Period 1: early fourth to early fifth century

Within the architectural remains, four phases of construction can be distinguished. In the first period, the main hall was established at the ground level of an elite house that succeeded earlier buildings of the imperial period on the site and extensively reused their building materials. The hall was inserted behind the large public-facing apse and along the north property wall. Internally, the hall was at first a fully rectangular space. Its current north and south walls belong to this first period, but the original east and west walls were lost when replaced in the next phase. Based on the extensive brick overhang, the hall was covered by a vaulted ceiling finished with stucco.

The main hall was adjoined to the east by a cistern and to the west by a vaulted portico, probably part of a larger open area or courtyard. Only the brick consoles and mouldings of this portico survive in the north wing of the tetrastyle courtyard. The presence of a cistern to the east is suggested by several features: a thick calcium carbonate deposit preserved on the floor and lower parts of the later Room 11; surviving patches of hydraulic mortar above its vault; and the cut for, and remains of, an earlier barrel vault in the north property wall above the later barrel vault of Room 11. Access to the cistern was provided by a narrow passage at its south-west corner. It was connected to the same water-supply system as the fountain in the east wall of Room 10 and seems to have been used for ritual activities in the main hall, particularly in relation to the unusual pit (Room 24) countersunk into the floor of the main hall (Room 12).

In this initial period, the main access into the underground complex was probably through a large arched opening at the west end of the main hall’s south wall. Invited participants could presumably have approached the complex from elsewhere within the residence, and an additional entrance from the street to the north is probable — perhaps via the predecessor of Rooms 9 and 17 which survive only in their late fifth-century form.

The bricks used for the consoles in the north wall of the tetrastyle courtyard, as noted, are unusual and exceptionally well made. They probably indicate a construction date in the period of decreased output from the city’s marble quarries after the mid-third century,Footnote ¹⁸ and may point to inspiration from central Italy. A better-dated example of faux-marble use can be seen in the slab-built and stuccoed columns surrounding a large porticoed space in the Governor’s House, archaeologically dated to around a.d. 300 or shortly thereafter. The Atrium House as well, located less than 90 m south of the House of Kybele, was being rebuilt and expanded at roughly the same time.Footnote ¹⁹ Its apsidal garden space (internal diameter of the apse: c. 14 m) provides a parallel on a larger scale to the apsed space south of the main hall (Room 14, internal diameter: c. 11.5 m), also presumably open. Both were constructed in reused petit appareil, were two-storeyed, faceted on the exterior and semi-circular on the interior, and incorporated interior niches (four in the Atrium House, one in the House of Kybele). Although further soundings will be required to clarify the building chronology of the House of Kybele, it is probable, on the basis of these parallels, that the complex formed part of a building boom that followed the elevation of Aphrodisias to the status of provincial capital of Caria in around a.d. 300.Footnote ²⁰ The later third and fourth centuries also saw the construction of private cult spaces elsewhere in the Empire.Footnote ²¹

Period 2: early to late fifth century

In its next, second phase, the underground complex more than doubled in size, with the addition of a series of connected rooms surrounding the main hall (Rooms 11, 21, 22 and the tunnel 25) and of finely decorated spaces to the west and southwest (Rooms 28 and 30). In this period, the main hall was given its current western wall and entrance from the tetrastyle courtyard. This entrance wall was probably located slightly further to the east than its predecessor. At this time too, the stairway into the new underground Room 11 was created — on the location of the first-phase cistern. The construction of the tunnel (Room 25) severed the ceramic pipe that had delivered water to the main cult hall. The removal of the cistern and the cutting of its water supply brought an end to the use of the underground pit (Room 24) in the main hall. A new beaten-earth occupation surface was put in place across Room 12, covering access to the pit. Instead, cult participants undertook a longer journey through the dark tunnel before ascending a short staircase and re-entering the hall through a low opening — only 95 cm high — cut into its south wall, requiring them to crouch and crawl (Fig. 8).

A new water system installed above the new spaces allowed calcium-rich water to seep through the vault of the tunnel (Room 25) and trickle down its walls, transforming it into an underground, cave-like environment. Another branch of the same water system supplied the fountain built into the east wall of Room 22.

In this second period, access to the underground complex was through the tetrastyle courtyard (Room 28, Fig. 27). Like the adjoining exedra (Room 30), it was first decorated with painted faux-marble revetment, soon replaced by the more elaborate scheme described above. The only architectural change that seems to have accompanied the second painted phase was the creation of a passage connecting the north-east corner of the courtyard directly with the older staircase to the upper storey.

Fig. 27.

House of Kybele. View into main hall (Room 12) from tetrastyle court (Room 28). Looking east.

Painted plaster in this second period seems to have characterised spaces that were more freely accessible, whereas the areas that we will see were used for cult activities had more subdued finishes, carefully designed to evoke a natural landscape of rock faces and cave ceilings. Because the main hall (Room 12) is earlier, this deliberately ‘naturalistic’ effect is here less pronounced. The destruction layers yielded no evidence for marble revetment or plaster finishes on its walls. Although the brick overhang at the springing of the vault allowed for stucco decoration, no traces were recovered from the debris. The rubble and brick masonry of the hall was largely left exposed — at least in the last main period of its use, and probably already from this conversion phase onward. This is supported by the presence of upright palm leaves scratched into the wet mortar coatings of the newly added Rooms 21 and 22.Footnote ²² An identical leaf discovered inside one of the bays on the north wall of the main cult room suggests that the walls of this room were provided with irregular mortar patches at the same time.

