Lydian Architectural Terracottas
Important factors must have caused the transition from thatched to tiled roofs. Architectural terracottas and roof tiles were developed in order to prevent humidity and damage to the wooden roof structures and mudbrick and wooden façades. They improved the durability of the wood and thus secured a longer life-span for buildings. But given the great difference in cost between thatched and tiled roofs, this development may also have been due to a measure against fire.1
According to ancient sources, a tiled roof was first used by Kinyras, the king of Cyprus2, though this has not been proved by archaeological finds. Exactly when roof tiles were introduced to Anatolia is not mentioned in ancient sources, but we learn from Pliny the Elder that Demaratus, a Corinthian aristocrat, took refuge along with three artisans in Etruria in 657 BC; he may have introduced tiled roof to Italy from Greece.3 In mainland Greece, tiled roofs first appear in the Temple of Apollo at Corinth, dated to the second quarter of the seventh century BC.4 In Sardis, whose high-quality finds suggest that it was an important production center in Anatolia, fragments indicate the use of roof tiles by the late seventh century BC.5
Wood is an essential part of the framework and structural components of every roof, but rain and damp undoubtedly shortened its life. Terracotta roof tiles protected the framework from damp, and drained rainwater falling on the roof. They also prevented the exposure of wooden structural elements to the air. In a peristyle building with a gabled roof, where the entablature extended beyond the walls, gutter (sima) tiles and spouts were sufficient to keep the water away without doing any harm. But on non-peristyle buildings, protruding spouts, gutter tiles or short roof gutters could not lessen the effect of damp. The long rain gutters on simas recovered from Gordion6, Sardis7, Düver8, Euromus9 and Hacıbayramlar10 show that the buildings in question were not peristyle in plan. Long rain gutters accelerated the water and drained it far from the entablature and sima.
Architectural Terracotta Production and Trade
Architectural terracotta production in Archaic Anatolia is spread over a wide area, but is particularly common in Phrygia, Lydia, Ionia and Aeolia. Miletus stands out as the largest production center in western Anatolia. Studies of the Archaic architectural terracottas from Euromos, a Carian city near to Ionia, reinforces its prominent position.11 Finds from Larisa and Phocaea in Aeolia indicate that Phocaea was another production center, and the Larisa assemblage was the work of the Phocaean craftsmen.12 Phrygian finds are more diverse in appearance, since Gordion is located on an east-west route, and in closer contact with the Near East through trade. Besides the limited contact with the western Anatolian centers like Miletus, Didyma, Sardis and Neandria13, close relationships with examples from Düver and Pazarlı suggest that there is more to be learned about the production of the Phrygian architectural terracottas.
Sardis must have been one of the important production centers in western Anatolia. It is among the earliest findspots of architectural terracottas known in Anatolia, and further studies of the Sardian examples14 will shed light on the questions of dating and production. Scholars have proposed different dates for the Sardian architectural terracottas, from the late seventh–early sixth centuries to ca. 570 BC.15 Different dates have also been proposed for the Larisan assemblage.16
There is no doubt that trade networks were crucial in the production and distribution of architectural terracottas. Rivers such as the Maeander (Menderes) and Hermus (Gediz), which flow into the sea on the western shore of Anatolia, provided routes for trade and communication between Ionia and Lydia. The use of these rivers also strengthened the artistic bonds between these two regions. The geographical location of Sardis and Lydia between Ionia and Phrygia made it a natural bridge in the production and use of architectural terracottas, where influences from both regions on figured and decorated examples can be observed. The similarities in dimensions and molds among terracottas that are thought to be from different provenances may suggest that most terracotta workshops were mobile. Although the headquarters of these mobile workshops may have been in urban centers, commissions in these centers might not have been sufficient sufficient and reliable enough to sustain the workshops in changing economic conditions. The workshops, which had been more regional in character before the middle of the sixth century BC, seems to have flourished after the coming of the Persians in the second half of the sixth century BC. In the light of current evidence, it appears that in the last quarter of the sixth century, especially in Ionia and Caria, new building activity flourished and older buildings were repaired. This boom in the market must have required close trade relations between Lydia and Ionia. As a matter of fact, there is evidence for this kind of contact between Lydia, Ionia and Caria.
In addition to inter-regional mobile workshops, there must have been local or regional ones. This is occasionally revealed by the differences in style and quality. We suggest that in Lydia there were workshops active only in that region, though it is quite possible that mobile workshops also took part in production through trade and communication networks along the Gediz and Büyük Menderes rivers and land routes. We should not imagine mobile workshops as a group of craftsmen concentrating on a single commission. Clients’ demands also caused workshops to diverge from their usual designs, a fact which should be taken into consideration while studying the still debated assemblages from Düver-Burdur.
