Chapter 5. The Sculpture of the Hellenistic, Roman, and Early Byzantine Periods. Literary, Epigraphic, and Archaeological Evidence.

A large proportion of our Hellenistic and Roman sculptural finds is sadly fragmentary. Some undoubtedly have been mutilated by falls in natural disasters, especially earthquakes — the destruction of A.D. 17 being the most serious.¹ Earthquake is almost certainly the explanation for the fallen architectural fragments discussed in this catalogue, of which the head-capitals form the most cohesive group (Cat. 197, Cat. 198, Cat. 199, Cat. 200, Cat. 201, Cat. 202, Cat. 203, Cat. 204, Cat. 205, Cat. 206, Cat. 207, Cat. 208, Cat. 209; Figs. 349, 350, 351, 352, 353, 354, 355, 356, 357, 358, 359, 360, 361, 362, 363, 364, 365, 366, 367, 368).

Other sculptural finds have been reused as building blocks, and were often smashed on purpose to a desired shape - usually in antiquity, but sometimes also in modern times. The mine of sculpture found in the piers and walls of the Synagogue is the most outstanding example of this practice. The Zeus head Cat. 107 (Figs. 231, 232) is one of many pieces used as rubble or building blocks in the long north hall of the Gymnasium. Damage to the Hellenistic relief from PN (Cat. 135 Figs. 269, 270) is due to reuse as paving in the fifth or sixth century A.D., when its center was cut out to form a well head.²

The most important reason for the loss of much of the sculpture of Sardis is undoubtedly the value of the stone which, when burned, is used as lime. Poignant testimony for this was the discovery in 1972³ of a limekiln in one of the rooms of the Gymnasium and, nearby, a number of partly burned statuary fragments (Cat. 95, Cat. 101, Cat. 139, Cat. 157, Cat. 222, and Cat. 216, Figs. 213, 214, 222, 274, 275, 304, 387, and 378, the last not burned but smashed and apparently also destined for the limekiln).⁴ Other partly burned fragments were also found elsewhere in the excavations (for example, Cat. 221, Figs. 386).⁵

Only a small proportion of the sculpture survived in good condition. A few examples were found in or near their original positions. (such as Cat. 225 Fig. 390); others were apparently put down on the ground and covered with earth for some unknown reason (Cat. 59, Cat. 67, Figs. 172, 173, 182, woman and man from HoB) and were thus well preserved; and occasionally, though mutilated, a section of a head or body may be in good condition, as Cat. 92 (Figs. 206, 207), a sage.

Relatively few of our pieces came from a context that suggests their original manner of display, but something can be made of the evidence. Least puzzling are the architectural fragments when found fallen from the buildings they decorated. Here we have an especially rich series of capitals from the screen colonnade of the Marble Court, the designation for the Imperial cult hall of the Gymnasium (MC on Fig. 4), all fallen within a few feet of that colonnade (Cat. 198, Cat. 199, Cat. 200, Cat. 201, Cat. 202, Cat. 203, Cat. 204, Cat. 205, Cat. 206, Figs. 351, 352, 353, 354, 355, 356, 357, 358, 359, 360, 361, 362, 363). I made efforts to ascertain some logical order from the findspots,⁶ which might have placed Dionysus, the central figure of the theme, in the center of the group, or to place all heads turned in one direction on one side of the colonnade. However, the findspots did not bear out these assumptions, and it seems as if the heads were placed at random. The reconstruction gives a fine impression of the original effect (Figs. 347, 348).⁷ The capital with an Anatolian goddess (Cat. 194, Fig. 344), found near floor level on the interior of the south hall (BSH) of the Gymnasium, should belong to that room.⁸ For other architectural pieces, we do not know the findspots (for example, the pier showing the birth of Zeus, Cat. 213, Figs. 373, 374, 375, which was found in the plain near Yılmaz Köy, Cat. 214 and Cat. 215, Figs. 376, 377), and we may know their function but not their place.

