Chapter 3
Three and a half kilometres as the crow flies south-east of Antioch a steep, round-topped hill of grey limestone, rising to a little over 1600 metres, marks the edge of the plain of Yalvaç. It is locally known as Karakuyu (Black Well) after anow almost dry water-source close to a small Byzantine church. Behind and beyond the hill to the east the ridges rise high and barren towards the peaks of Sultan Daǧ, which form a chain at a height of between 2200 and 2340 metres. The remains of the sanctuary of Mên Askaênos, the chief god or patrios theos of Antioch, occupy the summit of this hill, hidden from casual view by dense stands of pine, the tree sacred to Mên. The shrine was joined to the city by a track, which can still be recognised in stretches where the rock of the steep hill slope was levelled to allow men and animals to make their way up and down. Apart from a few quarry pits (see below), the first signs of human activity are smoothed rock faces to the east of the path, on which have been carved groups of small aediculae, decorated with the sign of the crescent moon, the characteristic symbol of the god. Beyond these the path leads along a valley running approximately from west to east, littered with small piles of uncut stone from ancient buildings now too ruined to be planned. At the eastern end are the remains of the Byzantine church (see below pp. 201–6). The path curls round this towards the south and after about 100 m reaches a small stadium. Beyond this, in various states of preservation, are a large enclosed temenos, two temples, and up to twenty house-like structures. These comprise the main features of a sanctuary which flourished, to judge from its inscriptions and from the surviving architecture, through the whole period of Antioch’s pagan prosperity from the second century BC to the third century AD (Fig. 5).
The temenos
The largest and most immediately striking architectural feature of the sanctuary, and the one that has attracted most attention in the few earlier discussions of the site, is the rectangular temenos enclosure which crowns the western ridge, overlooking the plain of Antioch some 400 metres below (Fig. 6).1 Because its south-west wall carries the highest concentration of inscriptions and votive dedications to Mên on the site (PLATES 16 and 17), it figures prominently in the accounts given by Ramsay and his associates.2 However, these earlier inscriptions are inaccurate in many important respects and give no impression of the sanctuary’s real character.
The enclosure is built on a NW–SE orientation, with short sides on the north-west and south-east and long ones on the south-west and north-east. As Ramsay observed, it is not exactly rectangular in shape. He gave the inside measurements as follows: SE 118 feet (36.03 m); SW 230 feet (70.23 m); NW 114 feet (34.8 m); and NE 224 feet (68.4 m).3 Our own surveyed figures correspond quite closely with these. Precise internal measurement is now only possible on the north-east side, which we reckoned at 68.8 metres. The external dimensions are: SE 44 m; SW 73.1 m; NW 41.7 m; NE 71.5 m. The wide disparity between our figures for the north-west and south-east sides and Ramsay’s internal measurements for the same sides is explained by his erroneous belief that the south-west wall was some twenty feet thick.
Fig. 5. The sanctuary of Mên Askaênos
The thickness of the external walls, where it can be observed, varies from one part of the enclosure to another. At the west end of the north-west side it measures 1.95 m; between the second and third buttresses on this side it is 2.15 m, but two metres before the third buttress is reached there is a pronounced kink and the wall reduces to 1.9 m. At the east end it is 1.8 m. The north-east wall, which is less carefully constructed than the north-west or south-west sides, is 1.18 m at the west end (where Ramsay gives a figure of 3’ 8" or 1.11 m) and 1.22 m in the middle. The south-east wall measures 1.23 at the central entrance and about 1.2 m at the west end (PLATE 18). The southernmost section of the south-west wall, which is less well built than the rest of it (see below), measures between 1.16 and 1.2 m thick, but the rear side of the rest of its length is obscured by piles of earth and rubble, much of it derived from Ramsay’s excavation dumps. Ramsay claimed that the west wall was about nineteen,4 or twenty feet thick,5 and in fact comprised two walls, each some five feet thick, with a great mass of rubble infilling. He believed that this breadth was necessary ‘to counteract the downward thrust caused by the structure on a sloping hill-side’. In fact it would be quite unnecessary to build a terrace wall of this thickness to support the foundations of the temple and the other interior structures of the temenos, since these were supported by the natural rock of the hill. None of our own observations bears out Ramsay’s theory. At one point slightly to the north of the entrance on the south-west side a foundation block is still visible in its original position, aligned parallel to the south-west wall, 6.18 m behind the face. This no doubt corresponds to Ramsay’s so-called inner face (20' = 6.11 m), but in reality this was certainly not the foundation of a casemate wall but of an internal portico, which ran along this side of the temenos as it did along the three others. Ramsay’s 5' (= 1.53 m) measurement for the outer section of the wall should serve as a rough guide to the total thickness of the wall here. Our own calculations suggest that 1.8 m, roughly equal to the thickness of the north-west wall, may be closer to the mark (see below).
16. South-west wall of the temenos of Mên Askaênos (Kelsey archive 7. 1578).
17. Buttress and votive aediculae on the southwest temenos wall.
Fig. 6. The temenos and temple of Mên Askaênos.
18. South-east and north-east walls of the temenos with the temple behind.
The construction of the temenos wall was not uniform. On the south-west and at the west end of the north-west side it is best preserved, surviving to a height of more than 2.5 m for lengthy stretches. The highest point above current ground level is 3.3 m at the second buttress of the north-west side (PLATE 19). Along much of the south-west wall the lowest visible blocks form a foundation course, more roughly finished than those above them, and never decorated with the carved aediculae that cover the upper courses. Moreover at several points this foundation course was made from outcrops of the natural rock (visible below the buttress and at the bottom right of PLATES 17). It is clear, therefore, that on this side no part of the wall is buried underground. The rock also served as the foundation for the pavement inside the temenos and for parts of the temple itself. The outer face of the wall on the south-west and north-west consists in the main of large, roughly rectangular blocks of dark grey limestone, finished with the point of a pick. The largest of these are as much as 90 cm square, but more typically their dimensions range between 40 and 70 cm. They were carefully dressed so as to fit closely with one another, and smaller blocks were sometimes inserted to fill the spaces between them. However, they were not generally cut to exact right angles, and only intermittent attempts were made to lay them in even courses (PLATE 20). The interior of the walls consists of large unshaped pieces of rubble, piled in and packed against the faces without any attempt at coursing or regularity. The overall effect is one of rough-hewn solidarity, not of rigidly geometrical symmetry. This was surely deliberate and in keeping with the character of a mountain sanctuary. The appearance of the temenos walls would blend with the natural grey limestone of the mountain top.
As currently preserved the maximum number of courses to survive on the south-west and north-west faces is five, but originally the wall must have been much higher. The uppermost surviving course stands hardly more than a metre above the foundations of the internal portico. Yet the temenos walls must have more than matched the column heights of this portico, so we may assume that they stood between three and five metres higher than they do at present.
The construction of the north-east and south-east sides is far less imposing. The walls here are thinner (1.1–1.2 compared to 1.8–1.9 m) and also made from rectangular or trapezoidal blocks, although these have been less carefully fitted to one another and there is a much higher proportion of small stones used to fill the gaps between them. The poorest quality of all is at the south-west corner, extending along the south-west wall to within 1.3 m of its southernmost buttress and at least as far as the second buttress of the south-east wall (PLATE 21). Here the external face is made from roughly shaped stones not laid in courses at all. No doubt the original wall had collapsed at this point, possibly due to earthquake damage which appears to have affected other parts of the site, and what survives today is a late repair.
The most conspicuous features of the wall are the buttresses which project from the face at roughly regular intervals: ten along the south-west, four on the north-west,6 nine along the north-east, and three along the south-east side, all to the west of the large entrance portico here. Their construction is virtually identical to that of the walls to which they belong, and they are certainly contemporary with them. On the south-west they are usually demonstrably bonded into the wall face, and in some cases a block from the wall itself has been cut so as to project and form part of the adjoining buttress (PLATE 22). On the other hand the bonding and attachment is rarely very solid and does not seem to have been designed in the first instance to increase the stability of the structure. The buttresses, therefore, were primarily ornamental. Comparable buttresses may be seen in the temenos walls that surround the temple of Apollo Clarius and the Doric temple which overlooks the upper agora at Sagalassus.7
19. North-west wall of the temenos.
20. Construction details of the south-west temenos wall.
21. South-west corner of the temenos showing low quality rebuild.
22. Buttress bonded into south-west temenos wall.
Where the buttresses survive to a considerable height on the south-west, it is clear that they taper from a broad base, approximately 1 m square, to narrower upper courses.8 On the north-east side, where the wall and buttresses stand hardly more than a metre above ground level, the dimensions are regularly larger, corresponding to the lower courses on the south-west side, with widths ranging from 98 to 116 and projections from 64 to 98 cm. The buttresses on the south-east are of comparable size. It is interesting that on this wall there are no buttresses east of the entrance to the temenos. At this point a row of rectangular oikoi approaches the sanctuary (see below), and surely for this reason buttresses were not added to the wall. This may imply that the oikoi were either built at the same time as, or before, the temenos boundary wall.
The south-west wall and its buttresses are liberally covered with the carved votive aediculae, with or without inscriptions, which are so typical of the sanctuary. The inscribed texts are all of the Roman imperial period, when the existing rough surface of the wall blocks was smoothed to take the texts and carvings.9 There are three entrances to the temenos. The most important of these was on the south-east side, facing the front of the temple itself. In its original form it was a simple opening, 1.77m wide,10 between two massive rectangular blocks (measuring 90 x 123 and 92 x 123 cm). At a later date apropylon was added in front of the entrance (Fig. 7). Its side walls were not bonded into the temenos proper and the construction, at least of the foundations of the colonnade at the front, was built to a more rectilinear specification than any part of the temenos wall. Moreover it is aligned 3o closer to the north (31oagainst 34o) than the south-east wall and is approximately parallel to the front of the temple itself. The remains at the front of the propylon comprise four large rectangular blocks,11 with smaller rectangular stones between them, set alongside a foundation course of rougher stones (PLATE 23). The large blocks at the two ends, each 43 cm high, belong to the course above the central ones and were probably the lowest visible part of the portico, while the middle blocks belong to the euthynteria. If we assume that the centres of the four columns stood exactly above the middle of each block, the interaxial measurements from west to east would have been 2.615, 3.295 and 2.6275 m, that is 8, 10 and 8 Doric feet. The west and east walls of the propylon were of uneven length, 6.11 and 5.75 m respectively, and between 1.05 and 1.2 m thick. They are made from large, roughly shaped blocks, which contrast with the rectangular foundations of the front colonnade and probably supported a wall similar in character to the temenos wall itself.
Fig. 7. The propylon of the temenos.
23. Foundations of the propylon to the temenos.
There were two smaller entrances. One, 1.44m wide between buttresses 12 and 13 on the north-east side, led by way of two rock-cut steps (the lower 1.62 m wide and 40 cm deep) to a small terrace outside the sanctuary (PLATE 24). The other, 2.25 m wide between buttresses 20 and 21 on the south-west side, led down by a longer stairway to a peristyle courtyard and another building complex, which is discussed in detail below (see PLATES 16 and 49).