It seems, then, that an invited participant in the cult activities of the complex in this major second period would undertake a journey from a grander environment within the house, to the more restrained decor of the tetrastyle courtyard, to the austere environment of the main hall, arriving ultimately at the natural, cave-like spaces of the new annexe shrine and the tunnel.

This large-scale, second-period transformation of the underground complex is likely to have taken place in the early fifth century. Soundings elsewhere in the house indicate that it underwent a major reorganisation and expansion at this time.Footnote ²³ Both phases of painted decoration could belong to this period: imitation-marble panelling again became popular in the late third and early fourth centuries, continuing into the fifth century.Footnote ²⁴ The combination of red flowers and birds from the second-phase painting programme in Rooms 28 and 30 finds parallels in fourth-century tombs from Sardis,Footnote ²⁵ but the combination of red flowers and birds continues into the late fifth century in the Harbour Necropolis at Ephesos.Footnote ²⁶ In broader historical terms, an early fifth-century date would accord with the increasingly troubled religious climate of the period. Aphrodisias is known to have retained a strong pagan community and in this context, the impact of anti-pagan legislation is unclear. Public sacrifice may have continued into the fifth century, but cult practices formerly conducted in publicly accessible temples were eventually relocated to private spaces within residences.Footnote ²⁷

Period 3: late fifth to mid-sixth century

In the later fifth century, Aphrodisias was struck by a major earthquake that caused widespread damage.Footnote ²⁸ Excavations in the city centre have demonstrated that the resulting debris was levelled and redistributed, after which a new use-level was established on top. Differences in height between the levels before and after the earthquake are highly variable. Soundings throughout the House of Kybele neighbourhood confirmed that the level of the street running through the area was raised by at least a metre, and soundings within the house show that the rise in occupation levels varied considerably across the building. The floor of the newly created Room 9 — a service entrance from the street — was raised to align with the contemporary street level. From here, once could walk straight into the newly created corridor (Room 10), on top of the vaults of the tunnel, which went from being half-subterranean to fully subterranean. The rise in ground level outside the main hall and its annexes transformed the rest of the complex into a semi-subterranean basement, concealed by an upper storey that in this (third) phase extended over the entire underground complex. The large number of late-fifth- to early-sixth-century sherds embedded in the mudbrick of the upper storey walls provides good internal evidence for dating this final, large-scale reconstruction phase.

The upper sections of walls, ceilings and vaults throughout the complex also had to be repaired and redone. In the north-west corner of the main cult hall, a ventilation system was integrated, possibly echoing a previous design. The tetrastyle courtyard (Room 28) was reconstructed, its piers re-set and new brick vaults added. The remains of marble columns found in debris layers show that it had an upper storey in this phase. There were at least two access points to the rooms on the upper storey, the first the low staircase found in the north-west corner of the walkway-corridor around the apse (Room 10), which most likely led to further service quarters, and the second the new staircase installed in the north-east corner of the tetrastyle courtyard, where it covered the wall paintings of the previous phase. The original staircase, Room 26, was taken out of use, with its lower steps removed and brick screens inserted to form the new Room 29.

Inside the new Room 29, the wall painting from the previous phase was intentionally removed and replaced with a fresh layer of mortar, aligning with the more austere character of the cult rooms. Everywhere else, the wall decoration from the previous phase was retained and, somewhat surprisingly, not replaced, even though it must have been damaged in the earthquake. Throughout the underground complex, the earlier floor surfaces were replaced by earthen floors. In Room 22, these sealed the installations associated with the previous phase, rendering them inaccessible and putting them out of use. Finally, once the corridor at the back of the apse (Room 10) was finished, the water-supply pipe that created the cave-like effect in the tunnel spaces below was rebuilt on top of its paved floor.

Period 4: mid-sixth to early seventh century

At some point between the large-scale renovation of the late-fifth century and the violent destruction of the complex in a.d. 617, further modifications were undertaken that transformed the subterranean complex from an exclusive into a concealed cult space. The shallow mortared basin at the centre of the tetrastyle courtyard was filled up and the floor levelled with a layer of soil, suggesting that rainwater requiring drainage no longer entered the space. The open-air centre of the courtyard was therefore closed with a wooden ceiling, the charred remains of which were found in the collapse layers above the floor.

A new access point was also created at this time in the space marked as Room 23 on the plan (Fig. 3). Unlike its predecessors, this was not a stone-built staircase, but a deep vertical shaft inserted between the north wall of the main hall and the north wall of the house proper, which was cut back and partially rebuilt to accommodate a new, shaft-like space (Fig. 28). Regular rows of cavities in its walls and a corresponding brick-and-mortar support indicate the presence of a steep wooden staircase descending from the upper storey above the main hall. At the base of the vertical shaft, an opening was cut through to provide direct access into the main hall. The creation of this new access point to and from the upper floor implies that the staircase in the north-east corner of the tetrastyle courtyard no longer functioned. It was probably blocked at its upper end, which would also have restricted the airflow to the underground cult hall. Any material evidence for this blockage, however, disappeared with the destruction of the upper storey.

Fig. 28.

House of Kybele. Sixth-century access shaft (Room 23). Looking north from Room 12.

