The Theatre at Kourion

The theatre lies on the south-eastern side of Kourion, cut into the ground where the plateau breaks toward the sea. What stands up today is mostly the lower bowl, rebuilt between 1961 and 1963 “up to the diazoma” in nineteen rows, with a back wall in very poor condition “completely repaired and restored” (Karageorghis 1963; 1964). Rebuilding the semicircular wall round the auditorium, planned for 1964, was put off because of that year’s troubles (Karageorghis 1965); we know of no report that it was resumed. The new blocks were cut from locally quarried limestone, “as had the original builders” (Soren & James 1988, p. 233). Colour therefore cannot separate modern from ancient. The working can: the blocks of the 1960s have crisp edges and regular joints, while the ancient ones are worn, irregular and, in the lowest rows, often “very rough and almost unshaped” (Stillwell 1961, p. 53).

Above the diazoma nothing was rebuilt. The podium wall that carried the upper seats survives as a moulded socle and pieces of its capping, and no seat of the upper section was found in place (pp. 52–53). The ring corridor and the outer wall behind it are robbed down, though two outer doorways stood “for about their original height” (p. 44). The stage building is at foundation level, and a modern timber deck covers the area of the ancient stage.

A laser survey made for an acoustic study in 2011 counted 17 steps on one line up the restored cavea, with treads of roughly 0.72 to 0.77 m (Economou & Charalampous 2013). That fits neither the nineteen rows of the reports nor Stillwell’s ancient run of 0.65 m; the model shows the bowl as it stands. The “thirteen rows” said in some accounts to have been rebuilt at Kourion belong to Soli, where the same annual report records thirteen new rows (Karageorghis 1964). A sign on the site credits the work to “Antonis” Travlos. His name was Ioannis, John in English.

Stillwell described the building by its own compass, with the cavea “north” and the stage “south”. A footnote gives the azimuth of the axis as 33° (p. 39 n. 4), and modern mapping and imagery put it between 36° and 40°; the model takes 36°. Either way the audience faced south-west, about 216°, although many descriptions say south. The nearest point of the coast lies almost exactly on that line, a little over half a kilometre away.

The theatre is “carved into the north side of a gully” (Soren & James 1988, p. 230). By our reckoning the orchestra lies about 62 m above the sea and the plateau behind the top of the cavea at about 70 m, a level fixed by the surveyed thresholds of the Earthquake House, a few metres away. The centre of the bowl is rock: Stillwell found cut bedding lines for the seats “near the center line of the cavea”. The wings rest on fill stiffened by rubble walls, one of them “nearly five meters” deep (p. 53). The east side sits lowest, on ground falling toward the bluff, and its foundations gave trouble. East and north-east of the theatre, on the higher ground (guidebooks sometimes put it to the west), lies the Annex of Eustolios, whose mosaic inscriptions belong to the late 4th or 5th century (Mitford 1971, nos. 201–204).

The sea is what people come for, and in antiquity most of the audience could not see it. The ground falls about 60 m to the shore over half a kilometre, so the water is nowhere near the foot of the stage, and in the Roman theatre a stage building some 12 m high stood between the seats and the view. Sight-lines drawn on the restored section show that only eyes in the top four rows of the upper cavea, and in the colonnade above them, cleared the crown of the stage front. From the highest seat the line of sight meets the water nearly two kilometres out: open sea and horizon, with the shore hidden. Everyone lower down looked at architecture. The view from every seat today exists because the stage building fell.

No part of the first theatre stands above ground. Stillwell assembled it from buried evidence (p. 66): stretches of an earlier outer wall inside the later masonry, a rock cutting for an earlier east entrance, seat blocks re-used in an early Roman wall, six half-columns built into the later stage foundations, the fill under the seats, and coins. In 1935 J. F. Daniel noted in the fill under the diazoma lamps, stamped black-glazed pottery and “4th century B.C. tiles. Nothing later” (p. 66). The earliest coins are Ptolemaic, among them a Ptolemy IX from rock level (pp. 62, 66). Because most of the Hellenistic coins from the area are of the 2nd century BC, Stillwell dated the theatre to “the latter part of the century, rather than the earlier” (p. 77). Nothing found carries it back into the Classical period (p. 43).

