Report 7: The Temple of Artemis at Sardis (2020)

by Nicholas Cahill

Editor’s Preface

Editor’s Preface

Nestled in its tranquil valley between the Acropolis and Necropolis hills alongside the gurgling Pactolus stream, the Temple of Artemis is without doubt the most beautiful monument of Sardis, and one of the most delightful ancient buildings in all of Turkey. The temple has been a major focus of archaeological work at Sardis since the earliest researchers came to the site. Clearing this enormous structure, buried under more than twelve meters of alluvium, was a heroic undertaking of Howard Crosby Butler. He accomplished the task in four long excavation seasons, with workforces of up to 250 men. Naturally, such a rapid pace of excavation could not be carried out with the precision or care of modern archaeology; and Butler’s practice of removing late features he considered of lesser interest, such as the mortared floors and walls of the Roman phase of the building, have made it difficult to answer a number of important questions.

The outbreak of World War I brought an unexpected end to Butler’s campaign, and although a brief season in 1922 allowed the team to fill gaps in their records, the excavation remained unfinished. Nonetheless, Butler’s two-volume report is in many ways an exemplary publication.1 He presents the history of exploration, careful descriptions of the surviving elements of the temple, analysis of the architectural elements and styles, and attempts at reconstructions of the building’s problematic superstructure, particularly the east and west ends. Gorgeous illustrations not achievable even with today’s technologies were created under difficult circumstances.2 Butler presents a historical narrative which, although incorrect in many respects, took the best account he could of the literary, epigraphic, numismatic, and other evidence available to him. He made every effort to fit the building’s history into the general history of the site from the Lydian through the Byzantine periods; but in doing so, he brought his own expectations to the building’s complex and unpredictable history. He assumed that, as at Ephesus and Samos, there must have been an Archaic temple of the time of Croesus beneath the later building, and therefore interpreted sandstone foundations of the Hellenistic building as having belonged to such an Archaic temple. He also believed that the marble foundations were laid after the destruction of the city at the beginning of the Ionian Revolt in 499 BC, during the Classical period. In his zeal to find remains from these prestigious periods of Greek architecture, Butler dismissed unmistakable signs of Roman construction—such as the mortared rubble of the peristyle—as mere later reinforcements.

When Prof. Hanfmann restarted research at Sardis in 1958, he continued to investigate the temple and sanctuary, in collaboration with architect Ken Frazer and others. They documented buildings and monuments of the sanctuary excavated but left unpublished by Butler as well as the two-phase “Lydian Altar,” which had mystified earlier excavators; and they conserved structures such as the central image base (“basis”), the Lydian Altar, and the northwest stairs. Frazer also excavated a series of test trenches in the temple in 1972 to resolve specific troublesome issues, such as the vexing absence of columns in antis. Unfortunately, Frazer was not familiar with stratigraphic excavation and, like Butler, cut through multiple strata and even the clayey bedrock in his search for stone architecture. In retrospect, we can see that they had much of the information needed to correctly outline the history of the columns in antis, among other problems, but did not pay sufficient attention to the archaeological evidence.

Hanfmann’s results were presented in the first Report of the Harvard-Cornell Expedition.3 Like Butler, he projected the temple’s chronology onto his own narrative of the history of the city. His conviction, for instance, that the destruction of Sardis by Antiochus III represented the major turning point in the city’s history led him to interpret a colossal head from the sanctuary as a Hellenistic portrait of Achaeus as Zeus, and to propose that the building was already divided into two back-to-back cellas in the Hellenistic period. In doing so, he disregarded the mundane but ultimately more secure evidence of clamps, dowels, lifting devices, and other architectural features of the temple itself, which had already led Gottfried Gruben to date the division, correctly, to the Roman period. The head is now more convincingly identified as a Roman portrait of Marcus Aurelius (as members of Butler’s team had already suggested).4

A major contribution to the study of the temple of Artemis was made by architectural historian Gottfried Gruben, who visited during the off-season in 1961. Although his research was conducted without the customary formalities and research permits, Gruben’s article on the temple, published that same year (!), transformed our understanding of this complex building.5 By carefully analyzing the forms and distribution of different types of clamps, dowels, lifting devices, chisels, and other technical aspects of the masonry, Gruben distinguished two sets of building techniques and concluded that they belonged to two major building phases, Hellenistic and Roman. Gruben’s focus on the details of masonry technologies, rather than the stylistic analysis, circumstantial evidence, and reliance on historical probability that had been fundamental to Butler’s and Hanfmann’s proposals, and his reliance on what he saw, rather than what he expected or wished to see, allowed him to place the study of the temple on a rigorous, scientific basis. Although he postulated a Hellenistic phase that did not in fact exist, Gruben’s observations have, for the most part, stood the test of time.