A sherd from a fifth-century wine jar recovered from one of the wall cavities provides a material terminus post quem for this intervention. These late alterations suggest a marked increase in the level of secrecy and may be linked to the increasingly hostile actions against pagan cults, even in a city like Aphrodisias, that are attested in the sixth century. Waves of anti-pagan persecution documented in Constantinople and elsewhere — for example, at Sardis — seem to have been intensified by Justinian’s harsh anti-pagan edict of 529.Footnote ²⁹

In spite of imperial legislation, cult activities within the house continued up to the time of its sudden destruction. The most striking aspect of the cult complex is that it was still functional when the House of Kybele, like other buildings in the city centre, was engulfed by fire in the second decade of the seventh century — events that can be linked to a Persian raid of 617.Footnote ³⁰ Inside the house, a destruction date of the second quarter of the seventh century is confirmed independently by the widespread finds of late-sixth- and early-seventh-century ceramics and copper-alloy tableware. The latest coin found on the floor of the main hall (Room 12) in 1986 was minted in 615/16.Footnote ³¹ A second coin minted in 615/16 had fallen from the upper storey above the tetrastyle courtyard (Room 28) and a third from the upper storey in Room 29.Footnote ³² Twelve more Heraclian coins, dated from 611 to 615, were associated with the final occupation phases of the complex.Footnote ³³

The fire that swept through the upper storey of the house must have been a true inferno, fuelled by air drawn in through the ventilation system from the levels below. The coins were partially melted; surviving botanical material is not just burned but has a vitrified appearance; and the marble columns of the courtyard’s upper storey were shattered into dozens of small friable fragments. The wooden ceiling of the courtyard caught fire, and burning debris fell through the opening, damaging the marble piers of the courtyard’s lower storey. The vault of the main hall partly collapsed, and fire consumed all wooden furniture and fittings in the underground complex.

The sudden destruction of the house provides a vivid snapshot of the artefacts that remained in the underground complex in the early-seventh century, at least in the tetrastyle courtyard and the main hall, where undisturbed assemblages were recovered. A series of marble cult figures and a dense body of other artefacts cast a strong light on activities in the house and on its owners. A selective overview of these artefacts is offered here; the marble figures are discussed in the next section.

The assemblage recovered from the tetrastyle courtyard indicates communal dining (Fig. 29). Spread over the surface of the courtyard were dozens of copper-alloy bowls, jugs and cooking pots, including a very large pot with a diameter of 45 cm and a bowl still filled with the vitrified remains of pulses. The vessels so far conserved show these were high-quality objects, with finely chased and repoussé decoration. Other finds include an iron folding stool, three incense burners, broken glass vessels, two finger-rings and a magical gemstone (Fig. 30). The gem is probably second- or third-century in date, mounted in a later silver pendant. Its main side represents a lion-headed figure holding a palm branch with one hand and throwing an offering onto a burning altar with the other, surrounded by magical symbols.Footnote ³⁴ Four lines of similar symbols occupy the other side of the stone. This gem is one of several items imported from Egypt. A soapstone scarab (eighth to sixth century b.c.) was found fallen from the upper-storey peristyle of the courtyard, and a shawabti (seventh to fifth century b.c.) and a green diorite statuette of a female goddess, probably Isis or Maat (seventh to fourth century b.c.), were already known from the older excavations of the north-east wing of the house.Footnote ³⁵ Since such objects are rare finds in late antique contexts, this concentration from one house has strong implications for the interests of its owners (discussed further, below).

Fig. 29.

House of Kybele. Ceramic and metal finds beneath destruction debris in tetrastyle court (Room 28). Looking south-west.

Fig. 30.

House of Kybele. Magical gem-stone found in tetrastyle court (Room 28).

In the main hall, small marble figures remained on display to the end, perhaps placed within the larger niches in the bays of the long sidewalls. Most of their fragments were recovered in two main locations: (a) beneath undisturbed debris at the eastern end of Room 12 (Fig. 31) and (b) from two locations at each end of the schist paved corridor behind the apse (labelled 10W and 10E in Fig. 3), where they had been moved in the years following the fire. The most evocative assemblage was found in the easternmost bay of the north wall: the body of the marble figure called here Kybele A (1) lay against the north wall, where it probably fell from the niche above. Nearby, amidst traces of charred wood, were a ceramic pitcher and two copper-alloy vessels, one badly damaged, the other a ewer complete with lid and a handle in the form of a leaping dog (Fig. 32). These items most likely stood on the wooden shelf set into the brickwork beneath the niche. On the floor within the same bay stood a local table amphora. Other finds from the floor or shelves in the room include further parts of marble figures, another table amphora and cooking pot, a ceramic jar, incense burners (Fig. 33), a glass unguentarium and a second glass vessel.

Fig. 31.

House of Kybele. Marble figure, Kybele A (1), with other finds and undisturbed debris at north wall of main hall (Room 12).

Fig. 32.

House of Kybele. Copper-alloy ewer with lid and handle in form of leaping dog from main hall (Room 12).

Fig. 33.

House of Kybele. Broken ceramic incense burners and candle sticks found in main hall (Room 12) and in schist-paved corridor (Room 10).

At the bottom of the latest access into the main hall (Room 23), a ceramic bowl, a miniature boat-shaped ceramic lamp and a large copper-alloy bronze oil lamp (with its wick preserved) were discovered. Additional finds in the destruction debris came from the upper storeys of the main hall and the tetrastyle courtyard and can be assigned with confidence to the early seventh century.