From this he drew a theatre on the same centre and axis as the later one, but smaller. Its outer wall ran on a radius of “only about 26 m” (p. 41). The orchestra was a full circle about 13.7 m across, cut in the rock at about the level the later arena would return to, some 0.7 m below the Roman floor. No Hellenistic orchestra surface survived, so the circle is his restoration: he added “three more rows of seats at the foot of the slope” below the lowest Roman row, finding nothing to suggest “a material change of pitch” (p. 66). The horns of the cavea ran “around beyond the half circle” (p. 68), to about 205° or 210°, and ended against oblique retaining walls, so that the side passages were “angled inward in the normal Greek manner” (p. 43). There was probably no diazoma.

The stage is the best evidenced part. The six re-used half-columns are Doric and polygonal, 0.32 m across at the foot, with a tongue behind for keying into a wall or holding the frames of painted panels. Two drums of 0.967 m and a capital give a column of about 2.09 m; add an entablature a quarter of that and the order stood “a little over 2.60 m” (pp. 67, 76). That is a high proskenion, the colonnaded platform of the later Hellenistic theatre, as at Priene. The building behind it has left almost nothing: “The first, most serious lack, is that of a recognizable skene” (p. 66). The model takes the proskenion’s width of about 12 m from Stillwell’s phase plan; the skene’s two storeys are a guess.

With its horns running past the half circle, Stillwell thought this theatre held “about the same” number as the enlarged Roman one (p. 68); our arithmetic gives about 3,700 at 40 cm a person. It belongs to Ptolemaic Cyprus. The building a visitor walks through is Roman, and calling the monument a Greek theatre names it after a phase that lies buried under it. Greco-Roman is fair.

Stillwell’s second period is a remodelling in the Roman manner, and his date for it, “after 15 B.C.”, is a guess hung on an earthquake. Cassius Dio (54.23.7) records a severe one in that year, after which Augustus came to the help of Paphos, “and though there is no mention of Kourion it is possible that sufficient damage was done ... to require some restoration” (p. 77). The works themselves are excavated. The horns were cut back until the cavea was a half circle, and the side passages were rebuilt straight and parallel to the stage. At least two seat blocks, “presumably from the ‘horns’ of the koilon”, went into the new north wall of the east parodos, and the passages were “most probably” left open to the sky (pp. 49, 77). The theatre “would have contained fewer persons” (p. 68). The orchestra was raised to about the later Roman level and the stage building lengthened to the orchestra’s edges (p. 67). The old proskenion was kept and refinished in “hard stucco in two coats”, “a deep Pompeian red”, and its painted panels were probably replaced by walls with doors (p. 67).

Nero’s share is three blocks. They come from one limestone entablature, architrave, frieze and cornice cut together in blocks 0.52 m high, carrying a bilingual text, Latin on the frieze and Greek on the architrave, that names the proconsul Q. Julius Cordus (Mitford 1971, no. 107). Mitford dated it to AD 64/65 and restored it, as an example only, as recording the building of the theatre for the city of the Kourians by Nero through Cordus. Stillwell read the modest scale of the blocks as “a façade in keeping with the restricted size of the theater”, and concluded that “all that Cordus did was to rebuild the stage” (p. 77). No plan, height or door count survives for that façade. In the model it is an optional front of one storey of columns, built round the inscribed blocks and marked speculative; switched off, the red proskenion shows instead.

Some popular accounts add an earthquake of AD 77 that damaged the theatre and led to Trajan’s rebuilding. We know of no evidence for it at the theatre. Stillwell, who ties each of his eight periods to its evidence, has no such event, and one of the popular books that repeats the date attaches it to the sanctuary of Apollo instead.

The fourth period made the building visitors see: “the enlargement of the Theater with the corridor and buttress walls, as well as the complete rebuilding of the parodoi, and the construction of the massive new stage building with its architecture still in limestone” (p. 77). Stillwell dated it AD 102–117. The evidence is circumstantial but consistent: a burst of building at Kourion under Trajan, above all at the sanctuary of Apollo; an inscription recording construction paid for from funds sanctioned by a proconsul, dated by its letters to the late 1st or early 2nd century (Mitford 1971, no. 116); and lamps of 1st-century types in the footing trenches of the new corridor and buttresses (p. 77).

