARUS (Innsbruck: Innsbruck University Press 2008) 215-224; “Also Ten Tribal Units – The Grouping of Cavalry on the Parthenon North Frieze,” AJA 115, 2011, 537-548. Olga Palagia is Professor of Classical Archaeology at the National and Kapodistrian University of Athens. She is a specialist in Greek sculpture and the art of Macedonia and has edited numerous collective volumes and conference proceedings. She is the author of The Pediments of the Parthenon (Leiden 1993). Her most recent edited volumes are Greek Sculpture: Function, Materials and Techniques in the Archaic and Classical Periods (Cambridge 2006) and Art in Athens during the Peloponnesian War (Cambridge 2009). András Patay-Horváth studied classical archaeology, ancient history and classical philology at the University Eötvös Loránd, Budapest, where he is currently employed as an assistant lecturer in the Institute for Ancient

278

Contributors

history. His research interest is focussed on the temples of Olympia and the Peloponnese. His recent publications include “Die Perserbeute von Plataia, die Anfänge der elischen Münzprägung und die finanziellen Grundlagen der Grossbaustelle Olympia”, Klio 95, 2013, 61-83 and “Hera in Olympia: Tempel, Kult und Münzprägung”, Thetis 20, 2013, 81-99. James Roy, after studies in Edinburgh and Cambridge, taught ancient history in The Universities of Sheffield and Nottingham. Since retirement he has been an Honorary Research Associate of the University of Nottinhgham. His research has focussed especially on the regions Arkadia and Elis (including Olympia), and recent publications include “On seeming backward: how the Arkadians did it.” in S. D. Lambert (ed.) Sociable man: essays on ancient Greek social behaviourin honour of Nick Fisher (Swansea, 2011); ‘Olympia, identity and integration: Elis, Eleia, and Hellas.’ in P. Funke and M. Haake (eds.) Greek Federal States and their Sanctuaries(Stuttgart, 2013); and ‘Autochthony in Ancient Greece’ in J. McInerney (ed.) A Companion to Ethnicity in the Ancient Mediterranean (Malden MA and Oxford, 2014) Kyoko Sengoku-Haga is an associate professor of Art History at Tohoku University, Japan. She obtained a specialization diploma in Greek and Roman Archaeology from the Italian Archaeological School at Athens (1998) and a Ph.D. degree in Art History from the University of Tokyo (2002). Her research focuses on both religious and technical aspects of Classical sculpture, and since 2007 she has been conducting a joint research project with the Computer Vision laboratory of prof. Katsushi Ikeuchi. Publications: K. Sengoku-Haga, Ancient Sculpture of Rhodes, Tokyo 2006 (in Japanese); K. Sengoku-Haga, M. Aoyagi, “Due statue marmoree da Somma Vesuviana: il Dioniso e la Peplophoros”, Amoenitas 1, 2010, 237-252. Wolfgang Sonntagbauer was born 1949. He studied Classical Philology and Sports at the University of Salzburg from 1968-1973. Afterwards he was teacher for 37 years at the Bundesrealgymnasium Salzburg. 1995 doctoral dissertation in Classical Archaeology about the Canon of Polyklet. His most recent publications include e.g. Das Eigentliche ist unaussprechbar. Der Kanon des Polyklet als "mathematische" Form (Frankfurt/Main 1995) --- Von der Hochzeit der Gegensätze. Zur frühgriechischen Seelenlehre in der 'Hochzeitszahl' der platonischen Politeia (Hildesheim 2012).