The similarities among lateral simas with long rain gutters demonstrates the contacts between Lydia, Ionia, Caria and Phrygia. Lotuses on spiral branches and motifs of tongue and cyma on both sides of the gutter were employed in Lydia17, Phrygia18, Ionia,19 and Caria.20 The workmanship of terracottas in Lydia, Ionia and Caria demonstrates such similarities that one can show the existence of very close master-apprentice relationships in the last quarter of the sixth century BC.21
Close contacts between Sardis, Ionia, and Caria can be observed in the last quarter of the sixth century BC. When one studies the similarities between the architectural terracottas with scroll motifs, the artistic affinities between the Sardis and Euromus finds, the latter being the work of Milesian craftsmen, become obvious.22
The “Lydian Building” at Sardis
The “Lydian Building,” which was reconstructed in the excavation house using reproductions of the original architectural terracottas from Sardis, demonstrates how they might have been used (No. 60, fig. 1).23 Chariots are depicted on the raking sima, while on the lateral simas, on both sides of the long rain gutter we see two rearing Pegasoi facing each other. Below them are partridges, a goddess carrying Anatolian panthers in each hand, and star-and-scroll motifs. One fragment of the partridge friezes, which appear prominently on the raking simas from Caria and Ionia, was also discovered on a plaque.24 The figure of a partridge walking towards the left closely resembles those from Ionia25, Caria26 and even Phrygia.27 The partridge frieze which inspired its counterpart on the Lydian House, is from a raking sima whose recorded provenance is Mylasa, and now in the Bodrum Museum.28 Although it was claimed that the plaque may be from Euromus,29 in fact it had been found at Hacıbayramlar in Caria and delivered to Milas Primary School, thence to the Bodrum Museum.30 The fact that the closest parallel for the partridge on this sima plaque is found on a cup by the Bochum Painter from Turgutlu (Lagina), in which even the meander border below the frieze was imitated, shows the close contact between Lydia and Caria, as well as between Fikellura ware and architectural decoration. During a process that began with the Persian invasion of Anatolia, a younger generation that had been trained by architectural terracotta masters from Ionia and perhaps Lydia had migrated to Italy, and employed figures and motifs they had learned from their masters on Caeretan hydriae, thus transferring the craftsmanship of architectural terracottas to another medium.31
Among the sima plaques from recent excavations at Sardis, one fragment is remarkable for its unusual depiction. The plaque belongs to a lateral sima with a long rain gutter in the middle. A komast figure below a rampant Pegasus is noteworthy (No. 59, Figs. 2, 3). The depiction of this mysterious and orgiastic dance of the male figure, who is almost in an ecstatic state, on the lateral sima must have been the artist’s innovation. This figured decoration, which gives the impression of a cultic dance, is unique among architectural terracottas. The painted Pegasus in relief and the komast figure in silhouette technique drawn in the space below indicate that this composition was the choice of the painter, rather than the mold-maker. A close parallel comes from a lateral sima from Koranza in Caria, on which men are dancing before rearing Pegasoi32, but these figures were rendered in relief. The scene is not the creation of the painter but the artisan that made the molds. The komast in silhouette immediately below the rearing Pegasus presents itself as the product of the painter of the Sardian sima plaque.
- 1Wikander 1988, 203.
- 2Pliny, Naturalis Historia 7.56, 195.
- 3Pliny, Naturalis Historia 35.152; Åkerström 1966, 257ff; Winter 1993b, 17.
- 4Winter 1993b, 17.
- 5Ramage 1978, 9, 38–41.
- 6Glendinning 2007, 183.
- 7Åkerström 1966, taf. 46. 1–2; taf. 48. 1; taf. 49. 1–2.
- 8Buzzi 1999, 58. Abb. 21; 102.
- 9Ateşlier 2009, 152.
- 10Ateşlier 2009, 155–157.
- 11Ateşlier 2006; Ateşlier 2009.
- 12Åkerström 1966, 200. Abb. 1–2; Akurgal 1962, pl. 101, fig. 24.
- 13Glendinning 2007, 184.
- 14Ramage 1978, 38–41.
- 15Ramage 1978, 35; Billot 1980, 292–293.
- 16Winter 1993a, 234–254; Winter 1993c, 30.
- 17Åkerström 19666, taf. 49. 1–2; taf. 50. 1–2; Ramage 1978, fig. 64–66.
- 18Glendinning 1996, fig. 22; Düver: Buzzi 1999, 60. Abb. 23.
- 19Ateşlier 2006, 62; Ateşlier 2009, 152.
- 20Ateşlier 2006, 73, fig. 1.
- 21Ramage 1978, fig. 88, 91; Ateşlier 2006, 68, 76, fig. 3, 16.
- 22Sardis: Ramage 1978, 27, fig. 88. (Sardis inventory no. T61.18:3285 (525–500); Euromos and Miletus: Ateşlier 2006, 68, 76, fig. 16.
- 23Hostetter 1994.
- 24Hostetter 1994, 10–11, fig. 25; Sardis inv. T61.76:3737. For another fragment, see Ratté 1994, 373. pl. 83.5.
- 25Åkerström 1966, taf. 53. 4; Åkerström 1978, pl. 86, fig. 4.
- 26Ateşlier 2006, 68, 77, fig. 17–18.
- 27FGU 2007, 236.
- 28Åkerström 1966, taf. 59. 1; Hostetter 1994, 10–11, fig. 24, 28; Ateşlier 2006, 68, 77, fig. 18.
- 29Hostetter 1994, 11 n. 22.
- 30Serdaroğlu 1972, 80.
- 31Hemelrijk 1984, 193.
- 32Tırpan and Söğüt 2005, 54, Fig. 61.