The builders kept the old cavea and wrapped it. Against its outer wall they built a facing 1.50 m thick, then a vaulted corridor 1.90 m wide, then a new outer wall 1.40 m thick: “Thus the radius of the building was extended 4.80 m” (p. 45), to about 30.7 m, an overall width of some 62 m. The corridor ran round the back of the cavea and climbed as it went, “more than seven meters” above the paving of the east parodos at its highest, in flights with risers of about 0.166 m set parallel to one another (p. 45). Five doorways 1.60 m wide led in from the streets (pp. 43–44). From the corridor five vaulted passages, the vomitoria, each about 5 m long and 1.65 m wide, opened onto the diazoma (p. 51).

The east side went wrong. The new wall there stood on “shockingly scamped” foundations over falling ground, the vault came high, and the wall leaned. Three buttresses went up against the south-east part of the wall “soon after the enlargement ... probably while it was still under construction”, with a fourth, 2.10 m thick, at the south-west angle, pierced by an archway onto an outer stair (pp. 43–44, 47–48). On the west the outer wall runs straight for about 7 m and then curves sharply, and Stillwell concluded that “the masons were working, for a time at least, without proper supervision” (p. 56 n.).

The parodoi were rebuilt as vaulted tunnels with seats over them. The evidence is voussoirs from a vault of about 1.68 m radius (p. 76) and a single block in the east parodos, 12 m in from the corner, whose top is worked to a slope so that the vault “slanted downward to fit below the rows of seats above it” (p. 49).

Rough totals computed from the model’s dimensions give the scale of the work. The enlargement and its stage building come to about 6,200 m³ of dressed stone and rubble core. The seats alone, about 430 m³, would have needed something like 5,000 to 15,000 person-days to quarry and shape, at the 19th-century handbook rates that archaeologists apply to ancient building. These are estimates, and they count quarrying and shaping only: they leave out haulage (the quarry is unknown, though the stone is local), lifting, mortar, carpentry and everything to do with marble.

Stillwell measured the seats that survive and calculated the rest. Each block is laid with its long side toward the orchestra and is normally 0.55 m wide. Each row is cut back below its front edge at 67° to 70°, under a nosing 9 or 10 cm high; there is no separate footrest. The run from the front of one row to the next was, “as nearly as can be calculated”, 0.65 m (p. 53). No regular foundations were laid: on the rock the blocks sit in cut beds, on the wings directly on the fill.

Stairs 0.70 m wide divided the lower cavea into five wedges, with no stair on the axis (pp. 53–54); the steps that survive all belong to the last period of use. Stillwell gives no count of rows for the lower cavea. His restored plan draws a field that holds nineteen at his run of 0.65 m, the number the restorers rebuilt, and the model uses nineteen.

The height of a row is a real difficulty. The blocks Stillwell measured average 0.35 to 0.36 m high, and at that rise nineteen rows reach only about 6.75 m above the orchestra. His restored section, the spot heights in the corridor behind, the three steps of Vomitorium III and the levels of the later arena all put the diazoma at about 8.2 m. A rise of about 0.43 m satisfies every one of those checks, and the model uses it. Nearly all the measured blocks are late re-sets, so either the first seats were taller or the late ones sat on bedding courses. The question stays open.

Behind the diazoma rose the podium, a wall with a moulded base and capping. The socle and first course survive to 0.75 m and the capping course is 0.35 m high, but 1.10 m in all is too low for the vomitoria, whose ceilings had to pass beneath it, so Stillwell restored it to “something over two meters” (p. 52). The model uses 2.20 m. On it ran a gangway and “only six rows” of upper seats (p. 52). No upper seat or stair survives in place (p. 53).

Of the “massive new stage building” only the foundations stayed where they were put. In its final form it measured 31.10 m long outside and 9.50 m deep including the stage, and from the north face of the outer wall to the south face of its back wall is 44 m (pp. 41, 68). About 4 m at each end was taken by a stair tower, “as at Aspendos”, and the flanks returned forward to meet the parodoi, with passages through them onto the stage (p. 68). South of the back wall the ground falls away, and Stillwell restores a long gallery whose floor stood “over four meters above the present remains of the back wall”, probably over a cellar (p. 68).