In the early 1980s Thomas Howe, then a graduate student at Harvard and the Expedition architect, refined Gruben’s phasing through further close examination of the techniques of construction. Moreover, while Gruben was not able to take precise measurements without a permit or the official support of the Expedition, Howe’s careful survey allowed him to document, for the first time, the building’s subtle curvature. Howe’s article on the curvature of the building contains many other observations.6

Gruben’s and Howe’s focus on architectural technologies of the temple, such as clamps, dowels, and lewises, solved many of the questions of the building’s history, and highlighted the fundamental advantages of such technical studies over stylistic analysis. The temple’s capitals, for instance, are very diverse stylistically, and Butler and others even assigned different dates to the two adjacent capitals on the standing columns; but technical aspects such as lewis holes and dowels clearly date both to the Roman period, and the five fallen, equally diverse capitals to the Hellenistic era. Further work, however, was hampered by the lack of adequate documentation. Butler’s 1:200 plan and other drawings remain essential records and, along with his extensive photographic documentation, are of extraordinary value to our understanding of the incompletely documented excavation; but these were not at a scale sufficient to distinguish critical technical details.

In 1987, therefore, at the invitation of Prof. Crawford H. Greenewalt, jr., director of the Sardis Expedition, Prof. Fikret Yegül initiated a new phase of the study of the temple of Artemis. An experienced architectural historian, Yegül has been a member of the expedition since 1963, and published one of the other great buildings of Sardis, the Bath-Gymnasium Complex.7 The goal of the project was to document the marble architecture of the temple at a scale (1:20) that would capture technical details of the masonry which were so important in unraveling the building’s tangled history. The scope of the project included representing the in situ marble masonry of the building in plan, sections, and elevation, with details of construction and ornament.8 A full study of all the fallen blocks from the superstructure was not envisioned or carried out. The drawings were to be accompanied by an explanatory text of 80–100 pages, and no excavation was planned.

The author’s elegant, precise, and detailed state drawings provide a record that must be among the most comprehensive graphic documentation of any ancient building. Recording the textures of different levels of finish of this incomplete structure, cuttings for clamps, dowels, lewises, and other technical features, even the cracks and flaws in the marble, the drawings are of inestimable value in furthering research on the temple and its controversial history, and as critical permanent records for the future. This effort occupied almost fifteen years, and Yegül’s devotion to the building is nowhere more evident than in the exquisite care he put into rendering every clean line of its marble blocks.

As the plan and elevation drawings moved toward completion in the early 2000s, it became clear that the Lydian Altar could not reasonably be omitted from the project. In 2005–2006, therefore, archaeologists cleaned and Felipe Rojas drew the Altar in plan and elevation at the same scale as the temple. The new excavations of this building, first uncovered in 1910 and re-excavated in 1970, produced significant new insights into its complex and confusing history.

In 2006–2008 Yegül drew a new series of reconstruction drawings of the building, including figs. 4.13–4.14, 4.17–4.18, and plates v–vii. These drawings forced the expedition to confront issues which were not yet sufficiently resolved, such as the history of the west end and the northwest stairs, the mysterious gap where one of the columns of the west porch should be, the complete absence within the cella of fragments of the interior colonnade, and the problematic lack of columns in antis, which had forced Butler to reluctantly conclude that the porches were hypaethral. These questions were the catalyst for two seasons of excavation, in 2010 and 2011.9 These excavations confirmed some conclusions already proposed in unpublished reports through analysis of previous research, such as the two-phase history of the northwest stairs and north peristyle, and that the east columns in antis had stood through most of the history of the building, and were only removed in the fourth or fifth century AD. Other results, however, were completely unexpected, such as a rich deposit of mid-first-century pottery from the foundations of a column of the east porch, leading archaeologists to conclude that this column foundation was laid in the Julio-Claudian period rather than in the second century, as Gruben and others had proposed.