Period 5: post-destruction

Apart from these assemblages, most spaces within the complex were much disturbed following the fire. A brief discussion of post-destruction activity may clarify the distribution of artefacts from the cult spaces and further informs our understanding of the activities taking place at the time of the fire.

In addition to the spread of finds in the tetrastyle courtyard and the main hall, piles of broken artefacts were excavated at the two extremities of the ground-level corridor behind the apse (Room 10) (Fig. 3). Already in 1986, the excavations had uncovered several important parts of marble figures in the eastern arm of the corridor (3, 6, and 7) (Fig. 3: 10E). They were found with six ceramic lamps, four pitchers, a miniature jug, two miniature jars, amphora sherds, unguentaria and other ceramic vessels.Footnote ³⁶ The new excavations found a pile of similar composition in the north-west corner of the corridor (Fig. 3: 10W): further broken parts of marble figures (3, 5, 7–9) were left here together with another miniature two-handled jar, broken ceramic plates and a large number of incense-burners and candlesticks.

Joining parts of the same marble figures found between the two piles suggests later sorting through of broken material from a common context, and one case — a small fragment of a figure found in 10E that joins a figure from the main hall (1) — makes it clear that these assemblages at 10W and 10E are the remains of material that came originally from the main hall. The original houseowners, or others, aware of the wealth of objects hidden beneath the rubble, returned to salvage them. Debris and objects were moved from the subterranean main hall, where the partly collapsed vault and upper storeys would have posed a real hazard, and were sifted through in the relative safety of the ground-level corridor (Room 10 — its roof had been damaged but remained largely intact). Metal vessels were clearly the main target. The plentiful vessels in the tetrastyle courtyard were under a heavy collapse and could not be reached; and some in the main hall also remained under collapsed debris to be found in excavation. Such valuable items are absent from the retrieval piles at 10W and 10E.

These collecting and retrieval activities were probably cut short by an earthquake that struck Aphrodisias around or shortly after 620.Footnote ³⁷ Already badly damaged sections of the upper storey of the complex collapsed onto the burnt debris, sealing and protecting them from further retrieval work. By contrast, rooms that had survived the disasters of the early seventh century more intact — certainly Rooms 21 and 22 — were reoccupied in subsequent decades, leading to the near-total loss of their original furnishings (Fig. 14).

The ruins of the house continued to serve as a quarry for building materials for centuries. The brick vaults of the tetrastyle courtyard, for example, were dismantled probably no earlier than the twelfth century. Throughout the house, deposits of discarded materials have been found, and walls from which the brick facing had been stripped are still visible in their bare mudbrick cores. Eventually, a medieval settlement was established on top of the remains of the house, beginning in the second half of the fourteenth century. The wall foundations of some of these new houses incorporated building materials from the ruined House of Kybele beneath, including an unusual aniconic bust in the form of a mountain (14), one of several that belonged somewhere in the underground complex.

1. Kybele A (Figs 35–36) Footnote ³⁹

Three of the best-preserved figures represent Kybele. The goddess is immediately recognisable in the first figure, Kybele A, by her crenellated mural crown and the lion at her feet. She sat on a thick cushion on a four-legged chair, and wears a thin, ankle-length chiton, tied tight under the breast with a tubular or rolled belt. A himation falls over one shoulder and covers lap and lower legs. The feet wear traditional sandals that rest on the back of the small crouching and visibly growling lion. Kybele’s lions more normally stand at her side.Footnote ⁴⁰ Here a footstool has been, as it were, pushed to one side, to allow the lion to take on its function as foot-rest. Kybele also had chthonic powers, represented here by a prominent snake wound around the right-hand leg of her seat.

Fig. 35.

House of Kybele. Marble figure, Kybele A (1).

Fig. 36.

a–b. House of Kybele. Marble figure, Kybele A (1), details.

Further attributes specify other aspects of Kybele’s powers. Perhaps most striking, the goddess wears a heavy imperial-style diadem that bears traces of gilding (Fig. 36a). It is represented as a linked jewel diadem with a central emblema rising above the forehead. This is the kind of diadem worn by imperial figures from the mid-fourth century.Footnote ⁴¹ It signified sole imperial–royal power, and here becomes a contemporary political metaphor for universal divine power.Footnote ⁴²

In her left hand, Kybele held a double cornucopia, carved in the round, until it joined a break at the shoulder. In her right hand, she holds a deep libation bowl beneath which there is perhaps an incense burner — similar in form to ceramic examples found in the house (Fig. 33). It was common for classical divine figures to be shown with libation bowls. They were not venerating themselves, but acting out their divine character. Also in her right hand, Kybele held a tall sceptre, now mostly broken, that rose from the plinth, supported by struts, to a large strut that emerged from the side of her head. The sceptre, too, was a traditional symbol of regal power in the civil realm, often carried by senior gods.

2. Kybele B (Fig. 37) Footnote ⁴³

The goddess is identified by the remains of a lion (front paw and back legs) seated on her lap on its hindquarters. She sits on a rocky mass and wears a cloak and long chiton to the feet that forms patterned swags between the lower legs. The chiton is tied under the breast by an undulating belt in the form of a snake. Although the breast is unformed, this is clearly a female figure.Footnote ⁴⁴ The cloak is fastened in the centre of the chest with a small round fibula, and instead of the classical sandals worn by Kybele A, the goddess wore the closed shoes of a contemporary citizen woman — both features of portrait busts and statues — used to increase the reality effect of the figure.Footnote ⁴⁵ They show the goddess wearing clothes and shoes of now.