The positions of a few pieces of statuary are known. The famous sarcophagus of Claudia Antonia Sabina (Cat. 243, Fig. 422) was set up near the ancient east-west highway in her mausoleum, of which a fair reconstruction can be made.⁹ The fragments of colossal statues of Antoninus Pius and Faustina (Cat. 79 and Cat. 251, Figs. 196, 197, 434) were found in or near the Temple of Artemis and were clearly set up as cult statues there - as may have been their predecessor, the colossal statue of Zeus (Cat. 102, Figs. 223, 224, 225), of which we also have fragments. It is possible that Zeus and Antoninus, and Artemis (Cat. 252, Fig. 435) and Faustina, sat side by side.¹⁰

Inscriptions also present tantalizing evidence for statues of which little or nothing remains. Of the standing portrait of Lucius Verus in the Gymnasium, recorded in a long inscription (Cat. 276, Fig. 469), we have only the foot preserved.¹¹ Another example is the base of the statues of the “children of Kore” (Cat. 277, Fig. 470), found in situ in BE-C of the Gymnasium but with no traces of the figures preserved.

Occasionally a statue found on or near floor level clearly had been set up in the room where it was found. In one instance, a table leg with a male figure (Cat. 225, Fig. 390) was found in its original position, standing in the corner of a room in the House of Bronzes (Fig. 1 No. 4). Two large statues found in the same area, a draped female and male (Cat. 59, Cat. 67, Figs. 172, 173, 182), were apparently purposely placed lying foot to foot for some unknown reason late in the Roman period. It seems likely that the two originally stood near each other.¹²

We also have evidence of earlier pieces being set up at a later date in secondary positions. The Lydian and Greek stelai set up by the Romans near the Temple of Artemis are examples of this.¹³ Another is Cat. 278 (Fig. 471), a fountain base with snakes that was, according to its inscription, moved from its original position "in the middle of the road" to a secondary place. There is also the third century head of a bearded sage (Cat. 92, Figs. 206, 207), found in the bedding for the Byzantine Road, which was perhaps set along the Main Avenue late in the Imperial period.

The vast majority of our pieces could not be dated by their archaeological context because so many were found reused as building materials, in piles doomed for the limekiln, or in other inconclusive positions. In a few instances there is external evidence for dating, as in the case of architectural pieces where the date of the building is known. For example, the inscription on the Marble Court places the dedication in A.D. 211-212,¹⁴ and thus dates the series of head-capitals which decorated the screen colonnade of this structure. Also, portraits of known persons (Cat. 78, Figs. 194, 195, Sabina) or types close in style to known portraits (as Cat. 93, Figs. 208, 209, 210, 211, a Diocletianic priest) can be dated fairly closely. In most cases however, the dating has been based on criteria such as style of drapery or technique of carving. Comparisons have been cited where possible. Dates cited here are meant to serve only as a starting point for further analysis, since stylistic dating is subject to personal interpretation.

Among our finds are a number of copies of well-known sculptured types, both Greek and Roman. A Roman copy of a bearded Zeus (Cat. 107, Figs. 231, 232) must ultimately derive from the Phidian Zeus at Olympia and has been related to a number of well-known types.¹⁵ An Amazonian relief (Cat. 188, Fig. 338) may also derive from a Phidian model, the famous shield of the Athena Parthenos.¹⁶ Our gilded Tyche, by the crown and the turn of the head, may be closely related to the Tyche of Antioch by Eutychides.¹⁷ A fragmentary Herakles is a copy of the Farnese Herakles type, which may reflect a Herakles made by Lysippus (Cat. 116, Fig. 242).¹⁸ We also have a number of fragments of the Venus pudica, too fragmentary or poor, however, to have been included in the catalogue, and a Venus Anadyomene (Cat. 111, Fig. 237). In addition to the standing figure of Asklepios in Manisa (Cat. 125, Figs. 254, 255, 256), a number of poorly preserved small bases with two feet and a snake and sometimes the end of a staff are evidence of his cult; most were probably small votive statuettes, dedicated as wishes or thanks for good health. Of Roman iconographical types we have examples of Aeneas carrying Anchises (Cat. 73, Fig. 189), Cybele or Tyche and a tree (Cat. 159, Fig. 306), and a bound captive, well known in Imperial reliefs (Cat. 191, Fig. 341). None are outstanding examples of their types, but all bear witness to the fact that sculptors at Sardis were well aware of the mainstream of Roman art.