An extensively published and award-winning architectural historian, Yegül brings years of experience and the most recent scholarship to his study of the temple. There are a few questions, such as the date of deposition of this first-century ceramic assemblage and hence of the east porch foundations, and the date of removal of the columns in antis, where his conclusions derived by the methods of architectural history differ from those of archaeology. In these instances, it is certainly salutary for each of us to reconsider our assumptions and conclusions, even those that seem very secure, and to always return to the basic facts and try to account for them all. In some instances and after long discussion, therefore, Prof. Yegül and I have simply agreed to disagree, and his conclusions are here presented.

The renewed attention to the Lydian Altar made us keenly aware of the difficulties of preserving this fragile building, and a conservation program in 2010–2012, supported by the J.M. Kaplan Fund and overseen by Hiroko Kariya and Catherine Williams, returned the altar to the condition in which it was found in 1910, and added protective cappings to the stair foundations and walls. Shortly thereafter, conservator Michael Morris proposed and developed a new technique for killing the biological growth that had infested the marble blocks of the temple in the century since it was excavated, and the J.M. Kaplan Fund generously supported a five-year program of conservation between 2014 and 2018. This was carried out by conservators Michael Morris and Hiroko Kariya, and by a dedicated crew of local women and men. Their work transformed the temple from a blotchy, blackened ruin to a gleaming, glowing wonder, and revealed previously unsuspected details of the building. Where possible we have used photographs of the temple’s current state; comparison with earlier photos will dramatically illustrate the metamorphosis, and will, we hope, preserve this magnificent monument for the future.

Like any archaeological project, research and publication of the temple would not have been possible without the efforts of many talented and dedicated members of the Sardis team. Expedition Editors Katherine Kiefer, from 1996 to 2016, and Kerri Sullivan, since 2019, contributed in more ways than can be enumerated here, working closely with the author to realize his vision for the publication. The manuscript was copy edited by Theresa Huntsman and proofread by Kerri Sullivan. Special mention must be made of Brianna Bricker, who, wearing many different hats as Senior Architect, Publications Coordinator, and Editor, played essential roles in many stages of research, documentation, and preparation of the volume. We are grateful to the anonymous reader for numerous helpful and constructive comments, and to other readers who commented on specific sections. Over the years Yegül was aided in his work on the field drawings by a number of architects and draftsmen, who are listed in the illustration credits; particular mention should be made of Catherine Alexander, Brianna Bricker, Felipe Rojas, and Philip Stinson. Philip Stinson was also responsible for the total station surveys, which brought new precision to the documentation of the building. He digitally stitched the individual 1:20 hand drawings into plans and elevations, making complex adjustments to correct for scanning and other distortions. Many photographers contributed their skills to the documentation of the temple, including Elizabeth Gombosi, Maria Daniels, Eliza Proctor, Ricky Taylor, Karen Heredia, Jessica Salley, Sara Champlin, Ellie Jordan, and Jivan Güner. Andrew Ramage, Series Editor of the Sardis Reports and Monographs series for twenty-three years, contributed in many ways to the development of this publication, and his wise and humorous voice is always present. And throughout the years, Sardis Expedition Administrators Elizabeth Gombosi and Bahadır Yıldırım, Associate Director Laura Gadbery, Sardis Coordinator Robin Woodman, George M. A. Hanfmann Curator of Ancient Art Susanne Ebbinghaus, and Teoman Yalçınkaya, representative of the Expedition in Izmir, have kept operations running in Cambridge and in Turkey, and each has been an indispensable source of wise advice and counsel. It is a privilege to work with such a talented group of scholars and friends.

Excavation and research at Sardis is conducted each year with the generous support and permission of the Republic of Turkey, particularly its Ministry of Foreign Affairs, Ministry of Culture and Tourism, and General Directorate of Cultural Heritage and Museums; as always we are grateful for the privilege of working at this fascinating and important ancient city.

In thanking the current Minister of Foreign Affairs Mevlüt Çavuşoğlu, Minister of Culture and Tourism Mehmet Nuri Ersoy, and Director General of Cultural Heritage and Museums Gökhan Yazgı for their support, we wish to also extend our thanks to the many dedicated Turkish civil servants whose professionalism and thoughtful care have facilitated work at the site for more than sixty years. Former Excavations Department Director Melik Ayaz and current director Umut Görgülü have offered friendly guidance for many years, and we are particularly indebted to him and his staff. Eren Sülek, director of the Archaeological and Ethnographic Museum in Manisa, and his predecessors Lütfi Ekinci, Harun Güllü, Sevgi Soyaker, Müyesser Tosunbaş, Hasan Dedeoğlu, and Kubilây Nayır, have given their constant support to research at Sardis over the decades, and it has been a pleasure working with them all.