The interior of the temenos was surrounded by a portico on all four sides. This is best preserved on the south-east and north-east, where the blocks of the uppermost foundation course, the euthynteria, are still visible (PLATE 25). Along the south-east side this comprises a row of rectangular grey limestone slabs, 95 cm deep and ranging in length from 88 to 137 cm.12 The blocks are approximately 28 cm high. There are cuttings visible in the tops of the foundation slabs for the attachment of the lowest visible blocks of the portico, the edge of which stood 4.37 m in front of the inner face of the temenos wall. On the north-east side the arrangement is very similar except that the rear edge formed by the blocks of the euthynteria is irregular (they range from 75 to 88 cm deep). The evidence for the south-west and north-west sides is much more tenuous. On the north-west no blocks from the euthynteria can be recognised, but a line of rougher stones, whose front edge is 4.24 m from the temenos wall, was doubtless part of the foundation packing for the portico. On the south-west side the only trace of the portico has already been noted, the single block, perhaps from the euthynteria, close to the entrance on the south-west side, which is 6.18 m from the outer face of the temenos wall. If the front of the portico here was about 4.38 m from the inner face of the temenos wall, as it is opposite, this would have been about 1.8 m thick (see p. 41 above).
24. North-east entrance to the temenos.
25. Euthynteria of the south-east portico of the temenos.
The columns of the porches would have stood on the course of blocks which rested on the euthynteria. Several fragments of Ionic fluted (PLATE 45) and unfluted white limestone column drums were noted in the debris which litters the temenos. Their diameters fall in the range 50 to 56 cm. In addition one white limestone fragment found close to the north colonnade seems to have been either the apophyge of a wall, perhaps the rear of the portico, or, preferably, the stylobate of the portico itself (see the catalogue of mouldings, Fig. 13 (i)–(l) ). If we are right to suppose that these belonged to the portico, the architects ensured that there was a deliberate contrast between the white columns and the grey limestone of the solid walls, as they did with the stonework of the temple itself (see below).
There is no reliable way to work out the internal dimensions of the porticos and the arrangement of the columns. Calculation suggests that the lengths of the four sides were approximately as follows: SW 61.3, NE 60.2, NW 30, and SE 32.2 m. This could allow an arrangement of twenty-two columns on the south-west and north-east and eleven on the north-west side, with interaxial spacing between 2.73 and 2.78 m. The columns themselves seem to have been about 50 cm in diameter, leaving intercolumnar spaces of about 2.25 m, the so-called eustyle arrangement favoured by the architect Hermogenes. But these figures necessarily rest on quite unproven assumptions about the numbers of columns along the sides of the temenos.
Ramsay noted two features associated with this portico which require discussion. He claimed that alongside the south-west wall a cistern had been excavated out of the rock at a later date to provide water for the sanctuary, which had been filled with ‘fragments of architecture, large inscribed columns, pieces of votive inscriptions etc’.13 In his first report he gave the dimensions as 136 x 8 feet; six years later the cistern has become ‘115 feet long, cut out of the rock of the hillside to a depth of 8 feet’. Absolutely no trace of this cistern can be seen today. Conceivably something may lie buried beneath the debris on this side of the sanctuary, but this is unlikely. The cistern would have had to be excavated underneath the south portico, which seems neither plausible nor architecturally possible. It is much more likely that Ramsay’s ‘cistern’ was simply the gap that he observed between the rear edge of the foundations of the portico and the inner edge of the temenos wall. The rock cuttings to which he refers could have been the cuttings made to level the natural rock on which the buildings stand. The figure of 8 feet must then refer not to the depth of this feature, as Ramsay’s later report suggests – for an 8-foot hole would certainly have resulted in some visible depression today – but to its width. We should thus dismiss the cistern as afigment of Ramsay’s imagination.
His other observation concerns the south-east corner of the temenos, where the excavators found the over-life-size statue of Cornelia Antonia, the plinth of asecond statue of a certain Tibereinos, as well as smaller items of sculpture.14
Ramsay claimed that 14' 4" (= 4.38 m) from the inner corner on the north-east side the wall widened at a right angle to a thickness of 6' 2" (= 1.88 m) and that the statues stood here on a slight platform. Traces of foundations can still be seen at this point, including a line of stones running roughly parallel to the inner face of the north-east wall and about 87 to 97 cm from it. However, it is unlikely that these are anything other than foundation material for the pavement of the portico, and the ‘platform’ seen by Ramsay was probably either part of this foundation or of the paving of the portico itself. Ramsay also alleged that there was a ‘chapel’ at the southern corner of the temenos, cut into the thickness of the south-west wall.15 Again we saw nothing to confirm this and presume that here too he was simply misinterpreting the visible remains of the portico.
The interior of the temenos was paved with closely fitted grey limestone slabs (PLATE 26).16 Most of these have been removed, showing that in several places they were laid on to the natural rock of the hill top, which was cut to take them. However, patches survive intact along the south-west side of the temple, in front of it and in the south-west corner of the enclosure. The level of this pavement immediately in front of the temple is higher than it is in the area close to the south-east portico, suggesting that there were steps or a terrace running across the area between these two points, where no paving stones have survived.
The quality of the finish of the paving, the portico, the propylon on the south-east side, and of course of the temple itself contrasts markedly and deliberately with the cruder and rougher finish on the exterior walls of the sanctuary. Although little of the ornament is now to be found, there is no doubt that this was designed as a sophisticated, symmetrical architectural ensemble, in the classic hellenistic manner (PLATE 27).
The inside of the temenos was liberally adorned with inscriptions, statues, votive monuments and carved bases, which supported statues and doubtless other offerings. Today only a small number of broken fragments can be seen on the site, some in the temenos itself, others in the neighbourhood of the probably fifth-century church which was largely built of stone taken from the temple and its surroundings. A selection of the carved bases, made from white marble or from white or grey limestone, is illustrated in the catalogue of mouldings (Fig. 13, (m)–(s) ). A better, although very imperfect, idea of the original position of some of this material can be gathered from the accounts of Ramsay’s excavation. J.G.C. Anderson, who published a series of agonistic inscriptions from the site, the majority of Roman imperial date, noted that most of them came from the temenos. They include bases set for athletic victors found in the ‘cistern’, that is in the south-west portico (JRS 3 (1913) 287 nos. 12, 13, 19, 22–3) and along the inside of the north-west wall (nos. 20–1). These bases include hexagonal columns inscribed on one face (no. 12 with Fig. 62, no. 13 with Fig. 63, no. 14 with Pl. XXI, no. 15, no. 16 with Pl. XXII), cylindrical cippi (no. 20 with Fig. 67, no. 21 with Fig. 68, no. 22 with Fig. 69), and an Ionic column drum, original diameter about 49 cm,17 with an inscription carved on a panel where the fluting had been smoothed flat (no. 23 with Fig. 70). This was probably one of the columns of the portico itself. We noted a hexagonal white limestone base, uninscribed at least on the visible sides, similar to those described by Anderson, in the south-west corner of the temenos, as well as a columnar white limestone cippus with a Latin dedication to Luna on the south-west side of the temple itself. From Anderson’s description of the find spots there is little doubt that most of these agonistic records were erected in the colonnade itself.
26. Paving of the temenos.
27. Moulded fragments from the temple and the temenos.
Anderson also published a series of agonistic inscriptions which had been carved on flat slabs or blocks, usually in a very fragmentary state (pp. 269–70 nos. 1–5 with Figs. 46–8, 281–2 nos. 8–10a with Figs. 58–9, 284 no. 11 with Fig. 1, 295 no. 24 with Fig. 71, and no. 25 with Fig. 72). He offers no descriptions of the stones, and his photographs and drawings give no idea of the original form of the monuments on which these texts were inscribed, except to suggest in one case that it was part of a rectangular block (p. 283 no. 10a, Fig. 59). However, one of his descriptions gives an important clue to the provenance of the whole series. A broken two-line fragment was ‘engraved on what is apparently a portion of a slab surrounded by a flat border on a lower level than the inscribed part’ (p. 295 no. 25). This suggests that Anderson was looking at one of the blocks from the cella wall of the temple itself, which have drafted margins and edges which are recessed from 5 to 7 mm below the face of the rest of the stone (see Fig. 13 (g) ). We noted several such blocks in the temenos area and others reused in the church, including two which carried fragmentary inscriptions along the drafted edge itself (PLATE 43).18 We may conclude that at least some of this last series of texts were inscribed on the walls of the temple, although we cannot exclude the possibility that others were fixed to or part of the internal walls of the porticos.
It is also clear that a further series of inscriptions, the small votive stelae of white marble carrying dedications to Mên Askaênos, were also displayed within the sanctuary. These were discovered by Ramsay’s team, but few were published at the time.19 A selection of these was taken to Konya and now forms part of the museum collection there. They were published separately by E.N. Lane and Barbara Levick,20 and then included in the third volume of Lane’s corpus of monuments relating to the cult of Mên.21 Lane later obtained access to Ramsay’s epigraphic notebooks which record many otherwise unpublished inscriptions from the sanctuary, and published them in CMRDM IV. None of these publications offers any discussion of the original location of the stones. Several of the marble stelae, which take the form of aediculae, have a tag projecting from the bottom by which they could have been fixed to a supporting block or to a wooden frame. Miss Hardie claimed to have seen niches both in the peribolos wall and in the rocks below the sanctuary where the votives might have been placed.22 We saw nothing that might have been suitable for this purpose, and it is overwhelmingly likely that these marble dedications, a more expensive type of offering than the crudely carved aediculae to be seen on the outside of the temenos wall or on the rock faces leading up to the sanctuary, were displayed within the temenos itself. This is confirmed by an observation of Ramsay, that the marble dedications were usually found inside the central sanctuary.23
The temple of Mên
The first discoverers of the sanctuary were perplexed by the building inside the temenos. Ramsay’s original view, followed by Margaret Hardie, was that ‘there was no temple but only a great altar, about 66 ft by 41 ft’.24 A year later Ramsay changed his mind and inclined to the view that it was ‘a small temple of quite unusual character’, and he noted that fragments of Ionic columns and capitals were to be seen.25 In the fuller report published six years later the building only merits the bare mention that it was a temple.26 This was recognised in Ismail Karamut’s brief report of 1989.27 Our own study leaves the matter in no doubt.
The ruins are clearly the foundations of an Ionic peripteral temple, standing on a stepped krepis, with 6 columns on the short and 11 on the long sides (PLATE 28). Far from being of unusual character, it is a classic Greek building of hellenistic date (Fig. 9).