Fig. 37.

a–b. House of Kybele. Marble figure, Kybele B (2), front and left side.

Kybele’s left hand held a thick sceptre, part of which is visible in relief against the mountain seat. It reached above the shoulder where it was supported by a large round-section strut. In her right hand, the goddess probably held either a libation bowl or a globe. The natural, rocky character of the throne represents Kybele as ‘mother of the mountains’ (mētēr oreia).Footnote ⁴⁶ The deity is seated in her own essence.

3. Kybele C (Figs 38–39) Footnote ⁴⁷

The high-quality figure was made in the same workshop as Kybele A and B, but a different sculptor carved the head. It is of considerable power and distinctive late antique effect. The goddess sits on a plain stool or chair, both forearms extended in front carrying important attributes.Footnote ⁴⁸ She wears a thin, sleeved chiton belted high under small breasts and a long cloak veiled over the top of her head. The chiton has an emphatic neckline and sleeves to the wrist, a widespread late antique costume fashion. The design and style of the chiton connect the figure closely to Kybele A and B.

Fig. 38.

a–b. House of Kybele. Marble figure, Kybele C (3), front and detail.

Fig. 39.

House of Kybele. Marble figure, Kybele C (3).

In contrast to Kybele A, the goddess has a plump face, full cheeks, long nose and arched brows above wide-open eyes with drilled ‘black’ pupils. The centre-parted hair forms a thick wreath around the face. The rounded face may channel contemporary female ideals seen in marble portraits.Footnote ⁴⁹ This Kybele also has a different accumulation of power attributes, both new and traditional. She wears her normal mural crown, that ‘grows’ through or out of the veiled cloak on top of her head. The crown was tall and narrow, and had a carefully marked wall-and-tower design. Over the front of her head, she wears a tall stephanē, a traditional attribute signalling female divinity. The missing attribute held in her left hand was most likely a cornucopia, attached at the break surface on the shoulder.

The recently discovered right lower arm and hand held a globe in its upturned palm. The thumb holding the globe at the side is inhumanly long, thin, and typical of the period. The globe is divided into quadrants by four vertical straps held by and meeting under a small boss at its top. The globe had been carried by god-like emperors since the early empire. In the late empire, it can be topped by a Victory, as carried by Honorius and Roma on diptychs, or by a cross, for example, carried by an archangel.Footnote ⁵⁰ In detailed representations it is often clear that the celestial globe of the heavens is meant, rather than the terrestrial globe; and in such cases, instead of a plain sphere, the globe is divided into segments decorated with stars (so on the Roma diptych).Footnote ⁵¹ Clear parallels for our figure carrying a segmented globe without an added element on top can be seen on the silver missorium of Theodosius, where the two seated, junior emperors each carry a segmented globe with stars.Footnote ⁵² Most likely, then, our Kybele carries a celestial globe of universal rule: she was thus marked as the truly all-powerful Great Mother Goddess by a well-known imperial symbol transferred to the divine sphere. Stars might have been added to the globe in paint. Cornucopia and celestial globe are signs of universal power that will bring wealth and prosperity to Kybele’s worshippers.

The three new Kybele figures presented here add to her late antique profile as an all-powerful divine mother. Though not necessarily shared by all, the ideas in Julian’s hymn to the goddess give a strong sense of Kybele as a pantheistic deity.Footnote ⁵³ She had a strong Athenian aspect, as well as her aspect as a goddess of the mountains in Asia. She had important mysteries and rites of initiation and purification.Footnote ⁵⁴ For Julian, she is the goddess that controls every form of life. Kybele was also a major divinity among Athenian religious intellectuals of the later fifth and early sixth centuries, whom we will see in close interaction with their Aphrodisian colleagues.Footnote ⁵⁵

4. Athena (Fig. 40) Footnote ⁵⁶

The helmeted head of Athena, turning in a dynamic posture, was broken from a full figure that once held in its left hand a tall spear supported by a thick round strut, now broken, projecting from the side of the head. The helmet is a drastically reduced version of the closed ‘Corinthian’-type helmet, pushed back from the face, as often worn by Athena. It sits on top of the hair like a small crested cap, and its distant descent from a classical Corinthian helmet is most easily understood in the two vestigial eye-holes near the front of the helmet. The crest rises from a holder, placed as if the helmet would never be worn over the face, and flows over the neck behind. The front of the crest, now broken off, projected forwards. The tall narrow face has distinctively late antique features, such as the slim nose and eyes drilled with small, round pupils and covered by wide upper eyelids. In terms of its technical handling and effect, the Athena comes from the same hand as Kybele A.

Fig. 40.

a–b. House of Kybele. Marble head of Athena (4).

In the fifth and sixth century, the goddess Athena often stood for the city of Athens and its philosophical activity. In this period, we will see pagan intellectuals at Aphrodisias closely connected to like-minded figures in Athens seeking a platonist route to the divine. When her cult statue was removed from the Parthenon, Athena was given a new home in the house of the philosopher Proklos, where she probably inhabited a small cult figure, similar to the one presented here.Footnote ⁵⁷ One of Proklos’s surviving hymns (5) is an exalted address to Athena. In the House of Kybele, the marble figure of Athena probably resonated with these associations.