The stage itself, the pulpitum, was “rather narrow” (p. 41): about 5.4 m from its front wall to the stage front. Stillwell’s restoration “shows a normal Vitruvian height for the pulpitum, but an alternate version is also outlined” (p. 68). He gives no figure for either. His section puts the floor about 1.3 m above the orchestra, in keeping with Vitruvius, who sets a Roman stage at no more than five feet (V.6.2). The model uses 1.25 m and offers, as a speculative option, a high stage of 3.00 m on the Greek pattern.

The first stage front, the scaenae frons, was limestone. Stillwell calls it “the straight, Asiatic, or Eastern type” with no curved or rectangular recesses (p. 41), and connects limestone Attic-Ionic column bases, for shafts about 0.47 m across, with “a limestone order for the scaena frons ... built in the early years of the second century” (p. 73). Whether this limestone front already had its niches and its second storey is uncertain; the model gives it both.

Toward the end of the 2nd century the limestone front was replaced in marble. “The back wall was retained, but the parts visible from the cavea were revetted in marble” (p. 77). What survives is a Corinthian capital 0.48 m high, spiral-fluted shafts (one preserved to 1.87 m, 0.43 m across at the foot), a few marble entablature blocks and a niche block with the bronze pin that held its revetment. From parallels and the style of the capital Stillwell put the marble “late in the second century, or even in the early years of the third” (p. 73). The marble shafts stood on the old limestone bases, which he calls unusual.

For this period he is firm: “It is certain that there were two storeys of columns” (p. 68). The upper shafts were slimmer, 0.26 to 0.27 m at the top (p. 73), and the entablature was broken forward over groups of columns, projecting 0.825 m from the wall at its lowest face. For the heights the model applies Vitruvius’s rule for the scaenae frons (V.6.6) to the orchestra diameter of 17.00 m: pedestals one-twelfth of it, a lower column one quarter, an upper storey three-quarters of the lower. The rule predicts a lower entablature of 0.85 m; the marble one measures 0.83 m. It gives a lower storey of 6.50 m and a crown 11.04 m above the stage, where Stillwell’s section shows about 12 m. No free number had to be invented.

The plan of the front comes from Stillwell’s restored plan: a central door 1.40 m wide between projecting podia, a side door on each side, and on each outer podium three columns and a half-domed niche, ten columns to a storey. His foundations are “not exactly symmetric” about the axis (p. 67), and the central opening sits about 0.45 m to the east. The model makes the front symmetrical.

This is the default state because it is the only one in which everything Stillwell restores stands together. Every fragment that fixes the heights of the orders is marble, the stair tower that makes the colonnade on the rim “certain” probably belongs to this period, and the marble work removed nothing of the Trajanic building.

Against the north-east curve, between the fourth and fifth buttresses, Stillwell found “a powerful, 3.00 m. rectangular foundation” of squared blocks in mortar (p. 48). He read the foundation as the core of a stair tower: with risers of about 16 cm, as in the corridor, “only four flights of steps are needed to reach the level of the colonnade” (p. 48). The tower and the colonnade hold each other up. The tower needs somewhere to go, and at the top of the cavea there is nothing else for it to reach: “If the stair tower is correctly interpreted such a colonnade is certain. There are, moreover, a quantity of plain limestone column drums (but no capitals) which could have belonged to such a feature” (p. 52).

The model follows his restored plan: 32 plain columns on a radius of about 27 m, sixteen to a side with a space on the axis, about 3.6 m tall, under a tiled lean-to roof against the top of the outer wall. The Tuscan capitals and the roof pitch are chosen. The tower’s date is also uncertain. Stillwell says only that it “may also belong” with the marble (p. 77), and Costello puts it under Septimius Severus (2014, p. 23), while Soren and James describe it with the Trajanic work (1988, p. 232). Before the tower, an outer stair at the south-west led to the upper seats (pp. 43–44).

The figure most often given is 3,500 seats, without a method. Mitford once gave 7,000 (Mitford 1980, p. 1316). Stillwell is the only author who shows his method: “about 3,300, based on an allowance of 40 centimeters (16 inches) per person” (p. 68). His alternative at the 36 cm allowed at Corinth is printed as 4,300, which does not follow from his own figure; the same bench at 36 cm gives about 3,670.