28. Temple of Mên Askaênos viewed from the south.
Most of the temple is made from the dark grey limestone, sometimes with fine white veins, which is quarried from the mountain on which it stands.28 However, certain parts of the building were made of a contrasting white limestone, perhaps the same as the fine limestone used for the imperial temple in the Roman colony. These included two fragments of white Ionic column capitals and asegment of the frame of the cella door (see catalogue of mouldings Fig. 13 (d) and (f); PLATE 27). On the other hand the apophyge of a column and the finely carved moulded socle and the blocks from the cella wall were made from the same grey limestone as most of the foundations (PLATE 29). The architects obviously intended a deliberate contrast between a sparse selection of details in white stone and the dark grey of the body of the building.29
Fig. 8. Temple of Mên. Cross section of krepis on the W side.
Fig. 9. Temple of Mên. Plan of the foundations.
29. Socle moulding of the cella wall of the temple.
The length of the krepis at the lowest visible course of steps is 31 m and the width 17.4m. However, at least one step lies buried on the long north-east side, and the width should be estimated at 17.8m or a little more. At the foundations of the seventh step, which may be followed around all four sides, the dimensions are 26.7 x 14.4 m, and the top of the stylobate measures 25.25 x 12.53 m, giving proportions of almost exactly 2:1. Within the peripteros the naos proper measures 18.52 x 8.63 m. There were 10 steps on the south-west side of the krepis, all of which can be traced, and certainly the same number on the north-west (PLATE 30). Only 6 of these, reckoned from the bottom, can now be seen, but the distance between the uppermost surviving step here, the sixth, and the centre of the foundation platforms for the columns at the north-west end is 2 m, exactly the same distance as that between the sixth step and the centre of the column foundations on the south-west side. We may assume that the total number and the arrangement of the steps on these two adjacent sides were identical. On the south-east side there were only six steps, at the same level as steps five to ten on the south-west side. This implies that the lowest four steps on the south-west led down to a lower level than those on the south-east, and this is confirmed by level readings taken on the surviving flagstones of the paving of the temenos, which show that the south-west was 1.019 m below the south-east area of paving. The measured heights of courses 2 to 5 of the stairway come to 1.07 m, thus closely approximating to this height differential.30 The krepis on the north-east side probably also had 6 steps, joining those on the south-east. Three of these can be traced, and a block close to the north-east corner of the building may belong to afourth. The others lie buried by earth and rubble. These observations on the krepis imply that there must have been a step or a terrace in the level of the courtyard itself across the area to the west or north-west of the temple, which linked the higher ground level on the south-east and north-east with the lower level on the south-west and north-west.
30. Foundation of the krepis on the south-west side.
At the north end of the south-east side steps 2 to 4 had been cut from the natural rock, which served as a foundation for the whole building, but every-where else the foundations of the steps were made from individual stone blocks, laid in roughly even courses, but with a rough, quarry-hewn face rather than asmooth finish ( PLATE 31). The one exception to this occurs at the top of the south-west side where a single block was cut to make two steps. The rough finish of the stones is itself enough to show that the visible remains did not make up the outer facing of the krepis. This is confirmed by the presence of a single example in situ of the original facing blocks, resting on the lowest foundation course close to the west corner of the building (PLATE 32). Two other examples were discovered displaced from their original positions, one in the middle of the south-west, the other on the north-west side.31 The dimensions of these stones and the indication of the foundation courses suggest that the steps generally measured about or a little more than 25 cm in height (Fig. 8).
31. Krepis foundation blocks.
32. Facing blocks on bottom step of the xkrepis.
At the top of the krepis the foundations of the peripteral columns can be traced on all sides of the naos. Along the south-west side the individual foundation platforms of 9 columns are still evident, excluding the fourth and the eleventh from the east end (Fig. 9). They consist of regular limestone blocks, and the foundations of the top step of the krepis butted up against them. This can be seen most clearly at columns 3, 5, 6 and 8. On the opposite long side the first column base from the east end has entirely disappeared. Numbers 2 to 6, like their counterparts opposite, were individual bases, numbers 2, 4 and 6 being made from a single, and numbers 3 and 5 from adjoining slabs. Columns 7 to 9 were built on a continuous foundation wall, which had larger blocks at the points where the columns would have stood. Columns 10 and 11 have disappeared. At the north-west end three original bases survive out of the original six, the two central ones and one to the west of them. Like most of those on the south-west side these are made from two adjoining slabs. Behind the two in the middle are two further bases of similar construction, which stood between the antae of the opisthodomos. There were also apparently individual foundation platforms at the south-east end of the temple, next to the top step of the krepis, but the traces here are indistinct. Between the second and third individual platforms of the north-east side there is a packing of smaller stones, which was probably repeated in all the intercolumnar spaces beneath the paved floor of the temple. Elsewhere, however, it has disappeared.
As already noted, the length of the perípteros on the long side is about 25.25 and the width 12.53 m. These are not exact measurements since we are dealing with much-damaged foundation courses not the finished superstructure of the building. However, these figures imply interaxial spacing of a little under 2.5 m between the columns, and this figure is confirmed by measurements made between the individual foundation platforms.32
The colonnade encloses the naos, which comprised an opisthodomos with 2columns between the antae, a cella, and a pronaos. There is space on the ground plan for a row of 4 columns at the front of the pronaos, with 2 between its antae, and it is in theory possible that there was a double colonnade at the front. The best parallels for 4 columns in front of the pronaos (tetrastylos) are the temple of Zeus Sosipolis at Magnesia on the Maeander, although this temple has no peristyle, and the later hellenistic temple of Apollo at Alabanda. On the other hand the hexastyle temple of Leto at Xanthus, the Ionic hexastyle temple of Artemis (?) at Termessus,33 and the larger octostyle temples of Apollo Smintheus in the Troad and of Hecate at Lagina all lack prostyle columns, although their designs allow room for one. In all these cases the wall of the pronaos begins in line with the third column of the peristyle, as at Antioch. Although in all these cases there were 2 columns between the side walls of the pronaos (distyle in antis), these parallels confirm that there were no prostyle columns in the temple of Mên. The pronaos of the Mên temple is almost clear of rubble and other debris, and we can say confidently that there were never any columns between its antae. The span was small (5.9 m; compare the 5.7 m span in the temple of Zeus Sosipolis at Magnesia) and no internal supports were necessary. The spacing of the columns conforms to Vitruvius’ eustyle plan.34
The foundations of the three rooms of the temple are continuous and were made from large rectangular limestone blocks. Along the south-east and south-west walls of the pronaos and the adjoining part of the south-west wall of the cella, blocks from the lowest visible course of the wall itself, not merely the foundations, are still in situ or very slightly displaced from their original position (PLATE 33). The exterior face of these walls had a continuous socle profile of the Attic type with a cavetto underneath (see catalogue of mouldings, Fig. 13 (a) ). The interior wall of the cella had no profile, but the equivalent face of the pronaos wall, which was in a sense part of the exterior of the building since the south-east end of the pronaos was open, has a carved design very similar to, but not identical with the external moulding.
33. Temple cella.
34. Socle blocks of the pronaos.
The threshold between the pronaos and the cella was made of three blocks, two of which are still in place. The doorposts were set in rectangular holes connected by grooves, and there are several holes presumably intended for fixing a wooden door-frame. The moulded course of the pronaos and the cella walls rests on foundations whose top course has been hollowed out at the top exterior corner to form a rim between 2 and 2.5 cm deep, on which the pavement or floor of the temple will have rested (PLATE 34). On the long north-east wall the uppermost foundation course has a hollowed rim about 28–30cm across, which would have accommodated the socle moulding placed above it. The arrangement is clear on the in situ block belonging to the long south-west wall at the south corner of the cella.
The internal dimensions of the rooms are as follows: pronaos 5.3, cella 7.9, opisthodomos 2.5 m. The thickness of the wall between the opisthodomos and the cella, and presumably the wall thickness throughout, was 70 cm. The internal width of all three rooms was 5.9 m.
The small temple
The small temple is situated on a rocky spur which overlooks the rest of the site from the north-east (PLATE 35; Fig. 10). Its orientation is roughly parallel to that of the temple of Mên and it was built on a terrace, cut from the rock on the north-east side, and shored up on the south-west by a terrace wall which com-prises three courses of rough-hewn blocks. This supporting wall probably continued along the north-west side at the rear of the temple, but the stones have fallen from their original position. On the other hand there was probably no terrace wall at the south-east end, where a gentle slope ran up to the building.
35. Peak behind the small temple.
36. Small temple. Interior view of north-east wall.
37. Small temple. North-west end.
Fig. 10. Small Temple. Plan of the foundations.
The temple was constructed from a mixture of grey and white limestone blocks, cut and laid for the most part in an irregular ashlar style. The foundation courses were placed on the natural whiteish limestone of the hill, which has been cut and smoothed in places to receive them. Not all the courses are horizontal, and the stones are of uneven height. Three qualities of finish may be detected. The bottom courses of the foundations, now visible at the base of the north-east wall on the interior side (PLATE 36), and on the exterior at the north-west end, are merely shaped with a pick and were certainly not intended to be visible. Above them, the lowest courses of the wall itself were finished with the point of achisel to give a stippled effect ( PLATE 37). Certain key features were prepared to a higher standard, notably the row of blocks at the south-east end of the building, which probably supported a step leading up to the level of the pronaos, and the white limestone threshold block or step which was found out of position in the ruins of the wall dividing the pronaos from the cella (PLATE 38). Here the finish is the work of a fine-pointed chisel. Doubtless much of the superstructure and of the decorative order of the building, which has now disappeared entirely, was completed in a similar fashion.
38. Small temple. Step leading to pronaos.
The overall measurements of the foundations are 14.8 x 7.6–7.7 m, but the actual dimensions of the built portion were less than this, 14.55 m long by 7.25 m wide, giving the same 2:1 proportions as the Mên temple (PLATE 39). The internal width of the building is 5 m, and it was divided into three chambers, with an opisthodomos of 2.5 and a cella of 4.5 m. The wall thickness varies. The south-west wall of the cella measures 1.12, the north-west wall of the opisthodomos 0.72, the north-east opisthodomos wall 0.88 and the wall dividing the cella from the pronaos 0.86 m.
39. Small temple viewed from south-east.
The layout of the pronaos is less clear. The south-east wall, which is made from smoothly finished blocks, projects beyond the north-east line of the building. It should probably therefore be seen as a step, which led up to a second step which was laid about 40 cm behind the outer face, along a line which is clearly visible on the block in the left foreground of PLATE 38. This second step would have reached the pavement level of the pronaos. The two flanking walls of the pronaos probably did not extend further south-east than the clearly defined blocks on the north side, 3.35 m from the front wall of the cella. There was a gap of about 1.65 m between this point and the second step, which presumably accommodated a row of four prostyle columns, forming the façade of the temple. The distance of 5 m from the cella wall to this step and to the front of the columns was thus approximately equal to the length of the cella and twice that of the opisthodomos, giving the ground plan proportions of 2:2:1.