5. Aphrodite (Fig. 41) Footnote ⁵⁸

The goddess is recognised by her provocative nakedness, more framed and highlighted than concealed by her garment. She stood with one knee forward and thighs pressed modestly together. Feet and ankles are somewhat formless, toes carved flatly in low relief on the plinth. The goddess held her dress out to one side with her left hand, of which three long and formless fingers are preserved, visible only in profile view (Fig. 41b). That is, the hand was probably not visible when the figure was in its planned setting, for example, in a small niche. In front view, the dress forms a kind of mannered hanging of repeated swags behind the figure on the (viewer’s) right. On the viewer’s left, the dress ‘flies’ out to the side, in corrugated horizontal ridges with little reference to observed reality. The Asklepios found in the house (below, 6) exhibits similar abstract dress patterns and poor attention to feet.

Aphrodite would be a likely and welcome participant in any gathering of pagan gods at Aphrodisias. A contemporary author describes the extreme effect that an Aphrodite statue had on him: he sweats, feels terror, astonishment (ekplēxis) and joy (euphrosynē) — and he is unable to keep away.Footnote ⁵⁹ This response may be only that of one viewer, but it is a precious sympathetic reaction to a pagan figure in the years around 500 that we otherwise lack. The perhaps startling intensity of the visual experience described is something few in the modern world would expect.

Fig. 41.

a–b. House of Kybele. Marble figure of Aphrodite (5), front and left side.

6–9. Asklepios group (Figs 42–45)

The following four pieces (6–9) were most likely parts of a single group, made up of Asklepios (6), Hygieia (7), Telesphoros (8), and a base with unusual covert symbols (9). Parts of the group were found in all three main locations: in the underground hall (Room 12) and in the two retrieval dumps in different parts of the apse walkway (10W and 10E).

The well-preserved figure of Asklepios (6) is recognised by its posture and the snake wound round its support (Fig. 42).Footnote ⁶⁰ The god stands with one leg pushed forward, one hand on hip (at the attachment break), the other arm bent sharply at the elbow to hold a thick scroll in front of his chest (Fig. 43). This swagger pose was a loose reception of a widely known Asklepios figure type.Footnote ⁶¹ In Late Antiquity, the closest complete version of the basic scheme is seen on the Asklepios and Hygiea diptych in LiverpoolFootnote ⁶² and in a small marble figure from Epidauros dedicated there by a priest of Asklepios in the early fourth century.Footnote ⁶³ The pose, clothing design, large head and thick hairstyle in the Aphrodisian Asklepios are clear reminiscences of the same statue scheme. The god was leaning on a tall staff whose path can be followed from the plinth to an attachment point under the elbow. He wears his usual citizen sandals, but the feet are worked in a summary manner. By contemporary instinct and without dress logic, the sculptor has turned the himation that falls over the left arm into a tight, wrist-length late-antique-style sleeve of a rather formless character. The Asklepios is close to the Kybele and Aphrodite figures in the handling of the body and clothes. The carving of the head, beard, and long hair, however, is different. They have a rough, matte surface that contrasts with the high polish of the body.

Fig. 42.

House of Kybele. Marble group of Asklepios and Hygieia(?) (6–9).

Fig. 43.

a–b. House of Kybele. Marble figure of Asklepios (6), details, front and left side.

Fig. 44.

a–b. House of Kybele. Marble head and feet of Telesphoros (8).

Fig. 45.

a–b. House of Kybele. Marble base of Asklepios group (9).

A figure of Hygieia (7), Asklepios’ daughter, often represented with him, most likely stood in the second cutting in the base. Surviving fragments of a plinth and tree support may belong to this figure.Footnote ⁶⁴ It was made from black-and-white marble, that is, from a quarry bed with adjoining black and white marble seams, so far documented only at the Göktepe quarries, some 40 km to the southwest of Aphrodisias.Footnote ⁶⁵

Telesphoros, the young son of Asklepios, can be restored at the proper right side of the god from two clear indications (Fig. 44): a broken strut at thigh level was intended to support such a figure, and the outline of its tiny feet, as usual set close together in a flat-footed stance, are preserved on the plinth. A small head (8) wearing a thick, hooded cowl over its head, was probably part of this Telesphoros.Footnote ⁶⁶ He is frequently represented with Asklepios and Hygieia on coins, gems and reliefs, always as a diminutive figure wearing a cowl of the same kind.Footnote ⁶⁷ In spite of his small scale and subservient posture, Telesphoros was an important part of Asklepios groups and an important deity in his own right. He appears on more than 290 coin types in Asia Minor, including more than twenty at Pergamon, where he had a major cult about which there is much in the writings of Aelius Aristeides.Footnote ⁶⁸

The base (9) was made to carry two figures set into bevelled cuttings (Fig. 45).Footnote ⁶⁹ Ten divine symbols are engraved on its front surface. Although puzzling to read as a connected utterance, their individual points of reference can for the most part be deduced. At the left end, a plain oval shape is perhaps an egg alluding to fertile generation — and sometimes an attribute of Asklepios.Footnote ⁷⁰ Next, a rectangular element carries what is most likely the double-headed axe (labrys) of Carian Zeus. The god’s local form, Zeus Nineudios, had been a popular deity at Aphrodisias since the late Hellenistic period.Footnote ⁷¹ Next, the two short stems topped by pine cones are Dionysian attributes, thyrsoi, to be understood in relation to the last symbol on the extreme right — a priestly headband decorated with four ivy tendrils, also a Dionysian attribute. Between, two snakes frame three central symbols: a crescent moon and a spoked sun stand for Selene and Helios, gods in their own right.Footnote ⁷² Sun and moon symbols on stones could have magical properties at this period,Footnote ⁷³ but their usual reference is to the cycle of day and night and the eternal passage of time.Footnote ⁷⁴ Here they should refer, at least, to the unending worship of Asklepios and Hygieia. The remaining symbol, to the left of the sun, looks at first like a comet, but comets normally have one trailing tail, not, as here, two fluttering tails. The symbol can better be taken as a headband with two free-hanging ties. It is possible, furthermore, to see that one of the streaming bands, though abraded, had a fringed end. This was then a Hellenistic-style royal diadem,Footnote ⁷⁵ and, like the imperial diadem of Kybele A and the celestial globe of Kybele C, it has the transferred or metaphorical meaning of supreme royal power.Footnote ⁷⁶