New Approaches to the Temple of Zeus at Olympia

279

Julia Taita achieved her PhD in Ancient History at the Universität Wien and held a four-year position as a Research Fellow in Ancient Greek History at the Università degli Studi di Milano. Now she is an independent scholar. Her publications include the monograph Olimpia e il suo vicinato in epoca arcaica (Milano 2007) and many papers concerning the sanctuary of Olympia, the administration of its cults and its relationships with the communities of the Elean region. She is also working on the edition of the unpublished inscriptions on bronze tablets from Olympia together with Peter Siewert (Vienna). Ulf Weber has studied classical archaeology, Greek philology and ancient history at the Friedrich-Schiller-University, Jena. Since 2004, he regularly took part in the excavations of the German Archaeological Institute at Didyma and in 2011 he completed his PhD dissertation entitled „Versatzmarken im antiken griechischen Bauwesen“ (published as volume no. 58 of the series “Philippika” in 2013). Since 2013 he works for the project „Kulte im Kult“ at the University Halle-Wittenberg. His publications include e.g. Der Plan des Didymaion – Buchstaben auf der Euthynterie enthüllen ihn und bestätigen die Verwendung des attischen Fußes, in: O. Pilz – M. Vonderstein (Hrsg.), Keraunia. Beiträge zu Mythos, Kult und Heiligtum in der Antike, Festschrift Uta Kron (Berlin 2011) 33–46. Franck Wojan, member of the CeTHiS, université François-Rabelais, Tours (France) and secretary of the Société française de Numismatique. He received his PhD in 2011 for a dissertation on the coinage and the history of the Eleans. His most recent articles include: 1) “The Civic Bronze Coinage of the Eleans: some Preliminary Remarks”, N. Holmes (ed.), Proceedings of the XIVth International Numismatic Congress – Glasgow 2009, Glasgow, 2011, p. 497-499 ; 2) with Michel Amandry, “Le monnayage d’Hadrien à Élis-Olympie”, Revue numismatique, 170 (2013), p. 279-328. He is now preparing the corpus of the Elean coinage.

INDEX Achilles, 5, 6, 7, 12, 13, 101, 102, 103, 188 Aeneas, 103 Agamemnon, 5, 6, 7, 12, 102 Agenor, 200 Agrippa, 77 Agrippina, 79, 81, 82 Aiakids, 105 Ajax, 101, 103 Alesion, 126 Alpheios, 114, 120, 125, 126, 146, 153 Anaitoi, 144 Andania, 127 Antipater, 200 Antonia Cleodice, 80 Aphrodite, 99 Apollo, 120 Apollonius of Tyana, 173 Argos, 7 Aristion, 200 Aristophanes, 123 Aristotle, 146 Arkadia, 143 Arsinoe II, 78 Artemis, 208 Artemis Ephesia, 119, 121 Artemis Leukophryene, 122 Atalante, 65 Athena, 41, 92, 103, 105 Athena Chalkioikos, 97 Athena Ergane, 104 Athena Ilias, 120 Athena Parthenos, 25, 28, 212 Athenaeus, 124 Athens, 104, 179 Atlas, 41 Augustus, 79 Aulos Sextus Eraton, 80 Autonoos, 104 Baiae, 201, 203

Briseis, 6 Caria, 175 Centauromachy, 102 Centaurs, 181 Chaladrioi, 143 Claudia Alcinoa, 80 Claudia Octavia, 79 Claudius, 79 Crete, 125, 135 Deianeira, 177 Delos, 10, 120, 121, 201 Delphoi, 9, 10, 97, 121, 122, 125, 136, 153, 189 Demeter, 4 Deukalion, 144 Didymaion, 57, 58, 60 Dio Chrysostom, 172 Diodorus Siculus, 171 Diogenes Laërtios, 172, 174 Diomedes, 99, 103, 212 Dionysios, 79 Dionysius of Syracusae, 171 Echetlaeos, 105 Elis, 4, 6, 7, 8, 9, 11, 30, 46, 81, 115, 118, 119, 121, 122, 123, 125, 126, 127, 133, 136, 137, 139, 140, 141, 142, 143, 144, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 171, 181 Eos, 101 Ephesos, 208 Epictetus, 112 Epidamnos, 135 Eros, 103 Euripides, 115, 181 Eurydice, 81 Ewa, 141, 143, 144, 146 Flavia Domitilla, 79 Gelon, 78 Glaukias, 78