The two internal walls differ in quality from one another. The front wall of the cella is comparable to the external walls, with rough foundations and the lowest visible courses stippled with the point of a pick. Slightly to the south-west of the centre of the wall there is a well-cut rectangular white limestone block (1.18 x 0.5 m) which appears to be from the threshold, or perhaps from a step (if the cella floor was higher than that of the pronaos). However, the stone cannot be in its original position since it lies at a lower level than the lowest visible course of the cella wall and even than the lower step leading to the pronaos. It must have been displaced from a higher point in the same wall.
The wall dividing the cella from the opisthodomos consists of four courses of roughly cut stones which had not been finished with the point of a pick. More-over, its highest surviving point, 80 cm above the present ground level, is only about 20 or 30 cm above the upper level of the lowest foundation course at the north corner of the building, and just below the upper edge of the bottom course of well finished stones there. It is, therefore, a foundation wall with no surviving superstructure. It also appears not to have been bonded in with the external walls of the temple.
The west wall of the temple consists of a single course of masonry intended to be visible, which rests on a very solid foundation 80 cm high and two or three courses deep. The inner face of the wall proper (distinct from the foundations) is very rough, while the upper course of the foundations has been pointed. There must, therefore, once have been another internal row of blocks or a layer of cladding belonging to the wall proper, which was placed against the interior side of the surviving portion.
The north-west end of the building might originally have been a single continuous wall, but comparison with the ground plan of the temple of Zeus Sosipolis at Magnesia makes it more likely that two columns were equally spaced between the projecting antae of the opisthodomos (Fig. 11). This colonnade will have begun at least one course above the existing remains, since the uppermost course at the east end appears to have been hollowed out to receive a further layer of stones on top of it (PLATE 40).
40. Small temple. Rear of opisthodomos
41. Octagonal base in the small temple.
There is no trace of any of the decorative mouldings of the temple, and the only ornamental feature that remains is an octagonal block of white limestone (1.045 m high, 0.68 m wide) which lies inside the cella (PLATE 41). It may have served to support a statue of the deity.
Fig. 11. Small Temple. Reconstructed plan.
Architectural comparisons
The temple of Mên at Antioch and its smaller companion can be placed in the evolution of temple architecture in hellenistic Asia Minor. Since so much of the decorated order and detailed carving has been lost, any attempt to place them precisely within this framework must rest principally on an analysis of their ground plans and comparisons with those of other buildings (Fig. 12).
The key building in the history of hellenistic Ionic temple architecture is the temple of Athena Polias at Priene, whose design is attributed to the architect Pythius. It was hexastyle and peripteral with eleven columns on the long sides. The design dates to the mid-fourth century BC when construction began, although most architectural historians believe that building continued for some two hundred years afterwards.35 The influence of the design of the Priene temple is immediately obvious in several other lesser-known buildings of western Asia Minor. The recently excavated temple of Leto at Xanthus in Lycia, which was probably completed by the mid-third century, is also hexastyle with eleven columns on the long sides. The differences from the Priene temple are obvious. The front of the pronaos was aligned with the third not the second column of the peristyle, it has no opisthodomos, and the cella has an unusual design with engaged half columns on the sides and engaged full columns in the rear wall. However, the relative proportions of pronaos and cella are close to those of its predecessor. In addition, the partition wall dividing the pronaos from the cella, like the two partition walls of the Priene naos, was not yet aligned with the columns of the peripteros.36
Another Ionic 6 x 11 temple which has been attributed to the third century BC is the temple of Zeus on the Kesbelion at Selge in Pisidia. Although the evidence is slight, this also seems to have had both a pronaos and an opisthodomos that were distyle in antis. The pronaos is relatively shorter than at Priene or Xanthus, contrasting with the later fashion for lengthened pronaoi, and this is an argument for the third-century date favoured by Machatschek and Schwarz in their publication. On the other hand the internal dividing walls of the naos are precisely aligned with the fourth and ninth columns of the pteron, and this appears to reflect the preference for a more rigid symmetry in temple design, prevalent among buildings of the later third and second centuries. A middle rather than an early hellenistic date is in any case perhaps preferable for abuilding in this highland area of southern Asia Minor.37
The Selge building, in fact, probably shows the influence of the architect Hermogenes, who was responsible for another important 6 x 11 temple, of Dionysus at Teus.38 Here the pronaos was extended almost to the length of the cella and the ground plan forms an axial rectangle of precisely 110 x 55 Attic feet, a symmetry that had not been aimed at in earlier buildings. Another refinement observable at Teus and in most later hellenistic temples is that the internal walls were strictly aligned with the external colonnade, in this case with the fifth and ninth columns of the long sides. Further, the ratio between the base diameter of the columns and their interaxial spacing has been increased from pycnostyle and systyle in Vitruvius’ classification to the eustyle arrangement which Vitruvius says was an invention of Hermogenes.39
Fig. 12. Comparative plans of temples in Asia Minor.
The Dionysus temple is one of three that have been attributed to Hermogenes. Its date is disputed but a convincing recent study argues that it should probably be placed in the last two decades of the third century, perhaps between 218 and 204 BC, when Teus was under Attalid domination. The later Hermogenian temples of Zeus Sosipolis and Artemis Leucophryene, both at Magnesia on the Maeander, may then be dated to c. 195–90 and 180–60 BC respectively.40 Comparison should also be made with the hellenistic Ionic temple at Termessus, sometimes called the temple of Zeus Solymius but more probably dedicated to Artemis. This had 6 x 11 columns, a pronaos with two columns in antis, but no opisthodomos.41
The broad similarities between the temple of Mên at Antioch and this group of hexastyle peripteral buildings is obvious. The most important difference is that the pronaos has no distyle in antis arrangement, probably because the span was too narrow to warrant it. Features which seem to betray Hermogenian influence are the eustyle spacing of the columns (the ratio is about 3.1:1) and the alignment of the internal walls with the fifth and ninth columns of the pteron.
The design may also be compared with that of several larger octostyle buildings of the third and second centuries. The basic proportions are close to those of the pseudo-dipteral temple of Apollo Smintheus in the Troad, which has been dated to the second quarter of the second century BC. In that case the relatively short pronaos and the pycnostyle spacing of the columns (2.50:1) hark back to those of third-century buildings.42 Hermogenes himself developed the basic ground plan of the Smintheum in the huge pseudo-dipteral octostyle temple of Artemis Leucophryene at Magnesia, his chef d’oeuvre 43 Here he increased the length of the pronaos in proportion to that of the cella and heightened this impression of length by adding an extra pair of columns between the antae of the pronaos, and three more pairs in the cella itself. The main purpose of these columns was doubtless to help sustain the thirteen-metre span of the roof, but they also served to give a processional feel to the approaches to the cella door and to the cult statue in the adyton. The Mên temple at Antioch resembles Artemis Leucophryene in its long pronaos, two-thirds the length of the cella. Amore revealing similarity between the two buildings is the use of the Doric foot (32.8 cm) as a unit of measurement, in place of the much more widely used Attic/ Ionic foot (29.4 cm). The blocks of the cella wall at Antioch are precisely one Doric foot high and the overall outline of the stylobate is 77 x 38.2 feet. The use of this unit at Magnesia may have been influenced by design considerations, while at Antioch, Magnesia’s colony, it may simply have been a matter of imitation.44 In general, however, the Antioch building was both more conservative and less ambitious than this illustrious model.
Two later temples of the second half of the second century which were influenced by Artemis Leucophryene should also be mentioned, the Ionic octastyle of Hecate at Lagina and the very similar Doric octostyle of Apollo at Alabanda, both pseudo-dipteral buildings without an opisthodomos, but with wider spacing of the central pair of columns in the front and each with the relatively long pronaos favoured in the mid-and later hellenistic periods.45 In this respect they offer parallels to the Antioch building.
These buildings provide a corpus of temple designs which were to remain influential during the Roman period. Two archaising octostyle temples of Imperial date, of Rome and Augustus at Ancyra, originally even attributed to Hermogenes but now acknowledged as early imperial, and of Zeus at Aezani, built under Hadrian and Antoninus Pius, have much in common with these hellenistic precursors.46 However, closest to Mên at Antioch is the corinthian hexastyle peripteral temple of Zeus Lepsinos at Euromus in caria, which is distyle in antis in the opisthodomos. Proportions, design and dimension are a close match for the Antioch building, except that it had four prostyle columns in front of the pronaos.47
These comparisons only provide a rough basis for dating the Antioch temple. The alignment of the walls of the naos with the columns of the pteron and the eustyle columnar spacing suggest a second- not a third-century date, later than Dionysus at Teus. The use of the Doric foot may have been copied from Zeus Sosipolis and Artemis Leucophryene at Magnesia on the Maeander, and the relatively long pronaos finds parallels in the middle and later second century. The high krepis is also a feature of buildings at this period, especially the temples of Dionysus and Demeter at Pergamum.48 A construction date between 175 and 125 BC seems probable.
The smaller of the two Antioch temples needs no lengthy discussion. It should already be clear from the earlier examination that its ground plan closely resembles that of Zeus Sosipolis at Magnesia. Given that Antioch was colonised by Magnesian settlers, it is difficult to resist the inference that the design was transferred directly from metropolis to colony. It may reasonably be dated to the same general period as the Mên temple. If so, during the second century BC the Magnesian settlers had adorned their chief sanctuary with two buildings whose architecture was directly inspired by models from their Ionic homeland. They are among the earliest hellenistic buildings yet identified on the central Anatolian PLATESau and are prime evidence for the hellenisation of the interior of Asia Minor.
It is possible to speculate further about the circumstances under which they were built. The quality of the workmanship, at least for the larger temple, is unmistakeable. These were not modest provincial works, but the products of craftsmen and architects completely in tune with the skills, building styles and traditions of the Aegean cities. Recent work on the cultural development of the interior has stressed the likely role of the Attalid kings of Pergamum as promoters of a new cultural outlook across this hitherto scarcely hellenised region, and it is not impossible that they had a hand in developments at Antioch. Since the treaty of Apamea in 189 BC they had been nominal masters of most of cis-Tauric Asia west of the Halys, including Antioch. The political history of the second century BC shows that they made continuous and strenuous efforts to turn this nominal suzerainty into palpable domination.