Symbola of the kind engraved on the base were thought to play an important part in this period in describing the indescribable divine.Footnote ⁷⁷ Some here allude to favoured deities — Asklepios (egg), Dionysos (thyrsoi, diadem, ivy headband), Carian Zeus (labrys), Helios (sun). The central group of diadem, sun and moon refers to the everlasting dominion of Asklepios who stood on the base above, a still potent and responsive god in this period. He appears in evocative texts of the fifth and early sixth century when the best doctors were thought to be religious thinkers and philosophers, and when his statues retained their vital power to dispense correct and effective medical cures.Footnote ⁷⁸

10. Apollo (Fig. 46) Footnote ⁷⁹

The small, roughly carved figure was made from the same Göktepe black-and-white marble as 8. Unlike other figures in this material, there was little attempt to employ the two colours in a logical manner.Footnote ⁸⁰ Body and plinth are carved from the black seam, and only head and right arm from the white seam. The figure is easily recognised as a loose reception of a figure scheme known in many media,Footnote ⁸¹ and called the Apollo Lykeios after a literary account of a statue in the Athenian Lykeion that mentions two identifying features: one arm leaning on a support, the other arm resting on top of the head.Footnote ⁸² The Aphrodisian figure has both features: left arm on a support, and the right arm on top of the head. The support is composed of the coiled snake itself, while other versions prefer a snake around a tree or coiled through a tripod as support.Footnote ⁸³

Fig. 46.

a–b. House of Kybele. Black-and-white marble figure of Apollo (10), front and left side.

Such a roughly carved figure would usually be labelled ‘unfinished’, as if it were a workshop piece awaiting further work. But if we have this artefact, excavated in a functioning complex, it is because it entered service, as it was, as an effective divine figure. We may find it surprising, but many ancient artefacts were put into use in varied states of incompleteness. In this case, if needed, we may imagine possible fifth- or sixth-century narratives. For example, the figure, ‘unfinished’ but smooth and worn, may have been carried around by its owner for protection, a much-loved personal talisman, and later been donated as a votive in a niche in the underground complex by its grateful owner when invited there as a guest. The other figures (1–9) have the aspect of cult figures that would receive prayers and veneration; this figure seems more like a treasured personal votive.

Apollo had contemporary and local resonance at an ancient cult site in the Meander valley, in a cave known as ‘Apollo’s Halls’ (Apollōnos Aulai), which we hear about when visited by the Aphrodisias-based philosopher Asklepiodotos on one of his religious field trips. Locally, this would perhaps have been the Apollo cult best known to fifth-century Aphrodisian pagans.Footnote ⁸⁴

11–12. Goddess busts A and B (Figs 47–8)

Puzzling busts, unusual and difficult to assess, come from the house but from higher (later) layers above Room 22. Two are fragmentary busts of a female divinity — ‘life-size’, close in scale, both now headless and fragmentary, and probably of the same deity.Footnote ⁸⁵ In origin, they had the same form, that is, they are hollowed out behind to turn them into the distinct category of ‘bust’. Busts had begun as a portrait format, but were soon also used for gods, with the meaning of ‘portrait’ of the divinity in question. The hollowing-out of the shoulders and chest from the back was a defining part of this curious format. Both busts were carved with a smooth lower bearing surface on which they could stand independently. They may have been displayed in the large niches in Room 22. The rough point-finish on the back of both busts and their complete lack of a central supporting ‘spine’ indicate a late antique date for both.Footnote ⁸⁶

Fig. 47.

House of Kybele. Marble bust, Goddess Bust A (11).

Fig. 48.

House of Kybele. Marble bust, Goddess Bust B (12), front and back.

The outer chest surface of both busts was carved with two or three rows of quite roughly chiselled conical mounds. They are insufficiently articulated to tell for sure, but by analogy with those of Kybele A and C, they most likely represent youthful female breasts. Perhaps closest are the rows of round bun-like elements on the apron of a small Hellenistic marble figure of Kybele, originally from Pergamon, now in Vienna, on whose base the goddess is named ‘Pergamenian Mother of the Gods’.Footnote ⁸⁷ Kybele, with emphasis on another aspect of her spheres of action (Great Mother), is then a possible, even likely identification for both busts.

Bust B preserves more of the ‘breasts’ than Bust A — in three rows and more widely spaced. Sometime later, the bust underwent three interventions. (1) A rectangular slot or notch was cut in relief from the lower border in the central fragment, shaving off part of a ‘breast’ to its right; the notch was not cut through to the back. (2) A large aperture was carved through the wall of the bust on the (viewer’s) right side, through to the void behind. This operation vertically bisected one of the surviving ‘breasts’. And (3) three ‘breasts’ below and to the sides, though not directly affected by the cutting of the aperture, were bisected horizontally. On the flat upper surface of two of the bisected ‘breasts’, small iron pins secured some kind of separately attached metal(?) tray, positioned directly under the aperture.