A unique case of archaeological and epigraphic evidence combined is the portrayal of the City Goddess of Sardis on the inscribed base found at Puteoli, which was erected by the cities that had suffered in the earthquake of A.D. 17 (see Appendix and Fig. 472). The goddess, accompanied by a child, certainly reflects a famous Greek type, the so-called Prokne and Itys by Alkamenes.

Very little light is shed on the sculpture from literary sources. The only direct references are Pausanias' mention of a statue of Adrastos of ca. 323-322 B.C.,¹⁹ and a statement in pseudo-Codinus, Patria 2.73, about images that stood in the Hippodrome of Constantinople, some of which were carried off from Sardis.²⁰ The epigraphic evidence sheds more light.²¹ Several of these inscriptions, summarized below, are included in the Catalogue (Cat. 275, Cat. 276, Cat. 277, Cat. 278, Figs. 467, 468, 469, 470, 471), but I have omitted the ordinary honorary statue bases. An important reference to the earthquake of A.D. 17 occurs on the base of a statue of Hera, restored by Julia Lydia after the earthquake (Cat. 275). Only the foot of Lucius Verus (Cat. 132, Fig. 264) remains of the statue which stood on the inscribed base of the Gymnasium (Cat. 276).²² A beautiful inscription honoring Caracalla fills the front of a base broad enough to support a statue; however, the inscription does not itself mention the statue, and there is no evidence of attachments on its top face.²³ Another dedication base (Cat. 277), this one to Caracalla and Geta, specifically mentions the images of the children of Kore which it supported.²⁴ Finally, a late inscription (Cat. 278) refers to a group of gilded bronze snakes from a fountain, set up by a governor of Lydia in the fourth century A.D.²⁵

Some of our statue types can be parallelled, and sometimes elucidated by numismatic evidence from Asia Minor.²⁶ For instance, the Anatolian moon-god, Men, who appears as a horseman on the Ashmolean relief Cat. 253 (Fig. 436), is shown on the reverse of a medallion of Marcus Aurelius in the same guise and in similar costume.²⁷ He is also shown with the crescent moon behind his head on a Sardian coin.²⁸ The relief of the Mother of the Gods in Berlin (Cat. 256, Fig. 442) our only definitely Roman representation of the famous indigenous goddess, is similar in dress and concept to a Severan coin of Smyrna, where, however, she is seen in side view.²⁹ (As with sculpture, there are surprisingly few coins of Cybele from Sardis.³⁰) The head of Tyche on Sardian coins³¹ is not in the same position as our frontal statue head Cat. 128 (Figs. 259, 260), but is additional testimony to the existence of the turreted city goddess. In the conventional position of river gods, just like our reliefs of the same (Cat. 189, Cat. 190, Figs. 339, 340), is a Sardian coin of the Hermus,³² and even our reclining Tmolus (Cat. 211, Fig. 371) has a numismatic parallel - a coin showing only his bearded head is clearly attested.³³ The openwork relief statuette of Helios as charioteer (Cat. 167, Fig. 315) is a frontal figure derived from the same source, probably a painting, as that which inspired a bronze coin of about A.D. 200.³⁴ That rendition of the rayed Helios, carrying a whip (or torch?) and drawn in a chariot by four horses, is an early example of the kind of composition that becomes standard for late antique and medieval art, and of which our statuette of the fifth or sixth century A.D. is a crude example.

Numismatic evidence is also important for statues which are now lost. A cult statue of Kore is shown on the reverse of Sardian coins, with the inscription Chrysanthia - a reference to the corn maiden and the golden flowers she brings forth.³⁵ An image of her (Cat. 194, Fig. 344) also appears on the figured capital from the south hall of the Gymnasium (BSH),³⁶ an important instance of a cult statue of an archaic date remembered or revived in Roman times. A statuary type may also be reflected in a coin of Sardis showing Herakles dragging a humped bull by the horns.³⁷

Zeus Lydios appears frequently on Sardian coins, both as a standing figure holding a scepter, with an eagle in his left hand, and as a bearded and filletted head.³⁸ We may have a copy of this figure, originally either a cult statue or a decorative image in the Gymnasium, in a head found during our excavations (Cat. 107, Figs. 231, 232).³⁹ He appears on a coin which shows also a monumental flaming altar with three caryatid figures - surely a copy of an actual altar to Zeus Lydios.⁴⁰ The features of Zeus Karios and Zeus Polieus are not known, unless the colossal fragment Cat. 102 (Figs. 223, 225), found near the Artemis Temple is from the cult statue of the latter.⁴¹