Although Pisidian Antioch was expressly identified by Strabo as a free city, outside direct Attalid control, it would have been vital to Attalid interests to maintain a strong presence in the city. Several documents help to build up apicture of their attempts to assert the Pergamene presence in Pisidia, and architectural influence from Pergamum itself has been detected in several cities, notably Sagalassus and Termessus.49 The endowment of sanctuaries was an obvious and recognised way to extend political influence. We know that it formed an important aspect of Pergamene diplomatic policy from the relationship of the Attalids with the temple state of Pessinus and its famous shrine of cybele.50 There Strabo tells us quite specifically that they built a famous temple for the goddess, with white marble columns, perhaps at the end of the third, but much more probably around the middle of the second century BC.51
If the Attalids did make a contribution towards the temple building at Antioch, this supplies an exact parallel to the relationship between Pergamum and Pessinus. The architectural arguments suggest a dating for the Antioch temple in the reigns of Eumenes II or Attalus II. It is very tempting to make a connection between the facts or the possibilities which can be established about the archaeological remains of the Mên sanctuary and the report that a royal letter of Eumenes II, discovered at the village of Arizli, on the territory of Phrygian Synnada, should be restored to include a mention of a temple of Mên. But for confirmation or refutation we shall have to wait for the full publication of this text.52
Catalogue of moulded fragments from the temple and the temenos (Fig. 13)
Very few of the original mouldings are preserved from the temple. In situ there are three blocks from the lowest course of the pronaos and the cella walls, which had an Attic profile, subtly distinguished on the exterior and interior faces of the pronaos. Several other blocks with the same mouldings had been incorporated into the structure of the Byzantine church. Other fragments which we noted included one of the original grey limestone steps of the krepis, the grey lime-stone block that formed the threshold to the cella, grey and white limestone moulded pieces that may have been part of the frame of the cella door, an Ionic column base with apophyge in grey limestone, two broken pieces of white limestone Ionic column capitals, and several grey limestone blocks from the cella wall. Nothing at all was noted of any part of the upper entablature.53 It is quite impossible, therefore, to reconstruct the original appearance of the temple significantly above the level of the ground plan.
In addition we recorded a selection of the more numerous carved fragments which came from the sanctuary enclosure, some belonging to the colonnades of the surrounding porticos, but the majority from the free-standing monuments which were abundant in the temenos.
Fig. 13. Moulded fragments from the temple and the temenos.
Mouldings from the temple
(a) Grey limestone socle blocks from the cella and the pronaos. Three are still in position, while at least eight more have been reused in the church, including a block either from the north-east or the north-west corner of the cella. The profile is Attic, comprising torus, scotia, fillet, torus. Height (block in wall dividing the cella from the pronaos) 0.575; width at top 0.84; height of moulded section 0.526. The design of the interior and exterior mouldings is the same but the dimensions differ slightly. On the upper sides of the blocks still in place there is a series of clamp holes, 0.027–32 square, 0.02–6 deep, connected by channels, length 0.042–72. Similar mouldings are a standard feature of third-century Ionic temples in Asia Minor.
(b) Grey limestone threshold block from the cella door, height 0.57. The curved lower section and the lower half of the vertical face on the left coincided with the curved lower section and vertical face of the socle blocks. Then there were two steps 12.4 and 12.6 cm high leading to the level of the threshold itself. Three holes for doorposts: 0.19 x 0.17 x 0.113; 0.24 x 0.186 x 0.08; 0.195 x 0.207 x 0.06; between them grooves 0.065 wide and 0.023 deep.
(c) Grey limestone door-frame fragment. The top of the moulding and the rear are broken.
(d) White limestone moulding, broken at the left and at the corners. Now in Yalvaç Museum. Probably not from the door-frame.
(e) Grey limestone column base fragment, found on the steps near the south-east corner of the temple. Height 0.232; width 0.42; estimated diameter 0.72, broken below (PLATE 42). Compare the profile with the similar but larger profile of the socle moulding.
(f) central section of volute of white limestone Ionic column capital. Length 0.165. Now in Yalvaç Museum.
(g) Profile of one of the grey limestone blocks of the cella walls (PLATE 43). One block found in the church has a similar profile on opposite faces and is 67 cm thick, probably indicating the thickness of the wall between the cella and the opisthodomos, since the external cella walls were thicker. The dimensions of the blocks with the socle moulding suggest that this external wall was 84 cm thick. The cella wall blocks measure between 32.5 and 32.8 cm high, one Doric foot. The example illustrated is 73 cm long (broken). Drafted edges 0.028 wide, recessed 0.008 from the face of the stone. Fine-tooled face, done with the point of the chisel, margins smoothed and sometimes inscribed with texts relating to the games held at the sanctuary (see p. 50). For very similar drafted margins com-pare the blocks of the cella walls of the temple of Zeus Sosipolis (c. Humann, Magnesia am Mäander, 145 Abb. 153, showing blocks 23.5 cm high with margins of 0.025, recessed 0.003–0.005).
(h) Grey limestone block from topmost course of cella wall, with apophyge below the level of the architrave (PLATE 44). There is a more elaborate moulding at the same point on the cella walls of the temple of Zeus Sosipolis, Magnesia am Mäander, Abb. 158.
42. Moulded column base.
43. Block from cella wall of the Mên temple.
44. Top course of the cella wall.
45. Ionic column fragment from the temenos portico.
Mouldings from the porticos
(i) White limestone column base without plinth. Height 0.16, estimated diameter 0.6.
(j) Bottom section with apophyge of white limestone column. Height 0.087, estimated diameter 0.56.
(k) Bottom section and apophyge of fluted white limestone Ionic column (PLATE 45). Height 0.112; estimated diameter 0.5.
(l) White limestone fragment from the north colonnade. Height 0.162, width 0.738, depth 0.593 (broken). Perhaps the apophyge of a wall or the stylobate of the portico. The apophyge, which has been finished with a claw chisel, is interrupted by a short curved section, perhaps to receive a column.
The other sanctuary buildings
Outside the walls of the temenos there is a considerable number of other buildings. Two, the smaller temple which has already been described, and the small stadium are quite distinctive. The use and function of the others is not obvious and needs some discussion. They fall into two groups. Firstly, there is aseries of five multiroomed building complexes, usually incorporating a small colonnaded portico into the design. Two of these are close to the south-west end of the stadium (PLATE 46); one, known to Ramsay as the ‘Hall of Initiation’, is on the south-west side of the temenos; two more are at the south end of the site east of the temenos. We would argue that all these structures should be seen, in essence, as houses, parallels for which can be found in hellenistic and pre-hellenistic domestic architecture elsewhere in Asia Minor and in other parts of the Greek world. They need not, however, have been used simply as dwellings, but may have served the needs of organisations and associations connected with the cult of Mên. The second group comprises a larger number of much simpler buildings, consisting of a simple rectangular chamber, with antae projecting at the front on either side of a main entrance. We noted thirteen of these around the south and east sides of the site, and these have been marked on the general plan. These were also probably designed to house small groups of worshippers, who spent extended periods at the sanctuary during festivals and will have used them especially for celebratory cult meals and banquets, an essential part of the activities of any Greek shrine.
46. General view of the stadium and houses 1 and 2 from the temenos.
Fig. 14. Houses 1 and 2.
Houses 1 and 2
Houses 1 and 2 lie close to one another immediately by the south and west end of the stadium (Fig. 14). The east wall of house 1 is barely 50 cm and the robbed east wall of house 2 less than a metre from the curved end wall of the stadium, which appears to have been built later and to have encroached on them. In neither case is it possible to reconstruct a full and coherent plan, partly because much of the stone, even of the foundations, has been removed, particularly from house 2, but also because both show clear signs of several building phases, which cannot now be properly distinguished from one another.
House 1 has an irregular trapezoidal shape, comprising a rectangular section at the east end, which adjoins a crudely triangular area to the west (PLATE 47). Its most striking feature is the external colonnade, formed by a row of massive rectangular piers on the east and south sides. These have square or rectangular shafts, placed on independent trapezoidal plinths, which measure up to 1.3 m along the sides. The blocks are carefully cut with rectangular corners, and are finished with the point of a pick, giving an attractive regular stippled effect. Three adjoining foundation slabs on the north side of the building appear to form the foundation course for a similar portico, and there are three more grey limestone blocks at the west end of the building, making the apex of the triangle, which appear to belong to the same period of construction and perhaps also supported rectangular piers. This external portico enclosed a structure with one or more central chambers. There are several large rectangular blocks in the currently visible walls of this enclosed area, but most of the interior lay-out is made from smaller stones and from obviously reused material. It is not safe to assume that any of the building, apart from the parts of the portico already described, dates to the original construction phase.
The interior is dominated by a square chamber at the east end, measuring 5.8 x 6 m. The corners are made from large blocks, the remainder of the walls from much smaller stones. For the most part these were laid without mortar, but there are traces of mortar on the north side, which may belong to a later phase of the building. Here it appears that the north wall was deliberately thickened and realigned at a later date in order to convert the room from a trapezoidal to asquare plan.
Three threshold blocks, all in reused contexts, show the position of the entrances in the building in its late phase. The door from the outside was in the north wall. A threshold stone 1.94 m long has a sill at the front edge, and there is a square hole to receive a door-post in the block which adjoins it on the east (PLATES 48). This entrance led to a narrow irregular courtyard. On the left, or east side another reused threshold block led to the square room already described. Athird doorway led from the exterior into a triangular room at the west end, which appears to have been inaccessible from the courtyard.
The second house to the south-west also shows evidence of at least two building phases. The earlier is probably roughly contemporary with the first phase of house 1, since it contains a row of massive stone piers similar to those of the portico of the first house. These are on the side of the building and appear to have led to a small rectangular courtyard. The wall currently visible at the rear of this courtyard is made of polygonal masonry, some four or five courses and 1.85 m high, but is a later addition to the plan. The remaining walls are made from roughly cut irregular limestone blocks, laid without mortar, which were probably built up against the original piered construction on the north-west side. There may have been a large rectangular vestibule on the north-west with an exterior doorway at the north corner, opposite a second doorway 1.65 m wide, which led to a rectangular room measuring 10.35 x 7.6 m. There was another room south-west of this, but its dimensions and lay-out are unclear since the west side is buried underneath the rubble and earth from Ramsay’s 1912 excavation. There does, however, seem to have been a ‘terrace’ in the centre of the area, conceivably the foundation of a stairway which could have led up the steep bank towards the temple enclosure. Details of the north-east side of the building have also been obscured by the dump of the old excavation. The edge of the stadium prevented the house from extending further in this direction.
47. Stone piers of house 1.
48. Entrance to house 1.
House 3
The well preserved south-west wall of the temenos, with its many dedications to Mên, overlooks another large building (Fig. 15). There was direct access to this from the temenos by the steps near the south corner of the south-west wall. Five steps led down from the doorway to a narrow terrace, and beyond this a further eight steps descended to a rectangular courtyard, 10 m wide and surrounded by a colonnaded portico (PLATE 49). None of the columns survives, but their positions are clearly marked by engraved circles and pointing marks incised on the uppermost blocks of the portico foundations (PLATE 50). These show that the columns were between 37 and 40 cm in diameter with interaxial spacing of 2.2 m. The rear wall of this portico, about 3.6 m behind the columns, was built of small stones laid in an irregular fashion. The quality of the construction here is much less good than the stonework of the walls of the two chambers to the north-west and suggests that this small portico was a later addition to the original building. The use of slender, almost delicate columns, contrasting markedly with the massive monoliths of houses 1 and 2, confirms this impression of secondary construction. Ramsay, who excavated this building in 1912, reported that the floor of this area was paved with large slabs. These are now entirely covered by soil.54
Fig. 15. House 3.