The Sardis Expedition is sponsored by Harvard Art Museums and Cornell University, and has had its headquarters at the Harvard Art Museums since its inception. We are very grateful to the Director of the Museums, Martha Tedeschi, for her generous interest in and support of the project, and for providing a welcoming home for research, publication, and documentation. Harvard’s Faculty Oversight Committee for the Sardis Expedition, and particularly Vice Provost for International Affairs Mark C. Elliott and Prof. Adrian Stähli, are strong supporters of the research and education missions of the Expedition, and we are very grateful for their counsel.

Archaeological fieldwork and research at Sardis is made possible by the generous sponsorship of many individuals and institutions. Research, conservation, and publication have been supported by numerous grants from the U.S. Department of State and the National Endowment for the Humanities.10

Among the private donors are Dr. Alexandre Balkanski, Mr. and Mrs. Max Barus, the Bollingen Foundation, the Ruth Covo Family Foundation, the J. Stephens Crawford Trust, the Ford Foundation, Mr. and Mrs. William Frederick, Mrs. Margaretta Frederick, Dr. and Mrs. David Greenewalt and the David Greenewalt Charitable Trust, Dr. Richard Hamilton, Mr. Patrick J. Healy, Hyacinth Foundation, the Institute for Aegean Prehistory, the J.M. Kaplan Fund, Nanette Kelekian, the John M. Kohler Foundation, the estate of William Kohler, Dr. Edwin H. Land and Mrs. Land, Mr. Thomas B. Lemann, the Lucius N. Littauer Foundation, the Loeb Classical Library Foundation of Harvard University, Rosemary Lonergan, Mr. Scott Malkin, Mr. Osman Mardin and his talented family, the Charles E. Merrill Trust, the Old Dominion Foundation, Mr. Donald I. Perry, the John and Emma Quint Memorial Fund, Mr. John J. Roche, the Billy Rose Foundation, the Rowland Foundation, Mr. Andrew Seager, Mrs. Valerie Smallwood, the Susan G. Soderquist Trust, the Eleanor Ransom Swift Trust, Edward Teppo, Richard and Genevieve Tucker, the Vila B. Webber Charitable Trust, the estate of W. C. Burriss Young, and the Zemurray Foundation, as well as several anonymous donors.

And finally we are grateful, as always, to the loyal Supporters of Sardis and to the many members of the expedition, whose interest in this project over the past sixty-two years is continually expressed in so many ways. It is a privilege to be a part of such a broad and devoted community.


  • 1Sardis II.
  • 2For instance, Lansing C. Holden measured and sketched the capitals of the temple, including the two in situ, by throwing a stone with a string attached over a standing column and hoisting up a bosun’s chair to ascend to its top: see Greenewalt et al. 2003, no. 22.
  • 3Sardis R1, pp. 53–117.
  • 4Sardis R2, no. 102 (inv. S61.27.14).
  • 5Gruben 1961.
  • 6Howe 1999.
  • 7Sardis R3.
  • 8The mortared rubble of the peristyle was drawn separately at 1:20 by Brianna Bricker in 2011; that drawing will be made available digitally.
  • 9Cahill and Greenewalt 2016.
  • 10The Department of State made grants in Turkish funds to Harvard under the Mutual Educational and Cultural Act Public Law 87-256 and Agricultural Trade Development and Assistance Act, Public Law 480 as amended (SCC 29-543). National Endowment for the Humanities grants to Harvard University for work at Sardis and for publication work in the Cambridge office: H67-0-56, H68-0-61, H69-0-23, RO-111-70-3966, RO-4999-71-171, RO-6435-72-264, RO-8359-73-217, RO-10405-74-319, RO-23511-76-541, RO-20047-81-0230, RO-20607-84, RO-21414-87, RP-10050-80-0387, RP-20247-81-2162, RP-20360-82, RP-20754-86; to Cornell University GM-21549-83. Recent conservation work at Sardis has been funded by a grant from the U.S. State Department, U.S. Embassy, Ankara: S-TU-150-17-GR-051-A01. The opinions, findings, and conclusions stated herein are those of the author and do not necessarily reflect those of the United States Department of State.