There must have been a large sculptors' workshop, or workshops, at Sardis. We have found not only great numbers of statues but also some unfinished pieces (Cat. 106, Cat. 117, Cat. 144, Figs. 229, 230, 243, 283)⁴² which would never have been brought to Sardis from elsewhere; they can only be explained by the presence of a local workshop. The unfinished pieces are of local marble and probably Roman, except a head of Zeus (Cat. 106 Figs. 229, 230) that seems to be early Hellenistic. Architectural sculpture is also testimony to a group of artists on hand to work together with the builders. In the head-capitals associated with the screen colonnade of the Marble Court and palaestra (Cat. 197, Cat. 198, Cat. 199, Cat. 200, Cat. 201, Cat. 202, Cat. 203, Cat. 204, Cat. 205, Cat. 206, Cat. 207, Cat. 208, Cat. 209, Figs. 349, 350, 351, 352, 353, 354, 355, 356, 357, 358, 359, 360, 361, 362, 363, 364, 365, 366, 367, 368), two schools can be distinguished: one trained in the Hellenistic tradition and one rooted in a typically Roman style.⁴³ Also interesting here is the fact that we have fragments of two head-capitals almost certainly carved by the same hand (Cat. 202, Cat. 203, Figs. 357, 358).

The large number of garland sarcophagus fragments is further evidence of local manufacture; in addition we have many pieces of an Asiatic sarcophagus whose architectural fragments (Cat. 182, Fig. 331) have patterns identical to those on the celebrated sarcophagus of Claudia Antonia Sabina (Cat. 243, Fig. 422), and must have been made in the same workshop. Morey has presented his case for a Lydian workshop in full detail,⁴⁴ but of late it has been much questioned.⁴⁵

Even in the later years of the Roman Empire, there is evidence for a local Sardian school in the outstanding head Cat. 95 (Figs. 213, 214) of the fifth or sixth century.⁴⁶ There is also other fine late Roman and early Byzantine sculpture, such as the decorative frieze with birds in the Synagogue forecourt, which is not, however, included in this study.⁴⁷

The Romans were not above helping themselves to a ready-made statue or inscription and reusing it for their own purposes. The most striking example of this is the Hellenistic stele with a moth (Cat. 137, Fig. 272), twice reused in Roman times. The subsequent owner did not trouble to alter the sculptural decoration, but simply added his own inscription; the third person inscribed yet another inscription and merely added a wreath. Another example of this reuse is the gorgon stele Cat. 150 (Fig. 295), where the date of the original inscription cannot be ascertained, but the present inscription is clearly superimposed on an earlier one, and the second user also added the gorgoneion.

In the luxurious late Roman House of Bronzes, a griffin pillar of much earlier date (Cat. 219, Fig. 384) and an inscription to the god Men were used to support a table, while the fountain relief of a river god (Cat. 190, Fig. 340) was reused as part of a basin.⁴⁸ A double reuse is attested for a stele which started out as an honorary inscription, was reused as a tombstone, and ended up as the front of a fountain in one of the Byzantine Shops, with a huge cross engraved upon it.⁴⁹

In statuary we have a late antique portrait head, recut to be used as a fountain piece (Cat. 94, Fig. 212). An inscribed base (Cat. 275, Figs. 467, 468) mentions a statue of Hera restored (that is, replaced) after the earthquake of A.D. 17. Another inscribed base (Cat. 278, Fig. 471) tells of a governor of Lydia, Basiliskos, who took gilded bronze serpents from a fountain which "stood in the middle of the road," and used them in another fountain. Apparently this second one, dated to the fourth century by the lettering, stood on top of the inscribed base, which was found in the aleipterion of the Gymnasium.⁵⁰

There are a few examples of statues repaired in antiquity, although most of the evidence of dowelling seems to have been for the original attachment of head or limbs. The male figure Cat. 67 (Fig. 182) from the House of Bronzes has a large section of the back right leg added on. The archaizing herm Cat. 110 (Fig. 236) has a smooth edge and a central dowel hole at the break on the shaft, suggesting an ancient repair. The tip of a leaf on the Zeus head-capital Cat. 197 (Figs. 349, 350) was broken off when found, but showed an ancient dowel hole, clearly for a repair. The fragment has been reattached by the expedition conservators.