49. Steps linking the temenos with the courtyard of house 3.
50. Column settings in the courtyard of house 3.
The earlier part of this building consisted of two rooms, a large one which is roughly square, measuring 15.7 x 15 m, with a rectangular antechamber, 14.5 x 4.5 m, on the south-east. Only the east side of this room is shown in Fig. 15. The main entrance, 2.35m wide, was in the south-east wall of the antechamber and led to another doorway, 2.3m wide, almost directly opposite it in the south-east wall of the main room. The walls themselves are substantial and range in thickness from 73 cm (south-east wall of the antechamber) to 1.2 m (south-west and north-west walls of the square room). They are constructed from large, even blocks of ashlar, between 45 and 70 cm high, which survive in places to the third or fourth course (PLATE 51). Mortar has been used only west of the entrance in the south-east wall of the large room, apparently to help level the courses rather than to give them solidity. The two rooms were built at the same time, to judge both by the style and quality of the building, and from the fact that the south-west wall is a continuous construction with no break. There is, however, an unbonded join on the north-east side between the wall of the antechamber and the east corner of the square room. The antechamber is divided by a cross wall, and a window links the small eastern section with the main square room.
51. Antechamber and south-east wall of the main room of house 3.
Ramsay’s excavation revealed a number of features in this building which are now covered up or only partially preserved. Fortunately, we are not entirely dependent on his written account and his published sketch plan, which seems suspect in a number of details, but on four good photographs, published as his PLATES 1. Three of these, 1, 2 and 4, show that along the north-east side of the large chamber was a low, rock-cut bench, 1’ 8" broad, abutting the wall. The interior of the room was paved with slabs and these surely overlaid the thin stratum of small, tightly packed stone chips, which Ramsay found covering the whole area. Close to the centre of the room were two rough-hewn stone sup-ports, about three feet high, which may have supported a table.55 Along the south-west side of the room and parallel to the outer wall there is a row of large stone blocks, still visible today, which Ramsay interpreted as an internal wall. These might possibly have been the bases for an internal colonnade, but the spaces between the individual blocks, ranging from 60 cm to 1 m, seem too narrow for this purpose. It is more likely that they too are simply the surviving blocks of the original pavement. Two other features of the room, seen both by Ramsay and by us, are a rectangular stone basin (0.78 x 0.5 m) beside the entrance, and a stone seat close to the north-east wall (height 0.67, width 0.55, depth at base 0.71, height of back 0.3, depth of seat 0.315), which carries the inscription Mηνί ευχήν, Mενέλαος Aτταου ὁ καί Κάρπος, ‘Menelaus son of Attaus, also called Carpus offered a vow to Mên’ (PLATE 52).
Ramsay took the building to be a ‘Hall of Initiation’, intimately associated with the rites which took place in the temple of Mên itself. The main room was taken to be a roofed, un-lit chamber or antron (cave); the two upright stones in the middle represented a symbolic doorway, through which the initiate passed to carry out ritual ablutions in a basin; the seat was the throne of the god himself.56 All this is quite unnecessary. The room was a paved hall with a bench along the north-east side; the two stone slabs in the centre may have been part of a table; the basin by the door was designed to collect water from the roof; the chair of Menelaus, although dedicated to Mên, was no more nor less than the regular seat of one of the devotees of the god, who used the room as a banqueting or meeting-house. If we disregard Seat of Menelaus, dedicated to Mên. the peristyle courtyard which was added to the two principal rooms at a later date, the ground plan is not dissimilar to that of the many simple rectangular oikoi on the site, with the addition of an antechamber. Discussion of its function belongs with that of those structures. Suffice to say that it may simply be interpreted as a large house, designed to accommodate the members of a cult association which might have held banquets or organised symposia associated with the god’s festivals. Ramsay reported that large quantities of animal bone, including pig, were found on the floor. That would be only natural.
52. Seat of Menelaus, dedicated to Mên.
Houses 4 and 5
House 4, which faces the south-east end of the temenos, is in a ruinous state (see Fig. 5). None of the walls stands more than two courses high, but its plan is clear, and unlike the other buildings it seems not to have undergone much rebuilding or modification. The core of the structure, perhaps indeed the only roofed part, was square, made up from a rectangular vestibule, about 16 x 4 m, which led to two adjoining rooms, the smaller to the south-west measuring 4.5 x 8 m, and alarger one, almost square, itself measuring 9.5 x 8.5 m. Further walls run off this complex on the north-west and north sides, probably enclosing an open courtyard. There are traces of an internal cross wall running parallel to the courtyard’s north-east boundary wall, about 5 m from it. This section might have been roofed to form a store-room or a byre for animals. The outline of this courtyard area should not be allowed to disguise the fact that the main element of the building had the form of a typical Greek house, with an antechamber leading to a suite of two rooms behind it.
Fig. 16. House 5.
House 5 which adjoins it on the east can be compared in building style to house 1 but is better preserved (Fig. 16). An external portico with massive rectangular piers can be traced around the north and west, and probably also continuing along the south side, although here it survives only at the west end. One of these piers has incised marks on its upper surface to indicate that it supported a large oval column, comparable to a roughly shaped oval column drum which was found in the central chamber of house 1. The walls inside this portico were excellently constructed out of rectangular or trapezoidal limestone blocks, 25–60 cm high, roughly finished but sometimes with a corner carefully cut to receive an adjacent stone. Small stones were also sometimes used for packing or for levelling the courses. On a smaller scale this construction technique is very similar to that used for the outer face of the walls of the main temenos, and the house was probably contemporary with it.
The entrance must have been through the south facing portico into an open courtyard. This led on the west to a large rectangular room, 6.6 x 9.8 m, and on the east to a smaller square chamber, 4.44 x 4.6 m. At the north-west corner of the portico there is a basin hollowed out from a single block of stone, presumably to collect water from the roof.
The Oikoi and the Andron
The remaining individual buildings on the site are much simpler in plan, consisting of a square or rectangular room entered by a doorway which stands between two projecting antae. The span between the antae was usually small and no columns were needed to support it. There is a row of five such oikoi running eastwards from the temenos wall as far as house 4, an isolated example south-east of house 5, at least two on the slope of the hill west of the small temple, and five on the ridge which overlooks the stadium from the north-east. Most of the walls have been reduced to a height of one or two courses and are made from blocks of undressed grey limestone. The chambers themselves are between three and four metres wide and four and five metres long. All of them are free-standing with asingle exception, the southernmost building of the series on the north-east edge overlooking the site. This is substantially larger than the rest and is joined to a smaller but similar building (Fig. 17). Its internal measurements are 8.25 x 7.8 m, and the south anta wall stands four courses and 2.5 m high, made from massive limestone blocks (PLATE 53). In this case the 9 m gap between the two anta walls was doubtless spanned with the aid of two columns (PLATE 54). There are remains of a threshold in the west wall, placed off centre. A striking and obvious parallel to this pair of buildings are the two oikoi which stand behind the temple of Zeus in the sanctuary of Labraunda in Caria, although these are finished to a much higher standard.57 The basic ground plans of the larger androns at Labraunda, built in the mid-fourth century BC, are also comparable to those of the oikoi at the Mên sanctuary.
Fig. 17. Andron and oikos.
53. Anta wall of the andron.
54. The andron.
The similarity between the largest of the Antioch buildings and the two adjoining Doric oikoi at Labraunda is the essential clue to the interpretation of all the secular buildings at the Mên sanctuary. Religious acts of ritual or devotion, narrowly defined, comprised only a small part of the activity of any Greek religious centre, whether it was a small shrine or a major sanctuary. Other aspects of community worship were more conspicuous and equally significant, and the most important of these involved banqueting. No festival, large or small, was complete without communal meals and the general good cheer which followed the consumption of the sacrificial meat. Indeed the point has been made that festival days at temples offered Greek communities their principal opportunities for social recreation.58 All this had to be accommodated at a temple site, and buildings to hold groups of diners or symposiasts are to be found either in the epigraphic or in the archaeological record of almost every temple or shrine. Labraunda is one of the clearest examples. Apart from the Doric oikoi there are also three large androns, occupying central positions to the south and south-west of the temple of Zeus. Excavation has shown that all three were designed as dining rooms, hestiatoria, with low benches around the walls to support the diners’ couches. Several other simple house-like structures distributed around the site no doubt fulfilled the same function for more modest groups of worshippers, and a large building 35 m square, with rooms built round a central courtyard, sited close to the propylaea and hitherto identified as a palace or an agora, has been convincingly reinterpreted as a banqueting complex.59
Different arrangements, designed with the same end in view, are well illustrated at the sanctuary of Demeter and Kore in Corinth between the sixth and fourth centuries BC. The topography of this sanctuary dictated that worshippers used about twenty-five oikoi tightly packed in three rows on terraces below the shrine. Each house measured about 5 x 4.35–4.75 m and could hold seven couches, dimensions which are virtually the same as those of the smallar oikoi at Antioch.60 At nearby Lerna the temple of Zeus looked across a piazza to nine oikoi erected at some expense by foreign cities, where their dignitaries could be accommodated, while beyond these there was a further block of less grand, but still substantial guest rooms, xenia. The most recent excavations have uncovered a row of houses with courtyards and several rooms behind these, which may be compared with the larger houses at Antioch.61 Larger dining rooms, designed to hold eleven couches, seem to have been about 6.3 m square or a little larger, roughly corresponding to the andron at Antioch. A pair with a vestibule in front was excavated at Perachora, and there are other parallels to be found in the Asclepieia at Corinth, Athens and Epidaurus, at the Argive Heraeum, and at the sanctuary of Aphaea on Aegina.62 Even more relevant are the buildings surveyed in the upper plain at Perachora, a sporadic scatter of structures of widely differing dates, grouped round a small temple, with notably elaborate provisions for the supply of water but no cemetery. This was not a regular settlement but catered for a transient and temporary population of worshippers, who probably camped in the sanctuary and used the oikoi for their festivities. Many of the buildings have the same simple plan found at Antioch, with a rectangular chamber and forecourt. In other instances two chambers stood side by side, approached from a vestibule in the front, much as the Antioch andron and its smaller adjoining chamber would have been. House 4 at Antioch also shows the typical arrangement of two equal-sized adjoining rooms behind a rectangular vestibule. The whole complex at Perachora despite its early date provides aparticularly appropriate parallel, for it is another instance of an extramural sanctuary some distance from the nearest permanent settlement, where the dining rooms could be sited here and there at random over a considerable area, and were not confined by the topography or by other more monumental buildings as they were at the sanctuary of Demeter and Kore at Corinth or at Labraunda. The oikoi were, in effect, the permanent equivalent of tents, skenai, which were to be found on festival days at other places of worship, such as the hill-top temenos of the Carian god Sinuri.63 At other sites which have been excavated the internal arrangement of these rooms shows precisely how they were fitted out for shared meals, with seven or eleven couches arranged around the walls.64 No doubt excavations at the Mên sanctuary would reveal similar fitments.