The restored rows let the sum be done again. Measured along the front edge of each row, the nineteen lower and six upper rows give about 1,326 m of bench round the half circle. The four radial stairs take out about 70 m, leaving about 1,256 m, or about 3,140 people at 40 cm each. The lower rows also run on past the half circle in straight lengths toward the parodoi, most of them over the parodos vaults, and counting those adds about 130 m and brings the total to about 3,460. The range is therefore about 3,100 to 3,500 at 40 cm a person, depending on how the ends were used, and about 3,500 to 3,850 at Corinth’s 36 cm. Stillwell’s 3,300 sits inside it, and the guidebooks’ 3,500 sits at its upper edge. All of these are estimates: the bench lengths come from a restoration, and the width allowed to each person is a convention.

Early in the 3rd century the orchestra became an arena, and Stillwell’s description is specific enough to build from. The builders cut away “the three or four lower rows of seats” and quarried the rock “some 70 or 80 cm” below the Roman floor, then spread it with 10 to 12 cm of clay and sand (pp. 41, 56, 66). The new edge is a rock-hewn wall on a radius of 10.80 m. Its height is fixed by the tribune, a small viewing box built over the axis, whose floor stood 3.00 m above the arena; “with the addition of a parapet the wall protecting the spectators would have measured some 3.70 m in height” (pp. 57–58). Stillwell had published the theatre at Corinth nine years earlier, where the arena wall measured 3.80 m (Stillwell 1961, p. 41 n.).

A crown block from the parapet, 0.35 m high with its face curved slightly inward, carries on top “a series of square holes, 0.38 m apart as for a grille” (p. 75). That is the only evidence for what stood on the wall: a grille of uprights, of unknown height and material (Soren and James call it a metal grill; 1988, p. 233). No net is attested, and popular accounts that remove two rows and give the arena a low wall understate what was done. In each parodos a timber barrier closed the passage, framed in rock cuttings 16 to 20 cm square (p. 50). Curved walls filled the corners between the arena and the stage front, faced with painted stucco in broad white panels 1.20 m wide between purplish-brown to black strips 0.33 m wide (p. 56). The stage platform was taken out.

On the axis under the tribune, a vaulted refuge 1.60 m wide and 2.20 m deep opened onto the arena at floor level, with a wall 0.60 m high across its door, “not so high ... that it cannot be easily stepped over” (pp. 56–57). Rock-cut cells off a passage under the north side had doors closed with locking bars; Stillwell allowed that they might have been prisons or pens for the animals and called the idea “entirely hypothetical” (p. 64).

Nothing found dates the conversion. Stillwell found it “tempting” to connect it with Caracalla’s years in the East, since Cassius Dio (78.9.7) says hunting theatres were built wherever the emperor wintered or expected to; hence his “about A.D. 214–217” (p. 78). The fittings are those of a theatre for wild-beast hunts, venationes, and Stillwell calls it a hunting theatre. We know of no evidence that gladiators fought here. The gladiator mosaic in the House of the Gladiators, in the north-west of the city, comes from a house built in the second half of the 3rd century (Costello 2014, p. 23), by which time the theatre had probably gone back to plays. It shows what one household liked on its floors.

Coins of Elagabalus, Philip the Arab and Philip’s empress Otacilia lay in the fill of the refuge and under the re-laid seats, so the return to theatre use came no earlier than the middle of the 3rd century (p. 78). It was done cheaply. The arena was filled about 0.70 m deep with two layers of broken stone, and a new orchestra was laid over it inside a curb of blocks on a radius of 7.20 m, with the lowest seats at 8.60 m. Its surface was a bed of broken stone and mortar, “presumably” for marble slabs, sloping toward the east parodos with no gutter (p. 56). A new front wall for the stage, 0.60 m thick, was founded on the rock (p. 58).

The lowest rows were re-laid from whatever came to hand, a few better blocks brought down from higher up among many “very rough and almost unshaped” ones, and the new stairs “hardly ever follow a truly radial line” (pp. 53–54). Almost every seat block Stillwell found in place belongs to this rebuilding (p. 53). The marble stage front stayed. Stillwell ruled out putting the marble in this phase: after “the very summary manner in which the building was reconverted from an amphitheatre”, “an elaborate marble decoration of the scaena seems improbable” (p. 73 n.).