Despite the ravages of earthquake and weather, the vicissitudes of taste, and the value of the material itself, we know that in the sixth or even seventh century A.D. pagan statuary was still on view at Sardis. The draped female and male from HoB (Cat. 59, Cat. 67, Figs. 172, 173, 182) were apparently buried by "post-Heraclian agency,"⁵¹that is, after the devastation of 616 which destroyed much of what remained of Roman Sardis. This implies that the statues, which we have dated to the first half of the first century A .D ., were still standing until that time. These statues, whose heads were inserted, were very likely given new portraits in later antiquity.⁵² Also in HoB was found a Dionysus (Cat. 225 Fig. 390) of the second or third century, which was in situ in a late Roman-early Byzantine context; it seems to have decorated a room used for domestic industry.⁵³ A statue group of Dionysus, a panther, and satyr (Cat. 122, Figs. 250, 251), a third or fourth century work, was found in the fill of a latrine by the Byzantine Shops along the south side of the Gymnasium together with Byzantine coins.⁵⁴ It is ironic to think of a largely Christian populace enjoying the decoration of such a clearly pagan subject in their latrines.⁵⁵ Statues such as these are testimony to what was undoubtedly a much larger group of classical statuary still on view in Byzantine times - a group which was gradually destroyed by the ravages of the limekilns.

These general conclusions made from the sum of the material we have found are only a beginning in the interpretation of the evidence for the sculptural taste and practice of the artisans of ancient Sardis. It is hoped, however, that the catalogue presented below will be useful in comparative studies, and that scholars will benefit from seeing the whole spectrum of sculpture from one of the most important cities of Asia Minor.