New Approaches to the Temple of Zeus at Olympia Gorgias, 173 Hadrian, 173 Hegias, 79 Helen, 103 Hellespont, 179 Hera, 11, 25, 101, 130, 132, 155, 156, 157, 174, 200, 211 Heracles, 41, 65, 91, 98, 105, 176, 177, 181 Herakles, 92 Herculaneum, 202, 204, 209 Hermes, 78, 80, 101 Herodes Atticus, 75, 81 Herodotus, 8, 10, 12, 68, 140, 165, 166, 167, 168, 169, 171, 172, 174, 175, 176, 177, 178, 179, 180, 182 Himera, 10 Hippodameia, 3, 6, 65, 90, 91, 93 Hippomenes, 65 Homeros, 6 Iliad, 99 Iliupersis, 103 Iphitos, 92 Isocrates, 123, 171 Isthmia, 60 Kalchas, 6 Kallikrates, 78 Kassandra, 103 Kladeos, 90 Kresilas, 208, 212 Kroisos, 178 Kyniskos, 200 Lapith, 75, 76, 178 Leonidas, 8 Lepreon, 123, 140, 143, 146, 147 Libya, 135 Livia, 79 Lucian, 164, 212 Lucius Vetlenus Florus, 80 Lysippos, 79 Magnesia, 122, 123 Marathon, 104 Marcellinus, 170 Mazi, 145 Memnon, 101, 103

281

Menelaos, 103 Metapioi, 144 Metapontum, 124, 228 Mikythos, 91, 92, 93, 96, 97 Naukydes, 200 Nemea, 122 Nero, 79 Nike, 9, 12, 92 Numisia Teisis, 80 Odysseus, 99, 101 Oinomaos, 2, 5, 7, 90, 91, 93, 102, 188 Olympias, 81 Onatas, 7 Oxylos, 7 Paris, 102 Parthenon, 5, 9, 25, 28, 29, 57, 58, 59, 60, 61, 66, 67, 68, 69, 103, 109, 174, 189, 269, 272, 274 Pasiteles, 201 Patrias, 142 Pausanias, 2, 4, 6, 7, 8, 9, 10, 12, 19, 25, 28, 29, 30, 41, 77, 79, 81, 90, 91, 92, 93, 94, 95, 96, 97, 105, 113, 115, 116, 117, 118, 121, 126, 130, 132, 134, 154, 159, 160, 167, 176, 178, 188, 189, 199, 200 Peirithoos, 178 Pelops, 2, 3, 5, 6, 7, 12, 13, 90, 91, 92, 93, 102, 188, 190 Penelope, 6 Peregrinus, 168, 174, 175, 181, 182 Pergamon, 122 Pericles, 212 Pheidias, 16, 17, 18, 22, 25, 28, 29, 31, 38, 43, 112, 160, 186, 199, 208, 211, 212, 252, 256 Philathenaios, 79 Philip II, 78, 81 Philostratus, 173 Phormis, 92 Phradmon, 208 Phylakos, 104 Pindar, 113 Pisa, 8

282 Plataia, 10, 12, 13 Pliny, 200, 205, 208 Plutarchus, 124, 125, 179, 208 Polemon, 124 Polybius, 115 Polykleitos, 199, 201, 202, 203, 207, 211 Pompeii, 204, 205, 207, 225, 227 Poppaea Sabina, 80 Poseidonia, 146 Praisos, 125 Praxiteles, 78, 80 Ps.-Andocides, 124 Ptolemy II, 78 Pulydamas, 79 Pythokles, 200 Rheneia, 125 Serdaioi, 146 Simonides, 13 Skillous, 7, 119, 125, 143 Sparta, 13, 97, 123, 140, 147, 148 Sterope, 3, 4, 90, 91, 93 Stoa Poikile, 105 Strabo, 115, 126

Index Sybaris, 146 Syracuse, 115, 228, 229 Taras, 229 Tegea, 7 Teisamenos, 10 Telemachos, 6 Themistius, 172 Theseus, 105, 179 Thetis, 6, 101 Thucydides, 116, 123, 140, 146, 157, 169, 170, 178 Thyreatis, 143 Tiberius Claudius Aphrodisius, 80 Timaeus, 115 Timokrates, 142 Titus, 79 Triphylia, 10, 145 Troy, 6 Varro, 202 Vespasian, 79 Vitruvius, 58 Xenokles, 200 Xenophon, 121, 140, 173, 176 Xerxes, 10, 179