Epigraphically the best documented religious centres of Graeco-Roman Asia Minor are those of Zeus at Panamara and Hecate at Lagina, both attached to the Carian city of Stratonicaea. A high proportion of the several hundred inscriptions recorded at these sites alludes to feasts and banquets provided by local benefactors, usually priests of the cults who were also members of the civic aristocracy.65 The copious epigraphic documentation from the Mên sanctuary at Antioch is largely confined to single dedications or refers to the competitions held in the stadium, but the archaeological remains show beyond question that communal meals were as much a feature of the worship of Mên as they were of the cults of other Greek deities. The site and its buildings offer an invaluable counterpart to the documented picture of the two Stratonicaean sanctuaries.
Although the function and purpose of the various buildings around the temple is thus clear, their chronology remains extremely imprecise. One might have looked for some useful indications from Ramsay’s excavations, but his reports contain no evidence of any substance and few reliable conclusions. He gives no date for the main temple or the temenos, although his working assumption seems to have been that both were hellenistic. Three periods were distinguished in the so-called ‘Hall of Initiation’, our house 3, with the main chamber coming first, followed by the rectangular vestibule and finally by the little colonnaded enclosure which he dubbed a pronaos. As noted above, the main chamber and vestibule appear, in fact, to have been contemporary. The stadium, for no given reason, was assumed to be ‘quite late’. There was no clear indication of the dates of houses 1 and 2, although they were seen to be earlier than the stadium. Four periods of development were identified in the church: the fourth century, the ninth century, the Turkish period, and the present day, but no grounds for these conclusions were offered. The pottery is described as mainly Roman with a little hellenistic material. No coins appear to have predated the first century BC and none, except in the area of the church, were later than the fourth century AD.66
Without the advantage of excavation it is difficult to progress beyond this. Only the two temples can be dated with confidence to around the middle of the second century BC. It is of course inconceivable that they should have stood as isolated buildings on an empty mountain top and there must have been other hellenistic structures. The temenos is probably roughly contemporary with the temple of Mên, and its careful although irregular construction technique supports this. There is no trace of the even courses or rectangular blocks that one might expect in monumental public building of the Roman imperial period. Moreover the avoidance of rigid rectangularity in the lay-out not only of the temenos but also between the central buildings of the whole sanctuary is also characteristic of hellenistic rather than Roman architectural planning. The two temples are not precisely aligned with one another; the temple of Mên does not sit exactly square within its temenos, and the four temenos walls are uneven, although not drastically so. House 3 is not built parallel to the south-west wall of the temenos but at a slight angle to it. One may compare the lay-out of the hellenistic city centre at Sagalassus, whose Doric temple stands slightly to one side of its temenos and at a different angle both to the adjoining heroon and to the bouleuterion.67
The construction technique of some of the other buildings supports the notion that the main part of the sanctuary is essentially hellenistic. The better walls of houses 3 and 5 resemble those of the temenos, although they are made from smaller blocks, and the main rectangular piers of houses 1, 2 and 5 seem to be early in date, perhaps comparable to some of the vernacular stone architecture of Lycia and Caria. In contrast the neat little colonnade of house 3 and certainly some of the rebuilding in houses 1 and 2 indicate that modifications took place in the Roman period, as would be expected. There seems to be no way of establishing a chronology for the smaller oikoi, which were built unpretentiously from local stone to a traditional design, but nothing stands in the way of the suggestion that they were contemporary with the temenos.
However, on one point Ramsay was decisive and with good reason. He argued strongly that the sanctuary, particularly the temple and its immediate surroundings, had been deliberately destroyed in antiquity by Christians. ‘The intention of the Christians who wrecked it was to leave it a howling wilderness into which no man could enter. The entrance was deliberately blocked up, and almost all the votive stelae which were erected in great numbers in the open space of the sanctuary were broken in pieces with the intention of deliberate wreckage. In one case, at least, parts of the same stele were found remote from one another, inside and outside the sanctuary.’ ‘Not a scrap of the temple above the stylobate was found, except two or three blocks of the lowest course. The stones (marble?) were carried away.’ 68 Anderson had made the same point: ‘In the hour of their final triumph the Christians entered in to destroy the holy place, and they left it a desolate ruin. Only the more massive monuments survived their revel of destruction and many even of these were smashed to fragments, while the lighter stones, slabs, tablets, statuettes, and the like were dashed to pieces and flung broadcast over the sacred enclosure.’69 As will be clear from the early part of this chapter Ramsay’s and Anderson’s archaeological observations should be treated with extreme reserve, but here they surely hit the mark. Delapidation and earthquake damage would account for the collapse but not for the dispersal or disappearance of the temple superstructure. A large and thriving community on the same or at a nearby site might take away the cut stone for its own building needs. But substantial public buildings on a remote mountain top, inaccessible to wheeled traffic in antiquity and up until the very recent past,70 ought to have been immune from wholesale stone robbing. Blocks from the cella walls, including much of the socle moulding, were reused in the construction of the church, and, if that is correctly dated to the later fourth or fifth centuries, it provides asolid terminus ante quem for the destruction of the temple. As for the remaining elements, the entablature, the columns and their capitals, and the decorative order from the surrounding temenos – all the features which would have helped earlier visitors to identify these hellenistic buildings for what they were – we must follow Ramsay’s view that they were smashed and perhaps buried by the Christians, in one of the more startling testimonies to the conflict between paganism and Christianity in late Roman Asia Minor.
Notes
1 Compare the description by L. Robert, A travers l’Asie Mineure (1980) 312–13 n. 16: ‘Sur la belle colline escarpée ou se dresse le sanctuaire de Mên Askaênos, près d’Antioche de Pisidie...quand nous y montâmes, elle servait de guette à un bekci [a look-out] muni d’un gourdin. Il était venu surveiller les vignobles de la plaine où les habitants d’un village étaient accusés d’avoir profité des vignes d’un autre. Le plaine s’étendait comme une carte, avec les villages, et on voyait d’en haut tout ce qui s’y passait.’
2 M. Hardie, JHS 32 (1912) 115–18; Ramsay, The Athenaeum 4372 (12 August 1911) 192–3; 4427 (31 August 1912) 226; 4428 (7 September 1912); 4454 (8 March 1913) 190; JRS 8 (1918) 115–17. Compare E.N. Lane, Berytus 15 (1964) 40 ff.; I. Karamut, TAD 28 (1989) 177–87.
3 The Athenaeum 4427 (31 August 1912) 226; in JRS 8 (1918) 115 these figures are averaged to produce a ‘roughly oblong space 227 by 135 feet in side measurement’.
4 The Athenaeum 4427 (31 August 1912) 226.
5 JRS 8 (1918) 116.
6 Probably originally six, as Ramsay remarked. The two at the east end have been covered up.
7 See M. Waelkens, AS 39 (1989) pl. XIV b and Sagalassos I (Leuven, 1993) 63 Fig. 36.
8 The top measurements of the buttresses on the south-west side, from north to south, are as follows (widths first, then projections): 66 x 77 cm; 66 x 68 cm; 56 x 57 cm; 75 x 43+ cm; 73 x 65 cm; 64 x 67 cm; 71 x 70 cm; 82 x 108 cm; 68 x 60+ cm; 55 x 69 cm. From the top the courses of the sixth buttress in this series measure 64 x 67, 71 x 79 and 84 x 96 cm.
9 The inscriptions were published by M.M. Hardie, JHS 32 (1912) 121 ff. nos. 1–67; E.N. Lane, CMRDM I (1971) nos. 179–244, with additional notes in CMRDM IV (1978) nos. 1–18 and p. 53 (a text copied by H. Waldmann in 1974).
10 Ramsay, JRS 8 (1918) 116 gives the measurement as 4’ 10" (1.48 m).
11 Measurements from west to east: 1.52 x 0.875; 1.24 x 0.87; 1.24 x 0.86; 1.48 x 0.875 m.
12 Their lengths vary on either side of 1 m: 108, 92, 88, 99, 137 and 105 cm.
13 The Athenaeum 4427 (31 August 1912) 226; JRS 8 (1918) 113.
14 The Athenaeum 4427 (31 August 1912) 226; JrS 8 (1918) 117. Anderson, JRS 3 (1913) 234–6 states that this ‘shrine’ was the find spot of two statuettes of Demeter and Mên, which he illustrates in figures 52–4. The statue of Cornelia Antonia, dating to the Antonine period (see above p. 13), is now in the Istanbul Museum (inv. no. 2645T; cf. G. Mendel, Cat. Mus. Imp. Ottomanes, 1377 no. 237) and has been published most recently by J. Inan and E. Rosenbaum, Roman and Early Byzantine Portrait Sculpture in Asia Minor (1966) 208–9 no. 287 with pl. CLXII.1 and CLXIII.
15 JRS 8 (1918) 118–19.
16 Some typical dimensions: 97 x 87 x 25 cm; 113 x 154 x 19+ cm; 145 x 105 x 38 cm; 173 x 105 x 37 cm.
17 Inferred from the circumference of 5’ 2" given by Anderson.
18 One block, found in the church, height 0.323, width 0.52, thickness 0.67 m, broken on the right. The inscription, letter heights 2 mm, is carved on the upper margin and reads πANKPATIONH[π]ATPISO... Another, found in the temenos on the north-east side of the temple, height 0.327, width 0.23 (broken right), thickness 0.46 (buried), height of central panel above rim 7 mm, width of margin 28 mm; the inscription is carved in clear, even letters 1.6–1.7 mm high, carved along the top margin, reading KAHPOY.. .(PLATES 43)
19 One exception is M. Hardie, JHS 2 (1912) 147 no. 68 Fig. 12 (a votive marble tablet), and nos. 69–70, two fragments, the latter with a tabula ansata.
20 E.N. Lane, Berytus 15 (1964) 36 ff.; B.M. Levick, AS 20 (1970) 37–50. Many of these monuments were studied again by M. Waelkens in the Konya Museum in 1984, and his copies suggest a number of minor alterations to the published texts.
21 CMRDM I, 165 ff., nos. 162 (in Konya), 175–6 (in Istanbul), 253–4 (in Yalvaç Museum), and 261–94. Lane’s claim that these came from Robinson’s 1924 excavation is clearly wrong. They are from Ramsay’s 1912/13 excavation of the sanctuary. See also Lane’s additional note on pp. 165–8.
22 JHS 32 (1912) 143.
23 JRS 8 (1918) 125 n. 2.
24 The Athenaeum 4372 (12 August 1911) 192; M. Hardie, JHS 32 (1912) 118–21 envisaged a high stepped altar facing towards the south, and vainly cited parallels to support this view. It is frustrating that she reported the discovery of a ‘pilaster which is 22 inches high and 45 inches broad’ without giving details of its architectural decoration.
25 The Athenaeum 4428 (7 September 1912).
26 JRS 8 (1918) 116.
27 TAD 28 (1989) 171–87.
28 Quarries were noted about two kilometres west of the sanctuary in the direction of Antioch, about one kilometre west, close to the visible part of the ‘sacred way’, and on the east side of the site, between houses 4 and 5. Moreover there are clear traces of a drag way on the south-west slope of the mountain, several metres wide and strewn with small stones and rubble chips.