• 1Tacitus Ann. 2.47 = Sardis M2 (1972) no. 220; cf. 275 (467, 468). See also Appendix, base from Puteoli (472).
• 2BASOR 166, 18.
• 3BASOR 211, 27-28, fig. 9. Located in BE-C on 4.
• 4BASOR 187, 20-21; Hanfmann, Letters, 182.
• 5Other kilns recorded at Sardis: limekilns starting at least by the middle Byzantine era in the Gymnasium, BASOR 154, 16, and at Roman Bath CG, Sardis R1 (1975) 142, 163; lime-slaking pit at PN, BASOR 177, 3; limekilns near AT, Sardis I (1922) 67; Sardis II (1925) 12; kilns operating as late as the 19th C., M. Schmidt, AM 6 (1881) 151. On the destruction of AT see Sardis XI (1916) viii and ix, where the dismantling of the temple in the early 7th C. A.D. is discussed; and Sardis I (1922) on the destruction since the 17th C.
• 6Hirschland, 21.
• 7BASOR 203, figs. 13-14; Hanfmann, Letters, 232, fig. 210.
• 8G.M.A. Hanfmann and M.S. Balmuth, "The Image of an Anatolian Goddess," in Bossert Festschrift, 267-268, discuss the relevance of placing such a figure in the Gymnasium
• 9Sardis V (1924) ill. 2.
• 10Sardis R1 (1975) 79-81, 87. “Artemis” is a portrait of Lucilla; cf. 252 435.
• 11BASOR 154, 14, fig. 4; Yegül, "Bath-Gymnasium," 102f.
• 12Hanfmann, Letters, fig. 38.
• 13Sardis R1 (1975) 68-73.
• 14BASOR, 42.
• 15Metraux, 156ff., and see "Coins," infra.
• 16Neda Leıpen, Athena Parthenos, A Reconstruction (Toronto 1971) 41-50, figs. 23-29, 81-83.
• 17Dohrn, Tyche, passim; Richter, Sculpture, figs. 753ff.
• 18Bieber, 37.
• 19Sardis M2 (1972) no. 270.
• 20Th. Praeger, ed., Scriptores originum Constantinopolitanarum II (Leipzig 1907) 189 = Sardis M4 (1976) 107, no. 6.
• 21Vermeule, Imp. Art, 69, 170.
• 22BASOR 154, 14, fig. 4; 158, 7-10; Hanfmann, Letters, 33, figs. 17-18.
• 23Vermeule, Imp. Art fig. 165.
• 24BASOR 211, 27, fig. 8; Sardis R1 (1975) 32.
• 25BASOR 211, 22, fig. 4; Sardis R1 (1975) 173, n.75; Sardis M4 (1976) 21, 114-115.
• 26The Greek and Roman coins found during the Harvard-Cornell excavations will be published in a forthcoming Sardis volume by A. E. M. Johnston and T. V. Buttrey respectively. For the coins found by the first Sardis expedition, see Sardis XI (1916).
• 27Vermeule, Imp. Art, 167, fig. 103.
• 28BMC Lydia, pl. 26:3.
• 29Vermeule, Imp. Art, 159, fig.90.
• 30BASOR 211, 35.
• 31BMC Lydia, pls. 24:15, 26:1-2; cf. also Appendix (Fig. 472), personification of Sardis with crown on base from Puteoli.
• 32Ibid., pls. 26:9, 27:13.
• 33Ibid., 249, no. 94, pl. 25:10
• 34Vermeule, Imp. Art, 164, fig. 98. Cf. Hanfmann, Roman Art, 242, 327, pl. XIV, with references to the mosaics in Rome and Hammath-Tiberias.
• 35BMC Lydia, cix f. and pls. 26:2, 27:2; G. M. A. Hanfmann and M. S. Balmuth in Bossert Festschrift, 268; Fleischer, 197; BASOR 211. 35. A . E. M . Johnston will treat in detail the role of Kore at Sardis in a forthcoming volume on coins in this series.
• 36Hanfmann-Balmuth, ibid.; Fleischer, 187, 193ff., pl. 77.
• 37BMC Lydia, pl. 27: 7. Herakles dragging a bull is also on Hellenistic relief ware fragments found at Sardis (P63.177:5204 and P65.283:6945) which will be published by I. Hanfmann in a forthcoming volume in this series.
• 38BMC Lydia, cix f., pls. 25:1, 3, 9; 26:6.
• 39Metraux, 155ff.; he gives a full discussion of the evidence and cites the bibliography.
• 40Price-Trell, fig. 244. We regret that this volume appeared too late to be fully utilized.
• 41On Zeus Karios and Zeus Polieus at Sardis, see Metraux, 157ff. Price-Trell still object to a temple to Zeus Polieus and Artemis, but an inscription (Sardis VII [1932] no. 8, XII, lines 133-134) explicitly proves the existence of a joint sanctuary. On Zeus Polieus, see also L. Robert, CRAI (Apr.-June 1975) 321.
• 42See supra, Ch. I, “Technique.”
• 43Hirschland, 20f.
• 44Sardis V (1924) 71ff.
• 45Most recently by H. Wiegartz in Mansel Mélanges, 376ff.
• 46Contra the theory of a late Sardian school, see Inan-Rosenbaum, 44.
• 47BASOR 187, 49f., fig. 60. To be published in a forthcoming volume in this series on the Synagogue.
• 48BASOR 175, 24; Robert, NIS, 33-34; Hanfmann, Letters, 56, figs. 34, 35 (inscription to Men), 63 (river god). Cf. Sardis M4 (1976) 43f.
• 49BASOR 157, 32-33, fig. 18; Hanfmann, Letters, 55-56, fig. 19. The suggestion made by Hanfmann that the shop served as a baptistry is weakened by the occurrence of similar crosses on the front of a street fountain at Ephesus.
• 50BASOR 211, 21-22, fig. 4 from hall BE-W, now designated BE-H on Gymnasium plan Fig. 4.
• 51BASOR 157, 27; Hanfmann, Letters, 59, fig. 38.
• 52Vermeule, Imp. Art, 7.
• 53BASOR 154, 27.
• 54BASOR 157, 34-35. The coins are listed in Sardis M1 (1971) nos. 120, 342, 370, 378, 463, 467, 482, 506, 507, 713, 719, 720, 824, 927, 1006, 1010. The Shops will be published by J. S. Crawford in a forthcoming volume in this series.
• 55See Brinkerhoff, Antioch, 56ff.