The rooms behind the centre of the stage building lay under “piles of fallen blocks”, and the latest coins beneath them are of Valentinian and Valens, AD 364–375. In Room S.1 a coin of about AD 372 lay a little above the floor, and a similar one on the floor of Room T.4 (pp. 63, 78). The theatre came down after that. Stillwell suggested that earthquakes recorded at Salamis in the 330s and 340s “may well mark the end of the Curium theater” (p. 78), which his own coins contradict; he never mentions AD 365. That date reached Kourion later, through the excavation of the Earthquake House, and Costello’s study of that house concludes that the city fell “sometime during the later 4th century AD, possibly ca. AD 370–380” (Costello 2014, p. 15). The model gives the collapse no year. The island’s run of earthquakes is set out under earthquakes in ancient Cyprus, and the province under Roman Cyprus.

The stage front fell outward, away from the seats: “The greatest number lay tumbled to the south of the stage where they had fallen” (p. 69). Nobody rebuilt. The ring corridor was divided up by rubble cross walls, one with a coin of Theodosius II beneath it, and lime-kilns were built in it, which is where the marble revetment went (pp. 68 n., 78). Squared blocks, Stillwell notes, are “dear to plunderer’s hearts” (p. 68). The theatre became “only a great ruin, used both as a quarry and as a dumping ground” (p. 78).

Some of it went uphill. A fragment of an early Severan inscription, in which the city honoured an archon, Aristo- son of Ariston, with his wife and son, came from the pool in the court of the Annex of Eustolios; other pieces of the same text lay over the stage building, in Vomitorium III, south of the east parodos and in the topsoil of the orchestra (Mitford 1971, no. 99, pp. 185–186).

The model is built from Stillwell’s report, still the only full publication of the building: its text, the restored plan, the phase plan and the north-south sections, measured at their printed scales. Where text and drawing disagree, the text wins for the dimensions it states, and the drawings supply what the text leaves out, with an error of 15 to 30 cm in plan and 20 to 40 cm in height.

Every layer carries one of five grades. Excavated means measured on the remains or read from Stillwell’s record of them, as for the arena, the corridor and the stage foundations. Inferred is a strong inference, such as the vaulted parodoi. Reconstruction covers Stillwell’s restorations and the model’s own, including the upper rows, the colonnade and the whole elevation of the stage front. Speculative marks values chosen so the model can exist at all, such as Nero’s façade and the height of the grille. Rejected readings are shown only for comparison.

The chief rejected alternative is a stage front of three storeys rising to the top of the cavea, as at Aspendos or Orange. Vitruvius allows a third storey, and drawn reconstructions of Roman theatres often supply one. At Kourion the evidence gives two storeys in the marble period, Stillwell’s section stops about 12 m above the stage, and a front that high would have hidden the sea from every seat. The alternatives plate also shows seats of marble (they are local limestone), a stage building as long as the cavea is wide (it is 31.10 m), a net over the arena wall (only grille sockets are attested) and an arena wall about 2 m high on the line of two removed rows. The awning that Soren and James say the stage building carried (1988, p. 232) is a speculative option: no mast sockets or brackets have been reported.

The ruin state separates the 1961–63 work from the ancient fabric by the sharpness of its edges and joints, as a visitor does, and an optional overlay washes the modern work grey. We know of no published record of which blocks in the lowest rows are ancient, so the ancient blocks in the ruin are placed schematically.

Several answers above are choices. The largest is the seat rise, 0.35 m as measured or about 0.43 m as the levels demand. The podium is “something over two meters” in Stillwell’s text but reads nearer 1.3 m on his section. The restored cavea has nineteen rows in the reports of the 1960s and seventeen steps in the survey of 2011. The height of the stage rests on Vitruvius alone. His overview calls the Trajanic stage front straight, without recesses (p. 41), while his detailed account gives the niches to “the second period” (p. 68), which could mean the limestone front or the marble one.

Smaller doubts run through the plan. The positions of the vomitoria, doorways, windows and buttresses are read from small drawings and may be out by a few degrees. The text gives the orchestra radius as 8.50 m in one place and 8.60 m in another, and the arena floor as 0.70 and 0.80 m below the late orchestra; the model takes 8.60 and 0.70, which agree with his other figures. The stair tower may be Trajanic or later.

The end has no date closer than the coins allow. Under the fallen blocks behind the stage lay a coin of Valens struck about AD 372, and the theatre fell after it was dropped.