29 For similar contrasts of stone colour, note the temple of Hemithea at Castabos, where limestone was used for the krepis, the orthostats, most of the cella walls and the antae, while white marble was chosen for the columns of the peristyle, the whole entablature, and to form a continuous decoration along the outer edge of the toichobate and the outer face of the threshold (W.H. Plommer, in Plommer and J.M. Cook, The Sanctuary of Hemithea at Castabos (1966) 77). At Labraunda local limestone was used for the walls of the temple of Zeus, marble for the columns, entablatures, floors and stylobates (A. Westholm, Labraunda I.2 (1963) 25; Labraunda I.3 (1982) 31).
30 The measured heights of the foundations of the steps on the south side are as follows: 1, 32 cm; 2, 29 cm; 3, 28,5 cm; 4, 23 cm; 5, 26,5 cm; 6, 25 cm; 7, 23,5 cm; 8, 26 cm; 9, 27 cm; 10, 25 cm.
31 Dimensions: a) height 25 cm, depth 52 cm; b) height 33 cm, depth 43 cm.
32 Along the south-west side the measurements, from one south-west face to the next, of each foundation plinth are as follows: bases 2–3, 2.35; 3–4, 2.5; 4–5, 2.42; 5–6, 2.4; 6–7, 2.38; 7–9, 4.8; 9–10, 2.5; 10–11, 2.4 m. On the north-west side the measurements between the south-west faces are: 2–3, 2.35; 3–4 2.5 m. The measurement between the bases in the opisthodomos is 2.35 m. On the north-east side the measurements from the south-east faces are: 2–3, 2.5; 3–4, 2.6, 4–5, 2.4 m.
33 Compare also the imperial temple of Antoninus Pius at Sagalassus and the so-called Traianeum at Pergamum.
34 Vitruvius III.3.1 defines interaxial spacing as follows: pycnostyle = 22 x lower diameter; systyle = 3 x lower diameter; eustyle = 34 x lower diameter combined with diastyle between central columns of façade; diastyle = 4 x lower diameter; araeostyle = anything larger.
35 T. Wiegand and H. Schräder, Priene. Ergebnisse der Ausgrabungen und Untersuchungen in den Jahren 1895–1898 (Berlin, 1904) 81–108; M. Schede, Die Ruinen von Priene (Berlin, 1964) 25–48; W. Koenigs, Ist. Mitt. 33 (1983) 162 ff.; J.C. Carter, The Sculpture of the Sanctuary of Athena Polias at Priene (1983) 24–43.
36 H. Metzger, Revue archéologique 1974, 317 ff.; Fouilles de Xanthos VI (1979) 17–28.
37 A. Machatschek and M. Schwarz, Bauforschungen in Selge, Öst. Ak. der Wissenschaften, phil.-hist. Klasse Denkschr. 152 (Vienna, 1981) 89–91; cf. S. Mitchell, ‘Hellenismus in Pisidien’, Asia Minor Studien 6 (1992) 7–8 and Mediterranean Archaeology (Sydney) 4 (1991) 126 with pl. 6 2–3 illustrating the decorated mouldings.
38 Y. Bequignon and A. Laumonier, BCH 48 (1924) 506 ff.; 49 (1925) 281–321 with pl. VIII; W. Hahland, ‘Der Fries des Dionysostempels in Teos’, JÖAI 38 (1950) 66–109; M. Duran Uz, III Araştirma Sonuçlari Toplantisi, Ankara 1985 (1986) 227–42.
39 Vitruvius III.3.6.
40 A. Davesne, ‘A propos de la chronologie du temple de Dionysos à Teos’, IVAraştirma Sonuçlari Toplantisi, Ankara 1986 (1987) 143–7.
41 See Lanckoronski, Städte Pamphyliens und Pisidiens I, Fig. 30, Taf. II (reconstructed façade), Taf. III (plan). Frieze blocks found in the vicinity of the building illustrate scenes from the story of Iphigeneia (dated by K. Stählin, Arch. Anz. 1968, 280–89 to the later second century BC).
42 R. Pullan, Antiquities of Ionia IV (London, 1881) 40 ff.; H. Weber, Ist. Mitt. 16 (1966) 100 ff.; C. Özgünel, IV. Kazi, Ankara 1982 (1983) 207–24; VI Kazi, Ankara 1984 (1985) 391–401; VIII Kazi, Ankara 1986 (1987) 47–57.
43 C. Humann, Magnesia am Mäander. Bericht über die Ergebnisse der Ausgrabungen der Jahre 1891–1893 (Berlin, 1904).
44 W. Dörpfeld, ‘Metrologische Beiträge’, Ath. Mitt. 1890, 167 ff.
45 Edhem Bey, CRAI 1905, 443 ff.; 1906, 407 ff.; G. Mendel, Musées Impériaux Ottomanes. Catalogue des sculptures grecques, romaines et byzantines I (Constantinople, 1912) 428 ff.; H. Schober, Der Fries des Hekataions von Lagina, Ist. Forsch. II (Vienna, 1933). T.N. Howe, AJA 90 (1986) 62 n. 68, has questioned whether the Alabanda temple should be restored with four prostyle columns as well as a distyle in antis pronaos. However, his note is confused. Howe wants to restore the temple without a prostyle but retaining the distyle in antis (see his illustration 7 d), and suggests that Edhem Bey’s reconstruction of the four columns in front is hypothetical. However, as Schober, Der Fries des Hekataions, 17 indicates, the bases for these columns were actually found, and if anything the two columns in antis, which are not actually aligned with those of the peristyle, seem out of place. The point needs to be checked on the spot.
46 D. Krencker and M. Schede, Der Tempel in Ankara (Berlin, 1936), with literature on the dating problem cited by S. Mitchell, Chiron 16 (1986) 17–31; R. Naumann, Der Tempel zu Aizanoi (Berlin, 1979).
47 E. Akurgal, Ancient Ruins and Cities of Turkey (1985) 246, publishes the plan made by Ü. Serdaroglu. The order of the temple, however, was Corinthian.
48 W. Radt, Pergamon. Geschichte und Bauten, Ausgrabungen und Erforschung einer kleinasiatischen Metropole (1988) 206–14 (Demeter), 218–22 (Dionysus), 374–5 (bibliography).
49 For Attalid building outside Pergamum see W. Radt, Pergamon. Geschichte und Bauten. Funde und Erforschung einer antiken Metropole (1988) 309–12. In general on Attalid cultural influence, H.-J. Schalles, Untersuchungen zur Kulturpolitik der pergamenischen Herrscher im 3. Jhdt. v. Chr., Ist. Forsch. 36 (1985). Inscriptions attesting Attalid activity and interest in Pisidia, include H. Swoboda and J. Keil, Denkmäler aus Lykaonien, Isaurien und Pisidien (1935) no. 74 (OGIS 751; Welles, Royal Correspondence no. 54), from Amblada; MAMA VI 173 (Apamea) and texts from Olbasa in western Pisidia published by A.S. Hall, IV Araş. Ankara 1986 (1987) 150 and R. Kearsley, AS 45 (1995) 43–57. For a possible Attalid garrison fortress controlling the roads from southern Pisidia to Pamphylia, see S. Mitchell et al., ‘Panemoteichos and Ören Tepe’, AS 47 (1997) 141–72. For Attalid influence on the architecture of Sagalassus, see M. Waelkens, Sagalassos I (1993) 42–5 (although the dating suggested there is too high). See too S. Mitchell, Mediterranean Archaeology 4 (1991) 131–2 (Sagalassus) and 143.
50 Strabo XII.5.3, 567; see B. Virgilio, Il ‘Tempio-Stato’ di Pessinunte fra Pergamo e Roma nel IIIsecolo a. c. (Pisa, 1981); J. Devreker, in Devreker and M. Waelkens, Les Fouilles de la Rijksuniversiteit te Gent a Pessinonte 1967–73, I (Bruges, 1984) 14–15.
51 J. Devreker, in Devreker and M. Waelkens (ed.), Les Fouilles de la Rijksuniversiteit te Gent a Pessinonte 1967–1973 (1984) 15–16 conjectures that work was begun around 205 BC in compensation for the removal of the sacred stone of Cybele to Rome in the previous year; Virgilio, Il ‘Tempio-Stato’ di Pessinunte, 70–1, suggests the period from 183–163 BC.
52 T. Drew-Bear, GRBS 17 (1976) 247 n. 2. Prof. Drew-Bear informs me that the text is very fragmentary and that it may be difficult to reach definite conclusions.
53 However, in July 1997 a section of Ionic cornice including dentils was visible at the east end of the temple.
54 ABSA 18 (1912) 37–79.
55 ABSA 18 (1912) 39–44.
56 See The Athenaeum, 25 January 1913, 106; 4454, 8 March 1913, 290. Ramsay supported this interpretation by invoking a supposed conduit which brought water from the (as we believe) non-existent cistern inside the temenos down to the ‘Hall of Initiation’, JRS 8 (1918) 114–15.
57 P. Hellström, ‘Labraunda 1983’, IIAraş., Izmir 1984 (1985) 131–8 with figs. 2 and 4.
58 R. MacMullen, Paganism in the Roman Empire (1982) 37.
59 P. Hellström, ‘Labraunda 1985’, IVAraş., Ankara 1986 (1987) 157–66 Fig. 1 repro-duces the site plan, also to be found in other recent Labraunda publications, which well illustrates these points.
60 N. Bookides and J. Fisher, Hesperia 41 (1972) 285–307; 43 (1974) 267–307.
61 S.G. Miller, Hesperia 46 (1977) 20–6; 57 (1988) 1–20 with the site plan on p. 2 Fig. 1.
62 Cited by R.A. Tomlinson, ABSA 64 (1969) 155–8.
63 L. Robert, Le Sanctuaire de Sinuri I (1945) 49 ff.
64 See also A. Furtwängler, Aegina. Das Heiligtum der Aphaia, 107 with pl. 70; F. Robert, Délos XX, 51 ff. esp. 64; M. Launey, BCH 61 (1937) 380–409.
65 For good accounts see H. Oppermann, Zeus Panamaros (1924); A. Laumonier, Les cultes indigènes en Carie (1958). MacMullen, Paganism in the Roman Empire, 37 ff., provides an evocative picture and the inscriptions are all collected in M. Çetin Sahin, Die Inschriften von Stratonikaia I and II, 1–2.
66 Ramsay, JRS 8 (1918) 110–21.
67 M. Waelkens, AS 38 (1988) 61 Fig. 2.
68 JRS 8 (1918) 121.
69 JRS 3 (1913) 209; a vivid and moving description, according to Syme, Anatolica. Studies in Strabo (1995) 344 n. 7.
70 The motorable road up to the site was constructed after the completion of our survey in 1983.