These interventions are without parallels but were perhaps connected to ritual use of the bust. Only guesses can be made. The tray(?) might have carried incense or other offerings to be passed through the aperture by the faithful to the divine beyond, ‘inside’ the goddess, with the probable aim of animating the deity within the bust. It seems that some kind of theurgic operation was being carried out in the veneration of this goddess.Footnote ⁸⁸

13–15. Mountain busts A–C (Figs 49–50)

The last marble finds from the house presented here are also difficult to parallel: three monumental, aniconic busts carved in the form of rocky mountains.Footnote ⁸⁹ They were possibly further representations of Kybele in her role as mother of mountains. For now, it remains unclear to which room they originally belonged.Footnote ⁹⁰

Fig. 49.

House of Kybele. Aniconic marble bust, Mountain Bust A (13).

Fig. 50.

a–b. House of Kybele. Three aniconic marble busts, Mountain Busts A–C (13–15), front and back.

All three objects are close in dimensions, technique and handling of the ‘subject’ matter. Their main carved features include: (1) a fully finished representation of a rocky surface, known from other contexts to represent the surface of a mountain or rocky seat;Footnote ⁹¹ (2) a flat, prepared surface, lightly corrugated with a claw-chisel and drilled with five-six attachment holes near the lower edge and a smaller dowel hole above;Footnote ⁹² and (3) behind, a rough, hollowed-out surface, point-chisel-finished inside a broad frame.

It is important to establish first how these objects were displayed, lying flat or standing up. At first it might seem that they should lie ‘flat’, with the hollow rear facing down and the prepared surface facing up. There are, however, good reasons for thinking the second possibility is the correct one: that is, the objects stood upright (as in Fig. 50), with the hollowed-out part as their back, and the prepared surface at their lower edge. There are two separate arguments. Firstly, the hollowing-out can have had little to do with reducing weight. The depth and area of stone removed would barely change the overall weight of the objects. Secondly, the five-six attachment holes are arranged in a shallow ‘swag’ formation that only makes sense if the prepared surface was vertical and the holes were for the attachment of some element that hung belong. The objects therefore stood upright, and the backs were hollowed out to turn them conceptually into the prestigious ‘portrait’ format of a bust. The hollowing-out, as often in late antique busts, is shallow, roughly executed and without spine, but it was sufficient to turn a carved, rocky-looking boulder into an aniconic ‘portrait’ bust of a mountain deity.

All three busts have prepared flat surfaces for the attachment of some subsidiary component or image. We may imagine, for example, a metal relief with repoussé design, perhaps of the divinity concerned, shown in a shrine or in a cave following the curved upper line of the attachment surface. Such a relief could have been of silver or gilded bronze, and was probably fixed by the small dowel hole in the upper part of the attachment surface (diam. 0.50 cm, depth 1.50 cm) and with glue for which the corrugated surface provided grip. The larger, deeper holes (diam. 1.50 cm, depth 3.5 cm) across the lowest part of the bust might have anchored both the thin metal relief and something more precious, such as a gold necklace or wreath. This element in all three busts was also anchored at both sides beyond the attachment surface, in the adjacent ‘rocks’.

The worship, procession, and rituals of Kybele as a plain black stone went back to the second century b.c.,Footnote ⁹³ and aniconic stones were of considerable interest to pagan thinkers of the late fifth and early sixth century. They discuss baityloi of various deities, and we hear of a large, black, unshaped (atupōtatos) cult stone at Petra that received sacrifices as Ares.Footnote ⁹⁴ Such plain stones are rather different, however, from the new mountain representations in bust form discussed here. Aniconic mountain gods are not otherwise known in ‘life-size’ marble busts, but they are frequent on city coins of the Roman east, where they represent famous local landmarks — for example, at Rhosos in Cilicia or Philippopolis in Thrace (both under Commodus).Footnote ⁹⁵ At Philippopolis a small shrine of a local deity is embedded into the face of the mountain in a way that helps to imagine the kind of relief that could have been added to the prepared surface on the marble busts.

The richest, most varied set of such mountain images is found on coins of Caesarea in Cappadocia in the later second and early third centuries.Footnote ⁹⁶ The sacred agalma of Mt Argaios appears, essentially as a stand-alone bust, in the following settings: worn on the mural crown of the city’s Tyche;Footnote ⁹⁷ set on top of an altar;Footnote ⁹⁸ enshrined in a columned naiskos;Footnote ⁹⁹ carried in sacred procession on top of a cart pulled by elephants;Footnote ¹⁰⁰ carried on the extended hand of a local goddess;Footnote ¹⁰¹ or carried in the crook of Serapis’ arm.Footnote ¹⁰² In all these coin types, the image of Mt Argaios is treated as an aniconic bust of ‘life-size’ scale that can be mounted, enshrined, manipulated, and handled in a variety of ways. Such a proliferation of mountain busts is helpful background. In the House of Kybele, similar busts of a local mountain god or of Kybele the Mountain Mother displayed in the same setting would be possible, especially if variations, perhaps addressing different aspects of the deity, were introduced in the added metal relief images.